Ash Syllabus Cast 137

All Rights Reserved. No part of this publication may be reproduced in any form without express written permission from CMEducation Resources. In some instances, ownership in specific articles (handouts, papers, etc.) is held by the specific author.

Dear Healthcare Colleague: Welcome and thank you for participating in the New Frontiers and Evolving Paradigms in Cancer and Thrombosis symposium jointly sponsored by CMEducation Resources, LLC and the University of Massachusetts Medical School (UMMS). The program is supported by an independent educational grant from Eisai, Inc. We are excited that you have joined us and we are confident that you will benefit from your participation. CMEducation Resources has put together an excellent program that will measurably enhance the quality of care you provide for patients with cancer and Thrombosis. The UMMS designates this continuing medical education activity for a maximum of 3.5 credit hours in Category 1 toward the Physician’s Recognition Award of the American Medical Association. With the cooperation of our distinguished faculty, this comprehensive course syllabus has been assembled to be an educational resource that you may consult during the course of this program, and as a guide to challenging clinical decisions that you must make in your day-to-day practice. In addition, we invite you to continue with our CME programming – including webcasts, HealthWRAPS®, SlideCASTs and ConsultCASTs – by linking on to www.CLINICALWEBCASTS.COM. To comply with continuing education requirements, the symposium evaluation form is enclosed in this notebook. At the conclusion of the symposium, please complete this form and return it to a CMEducation Resources representative at the registration desk. Please be sure to indicate additional program topics with specific focus areas that you wish to see covered in future CME programming. This information is important to us as it provides us with data to use in developing high-quality programs for the future as well as verifying your shipping address for your CME Certificate. Once again, we thank you for attending our symposium. If you have any questions during the symposium, please do not hesitate to contact one of the CMEducation Resources representatives at the registration desk. Sincerely, CMEducation Resources, LLC

CMEducation Resources, LLC New Frontiers and Evolving Paradigms

in Cancer and Thrombosis

Agenda Friday, December 3, 2010 Orange County Convention Centre – Orlando, FL

TIME LOCATION SESSION

6:00 – 6:30 p.m. Conference Room: 109B Program Registration, Distribution of Program Materials to ASH Satellite Symposium Attendees

6:30 – 6:50 p.m. Conference Room: 109B New Frontiers and Evolving Paradigms in Cancer and Thrombosis: The Journey From Science to Strategy, From Bench to Bed Side Samuel Z. Goldhaber, MD – Program Chairman

6:50 – 7:15 p.m. Conference Room: 109B The Complex Interface of Malignancy, Thrombosis, and Clinical End Points Frederick R. Rickles, MD

7:15 – 7:45 p.m. Conference Room: 109B Optimizing Risk Assessment and Management of Cancer Patients at Risk for Venous Thromboembolism (VTE): Reducing DVT Recurrence and Related Complications Craig Kessler, MD

7:45 – 8:15 p.m. Conference Room: 109B Risk Stratification Tools to Identify Patients For Primary and Secondary Prevention of VTE in Patients with Malignancy Alok A. Khorana, MD, FACP

8:15 – 8:30 p.m. Conference Room: 109B Program Chairman’s Concluding Vision Statement: Current and Near Future Perspectives of VTE Management in the Setting of Malignancy Samuel Z. Goldhaber, MD – Program Chairman

8:30 – 8:45 p.m. Conference Room: 109B Interactive Q & A Discussion Session



Faculty Roster Friday, December 3, 2010 Orange County Convention Centre – Orlando, FL Samuel Z. Goldhaber, MD – Program Chairman Senior Staff Physician, Director, VTE Research Group Brigham and Women’s Hospital Professor of Medicine Harvard Medical School Boston, MA ______________________________________________________________________________ Frederick R. Rickles, MD Professor of Medicine The George Washington University Washington, DC Craig Kessler, MD Professor of Medicine and Pathology Georgetown University Medical Center Washington, DC ______________________________________________________________________________ Alok A. Khorana, MD, FACP Associate Professor Vice Chief James P. Wilmot Cancer Center University of Rochester Rochester, NY LEARNING AND PROGRAM OBJECTIVES After participating in this program, physicians should be able to:

• Apply current guidelines for pharmacologic prophylaxis of DVT issued by national professional organizations (ASH, ASCO, NCCN, ACCP, ASHP) in at risk patients with cancer, medical and surgical conditions.

• Risk stratify medical, oncology, and surgical patients, evaluate their likelihood for incurring DVT, and learn how to assess and implement prophylaxis measures that can reduce the incidence of DVT in these patient populations.



Faculty Roster Friday, December 3, 2010 Orange County Convention Centre – Orlando, FL

• Assess and manage special needs of both inpatients and discharged outpatients at risk for DVT, with a focus on long-term prophylaxis against recurrent DVT in patients with cancer, medical disorders, and in cancer patients.

• Apply landmark clinical trials focusing on DVT prevention in medical and surgical patients—in particular, those with cancer—to their clinical practice.

• Evaluate, select among, and appropriately use the range of pharmacologic options available for DVT prophylaxis, including warfarin, unfractionated heparin, and LMWHs.

• Apply current guidelines issued by national professional organizations such as NCCN and ACCP in at risk patients with medical and surgical conditions.

• Risk stratify medical and surgical oncology patients, assess their likelihood for incurring DVT, and be aware of prophylaxis measures that can reduce the incidence of DVT in patients with a variety of tumor types, and with chemotherapy.

• Assess and manage special needs of cancer and critical care patients at risk for DVT, with a focus on protecting against recurrent DVT in patients undergoing surgery

ACCREDITATION STATEMENT This activity has been planned and implemented in accordance with the Essential Areas and policies of the Accreditation Council for Continuing Medical Education through the joint sponsorship of The University of Massachusetts Medical School, Office of CME and CMEducation Resources, LLC. The University of Massachusetts Medical School is accredited by the ACCME to provide continuing medical education for physicians. CREDIT DESIGNATION STATEMENT The University of Massachusetts Medical School designates this educational activity for a maximum of 3.5 AMA PRA Category 1 Credit(s). Physicians should only claim credit commensurate with the extent of their participation in the activity. POLICY ON FACULTY AND PROVIDER DISCLOSURE It is the policy of the University of Massachusetts Medical School to ensure fair balance, independence, objectivity and scientific rigor in all activities. All faculty participating in CME activities sponsored by the University of Massachusetts Medical School are required to present evidence-based data, identify and reference off-label product use and disclose all relevant financial relationships with those supporting the activity or others whose products or services are discussed. Faculty disclosure will be provided in the activity materials. Supported by an independent educational grant from Eisai, Inc.

The University of Massachusetts Medical School Office of Continuing Medical Education (UMMS‐OCME) has reviewed the appropriate documentation provided by the individuals who are in a position to control the content of this educational activity. The UMMS‐OCME has determined that any potential relevant conflict of interest has been resolved. For more information about faculty and planner disclosures, contact the UMMS‐OCME at [email protected]

University of Massachusetts Medical School Office of Continuing Medical Education Summary of Faculty Disclosure Information

Faculty Disclosures: As a sponsor accredited by the ACCME, The University of Massachusetts Medical School Office of Continuing Medical Education (UMMS‐OCME) must ensure balance, independence, objectivity, and scientific rigor in all its individually sponsored and jointly sponsored educational activities. All faculty participating in a sponsored activity are expected to disclose to the activity audience any discussion of off‐label use or investigations use of a product, and any relevant financial interest or other relationship which they, or their spouse/partner, have (a) with the manufacture(s) of any commercial product(s) and/or provider(s) of commercial services discussed in an educational presentation and (b) with any commercial supporters of the activity. (Relevant financial interest or other relationship can include such things as grants or research support, employee, consultant, major stockholder, member of speaker’s bureau, etc.)

The following faculty members have indicated their financial interests and/or relationships with commercial manufacture(s) (and/or those of their spouse/partner) below. Faculty with no relevant financial relationships are listed with N/A.

FINANCIAL INTERESTS OR RELATIONSHIPS

Faculty Member Relationship Corporation/Manufacturer Samuel Z. Goldhaber, MD

Consultant: Speaker’s Bureau: Grant/Research:

Sanofi‐Aventis, BMS, B‐I, EKOS, Medscape, Eisai N/A Sanofi‐Aventis, BMS, B‐I, Eisai, J&J

Frederick R. Rickles, MD

Consultant: Speaker’s Bureau: Grant/Research:

Genmab, Bayer/Ortho‐McNeil/J & J, Pharmacyclics, Leo Eisai

Craig Kessler, MD

Consultant:Speaker’s Bureau: Grant/Research:

Sanofi‐Aventis, Eisai Sanofi‐Aventis, Eisai

Alok A. Khorana, MD, FACP

Consultant:Speaker’s Bureau: Grant/Research:

Sanofi‐Aventis, Eisai, Leo PharmaSanofi‐Aventis, Leo Pharma Sanofi‐Aventis

The speaker must disclose any discussion of off‐label use and/or investigational products to the audience during the presentation.

Committee/Staff Disclosure

The following CME program planners have indicated their financial interests and/or relationships with commercial manufacturer(s) (and/or those of their spouse/partner below. Planners with no relevant financial relationships are listed with N/A.

COMMITTEE / STAFF RELATIONSHIPGideon Bosker N/AMilo Falcon N/A



New Frontiers and Evolving Paradigms in Cancer and Thrombosis: The Journey From Science to Strategy, From Bench to Bed Side

Faculty: Samuel Z. Goldhaber, MD – Program Chairman Time: 6:30 – 6:50 p.m. Notes:

1

New Frontiers and Evolving Paradigms in Cancer and Thrombosis

A Year 2010 Milestone Summit

Optimizing Prevention, Risk Assessment, and Management of Thrombotic Complications in Malignancy: What Do the Trials Teach Us?

How Should the Science Guide Us?

Program ChairmanSamuel Z. Goldhaber, MD

Cardiovascular DivisionBrigham and Women’s Hospital

Professor of MedicineHarvard Medical School

2

CME-certified symposium jointly d b th U i it f

Welcome and Program Overview Welcome and Program Overview

sponsored by the University of Massachusetts Medical School and CMEducation Resources, LLC

Commercial Support: Sponsored by an unrestricted educational grant from Eisai, Inc.

Faculty disclosures: Listed in program syllabus

We Request That You…

►PLEASE FILL OUT

QUESTION AND ANSWER (Q&A) CARDS as program proceedsso we can collect them and discuss during the Q&A session

COURSE SURVEY AND EVALUATION forms to obtain CME credit Pleaseforms to obtain CME credit. Please hand all survey forms to the staff at the desk outside following the program

3

Program Faculty

Program ChairmanSamuel Z. Goldhaber, MDCardiovascular DivisionBrigham and Women’s Hospital

Alok A. Khorana, MD, FACPVice-Chief, Division of Hematology/Oncology Associate Professor of Medicine and Oncology James P Wilmot Cancer CenterBrigham and Women s Hospital


Craig Kessler, MDProfessor of MedicineDepartment of HematologyAnticoagulation ServicesGeorgetown University Medical Center

James P. Wilmot Cancer CenterUniversity of RochesterRochester, NY

Frederick R. Rickles, MDClinical Professor of Medicine, Pediatrics, Pharmacology and Physiology

Division of Hematology-OncologyDepartment of Medicine

Washington, DCDepartment of MedicineThe George Washington University School of Medicine and Health Sciences

Washington, DC

Program Agenda

6:30 PM — 6:50 PMNew Frontiers and Evolving Paradigms in Cancer and Thrombosis: The Journey From Science to Strategy, From Bench to Bed Side

Program ChairmanSamuel Z. Goldhaber, MDCardiovascular Division │ Brigham and Women’s Hospital │ Professor of Medicine │ Harvard Medical School

6:50 PM — 7:15 PMThe Complex Interface of Malignancy, Thrombosis, and Clinical End Pointsand Clinical End Points

Frederick R. Rickles, MDClinical Professor of Medicine, Pediatrics, Pharmacology and Physiology │ Division of Hematology-Oncology │ Department of Medicine │ The George Washington University School of Medicine and Health Sciences │ Washington, DC

4

Program Agenda

7:15 PM — 7:45 PMOptimizing Risk Assessment and Management of Cancer Patients at Risk for Venous Thromboembolism (VTE): Reducing DVT Recurrence and Related Complications

Craig Kessler, MDProfessor of Medicine │ Department of Hematology │ Anticoagulation Services │ Georgetown University Medical Center │ Washington, DC

7:45 PM — 8:15 PMRisk Stratification Tools to Identify Patients For Primary and Secondary Prevention of VTE in Patients with yMalignancy

Alok A. Khorana, MD, FACPVice-Chief, Division of Hematology/Oncology Associate Professor of Medicine and Oncology │ James P. Wilmot Cancer Center │ University of Rochester │ Rochester, NY

Program Agenda

8:15 PM — 8:30 PMProgram Chairman’s Concluding Vision Statement: Current and Near Future Perspectives of VTE Management in the Setting of Malignancy

Translating Scientific Advances into Clinical Practice

Program ChairmanSamuel Z. Goldhaber, MDCardiovascular Division │ Brigham and Women’s Hospital │Professor of Medicine │ Harvard Medical School

8:30 PM — 8:45 PMInteractive Q&A and Discussion Session

5

A Virtual World Forum on Cancer and Thrombosis on Your Laptop and iPad

Please Visit www.iQandA-cme.com for the iQ&A interactive Medical Intelligence Zone

for Cancer and Thrombosis

More than 100 questions answered by experts and investigators

6

New Frontiers and Evolving Paradigms i C A d Th b i

A Year 2010 Milestone Summit

in Cancer And ThrombosisEpidemiology, Trials, Guidelines

Program ChairmanSamuel Z. Goldhaber, MD

Cardiovascular DivisionBrigham and Women’s Hospital


Disclosures

Research SupportBMS B h i I lh i Ei i J hBMS; Boehringer-Ingelheim; Eisai; Johnson & Johnson, Sanofi-Aventis

ConsultantB h i I lh i BMS Ei i EKOSBoehringer-Ingelheim; BMS; Eisai; EKOS: Medscape; Merck; Pfizer; Sanofi-Aventis

7

New Frontiers and Evolving Paradigms inCancer and Thrombosis

Epidemiology

As Number of Cancer Survivors Increase, VTE Rates Increase

Stein PD, et al. Am J Med 2006; 119: 60-68

8

VTE Risk and Cancer Type: “Solid and Liquid”

n4.54

Relative Risk of VTE Ranged From 1.02 to 4.34R

elat

ive

Ris

k of

VTE

inC

ance

r Pat

ient

s3.532.521.510.5

Stein PD, et al. Am J Med 2006; 119: 60-68

Pan

crea

sB

rain

Mye

lopr

olS

tom

ach

Lym

phom

aU

teru

sLu

ngE

soph

agus

Pro

stat

eR

ecta

lK

idne

yC

olon

Ova

ryLi

ver

Leuk

emia

Bre

ast

Cer

vix

Bla

dder

Rate of PE Diagnosis is Increasing in the USA

250,000

200 000

Total cohortSurgical patientsNon-surgical patients

229,637

200,000

150,000

100,000

Non surgical patients

163,096

66,541

90,468

126,546

CHEST 2009; 136: 983-990

50,000

0

1998 1999 2000 2001 2002 2003 2004 2005

36,078

9

Hospital Costs are Skyrocketing Hospital Costs are Skyrocketing

CHEST 2009; 136: 983CHEST 2009; 136: 983--990990

10

DVT: Ominous DVT: Ominous SequellaeSequellae

►► 30% recur over 10 years (after anticoagulation is 30% recur over 10 years (after anticoagulation is y ( gy ( gdiscontinued)discontinued)

►► More than ½ of DVTs result in chronic venous More than ½ of DVTs result in chronic venous insufficiencyinsufficiency

►► Leads to PE, potentially fatalLeads to PE, potentially fatal►► Leads to PE, potentially fatalLeads to PE, potentially fatal

►► 1% to 4% of PEs evolve chronic thromboembolic 1% to 4% of PEs evolve chronic thromboembolic pulmonary hypertension (CTEPH)pulmonary hypertension (CTEPH)

Recurrent VTE is Common After A First Recurrent VTE is Common After A First Episode of Symptomatic DVTEpisode of Symptomatic DVT

3030

355 patients followed for 8 years355 patients followed for 8 years

1010

1515

2020

2525

3030

CumulativeCumulativeIncidence (%)Incidence (%)

00 11 22 33 44 55 66 77 8800

55

YearsYearsPrandoni et al, Prandoni et al, Ann Intern MedAnn Intern Med 1996;125:11996;125:1--77

11

Stages of Chronic Venous Stages of Chronic Venous InsufficiencyInsufficiency

1.1. Varicose veinsVaricose veins

2.2. Ankle/ leg edemaAnkle/ leg edema

3.3. Stasis dermatitisStasis dermatitis

4.4. LipodermatosclerosisLipodermatosclerosis

5.5. Venous stasis ulcerVenous stasis ulcer

Progression of Progression of Chronic Venous InsufficiencyChronic Venous Insufficiency

From UpToDate 2006

12

U.S.A.U.S.A.SURGEON SURGEON GENERAL:GENERAL:

CALL TO ACTIONCALL TO ACTIONCALL TO ACTION CALL TO ACTION TO PREVENT TO PREVENT DVT AND PEDVT AND PE

September 15, September 15, 2008200820082008

100,000-180,000 Deaths/year in USA

13

CTEPHCTEPH

RECURRENT RECURRENT ACUTE PEACUTE PE

Lang, I. M. NEJM 2004;350:2236Lang, I. M. NEJM 2004;350:2236--22382238

ACUTE PEACUTE PE

DVT FREE RegistryDVT FREE Registry

5,451 patients enrolled prospectively5,451 patients enrolled prospectivelyC ti t DVT di d bC ti t DVT di d b●● Consecutive acute DVT diagnosed by venous Consecutive acute DVT diagnosed by venous ultrasonographyultrasonography

●● No exclusionsNo exclusions●● 183 participating sites in the U.S183 participating sites in the U.S..

Goldhaber SZ, Goldhaber SZ, TapsonTapson VF. Am J VF. Am J CardiolCardiol 2004;93:2592004;93:259--262.262.

14

DVT FREE Registry DVT FREE Registry ((N= 5,541):N= 5,541):TOP 5 Medical TOP 5 Medical ComorbiditiesComorbidities

1.1. HypertensionHypertension2.2. ImmobilityImmobility3.3. CancerCancer4.4. Obesity (BMI > 30)Obesity (BMI > 30)5.5. Cigarette SmokingCigarette Smoking

Am J Am J CardiolCardiol 2004; 93: 2592004; 93: 259--262262

New Frontiers New Frontiers andand Evolving Paradigms Evolving Paradigms ininCancer Cancer andand ThrombosisThrombosis

Pivotal VTE Pivotal VTE Primary Prevention TrialsPrimary Prevention Trials

15

Trials of VTE Prophylaxis in Trials of VTE Prophylaxis in Hospitalized Medical PatientsHospitalized Medical Patients

►►MEDENOXMEDENOX (enoxaparin 40 mg)(enoxaparin 40 mg)●● SamamaSamama MM, et al. N MM, et al. N EnglEngl J Med. J Med. ,, gg

1999;341:7931999;341:793--800.800.►►PREVENTPREVENT ((dalteparindalteparin 5000 IU)5000 IU)

●● LeizoroviczLeizorovicz A, et al. Circulation. A, et al. Circulation. 2004;110:8742004;110:874--879.879.

►►ARTEMISARTEMIS (fondaparinux 2.5 mg)(fondaparinux 2.5 mg)●● Cohen AT, et al. Cohen AT, et al.

●● BMJ 2006; 332: 325BMJ 2006; 332: 325..

PREVENTPREVENT-- Dalteparin Trial Dalteparin Trial ((N= 3,681)N= 3,681)

►►A multicenter, randomized, controlled study in A multicenter, randomized, controlled study in acutely ill medical patientsacutely ill medical patientsacutely ill medical patients acutely ill medical patients

►►Compared the incidence in the Compared the incidence in the dalteparindalteparin and and placebo groups ofplacebo groups of::●● Any symptomatic VTEAny symptomatic VTEy y py y p●● Asymptomatic proximal DVT Asymptomatic proximal DVT ●● Sudden deathSudden death

Circulation 2004; 110: 874Circulation 2004; 110: 874--879879

16

ion

PREVENT Study Design PREVENT Study Design (N= 3,681)(N= 3,681)

Follow-up periodTreatment period

D l i N d d

Rand

omiz

at

Day 14 Day 90

Day 21 /

Dalteparin

Placebo

No study drug

No study drug

•• Dalteparin 5000 Units SC once daily (12Dalteparin 5000 Units SC once daily (12--14 d)14 d)

•• Placebo SC once daily (12Placebo SC once daily (12--14 d)14 d)

Primary endpoint/Bilateral leg U/S

Primary Efficacy Endpoint: Primary Efficacy Endpoint: VTE (Day 21)VTE (Day 21)

RiskRatio

Difference in Incidence

(%)Dalteparin

N=1518 Placebo

73

4.96%42

2.77%

0.38 to 0.80-3.57 to -0.8195% CI

0.55-2.19

Ratio(%)N=1473

P = 0.0015

Circulation 2004; 110: 874Circulation 2004; 110: 874--879879

17

Dalteparin Benefit Similar Dalteparin Benefit Similar Across Subgroups Across Subgroups

►►AgeAge

►►GenderGender

►►Cancer Cancer

►►ObesityObesity

►► Previous DVTPrevious DVT

Quality Improvement Initiative to Quality Improvement Initiative to Improve VTE ProphylaxisImprove VTE Prophylaxis

►►Randomized controlled trial to issue or Randomized controlled trial to issue or ithh ld l t i l t t MD hithh ld l t i l t t MD hwithhold electronic alerts to MDs whose withhold electronic alerts to MDs whose

highhigh--risk patients were not receiving risk patients were not receiving VTE prophylaxisVTE prophylaxis

Kucher N et al. NEJM 2005; 352: 969

18

Computer ProgramComputer Program

►►We developed a computer program linked to We developed a computer program linked to the patient database that screened the the patient database that screened the system daily to identify highsystem daily to identify high--risk patientsrisk patients..

►►We included consecutive highWe included consecutive high--risk patients risk patients on medical and surgical services who were on medical and surgical services who were not receiving DVT prophylaxisnot receiving DVT prophylaxisnot receiving DVT prophylaxis.not receiving DVT prophylaxis.

Kucher N et al. NEJM 2005; 352: 969

Definition: “High Risk”Definition: “High Risk”

VTE risk score ≥ 4 points:VTE risk score ≥ 4 points:►► CancerCancer 33 (ICD codes)(ICD codes)►► Prior VTEPrior VTE 33 (ICD codes)(ICD codes)►► HypercoagulabilityHypercoagulability 33 (Leiden, ACLA)(Leiden, ACLA)►► Major surgeryMajor surgery 22 (> 60 minutes)(> 60 minutes)►► Bed restBed rest 11 (“bed rest” order)(“bed rest” order)►► Bed restBed rest 11 ( bed rest order)( bed rest order)►► Advanced ageAdvanced age 11 (> 70 years)(> 70 years)►► ObesityObesity 11 (BMI > 29 kg/m(BMI > 29 kg/m22))►► HRT/OCHRT/OC 11 (order entry)(order entry)

19

RandomizationRandomization

VTE risk score > 4No prophylaxis

N = 2,506

INTERVENTION: CONTROLN lSingle alert

N = 1,255No computer alert

N = 1,251

KucherKucher N, et al. NEJM 2005;352:969N, et al. NEJM 2005;352:969--977977

20

Baseline CharacteristicsBaseline Characteristics

►►Median age:Median age: 62.5 years62.5 yearsM di l iM di l i 83%83%►►Medical services:Medical services: 83%83%

►►Surgical services:Surgical services: 17%17%►►ComorbiditiesComorbidities

●● Cancer:Cancer: 80%80%●● Hypertension:Hypertension: 34%34%●● Infection:Infection: 30%30%●● Prior VTE:Prior VTE: 20%20%

KucherKucher N, et al. NEJM 2005;352:969N, et al. NEJM 2005;352:969--977977

Primary End Point

T/P

E

98

100

Intervention

Control

%Fr

eedo

m f

rom

DV

T

90

92

94

96

98

InterventionControl

Number at risk1255 977 900 8531251 976 893 839

Time (days)0 30 60 90

% 90

Kucher N, et al. NEJM 2005;352:969-977

21

l ll l


Pivotal VTE Treatment Trial Pivotal VTE Treatment Trial in Patients with Cancerin Patients with Cancer

Cancer and VTECancer and VTE

►► 33--fold higher recurrence and bleeding, fold higher recurrence and bleeding, when when treating cancer patients treating cancer patients (Prandoni. Blood 2002; (Prandoni. Blood 2002; 100: 3484100: 3484))

►► LMWH LMWH MonotherapyMonotherapy halves recurrence, halves recurrence, compared with warfarin. compared with warfarin.

FDA FDA approved May approved May 20072007

Lee AYY. NEJM 2003; 349:146Lee AYY. NEJM 2003; 349:146

22

“CLOT “CLOT Trial”Trial”

►►Dalteparin Dalteparin monotherapymonotherapy for 6 months was for 6 months was more effective (8 8% vs 17% recurrence)more effective (8 8% vs 17% recurrence)more effective (8.8% vs. 17% recurrence) more effective (8.8% vs. 17% recurrence) than warfarin in than warfarin in 672 cancer patients with 672 cancer patients with DVTDVT..

►►Dalteparin dose:Dalteparin dose: 200 u/kg daily 1200 u/kg daily 1stst monthmonth►►Dalteparin dose: Dalteparin dose: 200 u/kg daily 1200 u/kg daily 1stst month, month, then 150 u/kg daily.then 150 u/kg daily.

Agnes Lee, et al. NEJM 2003; 349:146Agnes Lee, et al. NEJM 2003; 349:146--153)153)

Dalteparin Reduces VTE Recurrence in Dalteparin Reduces VTE Recurrence in Cancer Patients Cancer Patients ((N = 676)N = 676)

CLOT TRIAL

NEJM 2003; 349:146NEJM 2003; 349:146--153153

23

LMWH LMWH MonotherapyMonotherapy

►► Cancer patients with DVT/PECancer patients with DVT/PE

A ti t h f il f i (hA ti t h f il f i (h►► Any patient who fails warfarin (has Any patient who fails warfarin (has recurrent DVT/PE) despite target INRrecurrent DVT/PE) despite target INR

►► Difficulty maintaining target INRDifficulty maintaining target INR

►► Poor GI absorption of medsPoor GI absorption of meds►► Poor GI absorption of medsPoor GI absorption of meds

►► Alopecia or rash from CoumadinAlopecia or rash from Coumadin

►► “Bridging”“Bridging”

ACCP ACCP VTE Rx in Cancer: Guidelines VTE Rx in Cancer: Guidelines 88thth EditionEdition

1.1. At least 3 months of LMWH.At least 3 months of LMWH.

2.2. Then administer LMWH or warfarin as Then administer LMWH or warfarin as long as the cancer is active. long as the cancer is active.

3.3. Indefinite duration anticoagulation after Indefinite duration anticoagulation after 22ndnd unprovoked VTEunprovoked VTE22 unprovoked VTE.unprovoked VTE.

CHEST 2008; 133: 454SCHEST 2008; 133: 454S

24

ConclusionsConclusions

1.1. Cancer and VTE are closely linked.Cancer and VTE are closely linked.

22 Cancer increases VTE risk and may beCancer increases VTE risk and may be2.2. Cancer increases VTE risk and may be Cancer increases VTE risk and may be occult when VTE is diagnosed.occult when VTE is diagnosed.

3.3. Cancer patients are at high risk for VTE Cancer patients are at high risk for VTE but receive less prophylaxis than any but receive less prophylaxis than any other atother at risk group of hospitalizedrisk group of hospitalizedother atother at--risk group of hospitalized risk group of hospitalized patients.patients.

4.4. Dalteparin 5,000 U/d is effective for Dalteparin 5,000 U/d is effective for VTE prophylaxis in cancer patients.VTE prophylaxis in cancer patients.

Conclusions (Continued)Conclusions (Continued)

5.5. Dalteparin 200 U/kg/day is effective Dalteparin 200 U/kg/day is effective fo t e tment of te VTEfo t e tment of te VTEfor treatment of acute VTE as for treatment of acute VTE as monotherapymonotherapy without warfarin. without warfarin.

6.6. NCCN, ASCO, and ACCP guidelines NCCN, ASCO, and ACCP guidelines endorse VTE prevention with LMWH endorse VTE prevention with LMWH and VTE treatment of cancer patients and VTE treatment of cancer patients with LMWH alone (with LMWH alone (monotherapymonotherapywithout warfarin).without warfarin).



The Complex Interface of Malignancy, Thrombosis, and Clinical End Points

Faculty: Frederick R. Rickles, MD Time: 6:50 – 7:15 p.m.

Notes:

25

The Role of the Coagulation The Role of the Coagulation Cascade in MalignantCascade in Malignant

The Role of the Coagulation The Role of the Coagulation Cascade in MalignantCascade in Malignant


Cascade in Malignant Cascade in Malignant Transformation Transformation

Can Anticoagulation Affect Cancer Survival?Can Anticoagulation Affect Cancer Survival?

Cascade in Malignant Cascade in Malignant Transformation Transformation

Can Anticoagulation Affect Cancer Survival?Can Anticoagulation Affect Cancer Survival?

Frederick R. Rickles, MDFrederick R. Rickles, MDProfessor of Medicine, Pediatrics, Professor of Medicine, Pediatrics,

Pharmacology and PhysiologyPharmacology and PhysiologyThe George Washington UniversityThe George Washington University

Washington, DCWashington, DC

Frederick R. Rickles, MDFrederick R. Rickles, MDProfessor of Medicine, Pediatrics, Professor of Medicine, Pediatrics,

Pharmacology and PhysiologyPharmacology and PhysiologyThe George Washington UniversityThe George Washington University

Washington, DCWashington, DC

26

DisclosuresDisclosures

ConsultantConsultantGenmabGenmab, Bayer/Ortho, Bayer/Ortho‐‐McNeil/J & J,McNeil/J & J,

PharmacyclicsPharmacyclics, Leo, Leo

Speaker’s BureauSpeaker’s BureauEisaiEisai

Fibrinolytic activities:t-PA u-PA u-PAR

Procoagulant Activities

Tumor cells

Angiogenesis,Basement matrix degradation

Interface of Coagulation and CancerInterface of Coagulation and Cancer

t-PA, u-PA, u-PAR, PAI-1, PAI-2

FIBRIN

IL-1, TNF-α, VEGF

PMN leukocyte

Activation of coagulation

degradation. TFTF--rich MPsrich MPs

Endothelial cellsMonocyte

Platelets

Falanga and Rickles, Falanga and Rickles, New Oncology:ThrombosisNew Oncology:Thrombosis, 2005; , 2005; Hematology, Hematology, ASH Education BookASH Education Book, , 20072007

27

Mechanisms of CancerMechanisms of Cancer-- Induced Induced Thrombosis: Clot and Cancer InterfaceThrombosis: Clot and Cancer Interface

1.1. Pathogenesis?Pathogenesis?

2.2. Biological significance?Biological significance?

3.3. Anticoagulation and cancer survival?Anticoagulation and cancer survival?

Activation of Blood Coagulation in CancerActivation of Blood Coagulation in CancerBiological Significance?Biological Significance?

►► EpiphenomenonEpiphenomenon? ? I thi i d t hI thi i d t hIs this a generic secondary event where Is this a generic secondary event where thrombosis is an incidental findingthrombosis is an incidental finding

oor, is clotting activation . . .r, is clotting activation . . .

►► A Primary Event?A Primary Event?Linked to malignant transformation Linked to malignant transformation

28

Blood CoagulationActivation

FVII/FVIIaFVII/FVIIaTumor Cell

TF

Interface of Clotting Activation Interface of Clotting Activation and Tumor Biology and Tumor Biology

VEGF

Angiogenesis

FIBRIN

THROMBINTHROMBIN

TF

Endothelial cellsEndothelial cells

ILIL--88PAR-2

Angiogenesis

Falanga and Rickles, New Oncology:Thrombosis, 2005;1:9-16; Ruf. J Thromb Haemost 2007 5 1584

Coagulation Cascade and Tumor BiologyCoagulation Cascade and Tumor Biology

TFTF ThrombinThrombinClottingClotting--

dependentdependentClottingClotting--

dependentdependentFibrinFibrin

ClottingClotting--independentindependent

ClottingClotting--dependentdependent

ClottingClotting--independentindependent

PARsPARs

VIIaVIIa XaXa

Fernandez, Patierno and Rickles. Sem Hem Thromb 2004;30:31; Ruf. J Thromb Haemost 2007;5:1584

Angiogenesis, Tumor Angiogenesis, Tumor Growth and MetastasisGrowth and Metastasis

29

In Situ In Situ Localization ofLocalization of Tissue Factor in Vascular Endothelium Tissue Factor in Vascular Endothelium of Human Lung of Human Lung AdenocarcinomaAdenocarcinoma –– coco-- localization with localization with vWFvWF

Shoji et al, Amer J Pathol 1998;152:399-411

In Situ In Situ localization of Tissue Factor in Tumor Vascular localization of Tissue Factor in Tumor Vascular Endothelium in Invasive Human Breast Cancer Endothelium in Invasive Human Breast Cancer

Contrino et al. Nature Med 1996;2:209-215

30

In Situ In Situ CoCo-- Localization of TF and VEGF Localization of TF and VEGF mRNA in Lung mRNA in Lung AdenocarcinomaAdenocarcinoma

H&EH&E

TFTF

Shoji et al. Amer J Pathol 1998;152:399-411

VEGFVEGF

Human melanoma cell lines grown Human melanoma cell lines grown as as xenogeneicxenogeneic tumorstumors in SCID micein SCID mice

TF high producerTF high producer

Abe K et al. PNAS 1999;96:8663-8668

©1999 by The National Academy of Sciences

TF low producerTF low producer

31

Regulation of Vascular Endothelial Growth Factor Production and Regulation of Vascular Endothelial Growth Factor Production and Angiogenesis by the Angiogenesis by the CytoplasmicCytoplasmic Tail of Tissue FactorTail of Tissue Factor

11 TF regulates VEGF expression inTF regulates VEGF expression in1.1. TF regulates VEGF expression in TF regulates VEGF expression in human cancer cell lineshuman cancer cell lines

2.2. Human cancer cells with increased TF Human cancer cells with increased TF are more are more angiogenicangiogenic (and, therefore, (and, therefore,

“ t t ti ’)“ t t ti ’) i ii i d t hi hd t hi hmore “metastatic’) more “metastatic’) in vivoin vivo due to high due to high VEGF productionVEGF production

Abe et al Abe et al Proc Nat Acad SciProc Nat Acad Sci 1999;96:86631999;96:8663--8668; Ruf et al 8668; Ruf et al Nature MedNature Med 2004;10:502-509

3.3. The The cytoplasmiccytoplasmic tail of TF, which contains three tail of TF, which contains three serine residues, appears to play a role in regulating serine residues, appears to play a role in regulating VEGF expression in human cancer cells perhaps byVEGF expression in human cancer cells perhaps by

Regulation of Vascular Endothelial Growth Factor Production Regulation of Vascular Endothelial Growth Factor Production and Angiogenesis by the and Angiogenesis by the CytoplasmicCytoplasmic Tail of Tissue FactorTail of Tissue Factor

VEGF expression in human cancer cells, perhaps by VEGF expression in human cancer cells, perhaps by mediating signal transductionmediating signal transduction

4.4. ThisThisaa and other data on signaling pathways activated and other data on signaling pathways activated by TF/by TF/VIIaVIIa engagement of PARengagement of PAR--22bb and/or thrombin and/or thrombin engagement of PARengagement of PAR--11cc suggest that clotting suggest that clotting pathways are directly involved in regulating tumorpathways are directly involved in regulating tumorpathways are directly involved in regulating tumor pathways are directly involved in regulating tumor growth, angiogenesis and metastasisgrowth, angiogenesis and metastasis

5.5. Is this a paradigm shift?Is this a paradigm shift?aa Abe et al Abe et al Proc Nat Acad SciProc Nat Acad Sci 1999;96:86631999;96:8663--68 68 bb Ruf et al Ruf et al Nature MedNature Med 2004;10:5022004;10:502--9 9 cc Karpatkin et al Karpatkin et al Cancer Res Cancer Res 2009;69:33742009;69:3374--8181

32

Activation of Blood Coagulation Activation of Blood Coagulation in Cancer and Malignant Transformationin Cancer and Malignant Transformation

Epiphenomenon vs. Malignant Transformation?Epiphenomenon vs. Malignant Transformation?Paradigm Shift (2005)Paradigm Shift (2005)

1.1. METMET oncogene induction produces DIC in human liveroncogene induction produces DIC in human liver1.1. METMET oncogene induction produces DIC in human liver oncogene induction produces DIC in human liver carcinomacarcinoma (Boccaccio lab)(Boccaccio lab)

(Boccaccio et al (Boccaccio et al Nature 2005;434:3962005;434:396--400)400)

2.2. PtenPten loss and loss and EGFREGFR amplification produce TF activation amplification produce TF activation and pseudopalisading necrosis through JunD/Activator and pseudopalisading necrosis through JunD/Activator ProteinProtein--1 in human glioblastoma 1 in human glioblastoma (Bratt lab)(Bratt lab)

(Rong et al (Rong et al Cancer Res 2005;65:14062005;65:1406--1413; 1413; Cancer Res 2009;69:25402009;69:2540--9)9)

3.3. KK--rasras oncogene, oncogene, p53p53 inactivation and TF induction in inactivation and TF induction in human colorectal carcinoma; TF and angiogenesis human colorectal carcinoma; TF and angiogenesis regulation in epithelial tumors by regulation in epithelial tumors by EGFR (ErbB1)EGFR (ErbB1) ––relationship to EMTsrelationship to EMTs (Rak lab)(Rak lab)

(Yu et al (Yu et al Blood 2005;105:17342005;105:1734--1741; Milson et al 1741; Milson et al Cancer Res 2008;68:100682008;68:10068--76)76)

“1. “1. METMET Oncogene Drives a Genetic Programme Oncogene Drives a Genetic Programme Linking Cancer to Haemostasis”Linking Cancer to Haemostasis”

Activation of Blood Coagulation Activation of Blood Coagulation in Cancer: Malignant Transformationin Cancer: Malignant Transformation

►► METMET encodes a tyrosine kinase receptor for hepatocyte encodes a tyrosine kinase receptor for hepatocyte growth factor/scatter factor (HGF/SF) growth factor/scatter factor (HGF/SF) →→●● Drives physiologicalDrives physiological cellular program of “invasive cellular program of “invasive

growth” (tissue morphogenesis, angiogenesis growth” (tissue morphogenesis, angiogenesis and repair)and repair)and repair)and repair)

●● Aberrant execution (e.g. hypoxiaAberrant execution (e.g. hypoxia--induced induced transcription) is associated with neoplastic transcription) is associated with neoplastic transformation, invasion, and metastasistransformation, invasion, and metastasis

Boccaccio et al Boccaccio et al Nature 2005;434:3962005;434:396--400400

33


2. “2. “PtenPten and Hypoxia Regulate Tissue Factor and Hypoxia Regulate Tissue Factor Expression and Plasma Coagulation ByExpression and Plasma Coagulation ByExpression and Plasma Coagulation By Expression and Plasma Coagulation By

Glioblastoma”Glioblastoma”►► PtenPten = tumor suppressor with lipid and protein = tumor suppressor with lipid and protein

phosphatase activityphosphatase activity►► Loss or inactivation of Loss or inactivation of Pten Pten (70(70--80% of 80% of

glioblastomas) leads to Akt activation andglioblastomas) leads to Akt activation andglioblastomas) leads to Akt activation and glioblastomas) leads to Akt activation and upregulation of upregulation of RasRas/MEK/ERK/MEK/ERK signaling cascade signaling cascade

Rong et al Ca Res 2005;65:1406-1413

►► Glioblastomas characterized histologically by Glioblastomas characterized histologically by “pseudopalisading necrosis”“pseudopalisading necrosis”

““PtenPten and Hypoxia Regulate Tissue Factor Expression and Hypoxia Regulate Tissue Factor Expression and Plasma Coagulation By and Plasma Coagulation By GlioblastomaGlioblastoma””

pseudopalisading necrosis pseudopalisading necrosis

►► Thought to be wave of tumor cells migrating Thought to be wave of tumor cells migrating away from a central hypoxic zone, perhaps away from a central hypoxic zone, perhaps created by thrombosiscreated by thrombosis

►► Pseudopalisading cells produce VEGF and ILPseudopalisading cells produce VEGF and IL--8 8 and drive angiogenesis and rapid tumor growth and drive angiogenesis and rapid tumor growth

►► TF expressed by >90% of grade 3 and 4 TF expressed by >90% of grade 3 and 4 malignant astrocytomas (but only 10% of malignant astrocytomas (but only 10% of grades 1 and 2)grades 1 and 2)

34

Results:Results:


1.1. Hypoxia and Hypoxia and PTEN PTEN loss loss →↑→↑ TF (mRNA, Ag and TF (mRNA, Ag and procoagulant activity); partially reversed with procoagulant activity); partially reversed with induction of induction of PTEN PTEN

2.2. Both Both AktAkt and and RasRas pathways modulated TF in pathways modulated TF in sequentially transformed astrocytes.sequentially transformed astrocytes.

3.3. Ex vivo Ex vivo data: data: ↑↑ TF (by IHTF (by IH--chemical staining) in chemical staining) in pseudopalisades of # 7 human glioblastoma pseudopalisades of # 7 human glioblastoma specimensspecimens

Pseudopalisading necrosis


H&EH&E

Vascular Endothelium

TF IHC

35

Activation of Blood Coagulation Activation of Blood Coagulation in Cancer: Malignant Transformain Cancer: Malignant Transformationtion

3. “Oncogenic Events Regulate Tissue Factor Expression 3. “Oncogenic Events Regulate Tissue Factor Expression In Colorectal Cancer Cells: Implications For Tumor In Colorectal Cancer Cells: Implications For Tumor P i A d A i i ”P i A d A i i ”Progression And Angiogenesis”Progression And Angiogenesis”

►► Activation of KActivation of K--ras ras oncogene and inactivation of oncogene and inactivation of p53 p53 tumor tumor suppressor suppressor ↑↑ TF expression in TF expression in human human colorectal cancer cellscolorectal cancer cells

►► Transforming events dependent on MEK/MAPK and PI3KTransforming events dependent on MEK/MAPK and PI3K►► CellCell--associated and MPassociated and MP--associated TF activity linked to genetic associated TF activity linked to genetic

status of cancer cellsstatus of cancer cellsstatus of cancer cellsstatus of cancer cells►► TF siRNA reduced cell surface TF expression, tumor growth and TF siRNA reduced cell surface TF expression, tumor growth and

angiogenesis angiogenesis ►► TF may be required for KTF may be required for K--rasras--driven phenotype driven phenotype

Yu et al Yu et al Blood 2005;105:17342005;105:1734--4141

“Oncogenic Events Regulate Tissue Factor “Oncogenic Events Regulate Tissue Factor Expression In Colorectal Cancer Cells: Expression In Colorectal Cancer Cells:

Implications For Tumor Progression AndImplications For Tumor Progression And


Implications For Tumor Progression And Implications For Tumor Progression And Angiogenesis”Angiogenesis”

Effect of TF si mRNA on tumor growth in vitro and in vivo

Yu et al Yu et al Blood 2005;105:17342005;105:1734--4141

36

“Oncogenic Events Regulate Tissue Factor Expression In “Oncogenic Events Regulate Tissue Factor Expression In Colorectal Cancer Cells: Implications For TumorColorectal Cancer Cells: Implications For Tumor


Colorectal Cancer Cells: Implications For Tumor Colorectal Cancer Cells: Implications For Tumor Progression And Angiogenesis”Progression And Angiogenesis”

Matrigel Assay: (D) HCT 116; (E) SI-3 cells – vWF immunohistology

Similar amplification of TF with upregulated VEGF induced by mutated EGFR in glioblastoma and lung Similar amplification of TF with upregulated VEGF induced by mutated EGFR in glioblastoma and lung cancer cells; accompanied by epithelialcancer cells; accompanied by epithelial--toto--mesenchymal transition (EMT)mesenchymal transition (EMT)

Milsom et al Milsom et al CA Res 2008;68:100682008;68:10068--7676Yu et al Yu et al Blood 2005;105:17342005;105:1734--4141

MicroparticlesMicroparticles

•• Originate directly from membrane surface of activated or Originate directly from membrane surface of activated or apoptotic cellapoptotic cell

•• Express surface antigens derived from parent cellExpress surface antigens derived from parent cellExpress surface antigens derived from parent cellExpress surface antigens derived from parent cell

•• AnucleateAnucleate

•• <1 µm in diameter<1 µm in diameter

•• Procoagulant activityProcoagulant activity

mediated by TF and/or PSmediated by TF and/or PS

Burnier L et al. Thromb Haemost 2009;101:439‐451

37

Cumulative incidence of VTE in cancer patients with (Cumulative incidence of VTE in cancer patients with (----) ) /without ( /without ( −− ) circulating TF) circulating TF--bearing bearing microparticlesmicroparticles

Zwicker et al. Clin Cancer Res 2009;15:6830-40

Microparticle TF PCA in Microparticle TF PCA in Cancer Patients Cancer Patients ±± VTEVTE

Manly DA, et al. Thromb Res 2010;125:511‐512

38


►► QQ: What do these experiments tell us?: What do these experiments tell us?

►► AA: They suggest two things:: They suggest two things:●● Tumor cellTumor cell--derived, derived, TFTF--rich rich microparticlesmicroparticles

(MPs) may be important as a predictive test (MPs) may be important as a predictive test for VTEfor VTEAll ti t ithAll ti t ith d id i●● All patients with All patients with oncogeneoncogene--driven cancer may driven cancer may need prophylactic anticoagulation need prophylactic anticoagulation

Mechanisms of CancerMechanisms of Cancer-- Induced Induced Thrombosis: ImplicationsThrombosis: Implications

1.1. Pathogenesis?Pathogenesis?

2.2. Biological significance?Biological significance?

3.3. Anticoagulation and cancer Anticoagulation and cancer survival ?survival ?

39

Anticoagulants and SurvivalAnticoagulants and Survival

►► Inconclusive evidence to dateInconclusive evidence to date►► Experimental data supportive of antitumor Experimental data supportive of antitumor

effects but exact mechanisms not establishedeffects but exact mechanisms not establishedeffects but exact mechanisms not establishedeffects but exact mechanisms not established►► Clinical trials provide supportive data for Clinical trials provide supportive data for

LMWH but are heterogeneous in design and LMWH but are heterogeneous in design and methodology:methodology:●● Tumour typesTumour types●● Stage or course of diseaseStage or course of diseasegg●● Treatment history or concurrent cancer therapiesTreatment history or concurrent cancer therapies●● LMWH agentsLMWH agents●● Doses and regimens of LMWHsDoses and regimens of LMWHs

A Lee ICTHIC, 2010A Lee ICTHIC, 2010

Survival Effect of AnticoagulantsSurvival Effect of Anticoagulants

Kuderer N et al. Cancer 2007;110:1149-60.

40

►► Multicentre, doubleMulticentre, double--blind, placeboblind, placebo--controlled RCTcontrolled RCT

►► Advanced lung, breast, GI, pancreas, ovary, H+NAdvanced lung, breast, GI, pancreas, ovary, H+N

PROTECHT StudyPROTECHT Study

►► NadroparinNadroparin vsvs placebo for duration of chemo (up to 4m)placebo for duration of chemo (up to 4m)

NadroparinNadroparin PlaceboPlacebo PP--valuevalue NNT/HNNT/H

No. PatientsNo. Patients 769769 381381

11°° endpoint: VTE + endpoint: VTE + ATEATE 2.0%2.0% 3.9%3.9% 0.02*0.02* 5454ATEATE

Major bleedingMajor bleeding 0.7%0.7% 00 0.180.18 154154

DeathDeath 4.3%4.3% 4.2%4.2%

11--yr mortalityyr mortality 43%43% 41%41%

Agnelli et al. Lancet 2009;10:943-949. *1-sided

Prophylaxis in Pancreatic CancerProphylaxis in Pancreatic Cancer

CONKO 004 FRAGEM

6%

8%

10%

no treatment enoxaparin

P<0.01

P=0.6 20%

30%

40%

no treatment dalteparin

P<0.02

No survival difference

Riess et al. ASCO May 2009 and ISTH July 2009. Maraveyas et al. Presented at ESMO 2009.

0%

2%

4%

VTE bleeding

0%

10%

VTE fatal PE Gr 3 bleed

P=0.03

NS

41

Key QuestionsKey QuestionsKey QuestionsKey Questions

Cancer and Thrombosis Cancer and Thrombosis Year 2010 StateYear 2010 State-- ofof-- thethe-- Science UpdateScience Update


1. 1. Does activation of blood coagulation affect the Does activation of blood coagulation affect the biology of cancer positively or negatively?biology of cancer positively or negatively?

2. 2. Can we treat tumors more effectively using Can we treat tumors more effectively using coagulation protein targets?coagulation protein targets?

3. 3. Can anticoagulation alter the biology of cancer?Can anticoagulation alter the biology of cancer?

1.1. Epidemiologic evidence isEpidemiologic evidence is suggestivesuggestive that VTE is a badthat VTE is a bad



Tentative AnswersTentative Answers1. 1. Epidemiologic evidence is Epidemiologic evidence is suggestivesuggestive that VTE is a bad that VTE is a bad

prognostic sign in cancerprognostic sign in cancer

2. 2. Experimental evidence is Experimental evidence is supportive supportive of the use of of the use of antithrombotic strategies for both prevention of antithrombotic strategies for both prevention of thrombosis and inhibition of tumor growth thrombosis and inhibition of tumor growth

3. 3. Results of recent, randomized clinical trials of LMWHs in Results of recent, randomized clinical trials of LMWHs in cancer patients indicate superiority to oral agents in cancer patients indicate superiority to oral agents in preventing recurrent VTE; increasing survival (preventing recurrent VTE; increasing survival (notnot due to due to prevention of VTE) not clearprevention of VTE) not clear

42

Survival StudiesSurvival Studies►► INPACT (NSCLC, prostate, pancreatic)INPACT (NSCLC, prostate, pancreatic)

LMWH in CancerLMWH in Cancer

( , p , p )( , p , p )●● nadroparinnadroparin + chemo vs. chemo+ chemo vs. chemo

►► ABEL (limited SCLC)ABEL (limited SCLC)●● bemiparinbemiparin + chemo vs. chemo+ chemo vs. chemo

►► TILT (TILT (nonsmallnonsmall cell lung cancer)cell lung cancer)●● tinzaparintinzaparin + chemo + chemo vsvs chemochemo

►► FRAGMATIC (newly diagnosed lung cancer)FRAGMATIC (newly diagnosed lung cancer)●● dalteparindalteparin + chemo + chemo vsvs chemochemo

A Lee ICTHIC, 2010A Lee ICTHIC, 2010Stay tuned !Stay tuned !



Optimizing Risk Assessment and Management of Cancer Patients at Risk for Venous Thromboembolism (VTE): Reducing DVT Recurrence

and Related Complications Faculty: Craig Kessler, MD Time: 7:15 – 7:45 p.m. Notes:

43

Optimizing Risk Assessment and Optimizing Risk Assessment and Management of Cancer Patients at RiskManagement of Cancer Patients at Risk


Management of Cancer Patients at Risk Management of Cancer Patients at Risk for Venous Thromboembolism (for Venous Thromboembolism (VTE)VTE)

Reducing Reducing DVT Recurrence and Related ComplicationsDVT Recurrence and Related Complications

Craig Kessler, MDCraig Kessler, MDProfessor of MedicineProfessor of Medicine

Department of HematologyDepartment of HematologyAnticoagulation ServicesAnticoagulation Services

Georgetown University Medical CenterGeorgetown University Medical CenterWashington, DCWashington, DC

COI Financial DisclosuresCOI Financial Disclosures

Grant/Research Support: Grant/Research Support: GlaxoSmithKlineGlaxoSmithKline sanofisanofi--aventisaventis EisaiEisaiGlaxoSmithKlineGlaxoSmithKline, , sanofisanofi aventisaventis, , EisaiEisai

Consultant: Consultant: sanofisanofi--aventisaventis, Eisai , Eisai

44

OutlineOutline

►►Guidelines Guidelines for VTE for VTE prevention in cancer prevention in cancer patientspatientspp

►►Performance to datePerformance to date►►Opportunities for improvementOpportunities for improvement►►Guidelines for VTE TreatmentGuidelines for VTE Treatment►►Performance toPerformance to datedate►►Performance to Performance to datedate►► LMWHsLMWHs——What Do the Trials, NCCN and What Do the Trials, NCCN and

ASCO Guidelines Teach Us About Duration ASCO Guidelines Teach Us About Duration of Therapy and Patients at Risk?of Therapy and Patients at Risk?

Recommendations for Venous Th b b li P h l i dThromboembolism Prophylaxis and Treatment in Patients with Cancer

ASCO Clinical Practice Guideline

•www.nccn.org

•NCCN Clinical Practice Guidelines in Oncology™

•Guidelines for supportiveGuidelines for supportive care

•“…the panel of experts includes a medical and surgical oncologists, hematologists, cardiologists, internists, radiologists. And a pharmacist.”

45

Importance of Guidelines to Clinical Importance of Guidelines to Clinical OutcomesOutcomes

“Clinicians armed with appropriate assessments “Clinicians armed with appropriate assessments and the best evidenceand the best evidence--based practice guidelines based practice guidelines

d f th l t d f td f th l t d f tcan reduce some of the unpleasant and frequent can reduce some of the unpleasant and frequent sideside--effects that often accompany cancer and effects that often accompany cancer and chemotherapy treatment, obtain the best chemotherapy treatment, obtain the best possible clinical outcomes, and avoid possible clinical outcomes, and avoid unnecessary costs.”unnecessary costs.”

Statement from Centers for Medicare and Medicaid Services, August 2005Statement from Centers for Medicare and Medicaid Services, August 2005

Incidence of VTE in US Patients Incidence of VTE in US Patients With With Cancer Has Risen Although the Overall Cancer Has Risen Although the Overall

Incidence of Cancer Has Not ChangedIncidence of Cancer Has Not Changed

4 Highest incidence of VTE: pancreatic CA (4.3%)

1

2

3

4

VTE

Inci

denc

e, %

Cancer No CancerLowest incidence of VTE: oral cavity, or pharynx

01979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999

Stein PD et al. Am J Med. 2006;119:60‐68.

National Hospital Discharge Survey data from 19 types of malignancies from 1979 through 1999 (non-age dependent)

46

Venous Thromboembolism in Cancer PatientsVenous Thromboembolism in Cancer Patients

Of all cases of VTE:Of all cases of VTE:20% occur in cancer patients20% occur in cancer patients

Of all cancer patients:Of all cancer patients:0 5% will have symptomatic VTE0 5% will have symptomatic VTE0.5% will have symptomatic VTE0.5% will have symptomatic VTEAs high as 50% have VTE at autopsyAs high as 50% have VTE at autopsy

Compared to patients without cancer:Compared to patients without cancer:Higher risk of first and recurrent VTEHigher risk of first and recurrent VTEHigher risk of bleeding on anticoagulantsHigher risk of bleeding on anticoagulantsHigher risk of dyingHigher risk of dying

VTE may be the presenting sign of occult malignancyVTE may be the presenting sign of occult malignancy10% with idiopathic VTE develop cancer within 2 10% with idiopathic VTE develop cancer within 2

yearsyears20% have recurrent idiopathic VTE20% have recurrent idiopathic VTE25% have bilateral DVT25% have bilateral DVT

Lee & Levine. Circulation 2003;107:I17 – I21; BuraBura et. al., J et. al., J ThrombThromb HaemostHaemost 2004;2:4452004;2:445--5151

Thrombosis and Survival:Likelihood of Death After Hospitalization

(Levitan N, et al. Medicine 1999;78:285)

1.00DVT/PE d M li

0.40

0.80

0.60

DVT/PE and Malignant Disease

Malignant Disease

DVT/PE Onlyroba

bilit

y of

D

eath

0 20 40 60 80 100 120 140 160 1800.00

0.20

y

Nonmalignant Disease

Number of Days

Pr

47

Cancer and Venous ThromboembolismThe Need for Risk Stratification

4.5 Chemotherapy End of Life

1 5

2

2.5

3

3.5

4

Diagnosis

Chemotherapy

Hospitalization

End of Life

Metastasis

ative Risk

0

0.5

1

1.5

1 2 3 4 5 6

Remission

Average Risk

Time

Rela

VTE in Hospitalized Cancer PatientsVTE in Hospitalized Cancer Patients

6.0VTE all patientsDVT all patientsPE all patients

7.0 VTE patients on chemotherapy

1 0

2.0

3.0

4.0

5.0

Rate of V

TE (%

)

p

Cancer 2007

0.0

1.0

1995 1996 1997 1998 1999 2000 2001 2002 2003Years

48

Effect of Malignancy on Risk of VTE• Population-based case-control (MEGA)

study• N=3220 consecutive patients with 1st

VTE vs. n=2131 control subjects• CA patients = OR 7x VTE risk vs. non-

CA patients

40

5053.5

ratio

MEGA = Multiple Environmentaland Genetic Assessment case‐control study

0

10

20

30al ng na

l

ast

ant

es hs hs ars

ars

ars

2822.2

20.3

4.9

19.8

14.3

2.6 1.13.6

Adjusted od

ds

VTE = venous thromboembolism; CA = cancer; OR = odds ratio. Silver In: The Hematologist ‐modified from Blom JW, et. al. JAMA. 2005;293:715‐722.

Hem

atol

ogic Lu

Gas

troin

test

in

Bre

a

Dis

tam

etas

tase

0 to

3 m

onth

3 to

12

mon

th

1 to

3 y

ea

5 to

10

yea

> 15

yea

Type of cancer Time since cancer diagnosis

The Importance of DVT Prophylaxis The Importance of DVT Prophylaxis in Patients With Cancer: ASCO Guidelinesin Patients With Cancer: ASCO Guidelines

►► VTE is a leading causes of death in CA, occurring in 4% to VTE is a leading causes of death in CA, occurring in 4% to 20% pts20% pts

►► Hospitalized CA pt and those on chemotx have greatest VTE Hospitalized CA pt and those on chemotx have greatest VTE risk risk ●● Cancer increased the risk of VTE 4.1Cancer increased the risk of VTE 4.1--foldfold●● Chemotherapy increased the risk 6.5Chemotherapy increased the risk 6.5--foldfold

►► Major risk factors: older age, comorbid conditions, recent Major risk factors: older age, comorbid conditions, recent surgery or hospitalization, active chemotherapy or hormonal surgery or hospitalization, active chemotherapy or hormonal g y p , pyg y p , pytherapytherapy

►► All hospitalized CA patients should be considered for All hospitalized CA patients should be considered for prophylaxisprophylaxis

►► Patients with cancer undergoing surgery should be Patients with cancer undergoing surgery should be considered for prophylaxisconsidered for prophylaxis

►► LMWH is the preferred drugLMWH is the preferred drugLyman GH, et al. J Clin Oncol. 2007;25:5490‐5505.

49

Updated ASCO Guidelines Updated ASCO Guidelines Hospitalized Patients with CancerHospitalized Patients with Cancer

Role of VTE Prophylaxis Evidence

Patients with cancer should be considered candidates for VTE prophylaxis with anticoagulants (UFH, LMWH, or fondaparinux) in the absence of bleeding or other contraindications to anticoagulation

Multiple RCTs of hospitalized medical patients with subgroups of patients with cancer. The 8th ACCP guidelines strongly recommend (1A) prophylaxis with either low-dose heparin or LMWH for bedridden patients with activecontraindications to anticoagulation bedridden patients with active cancer.

VOLUME 25 NUMBER 34 DECEMBER 1 2007

Prophylaxis in Cancer PatientsProphylaxis in Cancer Patients•• Cancer patients undergoing surgical procedures: routine Cancer patients undergoing surgical procedures: routine

thromboprophylaxis that is appropriate for the type of surgerythromboprophylaxis that is appropriate for the type of surgery

2008 ACCP Prevention of Venous Thromboembolism Practice Guidelines

thromboprophylaxis that is appropriate for the type of surgery thromboprophylaxis that is appropriate for the type of surgery (Grade 1A)(Grade 1A)

•• Cancer patients who are bedridden with an acute medical illness: Cancer patients who are bedridden with an acute medical illness: routine thromboprophylaxis as for other highroutine thromboprophylaxis as for other high--risk medical patients risk medical patients (Grade 1A)(Grade 1A)

•• Cancer patients receiving chemotherapy or hormonal therapy: Cancer patients receiving chemotherapy or hormonal therapy: recommend against the routine use of thromboprophylaxis for therecommend against the routine use of thromboprophylaxis for therecommend against the routine use of thromboprophylaxis for the recommend against the routine use of thromboprophylaxis for the primary prevention of VTE (Grade 1C)primary prevention of VTE (Grade 1C)

•• Cancer patients overall: recommend against the routine use Cancer patients overall: recommend against the routine use of primary thromboprophylaxis to try to improve survival of primary thromboprophylaxis to try to improve survival (Grade 1B) (Grade 1B)

Geerts WH, et al. Chest. 2008;133(6 suppl):381S‐453S.

50

Therapeutic Anticoagulation Treatment for Therapeutic Anticoagulation Treatment for VenousThromboembolismVenousThromboembolism--NCCN Update 2009NCCN Update 2009

►► The NCCN panel recommends VTE The NCCN panel recommends VTE thromboprophylaxisthromboprophylaxis for for all hospitalized patients with cancer who do not have all hospitalized patients with cancer who do not have cont aindications to s ch the apcont aindications to s ch the apcontraindications to such therapycontraindications to such therapy

►► Panel also emphasized that an increased level of clinical Panel also emphasized that an increased level of clinical suspicion of VTE should be maintained for cancer patients. suspicion of VTE should be maintained for cancer patients.

►► Following hospital discharge, it is recommended that Following hospital discharge, it is recommended that patients at highpatients at high--risk of VTE (e g cancer surgery patients)risk of VTE (e g cancer surgery patients)patients at highpatients at high risk of VTE (e.g. cancer surgery patients) risk of VTE (e.g. cancer surgery patients) continue to receive VTE prophylaxis for up to 4 weeks postcontinue to receive VTE prophylaxis for up to 4 weeks post--operation.operation.

http://www.nccn.org/professionals/physician_gls/f_guidelines.asp

CaveatsCaveats

►► No randomized controlled trials (RCTs) designed No randomized controlled trials (RCTs) designed ad hoc for hospitalized medical cancer patients ad hoc for hospitalized medical cancer patients are availableare availableare availableare available

►► Recommendations are based on RCTs of acutely Recommendations are based on RCTs of acutely ill medical patients, involving a small proportion ill medical patients, involving a small proportion of patients with cancerof patients with cancer

►► No bleeding data are reported specifically in the No bleeding data are reported specifically in the subgroup of patients with cancersubgroup of patients with cancersubgroup of patients with cancer subgroup of patients with cancer

51

Anticoagulant Prophylaxis to Prevent ScreenAnticoagulant Prophylaxis to Prevent Screen--Detected Detected VTE VTE HighHigh Risk Hospitalized Medical PatientsRisk Hospitalized Medical Patients

►► 3 large, randomized, 3 large, randomized, placeboplacebo--controlled,controlled,doubledouble blind trials in medical patients at highblind trials in medical patients at highdoubledouble--blind trials in medical patients at high blind trials in medical patients at high risk including cancerrisk including cancer●● MEDENOX (enoxaparin)MEDENOX (enoxaparin)11 ~ 15%~ 15%●● PREVENT (dalteparin)PREVENT (dalteparin)22 ~5%~5%●● ARTEMIS (fondaparinux)ARTEMIS (fondaparinux)33 ~15%~15%

S i f t ti DVT ithS i f t ti DVT ith►► Screening for asymptomatic DVT with Screening for asymptomatic DVT with venography venography or ultrasoundor ultrasound

1. Samama MM, et al. N Engl J Med. 1999;341:793-800.

2. Leizorovicz A, et al. Circulation. 2004;110:874-879.

3. Cohen AT, et al. BMJ. 2006;332:325-329.

Study RRR Thromboprophylaxis Patients with VTE (%)RRR

Anticoagulant Prophylaxis to Prevent Screen-Detected VTE High Risk Hospitalized Medical Patients

MEDENOX1 14.95.5

5.0

2.8

P < 0.001

P = 0.0015

63%

45%Placebo

Placebo

Enoxaparin 40 mg

Dalteparin 5,000 units

PREVENT2

10.5

5.6

1Samama MM, et al. N Engl J Med. 1999;341:793-800.2 Leizorovicz A, et al. Circulation. 2004;110:874-9.3Cohen AT, et al. BMJ 2006; 332: 325-329.

47%Placebo

units

Fondaparinux 2.5 mg

ARTEMIS3

52

1 6

1.8 1.7%g

(%)

Anticoagulant Prophylaxis to Prevent Screen-Detected VTE High Risk Hospitalized Medical Patients: Major Bleeding

0.4

0.6

0.8

1.0

1.2

1.4

1.6

LMWH

0.49%

0 16% 0.2%

1.1%

ce o

f M

ajor

Ble

edin

g

0.0

0.2

Medenox Prevent Artemis

0.16% 0.2%

Samama MM, et al. N Engl J Med. 1999;341:793-800.Leizorovicz A, et al. Circulation. 2004;110:874-9.Cohen AT, et al. BMJ 2006; 332: 325-329..

Inci

den

Study

EXCLAIM: EXCLAIM: ExtendedExtended--duration Enoxaparin duration Enoxaparin Prophylaxis in HighProphylaxis in High--risk Medical Patients risk Medical Patients

(Most benefit seen in Level 1 Disability Patients with bedrest or sedentary without BRP‐some with CA)

End points Outcome, extended prophylaxis, n=2052 (%)

Outcome, placebo, n=2062(%)

RR reduction (%)

p

VTE events 2.5 4.0 38% 0.001Major Bleeding 0.8% 0.3%

(Hull RD et al. Ann Intern Med 2010; 153:8)

%Symptomatic 0.3 1.1 73% 0.004

No Sxs 2.5 3.7 34% 0.032

53

PRODIGE: a randomized placeboPRODIGE: a randomized placebo--controlled trial of controlled trial of dalteparindalteparin for primary VTE prophylaxis in newly for primary VTE prophylaxis in newly

diagnosed malignant gliomadiagnosed malignant glioma

►► Reduced VTE for Reduced VTE for dalteparin 5,000 antidalteparin 5,000 anti--Xa Xa units qd for 6 mos: 11%units qd for 6 mos: 11%

Perry JR et al. JTH 2010;8;1959

units qd for 6 mos: 11% units qd for 6 mos: 11% vs 17% for placebo vs 17% for placebo

►► Increased ICH: 5.1% vs Increased ICH: 5.1% vs 1.2% for placebo 1.2% for placebo

►► Both NS significantBoth NS significant

2009 2009 NCCN Guidelines: NCCN Guidelines: DVT ProphylaxisDVT Prophylaxis

Pharmacologic ProphylaxisUFH

AdultCancer

InpatientContraindication to Anticoagulation Treatment

LMWHPentasaccharide

Mechanical ProphylaxisSequential Compression DevicesSequential Compression DevicesCompression Stockings

NCCN, National Comprehensive Cancer Network.NCCN. Venous Thromboembolic Disease: Version 1.2006. Available at: http://www.nccn.org/professionals/ physician_gls/PDF/vte.pdf.

54

Mechanical thromboprophylaxis in Mechanical thromboprophylaxis in critically ill patients: a systematic review critically ill patients: a systematic review

and metaand meta--analysisanalysisRESULTSRESULTS: 21 relevant studies (5 randomized controlled trials, 13 : 21 relevant studies (5 randomized controlled trials, 13 observational studies, and 3 surveys) were found. A total of 811 patients observational studies, and 3 surveys) were found. A total of 811 patients were randomized in the 5 randomized controlled trials; 3421 patients were randomized in the 5 randomized controlled trials; 3421 patients

ti i t d i th b ti l t diti i t d i th b ti l t diparticipated in the observational studies. participated in the observational studies. Trauma patients only were enrolled in 4 randomized controlled trials and Trauma patients only were enrolled in 4 randomized controlled trials and 4 observational studies. Meta4 observational studies. Meta--analysis of 2 randomized controlled trials analysis of 2 randomized controlled trials with similar populations and outcomes revealed that use of compression with similar populations and outcomes revealed that use of compression and pneumatic devices did not reduce the incidence of venous and pneumatic devices did not reduce the incidence of venous thromboembolism. The pooled risk ratio was 2.37 (CI,95% 0.57 thromboembolism. The pooled risk ratio was 2.37 (CI,95% 0.57 -- 9.90).9.90).A range of methodological issues, including bias and confounding A range of methodological issues, including bias and confounding variables, make meaningful interpretation of the observational studiesvariables, make meaningful interpretation of the observational studiesvariables, make meaningful interpretation of the observational studies variables, make meaningful interpretation of the observational studies difficult. difficult. CONCLUSIONSCONCLUSIONS: : The role of mechanical approaches to The role of mechanical approaches to thromboprophylaxis for intensive care patients remains uncertain.thromboprophylaxis for intensive care patients remains uncertain.

Limbus A et al. Am J Crit Care, 2006;15:402-10

The beneficial role for mechanical thromboprophylaxis in cancer pts is empiric and derived from benefits seen in surgical studies; No controlled studies in cancer patients

VTE Prophylaxis Is Underused VTE Prophylaxis Is Underused in Patients With Cancerin Patients With Cancer

90100

Cancer:FRONTLINE Survey1—3891 Clinician Respondentsis

, %

Major Surgery2

2030

40506070

8090

p

e of App

ropriate Proph

ylaxi

Major Abdominothoracic Surgery (Elderly)3

Medical Inpatients4

Confirmed DVT (Inpatients)5

Cancer:Surgical

Cancer: Medical

52

01020

FRONTLINE Surgical

FRONTLINE: Medical

Stratton Bratzler Rahim DVT FREE

1. Kakkar AK et al. Oncologist. 2003;8:381‐388.2. Stratton MA et al. Arch Intern Med. 2000;160:334‐340. 3. Bratzler DW et al. Arch Intern Med. 1998;158:1909‐1912.

Rate

4. Rahim SA et al. Thromb Res. 2003;111:215‐219.5. Goldhaber SZ et al. Am J Cardiol. 2004;93:259‐262.

5

55

Despite Evidence, Medical Patients Despite Evidence, Medical Patients at Risk Remain Unprotectedat Risk Remain Unprotected

ENDORSE1 IMPROVE2

Medical Surgical

No. of patients

37,356 30,827

At risk for VTE

42% 64%

United States

Other Countries

No. of patients

3,410 11,746

VTE prophylaxis

1852 (54%) 5788 (49%)

LMWH 4 6 (14%) 46 (40%)Receiving ACCP Tx

40% 59%

1. Cohen AT, et al. Presented at: 2007 Congress of the International Society on Thrombosis and Haemostasis; July 6‐12, 2007; Geneva, Switzerland.

2. Tapson VF, et al. Chest. 2007;132(3):936‐945.

LMWH 476 (14%) 4657 (40%)

UFH 717 (21%) 1014 (9%)

Electronic Alerts to Prevent VTE in Hospitalized Patients

98

100

T

92

94

Freedo

m From DVT

or PE (%

) 96

90

Intervention group

Control group

P<.001

110P<.001 by the log‐rank test for the comparison of the outcome between groups at 90 days.

Reprinted with permission from Kucher N, et al. N Engl J Med. 2005;352:969‐977.

0300 600 900

DaysNo. at RiskIntervention group 1255 977 900 853Control group 1251 976 893 839

56

Ambulatory Patients with Cancer Without VTE Ambulatory Patients with Cancer Without VTE Receiving Systemic Chemotherapy: Updated Receiving Systemic Chemotherapy: Updated

ASCO GuidelinesASCO Guidelines

Role of VTE Prophylaxis EvidenceRoutine prophylaxis in ambulatory

ti t i i h th i tRoutine prophylaxis with an antithrombotic agents is not recommended except as noted below

patients receiving chemotherapy is not recommended due to conflicting trials, potential bleeding, the need for laboratory monitoring and dose adjustment, and the relatively low incidence of VTE.

LMWH or adjusted dose warfarin (INR This recommendation is based on nonrandomized trial data and

~ 1.5) is recommended in myeloma patients on thalidomide or lenalidomide plus chemotherapy or dexamethasone

extrapolation from studies of postoperative prophylaxis in orthopedic surgery and a trial of adjusted-dose warfarin in breast cancer

VOLUME 25 NUMBER 34 DECEMBER 1 2007

Prospective Study of Adult Cancer PatientsProspective Study of Adult Cancer PatientsReceiving Systemic ChemotherapyReceiving Systemic Chemotherapy

• Prospective observational study conducted at 117 randomly selectedconducted at 117 randomly selected US practice sites.

• Data obtained on 4,458 consecutive adult patients initiating a new chemotherapy regimen between March 2003 and February 2006.

Prop

ortio

n with

VTE

.04

.03

.02

.01

Kuderer NM et al; J Clin Oncol 2008 (ASCO 2008).

• There were no exclusions for age, prior history or comorbid-ities with nearly 40% of patients age 65 and older.

Time (Days)

1514013012011010090807060504030201000.00

57

Reported Cause of Early Mortality Reported Cause of Early Mortality Cancer Patients Starting New ChemotherapyCancer Patients Starting New Chemotherapy

Cause of Death

No VTE N=4,365

VTE N=93

All N=4,4581.00

.99[HR=5.48, 95%CI: 2.21-13.61; P<.0001]

PD 2.1 2.2 2.1

Infection 0.3 0 0.3

PE 0 5.4 0.1

Pulmonary 0.2 0 0.2

Bleeding 0.1 0 0.1

Cum

lativ

e Su

rviv

al

.98

.97

.96

.95

.94

.93

.92

No VTE

Kuderer NM et al; J Clin Oncol 2008 (ASCO 2008)

Othervascular

0.2 0 0.2

Unknown 0.3 0 0.3

All 3.2 7.6 3.3Time (Days)

151401301201101009080706050403020100

.91

.90VTE

RCTs of Thromboprophylaxis in Ambulatory Cancer PatientsRCTs of Thromboprophylaxis in Ambulatory Cancer PatientsWarfarinWarfarin

DoubleDouble--blind, placeboblind, placebo--controlled RCT demonstrated the controlled RCT demonstrated the efficacy of lowefficacy of low--intensity intensity warfarinwarfarin (INR 1.3(INR 1.3--1.9) in patients 1.9) in patients yy yy (( ) p) preceiving chemotherapy for metastatic breast cancerreceiving chemotherapy for metastatic breast cancer

311 women with metastatic breast cancer on 311 women with metastatic breast cancer on 1st1st-- or 2ndor 2nd--line chemotherapyline chemotherapy

Randomized to 1 mg Randomized to 1 mg warfarinwarfarin for 6 weeks, then for 6 weeks, then warfarinwarfarintitrated to INR 1.3titrated to INR 1.3--1.9 or placebo1.9 or placebo

1 VTE in 1 VTE in warfarinwarfarin group group vsvs 7 in placebo arm 7 in placebo arm 85% risk reduction, 85% risk reduction, P P = .03, with no increased bleeding= .03, with no increased bleeding

Levine M, et al. Lancet. 1994;343:886-889.

INR=international normalized ratio

58

Trial N Treatment Chemo Duration VTE MajorBleeding

FAMOUSSolid tumors(Stage III/IV)

385 DalteparinPlacebo

64% 12 months 2.4%3.3%

0.5%0

Low Molecular Weight Heparin in RCTs of Low Molecular Weight Heparin in RCTs of Thromboprophylaxis in Ambulatory Cancer PatientsThromboprophylaxis in Ambulatory Cancer Patients

TOPIC-IBreast(Stage IV)

353 CertoparinPlacebo

100% 6 months 4%4%

1.7%0

TOPIC-2NSCLC(Stage IV)

547 CertoparinPlacebo

100% 6 months 4.5%†

8.3%3.7%2.2%

PRODIGEGlioma

186 DalteparinPlacebo

- 6-12 months 11%17%

5.1%1.2%

SIDERASSolid Tumors(Stage IV)

141 DalteparinPlacebo/Control

54% Indefinitely 5.9%7.1%

2.9%7.1%

PROTECHTSolid Tumors(Stage III/IV)

1166 Nadroparin 2:1 Placebo

100% < 4 monthswith chemo

1.4%2.9%

0.7%0

1. Kakkar AK, et al. J Clin Oncol. 2004;22:1944-1948. 2. Haas SK, et al. J Thromb Haemost. 2005(suppl 1):abstract OR059. 3. Perry JR et al. Proc ASCO 2007. 2011 4. Sideras K et al. Mayo Clin Proc 2006; 81:758-767. 5. Agnelli G et al. Am Soc Hemat Sunday December 7, 2008

The PROTECHT StudyThe PROTECHT StudyRCT of RCT of ThromboprophylaxisThromboprophylaxis in Cancer Patients Receiving in Cancer Patients Receiving

ChemotherapyChemotherapyDESIGNDESIGN

PlaceboPlacebo--controlled, double blind, multicenter RCTcontrolled, double blind, multicenter RCTNadroparin 3,800 anti Xa IU daily vs placebo: 2:1Nadroparin 3,800 anti Xa IU daily vs placebo: 2:1Nadroparin 3,800 anti Xa IU daily vs placebo: 2:1Nadroparin 3,800 anti Xa IU daily vs placebo: 2:11150 patients receiving chemotherapy for locally 1150 patients receiving chemotherapy for locally advanced or metastatic cancer. advanced or metastatic cancer. Start with new CTX; continue for maximum of 4 mosStart with new CTX; continue for maximum of 4 mosMean treatment duration: 90 daysMean treatment duration: 90 daysPrimary outcome: clinically detected thrombotic events, Primary outcome: clinically detected thrombotic events, i e composite of venous and arterial TE*i e composite of venous and arterial TE*i.e., composite of venous and arterial TEi.e., composite of venous and arterial TEMain safety outcome: Major bleedingMain safety outcome: Major bleeding

* deep vein thrombosis of the lower and upper limbs, visceral and cerebral venous thrombosis, pulmonary embolism, acute myocardial infarction, ischemic stroke, acute peripheral arterial thromboembolism, unexplained death of possible thromboembolic origin

Agnelli G et al: Lancet 2009;10, 930

59

The PROTECHT StudyThe PROTECHT StudyRCT of Thromboprophylaxis in Cancer Patients Receiving RCT of Thromboprophylaxis in Cancer Patients Receiving

ChemotherapyChemotherapyRESULTSRESULTS

►► Primary Efficacy Outcome: Any TE Event*Primary Efficacy Outcome: Any TE Event*●● Nadroparin: 16 of 769 (2.1%) Nadroparin: 16 of 769 (2.1%)

1 f 381 (3 9%)1 f 381 (3 9%)●● Placebo: 15 of 381 (3.9%) Placebo: 15 of 381 (3.9%) ●● Relative risk reduction: 47.2%, (interimRelative risk reduction: 47.2%, (interim--adjusted p=0.033)adjusted p=0.033)●● Absolute risk decrease: 1.8%; NNT = 53.8Absolute risk decrease: 1.8%; NNT = 53.8

►► Venous thromboembolism (VTE): Venous thromboembolism (VTE): ●● Nadroparin: 11 of 769 (1.4%)Nadroparin: 11 of 769 (1.4%)●● Placebo: 11 of 381 (2.9%) NSPlacebo: 11 of 381 (2.9%) NS

►► Major Bleeding:Major Bleeding:►► Major Bleeding: Major Bleeding: ●● Nadroparin: 5 (0.7%)Nadroparin: 5 (0.7%)●● Placebo: 0 (p= 0.177)Placebo: 0 (p= 0.177)●● Absolute risk increase: 0.7%; NNH = 153.8 Absolute risk increase: 0.7%; NNH = 153.8

Agnelli G et al: Lancet 2009;10:930

LMWH “halves” VTE in ambulatory patients with metastatic LMWH “halves” VTE in ambulatory patients with metastatic or locally advanced cancer who are receiving or locally advanced cancer who are receiving

chemotherapychemotherapy

PROTECHTPROTECHTAll cause thromboAll cause thrombo--embolic events: 2%embolic events: 2%Agnelli G et al. www.thelancet.com/Oncology Oct 2009 embolic events: 2% embolic events: 2% LMWH vs 3.9% in LMWH vs 3.9% in placeboplaceboMajor bleeding: Major bleeding: 0.7% LMWH vs 0.7% LMWH vs none in placebo none in placebo ((PP=0.18) =0.18) By the end of study By the end of study treatment, 33 treatment, 33 LMWH deaths vs 16LMWH deaths vs 161% DVT and 0 4% PE

2.1% DVT and 0.8% PE with placebo (N=381

pts)p=0.02

NNT=53.8

LMWH deaths vs 16 LMWH deaths vs 16 in placebo group; in placebo group; 48 of these deaths 48 of these deaths due to CA due to CA progression progression Benefits most Benefits most apparent in apparent in those with those with lung or GI CA lung or GI CA (not(not

1% DVT and 0.4% PE with LMWH (N=769 pts)

60

VTE Incidence In Various TumorsVTE Incidence In Various Tumors

Oncology SettingOncology Setting VTE VTE IncidenceIncidence

Breast cancer (Stage I & II) w/o further treatmentBreast cancer (Stage I & II) w/o further treatment 0.2%0.2%( g )( g )Breast cancer (Stage I & II) w/ chemoBreast cancer (Stage I & II) w/ chemo 2%2%Breast cancer (Stage IV) w/ chemoBreast cancer (Stage IV) w/ chemo 8%8%NonNon--Hodgkin’s lymphomas w/ chemoHodgkin’s lymphomas w/ chemo 3%3%Hodgkin’s disease w/ chemoHodgkin’s disease w/ chemo 6%6%Advanced cancer (1Advanced cancer (1--year survival=12%)year survival=12%) 9%9%HighHigh--grade gliomagrade glioma 26%26%

Otten, et al. Haemostasis 2000;30:72. Lee & Levine. Circulation 2003;107:I17Otten, et al. Haemostasis 2000;30:72. Lee & Levine. Circulation 2003;107:I17

Multiple myeloma (thalidomide + chemo)Multiple myeloma (thalidomide + chemo) 28%28%Renal cell carcinoma Renal cell carcinoma 43%43%Solid tumors (antiSolid tumors (anti--VEGF + chemo)VEGF + chemo) 47%47%Wilms tumor (cavoatrial extension) Wilms tumor (cavoatrial extension) 4%4%

Arterial Thrombotic Complications of MyelomaArterial Thrombotic Complications of Myeloma

VAD (n 6, 5.9%)TAD (n 2, 4.5%)PAD (n 3 6 4%)

N=195 ATE=11

5.6%

PAD (n 3, 6.4%)

4 developed thrombosis while on VKAs;

Libourel et al. Blood 2010;116:2

while on VKAs;2 on LMWH

61

LMWH Warfarin ASALMWH Warfarin ASA

939 14 18

15‐24 31 (LDW)

Palumbo A et al. Leukemia 2008;22:414

3‐14

These VTE prophylaxis regimens have not been assessed in any prospectiveprospective, randomized trial and are recommended based on anecdotal experience

Palumbo A et al. Leukemia 2008;22:414eukemia 008; :4 4

62

VTE Risk with Bevacizumab in Colorectal Cancer VTE Risk with Bevacizumab in Colorectal Cancer Approaches Risk of Antiangiogenesis in MyelomaApproaches Risk of Antiangiogenesis in Myeloma

Naluri SR et al. JAMA. 2008;300:2277

Tamoxifen and ChemotherapyTamoxifen and Chemotherapy

►► 705 postmenopeusal women with breast cancer705 postmenopeusal women with breast cancer

►► CMF regimenCMF regimen

►► Total thromboembolic eventsTotal thromboembolic events

68

10121416

Rate of thrombosis (%)p=0.0001

9.6%

►► Total thromboembolic eventsTotal thromboembolic events

►► 39 of 54 events occurred during chemotherapy39 of 54 events occurred during chemotherapy

Pritchard , J Clin Onc, 1996

024

Tamoxifen Tamoxifen + CT(n=352) (n=353)

1.4%

63

Treatment of Patients with Established Treatment of Patients with Established VTE to Prevent Recurrence VTE to Prevent Recurrence (continued)(continued)

Role of VTE Prophylaxis EvidenceIn the absence of clinical trials, benefits and risks of continuing LMWHAnticoagulation for an indefinite period

should be considered for patients with active cancer (metastatic disease, continuing chemotherapy)

benefits and risks of continuing LMWH beyond 6 months is a clinical judgment in the individual patient. Caution is urged in elderly patients and those with intracranial malignancy.

Inferior vena cava filters are reserved f th ith t i di ti t Consensus recommendations due tofor those with contraindications to anticoagulation or PE despite adequate long-term LMWH.

Consensus recommendations due to lack of date in cancer-specific populations

Treatment of Patients with Established Treatment of Patients with Established VTE to Prevent RecurrenceVTE to Prevent Recurrence

Role of VTE Prophylaxis Evidence

LMWH for 3-6 months isLMWH is the preferred approach for the initial 5-10 days in cancer patient with established VTE.

LMWH for 3 6 months is more effective than vitamin K antagonists given for a similar duration for preventing recurrent VTE.

LMWH for at least 6 months is preferred for long-term anticoagulant therapy. Vitamin K antagonists with a targeted INR of 2-3 are g gacceptable when LMWH is not available. The CLOT study demonstrated a relative risk reduction of 49% with LMWH vs. a vitamin K antagonist. Dalteparin sodium approved by the FDA for extended treatment of symptomatic VTE to reduce the risk of recurrence of VTE in patients with cancer (FDA 2007)

64

Control Group

The CLOT TrialThe CLOT TrialStudy SchemaStudy Schema

Dalteparin 200 IU/kg OD

Vitamin K antagonist (INR 2.0 to 3.0) x 6 mo

Experimental Group

5 to 7 days

Dalteparin 200 IU/kg OD x 1 mo then ~150 IU/kg OD x 5 mo

1 month 6 months

Lee AY, et al. N Engl J Med. 2003;349:146-153.

25

%

risk reduction = 52%HR 0.48 (95% CI 0.30, 0.77)log-rank p = 0.002

CLOT Trial:Results: Symptomatic Recurrent VTE

5

10

15

20

bilit

y of

Rec

urre

nt V

TE,

dalteparin, 9%

VKA, 17%

g p

0

5

Days Post Randomization0 30 60 90 120 150 180 210

Prob

ab


65

Dalteparin

N=338

VKA

N=335

p-value

CLOT Trial:Results: Bleeding

Major bleed 19 (5.6%) 12 (3.6%) 0.27

associated with death 1 0

critical site* 4 3

transfusion of > 2 units of RBC

d i Hb > 20 /L14 9

or drop in Hb > 20 g/L

Any bleed 46 (13.6%) 62 (18.5%) 0.09

*intracranial, intraspinal, pericardial, retroperitoneal, intra-ocular, intra-articular


Overall, these meta‐analyses and clinical trials do not conclusively establish the need for or value of prophylaxis of

CVC‐related thrombosis in cancer patients

Chaukiyal P et al. Thromb Haemost 2008;99:38

66

Influence of Influence of ThrombophiliaThrombophilia on Thrombotic on Thrombotic Complications of CVADs in CancerComplications of CVADs in Cancer

In 10 studies involving more than 1250 cancer patients with CVADs vs CA controls:

CA + FVL OR=5.18 (95% confidence interval: 3.0‐8.8)

CA + G20210A OR=3.95 (95% confidence interval: 1.5‐10.6)

The attributable risk of catheter associated thrombosis conferred by:

FVL 13.5%

G20210A 3.6%

CAVD = central venous access devicesDentali F, et al. JTH. 2007;5(Suppl 2):P‐S‐564.

Patient Group Recommended Not RecommendedHospitalized patients with cancer

VTE prophylaxis with anticoagulants If bleeding or contraindication to

ti l ti

ASCO Recommendations for VTE Prophylaxis in Patients with Cancer

Summary

anticoagulationAmbulatory patients with cancer receiving chemotherapy

Myeloma patients receiving thalidomide or lenalidomide + chemotherapy/ dexamethasone. LMWH or adjusted dose warfarin.

Otherwise, no routine prophylaxis

Patients with cancer undergoing surgery

Prophylaxis with low-dose UFH or LMWH Prophylaxis with mechanical methods for patients with contraindications to pharmacologic methods

Consider mechanical methods when contraindications to anticoagulation.pharmacologic methods g

Patients with cancer with established VTE

Pharmacologic treatment for at least 6 months. Consider continued anticoagulation beyond 6 months in those with active cancer.

-

To improve survival - Not recommended

Lyman GH et al: J Clin Oncol 2007; 25:5490-5505

67

What the ASCO/NCCN Guidelines What the ASCO/NCCN Guidelines Do Not Do Not Tell UsTell Us

►► What is the role for emerging novel anticoagulant What is the role for emerging novel anticoagulant medications? No comparisons with LMWHmedications? No comparisons with LMWH

►► Is there equivalent safety and efficacy between mIs there equivalent safety and efficacy between m--q y yq y yenoxaparinenoxaparin and and LovenoxLovenox??

►► Is there a survival advantage to the use of LMWH in cancer Is there a survival advantage to the use of LMWH in cancer patients?patients?

►► Is there a role for adjunctive Is there a role for adjunctive statinsstatins with anticoagulation in with anticoagulation in cancer patients?cancer patients?

►► Is there a role for monitoring Is there a role for monitoring hypercoagulabilityhypercoagulability markers in markers in i ?i ?cancer patients? cancer patients?

►► How does palliative care influence the survival and VTE How does palliative care influence the survival and VTE incidence data in cancer patients?incidence data in cancer patients?

►► How should incidental VTE be How should incidental VTE be anticoagulatedanticoagulated??►► What is the role for retrievable IVC filters in CA patientsWhat is the role for retrievable IVC filters in CA patients



Risk Stratification Tools to Identify Patients For Primary and Secondary Prevention of VTE in Patients with Malignancy

Faculty: Alok A. Khorana, MD, FACP Time: 7:45 – 8:15 p.m. Notes:

68

Risk Stratification Tools to Identify Patients Risk Stratification Tools to Identify Patients for Primary and Secondary Prevention of for Primary and Secondary Prevention of

VTE i th S tti f M liVTE i th S tti f M li


VTE in the Setting of MalignancyVTE in the Setting of Malignancy

Screening and VTE Risk Assessment Across the Screening and VTE Risk Assessment Across the Complex Spectrum of Malignant DisordersComplex Spectrum of Malignant Disorders——What What

Works? What Doesn’t?Works? What Doesn’t?

Alok A. Khorana, MD, FACPAlok A. Khorana, MD, FACPViceVice--Chief, Division of Hematology/OncologyChief, Division of Hematology/OncologyAssociate Professor of Medicine and OncologyAssociate Professor of Medicine and Oncology

James P. Wilmot Cancer CenterJames P. Wilmot Cancer CenterUniversity of RochesterUniversity of RochesterRochester, New YorkRochester, New York


ConsultantConsultantffsanofisanofi‐‐aventisaventis, Eisai, Leo , Eisai, Leo PharmaPharma

Speaker’s BureauSpeaker’s Bureausanofisanofi‐‐aventisaventis, Leo , Leo PharmaPharma

Grant/Research SupportGrant/Research Supportsanofisanofi‐‐aventisaventis

69

Risk assessment for VTE in cancer Risk assessment for VTE in cancer patientspatients

VTE in Cancer PatientsVTE in Cancer Patients

Clinical risk factorsClinical risk factorsBiomarkersBiomarkers

Risk assessment toolsRisk assessment tools

I li ti f th b h l iI li ti f th b h l iImplications for thromboprophylaxis Implications for thromboprophylaxis studiesstudies

Secondary prophylaxisSecondary prophylaxis

Risk Factors for VTERisk Factors for VTE

PatientPatient--related factorsrelated factorsOlder ageOlder age

TreatmentTreatment--related factorsrelated factorsH i li iH i li iOlder age Older age

Race, genderRace, genderComorbiditiesComorbidities

Hospitalization Hospitalization Chemotherapy Chemotherapy AntiAnti--angiogenicsangiogenicsMajor surgeryMajor surgeryErythropoiesisErythropoiesis--stimulating stimulating agentsagents

CancerCancer--related factorsrelated factors

Site of cancer Site of cancer

Ad d tAd d tagents agents TransfusionsTransfusions

Advanced stageAdvanced stage

Initial period after Initial period after diagnosis diagnosis

RaoRao et al., in et al., in CancerCancer--Associated Thrombosis. Associated Thrombosis. (Khorana and Francis, (Khorana and Francis, EdsEds)) 20072007

70

Type of cancer Adjusted OR (95% CI)

VTE and Site of CancerVTE and Site of Cancer

(95% CI)

Hematologic 28 (4-199.7)

Lung 22.2 (3.6-136.1)

GI 20.3 (4.9-83)

Breast 4.9 (2.3-10.5)

Prostate 2.2 (0.9-5.4)

BlomBlom JW et al. JW et al. JAMA JAMA 20052005

VTE With VTE With BevacizumabBevacizumab

14 RR=1.29 (95% CI, 1.03RR=1.29 (95% CI, 1.03--1.63)1.63)

4

6

8

10

12

Rat

e of

VTE

(%

)R

ate

of V

TE (

%)

13%13%

9.9%

6.2%6.2%

4 2%

RR=1.38 (95% CI, 1.12RR=1.38 (95% CI, 1.12--1.70)1.70)

0

2

Bevacizumab(n=1,196)

Control (n=1,083)

Bevacizumab (n=3,795)

Control (n=3,167)

4.2%

AllAll--Grade VTEGrade VTE((6 studies)6 studies)

HighHigh--Grade VTEGrade VTE((13 studies)13 studies)

However, when corrected for exposure time, RR =1.1 (95% CI, 0.89-1.36)

71

Biomarkers for Biomarkers for CancerCancer-- Associated VTEAssociated VTE

►►Blood countsBlood countsPlatelet countPlatelet countLeukocyte countLeukocyte countLeukocyte countLeukocyte countHemoglobinHemoglobin

►►DD--dimerdimer

►►Soluble PSoluble P--selectinselectin

►►Tissue factorTissue factor

►►CC--reactive proteinreactive protein

►►Factor VIIIFactor VIII

Incidence of VTE By Quartiles Of Incidence of VTE By Quartiles Of PrePre-- Chemotherapy Platelet CountChemotherapy Platelet Count

6%

s(%

)

s(%

)

1%

2%

3%

4%

5%

ce O

f VT

E O

ver

2.5

Mon

ths

ce O

f VT

E O

ver

2.5

Mon

ths

•P =0.005

Khorana AA et al. Khorana AA et al. Cancer Cancer 20052005

0%

1%

<250 250-300 300-350 >350

PrePre--chemotherapy Platelet Counts (x1000)chemotherapy Platelet Counts (x1000)

Inci

denc

Inci

denc

72

Incidence of VTE by PreIncidence of VTE by Pre--Chemotherapy Chemotherapy Leukocyte CountLeukocyte Count

6%6%

(%)

(%)

1%1%

2%2%

3%3%

4%4%

5%5%

nce

Of

VTE

Ove

r 2.

4 nc

e O

f VT

E O

ver

2.4

Mon

ths

Mon

ths

•P =0.0008

Khorana AA et al. Khorana AA et al. Blood Blood 20082008

0%0%

1%1%

<4.5 (n=342)<4.5 (n=342) 4.54.5--11 (n=3202)11 (n=3202) >11 (n=513)>11 (n=513)

PrePre--chemotherapy WBC Counts (x1000/mmchemotherapy WBC Counts (x1000/mm33))

Inci

den

Inci

den

Incidence of VTE by Type of LeukocyteIncidence of VTE by Type of Leukocyte

Absolute Absolute NeutrophilNeutrophilCount Count

Absolute Monocyte Absolute Monocyte Count Count

4.1

6.5

3 04.05.06.07.08.09.0 P=0.0001

P<0.0001

ortio

n w

ith V

TEor

tion

with

VTE

1.8 2.00.01.02.03.0

ANC<= 7.7 ANC >7.7 AMC <=1.2 AMC >1.2

Connolly et al ISTH 2009 Abs 1573Connolly et al ISTH 2009 Abs 1573

Prop

oPr

opo

73

Effect Effect of Leukocyte and Platelet Counts of Leukocyte and Platelet Counts on VTE Riskon VTE Risk

In the Vienna CATS registry, platelet count In the Vienna CATS registry, platelet count >>443,000 was associated with VTE (HR3.5)443,000 was associated with VTE (HR3.5), ( ), ( )

SimanekSimanek et alet al, J , J ThrombThromb HemostHemost 20092009

In the REALIn the REAL--2 study of advanced GEJ/gastric 2 study of advanced GEJ/gastric cancers, cancers, leukocytosisleukocytosis was associated with VTE was associated with VTE during chemotherapy (HR 2 0)during chemotherapy (HR 2 0)during chemotherapy (HR 2.0)during chemotherapy (HR 2.0)

Starling et al, Starling et al, J J ClinClin OncolOncol 20092009

Mortality by PreMortality by Pre--chemotherapy chemotherapy Leukocyte CountLeukocyte Count

WBC>11x10WBC>11x1099/L/L

0.200.200.180.180 160 16

14.0% (8.9%14.0% (8.9%--21.6%) 21.6%)

4.4% (3.2%4.4% (3.2%--6.1%) 6.1%)

P <0 0001P <0 0001

WBC>11x10WBC>11x1099/L/L

WBCWBC<<11x1011x1099/L/L

Prop

ortio

n D

ied

Prop

ortio

n D

ied 0.160.16

0.140.140.120.120.100.100.080.080.060.060.040.040.020.020.000.00

P <0.0001P <0.0001

MVA for early mortality: HR 2.0, p = 0.001MVA for early mortality: HR 2.0, p = 0.001

KudererKuderer et al ASH 2008et al ASH 2008Connolly et al ISTH 2009, Connolly et al ISTH 2009, ThromThrom Res Res 20102010

Time (Days)Time (Days)

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 1500 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

74

Tissue Factor in Cancer: Tissue Factor in Cancer: Lack of Standardized AssaysLack of Standardized Assays

►► ImmunohistochemistryImmunohistochemistry of tumor specimensof tumor specimens

►►TF ELISATF ELISA

►►TF MP TF MP procoagulantprocoagulant activity assayactivity assay

►► ImpedanceImpedance--based flow based flow cytometrycytometry

Tissue Factor Expression and VTETissue Factor Expression and VTE

10

15

20

25

30

ate

of V

TE (

%)

ate

of V

TE (

%)

P = 0.04

Khorana AA, et al. Clin Cancer Res. 2007;13:2870-2875.

0

5

Low TF High TF

RR

75

Cumulative Incidence of VTE for Cancer Patients Cumulative Incidence of VTE for Cancer Patients According to TFAccording to TF––bearing bearing MicroparticlesMicroparticles

L R k P 0 002L R k P 0 002VTE

VTE

0.60.6

0 50 5 Log Rank P=0.002Log Rank P=0.002

mul

ativ

e In

cide

nce

of V

mul

ativ

e In

cide

nce

of V 0.50.5

0.40.4

0.30.3

0.20.2

0.10.1

ZwickerZwicker J I et al. J I et al. ClinClin Cancer Res 2009;15:6830Cancer Res 2009;15:6830--68406840

MonthsMonths

Cum

Cum

0 5 10 15 20 250 5 10 15 20 250.00.0

FRAGEM and TF Biomarker DataFRAGEM and TF Biomarker Data

250250 ControlControl DalteparinDalteparin

BoxplotBoxplot of the percentage change of tissue factor antigen in the sera of of the percentage change of tissue factor antigen in the sera of pancreatic cancer patients in both the control and pancreatic cancer patients in both the control and dalteparindalteparin groupsgroups

250250

200200

150150

100100

5050

ControlControl DalteparinDalteparin

MaraveyasMaraveyas, et al. Blood , et al. Blood CoagulCoagul FibrinolysisFibrinolysis 20102010

5050

00

--5050

76

TF and Survival In Pancreatic CancerTF and Survival In Pancreatic Cancer

Median Survival in pts with TF MPMedian Survival in pts with TF MP--PCA >2.5 and </=2.5pg/ml. PCA >2.5 and </=2.5pg/ml.

Median survival was Median survival was 98.5 days 98.5 days for TF >2.5 for TF >2.5 pg/pg/mLmL vs.vs.231 days 231 days for TF </= for TF </= 2.5 pg/2.5 pg/mLmLp=< 0.0001p=< 0.0001

Prop

ortio

n su

rviv

ing

Prop

ortio

n su

rviv

ing

10100.90.90.80.80.70.70.60.60.50.50.40.40.30.30.20.20.10.1

00

BharthuarBharthuar et al et al ASCO GI ASCO GI 20102010

N=117 patients with N=117 patients with pancreaticobiliarypancreaticobiliarycancerscancers

Days on studyDays on study

TF (pg/TF (pg/mLmL) <2.5 >=2.5) <2.5 >=2.5

0 100 200 300 400 500 600 700 800 900 10000 100 200 300 400 500 600 700 800 900 1000

DD-- dimerdimer, F1/2 and VTE in Cancer, F1/2 and VTE in Cancer

Elevated DElevated D--d + F1/2d + F1/2

ility

)ili

ty)

0.250.25

Elevated F1/2Elevated F1/2

Elevated DElevated D--dimerdimer

umul

ativ

e Ri

sk (

prob

abi

umul

ativ

e Ri

sk (

prob

abi

0.200.20

0.150.15

0.100.10

0 050 05

Ay, C. et al. J Clin Oncol; 27:4124-4129 2009

NonelevaNonelevated D-dimerand F1/2

Observation Time (Days)Observation Time (Days)

CuCu

0 100 200 300 400 500 600 7000 100 200 300 400 500 600 700

0.050.05

77







VTE in Cancer OutpatientsVTE in Cancer Outpatients

►► The overwhelming majority of cancer patients The overwhelming majority of cancer patients are treated in the outpatient/ambulatory settingare treated in the outpatient/ambulatory settingp / y gp / y g

►► Which patients are most at risk?Which patients are most at risk?

►► Which patients will benefit most from Which patients will benefit most from prophylaxis?prophylaxis?

How do you define “high” risk?► Level of risk for which prophylaxis is considered

acceptable by both patients and oncologists

78

Risk ModelRisk Model

Patient Characteristic ScoreSite of CancerSite of CancerVery high risk (stomach, pancreas)High risk (lung, lymphoma, gynecologic, GU excluding prostate)

21

Platelet count > 350,000/mm3 1

Hb < 10g/dL or use of ESA 1Hb < 10g/dL or use of ESA 1

Leukocyte count > 11,000/mm3 1

BMI > 35 kg/m2 1

Khorana AA et al. Khorana AA et al. Blood Blood 20082008

Risk Model Validation

7%

8%

(%)

7.1%Development cohort 6.7%

2%

3%

4%

5%

6%

VTE

over

2.5

mos

1.8%

Development cohort

2.0%

Validation cohort

Risk Low (0) Intermediate(1-2) High(>3)

0%

1%

Rate

of

n=734 n=1627 n=340

0.8%0.3%

n=374 n=842 n=149

Khorana AA et al. Blood 2008

79

►► Full data available in 839 patientsFull data available in 839 patientsMedian observation time/followMedian observation time/follow up: 643 daysup: 643 days

Vienna CATS ValidationVienna CATS Validation

►► Median observation time/followMedian observation time/follow--up: 643 days up: 643 days

Score ≥317.7%

Number ofPatients Events

n n (%)

Score ≥3 96 16 (17%)

Score 0

Score 1

Score 2

6 months

1.5%

3.8%

9.4%Score 2 231 25 (11%)

Score 1 233 14 (6%)

Score 0 279 7 (3%)

Ay et al ISTH 2009 Abs Ay et al ISTH 2009 Abs

Expanded Risk Score withExpanded Risk Score withDD-- DimerDimer and and sPsP-- selectinselectin

Score Score ≥≥55

Score 4Score 4

Score 3Score 3

30.3%Number ofNumber of

Patients Patients EventsEventsn n n (%)n (%)

Score Score ≥≥55 3131 9 (29%)9 (29%)

Score 4Score 4 525210 10

(19%)(19%)1515Score 3Score 3

Score 2Score 2Score 1Score 1Score 0Score 01.0%

6 months

Score 3Score 3 13713715 15

(11%)(11%)

Score 2Score 2 226226 11 (5%) 11 (5%)

Score 1Score 1 192192 13 (7%)13 (7%)

Score 0Score 0 201201 4 (2%)4 (2%)

Ay et al ISTH 2009 Abs Ay et al ISTH 2009 Abs

80

Risk Score and ShortRisk Score and Short-- Term MortalityTerm Mortalityby VTE Risk Score Categoriesby VTE Risk Score Categories

1.00L

Ove

rall

Surv

ival

.95

.90

.85

Low

Intermediate

High

P < 0.0001

Time (Days)

121101009080706050403020100

.80

.75

KudererKuderer NM et al. NM et al. ASH ASH 20082008

International Myeloma Working Group International Myeloma Working Group Thromboprophylaxis RecommendationsThromboprophylaxis Recommendations

Individual risk factors: Individual risk factors: obesity (BMI ≥ 30), prior VTE, central venous catheterobesity (BMI ≥ 30), prior VTE, central venous catheterComorbid risk factors: Comorbid risk factors: cardiac disease, chronic renal disease, diabetes, acute cardiac disease, chronic renal disease, diabetes, acute infection, immobilizationinfection, immobilizationS i k f tS i k f t t l th it l th iSurgery risk factorsSurgery risk factors: trauma, general surgery or any anesthesia: trauma, general surgery or any anesthesiaMedications: Medications: erythropoietinerythropoietinMyelomaMyeloma--related risk factors: related risk factors: diagnosis, hyperviscositydiagnosis, hyperviscosityMyeloma therapy risk factors: Myeloma therapy risk factors: multiagent chemotherapy, doxorubicin, highmultiagent chemotherapy, doxorubicin, high--dose dose steroidssteroids••Patients with ≤ 1 VTE risk factor: Aspirin (81Patients with ≤ 1 VTE risk factor: Aspirin (81--325 mg daily)325 mg daily)••Patients with ≥ 2 VTE risk factors: LMWH (enoxaparin 40 mg/d) or fullPatients with ≥ 2 VTE risk factors: LMWH (enoxaparin 40 mg/d) or full--dose dose

f i l h h l i i i d f h lf i l h h l i i i d f h l

Palumbo A, Rajkumar SV, Dimopoulos MA, et al. Prevention of thalidomide- and lenalidomide associated thrombosis in myeloma. Leukemia. 2008 Feb;22(2): 414-23.

warfarin, although less existing supporting data for the latterwarfarin, although less existing supporting data for the latter••Patients receiving thalidomide/lenalidomide concurrently with highPatients receiving thalidomide/lenalidomide concurrently with high--dose dose dexamethasone or doxorubicin should receive LMWH thromboprophylaxis dexamethasone or doxorubicin should receive LMWH thromboprophylaxis ••Anticoagulant treatment can continue for 4 to 6 months or longer if additional risk Anticoagulant treatment can continue for 4 to 6 months or longer if additional risk factors are presentfactors are present

81







Rates of VTE in Recent Rates of VTE in Recent Prophylaxis StudiesProphylaxis Studies

p = 0.01931.0

30 0

35.0Control LMWH Aspirin Warfarin

p< 0.0114.5

5.0 5.0

12.0

8.0

p= NS

6.05 0

10.0

15.0

20.0

25.0

30.0

Rat

e of

VTE

(%

)

2.9 1.40.0

5.0

PROTECHT Myeloma CONKO‐004 FRAGEM

N=930 N=312 N=123N=1165

AgnelliAgnelli et al et al Lancet Lancet OncOnc 20092009Palumbo et al Palumbo et al ASH ASH 20092009RiessRiess et al Iet al ISTHSTH 20092009MaraveyasMaraveyas et al et al ESMOESMO 20092009

82

VTE in Lung Cancer: VTE in Lung Cancer: PROTECHT and TOPIC studiesPROTECHT and TOPIC studies

sVTE LMWH sVTEPlacebo

All VTE LMWH

All VTE PlaceboPlacebo LMWH Placebo

PROTECHT 3.5% 5% 4% 6.2%

TOPIC-2 3% 5.7% 4.5% 8.3%

All 3.2% 5.5% 4.3% 7.8%

Major Bleeding

Major Bleeding NNT=50 (sVTE)

Verso et al. JTH 2010 online

BleedingLMWH

Bleeding Placebo

PROTECHT 1% 0%

TOPIC-2 3.7% 2.2%

All 2.5% 1.7%

NNT=50 (sVTE)NNT=28 (allVTE)RRR=46%NNH=125

Ongoing Clinical TrialsOngoing Clinical Trials

Study (Agent) Criteria for inclusion* N Endpoints

PHACS(dalteparin x 12 wks) -Risk score >=3 404

Asymptomatic and symptomatic VTE

SAVE-ONCO (semuloparin up to 4 mos)

-Lung, bladder, GI, ovary-Metastatic or locally advanced

3200 DVT, PE, VTE-related death

-Lung colon pancreasMicroTEC(enoxaparin x 6 mos)

Lung, colon, pancreas -Metastatic or unresectable-Elevated TF MPs

227 VTE

* All studies enroll patients initiating a new chemotherapy regimen

83







Predictors of Recurrent VTE: Findings Predictors of Recurrent VTE: Findings from the RIETE Registryfrom the RIETE Registry

►► Recurrent PERecurrent PE●● Age < 65 (OR 3.0)Age < 65 (OR 3.0)

PE (OR 1 9)PE (OR 1 9)●● PE at entry (OR 1.9)PE at entry (OR 1.9)●● < 3 months from diagnosis of cancer (OR 2.0)< 3 months from diagnosis of cancer (OR 2.0)

►► Recurrent DVTRecurrent DVT●● Age < 65 (OR 1.6)Age < 65 (OR 1.6)●● < 3 months from diagnosis of cancer (OR 2.4)< 3 months from diagnosis of cancer (OR 2.4)gg

►► Patients with Patients with leukocytosisleukocytosis had increased risk of had increased risk of recurrent VTE and death (OR 2.7)recurrent VTE and death (OR 2.7)

Trujillo-Santos et al Thromb Haem 2008

84

CLOT Study:CLOT Study:Reduction in Recurrent VTEReduction in Recurrent VTE

2525

, %, %

Risk reduction = 52%Risk reduction = 52%pp--value = 0.0017value = 0.0017Recurrent VTERecurrent VTE

55

1010

1515

2020bi

lity

of R

ecur

rent

VTE

bilit

y of

Rec

urre

nt V

TE

DalteparinDalteparin

OACOAC

Lee et.al. Lee et.al. N N EnglEngl J Med, J Med, 2003;349:1462003;349:146

00

55

Days Post RandomizationDays Post Randomization

00 3030 6060 9090 120120 150150 180180 210210

Prob

aPr

oba

Treatment of Treatment of CancerCancer-- Associated VTEAssociated VTE

Study Design

Length of

Therapy N Recurrent VTE (%)

Major Bleeding

(%)

Death(%)

(Months) (%) (%)

CLOT Trial(Lee 2003)

DalteparinOAC

6 336336

917

64

3941

CANTHENOX(Meyer 2002)

EnoxaparinOAC

3 6771

1121

716

1123

NSNS0.002

0.09

0.030.09

LITE(Hull ISTH 2003)

TinzaparinOAC

3 8087

611

68

2322

ONCENOX(Deitcher ISTH 2003)

Enox (Low)Enox (High)OAC

6 323634

3.43.16.7

NS0.03

NS

NS

NS

NR

85


►► Cancer patients are clearly at increased risk for Cancer patients are clearly at increased risk for VTE but risk varies widelyVTE but risk varies widelyVTE but risk varies widelyVTE but risk varies widely

►► Yet, 53% of cancer patients are unaware that Yet, 53% of cancer patients are unaware that they are at high risk for VTEthey are at high risk for VTE

Sousou et al, Ca Inv 2010

►► HighHigh--risk subgroups can be identified based onrisk subgroups can be identified based on►► HighHigh risk subgroups can be identified based on risk subgroups can be identified based on clinical risk factors and biomarkersclinical risk factors and biomarkers

►► A recently validated risk model can predict risk A recently validated risk model can predict risk of VTE (and mortality) using 5 simple clinical of VTE (and mortality) using 5 simple clinical and laboratory variablesand laboratory variables


►►LMWHLMWH--based prophylaxis is safe and effective in based prophylaxis is safe and effective in certain highcertain high--risk settingsrisk settings●● Hospitalized and surgical patientsHospitalized and surgical patients●● Highly selected cancer outpatients (myeloma, Highly selected cancer outpatients (myeloma,

?pancreas, ?? lung)?pancreas, ?? lung)

►►Ongoing studies are adopting novel approaches Ongoing studies are adopting novel approaches to selecting patients for prophylaxisto selecting patients for prophylaxisto selecting patients for prophylaxisto selecting patients for prophylaxis

►►Clinicians need to conduct baseline and ongoing Clinicians need to conduct baseline and ongoing risk assessment for VTE in cancer patients risk assessment for VTE in cancer patients receiving chemotherapyreceiving chemotherapy


in Cancer and Thrombosis Program Chairman’s Concluding Vision Statement: Current and Near Future

Perspectives of VTE Management in the Setting of Malignancy

Faculty: Samuel Z. Goldhaber, MD – Program Chairman Time: 8:15 – 8:30 p.m. Notes:

86

VISION STATEMENTVISION STATEMENT


VTE in the Setting of MalignancyVTE in the Setting of Malignancy

Samuel Z. Goldhaber, MDSamuel Z. Goldhaber, MDCardiovascular DivisionCardiovascular Division

Brigham and Women’s HospitalBrigham and Women’s HospitalProfessor of MedicineProfessor of Medicine

Harvard Medical SchoolHarvard Medical School

87


Research Support:Research Support:BMS;BMS; BoehringerBoehringer--IngelheimIngelheim; Eisai; Johnson; Eisai; JohnsonBMS; BMS; BoehringerBoehringer IngelheimIngelheim; Eisai; Johnson ; Eisai; Johnson & Johnson, & Johnson, SanofiSanofi--Aventis Aventis

Consultant:Consultant:BoehringerBoehringer--IngelheimIngelheim; BMS; Eisai; EKOS:; BMS; Eisai; EKOS:BoehringerBoehringer IngelheimIngelheim; BMS; Eisai; EKOS: ; BMS; Eisai; EKOS: MedscapeMedscape; Merck; Pfizer; ; Merck; Pfizer; SanofiSanofi--AventisAventis

Toward Eradication of InToward Eradication of In--


Hospital VTE: Hospital VTE: The Promise of The Promise of

ProphylaxisProphylaxis

“VITAE” Studies“VITAE” Studies

88

VTE VTE Prophylaxis in 19,958 Medical Prophylaxis in 19,958 Medical Patients/9 Studies (MetaPatients/9 Studies (Meta-- analysis)analysis)

►► 62% reduction in fatal PE62% reduction in fatal PE►► 62% reduction in fatal PE62% reduction in fatal PE

►► 57% reduction in fatal or nonfatal PE57% reduction in fatal or nonfatal PE

►► 53% reduction in DVT53% reduction in DVT

DentaliDentali F, et al. Ann Intern Med 2007; 146: 278F, et al. Ann Intern Med 2007; 146: 278--288288

VITAE IVITAE I

►► VITAE I uses a Federal database to model VITAE I uses a Federal database to model Hospitalized Medical Patients with VTE. Hospitalized Medical Patients with VTE. pp

►► 2 of every 100 hospitalized Medical Service 2 of every 100 hospitalized Medical Service patients suffer VTE. patients suffer VTE.

With universal inWith universal in--hospital prophylaxis, the hospital prophylaxis, the VTE VTE rate would be cut by 58%.rate would be cut by 58%.

Thromb Haemost 2009; 102: 505-510

89

58% 58% Reduction in VTE with Universal Reduction in VTE with Universal Prophylaxis in Hospitalized Medical PatientsProphylaxis in Hospitalized Medical Patients

25000300003500040000

05000

10000150002000025000

Current prophylaxis rates 100% prophylaxis

Thromb Haemostas 2009; 102: 505-510

VITAE IIVITAE II

►► VITAE II models the 5VITAE II models the 5--year aftermath of initial year aftermath of initial VTE among these same Hospitalized MedicalVTE among these same Hospitalized MedicalVTE among these same Hospitalized Medical VTE among these same Hospitalized Medical Patients who were initially stricken. Patients who were initially stricken.

►► If universal prophylaxis had been utilized If universal prophylaxis had been utilized initially, the 5initially, the 5--year VTE complication rates ofyear VTE complication rates ofinitially, the 5initially, the 5 year VTE complication rates of year VTE complication rates of death, recurrence, PTS, and CTEPH would death, recurrence, PTS, and CTEPH would have been reduced by 60%. have been reduced by 60%.

ThrombThromb HaemostHaemost 2009; 102: 6882009; 102: 688--693693

90

VITAE IIVITAE II

Status Quo 100% VTE Prophylaxis

ThrombThromb HaemostHaemost 2009; 102: 6882009; 102: 688--693693


1.1. Electronic alerts can identify hospitalized Electronic alerts can identify hospitalized cancer patients at risk for VTE.cancer patients at risk for VTE.

2.2. Optimal prophylaxis for hospitalized Optimal prophylaxis for hospitalized cancer patients is LMWH.cancer patients is LMWH.

3.3. When VTE is diagnosed in cancer When VTE is diagnosed in cancer patients the only FDApatients the only FDA--approved LMWHapproved LMWHpatients, the only FDApatients, the only FDA--approved LMWH approved LMWH for Rx as for Rx as monotherapymonotherapy without warfarin without warfarin is is dalteparindalteparin. .

91

Conclusions (Conclusions (Continued)Continued)

4.4. ACCP guidelines state that “every hospital ACCP guidelines state that “every hospital should develop a formal strategy to should develop a formal strategy to p gyp gyprevent VTE.”prevent VTE.”

5.5. As cancer therapies improve, quality lifeAs cancer therapies improve, quality life--years will be extended.years will be extended.

66 DVT and PE will be mostly prevented inDVT and PE will be mostly prevented in6.6. DVT and PE will be mostly prevented in DVT and PE will be mostly prevented in cancer patients, and when necessary to cancer patients, and when necessary to treat, will be managed will LMWH treat, will be managed will LMWH monotherapymonotherapy..



Interactive Q &A Discussion Session Faculty: Faculty Panel Time: 8:30 – 8:45 p.m. Notes:

Full Paper

Treatment and secondary prevention of venousthromboembolism in cancer

R Coleman*,1 and P MacCallum2

1Academic Unit of Clinical Oncology, Cancer Research Centre, Weston Park Hospital, Whitham Road, Sheff ield S10 2SJ, UK; 2Barts and The LondonSchool of Medicine and Dentistry, Wolfson Institute of Preventive Medicine, Queen Mary University of London, Charterhouse Square, London EC1M 6BQ, UK

Patients with cancer who develop venous thromboembolism (VTE) are at elevated risk for recurrent thrombotic events, even duringanticoagulant therapy. The clinical picture is further complicated because these patients are also at increased risk of bleeding while onanticoagulants. In general, there are four key goals of treatment for VTE: preventing fatal pulmonary embolism (PE); reducing short-term morbidities associated with acute leg or lung thrombus; preventing recurrent VTE; and preventing the long-term sequelae ofVTE (e.g., post-thrombotic syndrome and chronic thromboembolic pulmonary hypertension). A fifth goal – minimising the risk forbleeding while on anticoagulation – is particularly warranted in patients with cancer. Traditionally, pharmacological treatment of VTEhas two phases, with the transition between phases marked by a switch from a rapid-acting, parenterally administered anticoagulant(such as unfractionated heparin (UFH), low-molecular-weight heparin (LMWH), or fondaparinux) to an oral vitamin K antagonist(e.g., warfarin). Recent clinical trials of established agents and the advent of new pharmacological options are changing this paradigm.Low-molecular-weight heparin continued for 6 months is more effective than warfarin in the secondary prevention of VTE in cancerpatients without increasing the risk of bleeding and is now the preferred treatment option. Given the impact of VTE on short-termand long-term outcomes in patients with cancer, a group of health-care providers based in the United Kingdom gathered in Londonin 2009 to discuss recent data on cancer-associated thrombosis and to evaluate how these recommendations can be integrated ortranslated into UK clinical practice. This article, which is the third of four articles covering key topics in cancer thrombosis, focuses ontreatment and secondary prevention of VTE in cancer patients.British Journal of Cancer (2010) 102, S17 – S23. doi:10.1038/sj.bjc.6605601 www.bjcancer.com& 2010 Cancer Research UK

Keywords: venous thromboembolism; heparin; warfarin

��

It has been clearly established that patients with cancer have anincreased risk of venous thromboembolism (VTE) compared withpatients without cancer. Venous thromboembolism, comprisingdeep vein thrombosis (DVT) and pulmonary embolism (PE), is animportant cause of morbidity and mortality in cancer patients andits treatment is particularly difficult because standard therapyis less effective in preventing recurrences and causes morebleeding than is the case in non-cancer patients with VTE. Theneed to perform surgery or invasive procedures in patients withcancer receiving anticoagulants, as well as the frequent presenceof thrombocytopenia due to chemotherapy or haematologicalmalignancy, widespread use of in-dwelling central lines, adminis-tration of multiple interacting drugs, and varying dietary intakeduring the course of cancer therapy, all add to the particularchallenges faced in VTE treatment.

Randomised clinical trials over the past decade have substan-tially altered the management of VTE in patients with cancer,leading to improved outcomes and quality of life. Unfractionatedheparin (UFH) and vitamin K antagonists (e.g., warfarin) havebeen the mainstay of management of VTE since the mid-twentiethcentury, followed by the widespread replacement of UFH bylow-molecular-weight heparins (LMWH) since the 1990s. Theappearance of new classes of oral anticoagulants that directlyinhibit specific clotting factor targets, such as thrombin and factor

Xa, may further transform the management of VTE in patientswith cancer in the years to come.

Venous thromboembolism is a major cause of death in patientswith cancer receiving chemotherapy (Khorana et al, 2007).Compared with patients without cancer, patients with cancer havea three-fold elevated risk of recurrent DVT or PE, followingan initial episode of VTE. For example, in a prospective study,Prandoni et al (2002) observed a 12-month cumulative incidenceof recurrent thromboembolism of 20.7% in 181 cancer patients,compared with 6.8% in patients without cancer. Both groupswere managed in the same standard manner with initial heparinfollowed by warfarin. Levitan et al (1999) showed that the cumu-lative probability of re-admission to the hospital with DVT amongpatients initially hospitalised for DVT and malignant disease wasapproximately three-fold higher than those initially hospitalisedwith DVT alone. A more recent retrospective study, conducted byElting et al (2004) using medical records from 529 consecutivecancer patients, found an overall rate of recurrence of 17%,ranging to as high as 32% in patients with inferior vena caval (IVC)filters. Similarly, in 2006, Blom et al (2006) found that the risk ofVTE within the first 6 months after a first thrombotic event was18.4/1000/0.5 year, with a 4.6-fold increased risk comparedwith cancer patients who did not have a thrombotic event in the6 months after cancer diagnosis. In this study, patients withleukaemia, brain cancer, or cancer of the bladder, ureter, or testeswere at highest risk for recurrence. In the RIETE registry of 15 520consecutive patients with VTE, cancer was present in 20% and was*Correspondence: Professor R Coleman; E-mail: [email protected]

British Journal of Cancer (2010) 102, S17 – S23

& 2010 Cancer Research UK All rights reserved 0007 – 0920/10 $32.00

www.bjcancer.com

http://dx.doi.org/10.1038/sj.bjc.6605601

http://www.bjcancer.com

mailto:[email protected]


associated with a two-fold increase in the risk of fatal PE within3 months (Laporte et al, 2008). Risk factors for the recurrenceon multivariate analysis were age o65 years or o3 months fromcancer diagnosis to VTE (Trujillo-Santos et al, 2008). Thosepresenting with PE were more likely to have a PE recurrence thanthose initially presenting with DVT.

The clinical picture is further complicated by the fact thatpatients with cancer are also at increased risk of bleedingduring anticoagulant therapy. In the study conducted byPrandoni et al (2002), referred to above, the 12-month cumulativeincidence of major bleeding was 12.4% in patients with cancer and4.9% in patients without cancer – a hazard ratio of 2.2. Notably,both recurrence and bleeding were directly related to cancerseverity, and could not be explained by over- or under-antic-oagulation (Prandoni et al, 2002). Similarly, in the study by Eltinget al (2004), the risk of major bleeding during anticoagulanttherapy was 12% (Elting et al, 2004). In the RIETE Registry, cancerwas independently associated with a 1.7-fold increased risk ofmajor bleeding (Ruiz-Gimenez et al, 2008). Risk factors forbleeding in cancer patients were immobility, metastases, recentbleeding, or creatinine clearance o30 ml min�1 (Trujillo-Santoset al, 2008).

A number of evidence-based guidelines on the management ofVTE in both cancer and non-cancer patients have recently beenpublished (Lyman et al, 2007; Kearon et al, 2008; Torbicki et al,2008; NCCN, 2009). Given the impact of VTE on short-term andlong-term outcomes in patients with cancer – and a potential gapbetween current guidelines and national clinical practice – a groupof health-care providers based in the United Kingdom gathered inLondon in 2009 to discuss recent data and guidelines on cancer-associated thrombosis and to evaluate how these recommenda-tions can be translated into best practices for the United Kingdom.This article, which is the third of four articles covering keytopics in cancer thrombosis, focuses on treatment and secondaryprevention of VTE in the cancer patient. The paper is structured asa brief review of key data and guidelines on acute treatment andsecondary prevention of VTE in patients with cancer, followedby an edited transcript of the discussion surrounding these data.

TREATMENT OF VTE IN CANCER PATIENTS

In general, the goals of VTE treatment can be summarised asfollows: (1) preventing fatal PE; (2) reducing short-term morbi-dities associated with acute leg or lung thrombus; (3) preventingrecurrent VTE; (4) preventing the long-term sequelae of VTE(e.g., post-thrombotic syndrome (PTS) and chronic thrombo-embolic pulmonary hypertension).

Anticoagulant therapy is the mainstay of treatment for VTE.Traditionally, pharmacological treatment of VTE has two phases,with the transition between phases marked by a switch from aparenterally administered anticoagulant with a rapid mechanismof action (e.g., UFH, low-molecular-weight heparin (LMWH), orfondaparinux (a synthetic pentasaccharide with the same antith-rombin-dependent inhibition of factor Xa as LMWH)) to an oralvitamin K antagonist (e.g., warfarin). As will be discussed, recentclinical trials of established agents and the advent of newpharmacological options may change this paradigm. In additionto anticoagulants, a number of additional options exist for acuteVTE, including systemic thrombolytic therapy, catheter-directedthrombolysis, mechanical thrombectomy, and use of vena cavalfilters.

Initial treatment of VTE

The use of UFH in the initial treatment of VTE is well established.The first and only trial that compared anticoagulant therapywith no therapy was published in 1960 (Barritt and Jordan, 1960,

see Kearon et al, 2008). The trial that enrolled patients withsymptomatic PE (with or without symptomatic DVT), who wererandomly allocated to 1.5 days of heparin and 14 days of vitamin Kantagonist therapy or no treatment, showed that anticoagulationsignificantly reduced recurrent PE and mortality. The need forinitial treatment with a rapid-acting anticoagulant such as heparinwas confirmed in a trial by Brandjes et al (1992). In this study,patients were assigned to either combination therapy (intravenousloading dose of 5000 U heparin, followed by an infusion of1250 U h�1 for a minimum of 7 days, in combination with vitaminK antagonist acenocoumarol) or to acenocoumarol alone (Brandjeset al, 1992). The study was ended early because of an excessof events in the acenocoumarol group (20%) compared with thecombined-therapy group (6.7%). Asymptomatic extension ofvenous thrombus was seen in 39.6% of patients in the aceno-coumarol group and in 8.2% of the combination group. Notably,the rate of major bleeding was comparable in the two groups(Brandjes et al, 1992).

Although UFH as an initial treatment for VTE is highly effective,it is associated with a number of important limitations, including ashort half-life, wide interpatient and intrapatient variability due topharmacokinetic shortcomings, the need for frequent monitoring, arisk for heparin-induced thrombocytopenia, and, with long-termtherapy, osteoporosis (a detailed discussion of the advantages andshortcomings of heparin can be found elsewhere in this supple-ment). Low-molecular-weight heparins have more predictablepharmacokinetics and greater bioavailability, which permitsbody-weight-adjusted dosing without the need for laboratorymonitoring for most patients. This makes therapy much simplerand allows for outpatient treatment of many patients, reducinghospitalization, and improving their quality of life.

A number of clinical studies and meta-analyses have comparedthe efficacy and safety of body-weight-adjusted LMWH, adminis-tered subcutaneously without monitoring, with monitored, dose-adjusted intravenous heparin (Quinlan et al, 2004; van Dongenet al, 2004) and found it to be more effective, with reduced majorbleeding during initial treatment and overall mortality at follow-up(van Dongen et al, 2004). Individual studies have not specificallyassessed the effect of LMWH as initial treatment in patients withcancer. However, a meta-analysis of 14 trials that included cancersubgroup data showed that LMWH was equivalent to UFH formortality (RR 0.89, 95% CI 0.61– 1.27) and for clinically suspectedDVT (RR 0.73, 95% CI 0.23–2.28) (Akl et al, 2008). In a post hocanalysis that included all studies that assessed DVT (irrespective ofdiagnostic strategy), LMWH was superior to UFH (RR 0.72, 95% CI0.55– 0.94). Rates of PE and minor or major bleeding were similarfor the two strategies.

Data on the subcutaneously administered factor Xa inhibitor,fondaparinux, in the initial treatment of VTE associated withcancer are more limited. Post hoc analyses of the cancer patientsubgroups of two randomised trials (Buller et al, 2003, 2004) inwhich fondaparinux was compared with UFH (for initial treatmentof PE) or LMWH (for initial treatment of DVT) suggest broadlysimilar efficacy and safety (van Doormaal et al, 2009). However,these data require confirmation.

In summary, LMWHs, or possibly fondaparinux, are the agentsof choice for the initial treatment of most episodes of VTEoccurring in patients with cancer. An exception is in the setting ofmassive PE characterised by shock or hypotension in which UFHremains the preferred mode of anticoagulation, as newer agentshave not been properly evaluated and an immediate anticoagulanteffect is required.

MANAGEMENT OF MASSIVE PE

Massive PE constitutes a medical emergency, and specialisedmedical advice should be sought. Patients presenting with PE

Treatment and secondary prevention of VTE in cancer

R Coleman and P MacCallum

S18

British Journal of Cancer (2010) 102(S1), S17 – S23 & 2010 Cancer Research UK

should be clinically risk stratified into high-risk and non-high-riskgroups. The former is characterised by systolic hypotension(BPo90 mm Hg) (Torbicki et al, 2008) and carries a short-termmortality of 415% (Torbicki et al, 2008). Pooled data fromrandomised trials in non-cancer patients suggest a significantreduction in mortality and PE recurrence with systemicallyadministered thrombolytic therapy in this subgroup (Wan et al,2004). Intravenous heparin should be started immediately(80 U kg�1 as a bolus, followed by infusion at the rate of18 U kg�1 h�1 and adjusted to keep the activated partial thrombo-plastin time (aPTT) ratio between 1.5 and 2.5 times control)(Torbicki et al, 2008). Oxygen should be administered if thepatient is hypoxic (Torbicki et al, 2008). A number of thrombolyticregimens are approved for PE. In practice, recombinant tissueplasminogen activator (rtPA) is commonly used at a dose of100 mg over 2 h, or at 0.6 mg kg�1 over 15 min (maximum dose50 mg) if the patient is rapidly deteriorating (Torbicki et al, 2008).In patients in whom thrombolytic therapy is contraindicated (e.g.,intracranial bleeding), alternative approaches, depending on localavailability and expertise, are surgical pulmonary embolectomyand percutaneous catheter embolectomy and fragmentation(Torbicki et al, 2008).

Secondary prevention of VTE

Heparins are used in the initial treatment of VTE to provideanticoagulant support during the period of time required toachieve an appropriate international normalised ratio (INR),generally 2.0–3.0, with oral warfarin. In patients without cancer,warfarin is subsequently continued for a period of up to 6 monthsor longer. The need for anticoagulation beyond the initial period ofheparinisation was demonstrated 30 years ago (Hull et al, 1979).Subsequently, a large, multicentre, randomised trial suggested thatlonger-duration warfarin (6 months) was superior to short-duration therapy (6 weeks). In this study, 902 patients wererandomly assigned to receive either 6 weeks or 6 months of oralanticoagulant therapy with a target INR of 2.0–2.85 (Schulmanet al, 1995). At 2 years, the rate of recurrence in the 6-week groupwas 18.1%, compared with 9.5% in the 6-month group (seeFigure 1). There was no difference between the groups in mortalityor in the rate of major haemorrhage (Schulman et al, 1995).

Although warfarin is effective in preventing VTE recurrence, itssafe use in cancer patients is complex. Warfarin has more than 200known drug, food, or botanical interactions that can result inirregular responses to treatment (Bick, 2006). In cancer patients,malnutrition, nausea, vomiting, and diarrhoea may make achiev-ing and maintaining a therapeutic INR with an oral agentchallenging. The need for therapeutic monitoring may also presenta considerable challenge and/or an undue burden to the patient.

Interruption of therapy is more likely in patients with cancer and,given the slow onset and offset of action of warfarin, this can leadto considerable gaps in anticoagulant coverage (Bick, 2006). Inaddition to these issues, some evidence suggests that patients withcancer who are receiving warfarin are not only at particular risk forrecurrent thromboembolic events but also have an increased riskfor bleeding (Prandoni et al, 2002).

As a result of the shortcomings of warfarin in cancer patients,the possible benefits of LMWH for prevention of recurrent VTEhave been evaluated in a number of randomised trials. The largestof these was the CLOT trial in which, after initial standard LMWHtreatment, oral anticoagulation and LMWH (dalteparin) werecompared for prevention of recurrent VTE in 672 patients withcancer (Lee et al, 2003). The recurrence rate over 6 months was 9%in the LMWH group compared with 17% in the oral anticoagulantgroup, a 52% (P¼ 0.002) reduction in favour of LMWH (Figure 2).There were no significant differences in major bleeding or deathbetween the two groups (Lee et al, 2003). In this trial, dalteparinwas administered at a full-treatment dose for the first month,followed by a 75% treatment dose for the remaining 5 months (Leeet al, 2003). This regimen is now licensed in the United Kingdomfor treatment of VTE in patients with cancer, and a number ofsystematic reviews and meta-analyses comparing LMWH with oralanticoagulation in the long-term treatment of VTE in patients withcancer have been published (Akl et al, 2008; Noble et al, 2008;Louzada et al, 2009). They show that, compared with oralanticoagulants, LMWH reduces the risk of recurrence by about50%, with no difference between treatment modalities in majorbleeding or mortality.

The use of anticoagulation in the thrombocytopenic patientremains challenging. Different options can be considered. Forexample, the label for dalteparin recommends a reduction in dailydose of 2500 IU if the platelet count falls to 50 –100� 109/l, until itrecovers to X100� 109/l, and discontinuation if the platelet countfalls to o50� 109/l. In a prospective cohort study, 203 patientswith metastatic cancer and VTE received LMWH at treatment dosefor 1 week (Monreal et al, 2006). The dose was then reduced to10 000 IU daily, irrespective of weight, for 3 months. The dose wasfurther reduced to 5000 IU daily at platelet counts o50� 109/l andto 2500 at counts o10� 109/l. Recurrent VTE developed in 9% ofpatients (fatal in two patients). Five percent had a major bleed(fatal in six patients) (Monreal et al, 2006). Although unsupportedby specific data, others have recommended a 50% dose reductionof LMWH if the platelet count is o50� 109/l and discontinuationif the platelet count is o20� 109/l (Falanga and Rickles, 2007).

Consideration should be given to insertion of a vena caval filterif anticoagulation is contraindicated because of thrombocytopeniaor active bleeding, but again firm data are lacking. Thus, theoptimal approach to management of VTE in this setting is

Cum

ulat

ive

prob

abili

tyof

rec

urre

nce

0.2

0.1

0.0

Months

6-month group

6-week group

0 2 4 6 8 10 12 14 16 18 20 22 24

Figure 1 Cumulative probability of recurrent VTE after a first episode,according to duration of anticoagulation with a vitamin K antagonist(Schulman et al, 1995). Copyright r [1995] Massacusetts Medical Society.All rights reserved.

Pro

babi

lity

of r

ecur

rent

ven

ous

thro

mbo

embo

lism

(%

)

25

20

15

10

5

0

No. at riskDalteparinOral anticoagulant

336 301 264 235 227 210 164336 242 154194200221280

Days after randomisation

Dalteparin

Oral anticoagulant

P =0.002

0 30 60 90 120 150 180 210

Figure 2 Probability of symptomatic recurrence of VTE among patientswith cancer, according to whether they received secondary prophylaxiswith dalteparin or oral anticoagulant therapy for acute VTE. Reproducedwith permission from Lee et al (2003).



S19

British Journal of Cancer (2010) 102(S1), S17 – S23& 2010 Cancer Research UK

uncertain. In individual circumstances in which there is a clearwish not to interrupt LMWH therapy, it may be reasonable tocontinue at a slightly reduced dose with platelet support tomaintain the count 450� 109/l and to reduce to a prophylacticdose, discontinue LMWH, or consider insertion of a vena cavalfilter if this is not possible.

Duration of anticoagulant therapy

The optimal duration of anticoagulant therapy for prevention ofrecurrent VTE in patients with cancer has not been specificallystudied. Guidelines recommend that it should continue as long asthe patient has active cancer (Lyman et al, 2007). In practice, thismeans indefinite treatments in patients with metastases or in thosereceiving anticancer therapy. However, each patient should beassessed individually in terms of the risk/benefit ratio ofcontinuing vs stopping anticoagulants, life expectancy, quality oflife, and patient preference. The CLOT trial data referred to earlierin favour of LMWH over oral anticoagulation did not extendbeyond 6 months. It seems likely that the benefit of LMWH wouldextend beyond this period if continuing anticoagulation isrequired, but consideration may be given to oral anticoagulanttherapy in individual instances.

Prevention of PTS

Post-thrombotic syndrome is a common cause of morbidityfollowing DVT. It occurs in 20–50% of patients overall aftersymptomatic DVT. Typical features include chronic pain, swelling,heaviness, oedema, and skin changes in the affected limb (Kahn,2006) and can be difficult to distinguish clinically from recurrentDVT. Use of graduated compression stockings providing an anklepressure of 40 mm Hg for 2 years has been compared with nomechanical support and shown to reduce the incidence of total andsevere PTS by approximately 50% (Brandjes et al, 1997).

Vena caval filters

There is only one open-label randomised trial of IVC filters inpatients (mostly without cancer) presenting with VTE (PREPIC,2005). In this trial, the filters were permanent and all patientsreceived standard anticoagulant therapy. Long-term follow-upsuggested that filter insertion was associated with a reduced risk ofPE, counter-balanced by an increased risk of DVT, and no overalleffect on mortality (PREPIC, 2005). Guidelines on the use of venacaval filters were published by the British Committee for Standardsin Haematology (Baglin et al, 2006). The consensus was that therewere no strong evidence-based indications for filters but theiruse should be considered in cases in which anticoagulation iscontraindicated, or has to be interrupted within the first monthof treatment to allow surgery, or in which there is occurrence ofPE despite adequate anticoagulation. In those cases in which anti-coagulation is contraindicated, it should be resumed as soon as thecontraindication is no longer present, and retrievable filters shouldbe considered in this setting. Such filters should be retrievedwithin 3 months of insertion where appropriate, or else left inpermanently.

Treatment of central venous catheter-relatedthromboembolism

No randomised, controlled trials have been reported that evaluatethe effects of particular therapeutic strategies on the outcomes ofcentral venous catheter (CVC)-related thrombosis. Indeed, boththe natural history and management of CVC-related thrombosishave been studied surprisingly little. In the RIETE Registry of 104patients with cancer and CVC-related thrombosis, 10% of patientspresented with symptomatic PE and 4% developed recurrent PE at

3-month follow-up, including one fatal event (Monreal et al, 2006).In routine practice, the catheter is commonly removed and thepatient is anticoagulated with LMWH in the standard way for VTEin patients with cancer (i.e., with LMWH) for at least 3 months.Whether it is necessary to remove the catheter is unclear. Currentrecommendations from the National Comprehensive CancerNetwork (NCCN) are to remove the catheter if it is not requiredand anticoagulate for at least 3 months. If the catheter is requiredand there is no contraindication to anticoagulation, it may bereasonable to leave the catheter in place and anticoagulate for atleast 3 months after it is eventually removed. If symptoms worsenwhile the catheter remains in place or if anticoagulation is contra-indicated, the catheter should be removed (NCCN, 2009).

Therapeutic anticoagulation failure

Therapeutic anticoagulation failure, defined as an extension ofDVT, or new DVT, or PE while on therapeutic levels of recom-mended anticoagulation therapy, is relatively common in patientswith cancer (Prandoni et al, 2002). The recurrence rate in theLMWH arm of the CLOT trial was 9%, much higher than wouldbe seen in a non-cancer population of patients with VTE. Manage-ment in this setting has not been systematically studied, and themeasures taken in practice depend in part on the individualcircumstances in which the recurrence took place. Among patientsreceiving vitamin K antagonist therapy, INR should be assessedand adjusted, if necessary, to a higher INR range after acutetherapy with UFH or LMWH (Streiff, 2006). Alternatively, andparticularly among patients who experience recurrence despitetherapeutic levels of a vitamin K antagonist, long-term LMWHrepresents a reasonable option. Vena caval filters representanother option in patients with therapeutic anticoagulation failure,particularly in those with VTE who have contraindicationsto anticoagulant therapy (NCCN, 2009). However, the hyper-coagulable state associated with cancer affects all vasculature;thus, regional approaches to preventing recurrent VTE, such asvena caval filters, may not provide adequate protection in thesepatients (Streiff, 2006). In fact, placement of an IVC filter may beassociated with an increased risk for recurrent thromboembolicdisease (NCCN, 2009).

When a recurrent thrombotic event occurs in a patient receivingheparin (UFH or LMWH), a diagnosis of heparin-inducedthrombocytopenia should be considered. Expert advice shouldbe sought, if necessary, for confirmation of diagnosis andprovision of non-heparin anticoagulants. More commonly, recur-rence represents therapeutic failure. Compliance issues should beconsidered along with escalation of the dose of LMWH orconverting from a once-daily to a twice-daily regimen (NCCN,2009). For patients receiving a 75–80% maintenance dose ofLMWH, this would mean increasing to a treatment dose. For thosereceiving a treatment dose at the time of recurrence, the dose couldbe increased by 20–25%. This approach was reported in a recentretrospective cohort study of 70 cancer patients with recurrentVTE despite anticoagulation (67% while on LMWH, 33% whilereceiving a vitamin K antagonist) (Carrier et al, 2009). Six (8.6%)patients had a second recurrence with this approach. The dose wasfurther escalated by 20–25% in three patients who had a furtherVTE event at 120% of a therapeutic dose. None of them hada thrombotic recurrence; however, bleeding was seen in 3 out of70 patients.

In conclusion, management of VTE in patients with cancerremains a challenge. Recent data support the use of LMWH bothas the initial treatment for VTE and as the preferred optionfor secondary prevention. It should continue for a minimum of6 months and longer if the cancer remains active. Both recurrentthrombosis and bleeding remain common problems, and furtherstudies are required to optimise management. The advent ofnew anticoagulants, many of them oral, will hopefully provide



S20


significant advances in therapy, although the challenge ofproviding anticoagulation without increasing the risk of bleedingis likely to remain.

DISCUSSION

Annie M Young: Should we be switching people that are onwarfarin to low-molecular-weight heparin? Specifically, howshould we manage patients who have had pulmonary emboli, areon warfarin, and have had treatment for cancer?

Peter K Maccallum: There are 2 issues. One is probably educa-tion of clinicians. We used to get a lot of patients who would comein under the relevant medical team and then be referred to theanticoagulant clinic on warfarin. In my experience, a smallproportion of patients are prepared to stay on warfarin, but themajority who are offered LMWH switch.

Annie M Young: I do not think outpatients are permitted achoice.

Ajay K Kakkar: There are 2 reasons for that. First, dalteparin wasnot available in the UK until recently for this indication. Thesecond issue is the point you raised. There will be a very largenumber of people who, beyond 6 months, will be on warfarin,because the general consensus view is that cancer patients withVTE should receive anticoagulant therapy to prevent recurrentthrombosis. Therefore, even in centres that use LMWH for the first6 months of therapy, they put them on warfarin because they donot want to promote self-injection. There is a study ongoing toevaluate whether 12 months of anticoagulation with LMWH ismore effective and as safe as 6 months of treatment. Anothercomplicating factor is the recent availability – or imminentavailability – of newer agents. The studies for these agents willinclude a small number of cancer patients, and it might beassumed that these agents should also be used in patients withcancer-associated thrombosis.

Faculty: Many haematologists are not aware of the issuessurrounding anticoagulation for the secondary prevention of VTE,and they tend to keep people on warfarin. There are also manyhaematologists who keep patients on LMWH, but reduce the doseby 75% at 6 months.

Peter K Maccallum: In the absence of evidence, one can have adiscussion with the patient to make a decision. There are somepatients who are, by the end of the 6 months, unwilling to continuedaily injections.

Faculty: Another problem is the drug-drug interactions seenwith warfarin, particularly in the cancer patient who is undergoingtreatment. In many ways, LMWH is easier to use.

Faculty: Qualitative data also suggest that warfarin has a nega-tive effect on quality of life in patients with cancer, in part becausethese patients require frequent venopuncture for monitoring, butalso because of the uncertainty and lack of freedom imposed by themonitoring regimen.

Ajay K Kakkar: Dosing presents a major issue with someLMWHs. How do you assess the appropriate de-escalated dose toensure that the safety profile of the regimen is maintained? LMWHis dosed close to the margin for an increased rate of bleeding, so itis important to be very regimen sensitive when using these agents.

John Pasi: Separately, recurrent thrombosis occurring in fullyanticoagulated cancer patients is a major issue.

Faculty: I agree. Most of the calls I get from the oncology unitare about patients who have had a recurrent event while onanticoagulation. At the moment, there are no good answers, andone has to make a common sense judgment regarding thetreatment of these patients. Another key issue is anticoagulationin thrombocytopenic patients.

Faculty: Can we discuss duration of therapy? Many believe thatpatients with advanced disease should stay on indefinite anti-coagulation. I want to challenge that, particularly in the typical breastor prostate cancer patient who has a median survival of 3–5 years.

Peter K Maccallum: The current recommendations suggest thatindefinite coagulation should be considered.

Faculty: But if a patient is in remission, is it safe to stop? I donot know of any physicians who keep cancer patients on anti-coagulation for 3–5 years after their first clot.

Annie M Young: It does happen in clinical practice.Faculty: I would suggest that even patients with stable disease

are at higher risk of VTE than patients without cancer.Peter K Maccallum: There is an opportunity to include this in

the consensus statement. If you look at global guidelines, the viewis generally that antithrombotic therapy should be continued whilethere is active cancer or active anticancer therapy. However, thereis always a question with breast cancer – what should physiciansdo with patients who are on tamoxifen for 5 years? I think this is avery difficult question, because warfarin is associated with a majorrisk for bleeding. In my opinion, warfarin at a full anticoagulantdose in such good-prognosis patients is problematic.

Faculty: Again, there is some role for discussion with patients. Inthe clinic, we often ask patients whether bleeding or havinganother VTE is their biggest worry. For many of them, thrombosisis a significant issue.

Ajay K Kakkar: This discussion clearly reflects the real problemof where we are in terms of treatment. We have solid data toguide treatment for the first 6 months, but no real guidancethereafter.

CONSENSUS STATEMENT

Following the development of VTE, cancer patients remain atelevated risk for recurrent thrombotic events. Treatment is furthercomplicated by the increased risk of bleeding while on anti-coagulants.

Acute management of VTE to prevent a fatal pulmonaryembolus, as well as reduction in both short-term and long-term

Table 1 Regimens and contraindications for LMWH in the United Kingdom

Agent Brand name Regimen for secondary prevention Key contraindications/cautions in patients with cancer

Dalteparinsodium

Fragmin For extended treatment:K 200 IU kg�1 (max 18 000 IU) s.c. once daily� 1 month

K Patients with cancer undergoing regional anaesthesia(Fragmin PI, 2009)

K Then 150 IU kg�1 (max 18 000 IU) s.c. oncedaily� 5 months (Fragmin PI, 2009)

K Dose reduction may be warranted in patients with cancer whoexperience thrombocytopenia orhave renal insufficiency (Fragmin PI, 2009)

Enoxaparinsodium

Clexane K 1.5 mg kg�1 once daily for extended prophylaxis(Meyer et al, 2002)

K Use with caution in patients with renal or hepatic impairmentand in low-weight patients (Clexane PI, 2009)

Tinzaparin Innohep K 175 U kg�1 once daily (Hull et al, 1979) K Use with caution in patients with renal impairment

Note that some regimens are based on studies conducted in non-cancer patients.



S21


sequelae of thrombosis, relies on the use of a rapid-acting,parenterally administered anticoagulant such as UFH, LMWH,or fondaparinux (see Table 1). In non-cancer patients, an oralvitamin K antagonist (e.g., warfarin) may be initiated andcontinued for at least 6 months, with indefinite treatmentconsidered in patients at increased risk of recurrence. In contrast,in patients with VTE and cancer, LMWH is now the preferredapproach because of better protection against recurrent thrombo-sis without increasing the risk of bleeding. It has additionaladvantages in this population in that it is more flexible in termsof interruption for invasive procedures or thrombocytopeniaand avoids the need for close INR monitoring that is essentialwith warfarin and is particularly challenging in cancer patients.The result is that blood tests and hospital visits can be minimised,thereby improving the quality of life. Warfarin may still be usedin cases in which there is a clear patient preference to avoidinjections. If the cancer is in remission and there are no additionalrisk factors for recurrence, anticoagulant therapy can generallybe stopped at 6 months. If the cancer remains active or there areongoing risk factors, consideration should be given to continuing

anticoagulation beyond 6 months, after discussing with the patient.Randomised trials have not been conducted beyond this stage inthese patients. Low-molecular-weight heparin is likely the optionof choice in cases in which the decision is taken to continueanticoagulation, but warfarin is an alternative approach in cases inwhich this would be preferred by the patient.

Massive pulmonary embolus, catheter-related thrombosis, andrecurrent VTE despite anticoagulation are clinical scenarios forwhich evidence on management in the setting of malignancy islimited. However, guidance on treatment, based on the limitedevidence, extrapolation from a non-malignant setting, and clinicalexpertise are provided. Management recommendations are basedon the NCCN Clinical Practice Guidelines in Oncology: VenousThromboembolic Disease (NCCN).

Conflict of interest

R Coleman has received lecture fees from Pfizer and grant supportfrom Boehringer Ingelheim. P MacCallum has received consultingfees from Pfizer and Boehringer Ingelheim.

REFERENCES

Akl EA, Terrenato I, Barba M, Sperati F, Sempos EV, Muti P, Cook DJ,Schunemann HJ (2008) Low-molecular-weight heparin vs unfractio-nated heparin for perioperative thromboprophylaxis in patients withcancer: a systematic review and meta-analysis. Arch Intern Med 168:1261 – 1269

Baglin TP, Brush J, Streiff M (2006) Guidelines on use of vena cava filters.Br J Haemotol 134: 590 – 595

Barritt DW, Jordan SC (1960) Anticoagulant drugs in the treatment ofpulmonary embolism. A controlled trial.. Lancet 1: 1309 – 1312

Bick RL (2006) Cancer-associated thrombosis: focus on extended therapywith dalteparin. J Support Oncol 4: 115 – 120

Blom JW, Vanderschoot JP, Oostindier MJ, Osanto S, van der Meer FJ,Rosendaal FR (2006) Incidence of venous thrombosis in a large cohort of66 329 cancer patients: results of a record linkage study. J ThrombHaemost 4: 529 – 535

Brandjes DP, Buller HR, Heijboer H, Huisman MV, de Rijk M, Jagt H,ten Cate JW (1997) Randomised trial of effect of compression stockingsin patients with symptomatic proximal-vein thrombosis. Lancet 349:759 – 762

Brandjes DP, Heijboer H, Buller HR, de Rijk M, Jagt H, ten Cate J (1992)Acenocoumarol and heparin compared with acenocoumarol alone inthe initial treatment of proximal-vein thrombosis. N Engl J Med 327:1485 – 1489

Buller HR, Davidson BL, Decousus H, Gallus A, Gent M, Piovella F,Prins MH, Raskob G, Segers AE, Cariou R, Leeuwenkamp O, Lensing AW(2004) Fondaparinux or enoxaparin for the initial treatment ofsymptomatic deep venous thrombosis: a randomized trial. Ann InternMed 140: 867 – 873

Buller HR, Davidson BL, Decousus H, Gallus A, Gent M, Piovella F, Prins MH,Raskob G, van den Berg-Segers AE, Cariou R, Leeuwenkamp O, Lensing AW(2003) Subcutaneous fondaparinux versus intravenous unfractionatedheparin in the initial treatment of pulmonary embolism. N Engl J Med349: 1695 – 1702

Carrier M, Le Gal G, Cho R, Tierney S, Rodger M, Lee AY (2009) Doseescalation of low molecular weight heparin to manage recurrent venousthromboembolic events despite systemic anticoagulation in cancerpatients. J Thromb Haemost 7: 760 – 765

Clexane [prescribing information]. Guildford. Sanofi-Aventis: Surrey, UK,2009

Elting LS, Escalante CP, Cooksley C, Avritscher EB, Kurtin D, Hamblin L,Khosla SG, Rivera E (2004) Outcomes and cost of deepvenous thrombosis among patients with cancer. Arch Intern Med 164:1653 – 1661

Falanga A, Rickles FR (2007) Management of thrombohemorrhagicsyndromes (THS) in hematologic malignancies. Hematology Am Soc HematolEduc Program 165 – 171, doi:2007/1/165 [pii] 10.1182/asheducation-2007.1.165

Fragmin [prescribing information], Pharmacia Limited: Sandwich, Kent,UK, 2009

Hull R, Delmore T, Genton E, Hirsh J, Gent M, Sackett D, McLoughlin D,Armstrong P (1979) Warfarin sodium versus low-dose heparin inthe long-term treatment of venous thrombosis. N Engl J Med 301: 855 –858

Kahn SR (2006) The post-thrombotic syndrome: progress and pitfalls.Br J Haematol 134: 357 – 365

Kearon C, Kahn SR, Agnelli G, Goldhaber S, Raskob GE, Comerota AJ(2008) Antithrombotic therapy for venous thromboembolic disease:American College of Chest Physicians Evidence-Based Clinical PracticeGuidelines (8th edn). Chest 133: 454S – 545S

Khorana AA, Francis CW, Culakova E, Kuderer NM, Lyman GH (2007)Thromboembolism is a leading cause of death in cancer patientsreceiving outpatient chemotherapy. J Thromb Haemost 5: 632 – 634

Laporte S, Mismetti P, Decousus H, Uresandi F, Otero R, Lobo JL, MonrealM (2008) Clinical predictors for fatal pulmonary embolism in 15 520patients with venous thromboembolism: findings from the RegistroInformatizado de la Enfermedad TromboEmbolica venosa (RIETE)Registry. Circulation 117: 1711 – 1716

Lee AY, Levine MN, Baker RI, Bowden C, Kakkar AK, Prins M, Rickles FR,Julian JA, Haley S, Kovacs MJ, Gent M (2003) Low-molecular-weightheparin versus a coumarin for the prevention of recurrent venousthromboembolism in patients with cancer. N Engl J Med 349: 146 – 153

Levitan N, Dowlati A, Remick SC, Tahsildar HI, Sivinski LD, Beyth R,Rimm AA (1999) Rates of initial and recurrent thromboembolic diseaseamong patients with malignancy versus those without malignancy: riskanalysis using Medicare claims data. Medicine (Baltimore) 78: 285 – 291

Louzada ML, Majeed H, Wells PS (2009) Efficacy of low- molecular- weight-heparin versus vitamin K antagonists for long term treatment of cancer-associated venous thromboembolism in adults: a systematic review ofrandomized controlled trials. Thromb Res 123: 837 – 844

Lyman GH, Khorana AA, Falanga A, Clarke-Pearson D, Flowers C,Jahanzeb M, Kakkar A, Kuderer NM, Levine MN, Liebman H,Mendelson D, Raskob G, Somerfield MR, Thodiyil P, Trent D,Francis CW (2007) American Society of Clinical Oncology guideline:recommendations for venous thromboembolism prophylaxis andtreatment in patients with cancer. J Clin Oncol 25: 5490 – 5505

Meyer G, Marjanovic Z, Valcke J, Lorcerie B, Gruel Y, Solal-Celigny P,Le Maignan C, Extra JM, Cottu P, Farge D (2002) Comparison oflow-molecular-weight heparin and warfarin for the secondary preventionof venous thromboembolism in patients with cancer: a randomizedcontrolled study. Arch Intern Med 162: 1729 – 1735

Monreal M, Falga C, Valdes M, Suarez C, Gabriel F, Tolosa C, Montes J(2006) Fatal pulmonary embolism and fatal bleeding in cancer patientswith venous thromboembolism: findings from the RIETE registry.J Thromb Haemost 4: 1950 – 1956



S22


NCCN Clinical Practice Guidelines in Oncology: Venous ThromboembolicDisease (V.1.2009) National Comprehensive Cancer Network Website. http://www.nccn.org/professionals/physician_gls/f_guidelines.asp.Accessed 28 August 2009

Noble SI, Shelley MD, Coles B, Williams SM, Wilcock A, Johnson MJ (2008)Management of venous thromboembolism in patients with advancedcancer: a systematic review and meta-analysis. Lancet Oncol 9: 577 – 584

Prandoni P, Lensing AW, Piccioli A, Bernardi E, Simioni P, Girolami B,Marchiori A, Sabbion P, Prins MH, Noventa F, Girolami A (2002)Recurrent venous thromboembolism and bleeding complications duringanticoagulant treatment in patients with cancer and venous thrombosis.Blood 100: 3484 – 3488

PREPIC Study Group (2005) Eight-year follow-up of patients withpermanent vena cava filters in the prevention of pulmonary embolism:the PREPIC (Prevention du Risque d0Embolie Pulmonaire par Interrup-tion Cave) randomized study. Circulation 112: 416 – 422

Quinlan DJ, McQuillan A, Eikelboom JW (2004) Low-molecular-weightheparin compared with intravenous unfractionated heparin for treat-ment of pulmonary embolism: a meta-analysis of randomized, controlledtrials. Ann Intern Med 140: 175 – 183

Ruiz-Gimenez N, Suarez C, Gonzalez R, Nieto JA, Todolı JA, Samperiz AL,Monreal M (2008) Predictive variables for major bleeding events inpatients presenting with documented acute venous thromboembolism.Findings from the RIETE Registry. Thromb Haemost 100: 26 – 31

Schulman S, Rhedin AS, Lindmarker P, Carlsson A, Larfars G, Nicol P,Loogna E, Svensson E, Ljungberg B, Walter H (1995) A comparison of sixweeks with six months of oral anticoagulant therapy after a first episodeof venous thromboembolism. Duration of Anticoagulation Trial StudyGroup. N Engl J Med 332: 1661 – 1665

Streiff MB (2006) Long-term therapy of venous thromboembolism incancer patients. J Natl Compr Cancer Netw 4: 903 – 910

Torbicki A, Perrier A, Konstantinides S, Agnelli G, Galie N, Pruszczyk P,Bengel F, Brady AJ, Ferreira D, Janssens U, Klepetko W, Mayer E, Remy-Jardin M, Bassand JP, Vahanian A, Camm J, De Caterina R, Dean V,Dickstein K, Filippatos G, Funck-Brentano C, Hellemans I, KristensenSD, McGregor K, Sechtem U, Silber S, Tendera M, Widimsky P,Zamorano JL, Zamorano JL, Andreotti F, Ascherman M, AthanassopoulosG, De Sutter J, Fitzmaurice D, Forster T, Heras M, Jondeau G, Kjeldsen K,Knuuti J, Lang I, Lenzen M, Lopez-Sendon J, Nihoyannopoulos P,Perez Isla L, Schwehr U, Torraca L, Vachiery JL (2008) Task Force forthe Diagnosis and Management of Acute Pulmonary Embolism ofthe European Society of Cardiology. Guidelines on the diagnosisand management of acute pulmonary embolism: the Task Force forthe Diagnosis and Management of Acute Pulmonary Embolism of theEuropean Society of Cardiology (ESC). Eur Heart J 29: 2276 – 2315

Trujillo-Santos J, Nieto JA, Tiberio G, Piccioli A, Di Micco P, Prandoni P,Monreal M (2008) Predicting recurrences or major bleeding in cancerpatients with venous thromboembolism. Findings from the RIETERegistry. Thromb Haemost 100: 435 – 439

van Dongen CJ, van den Belt AG, Prins MH, Lensing AW (2004) Fixeddose subcutaneous low molecular weight heparins versus adjusteddose unfractionated heparin for venous thromboembolism. CochraneDatabase Syst Rev Art. No. CD001100, doi:10.1002/14651858.CD001100.pub2

van Doormaal FF, Raskob GE, Davidson BL, Decousus H, Gallus A, LensingAW, Piovella F, Prins MH, Buller HR (2009) Treatment of venousthromboembolism in patients with cancer: subgroup analysis of theMatisse clinical trials. Thromb Haemost 101: 762 – 769

Wan S, Quinlan DJ, Agnelli G, Eikelboom JW (2004) Thrombolysiscompared with heparin for the initial treatment of pulmonary embolism:a meta-analysis of the randomized controlled trials. Circulation 110:744 – 749

CLINICAL SCENARIOS

Treatment of limb VTE

� Acute management* Commence LMWH in preference to UFH

— Dalteparin 200 U kg�1 once daily s.c.— Tinzaparin 175 U kg�1 once daily s.c.— Enoxaparin 1.5 mg kg�1 once daily s.c. (alternatively,

enoxaparin 1 mg kg�1 every 12 h s.c.)* Fondaparinux may be considered

— Fondaparinux 5 mg (o50 kg); 7.5 mg (50–100 kg); 10 mg(4100 kg) daily s.c.

� Maintenance treatment* Continue LMWH at 75– 80% treatment dose* Alternatively, administer warfarin at a dose to achieve an

INR of 2 –3 if not on myelosuppressive chemotherapy orcomplex supportive/concomitant medications

* Continue for at least 6 months. Indefinite treatment recom-mended if active metastatic disease is observed or there iscontinued exposure to potentially thrombogenic anticancertherapy

� Anticoagulation contraindicated* Recent central nervous system bleed* Active major bleed (42 U in 24 h)* Platelets o 50� 109/l* Severe platelet dysfunction* Known bleeding tendency, e.g., haemophilia* Elevated PT or aPTT above therapeutic target level* Severely limited life expectancy or no palliative benefit* Patient refusal

Pulmonary embolus

� Assess severity, including ECHO or CT angiography if necessaryfor right heart enlargement. Consider as high risk for death(15%) if systolic BP o90 mm Hg

� High-risk massive embolus

* Oxygen* Commence heparin 80 U kg�1 and infuse at 18 U kg�1 h�1 to

maintain aPTT between 2.0 and 2.5 (p 16)* Commence thrombolyis with rtPA 100 mg over 2 h or up to

50 mg over 15 min if rapidly deteriorating� Uncomplicated

* Treat as for VTE

CVC-related thrombosis

� Preservation of catheter access not essential* Remove line and anticoagulate for 3 months

� Preservation of catheter access clinically important* Anticoagulate with line in situ and for at least 3 months after

removal. Removal of line and reinsertion at a later date maybe necessary if symptoms worsen (e.g., recurrent emboli orincreasing arm swelling)

Vena caval filter placement

� Contraindications to anticoagulation (p 32)� Failure of anticoagulation� Non-compliance� Documented multiple PE and chronic pulmonary hypertension

Therapeutic anticoagulant failure

� During vitamin K antagonist therapy* Check compliance* Increase target INR to 3– 4* Change to LMWH* Consider vena caval filter

� During LMWH therapy* Check compliance* Increase 75 to 80% maintenance dose to full-treatment dose* Consider increase from treatment dose by 20– 25%* Convert once-daily to twice-daily administration* Consider vena caval filter



S23


http://www.nccn.org/professionals/physician_gls/f_guidelines.asp

therapy, variable nutrition status, greater difficultymaintaining an international normalized ratio(INR) of between 2 and 3 (43.3% time withinrange compared with 56.9% in noncancer patientcontrols), and limited venous access to supporttherapeutic monitoring.1-3 Bleeding during antico-agulation is of particular concern in patients withdisease-related and therapy-related thrombocy-topenia and cancer that involves the central nerv-ous system.

In the setting of acute VTE, the low-molecular-weight heparin (LMWH) enoxaparin has been shownto be equally effective and safe for initial antico-agulation compared with unfractionated heparin.4,5

Enoxaparin has the advantage of less nonspecificprotein binding, subcutaneous weight-based dos-ing without the need for monitoring in most cases,and less heparin-induced thrombocytopenia (HIT).6,7

Subcutaneous, weight-based enoxaparin has also

Venous thromboembolic events (VTEs), includingdeep venous thrombosis (DVT) and pulmonaryembolism, warrant prompt initiation of antithrom-botic therapy to prevent thrombus extension,embolization, and early as well as late recurrence.Challenges of VTE management in cancer patientscompared with noncancer patients include greaterheparin resistance due to excess circulating acute-phase proteins, increased recurrence and bleed-ing rates during standard-intensity oral warfarin

Secondary Prevention of Venous ThromboembolicEvents in Patients With Active Cancer: Enoxaparin

Alone Versus Initial Enoxaparin Followedby Warfarin for a 180-Day Period

*Steven R. Deitcher, MD, †Craig M. Kessler, MD, ‡Geno Merli, MD,§James R. Rigas, MD, **Roger M. Lyons, MD, and ‡‡Jawed Fareed, PhD,

for the ONCENOX Investigatorsa

*Section of Hematology and Coagulation Medicine, The Cleveland Clinic Foundation, Cleveland, Ohio;†Lombardi Cancer Center, Georgetown University Medical Center, Washington, District of Columbia;

‡Division of General Internal Medicine, Thomas Jefferson University, Philadelphia, Pennsylvania; §Division of Hematologyand Medical Oncology, Dartmouth Medical School, Lebanon, New Hampshire; **U.S. Oncology, San Antonio, Texas;

††Aventis Pharmaceuticals, Bridgewater, New Jersey; ‡‡Loyola University Medical Center, Maywood, Illinois.

Summary: This study evaluated enoxaparin alone versus initialenoxaparin followed by warfarin in secondary prevention of venousthromboembolic events in adults with active malignancy. Cancerpatients (n = 122) with acute symptomatic venous thromboembolicevents were randomly allocated to receive subcutaneous enoxaparin1.0 mg/kg every 12 hours for 5 days, followed by 1.0 mg/kg daily(group 1a) or 1.5 mg/kg daily (group 1b) for 175 days, or subcuta-neous enoxaparin 1.0 mg/kg every 12 hours for at least 5 days and

until a stable international normalized ratio of 2 to 3 was achievedon oral warfarin begun on day 2 and continued to day 180 (group 2).There were no significant differences in major and minor bleedingrates between treatment groups. No bleeding events were intracra-nial or fatal. Enoxaparin treatment was feasible, generally well toler-ated, and effective for a 180-day period in the secondary preventionof venous thromboembolic events in patients with active malignancy.Key Words: Enoxaparin—Deep venous thrombosis—Malignancy

Address correspondence to Steven R. Deitcher, MD,201 Industrial Road, Suite 310, San Carlos, CA 94070; e-mail:[email protected] full list of ONCENOX investigators is given in the Appendix.

Clinical and Applied Thrombosis/HemostasisVol. 12, No. 4, October 2006 389-396DOI: 10.1177/1076029606293692© 2006 Sage Publications

389 by guest on May 27, 2010 http://cat.sagepub.comDownloaded from

http://cat.sagepub.com

become a popular warfarin substitute in patientswith VTE who have difficult to regulate INRs orhave recurrent thrombosis despite an INR of between2 and 3 (“warfarin failure”), as is often seen incancer patients.

We performed this study to evaluate patientrecruitment, safety, efficacy, and compliance for a180-day period of enoxaparin alone versus initialenoxaparin followed by warfarin in the secondaryprevention of VTE in adult patients with activemalignancy. Baseline hemostatic profiling andassessment for inherited hypercoagulable stateswere performed to partially characterize the patho-genesis of thrombosis associated with cancer.Testing for heparin-associated antibodies was per-formed to evaluate the risk of HIT in enoxaparin-exposed cancer patients who are already prone tothrombocytopenia by virtue of their disease.

PATIENTS AND METHODS

This pilot feasibility study was conducted as a ran-domized, open-label, multidose, active comparator,parallel-design trial. Patients were randomizedto receive 1 of 3 treatments. Group 1a receivedsubcutaneous twice-daily enoxaparin (1.0 mg/kg)for 5 days, followed by once-daily enoxaparin(1.0 mg/kg) for 175 days; group 1b received sub-cutaneous twice-daily enoxaparin (1.0 mg/kg) for5 days, followed by once-daily enoxaparin (1.5mg/kg) for 175 days; and group 2 received subcu-taneous twice-daily enoxaparin (1.0 mg/kg) for aminimum of 5 days and until achievement of a sta-ble INR between 2 and 3 on oral warfarin begun onday 2 of enoxaparin and continued for a total of180 days of anticoagulation. Patients were moni-tored for 7 months from the date of randomization.

The primary objectives of this trial were toevaluate the feasibility of recruiting the neces-sary number of cancer patients (300 evaluablepatients) in a 12-month time frame and to evalu-ate the feasibility of compliance with long-termdaily subcutaneous enoxaparin injections in activecancer patients with acute VTE. Compliance wasdefined as the percentage of enoxaparin or war-farin doses dispensed that were actually taken bythe patient.

The secondary objectives included an evalua-tion of the safety of enoxaparin treatment alone(groups 1a and 1b) compared with enoxaparin fol-lowed by warfarin treatment (group 2) adminis-tered for 180 days to prevent secondary VTEin cancer patients, as determined by assessmentof major and minor bleeding rates and serious

adverse events (SAEs). A bleeding event was con-sidered major if it resulted in death, a serious, life-threatening clinical event requiring hospitalization,transfusion of at least 2 units of packed red bloodcells, a fall in hemoglobin of 2 grams or more thatwas attributable to the bleeding event, a retroperi-toneal, intracranial, or intraocular hemorrhage;the need for surgery or decompression of a closedspace; or an ecchymosis or hematoma greater than10 cm in diameter. Another secondary objectivewas to evaluate the efficacy of VTE treatment withenoxaparin alone (groups 1a and 1b) comparedwith enoxaparin followed by warfarin treatment(group 2).

Efficacy was determined by assessment of objec-tively confirmed recurrent VTE involving a not pre-viously involved venous segment and symptomaticVTE extension within the same venous segment asthe index event during treatment. Baseline testingfor acquired and inherited hypercoagulable stateswas performed on the first day of the study.Testing for the development of heparin-associatedantibodies was performed in groups 1a and 1b ateach monthly follow-up evaluation. The HemostasisResearch Laboratories of Loyola University MedicalCenter, Chicago, Illinois, performed all special coag-ulation testing.

To be eligible for enrollment in this study,patients had to be aged 18 years or older, weigh120 kg or less, and have a functional capacitybased on Karnofsky performance scale of 60 ormore, or an Eastern Cooperative Oncology Group(ECOG) score of 0, 1, or 28 based on the mostrecent performance status before their acute VTE.Patients could be enrolled or randomized within72 hours of VTE diagnosis and if LMWH or unfrac-tionated heparin had already been initiated as astandard of care therapy. All index VTEs had to beobjectively confirmed by appropriate imagingstudies. Catheter-associated VTE were not eligibleindex thrombotic events.

Patients had to have active, residual malig-nancy determined by the presence of measurabledisease, persistently elevated tumor markers,metastatic disease after tumor debulking, or his-tologically or cytologically confirmed cancer. Atstudy entry, a patient could not be a candidate forcurative intent surgery. Based on the investigators’judgment, all patients had to have an estimatedlength of survival that would allow for study com-pletion. The study imposed no general or dietaryrestrictions.

Patients were excluded from enrollment if theyhad an anticipated need for thrombolytic therapy,embolectomy, or placement of a new caval filter,

390 DEITCHER ET AL

by guest on May 27, 2010 http://cat.sagepub.comDownloaded from


as were those whose active cancer was acuteleukemia or a localized cutaneous malignancy.Patients with any contraindication to anticoagula-tion, including severe liver disease, known nonir-radiated intracerebral metastases, deep organbiopsy within 2 weeks, and major surgery within 1week were excluded. Patients with a baseline INRof 2 or more, known or suspected severe renalinsufficiency (creatinine clearance of 30 mL/minor less), history of HIT, history of warfarin-associated skin necrosis, and baseline plateletcount of less than 50,000/µL were not eligible.

Safety evaluations were performed on the safetypopulation defined as all randomized patientswho received at least 1 dose of study medica-tion. The intent-to-treat population included allpatients in the safety population who had at least1 follow-up measurement. Descriptive statisticswere used to summarize the incidence of majorand minor hemorrhagic events and the incidenceof recurrent VTE. Statistical analyses were per-formed using SAS 8.2 (SAS Institute, Cary, SC). Allstatistical tests performed used 2-sided hypothesistests at the overall 5% level of significance.

RESULTS

In the period from January 26, 2001, to March 28,2002, 102 patients from 27 sites were recruited.The appropriate institutional review board at each

investigative site approved this study, and allpatients signed an approved informed consentform. Each of the 3 treatment arms had approxi-mately equal numbers of subjects: 32 in theenoxaparin 1.0 mg/kg group, 36 in the enoxa-parin 1.5 mg/kg group, and 34 in the warfaringroup. One subject in the 1.0 mg/kg enoxaparingroup did not receive study drug. Of the 101patients in the safety sample, 91 were included inthe intent-to-treat analysis. Nine patients (5 in the1.0 mg/kg enoxaparin group and 4 in the 1.5mg/kg enoxaparin group) did not meet full entrycriteria and were enrolled after sponsor approval.

Table 1 summarizes the baseline demographiccharacteristics and index VTE diagnosis of thesafety population. Most patients in the 1.5 mg/kgenoxaparin group were female (63.9%) and aged51 years or older (58.3%). In contrast, the 1.0mg/kg enoxaparin group was 51.6% female and41.9% were aged 51 years or older. The warfaringroup was 47.1% female and younger than thosein the enoxaparin groups. Most patients in the 3treatment arms were diagnosed with DVT, andmore patients in the warfarin group had bothDVT and pulmonary embolism. Overall, 8.7% ofpatients had a history of VTE before the studyindex event. None of these differences were statis-tically significant.

Of the 91 patients in the intent-to-treat population,49 (53.8%) completed 180 days of study medica-tion. The primary reasons for discontinuation from

TREATMENT OF VENOUS THROMBOSIS IN CANCER 391

TABLE 1. Baseline Demographic Characteristics and Index Venous Thromboembolism Diagnosis: Safety Sample

Enoxaparin 1.0 mg/kg Enoxaparin 1.5 mg/kg Warfarin TotalCharacteristic (n = 31) (n = 36) (n = 34) (N = 101)

AgeMean ± SD 62.7 ± 13.4 64.0 ± 10.7 64.1 ± 12.4 63.7 ± 12.0Range 35-80 36-79 40-87 35-87

Age category (%)≤50 years 6 (19.4) 3 (8.3) 7 (20.6) 16 (15.8)51-60 years 6 (19.4) 11 (30.6) 2 (5.9) 19 (18.8)61-70 years 8 (25.8) 9 (25.0) 15 (44.1) 32 (31.7)>70 years 11 (35.5) 13 (36.1) 10 (29.4) 34 (33.7)

Race (%)Caucasian 25 (80.6) 29 (80.6) 32 (94.1) 86 (85.1)Black 5 (16.1) 4 (11.1) 2 (5.9) 11 (10.9)Asian 0 (0.0) 1 (2.8) 0 (0.0) 1 (1.0)Hispanic 0 (0.0) 1 (2.8) 0 (0.0) 1 (1.0)Other 1 (3.2) 1 (2.8) 0 (0.0) 2 (2.0)

Index VTE (%)PE 12 (38.7) 17 (47.2) 15 (44.1) 44 (43.6)DVT 24 (77.4) 29 (80.6) 31 (91.2) 84 (83.2)PE and DVT 7 (22.6) 10 (27.8) 13 (38.2) 30 (29.7)

SD = standard deviation; VTE = venous thromboembolic event; PE = pulmonary embolism; DVT = deep venous thrombosis.



the study differed among the treatment groups(Table 2). The discontinuation rate (58.3%) washighest in the enoxaparin group receiving 1.5mg/kg, with the most common reasons beingdeath (19.4%) and adverse events (13.9%). Thelowest discontinuation rate (41.9%) was in theenoxaparin group receiving 1.0 mg/kg. The pri-mary reason (12.9%) for discontinuation in thisgroup was reaching the study end point (recurrentVTE or major hemorrhage).

Table 3 summarizes the cancer stage at studyentry for the safety population. More subjects inthe enoxaparin group receiving 1.5 mg/kg pre-sented with stage IV cancer at the start of thestudy (66.7%) than in the other groups (54.8%and 52.9% for the 1.0 mg/kg and warfarin groups,respectively). It should be noted that there was asizeable percentage of subjects with “unknown”cancer stage in the 1.0-mg/kg enoxaparin (9.7%)and the warfarin (11.8%) groups. More subjectsin the 1.5-mg/kg enoxaparin group had alsobeen treated with radiation therapy (38.9%) andchemotherapy (58.3%) than those in the 1.0-mg/kgenoxaparin (32.4% and 55.9%, respectively) and

the warfarin groups (32.3% for both radiation andchemotherapy).

Compliance data were available for 98 of 101treated subjects. The overall compliance rate inthe 3 treatment groups averaged 95% throughoutthe study. Mean overall treatment compliance wasslightly lower in the warfarin group (90.1%) thanin the 1.0-mg/kg and 1.5-mg/kg enoxaparingroups (97.9% and 97.0%, respectively). Across allvisits, 92 patients (91.1%) took 81% to 100% oftheir study drug. Six subjects (5.9%) took lessthan 81% of the dispensed drug. Compliance waspoorest with warfarin (50% to 100% compliance)and best in the 1.0-mg/kg enoxaparin group (82%to 100% compliance).

Few patients experienced objectively confirmedsymptomatic extension of their index VTE or truerecurrent VTE (Table 4). More intent-to-treatpatients in the warfarin treatment arm experiencedVTE extension or recurrence than in either of theenoxaparin treatment arms. Three patients (10.0%)in the warfarin group, but only 2 patients in eachof the enoxaparin groups (6.9% of 1.0 mg/kggroup and 6.3% of 1.5 mg/kg group), experienced

392 DEITCHER ET AL

TABLE 2. Reasons for Patient Withdrawal or Termination

Enoxaparin 1.0 mg/kg, Enoxaparin 1.5 mg/kg, Warfarin, Total,Withdrawal Reasons n = 32 (%) n = 36 (%) n = 34 (%) N = 101 (%)

Adverse event 1 (3.2) 5 (13.9) 1 (2.9) 7 (6.9)Progressive disease 2 (6.5) 3 (8.3) 6 (17.6) 11 (10.9)Study end point 4 (12.9) 2 (5.6) 1 (2.9) 7 (6.9)Nonpermitted therapy 0 (0.0) 0 (0.0) 1 (2.9) 1 (1.0)Major protocol violation 1 (3.2) 0 (0.0) 0 (0.0) 1 (1.0)Lost to follow-up 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)Consent withdrawn 3 (9.7) 2 (5.6) 2 (5.9) 7 (6.9)Death 2 (6.5) 7 (19.4) 3 (8.8) 12 (11.9)

Malignancy 2 (6.5) 6 (16.7) 2 (5.9) 10 (9.9)Fatal VTE 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)Fatal hemorrhage 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)Other 0 (0.0) 1 (2.8) 1 (2.9) 2 (2.0)

Other 0 (0.0) 2 (5.6) 4 (11.8) 6 (5.9)

VTE = venous thromboembolic event.

TABLE 3. Cancer Stage at Study Entry: Safety Sample

Enoxaparin 1.0 mg/kg, Enoxaparin 1.5 mg/kg, Warfarin, Total,Cancer Stage n = 31 (%) n = 36 (%) n = 34 (%) N = 101 (%)

I 2 (6.5) 0 (0) 2 (5.9) 4 (4.0)II 2 (6.5) 4 (11.1) 1 (2.9) 7 (6.9)III 7 (22.6) 8 (22.2) 9 (26.5) 24 (23.8)IV 17 (54.8) 24 (66.7) 18 (52.9) 59 (58.4)Unknown 3 (9.7) 0 (0) 4 (11.8) 7 (6.9)



VTE during anticoagulant treatment. Overall,3.3% of study subjects experienced a symptomaticVTE extension and 4.4% experienced a new VTEfor a total thrombosis event rate of 7.7%. Nopatient was diagnosed with a new pulmonaryembolism during the study. Two patients not in theintent-to-treat sample, 1 in each of the enoxaparingroups, experienced recurrent DVT.

Because VTE developed in only a small numberof patients in the study, no trends or significancecould be observed. There was no effect of age,gender, race, clinical cancer stage, duration of can-cer diagnosis, or type of index VTE on the proba-bility that VTE would develop during the 6-monthtreatment period. Subjects with an ECOG per-formance status of 2, on the other hand, had a sig-nificantly increased risk of experiencing a VTEduring the study than did subjects with a score of0 or 1 (hazard ratio, 7.253; 95% confidence inter-val, 1.580 to 33.296; P =.011 from proportionalhazards model).

Table 5 summarizes adverse events for thesafety study population. Overall, 87.1% of patientsexperienced at least 1 nonserious adverse eventduring the study, and the number of events wasevenly distributed among the 3 treatment groups.Approximately half of all study patients experiencedat least 1 SAE during the study. The 1.0-mg/kg

enoxaparin and the warfarin groups had similarSAE rates (51.6% and 50.0%, respectively), andthe 1.5-mg/kg enoxaparin group had the highestincidence of SAEs (63.9% had at least 1 SAE).More cardiovascular SAEs occurred in the enoxa-parin groups than in the warfarin group (6 and 9versus 0), but the distribution of events in otherbody systems was nearly even across groups. Only3 subjects in the study had SAEs that were consid-ered to be “possibly” or “probably” related to thestudy drug: 2 in the warfarin group and 1 in the1.0-mg/kg enoxaparin group.

Overall, 58 patients experienced at least 1 hem-orrhagic event (major or minor) during the study.Seven patients experienced at least 1 major hem-orrhagic event: 1 in the warfarin group, 2 in the1.0-mg/kg enoxaparin group, and 4 in the 1.5-mg/kg enoxaparin group. All major hemorrhagicevents were considered SAEs but were analyzedseparately from other adverse events. The incidenceand types of adverse events and hemorrhagic eventsobserved with enoxaparin alone versus enoxa-parin followed by warfarin were similar to thosedescribed in the package inserts and the investiga-tor’s brochure.

Of the 33 deaths during the 7-month observationperiod, 29 (88%) were attributable to progression ofunderlying malignancy. Of the remaining 4 subjects,


TABLE 4. Analysis of Venous Thromboembolic Event Recurrence and Symptomatic Extension: Intent-to-Treat Sample

Enoxaparin 1.0 mg/kg, Enoxaparin 1.5 mg/kg, Warfarin, Total,VTE Event n = 29 (%) n = 32 (%) n = 30 (%) N = 91 (%)

Symptomatic extension of index VTEa 1 (3.4) 1 (3.1) 1 (3.3) 3 (3.3)Recurrent VTEb 1 (3.4) 1 (3.1) 2 (6.7) 4 (4.4)All VTE 2 (6.9) 2 (6.3) 3 (10.0) 7 (7.7)

VTE = venous thromboembolic event.a. Extension of the index thrombosis within the same venous segments that were originally involved.b. Detection of new thrombosis within a venous segment not previously involved.

TABLE 5. Summary of Adverse Events

Enoxaparin 1.0 mg/kg, Enoxaparin 1.5 mg/kg, Warfarin, Total,Adverse Event n = 31 (%) n = 36 (%) n = 34 (%) N = 101 (%)

Nonserious 27 (87.1) 32 (88.9) 29 (85.3) 88 (87.1)Treatment-related, nonserious 3 (9.7) 6 (16.7) 8 (23.5) 17 (16.8)Discontinued due to nonserious AE 1 (3.2) 0 (0.0) 2 (5.9) 3 (3.0)SAE 16 (51.6) 23 (63.9) 17 (50.0) 56 (55.4)Minor hemorrhage event 19 (61.3) 20 (55.6) 17 (50.0) 56 (55.4)Major hemorrhage event 2 (6.5) 4 (11.1) 1 (2.9) 7 (6.9)Died 7 (22.6) 15 (41.7) 11 (32.4) 33 (32.7)

AE = adverse event; treatment-related = probably or possibly related to the study drug, in the opinion of the investigator; SAE = seriousadverse event.



1 each died of presumed pulmonary embolism(1.0-mg/kg enoxaparin group), VTE (1.0-mg/kgenoxaparin group), cardiac arrest (warfaringroup), and heart failure (1.5-mg/kg enoxaparingroup). Only 12 patients were discontinued fromthe study due to death (Table 2). The remainder ofthe patients who died during the 7-month obser-vation period did so either after study discontinu-ation for another reason or after completion of the180-day period of study medication.

There were no adverse trends in platelet countduring the study. The mean and median values forplatelet count were within the normal limits for all3 groups at all study visits, including at baseline.Thrombocytopenia was reported in 7 patients: 5 inthe warfarin group and 1 in each of the enoxa-parin groups; however, a causal relationship tostudy medication was described for only 1 subjectin the warfarin group.

Mean antifactor Xa activity levels were between0.44 and 0.73 U/mL over the treatment period inthe 1.0-mg/kg enoxaparin group and between 0.59and 1.04 U/mL in the 1.5-mg/kg enoxaparin group.Overall, 25 samples had antifactor Xa activity levelsof more than 2.00 U/mL. Most samples had antifac-tor Xa activity levels between 0.5 and 1.5 U/mL.Table 6 summarizes the results of molecular coagu-lation studies performed in the 72 study patientswho provided informed consent for such test-ing. Factor V Leiden, prothrombin G20210A, andhomozgous MTHFR C677T prevalence rates were8%, 1%, and 8% respectively.

Table 7 summarizes the results of baseline coag-ulation studies. In addition, 74 patients underwentantigenic and functional tissue factor pathwayinhibitor (TFPI) testing. Antigenic TFPI levels in27% of tested patients were less than 70 ng/mL.As determined by functional TFPI testing, only 8%

of patients had decreased levels. Anticardiolipinantibody testing showed that 7% had elevatedtiters of immunoglobulin (Ig)G anticardiolipinantibody, and 3% had elevated titers of IgM anti-cardiolipin antibody. Of 514 samples tested foranti-heparin-platelet factor 4 antibodies by enzyme-linked immunosorbent assay (ELISA), 60 samplesfrom 31 different patients were positive. The titervaried from slightly positive (OD >0.40) tostrongly positive (OD >1.00, 10 samples). Sevenpositive samples were baseline samples. Whenthese 60 samples were tested by 14C serotoninrelease assay, none were positive.

DISCUSSION

This clinical trial was designed as a 3-arm, pilotfeasibility study to evaluate patient recruitment,compliance, safety, and efficacy for a 180-dayperiod of treatment with enoxaparin alone versusenoxaparin followed by warfarin in the secondaryprevention of VTE in patients with active malig-nancy. The objective to recruit the necessary num-ber of patients within a 12-month time frame wasnot met; however, compliance, safety, and efficacyresults were available for 102 enrolled patients.

The overall compliance rate in the 3 treatmentgroups was high, averaging 95% throughout the 6-month treatment period. Overall, average treatmentcompliance was slightly higher with enoxaparinalone (97.9% for 1.0-mg/kg and 97.0% for 1.5-mg/kg enoxaparin) than with warfarin (90.1%).This indicates that long-term subcutaneous adminis-tration of enoxaparin was generally well toleratedby patients. The need for daily subcutaneous injec-tion was not an obvious deterrent to study partici-pation or completion. The antifactor Xa activitylevels observed in most patients treated with bothdoses of enoxaparin were within a range believed tobe the desired target range for VTE treatment.

The incidence of recurrent VTE in the intent-to-treat population was numerically lower withenoxaparin therapy than with initial enoxaparinfollowed by oral warfarin. No numeric differencein the recurrent VTE rate was observed for the 2studied once-daily enoxaparin dosages. Althoughstatistical significance was not reached, the cumu-lative probability of being VTE-free at 6 monthswas numerically higher for enoxaparin alone thanfor enoxaparin followed by warfarin therapy.Standard of care included twice-daily subcuta-neous administration of enoxaparin at 1.0 mg/kgfor a minimum of 5 days, with warfarin therapystarted within 48 hours of enoxaparin initiation

394 DEITCHER ET AL

TABLE 6. Molecular Coagulation Studies

Laboratory Test ONCENOX Subjects, n (%)

Factor V LeidenNormal 66 (92)Heterozygous 6 (8)Homozygous 0 (0)

Prothrombin G20210ANormal 71 (99)Heterozygous 1 (1)Homozygous 0 (0)

MTHFR C677TNormal 42 (58)Heterozygous 24 (33)Homozygous 6 (8)

MTHFR = 5,10-methylene tetrahydrofolate reductase.



with a target INR of 2.5 for 2 consecutive days(range, 2.0 to 3.0).4 This was in accordance withrecommendations by the Sixth American Collegeof Chest Physician’s Consensus Conference onAntithrombotic Therapy for the treatment of VTE.9

The observed trend toward greater secondary VTEprevention with long-term enoxaparin comparedwith warfarin is in line with similar observationswith enoxaparin and other LMWHs.10-12

Overall, the incidence, type, and intensity ofadverse events observed with enoxaparin adminis-tered alone for 6 months in the current study wereno more extensive than those described for thestandard therapy of enoxaparin followed by war-farin. The higher rate of study discontinuation dueto death in the 1.5-mg/kg enoxaparin group mayreflect the greater number of patients with stageIV cancer and greater number of patients receiv-ing chemotherapy or radiation therapy in this armof the study. There were no unexpected adverseevents or other clinically significant safety findingsthat negated the use of long-term enoxaparin ther-apy as an alternative to oral warfarin.

The prevalence rates for factor V Leiden het-erozygosity, prothrombin gene G20210A heterozy-gosity, and homozygous MTHFR C677T did notdiffer significantly from rates reported for normalpopulations and are actually numerically less thanrates reported for patients with idiopathic VTE.13

Common molecular defects associated with hyper-coagulability in general were clearly not the pri-mary determinant of thrombosis in our cancerpopulation. The baseline laboratory characteriza-tion reflects the complex and variable nature ofthe hypercoagulability of malignancy. Acquired

activated protein C resistance based on testing ofsamples that have not been prediluted with factorV-deficient plasma (first generation) has beendescribed in association with several tumor his-tologies.14 Combinations of decreased levels ofnatural anticoagulants (eg, protein C, protein S,and TFPI), increased levels of procoagulant pro-teins (eg, factor VIII and von Willebrand factor),lupus anticoagulant activity, and increased levelsof fibrinolytic inhibitors (eg, plasminogen activa-tor inhibitor-1 and thrombin activatable fibrinoly-sis inhibitor) surely contributed to the originaldevelopment of thrombosis in this population.Nonetheless, the likely ongoing hypercoagulabilitywas adequately neutralized by once-daily enoxa-parin at 1.0 mg/kg and 1.5 mg/kg as reflected bylow VTE recurrence rates.

Ongoing and intermittent thrombocytopenia isoften seen in patients with cancer involving thebone marrow and in those receiving cycles ofcytotoxic therapy. The development of thrombocy-topenia during daily LMWH therapy may raisereasonable concerns about HIT. For this reason,we evaluated patients for the development ofheparin-associated antibodies. The detection ofsuch antibodies in 31 enoxaparin-treated patientswas not unexpected, because LMWH therapy mayresult in the generation of these antibodies.However, none of the detected antibodies werefunctional as determined by serotonin releaseassay. It seems prudent to recommend againstmaking the diagnosis of HIT in active cancerpatients receiving long-term enoxaparin solelyfrom the detection of heparin-associated antibod-ies by ELISA.


TABLE 7. Baseline Coagulation Studies

Laboratory Testa Normal Controls ONCENOX Subjects (range) % Abnormal

APC-ratioFirst generation > 2.0 2.2 ± 0.3 (1.3-2.7) 13Second generation > 2.0 2.6 ± 0.2 (1.8-3.2) 1

Antithrombin activity (%) 95.0 ± 9.5 104.4 ± 21.1 (47-154) 6Protein C activity (%) 96.0 ± 8.9 82.5 ± 41.4 (4-200) 36Protein S activity (%) 95.0 ± 8.3 69.3 ± 33.1 (18-172) 56vWF antigen (%) 135.0 ± 23.8 224.0 ± 83.0 (75.4-556.7) 71Factor VIII activity (%) 94.0 ± 7.8 111.0 ± 42.8 (1.0->200) 17dRVVT 33.7 ± 8.6 58.2 ± 15.6 (23.9-142.4) 36PAI-1 antigen (ng/mL) 32.1 ± 9.6 44.8 ± 19.6 (3.4-93.5) 34TAFI activity (%) 90.0 ± 4.7 93.9 ± 15.9 (44.2-139.8) 24TFPI antigen (ng/mL) 68 ± 13.7 74.0 ± 6.1 (62.5-92.5) 27

APC = activated protein C; vWF = von Willebrand factor; dRVVT = dilute Russell’s viper venom time; PAI-1 = plasminogen activa-tor inhibitor-1; TAFI = thrombin activatable fibrinolysis inhibitor.a. Data presented as mean ± standard deviation (range).



CONCLUSION

This study demonstrated that treatment withenoxaparin was feasible, generally well tolerated,and effective for a 180-day period in the secondaryprevention of VTE in patients with active cancer.

ACKNOWLEDGMENT

This clinical trial was sponsored by AventisPharmaceuticals.

APPENDIX 1.ONCENOX INVESTIGATORS

Wendy Breyer, Central Utah Medical Clinic, Provo, UTSteven R. Deitcher, Cleveland Clinic Foundation,

Cleveland, OHSteven Deitelzweig, Oshsner Foundation Hospital, New

Orleans, LAPaul Derderian, Community Cancer Care Specialists,

Mount Clemens, MIJames Fiorica, H. Lee Moffitt Cancer Center, Tampa, FLCharles W. Francis, University of Rochester Medical

Center, Rochester, NYDavid Hanson, Mary Bird Perkins Cancer Center, Baton

Rouge, LAMark Keaton, South Georgia Medical Center, Valdosta, GACraig Kessler, Georgetown University Medical Center,

Washington, DCPam Khosla, Rush Presbyterian St. Luke’s Hospital,

Chicago, ILManish Kohli, University of Arkansas, Little Rock, ARKapisthalam Kumar, Pasco Hernando Oncology, New

Port Richey, FLEdward Libby, University of New Mexico, Albuquerque,

NMEduardo Lim, West Suburban Center for Cancer Care,

River Forest, ILRoger Lyons, US Oncology, San Antonio, TXRobert McCroskey, Rainier Oncology Professional

Services, Puyallup, WARakesh P. Mehta, Scott & White Memorial Hospital and

Clinic, Temple, TXGeno Merli, Thomas Jefferson University, Philadelphia, PARobert Moss, Fountain Valley, CAJoe Muscato, Missouri Cancer Associates, Columbia, MOJohn Owen, Wake Forest University, Winston-Salem, NCAlbert Quiery, Geisinger Medical Center, Danville, PAJames R. Rigas, Dartmouth Hitchcock Medical Center,

Lebanon, NHWilliam R. Robinson, Don & Sybil Harrington Cancer

Center, Amarillo, TXHussain Saba, James A. Haley VA Hospital, Tampa, FLH. K. Shamasunder, Valley Tumor Medical Group,

Lancaster, CA

Brad Sherrill, Moses Cone Memorial Hospital,Greensboro, NC

Juliann Smith, Cancer and Blood Institute of the Desert,Rancho Mirage, CA

REFERENCES

1. Deitcher SR: Cancer-related deep venous thrombosis.Clinical importance, treatment challenges, and manage-ment strategies. Semin Thromb Haemost. 2003;29:247-258.

2. Luzzatto G, Schafer AI. The prethrombotic state in cancer.Semin Oncol. 1990;17:147-159.

3. Bona RD, Sivjee KY, Hickey AD, Wallace DM, Wajcs SB.The efficacy and safety of oral anticoagulation in patientswith cancer. Thromb Haemost. 1995;74:1055-1058.

4. Levine M, Gent M, Hirsch J, Leclerc J, Anderson D, WeitzJ, Ginsberg J, Turpie AG, Demers C, Kovacs M. A compar-ison of low-molecular-weight heparin administered pri-marily at home with unfractionated heparin administeredin the hospital for proximal deep-vein thrombosis. N EnglJ Med. 1996;334:677-681.

5. Merli G, Spiro TE, Olsson C-G, et al; Enoxaparin ClinicalTrial Group. Subcutaneous enoxaparin once or twice dailycompared with intravenous unfractionated heparin fortreatment of venous thromboembolic disease. Ann InternMed. 2001;134:191-202.

6. Weitz JI. Low-molecular-weight heparins. N Engl J Med.1997;337:688-698.

7. Warkentin TE, Levine MN, Hirsh J, et al. Heparin-inducedthrombocytopenia in patients treated with low-molecular-weight heparin or unfractionated heparin. N Engl J Med.1995;332:1330-1335.

8. Oken, MM, Creech, RH, Tormey, DC, et al. Toxicity andresponse criteria of the Eastern Cooperative OncologyGroup. Am J Clin Oncol. 1982;5:649-655.

9. Hyers TM, Agnelli A, Hull RD, et al. Antithrombotictherapy for venous thromboembolic disease. Chest. 2001;119(suppl):176S-193S.

10. Meyer G, Marjanovic Z, Valcke J, et al. Comparison oflow-molecular-weight heparin and warfarin for the sec-ondary prevention of venous thromboembolism inpatients with cancer: a randomized controlled study. ArchIntern Med. 2002;162:1729-1735.

11. Lee AY, Levine MN, Baker RI, et al; RandomizedComparison of Low-Molecular-Weight Heparin versus OralAnticoagulant Therapy for the Prevention of RecurrentVenous Thromboembolism in Patients with Cancer(CLOT) Investigators. Low-molecular-weight heparin ver-sus a coumarin for the prevention of recurrent venousthromboembolism in patients with cancer. N Engl J Med.2003;349:146-153.

12. Hull RD, Pineo GF, Mah AF, Brant RF, for the LITE studyInvestigators. A randomized trial evaluating long-termlow-molecular-weight heparin therapy for three monthsversus intravenous heparin followed by warfarin sodium.Blood. 2002;100:148a (abstract 556).

13. Rosendaal FR. Risk factors for venous thrombosis: prevalence,risk, and interaction. Semin Hematol. 1997;34:171-187

14. Deitcher SR, Choueiri T, Srkalovic G, Hussein MA. Acquiredactivated protein C resistance in myeloma patients withvenous thromboembolic events. Br J Haematol. 2003;123:959.

396 DEITCHER ET AL



Full Paper

Antithrombotic therapy and survival in patients with malignantdisease

AK Kakkar*,1 and F Macbeth2

1Thrombosis Research Institute and Queen Mary University of London, Emmanuel Kaye Building Manresa Road, London SW3 6LR, UK; 2NationalInstitute for Health and Clinical Excellence, MidCity Place, 71 High Holborn, London WC1V 6NA, UK

A broad range of studies suggest a two-way relationship between cancer and venous thromboembolism (VTE). Patients with cancerhave consistently been shown to be at elevated risk for VTE; this risk is partly driven by an intrinsic hypercoagulable state elicited bythe tumour itself. Conversely, thromboembolic events in patients without obvious risk factors are often the first clinical manifestationof an undiagnosed malignancy. The relationship between VTE and cancer is further supported by a number of trials and meta-analyseswhich, when taken together, strongly suggest that antithrombotic therapy can extend survival in patients with cancer by a mechanismthat extends beyond its effect in preventing VTE. Moreover, accumulating evidence from in vitro and in vivo studies has shown thattumour growth, invasion, and metastasis are governed, in part, by elements of the coagulation system. On 22 May 2009, a group ofhealth-care providers based in the United Kingdom met in London, England, to examine recent advances in cancer-associatedthrombosis and its implications for UK clinical practice. As part of the discussion, attendees evaluated evidence for and against aneffect of antithrombotic therapy on survival in cancer. This paper includes a summary of the data presented at the meeting andexplores potential mechanisms by which antithrombotic agents might exert antitumour effects. The summary is followed by aconsensus statement developed by the group.British Journal of Cancer (2010) 102, S24 – S29. doi:10.1038/sj.bjc.6605602 www.bjcancer.com& 2010 Cancer Research UK

Keywords: venous thromboembolism; thrombosis; antithrombotic therapy; cancer survival

��

As reviewed elsewhere in this supplement, patients with cancer areat increased risk of venous thromboembolism (VTE), not only as aresult of extrinsic factors (e.g., immobility, cancer treatment,surgery) but also as a result of an intrinsic hypercoagulable statecaused by the tumour itself. Conversely, it has been shown thatVTE often heralds an undiagnosed malignancy. Taken together,these data suggest that there is a two-way relationship betweencancer and VTE. This link has been clearly established bynumerous epidemiological studies; moreover, it has been shownthat antithrombotic therapy – particularly with heparins – mayextend survival in patients with cancer. Additional in vitro andin vivo data provide preliminary evidence that elements of thecoagulation system itself are critical determinants of cancergrowth, invasion, and metastasis. These data suggest that, beyondthe intuitively obvious benefit of preventing VTE in patients withcancer, antithrombotic therapies may have a direct antitumoureffect, thus prolonging life and perhaps suppressing metastasis.

On 22 May 2009, a group of health-care professionals based inthe United Kingdom met in London, England, to examine recentadvances in cancer-associated thrombosis and its implications forUK clinical practice. As part of the discussion, attendees reviewedclinical and experimental evidence suggesting an effect of anti-thrombotic therapy on survival in cancer patients. This paperincludes a summary of the data presented at the meeting andexplores potential mechanisms by which antithrombotic agents

might exert antitumour effects. The summary is followed by aconsensus statement developed by the group.

EFFECTS OF ANTITHROMBOTIC THERAPY ONSURVIVAL IN CANCER

A number of clinical trials, of varying quality, have assessed theimpact of antithrombotic therapy on cancer outcome. Whenconsidered together, the results of these studies suggest thatantithrombotic agents – including warfarin, unfractionatedheparin (UFH), and low-molecular-weight heparins (LMWH) –may prolong the survival of patients with malignant diseases.

Indirect evidence suggests that warfarin may have an impact oncancer occurrence. A prospective study randomised patients withVTE to either 6 weeks or 6 months of oral anticoagulation; patientswere questioned yearly thereafter about any newly diagnosedcancer (Schulman and Lindmarker, 2000). An analysis, conductedafter a mean follow-up of 8.1 years, found that cancer wasdiagnosed in 15.8% of patients who were treated for 6 weeks withoral anticoagulants, compared with 10.3% of those treated for6 months (OR 1.6; 95% CI 1.1– 2.4) (Schulman and Lindmarker,2000). Notably, the difference between the 6-week and 6-monthgroups was driven primarily by the occurrence of new urogenitalcancers (most commonly prostate cancer), which occurred in 6.7%of the 6-week group and in 2.8% of the 6-month group (Schulmanand Lindmarker, 2000). More recently, Tagalakis et al (2007)examined the effect of warfarin in patients with urogenital cancer.This nested, matched case–control study included 19 412 new*Correspondence: Professor AK Kakkar; E-mail: [email protected]

British Journal of Cancer (2010) 102, S24 – S29

& 2010 Cancer Research UK All rights reserved 0007 – 0920/10 $32.00

www.bjcancer.com

http://dx.doi.org/10.1038/sj.bjc.6605602




cases of urogenital cancer diagnosed over a 22-year period inCanada (Tagalakis et al, 2007). Four years of warfarin use in the5-year period immediately preceding the index date was associatedwith a rate ratio of 0.80; there was a trend towards a decreasingrate ratio for prostate cancer with increasing duration of warfarinuse through 5 years (Tagalakis et al, 2007). Although interesting,neither of these studies indicate whether warfarin therapy preventsor merely delays the onset of cancer, nor whether it affects overallmortality.

An early trial of warfarin – the VA Cooperative Study #75 –directly examined the effect of warfarin on survival in patients withcancer. In this study, 431 patients with malignancies wererandomised to either warfarin or placebo in addition to theirstandard cancer treatment (Zacharski et al, 1984). No differencesin survival were observed between treatment groups for patientswith advanced non-small-cell lung, colorectal, prostatic, or headand neck cancer (Zacharski et al, 1984). A trend towards improvedsurvival was seen in patients with non-small-cell lung cancer aftersurgical resection or potentially curative radiation therapy.Notably, warfarin was associated with a significant improvementin survival (P¼ 0.018) in a small subset of 50 patients with small-cell lung cancer (SCLC); these patients also had a significantlyincreased time to disease progression compared with controls(Zacharski et al, 1984).

A similar three-arm trial by CALGB (Chahinian et al, 1989)randomised 328 patients with extensive SCLC to a standardchemotherapy regimen (MACC), to the regimen with concurrentwarfarin, or to an alternating chemotherapy regimen. The overallresponse rate was significantly higher in the warfarin arm than inthe others (P¼ 0.012), but although there was also a trend towardsimproved survival, this did not reach statistical significance(Chahinian et al, 1989).

Another trial, published by Maurer et al (1997), also attemptedto confirm the results of the VA Cooperative Study in limited-stageSCLC patients. As in the previous study, patients were randomisedto receive warfarin or no warfarin, in addition to cancerchemotherapy and radiation therapy. There were no significantdifferences between the warfarin and no-warfarin groups in termsof response rates, survival, failure-free survival, disease-freesurvival, or patterns of relapse (Maurer et al, 1997). However,the trial was confounded by a protocol amendment after accrual of179 patients because of pulmonary toxicity with a reduction of thechemotherapy regimen from eight cycles to five cycles (Maureret al, 1997).

Only one study has reported on the effects of UFH on survival.In this study, 277 patients with SCLC were randomised to eitherreceive or not receive dose-adjusted subcutaneous UFH injectionsfor 5 weeks, in addition to chemotherapy (Lebeau et al, 1994). Inthis study, heparin was associated with improved completeresponse rates (37 vs 23%; P¼ 0.004); improved median survival(317 days vs 261 days; P¼ 0.01); and better survival rates at 1 year(40 vs 30%), 2 years (11 vs 9%), and 3 years (9 vs 6%) (Lebeau et al,1994). A subgroup analysis suggested that the effect on survivalwas only in patients with limited disease.

The effect of LMWH on survival has been more extensivelystudied. An early analysis of two studies, conducted by Green et al(1992), provided initial evidence that mortality was reduced inpatients treated with LMWH, compared with those who receivedUFH. A meta-analysis of trials comparing UFH with LMWH,published in 1996, provided additional evidence that LMWH isassociated with reduced mortality in patients with cancer (Siragusaet al, 1996). As expected, both UFH and LMWH were effective inpreventing recurrent VTE. An analysis of four studies in patientswith cancer found that mortality rates during the 16- to 90-dayfollow-up period of oral anticoagulant therapy were substantiallylower among patients assigned to the LMWH group (12%)compared with the UFH group (26%) (RR 0.33; 95% CI 0.1–0.8;P¼ 0.01) (Siragusa et al, 1996). The mortality rate in patients

without cancer was low and was not significantly different betweenthe two groups. Notably, significant reduction in mortality incancer patients who received LMWH was not observed during theinitial 15 days of treatment. Instead, the majority of deathsoccurred after ceasing LMWH or UFH, which suggests that theeffect of LMWH in preventing mortality was not because of itsantithrombotic effect (Siragusa et al, 1996). A second meta-analysis (1999) also found a striking 57% reduction in mortalitywith LMWHs compared with UFH in the small subgroup ofpatients with cancer (Gould et al, 1999).

The effect of LMWH on survival in cancer was tested directly inthe Fragmin Advanced Malignancy Outcome Study (FAMOUS)(Kakkar et al, 2004). In this study, 385 patients with advancedmalignancies (histologically confirmed, advanced stage III or IVdisease of the breast, lung, gastrointestinal tract, pancreas, liver,genitourinary tract, ovary, or uterus) were randomly assigned toreceive either dalteparin (5000 IU administered once daily) orplacebo (Kakkar et al, 2004). There were no restrictions onconcomitant use of chemotherapy or radiotherapy. Patients werefollowed for 1 year for the primary end point of mortality;secondary outcomes included objectively confirmed VTE andbleeding (Kakkar et al, 2004). Rates of symptomatic VTE were lowin both the dalteparin (2.4%) and placebo (3.3%) groups; bleedingwas seen in 4.7% of dalteparin patients and in 2.7% of placebopatients (Kakkar et al, 2004).

At 1 year after randomisation, survival estimates in thedalteparin and placebo groups were 46 and 41%, respectively(P¼ 0.19; Figure 1A). At 2 years, survival rates were 27 and 18%for the dalteparin and placebo groups, respectively, and at 3 yearsthe rates were 21 and 12%, respectively (Kakkar et al, 2004). Posthoc analysis of patients with a better prognosis who survived morethan 17 months showed a survival advantage for the dalteparingroup (Kakkar et al, 2004). At 2 and 3 years after randomisation,survival in the dalteparin and placebo groups was 78 vs 55% and60 vs 36%, respectively (P¼ 0.03; Figure 1B). Median survival inthe dalteparin group was 43.5 months, compared with 24.3 monthsin the placebo group (Kakkar et al, 2004).

The effect of LMWH on survival in patients with SCLC wasevaluated directly in a study conducted by Altinbas et al (2004). Inthis small study, 84 patients were randomised to receivecombination chemotherapy with or without dalteparin (5000 IUonce daily during the 18 weeks of combination chemotherapy)(Altinbas et al, 2004). Tumour response rates were substantiallyhigher among patients who received LMWH (69.2%) comparedwith those who did not (42.5%), but the difference between the twogroups was not statistically significant (P¼ 0.07). Median progres-sion-free survival was 10.0 and 6.0 months in the LMWH and no-LMWH groups, respectively (P¼ 0.01), with similar improvementsin survival with LMWH observed in patients with both limited andextensive disease stages. Overall, the hazard of death was reducedby 44% among patients who received LMWH (P¼ 0.012) (Altinbaset al, 2004).

A second recent trial prospectively examined the effect ofLMWH on survival in patients with advanced malignancies (Klerket al, 2005). Patients with metastatic or locally advanced solidtumours who could not be treated curatively were randomlyassigned to receive a 6-week course of weight-adjusted nadroparin(administered twice daily during the initial 14 days of treatmentand once daily thereafter for an additional 4 weeks) or placebo(Klerk et al, 2005); concomitant chemotherapy or radiotherapywas permitted. The primary end point was all-cause mortality, thesecondary end point was major and clinically relevant non-majorbleeding (Klerk et al, 2005).

At 6 months, survival was 61% among patients randomlyallocated to nadroparin, compared with 56% in the placebo group(Klerk et al, 2005). At 12 and 24 months, the corresponding valueswere 39 vs 27% and 21 vs 11% (Figure 2A) (Klerk et al, 2005).Among all patients, median survival was significantly longer in the

Antithrombotic therapy and survival

AK Kakkar and F Macbeth

S25


nadroparin group (8.0 months) than in the placebo group(6.6 months; hazard ratio 0.75; 95% CI 0.59–0.96; P¼ 0.021).After adjustment for life expectancy, WHO performance status,concomitant treatment, and type and histology of cancer, therelationship between nadroparin and improved survival remainedstatistically significant (hazard ratio, 0.76; 95% CI 0.58–0.99)(Klerk et al, 2005).

Consistent with the results seen in the FAMOUS trial, the effectof nadroparin on survival was most apparent among patients witha better prognosis at enrollment (defined as an estimated lifeexpectancy of X6 months; Figure 2B) (Klerk et al, 2005). In thisgroup, the hazard ratio was 0.64 (95% CI 0.45– 0.90; P¼ 0.010),compared with 0.88 (95% CI 0.62–1.25) in those with a lifeexpectancy of o6 months. Median survival in the good-prognosisgroup was 15.4 months and 9.4 months for the nadroparin andplacebo groups, respectively. There was no significant difference inthe rate of major bleeding between the nadroparin group (fiveevents) and the placebo group (one event; P¼ 0.12) (Klerk et al,2005).

In contrast to these data, a trial conducted by Sideras et al(2006) found that LMWH did not influence survival times inpatients with advanced cancer. This small study, including 141participants, initially randomised patients with advanced cancerto treatment with LMWH or saline. Because of low accrual, theplacebo injection arm was eliminated and the study became openlabel, with patients receiving either LMWH plus standard clinicalcare or standard clinical care alone (Sideras et al, 2006). Median

survival was 10.5 months in the combined standard care andplacebo groups and 7.3 months in the combined LMWH arms.When the two arms from the initial, blinded phase of the studywere examined, the median survival times were 6.2 months in theLMWH arm and 10.3 months in the placebo arm (Sideras et al,2006).

A recent systematic meta-analysis of randomised trials suggeststhat LMWH, on balance, improves overall survival in cancerpatients, including those with advanced disease (Lazo-Langneret al, 2007). This meta-analysis included four studies, enrolling 898patients with solid tumours who were randomly allocated to eitherLMWH or placebo (Lazo-Langner et al, 2007). Three studies useddalteparin (5000 IU daily) for 18 weeks, 1 year, or 2 years; one used6 weeks of weight-adjusted nadroparin, with a high dose duringthe first 2 weeks (Lazo-Langner et al, 2007).

At 1 year, the pooled results of the studies showed a 30%reduction in the hazard of death in favour of the LMWH group(P¼ 0.05; Figure 3); for patients with less advanced disease, therewas a 25% reduction in the hazard of death (P¼ 0.04) (p 731). At2 years, LMWH reduced mortality in all patients by 43% (P¼ 0.03),and by 41% (P¼ 0.004) in patients with advanced disease. LMWHconferred a statistically significant improvement in survival atboth 12 and 24 months (Lazo-Langner et al, 2007).

In summary, there is accumulating evidence that anticoagulanttherapy may increase survival in some patients with cancer.Further studies are currently ongoing to confirm the effects ofanticoagulant therapy in a range of tumour types. The largest of

Kap

lan–

Mei

er s

urvi

val

dist

ribut

ion

func

tion

estim

ate

Placebo

Placebo

Placebo

No. at risk55 31 26 22 20 13 8 5 5 5 3

8 5 23 089 9101747

17 23 29 35 41 47 53 59 65 71 77 83

No. at risk190 85 30 22 12 5 4

184 72 9 8 5 215

0 12 24 36 48 60 72 84Time to event from randomisation (months)

Time to event from randomisation (months)

Delteparin

Delteparin

Delteparin

Placebo

Delteparin

Median survival (months) and its 95% CIDalteparin: 10.8. (9.27, 13.5)Placebo: 9.14 (7.33, 11.5)

Median survival (months) and its 95% CIDalteparin: 43.5 (33, 52.3)Placebo: 24.3 (22.4, 41.5)

P =0.19

P =0.03

1.00.90.80.70.60.50.40.30.20.10.0

Kap

lan–

Mei

er s

urvi

val

dist

ribut

ion

func

tion

estim

ate

1.00.90.80.70.60.50.40.30.20.10.0

Figure 1 (A) Survival curves for the intent-to-treat population enrolled in the FAMOUS trial; (B) survival curves for the subgroup of patients with betterprognosis who survived beyond 17 months after randomisation (Kakkar et al, 2004).



S26


these is the FRAGMATIC (FRAGMin Added to standard TherapyIn patients with lung Cancer) trial (Wales Cancer Trials Unit,2009), which is currently enrolling patients with pathologicallyconfirmed lung cancer (of all histological types and all stages) who

will be randomly allocated to standard care or standard care plus6 months of treatment with dalteparin. This trial aims to recruit 2200patients and is powered to detect a 5% increase in 1-year survival.The GASTRANOX Study (ClinicalTrials.gov, 2009) is enrolling up to

Nadroparin

Nadroparin

Placebo

Placebo

No. at riskNadroparinPlacebo

Months after randomisation

148 51 23 15 7 4 3154 36 12 4 3 2 1

0 12 24 36 48 60 72 84 96

No. at riskNadroparinPlacebo

Months after randomisation

79 41 20 12 5 4 385 31 11 3 3 2 1

0 12 24 36 48 60 72 84 96

P=0.021

P=0.010

Pro

babi

lity

of s

urvi

val

1.0

0.8

0.6

0.4

0.2

0

Pro

babi

lity

of s

urvi

val

1.0

0.8

0.6

0.4

0.2

0

Figure 2 (A) Probability of survival in all patients with advanced solid malignancy according to assignment to nadroparin or placebo; (B) probability ofsurvival in patients with advanced solid malignancy with a life expectancy of X6 months at enrollment, according to assignment to nadroparin or placebo(Klerk et al, 2005).

Patients with limited and advanceddisease or (random)

95% CI

One-year mortality

Two-year mortality

Altinbas (2004)

Altinbas (2004)

Kakkar (2004)

Kakkar (2004)

Klerk (2005)

Klerk (2005)

Sideras (2006)

Sideras (2006)

0.70 (0.49, 1.00) 0.75 (0.57, 0.99)P=0.04

0.59 (0.42, 0.84)P=0.004

0.57 (0.34, 0.96)

P=0.05

P=0.03

0.01 0.1 1 10 100 0.01 0.1 1 10 100No LMWH betterLMWH betterNo LMWH betterLMWH better

Patients with advanceddisease or (random)

95% CI

Figure 3 One- and 2-year mortality in cancer patients randomised to LMWH vs placebo/no intervention (Lazo-Langner et al, 2007).



S27


740 patients with advanced gastric cancer who will be randomlyallocated to LMWH enoxaparin (1 mg kg�1 day�1) for 6 months withchemotherapy or chemotherapy alone. The primary end point of thetrial is the composite of all-cause mortality and symptomatic VTE.

POTENTIAL MECHANISMS FOR THE EFFECT OFANTITHROMBOTIC THERAPY ON SURVIVAL INCANCER

The results of individual clinical trials and meta-analyses, onbalance, seem to indicate that antithrombotic therapy – andparticularly LMWH – has an effect on survival in patients withcancer. Although this observation is relatively consistent acrossstudies, it is difficult to understand how a short course of LMWHcan provide a substantial survival advantage in patients withcancer. Recent data suggest that antithrombotic therapy may havea direct tumour biology-modifying effect.

A large number of circulating proteins, usually in inactive form,are involved in the haemostatic cascade. The activation of thesefactors culminates in the formation of the fibrin network of theclot and is countered by the fibrinolytic cascade, which, whenactivated, results in thrombus degradation. Cancer itself seems toelicit a systemic hypercoagulable state. Production of a wide rangeof procoagulant molecules, including tissue factor (TF) and cancerprocoagulant (CP), a cysteine protease growth factor, has beendemonstrated in patients with cancer, as well as increased levels ofprocoagulant markers, including TF, activated factor VII (FVIIa),prothrombin-activation peptide, and thrombin–antithrombincomplexes (a comprehensive review can be found in Petraliaet al (2005). Recent evidence suggests that the production ofprocoagulant molecules in patients with cancer is more than amere side effect of cancer growth; instead, data suggest that thesefactors have an integral role in malignancy through elicitingtumour growth, invasion, metastasis, and angiogenesis. These dataprovide a potential link between the use of anticoagulant therapiesand improved survival in patients with cancer.

Tissue factor, a cell-surface bound, transmembrane glycoprotein,has been shown to be expressed on a variety of tumours derived fromthe epithelium. Tissue factor interacts with FVIIa to form the TF–VIIacomplex – the primary activator of coagulation (Petralia et al, 2005).This complex seems to have a role in cell adhesion and migrationthrough the recruitment of actin-binding protein 280 (filamin A), anintracellular protein implicated in cell motility (Ott et al, 1998).In vitro, immobilised ligands for TF specifically support cell adhesion,spreading, and intracellular signalling, suggesting that the interactionbetween the cytoplasmic domain and filamin A may support tumourcell metastasis and vascular remodelling. Cancer procoagulant is alsoexpressed by a wide range of tumours. It has the ability to initiatethe haemostatic cascade directly by activating FX independently ofthe TF–VIIa complex.

Tissue factor expression has been shown to dramatically changetumour behaviour. In animal models, overexpression of TF incancer cells has been shown to significantly upregulate over 40genes and downregulate nearly 230 genes involved in transcrip-tion, translation, intercellular signalling, cell growth, and apoptosis(Wang et al, 2004). Tissue factor expression correlates withhistological grade and heralds the transformation from benign tomalignant phenotype (Kakkar et al, 1995). Overexpression of TF inexperimental models of pancreatic adenocarcinoma enhancesin vitro invasion and primary tumour growth (Kakkar et al,1999). Similarly, overexpression in a mouse sarcoma model resultedin increased levels of proangiogenic vascular endothelial growthfactor (VEGF) and suppression of thrombospondin, an antiangio-genic regulatory protein (Zhang et al, 1994). In human breastcancer, TF expression has been shown to correlate with an invasivephenotype and initiation of angiogenesis (Contrino et al, 1996).Tissue factor expression in hepatocellular cancer is strongly

associated with VEGF-induced angiogenesis and venous invasion(Poon et al, 2003), potentially mediated by an interaction withintegrin aIIIbI. Moreover, TF-expressing cells seem to be protectedfrom apoptosis induced by serum deprivation and loss of adhesion,suggesting a potential mechanism by which TF may promotemetastasis (Versteeg et al, 2004).

These data are consistent with the known role of TF inenhancing wound healing, in which it indirectly induces prolifera-tion of human vascular endothelial cells and promotes endothelialcell alignment through the production of thrombin (Carney et al,1992; Haralabopoulos et al, 1997). Protease-activated receptor-1(PAR-1) is a receptor for thrombin that is overexpressed in a rangeof tumour cell lines, particularly metastatic cell lines. Thrombin–PAR signalling has been shown to upregulate expression of TF andurokinase plasminogen in prostate cancer, increase invasiveness ofbreast and pancreatic cell lines, and enhance procoagulant activityin colon cancer (Petralia et al, 2005).

Together, it is clear that there is a role for elements ofthe haemostatic system in cancer progression that extendsbeyond their role in fibrin generation. These data also suggestmechanisms through which antithrombotic treatments, such asLMWH, may have a direct impact on tumour phenotype, as wellas influence survival in patients with cancer, beyond their effectin suppressing VTE.

CONSENSUS STATEMENT

Although the association between cancer and an increased risk ofthromboembolism is well understood, the evidence that anti-coagulant therapy may improve survival in cancer patients hasbeen slow to accumulate. The results of studies have been quiteheterogeneous in both the patient population and the anti-coagulant used. Many of the studies have also been underpoweredto show clinically significant improvements in survival. However,it does seem that there is a beneficial effect even from quite shortcourses of treatment, and it seems that heparins may be moreeffective than coumarins, and that this effect may be more markedin patients with better prognosis. The beneficial effects do notseem to come at the expense of a significant risk of adverse effects,haemorrhage in particular.

Together, these findings suggest that, although there may be ashort-term benefit in preventing thromboembolism in a group ofpatients at increased risk, there may also be a longer-term effect onthe cancer itself. This hypothesis is supported by a number ofin vitro and animal studies indicating a link between key factors inthe coagulation cascade and tumour growth and metastasis.

In addition to continuing investigation of the detailed mecha-nisms of the complex interactions between growing tumour cellsand the coagulation cascade, there are a number of importantquestions to be answered before routine anticoagulant therapy isintegrated into cancer therapy:

� How large and how consistent is any survival benefit?� Which patients (tumour type, stage, prognostic category) are

most likely to benefit from therapy?� Which anticoagulant is most effective?� What duration of treatment is needed?� How large is the risk of significant adverse effects?

We would therefore urge participation in the ongoing clinicaltrials and in planning of future studies to make the most of thisfascinating and potentially important area of cancer treatment.

Conflict of interest

AK Kakkar has received consulting fees from Bayer, Sanofi-aventis,Boehringer Ingelheim, Pfizer, Bristol-Myers Squibb, and Eisai.F Macbeth has received grant support from CRUK and Pfizer.



S28


REFERENCES

Altinbas M, Coskun HS, Er O, Ozkan M, Eser B, Unal A, Cetin M, Soyuer S(2004) A randomized clinical trial of combination chemotherapy withand without low-molecular-weight heparin in small cell lung cancer.J Thromb Haemost 2: 1266 – 1271

Carney DH, Mann R, Redin WR, Pernia SD, Berry D, Heggers JP,Hayward PG, Robson MC, Christie J, Annable C (1992) Enhancementof incisional wound healing and neovascularization in normal rats bythrombin and synthetic thrombin receptor-activating peptides. J ClinInvest 89: 1469 – 1477

Chahinian AP, Propert KJ, Ware JH, Zimmer B, Perry MC, Hirsh V,Skarin A, Kopel S, Holland JF, Comis RL (1989) A randomized trial ofanticoagulation with warfarin and of alternating chemotherapy inextensive small-cell lung cancer by the Cancer and Leukemia Group B.J Clin Oncol 7: 993 – 1002

ClinicalTrials.gov. Overall survival of inoperable gastric/gastrooeso-phageal cancer subjects on treating with LMWH + chemotherapy (CT)versus standard CT (GASTRANOX). http://clinicaltrials.gov/ct2/show/NCT007183540. Accessed 22 September 2009

Contrino J, Hair G, Kreutzer DL, Rickles FR (1996) In situ detection oftissue factor in vascular endothelial cells: correlation with the malignantphenotype of human breast disease. Nat Med 2: 209 – 215

Gould MK, Dembitzer AD, Doyle RL, Hastie TJ, Garber AM (1999) Low-molecular-weight heparins compared with unfractionated heparin fortreatment of acute deep venous thrombosis. A meta-analysis ofrandomized, controlled trials. Ann Intern Med 130: 800 – 809

Green D, Hull RD, Brant R, Pineo GF (1992) Lower mortality in cancerpatients treated with low-molecular-weight versus standard heparin.Lancet 339: 1476

Haralabopoulos GC, Grant DS, Kleinman HK, Maragoudakis ME (1997)Thrombin promotes endothelial cell alignment in Matrigel in vitro andangiogenesis in vivo. Am J Physiol 273: C239 – C245

Kakkar AK, Chinswangwatanakul V, Lemoine NR, Tebbut S, WilliamsonRC (1999) Role of tissue factor expression on tumour cell invasionand growth of experimental pancreatic adenocarcinoma. Br J Surg 86:890 – 894

Kakkar AK, Lemoine NR, Scully MF, Tebbutt S, Williamson RC (1995)Tissue factor expression correlates with histological grade in humanpancreatic cancer. Br J Surg 82: 1101 – 1104

Kakkar AK, Levine MN, Kadziola Z, Lemoine NR, Low V, Patel HK,Rustin G, Thomas M, Quigley M, Williamson RC (2004) Low molecularweight heparin, therapy with dalteparin, and survival in advancedcancer: the fragmin advanced malignancy outcome study (FAMOUS).J Clin Oncol 22: 1944 – 1948

Klerk CP, Smorenburg SM, Otten HM, Lensing AW, Prins MH, Piovella F,Prandoni P, Bos MM, Richel DJ, van Tienhoven G, Buller HR (2005) Theeffect of low molecular weight heparin on survival in patients withadvanced malignancy. J Clin Oncol 23: 2130 – 2135

Lazo-Langner A, Goss GD, Spaans JN, Rodger MA (2007) The effect of low-molecular-weight heparin on cancer survival. A systematic review andmeta-analysis of randomized trials. J Thromb Haemost 5: 729 – 737

Lebeau B, Chastang C, Brechot JM, Capron F, Dautzenberg B, DelaisementsC, Mornet M, Brun J, Hurdebourcq JP, Lemarie E, for ‘Petites Cellules’Group (1994) Subcutaneous heparin treatment increases survival insmall cell lung cancer. Cancer 4: 38 – 45

Maurer LH, Herndon II JE, Hollis DR, Aisner J, Carey RW, Skarin AT,Perry MC, Eaton WL, Zacharski LL, Hammond S, Green MR (1997)Randomized trial of chemotherapy and radiation therapy with or withoutwarfarin for limited-stage small-cell lung cancer: a Cancer and LeukemiaGroup B study. J Clin Oncol 15: 3378 – 3387

Ott I, Fischer EG, Miyagi Y, Mueller BM, Ruf W (1998) A role for tissuefactor in cell adhesion and migration mediated by interaction with actin-binding protein 280. J Cell Biol 140: 1241 – 1253

Petralia GA, Lemoine NR, Kakkar AK (2005) Mechanisms of disease: theimpact of antithrombotic therapy in cancer patients. Nat Clin PractOncol 2: 356 – 363

Poon RT, Lau CP, Ho JW, Yu WC, Fan ST, Wong J (2003) Tissue factorexpression correlates with tumor angiogenesis and invasiveness inhuman hepatocellular carcinoma. Clin Cancer Res 9: 5339 – 5345

Schulman S, Lindmarker P (2000) Incidence of cancer after prophylaxiswith warfarin against recurrent venous thromboembolism. N Engl J Med342: 1953 – 1958

Sideras K, Schaefer PL, Okuno SH, Sloan JA, Kutteh L, Fitch TR, Dakhil SR,Levitt R, Alberts SR, Morton RF, Rowland KM, Novotny PJ, Loprinzi CL(2006) Low-molecular-weight heparin in patients with advanced cancer:a phase 3 clinical trial. Mayo Clin Proc 81: 758 – 767

Siragusa S, Cosmi B, Piovella F, Hirsh J, Ginsberg JS (1996) Low-molecular-weight heparins and unfractionated heparin in the treatment of patientswith acute venous thromboembolism: results of a meta-analysis. Am JMed 100: 269 – 277

Tagalakis V, Tamim H, Blostein M, Collet JP, Hanley JA, Kahn SR (2007)Use of warfarin and risk of urogenital cancer: a population-based, nestedcase-control study. Lancet Oncol 8: 395 – 402

Versteeg HH, Spek CA, Richel DJ, Peppelenbosch MP (2004) Coagulationfactors VIIa and Xa inhibit apoptosis and anoikis. Oncogene 23: 410 – 417

Wales Cancer Trials Unit. FRAGMATIC: a randomised phase III clinicaltrial investigating the effect of FRAGMin Added to standard Therapy Inpatients with lung Cancer. http://www.wctu.org.uk/fragh.htm. Accessed22 September 2009

Wang X, Wang M, Amarzguioui M, Liu F, Fodstad O, Prydz H (2004)Downregulation of tissue factor by RNA interference in humanmelanoma LOX-L cells reduces pulmonary metastasis in nude mice.Int J Cancer 112: 994 – 1002

Zacharski LR, Henderson WG, Rickles RR, Forman WB, Cornell Jr CJ,Forcier AJ, Edwards RL, Headley E, Kim S-H, O’Donnell JF, O’Dell R,Tornyos K, Kwaan HC (1984) Effect of warfarin anticoagulation onsurvival in carcinoma of the lung, colon, head and neck, and prostate.Final report of VA Cooperative Study #75. Cancer 53: 2046 – 2052

Zhang Y, Deng Y, Luther T, Muller M, Ziegler R, Waldherr R, Stern DM,Nawroth PP (1994) Tissue factor controls the balance of angiogenic andantiangiogenic properties of tumor cells in mice. J Clin Invest 94: 1320 – 1327



S29


http://clinicaltrials.gov/ct2/show/NCT00718354

http://clinicaltrials.gov/ct2/show/NCT00718354

http://www.wctu.org.uk/fragh.htm

Pharmacoeconomics 2006; 24 (6): 593-607ORIGINAL RESEARCH ARTICLE 1170-7690/06/0006-0593/$39.95/0

2006 Adis Data Information BV. All rights reserved.

Dalteparin versus Warfarin for thePrevention of Recurrent VenousThromboembolic Events inCancer PatientsA Pharmacoeconomic Analysis

George Dranitsaris,1 Mark Vincent2 and Mark Crowther3

1 Augmentium Pharma Consulting, Toronto, Ontario, Canada2 London Regional Cancer Centre, London, Ontario, Canada3 St Joseph’s Hospital, Hamilton, Ontario, Canada

Objective: In a recent randomised trial (CLOT [Comparison of Low molecularAbstractweight heparin versus Oral anticoagulant Therapy for long term anticoagulation incancer patients with venous thromboembolism]), which evaluated secondaryprophylaxis of venous thromboembolism (VTE) in cancer patients, dalteparinreduced the relative risk of recurrent VTEs by 52% compared with oral anticoagu-lation therapy (p = 0.002). A Canadian pharmacoeconomic analysis was conduct-ed to measure the economic value of dalteparin for this indication.Design: The study was conducted from the Canadian healthcare system. The firstpart of this study utilised the CLOT trial database, from which resource utilisationdata were converted into Canadian cost estimates ($Can, year 2005 values).Univariate and multivariate regression analyses were conducted to compare thetotal cost of therapy between patients randomised to treatment with dalteparin ororal therapy. Health state utilities and treatment preferences were then measuredin 24 oncology care providers using the time trade-off technique.Results: When all of the cost components were combined for the entire popula-tion (n = 676), patients in the dalteparin group had significantly higher overallcosts than the control group ($Can4162 vs $Can2003; p < 0.001). The preferenceassessment revealed that 23 of 24 respondents (96%) selected dalteparin overwarfarin, with an associated gain of 0.157 QALYs. When the incremental cost ofdalteparin ($Can2159 per patient) was combined with the QALY gain, thefindings revealed that dalteparin was associated with a cost of approximately$Can13 800 (95% CI 12 400, 15 100) per QALY gained.Conclusions: Given the practical advantages of dalteparin in terms of conve-nience, improved efficacy and the acceptable economic value, this analysis

594 Dranitsaris et al.

suggests that long-term dalteparin therapy is a sound alternative to warfarin for theprevention of recurrent VTEs in patients with cancer.

Over the past decade, considerable cancer sup- treatment can be problematic because of the poten-portive care research has been devoted to febrile tial for drug-drug interactions, and unpredictableneutropenia, emesis and anaemia. In contrast, other anticoagulation levels secondary to liver dysfunc-adverse consequences of cancer such as venous tion and gastrointestinal dysfunction. The resultingthromboembolism (VTE), manifesting as deep vein poor quality anticoagulant care can contribute tothrombosis (DVT) and pulmonary embolism (PE), both bleeding and recurrent VTE.[2] In addition,has received less attention. Compared with non- there is a need to closely monitor patients receivingcancer patients, the incidence of VTE has been oral anticoagulation therapy; laboratory monitoringincreasing in cancer patients over the past 10 years is not only costly but also requires patients to under-and the risk of a recurrent DVT and subsequent PE take additional hospital or clinic visits.remains elevated, which may be related to more Dalteparin is a low-molecular-weight heparinaggressive chemotherapy and improvements in (LMWH) that has been used for many years for theoverall survival.[1,2] In one large epidemiological treatment and prevention of VTE. The advantage ofstudy, Sallah et al.[3] evaluated VTE in 1041 patients dalteparin over warfarin is the reduced variability inwith solid tumours admitted to three major medical the anticoagulation response, eliminating the needcentres. The investigators identified 81 patients with for laboratory monitoring.VTE for an overall prevalence of 7.8% (95% CI 6.2,

Dalteparin has been shown to be more effective9.4), or one event per 12.8 patients.

than oral anticoagulation therapy for the secondaryThis high incidence is of concern to oncologists

prophylaxis of VTE in patients with cancer, with anand hospital administrators because the North

overall relative risk reduction of approximatelyAmerican and European populations are aging and

52%.[5] Furthermore, evidence suggests a survivalcancer is a disease that usually occurs later in life.[4]

benefit in cancer patients treated with LMWH ver-From the perspective of a cancer treatment centre,

sus a non-LMWH control group.[6-8]

these changing demographics have resulted in in-One of the potential barriers to the use ofcreases in the number of patients seeking treatment.

dalteparin for secondary VTE prophylaxis in cancerAn inability to adequately meet these demandspatients is its high acquisition cost compared withtranslates into longer patient waiting times and, po-warfarin, and the need for subcutaneous administra-tentially, suboptimal care.tion. However, other characteristics of dalteparin –One approach to help oncology centres meet thissuch as improved efficacy, a reduced need for pa-increased volume would be to identify existing sup-tient monitoring and a lower clinical burden –portive care agents that provide superior efficacy toshould be considered when discussing the cost ofpatients, versus traditional agents, thus reducing un-therapy. These factors were considered in previousscheduled clinic visits by preventing predictableeconomic evaluations of LMWHs in patients such ascomplications of cancer or its therapy. Thus, in thethose undergoing orthopaedic surgery.[9-11]

case of VTE, prolonged warfarin prophylaxis istypically offered to reduce the risk of secondary Therefore, there are two questions that formularyevents such as DVT and PE. However, warfarin committees have to consider:

2006 Adis Data Information BV. All rights reserved. Pharmacoeconomics 2006; 24 (6)

Cost Effectiveness of Dalteparin for Preventing VTEs in Cancer 595

1. Does the use of dalteparin for secondary prophy- reach statistical significance (dalteparin = 14%, orallaxis of VTE in cancer patients provide good eco- therapy = 19%; p = 0.09).nomic value when all of the clinical, economic and In a post hoc analysis that evaluated variouspatient factors are quantified? patient subgroups, the occurrence of death at 1 year2. How does the economic value of dalteparin com- in patients with no metastatic disease was 20% in thepare with other drugs currently used in cancer sup- dalteparin group compared with 36% in the oralportive care? anticoagulant group (hazard ratio = 0.50; p =

0.03).[6]In this study, a pharmacoeconomic analysis wasconducted to determine whether dalteparin is an Based on the CLOT trial,[5] dalteparin was ap-economically reasonable alternative to warfarin proved by the Canadian regulatory authority as anfrom the perspective of the publicly funded Canadi- acceptable agent for long-term secondary prophy-an healthcare system. laxis of DVT in cancer patients.

Extraction of Resource Utilisation InformationMethodsfrom the CLOT Trial Database

The CLOT study database contained healthcareClinical Trialresource use data for 338 patients randomised into

The clinical trial data for the pharmacoeconomic each group.[5] In the original trial, two patients fromevaluation were obtained from a multicentre, each group were excluded because they did not haverandomised, nonblind Canadian trial (CLOT [Com- a qualifying thrombotic event. In the current analy-parison of Low molecular weight heparin versus sis, all randomised patients were included. AlthoughOral anticoagulant Therapy for long term anticoagu- the original CLOT trial was not designed to formallylation in cancer patients with venous thromboembo- collect healthcare resource data for an economiclism]), which compared dalteparin with oral antico- evaluation, important healthcare resource itemsagulation therapy for the secondary prevention of were collected about each patient on the followingVTEs in cancer patients.[5] In that study, Lee et al.[5] variables: (i) dosage and duration of therapy ofrandomised 676 cancer patients with a newly diag- study drugs; (ii) routine laboratory tests; (iii) patientnosed VTE to receive 6 months’ prophylaxis with international normalisation ratio (INR) in the case ofsubcutaneous dalteparin (200 IU/kg once daily in warfarin; (iv) patient telephone contact; (v) un-the first month, then 150 IU/kg once daily from scheduled clinical visits; (vi) diagnostic tests rele-months 2 to 6) or 7 days of dalteparin followed by an vant for VTE; (vii) blood transfusions; and (viii) theoral anticoagulant. occurrence of bleeding-related events that were pos-

sibly or probably due to the study drugs.During the study period, 27 of 336 (8.0%) pa-tients evaluated in the dalteparin group developed There were no direct hospital length of stay orrecurrent VTE compared with 53 of 336 (15.8%) individualised treatment data for the management ofpatients treated with warfarin (overall 52% relative patients who actually developed recurrent VTEs.risk reduction, p = 0.002). In this study, there were Therefore, the Canadian and international literatureno significant differences in major bleeding events was used to obtain cost estimates for treating a DVT,between groups (dalteparin = 6%, oral therapy = PE, fatal PE as well as the following adverse events:4%; p = 0.27).[5] Overall rates of bleeding were heparin-induced thrombocytopenia, haematochezia,higher with warfarin, but the difference did not haematoma, haematuria, haematemesis, maleana,



retroperitoneal bleeding, intracranial bleeding and Treatment Preferences andHealth-State Utilitiesintraperitoneal bleeding.[12-14] Only events of grade

III or higher severity were included in the economicThe health-related quality-of-life (HR-QOL) val-analysis (table I). Unless specified, it was assumed

ues measured in the analysis were patient prefer-that patients developing DVTs were treated in theences for alternative health states. The two healthoutpatient setting.states were long-term secondary treatment with war-

Unit costs were obtained from the University farin for 6 months to prevent a recurrent VTE, orHealth Network and the Toronto Sunnybrook Re- dalteparin for the same indication and duration. Ingional Cancer Center in Toronto (table I). All costs the current study, QALYs were measured asin the current study were reported in 2005 $Can, and ‘healthy months equivalence’ for the time spent incost estimates from previous years were converted each health state.[17,18] The scores in months wereto 2005 dollars using the consumer price index for then converted to utility measures between 0 and 1,healthcare as reported by Statistics Canada.[16] The where 0 represented death and 1 was a state ofprotocol evaluated in the pivotal randomised trial perfect health or optimal quality of life (QOL).and used in the current analysis was subcutaneous Gains in healthy month equivalence were also con-dalteparin 200 IU/kg once daily in the first month, verted into QALYs by dividing by 12 months.then 150 IU/kg once daily from months 2 to 6. The The ideal population for measuring health-statedaily cost of supplies (e.g. syringes) for dalteparin utilities and treatment preferences would have beenadministration was also included in the analysis. cancer patients who had already experienced a VTE

and were about to receive extended prophylaxis.Care was used to ensure the patient demographic,However, in this study, a patient surrogate groupclinical and associated outcomes data had beenconsisting of 24 oncology nurses and pharmacistsproperly linked with the appropriate costs in allwith front line clinical experience was used. There israndomised patients. The final outcome of this pro-evidence in the oncology literature to suggest thatcess allowed an overall estimate of resource use andnurses and clinical pharmacists are suitable patientcost between groups.surrogates for objective outcomes, and that utilityestimates derived from nurses and pharmacists do

Design of Pharmacoeconomic Analysisnot substantially alter the findings of cost-utilitystudies compared with estimates reported by pa-

When the relevant healthcare resources were ex- tients.[19,20] Furthermore, we ensured that the surro-tracted from the CLOT database (see the Extraction gate sample had clinical experience in cancer sup-of Resource Utilisation Information from the CLOT portive care, in which VTE management representsTrial Database section), unit cost estimates were an important component. With a sample of 24 re-applied to each component. A univariate analysis on spondents, healthy months equivalence was mea-the overall cost was initially conducted to determine sured with a precision of ±1.0 month, with a 95%the magnitude of the cost difference between pa- probability.tients treated with dalteparin relative to those who After informed consent was obtained, each pa-received oral anticoagulation therapy. This cost dif- tient surrogate (i.e. healthcare provider) was inter-ference was then used in the subsequent cost-utility viewed. Respondents were presented with informa-analysis to estimate the incremental cost per QALY tion about the natural history of VTEs in cancer,gained with dalteparin. followed by a description of the warfarin and



Table I. Unit costs used in the economic analysis ($Can, year 2005 values)

Parameter Cost estimate Source

Drugs

Dalteparin 22.05/daya Pfizer, Canada

Warfarin (30 days) 17.36 Outpatient pharmacy, UHN

Diagnostic tests

CBC with differential 10.95 each Biochemistry Department, UHN

PT/APTT 7.50 each Biochemistry Department, UHN

Liver function tests (AST, ALT, ALP, GGTP) 5.17 each Biochemistry Department, UHN

Serum creatinine 5.10 each Biochemistry Department, UHN

Electrolytes (K, Na, Cl) 5.10 total Biochemistry Department, UHN

Urea 3.18 each Biochemistry Department, UHN

Bilirubin 8.50 each Biochemistry Department, UHN

Albumin 5.17 each Biochemistry Department, UHN

Diagnostic tests

Compression ultrasonography 83.82 each Diagnostic Imaging Department, UHN

Lung perfusion scan 521.81 each Diagnostic Imaging Department, UHN

Pulmonary angiography (four vessels) 1108.82 each Diagnostic Imaging Department, UHN

Contralateral venography 521.81 each Diagnostic Imaging Department, UHN

CT scan of the lung with contrast 87.13 each Diagnostic Imaging Department, UHN

CT scan of the lung with without contrast 116.17 each Diagnostic Imaging Department, UHN

Unscheduled patient contact

Telephone consultation 17.10 UHN and Schedule of benefits, Ontario, 2002

Clinic visit 161b

Blood transfusions

One unit of red blood cells 305 Dranitsaris[15]

Physical cost of transfusion 278

Treatment of venous thromboembolisms and other events

Fatal PE 1162 Gordois et al.[13]

Nonfatal PE 4237

Outpatient DVT management 2445

Inpatient DVT management 3473

Heparin-induced thrombocytopenia 2417

a Based on mean weight and duration data reported in the trial and using a dalteparin drug cost of $Can1.56/1000IU for the first 30days and then $Can1.976/1000IU from day 31 to 125. Drug cost differences were based on variations in drug vial sizes. Cost ofsupplies for dalteparin administration were also included in the daily cost.

b Consists of a clinic visit and a partial patient assessment.

ALP = alkaline phosphatase; ALT = alanine aminotransferase; APTT = activated partial thromboplastin time; AST = aspartateaminotransferase; CBC = complete blood count; DVT = deep vein thrombosis; GGTP = gamma glutamyl transpeptidase; PE = pulmonaryembolism; PT = prothrombin time; UHN = University Health Network.

dalteparin administration protocol. This included in- events associated with each therapy. The final pieceformation on the method of administration, monitor- of information presented to respondents was theing requirements and the associated risks and bene- results of the post hoc subgroup survival analysis,fits. During the final part of the interview, the

which suggested a potential survival benefit in pa-clinical outcomes from the CLOT trial weretients treated with dalteparin.[6] However, respon-presented. This consisted of the risk for recurrent

DVTs, fatal and non-fatal PEs, and major bleeding dents were told that this result was derived from a



post hoc subgroup analysis and the results needed to agement of cancer patients with VTEs, involvementbe confirmed in a prospective study. in guideline development, level of familiarity with

the cost of drugs used in cancer and whether or notRespondents were then asked how many monthsfamily members had developed a VTE.of ‘optimal health’ they considered being equivalent

to the time spent in each of the less than optimalhealth states described. These measures were used Cost-Utility Analysisto weigh the duration within each health state by theQOL experienced by a patient living through that The clinical, economic and QALY estimatestime period. The utility value for each health state were then used to conduct a cost-utility analysiswas based on scores from the interviews. It was

comparing dalteparin and warfarin. The primarydefined as the ratio of the equivalent time in optimal

outcome was the cost per QALY gained withhealth to the months receiving treatment with either

dalteparin, which was calculated by dividing thedalteparin or warfarin (e.g. 3 healthy monthsdifference in cost relative to warfarin therapy (nu-equivalent to 6 treatment months = 0.50). This pro-merator) by the difference in QALYs gained (de-vided a utility value in the range of 0–1, where 0nominator). Future costs and benefits were not dis-represented death and 1 was a state of optimalcounted, because of the short time periods involved.health.However, the stability of the baseline results wasPrinted interview tools were used to facilitate thetested through a sensitivity analysis. This procedureparticipant’s understanding of the time trade-offincluded re-analysing the data using the upper andtechnique.[17,18] Demographic data were also collect-lower 95% confidence intervals for costs and healthed from each participant and consisted of age, years

of oncology experience and experience in the man- state utilities.

Table II. Patient demographic and clinical characteristics at randomisation in the CLOT (Comparison of Low molecular weight heparinversus Oral anticoagulant Therapy for long term anticoagulation in cancer patients with venous thromboembolism) study[5]

Parameter Dalteparin (n = 338) Oral therapy (n = 338)

M/F 159/179 169/169

Median age [y] (range) 64.0 (22–86) 64.2 (28–89)

Mean weight [kg] (range) 73.6 (39–132) 74.7 (40–128)

Major illness in past 3 months (%) 51.8 59.8

ECOG score (%)

0 23.7 18.6

1 39.9 44.4

2 34.9 36.1

3 1.5 0.9

Patient was in hospital (%) 50.0 53.9

Solid tumour with metastatic disease (%) 65.9 68.6

Haematological cancer (%) 11.8 8.9

History of VTE (%) 11.5 10.6

Mean baseline INR 1.1 1.2

Mean baseline platelets (103 × mm3) 260.3 250.6

Mean baseline serum creatinine (mg/dL) 0.84 1.0ECOG = European Cooperative Oncology Group Performance scale; F = female; INR = international normalisation ratio; M = male;VTE = venous thromboembolism.



Table III. Comparison of healthcare resource utilisation; data from the CLOT (Comparison of Low molecular weight heparin versus Oralanticoagulant Therapy for long term anticoagulation in cancer patients with venous thromboembolism) database[5]

Parameter (mean number/patient) Dalteparin (n = 338) Oral therapy (n = 338)

Duration of therapy (days) 126.3 116.9 a

Treatment complianceb (%) 98.2 88.7

Mean dose of dalteparin (IU/kg)

First 7 days 200.0

First 30 days 200.6 0.0

Beyond day 30 165.1 0.0

Routine laboratory monitoring

CBC 4.5 4.1

APTT 3.4 3.2

INR measurements (PT) 0.0 22.0

Sodium 4.2 3.8

Potassium 4.2 3.8

Chloride 3.9 3.4

Urea 4.1 3.7

Serum creatinine 4.2 4.1

ALT 3.9 3.4

AST 3.5 3.1

ALP 3.9 3.4

GGTP 3.4 3.0

Bilirubin 3.9 3.4

Albumin 3.7 3.2

Diagnostic tests

Compression ultrasonography 0.87 0.85

Contralateral venography 0.02 0.04

Spiral CT scan 0.11 0.10

Lung scan 0.27 0.26

Pulmonary angiography 0.01 0.03

Unscheduled patient contact

Telephone consultation 6.9 6.8

Clinic visit 1.0 1.1

Blood transfusions

Total RBC units given 91 119

Total number of transfusions of ≥II units 27 40

Mean number APTT measurements 7.0 11.0

Mean number INR measurements 0.27 1.84

a Patients in the oral therapy group received dalteparin for a mean of 8 days, as indicated in the protocol.

b p < 0.001 between treatment groups.

ALP = alkaline phosphatase; ALT = alanine aminotransferase; APTT = activated partial thromboplastin time; AST = aspartateaminotransferase; CBC = complete blood count; GGTP = gamma glutamyl transpeptidase; INR = international normalisation ratio; PT =prothrombin time; RBC = red blood cell.

Statistical Analysis pare the overall cost between groups (see Mul-

tivariate Analysis section). A paired t-test analysisDemographic data and utility estimates werewas then utilised to evaluate healthy month equiva-presented as descriptive statistics as means, medians

or proportions. The unpaired t-test was used to com- lence scores and utility estimates for dalteparin and



Table IV. Treatment-associated clinical outcomes related to resource utilisation

Parameter Dalteparin (n = 338) Oral therapy (n = 338)

DVT alone 14 37

Nonfatal PE 8 9

Fatal PE 5 7

Hospital admission rate regardless of cause* (%) 25.1 28.5

Hospital admission rate for VTE, bleeding or HIT* (%) 3.2 3.8

Total number of hospital days for VTE, bleeding or HIT 32 40

DVT = deep vein thrombosis; HIT = heparin-induced thrombocytopenia; PE = pulmonary embolism; VTE = venous thromboembolism;* p = not significant between treatment groups.

warfarin. A multivariable regression analysis was characteristics and risk factors (table II). Given thisalso conducted in an exploratory manner to compare balance between groups, it was reasonable to pro-the cost of treatment with dalteparin relative to the ceed with a comparison of overall resource usecontrol group. This included a main effects-only whereby a reduction in resource use in one groupmodel as well as an evaluation of patient subgroups relative to another would likely be due to efficacy inthrough the application of interaction effects. The reducing the recurrence of VTEs and possibly acut-off for significance for all of the statistical pro- reduction in the rate of cancer progression.cedures was at the p = 0.05 level. All of the statisti-

Healthcare resource utilisation data were collect-cal analyses were performed using Stata, release 7.0

ed as part of the CLOT trial database. The mean(Stata Corp., College Station, Texas, USA).

duration of dalteparin therapy in the experimentalgroup was 126.3 days compared with 8 days in theResultscontrol group (table III). Patients randomised to theoral anticoagulation therapy group received treat-The results of the economic analysis were basedment for a mean of 116.9 days. Compliance withon data from 676 patients randomised to receiveanticoagulation therapy was significantly higher inextended prophylaxis with dalteparin or oral antico-the dalteparin group by approximately 10%agulation therapy. As was reported in the original(p < 0.001). The utilisation of laboratory tests wasrandomised clinical trial,[5] treatment arms were we-comparable between groups, with the exception ofll balanced with respect to demographic, clinical

Table V. Comparison of costs between groups ($Can, year 2005 values)

Cost parameter (mean) Dalteparin (n = 338) Oral therapy (n = 338)

Drug therapya 2852 269

Laboratory monitoring 303 437

Diagnostic tests 253 267

Unscheduled patient contact 286 300

Blood transfusions 143 208

Treatment of major bleeding events and other complicationsb,c 97.5 92.3

VTE recurrence management 228 429

Mean cost per patientd (95% CI) 4162 (3910, 4413) 2003 (1910, 2096)

a Cost of supplies for dalteparin administration included in the daily cost.

b NCI grade III or higher that were possibly or probably related to treatment.

c Costs obtained from Hull et al.,[12] and converted into 2005 values.

d p < 0.001 between treatment groups.

NCI = National Cancer Institute; VTE = venous thromboembolism.



mission were 3.2% for dalteparin and 3.8% in theoral therapy group (p = 0.68). Overall, these ratestranslated to 32 additional hospital days fordalteparin patients and 40 days in the control group(table IV).

A cost comparison was then performed using theresource use estimates in tables III and IV alongwith the unit cost estimates found in table I. Patientsrandomised to dalteparin had higher costs for drugtherapy and for the treatment of major bleedingevents (table V). In contrast, the warfarin group hadelevated costs for laboratory monitoring, diagnostictests, unscheduled patient contact, blood transfu-sions and VTE recurrence management. When all ofthe costs were considered, patients in the dalteparingroup were approximately $Can2159 more costly totreat than patients in the control group (table V). The

Dalteparin Oral therapy

0

20

40

60

80

100 VTE treatmentMajor bleedingBlood transfusionsPatient contactDiagnostic testsLaboratory monitoringDrug therapy

Per

cent

age

of c

ontr

ibut

ion

Fig. 1. Contributors to the overall cost of dalteparin and warfarintherapy. VTE = venous thromboembolism.

largest cost contributor in the dalteparin group wasINR measurements, which are required for oral anti- drug acquisition, at 67% (vs 13% in the control),coagulation therapy. An average of 22 INRs were while VTE treatment and laboratory monitoringmeasured per patient randomised to oral anticoagu- were the largest cost components in the oral antico-

agulation therapy group (figure 1).lation therapy. Furthermore, there were more un-scheduled clinic visits (absolute number = 352 vs

Treatment Preferences and387), telephone consultations, blood transfusionsHealth-State Utilitiesand activated partial thromboplastin time/INR mea-

surements (associated with transfusions) in the war-Treatment preferences and health-state utilitiesfarin group (table III).

for each outcome were estimated from a sample ofThe occurrence of VTE-related events was then 24 oncology healthcare professionals. The mean age

extracted from the database and compared between of respondents was 44.3 years (range 24–62 years)groups (table IV). As reported in the original trial, with an average of 8.6 years of direct oncologythere was a statistically significant reduction in the experience (range 1–27 years). Furthermore, 17 ofoccurrence of VTEs in patients randomised to the 24 (70.8%) participants had direct experience in thedalteparin group. Overall, there were 23 fewer treatment of VTEs in cancer patients, with an aver-DVTs and 3 fewer PE events in the dalteparin group. age length of experience of 6.7 years. In addition,In patients who were receiving therapy in the outpa- 8.3% of the sample had been involved in the devel-tient setting, there was a hospital admission rate opment of practice guidelines/algorithms specific to(regardless of cause) of 25.1% in the dalteparin VTE management over the previous 2-year period.group compared with 28.5% in the control group Lack of drug cost knowledge could affect treatment(p = 0.33). When the cause for the admission was preferences. Respondents were asked to state theirlimited to recurrent VTE, bleeding or heparin-in- knowledge of costs for oncology drugs. The find-duced thrombocytopenia, the rates of hospital ad- ings revealed that 41.7% and 45.8% of the sample



Table VI. Cost-effectiveness and cost-utility analysis (all cost-effectiveness ratios were rounded to the nearest hundred) [$Can, year 2005values]

Parameter Outcome (95% CI)

dalteparin oral therapy difference

Utilities

Treatment preferences from respondents 23 (96) 1 (4)a

[n = 24] (%)

Mean health month equivalence 3.94 (3.32, 4.56) 2.06 (1.55, 2.58) 1.88a (1.37, 2.49)

Mean health state utilitya,b 0.66 (0.55, 0.76) 0.34 (0.26, 0.43)

Mean cost per patient 4162 (3910, 4413) 2003 (1910, 2096) 2159a (1943, 2373)

Cost-effectiveness analysisc

Recurrence of VTE (%) 8.0 15.8 7.8a

Cost per VTE avoided 27 700 (24 400, 30 400)

Cost-utility analysisc

QALY gained 0.157d (0.114, 0.208)

Cost per QALY gained 13 751c,e (12 400, 15 100)

a Differences that were statistically significant; p < 0.05.

b A quality-of-life score for a health state between 0 and 1, with 0 = death and 1 = optimal health. In this case, the duration of thehealth state was 6 months.

c 95% confidence intervals were crudely estimated using the 95% CI limits for the difference in cost (i.e. in the numerator), whilekeeping the gains in VTE avoidance and QALYs (i.e. in the denominator) constant.

d Difference in healthy month equivalence divided by 12 months.

e Incremental cost of dalteparin divided by gain in QALYs.

VTE = venous thromboembolism.

were ‘very familiar’ or ‘somewhat familiar’, respec- in healthy month equivalence was approximately1.88 months with dalteparin, which corresponded totively, with the cost of drugs used to treat cancer.an additional gain of approximately 0.157 QALYs.The final series of demographic questions focused

on respondents’ family experience with VTEs. ThePharmacoeconomic Analysisdata revealed that 3 of 24 patients (12.5%) had a

positive family history for VTEs.Using the differential in the cost and VTE recur-

Once all of the information had been presented rence rate, the initial analysis estimated the incre-on both treatments, respondents were asked to select mental cost per VTE avoided with dalteparin. Over-their preferred prophylactic intervention. Overall, all, the use of dalteparin as an alternative to warfarin23 of 24 (96%) respondents selected dalteparin over was associated with an incremental cost of approxi-warfarin. Healthy month equivalence scores and mately $Can27 700 (95% CI 24 400, 30 400) perhealth state utilities for each alternative were then VTE avoided (table VI). However, this cost-effec-estimated from the sample. The utility of the tiveness ratio fails to consider the gain in QALYsdalteparin health state was almost 2-fold higher than associated with dalteparin. To provide such infor-treatment with warfarin, indicating a QOL benefit mation, a cost-utility analysis was performed where(table VI). These preferences and utility scores for the incremental cost of dalteparin ($Can2159) wasdalteparin were likely due to a combination of fac- combined with the 0.157 QALY gain (a gain of 1.88tors such as improved efficacy, demonstrated safety health months equivalence corresponds to a QALYprofile with extended use and the ability to eliminate gain of 1.88 months per 12 months). The findingscontinuous laboratory monitoring. Overall, the gain revealed an incremental cost of $Can13 751 per



QALY gained, consistent with the hypothesis that Multivariate Analysis

dalteparin is a cost-effective intervention in cancerpatients at risk for recurrent VTEs (table VI). The primary objective of the multivariable analy-

sis was to identify patient subgroups in which theuse of dalteparin could be potentially cost saving.Sensitivity AnalysisAn initial assessment of the total cost (dependentvariable) revealed that it was highly skewed by a

A series of one-way sensitivity analyses were small number of high-cost cases. This is a commonthen conducted using the 95% CI of the health occurrence in cost of illness studies; usual practice ismonths equivalent scores (1.37, 2.49) and the cost to log-Normalise the data in this case. The adequacydifferential ($Can1943, $Can2373) between the two of the procedure was verified by inspection of thetreatments. Under the worst case scenario, in which normal plots and application of the Skew test. Thethe highest cost difference between the two treat- residuals from the final models were also examinedments (i.e. $2373) as well as the lowest gain in to ensure that they were Normally distributed.healthy months equivalence (i.e. 1.37) was used, the The adjusted R2 is a multivariate regression indi-cost per QALY gained with dalteparin would be cator of model goodness of fit, with a range betweenapproximately $Can21 000. Under the best case sce- 0 and 100%. The final regression model generatednario, in which the lowest cost difference (i.e. an adjusted R2 value of 0.48, suggesting that 48% of$Can1943) along with the highest gain in healthy the variability observed in the dependent variablemonths equivalence (i.e. 2.49) was used, the cost per was explained by the independent variables retainedQALY gained with dalteparin would be cost effec- in the model. The results of the multivariable analy-tive at $Can9400. These findings imply that the cost sis determined that treatment group, male gender,per QALY gain estimated in the current study was patient age and performance status at the initiationstable, despite reasonable variations in the point of therapy were significantly associated with theestimates. cost of therapy (table VII).

Table VII. Multivariable regression analysis on overall cost of therapy

Variable Parameter estimate SEM p-Value Effect on total cost

Intercept 7.65

Dalteparin group 0.89 0.66 <0.001 Overall 2.4-fold increase

Male 0.077 0.031 0.012 8% increase in men

Age –0.003 0.001 0.018 Overall decrease with increasing age

ECOG 1 0.019 0.060 0.74 NS

ECOG 2 0.15 0.062 0.81 NS

ECOG 3 0.24 0.24 0.30 NS

Interaction effects (vs ECOG 0)a

ECOG 1 × dalteparin –0.13 0.081 0.096 NS

ECOG 2 × dalteparin –0.26 0.83 0.002 23% decrease in dalteparin cost

ECOG 3 × dalteparin –0.71 0.29 0.016 51% decrease in dalteparin cost

Adjustedb R2 0.48

a The subgroup analysis was relative to an ECOG Performance Scale score of 0, meaning no physical impairment. For example, forECOG 1 × dalteparin = cost of dalteparin in ECOG 1 patients vs ECOG 0 patients who received dalteparin.

b Proportion of variability in the dependent variable that is accounted for by the model.

ECOG = European Cooperative Oncology Group; NS = not significant; SEM = standard error of the mean.



A closer examination of the variable for the invaluable in many situations, the data required todalteparin group revealed that, when controlling for construct valid models are often not available. As athe other factors, treatment with dalteparin was ap- result, analysts are forced to rely on data fromproximately 2.4-fold (anti-log of 0.89 = 2.4) more multiple sources and invariably have to make impor-costly than with oral anticoagulation therapy tant assumptions. Such assumptions are often based(p < 0.001). The results of the multivariable proce- on judgment rather than validated data. As an alter-dure confirmed those of the univariate analysis, that native to computer modelling, economic analysesthe cost of treatment with dalteparin was significant- based on actual resource utilisation data collected asly greater than that of warfarin. The negative value part of a prospective randomised trial offers realfor patient age implied that, as patient age increased, world information for health policy decision mak-overall cost of management decreased. Furthermore,

ing.the positive association with male sex and overall

In the current analysis, resource utilisation datacost implied that men accrued more costs than wo-collected as part of the CLOT study[5] were used tomen (table VII).estimate the incremental cost per VTE avoided andThe regression model provided some interestingQALY gained with dalteparin as an alternative toinsight related to the statistically significant andwarfarin. Within the framework of the treatmentnegative interaction effects between patients treatedregimen used in the Canadian-led CLOT trial, thewith dalteparin and performance status where pro-results of the univariate and multivariate economicgressive cost reductions were observed in poor per-analyses suggest that patients randomised to receiveformance status patients (table VII). The interpreta-

tion is that dalteparin-treated patients with a per- dalteparin had significantly higher overall costsformance status of 2 or 3 (i.e. progressively being compared with those treated with warfarin. Whenmore bed ridden) had significantly lower overall the absolute reduction in the VTE rate was com-management costs than performance status 0 pa- pared with the additional expenditure, dalteparintients (i.e. fully mobile) randomised to the same was associated with an incremental cost of approxi-therapy. As a result, the overall cost differential mately $Can27 700 per VTE avoided. However, thisbetween dalteparin and warfarin was reduced in estimate fails to consider treatment preferences andpoorer performance status patients. the associated patient utility gain secondary to re-

ceiving extended prophylaxis with dalteparin, whichDiscussion has several advantages over warfarin, including im-

proved efficacy, potentially enhanced safety, addedThe basic premise of health economic studiesconvenience, and, possibly, a survival advantage inpredicated on randomised controlled trials is to com-selected patient subgroups. When differences inpare the costs and consequences of alternative phar-treatment preferences and health-state utilities weremaceutical interventions and determine which treat-combined with the additional cost, dalteparin wasment offers the best value for money. There areassociated with a favourable incremental cost ofseveral approaches available to evaluate economic$Can13 751 per QALY gained. A $Can50 000 costefficiency. Decision analysis, or Markov computerper QALY has been suggested as a threshold, at ormodelling, is one of the most commonly used meth-below which new medical interventions should beods for conducting cost-effectiveness analyses. Us-considered for adoption by healthcare systems be-ing this technique, outcomes are typically presentedcause they are considered to have ‘acceptable’ eco-as incremental cost per QALY gained. Even though



nomic value.[21] Dalteparin appears to meet this cri- and patients who had a baseline ECOG performanceterion easily. status of 2 or 3. These results can be interpreted in

the following way. Patients with a baseline perform-Interestingly, the findings of the current study areance status of 2 or 3 who were treated withconsistent with those of Aujesky et al.,[22] who ap-dalteparin required less hospital resources than per-plied decision analysis computer modelling. In thatformance status 0 patients who received the samestudy, the investigators determined that extendedtherapy. Hence, the incremental cost of usingprophylaxis with dalteparin would be considereddalteparin in such patients would be reduced com-‘cost effective’ (i.e. a cost per QALY <$Can50 000)pared with performance status 0 patients. There areif the daily cost of the drug were reduced to less thantwo possible explanations for this phenomenon. Ear-$US18 per day for a dose of 200 IU/kg/day (2002ly death in dalteparin patients with performancevalues). As reported by the investigators, the cost instatus of 2 or 3 may have contributed to lowerCanada for dalteparin is approximately $US13 perhealthcare resource utilisation. However, if this hy-day (2002 values, equivalent to $Can18/day), whichpothesis was correct, then the same effect shouldwould generate a cost per QALY below thehave also been observed in the oral therapy group.$Can50 000 threshold for economic value.Alternatively, immobility, which is associated with

The determination of a cost per QALY gaineda performance status of 2 or 3, has been identified as

with dalteparin allows an economic value compari-a risk factor for VTEs.[25] It is tempting to speculate

son versus other agents used in cancer supportivethat the use of dalteparin in patients with a poorer

care. Two agents commonly used in cancer support-performance status may result in a lower cost differ-

ive care include recombinant erythropoietin and fil-ential relative to warfarin secondary to reducing the

grastim. In one study, which used decision analysisnumber of recurrent VTEs, compared with a similar

modelling in a stage IV breast cancer population, thegroup of performance status 0 patients. Neverthe-

incremental cost per QALY gained with prophylac-less, it is important to recall that the multivariate

tic recombinant erythropoietin was approximatelyanalysis was exploratory and that its findings need

$Can21 000.[23] In another decision analysis con-to be confirmed through a prospective study.

ducted in the US, the prophylactic use of filgrastimin breast cancer patients receiving adjuvant chemo-

Study Limitationstherapy was associated with a cost of $Can46 600per life-year gained (converted to $Can, using an There are several limitations in this analysis thatexchange rate of $US1 = $Can1.24, as of May have to be addressed. Some of the hospital resources2005).[24] A comparison with these agents indicates such as specific diagnostic tests utilised were proto-that prophylaxis with dalteparin offers at least com- col driven and may not completely reflect standardparable value for money. practice. These considerations may reduce the

Exploratory multivariable analysis provided generalisability (external validity) of the results tosome interesting insight into factors contributing to patients outside of the setting created by a randomis-the overall cost of therapy. There was a negative ed trial. Detailed healthcare utilisation data for theassociation between increasing age and overall cost management of recurrent VTEs and for the treat-of therapy, which could be related to increased ment of bleeding events were not available. There-cancer-related mortality with advancing age. Statis- fore, we had to rely on estimates reported in thetically significant and negative interaction effects literature for managing such events in both the hos-were identified between treatment with dalteparin pitalised and outpatient setting. Oncology nurses



M. Crowther: provided clinical expertise and guidance onand pharmacists were used as patient surrogates inthe analysis. Contributed to the development and review ofthe utility assessments. Even though respondentsthe manuscript.

had an average of 8.6 years of cancer experience and70.8% also had VTE management experience, pa-

Referencestients are preferable to surrogates.1. Kearon C. Natural history of venous thromboembolism. Circu-

lation 2003; 107 Suppl. 1: I22-302. Deitcher SR. Cancer-related deep venous thrombosis: clinicalConclusion

importance, treatment challenges, and management strategies.Semin Thromb Hemostasis 2003; 29: 247-58This pharmacoeconomic analysis suggests dal-

3. Sallah S, Wan JY, Nguyen NP. Venous thrombosis in patientsteparin, when used for long-term secondary VTE with solid tumors: determination of frequency and characteris-

tics. Thromb Haemost 2002; 87: 575-9prophylaxis in patients with cancer, is associated4. National Cancer Institute of Canada. Canadian Cancer Statistics

with an estimated incremental cost of approximately 2004. Toronto, 2004 [online]. Available from URL: http://www.cancer.ca [Accessed 2005 Feb 4]$Can2159 per patient. When treatment preferences

5. Lee AY, Levine MN, Baker RI, et al. Low-molecular-weightand health-state utility gains were combined withheparin versus a coumarin for the prevention of recurrent

the additional cost, the use of dalteparin for this venous thromboembolism in patients with cancer. N Engl JMed 2003; 349: 146-53indication was associated with an incremental cost

6. Lee AYY, Rickles FR, Julian JA, et al. Randomized comparisonof $Can13 751 per QALY gained. Given the practi- of low molecular weight heparin and coumarin derivatives onthe survival of patients with cancer and venous thromboembo-cal advantages of this drug in terms of safety, im-lism. J Clin Oncol 2005; 23: 2123-9proved efficacy, convenience and the acceptable

7. Klerk CP, Smorenburg SM, Otten HM, et al. The effect of loweconomic profile, it is a clinically and economically molecular weight heparin on survival in patients with ad-

vanced malignancy. J Clin Oncol 2005; 23: 2130-5sound alternative to warfarin for the prevention of8. Kakkar AK, Levine MN, Kadziola Z, et al. Low molecularrecurrent VTEs in cancer patients. weight heparin, therapy with dalteparin, and survival in ad-

vanced cancer: the Fragmin Advanced Malignancy OutcomeStudy (FAMOUS). J Clin Oncol 2004; 22: 1944-8Acknowledgements

9. O’Brien BJ, Anderson DR, Goeree R. Cost-effectiveness ofenoxaparin versus warfarin prophylaxis against deep-veinThe authors would like to express their gratitude to thethrombosis after total hip replacement. CMAJ 1994; 150:

local investigators who administered the health-utility assess-1083-90

ment. This study was funded via a research grant from Pfizer 10. Haentjens P, De Groote K, Annemans L. Prolonged enoxaparinCanada, Inc. to the principal investigator, following an RFP therapy to prevent venous thromboembolism after primary hip

or knee replacement: a cost-utility analysis. Arch Orthop Trau-(request for proposal).ma Surg 2004; 124: 507-17Once the study had begun, an arms-length approach was

11. Botteman MF, Caprini J, Stephens JM, et al. Results of antaken with the sponsor. The principal investigator retained theeconomic model to assess the cost-effectiveness of enoxaparin,

right to publish the findings of a study in a journal of his a low-molecular-weight heparin, versus warfarin for the pro-choosing. phylaxis of deep vein thrombosis and associated long-term

G. Dranitsaris, M. Vincent and M. Crowther have acted as complications in total hip replacement surgery in the UnitedStates. Clin Ther 2002; 24: 1960-86consultants to Pfizer Canada, Inc. The corresponding author

12. Hull RD, Pineo GF, Raskob GE. The economic impact ofhad full access to all the data in the study and final responsi-treating deep vein thrombosis with low molecular weightbility for the decision to submit the paper.heparin: outcome of therapy and health economic aspects.

Contribution of each author. Haemostasis 1998; 28 Suppl. 3: 8-16G. Dranitsaris: study design, development of data collec- 13. Gordois A, Posnett J, Borris L, et al. The cost effectiveness of

fondaparinux compared with enoxaparin as prophylaxistion instrument, statistical analysis and preparation of manu-against thromboembolism following major orthopedic surgery.script.J Thromb Haemost 2003; 1: 2167-74M. Vincent: study design, development of data collection

14. Dranitsaris G, Kahn S, Stumpo C, et al. Pharmacoeconomicinstrument and contribution to manuscript development and analysis of fondaparinux versus enoxaparin for the preventionreview. Also provided clinical expertise and guidance on the of thromboembolic events in orthopedic surgery patients. Am Janalysis. Cardiovasc Drugs 2004; 4: 325-33



15. Dranitsaris G. The cost of blood transfusions in cancer patients: 22. Aujesky D, Smith KJ, Cornuz J, et al. Cost effectiveness of lowa reanalysis of a Canadian economic evaluation. J Oncol molecular weight heparin for secondary prophylaxis of cancerPharm Prac 2000; 6: 1-6 related venous thromboembolism. Thromb Haemost 2005; 93:

16. Statistics Canada [online]. Available from URL: http:// 592-9www.statcan.ca [Accessed 2005 Feb 4]

23. Martin SC, Gagnon DD, Zhang L, et al. Cost-utility analysis of17. Torrance GW. Utility approach to measuring health-related survival with epoetin-alfa versus placebo in stage IV breast

quality of life. J Chronic Dis 1987; 40: 593-600cancer. Pharmacoeconomics 2003; 21: 1153-69

18. Gafni A. Alternatives to the QALY measure for economic24. Silber JH, Fridman M, Shpilsky A, et al. Modeling the cost-evaluations. Support Care Cancer 1997; 5: 105-11

effectiveness of granulocyte colony-stimulating factor use in19. Leung P, Tanock IF, Oza AM, et al. Cost utility analysis ofearly-stage breast cancer. J Clin Oncol 1998; 16: 2435-44chemotherapy using paclitaxel, docetaxel or vinorelbine for

patients with anthracycline-resistant breast cancer. J Clin 25. Anderson FA, Spencer FA. Risk factors for venous thromboem-Oncol 1999; 17: 3082-90 bolism. Circulation 2003; 107: I9-I16

20. Ortega A, Dranitsaris G, Sturgeon J, et al. Cost utility analysis ofpaclitaxel in combination with cisplatin for patients with ad-vanced ovarian cancer. Gynecol Oncol 1997; 66: 454-63 Correspondence and offprints: George Dranitsaris, Aug-

21. Laupacis A, Feeny D, Detsky AS, et al. How attractive does amentium Pharma Consulting, 283 Danforth Ave, Suite 448,new technology have to be to warrant adoption and utilization?Toronto, Ontario M4K 1N2, Canada.Tentative guidelines for using clinical and economic evalua-E-mail: [email protected]. CMAJ 1992; 146: 473-81


original article

The

new england journal

of

medicine

n engl j med

349;2

www.nejm.org july

10, 2003

146

Low-Molecular-Weight Heparinversus a Coumarin for the Prevention

of Recurrent Venous Thromboembolism in Patients with Cancer

Agnes Y.Y. Lee, M.D., Mark N. Levine, M.D., Ross I. Baker, M.D.,Chris Bowden, M.D., Ajay K. Kakkar, M.B., Martin Prins, M.D.,

Frederick R. Rickles, M.D., Jim A. Julian, M.Math., Susan Haley, B.Sc., Michael J. Kovacs, M.D., and Michael Gent, D.Sc.,

for the Randomized Comparison of Low-Molecular-Weight Heparin versus Oral Anticoagulant Therapy for the Prevention of Recurrent Venous

Thromboembolism in Patients with Cancer (CLOT) Investigators*

From the Departments of Medicine(A.Y.Y.L., M.N.L.) and Clinical Epidemiologyand Biostatistics (M.N.L., J.A.J., S.H., M.G.),McMaster University, Hamilton, Ont., Can-ada; Henderson Research Centre, HamiltonHealth Sciences and McMaster Universi-ty, Hamilton, Ont., Canada (M.N.L., J.A.J.,S.H., M.G.); the Department of Medicine,University of Western Australia, Perth,Australia (R.I.B.); Pharmacia Corporation,Peapack, N.J. (C.B.); the Department ofSurgical Oncology and Technology, ImperialCollege, London (A.K.K.); Academic Hos-pital Maastricht, Maastricht, the Nether-lands (M.P.); the Department of Medicineand Pediatrics, George Washington Univer-sity, Washington, D.C. (F.R.R.); and the De-partment of Medicine, University of West-ern Ontario, London, Ont., Canada (M.J.K.).Address reprint requests to Dr. Levine atHamilton Health Sciences, HendersonHospital, Rm. 9, 90 Wing, 711 ConcessionSt., Hamilton, ON L8V 1C3, Canada.

*Participating investigators and institu-tions are listed in the Appendix.

N Engl J Med 2003;349:146-53.

Copyright © 2003 Massachusetts Medical Society.

background

Patients with cancer have a substantial risk of recurrent thrombosis despite the use oforal anticoagulant therapy. We compared the efficacy of a low-molecular-weight heparinwith that of an oral anticoagulant agent in preventing recurrent thrombosis in patientswith cancer.

methods

Patients with cancer who had acute, symptomatic proximal deep-vein thrombosis, pul-monary embolism, or both were randomly assigned to receive low-molecular-weightheparin (dalteparin) at a dose of 200 IU per kilogram of body weight subcutaneouslyonce daily for five to seven days and a coumarin derivative for six months (target interna-tional normalized ratio, 2.5) or dalteparin alone for six months (200 IU per kilogramonce daily for one month, followed by a daily dose of approximately 150 IU per kilogramfor five months).

results

During the six-month study period, 27 of 336 patients in the dalteparin group had recur-rent venous thromboembolism, as compared with 53 of 336 patients in the oral-antico-agulant group (hazard ratio, 0.48; P=0.002). The probability of recurrent thromboem-bolism at six months was 17 percent in the oral-anticoagulant group and 9 percent in thedalteparin group. No significant difference between the dalteparin group and the oral-anticoagulant group was detected in the rate of major bleeding (6 percent and 4 percent,respectively) or any bleeding (14 percent and 19 percent, respectively). The mortality rateat six months was 39 percent in the dalteparin group and 41 percent in the oral-anticoag-ulant group.

conclusions

In patients with cancer and acute venous thromboembolism, dalteparin was more effec-tive than an oral anticoagulant in reducing the risk of recurrent thromboembolism with-out increasing the risk of bleeding.

abstract

Copyright © 2003 Massachusetts Medical Society. All rights reserved. Downloaded from www.nejm.org by CAMILO FALCON on May 21, 2007 .

n engl j med

349;2

www.nejm.org july

10, 2003

long-term anticoagulation in patients with cancer

147

he standard treatment for acute

venous thromboembolism consists of ini-tial therapy with low-molecular-weight

heparin or unfractionated heparin followed by long-term therapy with an oral anticoagulant.

1

This ap-proach is highly effective in most patients, but pa-tients with cancer have a substantial risk of recurrentthromboembolism and hemorrhagic complica-tions.

2,3

Furthermore, oral anticoagulant therapy isproblematic in patients with cancer. Drug interac-tions, malnutrition, vomiting, and liver dysfunctioncan lead to unpredictable levels of anticoagulation.Invasive procedures and thrombocytopenia causedby chemotherapy often require interruption of an-ticoagulant therapy, and poor venous access canmake laboratory monitoring difficult. These limi-tations may contribute to the higher risk of recur-rent thromboembolism and bleeding in patientswith cancer than in patients without cancer.

2,3

Secondary prophylaxis with low-molecular-weight heparin may be a more effective and practicalalternative to oral anticoagulant therapy. Unlike vi-tamin K antagonists, low-molecular-weight hepa-rins have predictable pharmacokinetic propertiesand drug interactions,

4

and they can be effective inpatients with cancer who have recurrent thrombo-embolism while receiving warfarin.

5-7

Poor gastro-intestinal absorption is not a concern with subcu-taneously injected low-molecular-weight heparins.The therapeutic dosage is based on the patient’sweight, and laboratory monitoring is not routinelyrequired. With a rapid onset of action and predict-able clearance, they are also convenient for patientswho require frequent interruptions of anticoagu-lant therapy.

We performed a multicenter, randomized, open-label clinical trial to investigate whether the low-molecular-weight heparin dalteparin is more effec-tive and safer than oral anticoagulant therapy inpreventing recurrent thromboembolism in patientswith cancer who have acute venous thromboem-bolism.

study population

Adult patients with active cancer and newly diag-nosed, symptomatic proximal deep-vein thrombo-sis, pulmonary embolism, or both were eligible.Active cancer was defined as a diagnosis of cancer,other than basal-cell or squamous-cell carcinoma ofthe skin, within six months before enrollment, any

treatment for cancer within the previous six months,or recurrent or metastatic cancer. Proximal deep-vein thrombosis was diagnosed on the basis of evi-dence of thrombus in the popliteal or more proximalveins on compression ultrasonography or contrastvenography.

8

A diagnosis of pulmonary embolismrequired verification by ventilation–perfusion lungscanning, helical computed tomography, or pulmo-nary angiography.

9-11

Patients were excluded if they weighed 40 kg orless, had an Eastern Cooperative Oncology Group(ECOG) performance status score of 3 or 4,

12

hadreceived therapeutic doses of any heparin for morethan 48 hours before randomization, were alreadyreceiving oral anticoagulant therapy, had had activeor serious bleeding within the previous two weeks,had conditions associated with a high risk of seriousbleeding (e.g., active peptic ulcer or recent neuro-surgery), had a platelet count of less than 75,000 percubic millimeter; had contraindications to heparintherapy (e.g., heparin-induced thrombocytopenia)or the use of contrast medium, had a creatinine levelthat was at least three times the upper limit of thenormal range, were pregnant, or could not returnto the clinical center for follow-up.

At base line, a complete blood count was ob-tained and the prothrombin time, activated partial-thromboplastin time, and serum creatinine and liv-er enzyme levels were measured. The study protocolwas reviewed and approved by the institutional re-view boards of each participating center. Writteninformed consent was obtained from all patients.

treatment regimens

Patients were assigned to receive subcutaneousdalteparin or an oral anticoagulant. Randomizationwas stratified according to the clinical center andcentralized at the coordinating and methods centerat the Henderson Research Centre, Hamilton, On-tario, Canada. The patients assigned to the oral-anti-coagulant group received a low-molecular-weightheparin, dalteparin (Fragmin, Pharmacia), initiallyfor five to seven days and a vitamin K antagonist forsix months. Dalteparin was supplied in 3.8-ml mul-tidose vials containing 25,000 IU of dalteparin permilliliter. A dose of 200 IU per kilogram of bodyweight (maximal daily dose, 18,000 IU) was admin-istered once daily. Within 24 hours after random-ization, patients in this group also began takingwarfarin or acenocoumarol. Warfarin was used inall participating centers except those in the Nether-lands and Spain. All doses were adjusted to achieve

t

methods


n engl j med

349;2

www.nejm.org july

10

,

2003

The

new england journal

of

medicine

148

a target international normalized ratio (INR) of 2.5(therapeutic range, 2.0 to 3.0). Dalteparin was dis-continued after a minimum of five days and oncethe INR had remained above 2.0 for two consecu-tive days. The INR was measured at least once everytwo weeks thereafter.

The patients assigned to the dalteparin group re-ceived 200 IU of dalteparin per kilogram (maximaldaily dose, 18,000 IU) from multidose vials oncedaily for the first month. For the remaining fivemonths, patients were treated with 75 to 83 percentof the full dose (approximately 150 IU per kilogram)with the use of prefilled syringes. These syringeswere supplied according to the patient’s weight:7500 IU for those weighing 56 kg or less, 10,000 IUfor those weighing 57 to 68 kg, 12,500 IU for thoseweighing 69 to 82 kg, 15,000 IU for those weighing83 to 98 kg, and 18,000 IU for those weighing 99 kgor more. Patients were instructed to inject the entirecontents of one syringe once daily. The practice ofmeasuring the anticoagulant effect against activatedfactor X was discouraged. The only exception wasin the cases of patients in whom clinically signifi-cant renal insufficiency developed.

Dose adjustment was recommended for patientswith thrombocytopenia. Study drug was withheldfrom patients with a platelet count of less than50,000 per cubic millimeter and was resumed atthe scheduled dose when the count was 100,000 percubic millimeter or higher. When the platelet countwas 50,000 to 99,000 per cubic millimeter, the nextlower dose of prefilled syringe was used in the dal-teparin group, whereas the target INR was reducedto 2.0 (range, 1.5 to 2.5) in the oral-anticoagulantgroup.

The assigned study treatment was administeredat home whenever possible, and it was continuedduring hospitalization. Patients, family members,or both were taught how to inject the medication,but home care or equivalent nursing services werearranged if necessary.

follow-up

During the six-month study period, patients werecontacted by telephone every two weeks and wereseen in the clinic one week and one, three, and sixmonths after randomization. Each clinic visit in-cluded a history taking, physical examination, as-sessment of compliance, and blood drawing for thecalculation of a complete blood count and measure-ment of liver enzymes and creatinine. Scheduledcalls and visits included a standardized assessment

of the signs and symptoms of recurrent thrombo-embolism, bleeding episodes, and adverse reac-tions. Patients were instructed to report to the clinicimmediately if they had any bleeding or symptomsof recurrent deep-vein thrombosis, pulmonary em-bolism, or both. All suspected episodes of recurrentthrombosis were investigated with the use of objec-tive tests, according to prespecified diagnostic al-gorithms.

13

All patients were followed until thesix-month visit, death, or withdrawal of consent,whichever came first.

outcome measures

The primary efficacy outcome was the first epi-sode of objectively documented, symptomatic, re-current deep-vein thrombosis, pulmonary embo-lism, or both during the six-month study period.Recurrent deep-vein thrombosis was diagnosed ifa previously compressible proximal venous segmentor segments could no longer be compressed on ul-trasonography or if there were constant intralumi-nal filling defects in two or more projections on ve-nography. Unequivocal extension of the thrombuswas required for the diagnosis of recurrence if theresults were abnormal on previous testing.

8,14,15

Ve-nography was required to confirm distal deep-veinthrombosis. Pulmonary embolism was diagnosedon the basis of a lung scan indicating a high proba-bility of its presence, as indicated by the presence ofnew or enlarged areas of segmental perfusion de-fects with ventilation–perfusion mismatch; an ab-normal perfusion scan with documentation of newor recurrent deep-vein thrombosis; the presence ofnonenhancing filling defects in the central pulmo-nary vasculature on helical computed tomography;a finding of intraluminal filling defects on pulmo-nary angiography; or evidence of fresh pulmonaryembolism at autopsy.

9,14,15

Secondary outcome events included clinicallyovert bleeding (both major bleeding and any bleed-ing) and death. A bleeding event was classified asmajor if it was associated with death, occurred at acritical site (intracranial, intraspinal, intraocular,retroperitoneal, or pericardial area), resulted in aneed for a transfusion of at least 2 units of blood,or led to a drop in hemoglobin of at least 2.0 g perdeciliter.

14,15

All suspected events were reviewed by a centraladjudication committee whose members were un-aware of the patients’ treatment assignments. Sup-porting documents, including clinical notes, imag-ing studies, and the results of laboratory tests, were


n engl j med

349;2

www.nejm.org july

10, 2003


149

forwarded to the coordinating and methods centerfor adjudication. All reported episodes that weresuggestive of recurrent thrombosis were evaluatedand confirmed or rejected as representing recur-rence, overt bleeding events were classified as ma-jor or minor, and all reported deaths were reviewedto determine the cause of death.

statistical analysis

The initial calculation of the sample size was basedon an estimated risk of recurrent thrombosis of 20percent at six months among patients treated withoral anticoagulant therapy. In order to detect a 50percent reduction in risk with a power of 0.85 anda two-sided alpha of 0.05, it was determined that70 primary efficacy outcome events were required.In order to adjust for the loss to follow-up from earlydeath, the sample size was increased by 20 percent.A blinded reassessment of the sample size that wasspecified in the protocol led us to increase the tar-geted enrollment by an additional 90 patients. Ac-cordingly, we determined that 676 patients wouldbe required.

An analysis of efficacy end points was performedaccording to the intention-to-treat principle andincluded all randomized patients who had a con-firmed, qualifying thrombotic event and active can-cer. The primary analysis of efficacy was based onthe time from randomization to the first recurrentthromboembolic event. Data on patients withoutevents were censored at the time of the six-monthvisit or death, whichever occurred first. The risk ofrecurrence over time was estimated according to theKaplan–Meier method, and the treatment groupswere compared with use of the two-sided log-ranktest.

16,17

A Cox proportional-hazards regressionmodel was used to examine the influence of poten-tially prognostic base-line factors (e.g., age, ECOGstatus, type of qualifying thrombotic event, andpresence or absence of metastases) on the risk ofrecurrent thromboembolism. Interactions betweentreatment group and covariates were assessed inthe model.

Death from all causes was also calculated andcompared with use of the Kaplan–Meier methodand the two-sided log-rank test, respectively. Pa-tients who received at least one dose of the studydrug were included in the safety analyses. The pro-portions of patients in each group who had a majorbleeding event after the first dose and up to 48 hoursafter the permanent discontinuation of the studydrug were compared with use of a two-sided Fish-

er’s exact test. Similarly, the proportions of patientsin each group with any bleeding were compared.

Two investigators designed and developed theoriginal protocol, which was revised and approvedby the steering committee. This committee, com-posed of seven academic members and one repre-sentative of the sponsor, was responsible for over-seeing the conduct of the study, formulating thestatistical-analysis plan, reviewing and interpretingthe data, and preparing the manuscript. The centraladjudication committee and data-monitoring com-mittee operated independently of the sponsor. TheClinical Trials Methodology Group at the Hender-son Research Centre, Hamilton Health Sciences,was responsible for study coordination, data man-agement, statistical analyses, and administrative

* Plus–minus values are means ±SD. ECOG denotes Eastern Cooperative On-cology Group, DVT deep-vein thrombosis, and PE pulmonary embolism.

† Eight patients were included in the study before the protocol was amended to exclude patients with an ECOG score of 3 or 4.

‡ Antineoplastic treatment included chemotherapy, radiation, and surgery.

Table 1. Base-Line Characteristics of the Patients.*

CharacteristicDalteparin(N=338)

Oral Anticoagulant(N=338)

Mean age (yr) 62±12 63±13

Female sex (no. of patients) 179 169

ECOG performance score (no. of patients)0123†

80135118

5

63150122

3

Hospitalization status (no. of patients)OutpatientInpatient

169169

156182

Hematologic cancer (no. of patients) 40 30

Solid tumor (no. of patients)No clinical evidence of diseaseLocalized diseaseMetastatic disease

3639

223

3343

232

Antineoplastic treatment (no. of patients)‡ 266 259

Current smoker (no. of patients) 33 42

History of DVT or PE (no. of patients) 39 36

Recent major surgery (no. of patients) 62 67

Central venous catheter (no. of patients) 46 40

Qualifying thrombotic event (no. of patients)DVT alonePE, with or without DVT

235103

230108


n engl j med

349;2

www.nejm.org july

10

,

2003

The

new england journal

of

medicine

150

activities. Pharmacia provided funding and thestudy drug.

study population

Forty-eight clinical centers in eight countries par-ticipated (see the Appendix). Recruitment began inMay 1999 and was completed in October 2001. Ofthe 1303 patients who met the inclusion criteria,439 also met one or more of the exclusion criteriaand were not considered eligible. The three mostfrequent reasons for exclusion were an ECOG scoreof 3 or 4 (169 patients), treatment with any heparinfor more than 48 hours (107), and an inability toreach the clinical center easily (43). Of the remain-ing 864 eligible patients, 676 provided written in-formed consent. Eight patients with an ECOG scoreof 3 were enrolled before the protocol was amend-ed to exclude patients with such a score.

Of the 676 consenting patients, 338 were allo-cated to receive dalteparin and 338 were assignedto oral anticoagulant therapy, each for six months.Patients in the two groups had similar base-linecharacteristics (Table 1). Ninety percent of the pa-tients had solid tumors (Table 2), and 67 percent hadmetastatic disease at the time of randomization.

anticoagulant therapy

The mean duration of study treatment was 125 daysin the dalteparin group and 115 days in the oral-anticoagulant group. For patients who did not havean outcome event throughout the study period, themean duration of study treatment was 170 days inboth groups.

In the oral-anticoagulant group, the mean (

±

SD)INR was 2.5

±

0.75. Using linear interpolation overtime, we estimated that the INR was in the therapeu-tic range 46 percent of the time, below the range 30percent of the time, and above the range 24 percentof the time.

recurrent venous thromboembolism

Two patients in each group were excluded from theefficacy analysis because they did not have a quali-fying thrombotic event: one patient had a thrombo-sis in an arm vein, one had an asymptomatic throm-bus in the leg, and the other two did not have aconfirmed pulmonary embolism. Symptomatic, re-current deep-vein thrombosis, pulmonary embo-lism, or both occurred in 27 of 336 patients in thedalteparin group and 53 of 336 patients in the oral-

anticoagulant group (Table 3). The hazard ratio forrecurrent thromboembolism in the dalteparin groupas compared with the oral-anticoagulant group was0.48 (95 percent confidence interval, 0.30 to 0.77;P=0.002) over the six-month study period (Fig. 1).The Kaplan–Meier estimate of the probability of re-current thrombosis at six months was 9 percent inthe dalteparin group, as compared with 17 percentin the oral-anticoagulant group. All recurrent deep-vein thromboses were proximal. No significant in-teractions between treatment group and risk factorswere detected. Of the 53 thrombotic events in theoral-anticoagulant group, 20 occurred when the INRwas below 2.0.

bleeding

Three patients assigned to oral anticoagulant ther-apy did not receive the study drug and were exclud-

results

Table 2. Sites of Solid Tumors.

Tumor SiteDalteparin(N=298)

Oral Anticoagulant(N=308)

no. of patients

Breast 59 49

Colorectal area 54 54

Lung 40 50

Genitourinary tract 39 47

Gynecologic system 38 30

Pancreas 13 16

Brain 14 13

Other 41 49

Table 3. Primary Efficacy Outcome Events.

EventDalteparin(N=336)

OralAnticoagulant

(N=336)

no. of patients

Deep-vein thrombosis alone 14 37

Nonfatal pulmonary embolism

8 9

Fatal pulmonary embolism 5 7

Total 27 53


n engl j med

349;2

www.nejm.org july

10, 2003


151

ed from the safety analyses. Nineteen of 338 patientsin the dalteparin group (6 percent) and 12 of 335patients who received oral anticoagulant therapy(4 percent) had major bleeding (P=0.27). The re-spective rates of any bleeding were 14 percent and19 percent (P=0.09). At the time of a major bleed-ing event, two patients in the dalteparin group hadthrombocytopenia. Major bleeding was associatedwith an INR of more than 3.0 in six patients in theoral-anticoagulant group.

In the dalteparin group, one patient died frommassive hemoptysis related to metastatic lung can-cer and three patients bled at a critical site: one pa-tient with a brain tumor had intracranial bleeding,one patient with prostate cancer had retroperitonealbleeding, and one patient with lung cancer had peri-cardial bleeding. In the oral-anticoagulant group,there were no fatal bleeding events and four patientsbled at a critical site: two patients, one with breastcancer and one with prostate cancer, had intracrani-al bleeding, and two patients, one with a brain tu-mor and one with prostate cancer, had retroperito-neal bleeding.

mortality

During the six-month study period, 130 patientsdied in the dalteparin group and 136 patients died inthe oral-anticoagulant group. The respective mor-tality rates at six months were 39 percent and 41 per-cent (P=0.53) (Fig. 2). Ninety percent of the deathsin each group were due to progressive cancer.

In patients with cancer, recurrent thromboembo-lism complicates management and diminishes thepatients’ quality of life. Our study shows that therisk of symptomatic, recurrent thromboembolismamong patients with active cancer is significantlylower with dalteparin therapy than with oral antico-agulant therapy. Although previous trials comparinglow-molecular-weight heparins with warfarin forthe secondary prophylaxis of venous thromboembo-lism did not find a difference in the risk of recurrentthrombosis, most of the trials were small and con-ducted primarily in patients without cancer.

18-24

We did not detect a significant difference in therates of major bleeding or any bleeding between thetreatment groups. Given the limitations of cross-study comparisons, the rates of bleeding in the oral-anticoagulant group are consistent with those inprevious studies.

2,25

However, our rates of major

discussion

Figure 1. Kaplan–Meier Estimates of the Probability of Symptomatic Recur-rent Venous Thromboembolism among Patients with Cancer, According to Whether They Received Secondary Prophylaxis with Dalteparin or Oral Anti-coagulant Therapy for Acute Venous Thromboembolism.

An event was defined as an objectively verified, symptomatic episode of recur-rent deep-vein thrombosis, pulmonary embolism, or both during the six-month study period. The hazard ratio for recurrent thromboembolism in the dalteparin group as compared with the oral-anticoagulant group was 0.48 (95 percent confidence interval, 0.30 to 0.77; P=0.002 by the log-rank test).

No. at RiskDalteparinOral anticoagulant

336 336

301 280

264 242

235 221

227 200

210 194

164 154

Prob

abili

ty o

f Rec

urre

nt V

enou

sTh

rom

boem

bolis

m (%

)

Days after Randomization

25

20

15

10

5

00 30 60 90 120 150 180 210

Oral anticoagulant

Dalteparin

P=0.002

Figure 2. Kaplan–Meier Estimates of the Probability of Death from All Causes among Patients with Cancer, According to Whether They Received Secondary Prophylaxis with Dalteparin or Oral Anticoagulant Therapy for Acute Venous Thromboembolism.

There was no significant difference between the groups (P=0.53 by the log-rank test).

No. at RiskDalteparinOral anticoagulant

336 336

310 301

274 268

248 240

237 220

220 211

206 194

Prob

abili

ty o

f Dea

th (%

)

Days after Randomization

50

40

30

20

10

00 30 60 90 120 150 180 210

Oral anticoagulant

Dalteparin

P=0.53


n engl j med

349;2

www.nejm.org july

10

,

2003

The

new england journal

of

medicine

152

bleeding were lower than those reported in anotherrandomized trial involving patients with cancer; inthat study, 7 percent of patients who received low-molecular-weight heparin and 16 percent of thosewho received warfarin had major bleeding over athree-month period.

18

Differences in the patientpopulation, the INR control, and outcome assess-ment may explain some of the variations in theseresults.

The open-label design could be a potential sourceof bias in our trial. We believed that a double-blinddesign would not be logistically feasible or safe inpatients with cancer who had many other seriousconditions and who were taking multiple drugs,potentially increasing the risk of drug interactions.We tried to minimize reporting and diagnostic biasby contacting patients in both groups at frequentand regular intervals, using standardized follow-upprocedures, using objective tests to evaluate sus-pected events, and having all suspected outcomesevaluated by a central committee whose memberswere unaware of the patients’ treatment assign-ments. Also, substantial bias related to treatmentmanagement is unlikely because the level of INR

control achieved was similar to that in other stud-ies and showed that patients who received oral an-ticoagulant therapy were treated adequately.

2,18

Inaddition, the mean duration of treatment in pa-tients who did not have any outcome event was thesame in the two groups.

When we planned the study, there was little in-formation on the optimal dose of low-molecular-weight heparin for secondary prophylaxis. We de-signed a regimen that would provide intensiveanticoagulation initially and potentially reduce therisk of anticoagulant-related bleeding over the longterm. Practical issues regarding the long-term useof low-molecular-weight heparin include the cost ofthe drug and the feasibility of self-injection. Long-term self-injection of dalteparin was acceptable toour patients, and it significantly reduced the risk ofrecurrent venous thromboembolism without in-creasing the risk of bleeding.

Funded by Pharmacia, Peapack, N.J., which also supplied thestudy drug. Dr. Lee is the recipient of a New Investigator Award fromthe Canadian Institutes of Health Research, Drug Research and De-velopment Program; Dr. Levine is the Buffett Taylor Chair in BreastCancer Research, McMaster University, Hamilton, Ont., Canada;and Dr. Kovacs is an Internal Scholar of the Department of Medi-cine, University of Western Ontario, London, Ont., Canada.

appendix

The following investigators and institutions participated in the CLOT Trial:

Steering Committee:

M. Levine (chair), R. Baker, C. Bowden, M.Gent, A. Kakkar, A. Lee, M. Prins, F. Rickles;

External Safety and Efficacy Monitoring Committee:

J. Pater (chair), H. Büller, S. Goldhaber;

CentralAdjudication Committee:

J. Ginsberg, J. Hirsh, C. Kearon, G. Thomson, J. Weitz;

Coordinating and Methods Center:

Clinical Trials MethodologyGroup, Henderson Research Centre, Hamilton, Ont., Canada — J. Julian, S. Haley, A. Ling, B. Rush, T. Finch, L. Bonilla-Escobedo, L. Mat-thews, J. Windsor, C. Tavormina, H. Nelson, G. Lewis, J. Sicurella;

Clinical Centers

(the numbers of patients enrolled in each country are givenin parentheses) —

Canada (255)

: Hamilton Health Sciences, Henderson Hospital, Hamilton, Ont. — A. Lee, N. Booker, S. Schmidt; LondonHealth Sciences Centre, London, Ont. — M. Kovacs, B. Morrow; Queen Elizabeth II Health Sciences Centre, Halifax, N.S. — B. McCarron,S. Pleasance; Toronto General Hospital, Toronto — W.F. Brien, S. Boross-Harmer; St. Jospeh’s Hospital, Hamilton, Ont. — J.D. Douketis,T. Schnurr; Montreal General Hospital, Montreal — S. Solymoss, B. St. Jacques; Sunnybrook and Women’s College Health Sciences Centre,Toronto — W. Geerts, K. Code; British Columbia Cancer Agency, Vancouver Cancer Centre, Vancouver, B.C. — S. Chia, S. Monkman;Hamilton Health Sciences, Hamilton General Hospital, Hamilton, Ont. — A.G.G. Turpie, J. Johnson; Kelowna General Hospital, Kelowna,B.C. — J. Sutherland, S. Shori;

Australia (144) and New Zealand (16), Australasian Society of Thrombosis and Haemostasis

: Royal Perth Hospital,Perth, W.A. — R. Baker, J. Smith; Flinders Medical Centre, Bedford Park, S.A. — D.W. Coghlan, J.M. Osmond; Prince of Wales Hospital,Randwick, N.S.W. — S. Dunkley, B. Chong; Box Hill Hospital, Monash University, Box Hill, Victoria — H. Salem, L. Poulton; WestmeadHospital, Westmead, N.S.W. — M. Hertzberg, P. Stavros; Auckland Hospital, Auckland — P. Ockelford, V. Rolfe-Vyson; St. George Hospi-tal, Kogarah, N.S.W. — T.A. Brighton, R. Ristuccia; Royal North Shore Hospital, University of Sydney, Sydney, N.S.W. — C.M. Ward, K.Sheather; Royal Adelaide Hospital, Adelaide, S.A. — I.N. Olver, T. Marafioti; St. Vincent’s Hospital, Sydney, N.S.W. — D. Ma; Monash Med-ical Centre, Clayton, Victoria — T.E. Gan, A. Cummins; Royal Melbourne Hospital, Parkville, Victoria — A. Grigg, E. Cinc;

United States(118)

: University of Southern California, Keck School of Medicine, Los Angeles — H. Liebman, I. Weitz; University of Texas, M.D. AndersonCancer Center, Houston — C.P. Escalante, P. Horace; Northwestern University, Chicago — D. Green, M. Calimaran; University of NorthCarolina at Chapel Hill, Chapel Hill — S. Moll, S.K. Jones; Arizona Cancer Center, University of Arizona, Tucson — A. Stopeck, K. Glennie;Atlanta Veterans Affairs Medical Center–Emory University, Atlanta — M. Ribeiro, L. Starke; Cleveland Clinic Foundation, Cleveland — S.R.Deitcher; Mt. Sinai Medical Center, New York — L. Lipsey; St. Joseph Mercy Oakland, Pontiac, Mich. — A. Brady, R. Krishnan; University ofVermont and Fletcher Allen Health Care, Burlington — M. Cushman, L. Chassereau; University of Virginia Health System, Charlottesville— B.G. Macik, L. Newton; Lovelace Health System, Albuquerque, N.M. — A. Tarnower, R.J. Weiler; Newark Beth Israel Medical Center,Newark, N.J. — A.J. Cohen, E. White; University of Connecticut, Farmington — R. Bona, K. Jennings;

Italy (67)

: Ospedali Riuniti, Bergamo— A. Falanga, R. Labianca; Clinica Medica II, University of Padua, Padua — P. Prandoni, A. Piccioli, E. Zanon; Angelo Bianchi Bonomi He-mophilia Thrombosis Center, University of Milan and National Cancer Institute of Milan, Milan — A.B. Federici, G. Pizzocaro;

the Nether-lands (41)

: Academic Medical Center of the University of Amsterdam, Amsterdam — S.M. Smorenburg, C.P.W. Klerk; University HospitalNymegen, Nymegen — F. Berkmortel, D.J.T. Wagener; Maasland Hospital, Sittard — F.L.G. Erdkamp; St. Elisabeth Hospital, Tilburg — C.van der Heul, C. Post; St. Antonius Hospital, Nieuwegein — D.H. Biesma; Van Weel Bethesda Hospital, Dirksland — C. Kroon, M. Kam-phuis van der Poel;

Spain (33)

: Hospital Universitari Germans Trias i Pujol, Badalona — E. Davant, M. Monreal;

United Kingdom (2)

: Old-church Hospital, Romford — M. Quigley; Mount Vernon Cancer Centre, Northwood — G.J.S. Rustin, J. Boxall.


n engl j med

349;2

www.nejm.org july

10, 2003


153

references

1.

Hyers TM, Agnelli G, Hull RD, et al. An-tithrombotic therapy for venous thrombo-embolic disease. Chest 2001;119:Suppl:176S-93S.

2.

Hutten BA, Prins MH, Gent M, GinsbergJ, Tijssen JG, Buller HR. Incidence of recur-rent thromboembolic and bleeding compli-cations among patients with venous throm-boembolism in relation to both malignancyand achieved international normalized ra-tio: a retrospective analysis. J Clin Oncol2000;18:3078-83.

3.

Prandoni P, Lensing AW, Piccioli A, et al.Recurrent venous thromboembolism andbleeding complications during anticoagu-lant treatment in patients with cancer andvenous thrombosis. Blood 2002;100:3484-8.

4.

Weitz JI. Low-molecular-weight hepa-rins. N Engl J Med 1997;337:688-98. [Erra-tum, N Engl J Med 1997;337:1567.]

5.

Walsh-McMonagle D, Green D. Low-molecular-weight heparin in the manage-ment of Trousseau’s syndrome. Cancer1997;80:649-55.

6.

Eikelboom JW, Baker RI. Low-molecu-lar-weight heparin for the treatment ofvenous thrombosis in patients with adeno-carcinoma. Am J Hematol 1998;59:260-1.

7.

Luk C, Wells PS, Anderson D, KovacsMJ. Extended outpatient therapy with lowmolecular weight heparin for the treatmentof recurrent venous thromboembolism de-spite warfarin therapy. Am J Med 2001;111:270-3.

8.

Lensing AWA, Prandoni P, Brandjes D,et al. Detection of deep-vein thrombosis byreal-time B-mode ultrasonography. N Engl JMed 1989;320:342-5.

9.

Hull RD, Hirsh J, Carter CJ, et al. Pulmo-nary angiography, ventilation lung scanning,and venography for clinically suspected pul-monary embolism with abnormal perfusionlung scan. Ann Intern Med 1983;98:891-9.

10.

Remy-Jardin M, Remy J, Deschildre F, etal. Diagnosis of pulmonary embolism withspiral CT: comparison with pulmonary an-giography and scintigraphy. Radiology 1996;200:699-706.

11.

Qanadli SD, Hajjam ME, Mesurolle B, etal. Pulmonary embolism detection: prospec-tive evaluation of dual-section helical CTversus selective pulmonary arteriography in157 patients. Radiology 2000;217:447-55.

12.

Zubrod CG, Schneiderman M, Frei E III,et al. Appraisal of methods for the study ofchemotherapy of cancer in man: compara-tive therapeutic trial of nitrogen mustard andtriethylene thiophosphoramide. J ChronicDis 1960;11:7-33.

13.

Lee AY, Levine MN. Management ofvenous thromboembolism in cancer pa-tients. Oncology (Huntingt) 2000;14:409-17, 421.

14.

The Columbus Investigators. Low-molecular-weight heparin in the treatmentof patients with venous thromboembolism.N Engl J Med 1997;337:657-62.

15.

Levine M, Gent M, Hirsh J, et al. A com-parison of low-molecular-weight heparin ad-ministered primarily at home with unfrac-tionated heparin administered in the hospitalfor proximal deep-vein thrombosis. N Engl JMed 1996;334:677-81.

16.

Kaplan EL, Meier P. Nonparametric esti-mation from incomplete observations. J AmStat Assoc 1958;53:457-81.

17.

Mantel N. Evaluation of survival data andtwo new rank order statistics arising in itsconsideration. Cancer Chemother Rep 1966;50:163-70.

18.

Meyer G, Marjanovic Z, Valcke J, et al.Comparison of low-molecular-weight hepa-rin and warfarin for the secondary preven-tion of venous thromboembolism in patientswith cancer: a randomized controlled study.Arch Intern Med 2002;162:1729-35.

19.

Hull R, Pineo G, Mah A, Brant R. Long-term low molecular weight heparin treat-ment versus oral anticoagulant therapy forproximal deep vein thrombosis. Blood 2000;96:449a. abstract.

20.

Lopaciuk S, Bielska-Falda H, NoszczykW, et al. Low molecular weight heparin ver-sus acenocoumarol in the secondary pro-phylaxis of deep vein thrombosis. ThrombHaemost 1999;81:26-31.

21.

Pini M, Aiello S, Manotti C, et al. Lowmolecular weight heparin versus warfarin inthe prevention of recurrences after deep veinthrombosis. Thromb Haemost 1994;72:191-7.

22.

Veiga F, Escriba A, Maluenda MP, et al.Low molecular weight heparin (enoxaparin)versus oral anticoagulant therapy (aceno-coumarol) in the long-term treatment ofdeep venous thrombosis in the elderly: arandomized trial. Thromb Haemost 2000;84:559-64.

23.

Das SK, Cohen AT, Edmondson RA, Me-lissari E, Kakkar VV. Low-molecular-weightheparin versus warfarin for prevention of re-current venous thromboembolism: a ran-domized trial. World J Surg 1996;20:521-6.

24.

Gonzalez-Fajardo JA, Arreba E, Castro-deza J, et al. Venographic comparison of sub-cutaneous low-molecular weight heparinwith oral anticoagulant therapy in the long-term treatment of deep venous thrombosis.J Vasc Surg 1999;30:283-92.

25.

Palareti G, Legnani C, Lee A, et al. A com-parison of the safety and efficacy of oral anti-coagulation for the treatment of venousthromboembolic disease in patients with orwithout malignancy. Thromb Haemost 2000;84:805-10.

Copyright © 2003 Massachusetts Medical Society.


Randomized Comparison of Low Molecular WeightHeparin and Coumarin Derivatives on the Survival ofPatients With Cancer and Venous ThromboembolismAgnes Y.Y. Lee, Frederick R. Rickles, Jim A. Julian, Michael Gent, Ross I. Baker, Chris Bowden,Ajay K. Kakkar, Martin Prins, and Mark N. Levine

A B S T R A C T

PurposeExperimental studies and indirect clinical evidence suggest that low molecular weightheparins may have antineoplastic effects. We investigated the influence of a lowmolecular weight heparin dalteparin on the survival of patients with active cancer andacute venous thromboembolism.

Patients and MethodsSurvival data were examined in a posthoc analysis in patients with solid tumors andvenous thromboembolism who were randomly assigned to dalteparin or a coumarinderivative for 6 months in a multicenter, open-label, randomized, controlled trial.All-cause mortality at 12 months was compared between treatment groups in patientswith and without metastatic malignancy. The effect of dalteparin on survival wascompared between the two patient subgroups.

ResultsDuring the 12-month follow-up period, 356 of 602 patients with solid tumors and acutevenous thromboembolism died. Among patients without metastatic disease, the probabilityof death at 12 months was 20% in the dalteparin group, as compared with 36% in the oralanticoagulant group (hazard ratio, 0.50; 95% CI, 0.27 to 0.95; P � .03). In patients withmetastatic cancer, no difference in mortality between the treatment groups was observed(72% and 69%, respectively; hazard ratio, 1.1; 95% CI, 0.87 to 1.4; P � .46). The observedeffects of dalteparin on survival were statistically significantly different between patientswith and without metastatic disease (P � .02).

ConclusionThe use of dalteparin relative to coumarin derivatives was associated with improvedsurvival in patients with solid tumors who did not have metastatic disease at the time ofan acute venous thromboembolic event. Additional studies are warranted to investigatethese findings.

J Clin Oncol 23:2123-2129. © 2005 by American Society of Clinical Oncology

INTRODUCTION

Clinical evidence in support of anticoagu-lants having an antitumor effect was firstreported in a multicenter, randomized, con-trolled trial in 1981.1 In the Veterans AffairsResearch Service Cooperative Study 75, war-farin was found to be associated with animprovement in median survival in patients

with small-cell lung cancer who were receivingchemotherapy. Similarly, a randomized trialin the same patient population demonstrateda survival advantage for those patients treatedwith subcutaneous injections of unfraction-ated heparin.2 However, despite compellingexperimental evidence for a pathogenic roleof blood coagulation in tumor growth andmetastasis,3-6 other studies in patients with

From the McMaster University; TheHenderson Research Centre, Hamilton,ON, Canada; The George WashingtonUniversity and the Children’s NationalMedical Center, Washington, DC;University of Western Australia, Perth,Australia; Pfizer Inc, New York, NY;Imperial College, London, UnitedKingdom; and Academic Hospital ofMaastricht, Maastricht, the Netherlands.

Submitted March 18, 2004; acceptedAugust 23, 2004.

Supported by a New InvestigatorAward from the Canadian Institutes ofHealth Research/Rx&D ResearchProgram (A.Y.Y.L). M.N.L. is the BuffettTaylor Chair in Breast Cancer Research,McMaster University, Hamilton,Ontario, Canada.

Presented in part as a poster at the39th Annual Meeting of the AmericanSociety of Clinical Oncology,May 31-June 3, 2003, Chicago, IL.

Authors’ disclosures of potential con-flicts of interest are found at the end ofthis article.

Address reprint requests to Mark N.Levine, MD, Hamilton Health Science,Henderson Hospital, Room 9, 90Wing, 711 Concession St, Hamilton,ON L8V 1C3, Canada; e-mail:[email protected].

© 2005 by American Society of ClinicalOncology

0732-183X/05/2310-2123/$20.00

DOI: 10.1200/JCO.2005.03.133

JOURNAL OF CLINICAL ONCOLOGY O R I G I N A L R E P O R T

VOLUME 23 � NUMBER 10 � APRIL 1 2005

2123

Copyright © 2005 by the American Society of Clinical Oncology. All rights reserved. Downloaded from jco.ascopubs.org on May 21, 2007 . For personal use only. No other uses without permission.

solid tumors have failed to confirm a survival benefit for pa-tients treated with anticoagulants.7-10

More recently, the question of whether anticoagulantscan favorably influence the natural history of cancer hasreceived renewed attention. Randomized controlled trialsand meta-analyses of studies that compared low molecularweight heparins with unfractionated heparin for the initialtreatment of venous thromboembolism have reported areduction in the overall mortality of patients with cancerwho were randomly assigned to receive a low molecularweight heparin.11-15 Although the reduction in mortalityhas been consistent across studies and could not be attrib-uted to differences in fatal pulmonary embolism or bleed-ing, the observation that 5 to 7 days of low molecular weightheparin treatment reduced cancer mortality has been diffi-cult to explain. A plausible biologic mechanism, however, isnow emerging from experimental studies that show lowmolecular weight heparins can inhibit angiogenesis, a pro-cess that is critical for tumor growth and metastasis, in adose-dependent fashion.3,4,16,17

To date, two randomized, placebo-controlled trials de-signed to evaluate whether low molecular weight heparinscan improve survival in patients with advanced or incurablemalignancies have been completed.18,19 To examine theinfluence of a low molecular weight heparin relative tocoumarin derivatives on the survival of cancer patients withvenous thromboembolism and to investigate the hypothesisthat low molecular weight heparins have a greater impacton survival in cancer patients with limited disease than inthose with disseminated cancer, we performed a posthocanalysis of the mortality data in patients with solid tumorswho participated in the Comparison of Low MolecularWeight Heparin Versus Oral Anticoagulant Therapy forLong Term Anticoagulation in Cancer Patients With Ve-nous Thromboembolism (CLOT) trial.20

METHODS

Study Population

The CLOT trial was an international, multicenter, open-label, randomized trial that evaluated the relative efficacy andsafety of dalteparin (Fragmin; Pfizer, New York, NY), a low mo-lecular weight heparin, with oral anticoagulant therapy for theprevention of recurrent venous thromboembolism in patientswith cancer.20 Briefly, patients with cancer and acute venousthromboembolism were randomly assigned to 6 months of treat-ment with dalteparin alone or dalteparin followed by a coumarinderivative (warfarin or acenocoumarol). In the dalteparin group,patients received dalteparin once daily by subcutaneous injectionsat 200 U/kg for the first month followed by approximately 150U/kg for the subsequent 5 months. In the oral anticoagulantgroup, patients received dalteparin 200 U/kg once daily for the firstweek followed by an oral anticoagulant at doses that maintainedthe international normalized ratio between 2.0 and 3.0. The pri-mary efficacy outcome was symptomatic, recurrent venous

thromboembolism up to 6 months and patients were observed fordeath for up to 12 months. The study showed that dalteparinsignificantly reduced the risk of symptomatic, recurrent venousthromboembolism by 52% (P � .002) without increasing bleed-ing. At 6 months, 40% of the patients were dead and a difference inoverall mortality between treatment groups was not observed (P �.53). Ninety percent of the deaths in both groups were attributedto progressive cancer.

Statistical Analysis

The analysis was performed according to the intention-to-treat principle and based on the time from random assignment todeath. The extent of cancer in patients with solid tumors wasclassified at study enrollment by local investigators as metastaticdisease, localized disease without evidence of metastases, and noclinical evidence of active malignancy. Staging according to TNMclassification was not required at random assignment. An a prioridecision was made by the Steering Committee to combine the twolatter patient groups into a single group of patients without met-astatic disease for the purpose of this analysis because of the smallnumber of patients in each group.

The probability of death during the 12 months after ran-dom assignment was estimated according to the Kaplan-Meiermethod for each treatment group in patients with and withoutmetastatic disease.21 The difference in treatment-related sur-vival within the two subgroups of patients was compared usingthe two-sided log-rank test.22 Treatment-related survival wasalso analyzed with the exclusion of patients with specific tu-mor types if there were statistically significant imbalances inthe number of patients with such tumor types between thetreatment groups.

Cox proportional hazards regression models were used toadjust the treatment effect on survival for baseline factors for allpatients with solid tumors, and for the subgroups with and with-out metastases. These variables, identified a priori as potentiallyimportant predictors, and recorded at the time of randomization,included age (either continuous or by decade), sex, Eastern Coop-erative Oncology Group performance status, smoking status (everv never), qualifying episode of venous thromboembolism (pulmo-nary embolism v deep vein thrombosis), type of cancer treatment(radiation v none, chemotherapy v none), and major primarytumor site (breast, colorectal, lung, gynecologic, genitourinary,brain, pancreas, and other). Treatment and tumor site were forcedinto all regression models. Using a manual backward elimina-tion modeling strategy, a variable remained in the model if theassociated P value was less than .10 using both the Wald andscore tests. The residuals from the final models were inspectedfor possible outliers, influential observations, and unusual pat-terns. Hazard ratios and their corresponding 95% CIs wereestimated in the modeling process. To determine whether therewas a difference in the influence of dalteparin on mortalitybetween patients with and without metastatic disease, the un-adjusted hazard ratios were compared using a two-sided z test.A two-sided P value of less than .05 was considered to bestatistically significant. The Clinical Trials Methodology Groupat the Henderson Research Centre, Hamilton Health Sciences(Hamilton, ON, Canada) was responsible for data managementand statistical analyses. The Steering Committee was responsi-ble for supervising and providing final approval of these activ-ities and preparation of this article.

Lee et al

2124 JOURNAL OF CLINICAL ONCOLOGY


RESULTS

A total of 602 patients with solid tumors were included inthe CLOT study and data on their survival at 12 monthsafter random assignment were available for inclusion in theanalysis. The baseline characteristics of the 452 patientswith metastatic disease and the 150 patients without metas-tases are provided in Table 1. An assessment of the po-tential imbalance in each of the prognostic baselinefactors between treatment groups was undertaken foreach subgroup of patients with and without metastases.No statistically significant differences were observed be-tween treatment groups for baseline variables in patientswith metastatic disease. In the subgroup of patients with-out metastases, statistically significantly fewer patientswith lung cancer (P � .04) were treated with dalteparinthan with oral anticoagulant.

During the 12-month follow-up period, 174 of 296patients in the dalteparin group died, compared with 182 of306 patients in the oral anticoagulant group. The differencewas not statistically significant (P � .62; Fig 1). In patientswithout known metastases, 15 of 75 patients in the daltepa-rin group and 26 of 75 patients in the oral anticoagulantgroup died. The Kaplan-Meier estimate of the probability of

death at 12 months was 20% in the dalteparin group, com-pared with 36% in the oral anticoagulant group (Fig 2). Thedifference is statistically significant, with a hazard ratio of0.50 (95% CI, 0.27 to 0.95; P � .03), in favor of dalteparin.Adjusting for baseline prognostic factors did not change thefindings dramatically (adjusted hazard ratio, 0.41; 95% CI,0.19 to 0.86; P � .02). When patients with lung cancer wereexcluded from the analysis, the hazard ratio remained infavor of dalteparin (unadjusted hazard ratio, 0.63; 95% CI,

Table 1. Baseline Characteristics of the Patients

Characteristic

% Without Metastases % With Metastases

Dalteparin(n � 75)

OralAnticoagulant

(n � 75)Dalteparin(n � 221)

OralAnticoagulant

(n � 231)

Age, yearsMean 61 63 63 62Standard deviation 12 12 11 13

Female sex 51 49 59 52Current smoker 5 9 11 13Cancer treatment 67 67 71 68Qualifying DVT 59 60 73 71ECOG score

0 41 28 15 151 36 45 42 432 23 25 43 42

Tumor siteBreast 29 25 16 13Colorectal 15 13 19 19Lung 5� 17� 16 15Gynecologic 5 1 15 13Genitourinary 12 21 14 13Brain 17 15 1 1Pancreas 4 0 4 7Other 12 7 14 19

Abbreviations: DVT, deep vein thrombosis; ECOG, Eastern CooperativeOncology Group.

�Statistically significant fewer patients with lung cancer were treatedwith dalteparin than with oral anticoagulant in the group of patientswithout metastases (P � .04). No statistically significant differencesbetween treatment groups for baseline factors were detected in patientswith metastases.

Fig 1. Survival in patients with solid tumors. (– – – –) Oral anticoagulant(OAC); (———) dalteparin. Log-rank test P � .62 (two-sided) for the overallcomparison between the treatment groups.

Fig 2. Survival in patients with solid tumors according to the presence orabsence of metastatic disease. (– – – –) Oral anticoagulant (OAC); (———)dalteparin. For patients without metastatic disease, the hazard ratio was0.50 (95% CI, 0.27 to 0.95; P � .03) for the overall comparison between thetreatment groups. For patients with metastatic disease, the hazard ratio was1.1 (95% CI, 0.87 to 1.4; P � .46) for the overall comparison between thetreatment groups.

Anticoagulants and Cancer Survival

www.jco.org 2125


0.31 to 1.3; P � .19; adjusted hazard ratio, 0.37; 95% CI,0.17 to 0.83; P � .02).

In contrast, in patients with known metastatic malig-nancy, 159 of 221 patients assigned to dalteparin and 156 of231 patients allocated to oral anticoagulant died. The proba-bility of mortality at 12 months was 72% and 69%, respectively(Fig 2). The hazard ratio in the dalteparin group as comparedwith the oral anticoagulant group in patients with metastaticdisease was 1.1 (95% CI, 0.87 to 1.4; P � .46). A comparison ofthe two hazard ratios of dalteparin to oral anticoagulant be-tween the subgroups of patients with and without metastaticdisease was statistically significant (P � .02).

In the best-fitting regression model, the treatment ef-fect for dalteparin versus oral anticoagulant adjusted forstatistically significant baseline risk factors was 0.43 (95%CI, 0.21 to 0.89; P � .02) for patients without metastases,and 1.1 (95% CI, 0.91 to 1.4; P � .24) for patients withknown metastases. In addition, performance status, smok-ing, and treatment with chemotherapy were also indepen-dent predictors for survival (Table 2).

DISCUSSION

In this subgroup analysis of the CLOT trial, we examinedthe effect of a low molecular weight heparin dalteparin onthe survival of patients with cancer and venous thrombo-embolism. Although a difference in survival at 12 monthswas not observed for the entire study population and pa-tients with known metastases, we demonstrated a statisti-cally significant improvement in overall survival associatedwith dalteparin, relative to oral anticoagulant therapy, inpatients with solid tumors who were not known to havemetastatic disease at the time of their thromboembolicevent. The 50% relative risk reduction in the 12-monthmortality remained significant after adjusting for knownprognostic factors. A higher Eastern Cooperative OncologyGroup status and smoking predicted for higher mortality,whereas age, sex, and the type of thrombotic event did not.

The lack of effect on survival in patients with metastaticdisease suggests that the mechanisms of action of dalteparinmay be dependent on or interact with the stage of cancer orextent of tumor burden.

Although our results do not confirm a causal relation-ship between low molecular weight heparin and inhibitionof tumor growth or progression, they are consistent withprevious observations from clinical studies, including tworecently completed randomized, placebo-controlled trialsdesigned to investigate the influence of low molecularweight heparin on cancer survival.18,19 In the study byKakkar et al18 (Fragmin Advanced Malignancy OutcomeStudy [FAMOUS] trial), which randomly assigned 382 pa-tients with metastatic or advanced solid malignancies toonce-daily injections of dalteparin or placebo for 1 year, asurvival difference between treatment groups was not ob-served. The 1-year survival was 45% and 42%, respectively(P � .29). However, in those patients who survived beyond17 months, an improvement in survival was observed inpatients who received dalteparin (P � .04). In the Malig-nancy and Low Molecular Weight Heparin Therapy(MALT) trial reported by Klerk et al,19 approximately 300patients with incurable solid tumors were randomly as-signed to the low molecular weight heparin nadroparinor placebo for 6 weeks. A statistically significant im-provement in overall survival was observed for nadropa-rin relative to placebo. The reduction in mortality wasalso in favor of nadroparin in the subgroup of patientswho were identified as having a life expectancy of greaterthan 6 months. In all three studies, the use of low molec-ular weight heparin was associated with improved sur-vival in patients with relatively good prognosis.

There are also important differences between our studyand these additional trials. First, patients in the above-mentioned two studies did not have acute venous throm-boembolism at the time of random assignment. Given thestrong association between coagulation and tumor biology,the presence of venous thromboembolism could have an

Table 2. Hazard Ratios� Based on Multivariable Cox Proportional Hazards Modeling of Clinical Risk Factors for Mortality During AnticoagulantTherapy With Low Molecular Weight Heparin or Oral Anticoagulant

Risk Factor

Without Metastases With Metastases

Hazard Ratio 95% CI P † Hazard Ratio 95% CI P †

Dalteparin versus oral anticoagulants‡ 0.4 0.2 to 0.9 .02 1.1 0.9 to 1.4 .24ECOG (0, 1, 2) 2.0 1.2 to 3.3 .007 1.6 1.3 to 1.8 .0001Smoker (ever v never) 2.8 0.96 to 8.3 .06 1.5 1.1 to 2.1 .01Chemotherapy (yes v no) —§ — — 0.8 0.6 to 0.95 .02

Abbreviations: VTE, venous thromboembolism; ECOG, Eastern Cooperative Oncology Group.�Hazard ratios for age, gender, qualifying episode of VTE, and radiotherapy did not meet statistical significance in either subgroup of patients.†Two-sided Wald test.‡Comparison was also adjusted for major tumor sites (see Statistical Analysis).§The hazard ratio for chemotherapy was not statistically significant in patients without metastases.

Lee et al



influence on the effect, if any, of low molecular weightheparins on tumor progression. Second, our patients mighthave had a poorer prognosis than those in the other trialsbecause cancer patients with venous thromboembolismhave a shorter life expectancy than similar cancer patientswithout thrombosis.23,24 The short life expectancy of ourpatients with metastatic disease could have limited the de-tection of a survival benefit associated with any therapy.Another difference among the studies is the treatment reg-imens used in the experimental and control groups. In ourstudy, full therapeutic doses of dalteparin were adminis-tered for the first month followed by 75% of the full dose for5 months, whereas in the FAMOUS study prophylactic doseof dalteparin was used, and in the MALT study nadroparinwas given at therapeutic doses for 2 weeks followed by halfof this dose for 4 weeks. Lastly, the CLOT study differs fromthe FAMOUS and MALT trials in the duration of follow-up.In the latter trials, improvement in overall survival was notobserved until at least 1 year after randomization. There-fore, it is possible that a survival benefit might becomeevident in patients with metastatic disease in the CLOT trialwith longer follow-up.

The major limitation of our analysis is the potentialimbalance of prognostic factors between treatment groups.Although our study was a randomized trial, stratificationfor tumor type and extent of disease was not performedbecause the primary outcome in the CLOT trial was notsurvival. In addition, we could not control for differences inprevious or concurrent antineoplastic therapy. We did ex-amine the possibility that the difference in the number ofpatients with lung cancer treated with dalteparin or oralanticoagulant in patients without metastases might haveinfluenced the findings. Reanalyzing the results with exclu-sion of patients with lung cancer also produced a statisti-cally significant hazard ratio in favor of dalteparin whenadjusted for prognostic baseline factors. Lastly, we cannotexclude the possibility that our observations are due tochance and imbalance of unknown prognostic variables.Therefore, the results of this subgroup analysis should beinterpreted with caution.25

The mechanisms for a potential antineoplastic effect oflow molecular weight heparins remain unknown and willrequire further investigations in well-designed experimen-tal studies. An antiangiogenic effect is an appealing possi-bility and is compatible with our observation that a survivalbenefit was evident in patients with limited disease andpersisted beyond the administration of the agent.26 It can behypothesized that in patients with disseminated cancer,tumor-related vasculature is sufficiently developed so thatan antiangiogenic agent would have minimal impact,whereas impairing the establishment of such vasculature byan antiangiogenic agent could exert an inhibitory effect ontumor growth even beyond the time of drug exposure.Although it has been suggested that the improvement in

cancer survival associated with low molecular weight hepa-rins observed in previous trials may be due to a reduction infatal pulmonary embolism as compared with unfraction-ated heparin, the survival benefit beyond the period of lowmolecular weight heparin administration observed in ourstudy would argue against this hypothesis.

A strong association between cancer and thrombosishas been demonstrated consistently in experimental andclinical studies. Our results offer additional evidence thatthe coagulation system is intrinsically involved in tumor-igenesis or tumor progression. Future studies designed toconfirm the antitumor effects of low molecular weightheparins and explore the pathophysiological mecha-nisms are awaited.

■ ■ ■

Acknowledgment

We thank C.P.W. Klerk and H.R. Buller for their help-ful review of the manuscript.

Appendix

The following investigators and institutions participatedin the CLOT Trial: Steering Committee: M. Levine (Chair),R. Baker, C. Bowden, M. Gent, A. Kakkar, A. Lee, M. Prins,F. Rickles. External Safety and Efficacy Monitoring Com-mittee: J. Pater (Chair), H. Buller, S. Goldhaber. CentralAdjudication Committee: J. Ginsberg, J. Hirsh, C. Kearon,G. Thomson, J. Weitz. Coordinating and Methods Centre:Clinical Trials Methodology Group, Henderson ResearchCentre, Hamilton, ON, Canada–J. Julian, S. Haley, A. Ling,Q. Guo, B. Rush, T. Finch, L. Bonilla-Escobedo, L.Matthews, J. Windsor, C. Tavormina, H. Nelson, G. Lewis,J. Sicurella. Clinical Centers (the number of patients contrib-uted from each country follows the country): Canada (225)—Hamilton Health Sciences, Henderson Hospital, Hamilton,ON—A. Lee, N. Booker, S. Schmidt; London Health Sci-ences Centre, London, ON—M. Kovacs, B. Morrow; QueenElizabeth II Health Sciences Centre, Halifax, NS—B.McCarron, S. Pleasance; Toronto General Hospital, To-ronto, ON—W.F. Brien, S. Boross-Harmer; St. Joseph’sHospital, Hamilton, ON—J.D. Douketis, T. Schnurr; TheMontreal General Hospital, Montreal, PQ—S. Solymoss, B.St. Jacques; Sunnybrook & Women’s College Health Sci-ences Centre, Toronto, ON—W. Geerts, K. Code; BritishColumbia Cancer Agency, Vancouver Cancer Centre, Van-couver, BC—S. Chia, S. Monkman; Hamilton Health Sci-ences, Hamilton General Hospital, Hamilton, ON—A.G.G.Turpie, J. Johnson; Kelowna General Hospital, Kelowna,BC—J. Sutherland, S. Shori; Australia (144) and New Zea-land (16)—Australasian Society of Thrombosis and He-mostasis—Royal Perth Hospital, Perth, WA—R. Baker, J.Smith; Flinders Medical Centre, Bedford Park, SA—D.W.Coghlan, J.M. Osmond; Prince of Wales Hospital, Rand-wick, NSW—S. Dunkley, B. Chong; Box Hill Hospital,


www.jco.org 2127


Monash University, Box Hill—H. Salem, L. Poulton; West-mead Hospital, Westmead, NSW—M. Hertzberg, P.Stavros; Auckland Hospital, Auckland—P. Ockelford, V.Rolfe-Vyson; St. George Hospital, Kogarah, NSW—T. A.Brighton, R. Ristuccia; Royal North Shore Hospital, Uni-versity of Sydney, Sydney, NSW—C.M. Ward, K. Sheather;Royal Adelaide Hospital, Adelaide, SA—I.N. Olver, T.Marafioti; St. Vincent’s Hospital, Sydney, NSW—D. Ma;Monash Medical Centre, Clayton, VIC—T. E. Gan, A.Cummins; Royal Melbourne Hospital, Parkville, VIC—A.Grigg, E. Cinc; United States (118) —University of South-ern California, Keck School of Medicine, Los Angeles,CA—H. Liebman, I. Weitz; University of Texas, M.D.Anderson Cancer Center, Houston, TX—C.P. Escalante, P.Horace; Northwestern University, Chicago, IL—D. Green,M. Calimaran; University of North Carolina at Chapel Hill,Chapel Hill, NC—S. Moll, S.K. Jones; Arizona Cancer Cen-tre, University of Arizona, Tucson, AZ—A. Stopeck, K.Glennie; Atlanta VA Medical Centre/Emory Universi-ty—M. Ribeiro, L. Starke; The Cleveland Clinic Founda-tion, Cleveland, OH—S.R. Deitcher; Mt. Sinai MedicalCenter, New York, NY—L. Lipsey; St. Joseph Mercy Oak-land, Pontiac, MI—A. Brady, R. Krishnan; University ofVermont & Fletcher Allen Health Care, Burlington,VT—M. Cushman, L. Chassereau; University of VirginiaHealth System, Charlottesville, VA—B.G. Macik, L. New-ton; Lovelace Health System, Albequrque, NM—A.Tarnower, R.J. Weiler; Newark Beth Israel Medical Center,Newark, NJ—A.J. Cohen, E. White; University of Connect-icut, Farmington, CT—R. Bona, K. Jennings; Italy (67)—Ospedali Riuniti, Bergamo—A. Falanga, R. Labianca;Clinica Medica II, University of Padua, Padua—P.Prandoni, A. Piccioli, E. Zanon; Angelo Bianchi BonomiHemophilia Thrombosis Center, University of Milan and

National Cancer Institute of Milan—A.B. Federici, G.Pizzocaro; the Netherlands (41)—Academic Medical Cen-ter of the University of Amsterdam, Amsterdam—S.M.Smorenburg, C.P.W. Klerk; University Hospital Nymegen,Nymegen—F. v.d. Berkmortel, DJTh Wagener; MaaslandHospital, Sittard—F.L.G. Erdkamp; St. Elisabeth Hospital,Tilburg—C. van der Heul, C. Post; St. Antonius Hospi-tal, Nieuwegein—D.H. Biesma; Van Weel BethesdaHospital, Dirksland—C. Kroon, M. Kamphuis van derPoel; Spain(33)—Hospital Universitari Germans Trias iPujol, Badalona—E. Davant, M. Monreal; United Kingdom(2)—Oldchurch Hospital, Romford, Essex—M. Quigley;Mount Vernon Cancer Centre, Northwood, Middlesex—G.J.S. Rustin, J. Boxall.

Authors’ Disclosures of Potential

Conflicts of Interest

The following authors or their immediate family mem-bers have indicated a financial interest. No conflict exists fordrugs or devices used in a study if they are not being evalu-ated as part of the investigation. Owns stock (not includingshares held through a public mutual fund): Chris Bowden,Bristol-Myers Squibb. Acted as a consultant within the last2 years: Frederick R. Rickles, Pharmacia/Pfizer; Ajay K.Kakkar, Aventis, Pfizer; Agnes Y.Y. Lee, Aventis, LEOPharma, Pharmacia/Pfizer, Sanofi, Wyeth; Mark N. Levine,Pfizer. Performed contract work within the last 2 years:Frederick R. Rickles, Pharmacia/Pfizer. Received more than$2,000 a year from a company for either of the last 2 years:Frederick R. Rickles, Pharmacia/Pfizer; Ajay K. Kakkar,AstraZeneca, Aventis, Pfizer; Agnes Y.Y. Lee, Pharmacia/Pfizer; Chris Bowden, Bristol-Myers Squibb; Mark N.Levine, Pharmacia/Pfizer.

REFERENCES

1. Zacharski LR, Henderson WG, Rickles FR,et al: Effect of warfarin on survival in small cellcarcinoma of the lung: Veterans AdministrationStudy No. 75. JAMA 245:831-835, 1981

2. Lebeau B, Chastang C, Brechot JM, et al:Subcutaneous heparin treatment increases sur-vival in small cell lung cancer: “Petites Cellules”Group. Cancer 74:38-45, 1994

3. Nash GF, Walsh DC, Kakkar AK: The roleof the coagulation system in tumour angiogene-sis. Lancet Oncol 2:608-613, 2001

4. Mousa SA: Anticoagulants in thrombosisand cancer: The missing link. Semin ThrombHemost 28:45-52, 2002

5. Rickles FR, Patierno S, Fernandez PM:Tissue factor, thrombin, and cancer. Chest 124:58S-68S, 2003 (suppl 3)

6. Rickles FR, Levine MN, Dvorak HF: Abnor-malities of hemostasis in malignancy, in ColmanRW, Hirsh J, Marder VJ, et al (eds): Hemostsisand Thrombosis. Philadelphia, PA, LippincottWilliams & Wilkins, 2001, pp 1131-1152

7. Maurer LH, Herndon JE, Hollis DR, et al:Randomized trial of chemotherapy and radiationtherapy with or without warfarin for limited-stagesmall-cell lung cancer: A Cancer and LeukemiaGroup B study. J Clin Oncol 15:3378-3387, 1997

8. Levine M, Hirsh J, Gent M, et al: Double-blind randomised trial of a very-low-dose warfa-rin for prevention of thromboembolism in stageIV breast cancer. Lancet 343:886-889, 1994

9. Smorenburg SM, Hettiarachchi RJ, Vink R,et al: The effects of unfractionated heparin onsurvival in patients with malignancy: A system-atic review. Thromb Haemost 82:1600-1604,1999

10. Smorenburg SM, Vink R, Otten HM, et al:The effects of vitamin K-antagonists on survivalof patients with malignancy: A systematic anal-ysis. Thromb Haemost 86:1586-1587, 2001

11. Prandoni P, Lensing AW, Buller HR, et al:Comparison of subcutaneous low-molecular-weight heparin with intravenous standard hepa-rin in proximal deep-vein thrombosis. Lancet339:441-444, 1992

12. Green D, Hull RD, Brant R, et al: Lowermortality in cancer patients treated with low-

molecular-weight versus standard heparin. Lan-cet 339:1476, 1992

13. Hettiarachchi RJ, Smorenburg SM, GinsbergJ, et al: Do heparins do more than just treatthrombosis? The influence of heparins on cancerspread. Thromb Haemost 82:947-952, 1999

14. Gould MK, Dembitzer AD, Doyle RL, et al:Low-molecular-weight heparins compared withunfractionated heparin for treatment of acutedeep venous thrombosis: A meta-analysis ofrandomized, controlled trials. Ann Intern Med130:800-809, 1999

15. Lensing AW, Prins MH, Davidson BL, et al:Treatment of deep venous thrombosis with low-molecular-weight heparins: A meta-analysis ArchIntern Med 155:601-607, 1995

16. Norrby K: Heparin and angiogenesis: Alow-molecular-weight fraction inhibits and a high-molecular-weight fraction stimulates angiogene-sis systemically. Haemostasis 23:141-149, 1993(suppl 1)

17. Hejna M, Raderer M, Zielinski CC: Inhibi-tion of metastases by anticoagulants. J NatlCancer Inst 91:22-36, 1999

Lee et al



18. Kakkar AK, Levine MN, Kadziola Z, et al:Low molecular weight heparin, therapy withdalteparin, and survival in advanced cancer: TheFragmin Advanced Malignancy Outcome Study(FAMOUS). J Clin Oncol 22:1944-1948, 2004

19. Klerk CPW, Smorenburg SM, OttenJMMB, et al: Malignancy and low-molecular-weight heparin therapy: The MALT trial. Pre-sented at Intl Soc Thromb Haemost XIXInternational Congress, Birmingham, UK, July12-18, 2003

20. Lee AY, Levine MN, Baker RI, et al: Low-molecular-weight heparin versus a coumarin for

the prevention of recurrent venous thromboem-bolism in patients with cancer. N Engl J Med349:146-153, 2003

21. Kaplan E, Meier P: Nonparametric estima-tion from incomplete observations. J Am StatAssoc 53:457-481, 1958

22. Mantel N: Evaluation of survival data andtwo new rank order statistics arising in its consid-eration. Cancer Chemother Rep 50:163-170, 1966

23. Levitan N, Dowlati A, Remick SC, et al: Ratesof initial and recurrent thromboembolic diseaseamong patients with malignancy versus those with-

out malignancy: Risk analysis using Medicare claimsdata. Medicine (Baltimore) 78:285-291, 1999

24. Sorensen HT, Mellemkjaer L, Olsen JH, et al:Prognosis of cancers associated with venous throm-boembolism. N Engl J Med 343:1846-1850, 2000

25. Assmann SF, Pocock SJ, Enos LE, et al:Subgroup analysis and other (mis)uses of baselinedata in clinical trials. Lancet 355:1064-1069, 2000

26. Boehm T, Folkman J, Browder T, et al:Antiangiogenic therapy of experimental cancerdoes not induce acquired drug resistance. Na-ture 390:404-407, 1997


www.jco.org 2129


Division of Hematology-Oncology, Department of Medicine, GWU Medical Faculty Associates, 2150 Pennsylvania Ave, NW, Washington, DC 20037, USA Telephone: (202)741-2478 Fax: (202)741-2487Email: [email protected]

P103

Journal of Pathophysiology of Haemostasis and Thrombosis

Mechanisms of Cancer-Induced Throm-bosis

Frederick R. Rickles

The George Washington University School of Medicine and Health Sciences Washington, DC 20037, USA

Key Words Platelets · Fibrin · Thrombus · Murine · Atherothrobsis · Microcirculation · Artery · Vessel · Wall Abstract Substantial epidemiologic, laboratory, pathologic and clinical evidence supports the historic association be-tween activation of blood coagulation and progression of cancer. The increased risk for venous throm-boembolism (VTE) in cancer has been considered an epiphenomenon. However, recent studies from sev-eral laboratories have linked more closely malignant transformation (oncogenesis), tumor angiogenesis and metastasis to the generation of clotting interme-diates (e.g. tissue factor [TF], factor Xa and throm-bin), clotting or platelet function inhibitors (e.g. COX-2) or fibrinolysis inhibitors (e.g. plasminogen activator inhibitor, type 1 [PAI-1]). Furthermore, TF, Xa and thrombin may induce important tumor cell signaling cascades in a clotting-dependent and/or clotting-

independent manner (e.g. thru engagement of prote-ase-activated receptors [PARs]). Targeting blood clotting reactions in cancer may provide a unique ap-proach to cancer treatment. Copyright © 2006 S. Karger AG, Basel

Introduction

Activation of blood coagulation occurs commonly in patients with cancer and leads to the development of ve-nous thromboembolism (VTE) at an overall rate as high as seven times as often as in age-matched, sex-matched, con-trol subjects (1). The pathophysiology of thrombosis in cancer is complex but can be viewed classically as related to abnormalities of so-called Virchow’s triad: [1] stasis of the blood; [2] vascular injury; [3] hypercoagulability (or, as described by Virchow, “abnormalities of the fixed ele-ments of the blood”), the terminology for which has been updated and fits well the situation with many cancer pa-tients (2). Substantial epidemiologic, laboratory, patho-

P104

logic and clinical evidence supports this important associa-tion but, until recently, VTE was thought largely to be an epiphenomenon in cancer. That is, cancer and the treat-ment of cancer were assumed to be the proximate causes of the increased risk for thrombosis but thrombus formation itself was not considered an intrinsic molecular event in oncogenesis.

Recent evidence from several laboratories (3-5), how-ever, has linked more closely malignant transformation, tumor angiogenesis and metastasis to thrombus formation, mediated perhaps by signaling cascades that can be trig-gered in a clotting-dependent and/or clotting-independent manner. Tissue factor (TF), the ubiquitous activator of blood clotting, has been shown, for example, to induce production of vascular endothelial growth factor (VEGF) in human tumor cells independent of its ability to activate factor Xa-catalyzed conversion of prothrombin (6). The TF-VIIa complex and factor Xa are among known activa-tors of G-protein-coupled protease-activated receptor-2 (PAR-2) in tumor cells, while the TF-VIIa-Xa complex and thrombin efficiently activate PAR-1. Both PARs have been implicated in signaling pathways leading to angio-genesis and metastasis (7-9). The precise role of the cyto-plasmic domain of TF, which has been targeted as the likely signaling region of the molecule (perhaps due to the presence of 3 serine residues in the cytoplasmic tail of TF), remains controversial (6,8). Nevertheless, multiple cell-signaling cascades are triggered during the generation of these clotting intermediates that are believed to influence tumor cell migration, adhesion, cell-cell interaction, dia-pedesis and replication, as well as new vessel growth.

Taking advantage of epidemiologic information on the negative impact of activation of blood clotting in cancer patients (10,11), investigators are now exploring with re-newed interest the potential that inhibitors of blood clotting can impact on cancer survival. Recently published ran-domized, controlled trials have documented increased sur-vival in patients with advanced cancer treated with various forms of low-molecular-weight heparin (LMWH) (12-16). While the mechanisms for the salutary effects of antico-agulants on cancer survival have yet to be elucidated, some evidence suggests that various heparin fractions can inter-fere with a variety of important tumor functions, including, for example: [1] tumor angiogenesis; [2] heparanase-mediated extravasation of blood-borne tumor cells; and, [3] interactions between carcinoma mucins and selectins.

Targeting blood clotting reactions in cancer may pro-vide a unique approach to treatment. New agents are being developed that can activate clotting selectively only in tu-mor vessels, thus producing localized infarction and reduc-tion in tumor size (17-20). Other approaches target tumor cell and tumor-associated endothelial cell TF to deliver immuno-toxins or selective inhibitors of tumor growth with specificity to spare the surrounding normal tissue

(21,22). Both selective activation and inhibition of clotting may prove to be useful adjuncts to traditional therapeutic approaches in cancer patients.

Epidemiology

As noted, cancer significantly increases the risk for VTE and this risk is greatest within the first 3-6 months following the diagnosis of cancer (1). Among all patients in whom the diagnosis of VTE is established, cancer is re-sponsible for approximately 20% of the attributable risk (23) and patients with cancer are at nearly 3 times the risk for recurrence of VTE as are matched control subjects after a first episode (24). Although the absolute risk of VTE in cancer patients remains relatively small (e.g. 0.6% in one large population-based study) (11), cancer patients under-going surgery have at least a 2-fold increase risk of postop-erative thrombosis compared to non-cancer controls under-going the same procedures (25), and even without the added stress of surgery, the adjusted odds ratio for VTE is nearly 30 in patients with some tumor types (1). Finally, the risk of death due to cancer appears to increase signifi-cantly if the diagnosis of VTE is established simultane-ously or within the first year (10,11), supporting long-standing experimental data that suggest that thrombus for-mation provides some biological advantage to the growth and dissemination of cancer (2,26).

Mechanisms of Cancer-Induced Thrombosis (Table 1)

Stasis, vascular injury and hypercoagulability, the 3 parts of Virchow’s Triad (27), all play roles in the patho-genesis of thrombus formation in cancer patients (Table 1). Some or all of the following, each grouped under one of aspect of the triad, can complicate the course of disease in the cancer patient. STASIS: [1] bed rest; [2] extrinsic compression of blood vessels by tumor; VASCULAR IN-JURY: [1] direct invasion of vessels by tumor; [2] pro-longed use of central venous catheters; [3] endothelial damage secondary to chemotherapy; HYPERCOAGULA-BILITY:[1] release of tumor-associated procoagulants and cytokines; [2] impaired endothelial cell defense mecha-nisms and reduction of naturally occurring inhibitors (e.g. antithrombin, protein C or protein S deficiency; activated protein C resistance); [3] increased adhesive interactions between tumor cells, vascular endothelial cells, platelets and host monocyte/macrophages, further enhanced in mucin-secreting tumors (e.g. selectin-mediated). These

P105

and other probable mechanisms for thrombosis in cancer patients have been examined in detail elsewhere (2,26,28). Suffice it to say that the pathogenesis of thrombosis in can-cer is complex and most likely involves multiple mecha-nisms that may differ dramatically from patient to patient. However, until recently, it has been presumed that all of these mechanisms are secondary and have no primary role in the molecular events leading to the development of can-cer.

Molecular Pathogenesis of Thrombosis in Cancer – Direct Link to Oncogenesis (Table 2)

Boccaccio and her colleagues developed a model for human liver carcinoma by targeting activated human MET oncogene to mouse liver with a lentivrial vector and liver-specific promoter (3). The animals developed slowly, pro-gressive hepatocarcinogenesis, which was preceded and accompanied by a thrombohemorrhagic syndrome ulti-mately indistinguishable from Trousseau’s Syndrome with disseminated intravascular coagulation (DIC). Venous thrombosis in the tail vein of the mouse occurred early and was followed by a progressive coagulopathy and fatal in-ternal hemorrhage. The syndrome was characterized in the animals by elevated blood levels of fibrin d-dimer, a pro-longed prothrombin time and a marked reduction of the platelet count (i.e. DIC). Genome-wide expression profil-ing of hepatocytes expressing the MET oncogene demon-strated impressive upregulation of both the plasminogen activator inhibitor 1 (PAI-1) and cyclooxygenase-2 (COX-2) genes with two to three-fold increase in circulating pro-tein levels. Inhibitors of either PAI-1 (XR5118) or COX-2 (Rofecoxib®) prevented both laboratory and clinical evi-dence of DIC in the mice.

Based on their observations in this mouse model the in-vestigators postulated a 5-step process (Figure 1), whereby MET induction by hypoxia results in increased expression of the tyrosine kinase receptor for hepatocyte growth fac-tor/scatter factor (step 1), activation of prothrombotic, he-mostasis genes (e.g. COX-2 and PAI-1) by MET signaling (step 2) with fibrin nesting of the tumor cells, formation of a fibrin provisional matrix for the tumor (step 3), induction of neo-angiogenesis by the fibrin matrix and other coagula-tion proteases (step 4) and, finally, “scattering” of tumor cells with invasion (step 5) (29). Thus, for the first time, experimental evidence has been generated linking directly activation of hemostasis with oncogenesis.

Inactivation of the tumor suppressor gene Pten, to-gether with hypoxia, which leads to Akt activation and upregulation of the Ras/MEK/ERK signaling cascade, has

recently been shown to induce TF gene expression in hu-man astrocytoma cell lines (4), an in vitro model for ma-lignant transformation. Cells transformed with Akt showed the greatest incremental increase in hypoxia-induced TF expression and secretion, exhibiting procoagu-lant activity which was factor VII-dependent and inhibited with anti-TF antibodies. These findings were partially re-versed by induction of PTEN. PTEN is inactivated by mu-tation, promoter methylation, or by other mechanisms in upwards of 80% of human glioblastomas and TF is ex-pressed in >90% of such tumors. In low grade astrocy-tomas, however, only 10% of tumors express TF and evi-dence for PTEN inactivation is rarely found. The histopa-thologic hallmark of high grade astrocytomas (glioblas-toma multiforme) is so-called pseudopalisading necrosis, thought to represent a “wave of tumor cells actively mi-grating away from a central hypoxic zone that is created following vascular compromise and associated with in-travascular thrombosis.” (4). Thus, the finding that pseu-dopalisading cells produced increased amounts of TF in 7 human glioblastoma specimens further supports the mo-lecular relationship between malignant transformation and clotting activation.

Yu and colleagues (5) examined another model of hu-man oncogenesis to test the hypothesis that activation of clotting is a direct result of the molecular transforming events. In the well-described, step-wise model of human colorectal cancer (CRC), in which activation of mutant K-ras and subsequent inactivation/loss of p53 “drive many interrelated aspects of the malignant phenotype….”, the au-thors demonstrated that “TF is required for full expression of the K-ras-dependent tumorigenic and angiogenic pheno-type of CRC cells in vivo, but not for cellular transforma-tion in vitro.”(5) Conversely, activation of K-ras and loss of p53 were both necessary to achieve full expression of TF both on the cell membrane and on the surface of mi-crovesicles shed into the circulation of the mouse. With each subsequent mutation, increasing TF levels were documented in the respective cell lines in vitro and in vivo in the more aggressive tumors produced in SCID mice, which were shown to have acquired a new K-ras mutation. Finally, the investigators showed that TF gene-silencing, small interfering RNA (siRNA) markedly inhibited in vivo tumor growth and angiogenesis in Matrigel plugs in vivo.

Thus, 3 different tumor model systems have now pro-vided complimentary evidence that oncogene activation and/or tumor suppressor gene inactivation upregulates blood clotting in vivo (by increasing TF, PAI-1 and COX-2), strongly implicating clotting pathways in the basic bi-ology of cancer. Furthermore, this data implies that target-ing clotting intermediates might prove to be a rationale strategy for both reducing the thrombotic risk in cancer and, perhaps, impairing tumor growth.

P106

Targeting Tissue Factor and other Coagula-tion Intermediates in Cancer Therapy

Two more recent therapeutic approaches, one applied to experimental tumor models and one applied to patients with advanced malignancy, support the rationale for further testing of “anticoagulant”, anti-tissue factor strategies for treating cancer. One approach utilizes either a modified TF construct to induce local tumor infarction (20), or tar-gets TF in tumor cells and/or tumor vascular endothelium to induce direct tumor lysis (22). This method exploits dif-ferential TF expression in tumors and tumor-associated en-dothelium. Modifications of this therapeutic strategy takes advantage of the high affinity binding of factor VIIa to TF and the differential expression of TF on tumor cells and tumor-associated vascular endothelium (31,32). Along with other endothelial-specific targets, investigators have succeeded in treating effectively experimental tumors by inducing local infarction (17-21) (Table 3). The second approach utilizes a more traditional anticoagulant strategy to reduce the effects of thrombin generation and other clot-ting intermediates on tumor growth, angiogenesis and me-tastasis.

The first approach, which has been effective, is to de-liver a “toxic” construct that binds to TF in the neoangio-genic vasculature of the tumor, activates local blood coagu-lation, produces an obstructive thrombus and tumor infarc-tion. Reported most recently by El-Sheikh and colleagues (20), the investigators linked a soluble, truncated form of TF (tTF) to the heparin-binding domain of VEGF, the lat-ter of which targets at least 3 receptors hyperexpressed on tumor endothelium – the VEGF receptor, neuropilin 1 and chondroitin 6 sulfate proteoglycan (Table 3). This ap-proach was successful in inducing selective tumor necrosis quite rapidly (5 minutes) after infusion of the construct, to-gether with recombinant factor VIIa, into mice harboring the CT26 colon cancer (20), apparently sparing non-tumor vasculature and avoiding systemic activation of coagula-tion.

Utilizing another modification of this strategy, which is not designed to activate TF in the tumor endothelium but instead targets the genes for TF and VEGF, we have dem-onstrated regression of human breast tumors in nude mice (Figure 2). We reported the use of synthetic analogs of curcumin, which are potent inhibitors of TF and VEGF ex-pression in both tumor cells and vascular endothelial cells. We demonstrated marked reduction of the expression of both gene products in MDA-MB-231 cells and in other human tumor cell lines, as well as in endothelial cells (22). More recent data demonstrates the feasibility of linking the lead compound among these synthetic analogs to rVIIai to target the tumor endothelium and tumor cells, whereby we achieved upwards of 60% uptake of the labeled complex

into malignant cells. One or more of these targeting strate-gies may prove useful in cancer chemotherapy. Of interest, the antiangiogenic rationale, using a factor VIIa cassette for targeting TF in the neoangiogenic endothelium, is cur-rently being tested in Phase II studies of patients with “wet” macular degeneration (33)

Fig. 2. Regression of human breast tumors in nude mice after treatment with compound 14. MDA-MB-231 (1 x 106 cells) were inoculated int eh flanks of nude mice and tumors were allowed to grow for 3 weeks. Compound 14 was adminis-tered by injection near the base of the tumors 3x/wk x 2 weeks. The bar graph dem-onstrates a dose-dependent decrease in tumor weight after treatment (*p < 0.05; n = 5 mice/group). (Reproduced from reference 22, with permission of the publishers).

The direct anticoagulant strategy of cancer treatment

focuses typically on LMWH compounds or other newer anticoagulants. The initial impetus for the use of these drugs was for the treatment and/or prevention of VTE and the studies generally compared their safety and efficacy to unfractionated heparin (UFH). Retrospective analyses of the cancer patient subset in many of the larger trials re-vealed that patients treated for a short time with relatively low (prophylactic) doses of LMWH, demonstrated a lower 3 month mortality rate than those treated with UFH (34). Based on this type of “hypothesis-generating” data, a num-ber of randomized controlled trials (RCTs) have been un-dertaken to test the benefit of LMWH in cancer patients, with and without VTE, and results of 6 studies have been published within the past few years (12-15, 35). In total, 1,787 patients with advanced cancer were randomized to treatment with a LMWH or control (no anticoagulant, pla-cebo, unfractionated heparin or warfarin). While the study designs were different, one common outcome was that in each study, the mortality rate for the patients treated with LMWH was improved. In the two trials in which survival was the primary endpoint, the control group received no anticoagulant treatment and the apriori statistical design was adequate to support the conclusion (14,16), patients treated with LMWH had a significantly prolonged median progression-free survival (p = 0.01) and median survival (p = 0.021), respectively.

More well-designed studies are needed but the results of these 6 trials across a broad group of patients with poor prognosis cancers have stimulated renewed interest in the

P107

anticoagulant approach, particularly since the increased survival observed was not due to prevention of fatal pul-monary emboli or other thrombotic complications. In-stead, these patients lived longer with their tumors and ul-timately died of cancer, lending support to the concept that LMWH affects tumor biology.

Less certain is the mechanism(s) by which LMWH ef-fects this change in tumor behavior. As indicated earlier, possible mechanisms include (but are not limited to): [1] inhibition of angiogenesis (Figure 3) (2,26,30,36-38); inhi-bition of tumor proteases (e.g. heparanase)(39,40); inhibi-tion of selectins (41). It is very likely that these and other mechanisms are impacted by LMWH and are overlapping – e.g. inhibition of thrombin generation and thrombin ac-tion, for example, likely reduces not only its effects on an-giogenesis but also effects of thrombin on cell growth genes, selectin expression, etc. We and others have demon-strated a consistent effect of LMWH on angiogenesis in a variety of in vitro, ex vivo and in vivo models (42). As demonstrated in Figure 4, one preparation of LMWH was quite effective in inhibiting the proliferation of blood ves-sels in the embryonic chick aortic ring model, and similar effects have been seen in virtually every model of angio-genesis.

COAGULATION CASCADE

TF ThrombinClotting-

dependentClotting-

dependent

Clotting-independent

Clotting-dependent

ANGIOGENESIS

Fibrin

Clotting-independent

PARs

TUMOR GROWTH AND METASTASIS

OncogeneActivation

↓

LMWH?

??

VEGF

??

platelet

Fig. 3. Possible Effects of Low-Molecular-Weight Heparins on Tumor Biol-ogy. See text for details. (Modified from Reference 30, with permission of the pub-lishers).

Fig. 4. Ex Vivo Angiogenesis. Effect of Low-Molecular-Weight Heparin (Dal-teparin/Fragmin; continuous exposure) on chick aortic ring angiogenesis in response to maximal stimulation with EGM-2 complete medium (containing 2% fetal bovine serum, hydrocortisone, human fibroblast growth factor-β, vascular endothelial growth factor and insulin-like growth factor-1). This experiment is representative of 3 experiments performed to date with a dose response mean inhibition as follows: 0.1 IU/ml – 25%; 1.0 IU/ml - 48% ; 5 IU/ml – 75%; 10 IU/ml –80%; all values p < 0.001 compared to controls.

Summary

In summary, blood clotting reactions are intimately in-volved in the biology of cancer, not just rendering the can-cer patient susceptible to thrombosis but stimulating tumor growth genes, tumor cell diapedesis, adhesion to the endo-thelium, angiogenesis, and a host of other processes critical to both primary tumor growth and metastasis. Rational strategies to downregulation of these coagulation pathways may provide a duality of beneficial effect – both reduction in thrombosis potential and tumor proliferation.

P109

Table 1. Pathogenesis of Thrombosis in Cancer - Modification of Virchow’s Triad. 1. STASIS

a. Prolonged bed rest b. Extrinsic compression of blood vessels by tumor

2. VASCULAR INJURY a. Direct invasion by tumor b. Prolonged use of central venous catheters c. Endothelial damage by chemotherapy drugs, tumor-associated cytokines, etc.

3. HYPERCOAGULABILITY a. Release of tumor-associated procoagulants (e.g. tissue factor) from malignant cells and/or inflammatory cells

(e.g. macrophages) b. Impaired endothelial defense mechanisms and naturally occurring inhibitors (e.g. protein C, protein S, anti-

thrombin, APC resistance) c. Enhanced adhesive interaction between tumor cells, vascular endothelial cells, platelets and host macrophages

(e.g. mucin-secreting adenocarcinomas and leukocyte L-selection and platelet P-selectin)

Table 2. Oncogene Activation of Blood Coagulation Oncogene or/ Tumor Signaling Pathway (s) Human Tumor Effect on Coagulation Suppressor Gene or Vascular Biology (reference number) (genes regulated) MET (5) Tyrosine kinase receptor Hepatoma Thrombosis; DIC (hepatocyte growth factor/ (PAI-1; COX-2) Scatter factor) PTEN (6) Ras/MEK/ERK Glioblastoma Thrombosis;necrosis (TF) K-ras & p53 (7) MEK/MAPK/PI3K Colon Cancer Angiogenesis (TF; VEGF; TSP-1/2)

DIC = disseminated intravascular coagulation; PAI-1 = plasminogen activator inhibitor-1; COX-2 = cyclooxygenase-2; TF=tissue factor; VEGF=vascular endothelial growth factor; TSP=thrombospondin

References

1. Blom JW, Doggen CJM, Osanto S, Rosendaal FR. Malignancies, prothrombotic mutations and the risk of venous thrombosis. J Amer Med Assoc 2005;293:715-722.

2. Falanga A, Rickles FR. The pathogenesis of thrombosis in cancer. New Oncol: Thrombosis 2005;1:9-16.

3. Boccaccio C, Sabatino G, Medico E, Girolami F, Follenzi A, Reato G, Sottile A, Naldini L, Comoglio PM. The MET oncogene drives a genetic programme linking cancer to haemostasis. Nature 2005;434:396-400.

4. Rong Y, Post DE, Pieper RO, Burden DL, Van Meir EG, Brat DJ. PTEN and hypoxia regulate tissue factor expression and plasma coagulation by glioblastoma. Cancer Res 2005;65:1406-1413.

5. Yu JL, May L, Lhotak V, Shahrzad S, Shirasawa S, Weitz JI, Coomber BL, Mackman N, Rak JW. Oncogenic events regulate tissue factor expres-sion in colorectal cancer cells: implications for tumor progression and angiogenesis. Blood 2005;105:1734-1741.

6. Abe K, Shoji M, Chen J, Bierhaus A, Danave I, Micko C, Casper K, Dillehay D, Nawroth PP, Rickles FR. Regulation of vascular endothelial growth factor production and angiogenesis by the cytoplasmic tail of tissue factor. Proc Nat Acad Sci USA 1999;96:8663-8668.

7. Ruf W, Dorfleutner A, Riewald M. Specificity of coagulation factor signaling. J Thromb Haemost 2003;1:1495-1503.

8. Belting M, Dorrell MI, Sandgren S, Aguilar E, Ahamed J, Dorfleutner A, Carmeliet P, Mueller BM, Friedlander M, Ruf W. Regulation of an-giogenesis by tissue factor cytoplasmic domain signaling. Nat Med 2004;10:502-509.

9. Hu L, Lee M, Campbell W, Perez-Soler R, Kar-patkin S. Role of endogenous thrombin in tumor implantation, seeding, and spontaneous metasta-sis. Blood 2004;104:2746-2751.

10. Levitan N, Dowlati A, Remick SC, Tahsildar HI, Sivinski LD, Beyth R, Rimm AA. Rates of initial and recurrent thromboembolic disease among pa-tients with malignancy versus those without ma-lignancy: risk anlaysis using Medicare claims data. Medicine 1999;78:285-291.

11. Sorensen HT, Mellemkjaer L, Olsen JH, Baron JA. Prognosis of cancers associated with venous thromboembolism. N Engl J Med 2000;343:1846-1850.

12. von Templehoff GF, Harenberg J, Niemann F, Hommel G, Kirkpatrick CJ, Heilmann L. Effect of low molecular weight heparin (Certoparin) versus unfractinated heparin on cancer survival following breast and pelvic surgery: a prospective randomized, double-blind trial. Int J Oncol 2000;16:815

P110

13. Kakkar AK, Levine MN, Kadziola Z, Lemoine NR, Low V, Patel HK, Rustin G, Thomas M, Quigley M, Williamson RCN. Low molecular weight heparin, therapy with Dalteparin, and sur-vival in advanced cancer: The Fragmin Advanced Malignancy Outcome Study (FAMOUS). J Clin Oncol 2004;21:1944-1948.

14. Altinbass M, Coskun HS, Er O, Ozkan M, Eser B, Unal A, Cetin M, Soyuer S. A randomized clini-cal trial of combination chemotherapy with and without low-molecular-weight heparin in small cell lung cancer. J Thromb Haemost 2004;2:1266-1271.

15. Lee AYY, Rickles FR, Julian JA, Gent M, Baker RI, Bowden C, Kakkar AK, Prins M, Levine MN. Randomized comparison of low molecular weight heparin and coumarin derivatives on the survival of patients with cancer and venous thromboem-bolism. J Clin Oncol 2005;23:2123-2129.

16. Klerk CPW, Smorenburg SM, Otten H-M, Lensing AWA, Prins MH, Piovella F, Prandoni P, Bos MMEM, Richel DJ, van Tienhoven G, Buller HR. The effect of low molecular weight heparin on survival in patients with advanced malignancy. J Clin Oncol 2005;23:2130-2135.

17. Huang X, Molema G, King S, Watkins L, Edging-ton TS, Thorpe PE. Tumor infarction in mice by antibody-directed targeting of tissue factor to tu-mor vasculature. Science 1997;275:547-550.

18. Hu Z, Sun Y, Garen A. Targeting tumor vascula-ture endothelial cells and tumor cells for immuno-therapy of human melanoma in a mouse xenograft model. Proc Nat Acad Sci USA 1999;96:8161-8166.

19. Nilsson F, Kosmehl H, Zardi L, Neri D. Targeted delivery of tissue factor to the ED-B domain of fibronectin, a marker of angiogenesis, mediates the infarction of solid tumors in mice. Cancer Res 2001;61:711-716.

20. El-Sheikh A, Borgstrom P, Bhattacharjee G, Belt-ing M, Edgington TS. A selective tumor mi-crovasculature thrombogen that targets a novel receptor complex in the tumor angiogenic micro-environment. Cancer Res 2005;65:11109-11117.

21. Hu Z, Garen A. Targeting tissue factor on tumor vascular endothelial cells and tumor cells for im-munotherapy in mouse models of prostatic can-cer. Proc Nat Acad Sci USA 2001;98:12180-12185.

22. Adams BK, Ferstl EM, Davis MC, Herold M, Kurtkaya S, Camalier RF, Hollingshead MG, Kaur G, Sausville EA, Rickles FR, Snyder JP, Li-otta DC, Shoji M. Synthesis and biological evaluation of novel curcumin analogs as anti-cancer and anti-angiogenesis agents. Bioorg Med Chem 2004;12:3871-3883.

23. Heit JA, Silverstein MD, Mohr DN, Petterson TM, O’Fallon WM, Melton LJ. Relative impact of risk factors for deep vein thrombosis and pul-monary embolism: a population-based study. Arch Int Med 2002;162:1245-1248.

24. Prandoni P, Lensing AWA, PIccioli A, Bernardi E, Simioni P, Girolami B, Marchiori A, Sabbion P, Prins MH, Noventa F, Girolami A. Recurrent venous thromboembolism and bleeding complica-tions during anticoagulant treatment in patients with cancer and venous thrombosis. Blood 2002;100;3484-3488.

25. Rickles FR, Levine MN. Epidemiology of thrombosis in cancer. ACTA Haematol (Suppl) 2001;106:6-12.

26. Dvorak HF, Rickles FR. Malignancy and Hemo-stasis. In: Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 5th edition (Eds. R.W. Colman, J Hirsh, VJ Marder, AW Clowes, and JN George). Lippincott-Raven Publishers, Philadelphia, PA, Chapter 57, 2006;851-873.

27. Virchow R. Weitere Unter-suchungen ϋber die Verstopgung der Lungenarterie und ihre Folgen. In: Virchow R, ed. Gesammelte Abhandlungen zur wissenschaftlichen Medizin. Meidinger Sohn: Frankfurt am Main, 1856, pp. 227-380.

28. Wahrenbrock M, Borsig L, Le D, Varki N, Varki A. Selectin-mucin interactions as a probable mo-lecular explanation for the association of Trous-seau syndrome with mucinous adenocarcinomas. J Clin Invest 2003;112:853-862.

29. Boccaccio C, Comoglio PM. A functional role for hemostasis in early cancer development. Cancer Res 2005;65:8579-8582.

30. Rickles FR, Patierno S, Fernandez PM. Tissue factor, thrombin and cancer. Chest (Supplement) 2003;124:58S-68S.

31. Contrino J, Hair GA, Schmeizl M, Rickles FR, Kreutzer DL. In situ characterization of antigenic and functional tissue factor expression in human tumors utilizing monoclonal antibodies and re-combinant factor VIIa as probes. Amer J Pathol 1994;145:1315-1322.

32. Contrino J, Hair GA, Kreutzer DL, Rickles FR. In situ expression of antigenic and functional tis-sue factor in vascular endothelial cells: correla-tion with the malignant phenotype of human breast disease. Nat Med 1996;2:209-215.

33. Bora PS, Hu Z, Tezel TH, Sohn J-H, Kang, SG, Cruz JMC, Bora NS, Garen A, Kaplan HJ. Im-munotherapy for choroidal neovascularization in a laser-induced mouse model simulating exuda-tive (wet) macular degeneration. Proc Nat Acad Sci USA 2003;100:2679-2684.

34. Hettiarachchi RJK, Smorenburg SM, Ginsberg J, Levine M, Prins MH, Buller HR. Do heparins do more than just treat thrombosis? The influence of heparins on cancer spread. Thromb Haemost 1999;82:947-952.

35. Meyer G, Marjanovic Z, Valcke J, Lorcerie B, Gruel Y, Solal-Celigny P, Le Maignan C, Extra JM, Cottu P, Farge D. Comparison of low-molecular-weight heparin and warfarin for the secondary prevention of venous thromboem-bolism in patients with cancer. Arch Intern Med 2002;162:1729-1735.

36. Folkman J, Langer R, Linhardt RJ, Haudenschild C, Taylor S. Angiogenesis inhibition and tumor regression caused by heparin or a heparin frag-ment in the presence of cortisone. Science 1983;221:719-725.

37. Collen A, Smorenburg SM, Peters E, Lupu F, Koolwijk P, Van Noorden C, van Hinsbergh VWM. Unfractionated and low molecular weight heparin affect fibrin structure and angiogenesis in vitro. Cancer Research 2001;60:6196-6200.

38. Amirkhosravi A, Mousa SA, Amaya M, Francis JL. Antimetastatic effect of tinzaparin, a low-molecular-weight heparin. J Thromb Haemost 2003;1:1972-1976.

39. Vlodavsky I, Friedmann Y. Molecular properties and involvement of heparase in cancer metastasis and angiogenesis. J Clin Invest 2001;108:341-347.

40. Hasan J, Shnyder SD, Clamp AR, McGown AT, Bicknell R, Presta M, Bibby M, Double J, Craig S, Leeming D, Stevenson K, Gallagher JT, Jayson GC. Heparin octasaccharides inhibit angiogene-sis in vivo. Clin Cancer Res 2005;11:8172-8179.

41. Stevenson JL, Choi SH, Varki A. Differential me-tastasis inhibition by clinically relevant levels of heparins – correlation with selectin inhibition, not antithrombotic activity. Clin Cancer Res 2005;11:7003-7011.

42. Fernandez PM, Chou DS, Aquilina JW, Patierno SR, Rickles FR. Unfractionated heparin (UFH) and a low molecular weight heparin (dalteparin) exhibit antiangiogenic effects using in vitro, ex vivo and in vivo angiogenesis models. Proc Amer Assoc Cancer Res 2003;44:698-699.

Mayo Clin Proc. • June 2006;81(6):758-767 • www.mayoclinicproceedings.com758

LOW-MOLECULAR-WEIGHT HEPARIN IN PATIENTS WITH ADVANCED CANCERORIGINAL ARTICLE

From the Division of Hematology (K.S.), Department of Oncology (K.S., S.H.O.,S.R.A., C.L.L.), and Department of Health Sciences Research (J.A.S., P.J.N.),Mayo Clinic College of Medicine, Rochester, Minn; Toledo Community Hospi-tal Community Clinical Oncology Program (CCOP), Toledo, Ohio (P.L.S.); CedarRapids Oncology Project CCOP, Cedar Rapids, Iowa (L.K.); Division of Hema-tology/Oncology, Mayo Clinic College of Medicine, Scottsdale, Ariz (T.R.F.);Wichita Community Clinical Oncology Program, Wichita, Kan (S.R.D.);Meritcare Hospital CCOP, Fargo, ND (R.L.); Iowa Oncology Research Associa-tion CCOP, Des Moines (R.F.M.); and Carle Cancer Center CCOP, Urbana, Ill(K.M.R.). A list of additional participating institutions appears at the end ofthis article.

This study was conducted as a trial of the North Central Cancer TreatmentGroup and Mayo Clinic and was supported in part by Public Health Servicegrants CA-25224, CA-37404, CA-15083, CA-63826, CA-52352, CA-60276,CA-35431, CA-37417, CA-35101, and CA-35195 from the National CancerInstitute, Department of Health and Human Services.

Both the low-molecular-weight heparin (dalteparin) and the placebo wereprovided by Pharmacia & Upjohn, Inc.

Address reprint requests and correspondence to Charles L. Loprinzi, MD,Department of Oncology, Mayo Clinic College of Medicine, 200 First St SW,Rochester, MN 55905 (e-mail: [email protected]).

© 2006 Mayo Foundation for Medical Education and Research

Low-Molecular-Weight Heparin in Patients With Advanced Cancer:A Phase 3 Clinical Trial

KOSTANDINOS SIDERAS, MD; PAUL L. SCHAEFER, MD; SCOTT H. OKUNO, MD; JEFF A. SLOAN, PHD;LEILA KUTTEH, MD; TOM R. FITCH, MD; SHAKER R. DAKHIL , MD; RALPH LEVITT, MD; STEVEN R. ALBERTS, MD;ROSCOE F. MORTON, MD; KENDRITH M. ROWLAND, MD; PAUL J. NOVOTNY, MS; AND CHARLES L. LOPRINZI, MD

OBJECTIVE: To prospectively assess whether low-molecular-weight heparin (LMWH) provides a survival benefit in patients withadvanced cancer.

PATIENTS AND METHODS: Between December 1998 and June2001, we performed a randomized controlled study of patientswith advanced cancer. Initially, the study was double blinded andplacebo controlled, with the patients receiving daily injections of5000 U of LMWH or saline. However, because of low accrualmidway through the study, the placebo injection arm was elimi-nated, and the study became open labeled, with patients receivingeither LMWH injections plus standard clinical care or standardclinical care alone. The primary study end point was overallsurvival.

RESULTS: Of 141 patients randomized to this clinical trial, 3dropped out, leaving 138 patients. The median survival time was10.5 months (95% confidence interval, 7.6-12.2 months) for thecombined standard care and placebo groups. The median survivaltime for the combined LMWH arms was 7.3 months (95% confi-dence interval, 4.8-12.2 months). These median survival timeswere not significantly different (log-rank P=.46). The median sur-vival times for the blinded and unblinded LMWH groups were 6.2months and 9.0 months, respectively. The median survival timeswere 10.3 months for the blinded placebo arm and 10.5 monthsfor the standard care arm. The rate of severe or life-threateningvenous thromboembolism was 6% in the LMWH arms and 7% in thecontrol arms. The rate of severe or life-threatening bleeding was3% in the LMWH arms and 7% in the control arms.

CONCLUSION: This trial was unable to demonstrate any survivalbenefit for LMWH in patients with advanced cancer.

Mayo Clin Proc. 2006;81(6):758-767

CI = confidence interval; LMWH = low-molecular-weight heparin;NCCTG = North Central Cancer Treatment Group; QOL = quality of life;SDS = symptom distress scale; UFH = unfractionated heparin; ULN =upper limit of normal

L ow-molecular-weight heparin (LMWH), manufac-tured by the depolymerization of standard heparin, has

excellent bioavailability and a long biological half-life whengiven subcutaneously.1-3 These factors enable LMWH to begiven subcutaneously in a fixed dose, based on body weight,once or twice daily. In addition, this treatment can be admin-istered on an outpatient basis without the need for intrave-nous lines or laboratory monitoring. Two trials, reportedbefore the development of this current clinical trial, demon-strated that LMWH can be used safely and effectively totreat patients with proximal-vein thrombosis on an outpa-tient basis.4,5

Prandoni et al6 performed a trial that compared subcuta-neous LMWH with intravenous unfractionated heparin(UFH) in patients with proximal deep venous thrombosis.Fifteen of 85 patients in the LMWH arm and 18 of 85patients in the intravenous UFH arm had malignant dis-ease. Interestingly, 8 (44%) of the 18 patients with malig-nant disease treated with UFH had died when the study wasreported compared with only 1 (7%) of the 15 treated withLMWH. Green et al7 performed a similar study comparingLMWH vs UFH for the treatment of proximal deep venousthrombosis. In their study, 49 of 219 patients in the UFHgroup and 46 of 213 patients in the LMWH group hadcancer. Fourteen deaths due to cancer occurred in the UFHgroup compared with only 7 cancer deaths in the LMWHgroup. Moreover, at least 2 meta-analyses, intended tocompare LMWH to UFH for the treatment of deep venousthrombosis, have reported an improved survival in cancerpatients treated with LMWH.8,9 Thus, these data suggestedthat the use of LMWH might improve survival in patientswith cancer.

A possible association between the hemostatic systemand cancer cells has been demonstrated in several experi-mental and clinical studies.10,11 In one trial, the addition ofwarfarin to chemotherapy was associated with improved

For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.For personal use. Mass reproduce only with permission from Mayo Clinic Proceedings.

Mayo Clin Proc. • June 2006;81(6):758-767 • www.mayoclinicproceedings.com 759

LOW-MOLECULAR-WEIGHT HEPARIN IN PATIENTS WITH ADVANCED CANCER

survival in patients with small cell lung cancer,12 whereas,in a multicenter study, warfarin demonstrated an improvedresponse rate.13 Additionally, the use of subcutaneous UFHin the perioperative period for patients with resectedcolorectal cancer was associated with an improved survivalover no UFH.14

Lebeau et al15 randomized 277 patients with small celllung cancer to receive subcutaneous heparin or nothingduring the first 5 weeks of chemotherapy. There were morecomplete responses (37% vs 23%), an improved mediansurvival (317 days vs 261 days), and improved survivalrates at 1, 2, and 3 years (40% vs 30%, 11% vs 9%, and 9%vs 6%, respectively) in the patients treated with LMWH.This benefit was seen only in patients with limited stagedisease.

Although the theoretic mechanism to account for thepotentially improved response rate and survival in patientswith cancer by modifying the hemostatic system is notclearly evident, several theoretical possibilities exist. Onesuch possibility is a direct interference by anticoagulantagents on the local microthrombus formation and fibrindeposition, which has been hypothesized to protect cancercells from chemotherapy.12,16,17 Another possibility mightbe heparin’s antithrombin properties. Thrombin is knownto act as a growth factor and an initiator of DNA synthesis.Thus, interference with thrombin might have antitumorgrowth effects.18 In addition, heparins, specifically LMWH,have been shown to have direct antigrowth, antiangiogen-esis, and antimetastatic effects.19-22

We performed this study to prospectively assesswhether LMWH provides a survival benefit in patientswith advanced cancer.

PATIENTS AND METHODS

This study, performed between December 1998 and June2001, was initially developed as a randomized, double-blinded, placebo-controlled clinical trial to compare subcu-taneous LMWH once daily (dalteparin, Pharmacia &Upjohn, Inc, Kalamazoo, Mich), 5000 U, to placebo injec-tions in patients with advanced incurable cancer. Becauseof a low accrual rate (after 52 accrued patients) and con-cerns that the low accrual rate was related to the require-ments for placebo injections, the study was modified inFebruary 2000, and the saline placebo injections wereeliminated. Then, unblinded LMWH was compared withstandard clinical care (with 89 more patients accrued afterthat point).

Patients considered for this protocol had advancedbreast cancer and failed first-line chemotherapy, ad-vanced prostate cancer and failed primary hormonaltherapy, advanced lung cancer, or advanced colorectal

cancer. Other eligibility criteria included age older than18 years, Eastern Cooperative Oncology Group perfor-mance status of 0 to 2, physician-judged life expectancyof more than 12 weeks, and fulfillment of certain labora-tory criteria (white blood cell count, >3500 × 109/L; plate-let count, >150,000 × 109/L; total bilirubin, <1.5 times theupper limit of normal [ULN]; aspartate aminotransferase,<3 times the ULN; creatinine, <1.5 times the ULN; pro-thrombin time, <1.5 times the ULN; activated partialthromboplastin time, <1.5 times the ULN). Exclusioncriteria included known central nervous system metasta-sis, a history of heparin-induced thrombocytopenia, cur-rent anticoagulation therapy, a known intolerance to hep-arin, enrollment in a chemotherapy or radiation therapyprotocol, and a history of deep venous thrombosis, pul-monary embolism, or a clotted catheter within the pastyear. However, patients were allowed to receive chemo-therapy or radiation therapy if not part of an investiga-tional protocol.

Patient evaluations at study initiation included a historyand physical examination, blood tests (hematology group,liver and renal function tests, prothrombin time, partialthromboplastin time), and quality-of-life (QOL) question-naires. For QOL assessments, the single-item visual analogUniscale23 and a 5-item series of linear analog self-assess-ment measures developed by the Mayo Cancer Center Sta-tistics Unit24 were used. These questionnaires were supple-mented by a 13-item symptom distress scale (SDS), whichis one of the longest-standing QOL tools for oncologypatients, has been well validated, has been proved to bereliable,25 and is prognostic of survival in patients withcancer.26 Eligible patients were stratified before random-ization according to age (>50 years or ≤50 years), sex,disease site, prior thrombotic episodes, good, bad, or un-sure prognostic index scores,27 and current therapy (sys-temic vs radiation vs both vs none).

Protocol therapy consisted of dalteparin, 5000 U subcu-taneously once a day (or placebo in the first 52 patients).After study initiation, all patients were to be seen for ahistory and physical examination monthly for the first yearand then every 3 months for 2 years. The QOL question-naires were to be collected on the same schedule. Patientswho received injections (LMWH or placebo) were sched-uled to have a hematology group test weekly for the firstmonth, then monthly for the first year, and then every 3months for 2 years.

The study was originally designed as a randomized,placebo-controlled, double-blinded phase 3 trial with 2treatment arms (placebo and LMWH). The randomizationprocesses applied were handled through the North CentralCancer Treatment Group (NCCTG) Randomization Of-fice. Patients were randomized to each of the 2 treatment




groups using a dynamic allocation method that incorpo-rated the stratification factors mentioned earlier to bal-ance the marginal distributions.28 The study was moni-tored in concordance with the NCCTG standard operatingprocedures, which include biannual review of safety andblinded efficacy by the NCCTG Data Monitoring Com-mittee. The primary study end point was overall survival.Secondary end points included incidence of toxic effects,incidence of thromboembolic events, and changes inQOL measured by the Uniscale, SDS, and linear analogself-assessment.

Log-rank testing formed the basis for analysis of theprimary end point of survival.29,30 Analysis of the secondaryend points was performed by simple comparison of fre-quency distributions via standard t test or Wilcoxon proce-dures and Fisher exact test for the ordinal and categoricallevel variables, respectively. The protocol was originallydesigned to accrue 265 patients in each of the original 2treatment arms. This would have provided 80% power todetect a 34% improvement in 12-month median survivalusing a 2-sided log-rank test. This power calculation as-sumed a median survival of 12 months using the currentstandard therapy, a minimum of 18 months of follow-up foreach patient, and an α level of .05.

RESULTS

BASELINE CHARACTERISTICS

A total of 141 patients were randomized to this clinicaltrial. Three patients, 1 randomized to blinded LMWH and 2to unblinded LMWH, dropped out before receiving anyprotocol therapy. Of the 138 remaining patients, 68 re-ceived LMWH (24 initially in the blinded LMWH groupand 44 after the placebo arm was eliminated), and 70patients were in the standard clinical care group (26 ini-tially receiving placebo injections and 44 more after theplacebo treatment was discontinued). Of the 26 patientsinitially receiving placebo injections, 10 were alive by thetime of unblinding in February 2000. These 10 patientswere told to discontinue the injections and were given theoption to stay in the study as controls (standard clinicalcare) the next time they came in contact with the studypersonnel. These 10 patients discontinued taking placeboinjections anywhere from 2 to 11 months after their studyinitiation.

The protocol accrual was stopped before reaching theprestudy planned accrual goal by the NCCTG Data Moni-toring Committee because of a slower than predicted proto-col accrual rate, with the knowledge (provided by an in-terim analysis report) that the patient survival rates werenumerically worse on one arm of the blinded study. Thedistribution of baseline values for all 4 groups is given in

Table 1, demonstrating that the treatment groups were wellbalanced. In particular, there seemed to be no difference inthe blinded and unblinded LMWH arms in terms of patientcharacteristics.

EFFICACY ANALYSIS: SURVIVAL

Kaplan-Meier survival curves, based on the primary dis-ease types, are illustrated in Figure 1, revealing that thepatient groups were relatively similar, except that the pa-tients with prostate cancer fared slightly better than theother patient groups. The Kaplan-Meier survival curves forthe patients randomized and not randomized to receiveLMWH are illustrated in Figure 2, illustrating no signifi-cant survival differences. Figure 3 demonstrates survivalcurves with patients divided into 4 groups based on thetreatment arm they were assigned (LMWH vs no LMWH)and whether they were part of the initial double-blindedportion of the study or the latter unblinded portion, againdemonstrating no significant survival differences amongthe 4 arms. Examining survival curves for patients random-ized to receive LMWH and those not randomized to receiveLMWH for each individual primary disease type did notsuggest that LMWH was beneficial in any of these patientsubsets. The median survival time was 10.5 months (95%confidence interval [CI], 7.6-12.2 months) for the com-bined standard care and placebo groups. The median sur-vival for the combined LMWH arms was 7.3 months(95% CI, 4.8-12.2 months). These median survival timeswere not significantly different (log-rank P=.46). The me-dian survival times for the blinded and unblinded LMWHgroups were 6.2 months and 9.0 months, respectively. Themedian survival times were 10.3 months for the blindedplacebo arm and 10.5 months for the standard care arm.

When a subpopulation of patients who lived longer than6 months was examined, no significant differences in sur-vival occurred between the 2 treatment groups (Figure 4).The median survival times in this situation were 16.6months in the combined LMWH arms and 12 months in theplacebo and standard clinical care arms (log-rank P=.46).This slight difference in median survival rates was due tothe patients in the unblinded LMWH arm, who tended tosurvive longer than the patients in the other 3 arms in thissubpopulation of patients who lived longer than 6 months.To investigate this issue further, we stratified patients ac-cording to a prognostic score using baseline covariatesonly. This way, treatment or no treatment with LMWH isnot directly related to the stratification, and, at least theo-retically, a better assessment of the treatment effect in thissubgroup can be obtained. When this analysis was per-formed in the current trial, again no statistically significantdifference was observed, and the survival curves weremore overlapping.




The primary analysis for survival was examined forpotential confounding variables by concomitant covariateswith use of Cox proportional hazards models (Table 2).Variables included in the modeling process were treatmentarm; baseline good, bad, or unsure status; tumor site; andbaseline performance score. The adjusted P value for differ-ences between the treatment arms was .25, indicating thatcovariates did not modify the results of the primary analysis.

QUALITY OF LIFE

The QOL and SDS scores were similar, both at baselineand during the protocol period, in patients randomized toreceive LMWH vs those not randomized to receiveLMWH. The average baseline patient QOL was 74, indi-cating a high level of QOL for this patient sample.Thirty-seven and 36 patients in the combined LMWHand placebo or standard clinical care arms, respectively,

TABLE 1. Distribution of Baseline Factors in Randomized Patients*

Blinded Unblinded StandardLMWH Placebo LMWH care P value P value

Baseline factors (n=24) (n=26) (n=44) (n=44) (among all 4 groups) (LMWH vs none)

Median age (y) 64.5 63.5 68.5 70.5 .10 .38Age range (y)† .68 .95

33-50 4 (17) 4 (15) 4 (9) 4 (9)51-86 20 (83) 22 (85) 40 (91) 40 (91)

Sex (%)† .09 .89Male 12 (50) 11 (42) 28 (64) 31 (70)Female 12 (50) 15 (58) 16 (36) 13 (30)

Disease site† .78 .95Breast 4 (17) 5 (19) 4 (9) 2 (5)Colon 5 (21) 7 (27) 12 (27) 12 (27)Small cell lung 3 (12) 2 (8) 3 (7) 2 (5)Non–small cell lung 9 (38) 8 (31) 19 (43) 23 (52)Prostate 3 (12) 4 (15) 6 (14) 5 (11)

Current therapy† .85 .55Systemic 12 (50) 12 (46) 21 (48) 23 (52)Radiation 2 (8) 0 (0) 2 (5) 1 (2)Both 1 (4) 2 (8) 1 (2) 1 (2)None 9 (38) 12 (46) 20 (45) 19 (43)

Prior thrombotic event .58 .30Yes 1 (4) 1 (4) 2 (5) 0 (0)No 23 (96) 25 (96) 42 (95) 44 (100)

Prognostic index scores† .93 .79Good 9 (38) 10 (38) 18 (41) 16 (36)Bad 2 (8) 2 (8) 1 (2) 3 (7)Unsure 13 (54) 14 (54) 25 (57) 25 (57)

ECOG performance status .72 .690 7 (29) 8 (31) 15 (34) 14 (32)1 15 (63) 15 (58) 22 (50) 20 (45)2 2 (8) 3 (12) 7 (16) 10 (23)

Appetite rating‡ .96 .48Increased 3 (13) 2 (8) 2 (5) 3 (7)Same 8 (33) 12 (46) 14 (34) 18 (43)Slightly reduced 7 (29) 5 (19) 12 (29) 9 (21)Moderately reduced 4 (17) 6 (23) 9 (22) 7 (17)Markedly reduced 2 (8) 1 (4) 4 (10) 5 (12)

Hemoglobin (mg/dL) .35 .15Median 11.8 11.9 12.1 12.7IQR 2.0 1.8 2.1 2.4

Aspartate aminotransferase (U/L) .54 .40Median 20.5 22.0 22.0 21.5IQR 11.0 10.0 10.5 12.5

Creatinine (mg/dL) .50 .53Median 0.9 0.8 0.9 0.9IQR 0.25 0.20 0.4 0.4

Uniscale QOL .53 .38Median 81 83 74 79IQR 50.0 25.0 50.0 50.0

*Values are number (percentage) unless indicated otherwise. ECOG = Eastern Cooperative Oncology Group; IQR = interquartile range;LMWH = low-molecular-weight heparin; QOL = quality of life.

†Stratification factor.‡Data were available for only 41 and 42 patients in the unblinded LMWH and standard care columns, respectively.




reported decreases in QOL during the treatment periodof a clinically meaningful amount of 10 points on the100-point scale (where 0 is worst QOL and 100 is bestQOL).

The SDS scores indicated that fatigue was a major con-cern in this patient population. The mean baseline fatiguelevel was 63 of 100 (where 0 is worst and 100 is best). Theother major concerns were pain frequency (mean score,

0

20

40

60

80

100

0 1 2

Prob

abili

ty o

f su

rviv

al (

%)

Time from randomization (y)

Breast

Colon

Small cell lung

Non–small cell lung

Prostate

P=.02

No. at riskBreast 15 6 2Colon 36 17 2Small cell lung 10 1 1Non–small cell lung 58 14 0Prostate 18 11 0

FIGURE 1. Survival curves for the different study subpopulations.

FIGURE 2. Survival curves for patients randomized to receive low-molecular-weight heparin (LMWH) vsthose randomized to receive placebo or standard care.

0

20

40

60

80

100

0 1 2

Prob

abili

ty o

f su

rviv

al (%

)


LMWH

Placebo or standard care

P=.46

No. at riskLMWH 68 23 2Placebo or standard care 69 28 6




70), insomnia (mean score, 75), pain severity (mean score,78), bowel dysfunction (mean score, 78), and cough (meanscore, 80).

TOXIC EFFECTS

Table 3 provides information on several toxic effects thatmight have been considered potentially related to LMWH.

FIGURE 3. Survival curves for patients randomized to receive blinded low-molecular-weight heparin(LMWH), blinded placebo injections, unblinded LMWH, or standard care alone.

FIGURE 4. Survival curves for patients living at least 6 months who were randomized toreceive low-molecular-weight heparin (LMWH) vs those randomized to receive placebo orstandard care.

0

20

40

60

80

100

0 1 2

Prob

abili

ty o

f su

rviv

al (

%)


Blinded LMWH

Placebo injection

Unblinded LMWH

Standard care

P=.78

No. at riskBlinded LMWH 84 8 2Placebo injection 86 11 3Unblinded LMWH 44 15 0Standard care 43 17 3

0

20

40

60

80

100

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Prob

abili

ty o

f su

rviv

al (%

)


LMWH

Placebo or standard care

P=.46

No. at riskLMWH 37 23 2Placebo or standard care 50 27 6




In addition to these factors, more than 100 other less likelyrelated toxic effects were also tabulated in the protocolpatients. There was no suggestion that LMWH caused anyof these other potential toxic effects. The most prevalenttoxic effects observed were fatigue (reported by 48% ofLMWH-treated patients and 69% of other patients), an-orexia (reported by 43% of LMWH-treated patients and49% of other patients), and nausea (reported by 31% ofLMWH-treated patients and 40% of other patients). Withregard to bleeding, 2 (3%) of 68 patients in the LMWHarms and 5 (7%) of 70 patients in the control groups experi-enced severe, life-threatening, or lethal hemorrhagic com-plications (grades 3, 4, or 5 toxic effects). One of thesepatients, who was in the standard clinical care group, diedas a direct result of a hemorrhagic stroke. With regard tothrombosis, 6% in the LMWH arms and 7% in the controlarms experienced grade 3 to 4 thrombotic complications.

DISCUSSION

This trial found no survival benefit with use of LMWH inpatients with advanced cancer. A similar trial by Kakkaret al31 also found no significant difference in patientstreated with LMWH compared with placebo (385 pa-tients; median survival, 10.8 vs 9.1 months; P=.19). In

contrast, a trial by Klerk et al32 showed a statistically sig-nificant difference in patients treated with LMWH vsplacebo (302 patients; median survival, 8 vs 6.6 months;P=.02).

In the study by Kakkar et al, an analysis was performedin patients with better prognosis (not defined a priori),defined as patients who lived more than 17 months. Thepatients treated with dalteparin in this subgroup were re-ported to have had a statistically significant increase inmedian survival (43.5 months; 95% CI, 33-52.3 months)compared with the placebo group (24.3 months; 95% CI,22.4-41.5 months). This subpopulation was examined be-cause of the hypothesis that LMWH may be beneficial inpatients with less advanced disease and better prognosisbecause it may simply be too late for LMWH to offer asurvival benefit in patients with a short life expectancy. Inthe study by Klerk et al,32 an a priori specified subgroup ofpatients with more than 6 months of expected survival wasexamined, and LMWH therapy appeared to be beneficial(median survival, 15.4 vs 9.4 months; P=.10), as was seenin the whole study population.

In the current study, the LMWH group survival ratewas numerically better for a similar subgroup of pa-tients (also not defined a priori) who survived at least 6months (Figure 4), although this difference was not statis-tically significant. Reanalyzing these data by stratifyingpatients according to a prognostic score using baselinecovariates, a less biased analysis method, again demon-strated no survival benefit for patients with better prog-noses. Thus, the results of the current study do not supportthe theory that LMWH is beneficial for patients withbetter prognoses.

The results of the current trial demonstrated the safetyof LMWH in patients with advanced incurable cancers.Low-molecular-weight heparin did not appear to sig-nificantly increase the incidence of major bleeding.

It is reasonable to ask whether the results of this clini-cal trial were flawed based on the use of 4 differentprimary disease types, as opposed to the use of a single,more uniform patient population. Several specific diseasesites could have been chosen for study, ranging from asingle disease site, such as breast, prostate, colon, orlung cancer, vs all patients with advanced cancer, re-gardless of the primary tumor type. Likewise, differentdisease stages, ranging from precancers to widespreadincurable metastatic disease, could have been chosen forstudy.

To address potential concerns related to the use of dif-ferent tumor types and patients receiving different thera-pies, this issue was studied in some depth before the initia-tion of this trial. First, several trials that involved use ofantianorexia agents across a broad spectrum of cancer pa-

TABLE 2. Survival Analyses by Baseline Factors

MedianNo. of No. of survival Log-rank

Factor patients deaths time (d) P value

Prognostic index score .01Good 54 43 371Bad 8 7 76Unsure 75 63 230

Performance score .020 44 34 3901 71 60 2192 22 19 222

Current therapy .04Systemic 67 56 308Radiation 5 5 95Both 5 3 312None 60 49 294

Disease group .02Breast 15 13 346Colon 36 31 319Small cell lung 10 8 284Non–small cell lung 58 50 196Prostate 18 11 479

Age group (y) .88≤50 15 12 170>50 122 101 309

Prior thrombotic events .88Yes 4 2 371No 133 111 300

Sex .73Male 82 67 284Female 55 46 332




tient subpopulations were examined.33-36 Extensive analysisof these trials indicated that even though the survivalcurves for the varying disease sites varied, balancing thedistributions of disease site across treatment arms allowedfor an unbiased evaluation of the effect of each agent understudy with regard to survival.

To demonstrate the effect of randomization in this ex-perimental environment, the results of an analysis based on496 patients with advanced disease from lung (40%), gas-trointestinal (36%), and other sites (24%), which wereaccrued to a cancer control protocol involving therapeutic

agents for anorexia or cachexia, were evaluated.37 Thesurvival curves for the treatment arms (which would not beexpected to produce survival differences) had an associatedlog-rank P=.86.

Second, the sample of 496 patients was randomly di-vided into 2 groups (group A and group B), with no stratifi-cation factors being used, and survival curves were gener-ated from these groups. These survival curves overlapped.Subsequently, this was repeated 10,000 times usingbootstrapped samples. For each of these 10,000 samples,the question was asked, “How often were statistically sig-

TABLE 3. Incidences of Toxic Effects*

Blinded Unblinded StandardLMWH Placebo LMWH care P value P value

Toxic effects (n=24) (n=26) (n=44) (n=44) (among all 4 groups) (between blinded groups)

Soreness at injection site 6 (25) 3 (12) ND ND .28 NDBruising at injection site 12 (50) 5 (19) ND ND .03 NDBruising without

thrombocytopenia .22 .05Mild 4 (17) 1 (4) 6 (14) 3 (7)Moderate 0 (0) 0 (0) 2 (5) 0 (0)

Epistaxis .86 .35Mild 1 (4) 1 (4) 2 (5) 2 (5)Severe 0 (0) 1 (4) 0 (0) 1 (2)

Hematuria .23 .05Mild 0 (0) 0 (0) 1 (2) 1 (2)Moderate 0 (0) 1 (4) 0 (0) 3 (7)

Hemoptysis .36 .30Mild 0 (0) 2 (8) 1 (2) 0 (0)Severe 0 (0) 0 (0) 0 (0) 1 (2)

Hemorrhage, CNS .49 .75Severe 1 (4) 0 (0) 0 (0) 0 (0)Lethal 0 (0) 0 (0) 0 (0) 1 (2)

Hemorrhage, other, mild 1 (4) 0 (0) 1 (2) 0 (0) .79 .99Hemorrhage with grade 3-4

thrombocytopenia .26 .50Mild 0 (0) 0 (0) 1 (2) 0 (0)Severe 0 (0) 0 (0) 1 (2) 0 (0)

Melena .77 .75Mild 0 (0) 0 (0) 1 (2) 0 (0)Severe 0 (0) 0 (0) 0 (0) 1 (2)

Petechiae .32 .13Mild 1 (4) 0 (0) 4 (9) 2 (5)Moderate 1 (4) 0 (0) 1 (2) 0 (0)

Rectal bleeding .45 .55Mild 1 (4) 0 (0) 1 (2) 2 (5)Moderate 0 (0) 0 (0) 0 (0) 1 (2)

Vaginal bleeding .51 .50Mild 1 (4) 0 (0) 0 (0) 0 (0)Moderate 0 (0) 0 (0) 1 (2) 0 (0)

Thrombocytopenia .48 .93Mild 1 (4) 1 (4) 1 (2) 0 (0)Moderate 0 (0) 1 (4) 0 (0) 0 (0)Severe 1 (4) 1 (4) 2 (5) 0 (0)Life-threatening 0 (0) 0 (0) 0 (0) 1 (2)

Thrombosis .70 .70Severe 0 (0) 1 (4) 2 (5) 2 (5)Life-threatening 2 (8) 0 (0) 0 (0) 2 (5)

Transfusion, red blood cells,severe 1 (4) 1 (4) 1 (2) 1 (2) .94 .99

*Values are number (percentage). CNS = central nervous system; LMWH = low-molecular-weight heparin; ND = not done.




nificant (P≤.05) survival results seen in the 2 randomlyallocated groups?” The answer was 5% of the time, exactlywhat would have been expected by chance alone. Whenthis experiment was repeated using the stratification factorsidentified in the study protocol, statistically significant(P≤.05) survival curves were observed 4% of the time,demonstrating only a slight improvement with the use ofstratification. Thus, this evidence indicates that it is pos-sible to perform a clinical trial across a known heteroge-neous population, especially if stratification is used to re-move the potentially confounding effect of any covariatethat is strongly related to the primary efficacy end point.

An additional reason for using 4 different common pri-mary disease types in the current trial is that it provided awider net to potentially detect a treatment effect, especiallywhen hypothesis-generating subgroup analyses were per-formed to explore for a potential benefit in 1 of the primarydisease type subgroups. One could argue that it might havebeen better to pick 1 of the 4 disease types for studybecause the LMWH might have been efficacious in only 1of the 4 disease types. If this were the case, then, right upfront, investigators opting to choose 1 primary disease typefor study would have only a 25% chance of picking thecorrect disease type.

In the development of this trial, the issue of whether toallow concomitant cytotoxic chemotherapy in study pa-tients was also seriously considered. At first glance, itmight have seemed preferable to not allow concomitantchemotherapy. With this approach, there would have beenno chance of possible detrimental effects of LMWH onchemotherapy because of toxic effects or potential pharma-cological interactions. However, LMWH is used in patientsreceiving concomitant chemotherapy. In addition, experi-ence with more than 2000 patients who have enteredNCCTG anorexia or cachexia trials has illustrated that thepatients who are no longer receiving chemotherapy have apoorer prognosis than patients who continue to receivechemotherapy at the time of study entry. Thus, patientswere allowed to receive concomitant chemotherapy whilein this study.

Another issue to examine, regarding the validity of theresults of this clinical trial, relates to whether the studyconclusions are compromised because of the lower thanplanned patient accrual in this trial. To address this issue, itis first worth noting that the decision of the NCCTG DataMonitoring Committee, made independently from the pro-tocol principle investigators, certainly appeared to be wise.The final sample size of 138 patients provided 46% powerto detect a 50% improvement in median survival times.Given the results from this study that indicate a 3.2-monthsuperiority in median survival for patients receiving pla-cebo, the probability is less than 9% of obtaining results

REFERENCES1. Holmer E, Mattsson C, Nilsson S. Anticoagulant and antithrombotic

effects of heparin and low molecular weight heparin fragments in rabbits.Thromb Res. 1982;25:475-485.

2. Kakkar VV. Prevention and management of venous thrombosis. Br MedBull. 1994;50:871-903.

3. Hull RD, Pineo GF. Low-molecular-weight heparins for the treatment ofvenous thromboembolism. Ann Med. 1993;25:457-462.

4. Levine M, Gent M, Hirsh J, et al. A comparison of low-molecular-weightheparin administered primarily at home with unfractionated heparin adminis-tered in the hospital for proximal deep-vein thrombosis. N Engl J Med. 1996;334:677-681.

5. Koopman MM, Prandoni P, Piovella F, et al, Tasman Group. Treatmentof venous thrombosis with intravenous unfractionated heparin administered inthe hospital as compared with subcutaneous low-molecular-weight heparinadministered at home. N Engl J Med. 1996;334:682-687.

6. Prandoni P, Lensing AW, Buller HR, et al. Comparison of subcutaneouslow-molecular-weight heparin with intravenous standard heparin in proximaldeep-vein thrombosis. Lancet. 1992;339:441-445.

7. Green D, Hull RD, Brant R, Pineo GF. Lower mortality in cancer patientstreated with low-molecular-weight heparin versus standard heparin [letter].Lancet. 1992;339:1476.

8. Siragusa S, Cosmi B, Piovella F, Hirsh J, Ginsberg JS. Low-molecular-weight heparins and unfractionated heparin in the treatment of patients withacute venous thromboembolism: results of meta-analysis. Am J Med.1996;100:269-277.

9. Gould M, Dembitzer A, Doyle RL, Hastie TJ, Garber AM. Low-molecu-lar-weight heparins compared with unfractionated heparin of treatment ofacute deep venous thrombosis: a meta-analysis of randomized, controlledtrials. Ann Intern Med. 1999;130:800-809.

10. Donati MB, Poggi A. Malignancy and haemostasis. Br J Haematol.1980;44:173-182.

11. Zacharski LR, Henderson WG, Rickles FR, et al. Rationale and experi-mental design for the VA Cooperative Study of Anticoagulation (Warfarin) inthe Treatment of Cancer. Cancer. 1979;44:732-741.

12. Zacharski LR, Henderson WG, Rickles FR, et al. Effect of warfarinanticoagulation on survival in carcinoma of the lung, colon, head and neck, andprostate: final report of VA Cooperative Study #75. Cancer. 1984;53:2046-2052.

such as would be observed if LMWH were actually 10%superior in terms of median survival. Hence, despite havingstopped the study early, our results minimize the possibilityof any real survival advantage due to LMWH.

CONCLUSION

This trial demonstrated the safety of the administeredLMWH in patients with advanced incurable cancer butfailed to provide any information to suggest that theLMWH prolonged survival or improved patients’ QOL.

Additional Participating Institutions. Duluth CommunityClinical Oncology Program (CCOP), Duluth, Minn (Daniel A.Nikcevich, MD); Medcenter One Health Systems, Bismarck,ND, Mid Dakota Clinic, Bismarck, ND (Edward Wos, MD);CentraCare Clinic, St Cloud, Minn (Harold E. Windschitl, MD);Sioux Community Cancer Consortium, Sioux Falls, SD (Loren K.Tschetter, MD); Geisinger Clinic and Medical Center CCOP,Danville, Pa (Albert Bernath, MD); Michigan Cancer Con-sortium, Ann Arbor (Philip J. Stella, MD); Missouri ValleyCancer Consortium, Omaha, Neb (James A. Mailliard, MD).




13. Chahinian AP, Propert KJ, Ware JH, et al. A randomized trial of antico-agulation with warfarin and of alternating chemotherapy in extensive small-cell lung cancer by the Cancer and Leukemia Group B. J Clin Oncol. 1989;7:993-1002.

14. Kingston RD, Fielding JW, Palmer MK. Peri-operative heparin: a pos-sible adjuvant to surgery in colo-rectal cancer? Int J Colorectal Dis. 1993;8:111-115.

15. Lebeau B, Chastang C, Berchot JM, et al, “Petites Cellules” Group.Subcutaneous heparin treatment increases survival in small cell lung cancer.Cancer. 1994;74:38-45.

16. Rickles FR, Edwards RL. Activation of blood coagulation in cancer:Trousseau’s syndrome revisited. Blood. 1983;62:14-31.

17. Thornes RD. Fibrinogen and the interstitial behaviour of cancer. In:Wissler RW, Dao TL, Wood S Jr, eds. Endogenous Factors Influencing Host-Tumor Balance. Chicago, Ill: University of Chicago Press; 1967:255-266.

18. Bar-Shavit R, Kahn AJ, Mann KG, Wilner GD. Identification of athrombin sequence with growth factor activity on macrophages. Proc NatlAcad Sci U S A. 1986;83:976-980.

19. Rosenbaum J, Tobelem G, Molho P, Barzu T, Caen JP. Modulation ofendothelial cells growth induced by heparin. Cell Biol Int Rep. 1986;10:437-446.

20. Folkman J, Langer R, Lindhart RJ, Haudenschild C, Taylor S. Angiogen-esis inhibition and tumor regression caused by heparin or a heparin fragment inthe presence of cortisone. Science. 1983;221:719-725.

21. Crum R, Szabo S, Folkman J. A new class of steroids inhibits angiogenesisin the presence of heparin or a heparin fragment. Science. 1985;230:1375-1378.

22. Hilgard P, Thornes RD. Anticoagulants in the treatment of cancer. Eur JCancer. 1976;12:755-762.

23. Spitzer WO, Dobson AJ, Hall J, et al. Measuring the quality of life ofcancer patients: a concise QL-index for use by physicians. J Chronic Dis.1981;34:585-597.

24. Bretscher M, Rummans T, Sloan J, et al. Quality of life in hospicepatients: a pilot study. Psychosomatics. 1999;40:309-313.

25. McCorkle R, Quint-Benoliel J. Symptom distress, current concerns andmood disturbances after diagnosis of life-threatening disease. Soc Sci Med.1983;17:431-438.

26. Degner LF, Sloan JA. Symptom distress in newly diagnosed ambulatorycancer patients and as a predictor of survival in lung cancer. J Pain SymptomManage. 1995;10:423-431.

27. Sloan JA, Loprinzi CL, Laurie JA, et al. A simple stratification factorprognostic for survival in advanced cancer: the Good/Bad/Uncertain index. JClin Oncol. 2001;19:3539-3546.

28. Pocock SJ, Simon R. Sequential treatment assignment with balancingfor prognostic factors in the controlled clinical trial. Biometrics. 1975;31:103-115.

29. Peto R, Peto J. Asymptotically efficient rank invariant test procedures. JR Stat Soc [A]. 1972;135:185-206.

30. Kaplan EL, Meier P. Nonparametric estimation from incomplete obser-vations. J Am Stat Assoc. 1958;53:457-481.

31. Kakkar AK, Levine MN, Kadziola Z, et al. Low molecular weightheparin, therapy with dalteparin, and survival in advanced cancer: the FragminAdvanced Malignancy Outcome Study (FAMOUS). J Clin Oncol. 2004;22:1944-1948.

32. Klerk CP, Smorenburg SM, Otten HM, et al. The effect of low molecularweight heparin on survival in patients with advanced malignancy. J ClinOncol. 2005;23:2130-2135.

33. Kardinal CG, Loprinzi CL, Schaid DJ, et al. A controlled trial of cypro-heptadine in cancer patients with anorexia and/or cachexia. Cancer.1990;65:2657-2662.

34. Loprinzi CL, Ellison NM, Schaid DJ, et al. Controlled trial of megestrolacetate for the treatment of cancer anorexia and cachexia. J Natl Cancer Inst.1990;82:1127-1132.

35. Loprinzi CL, Michalak JC, Schaid DJ, et al. Phase III evaluation of fourdoses of megestrol acetate as therapy for patients with cancer anorexia and/orcachexia. J Clin Oncol. 1993;11:762-767.

36. Goldberg RM, Loprinzi CL, Mailliard JA, et al. Pentoxifylline for treat-ment of cancer anorexia and cachexia? a randomized, double-blind, placebo-controlled trial. J Clin Oncol. 1995;13:2856-2859.

37. Loprinzi CL, Kugler JW, Sloan JA, et al. Randomized comparisonof megestrol acetate versus dexamethasone versus fluoxymesterone forthe treatment of cancer anorexia/cachexia. J Clin Oncol. 1999;17:3299-3306.


78 The Open Cardiovascular Medicine Journal, 2010, 4, 78-82

1874-1924/10 2010 Bentham Open

Open Access

Cancer-Associated Thrombosis

Mehran Karimi* and Nader Cohan

Hematology Research Center, Shiraz University of Medical Sciences, Shiraz, Iran

Abstract: Thrombosis is a common complication in patients with cancer and it is estimated that about 20% of patients with cancer experience venous thromboembolism (VTE). This complication is associated with high rate of morbidity and mortality and is sometimes the first manifestation of an occult cancer. The risk profiles and markers involved in cancer-associated thrombosis share similarities with inflammation-induced atherosclerosis and thrombosis. The type of cancer, chemotherapy, surgery, central venous catheters, pre-chemotherapy platelet and leukocyte count are associated with high risk of VTE in cancer patients. Landmark studies demonstrated that effective prophylaxis and treatment of VTE reduced morbidity and increased survival. Low-molecular-weight heparin (LMWH) is preferred as an effective and safe means for prophylaxis and treatment of VTE. It has largely replaced unfractionated heparin and vitamin K antagonists. The advan-tages of LMWH include increased survival and quality of life, decreased rate of VTE, low incidence of thrombocytopenia. New guidelines for prophylaxis and treatment are now available and prophylaxis is recommended in hospitalized cancer patients and patients undergoing major surgery. Treatment with LMWH should be considered as the first line of therapy for established VTE and to prevent recurrent thrombosis in patients with cancer.

Keywords: Cancer, Thrombosis, Low-molecular-weight heparin.

INTRODUCTION

Thrombosis, a well-recognized complication of cancer, may be the first manifestation of malignancy and is associ-ated with a high rate of morbidity and mortality [1-4]. This association was first described by Armand Trousseau in 1865 and the condition still often called Trousseau’s syndrome [5-7]. The clinical manifestations of thrombosis in cancer vary from venous thromboembolism (VTE) to disseminated in-travascular coagulation, which is more commonly seen in hematological malignancies [8]. Venous thromboembolism in patients with cancer may present as a vast range of clini-cally significant thrombotic complications including deep vein thrombosis, pulmonary embolism, arterial thrombosis, nonbacterial thrombotic endocarditis, superficial thrombo-phlebitis, catheter-related thrombosis and hepatic venoocclu-sive disease [9-11]. Certain malignancies, particularly mucin-secreting ade-nocarcinomas of the ovary, pancreas, stomach, brain tumors and hematological malignancies, are associated with a higher risk of VTE [12-16]. Some conditions are well-known risk factors for increased risk of thrombosis in patients with can-cer. Chemotherapy is one of the most important risk factors for increased risk of VTE [10-12]. The strongest clinical relationship between chemotherapy and thrombosis was found in patients with breast cancer receiving chemotherapy [10, 12-15]. In a study by the Eastern Cooperative Oncology Group, VTE was significantly more common in patients with

*Address correspondence to this author at the Hematology Research Center, Nemazee Hospital of Shiraz University of Medical Sciences, Zand st, Shiraz, Iran; Tel/Fax: +98 711 6473239; E-mail: [email protected]

breast cancer who underwent chemotherapy and hormonal therapy than in the control group [16]. A high incidence of VTE following chemotherapy was also reported in other cancers [17, 18]. Chemotherapy increased the risk of VTE and recurrent VTE 6-fold and 2-fold, respectively in patients with cancer, and it is estimated that the annual incidence of VTE in cancer patients undergoing chemotherapy is about 10.9% [10]. Surgery is estimated to increase the risk of postoperative VTE about 2-fold in patients with cancer compared to pa-tients without cancer who underwent surgery and was asso-ciated with a 3-fold to 4-fold increase in the likelihood of developing pulmonary embolism after surgery [19-21]. In contrast, some studies did not show an increased risk of VTE associated with surgery in patients with cancer [14, 22]. One study that analyzed the effect of neurosurgery on the risk of thrombosis in patients with glioma revealed that these patients were 70% more likely to develope VTE compared to patients who did not undergo surgery [14]. Other risk factors for thrombosis in cancer are central venous catheters, immobilization, oral contraceptive use, trauma, previous vein thrombosis, hormonal therapy, pregnancy, older age, prothrombotic mutations such as factor V leiden and prothrombin 20210A, elevated D-dimer levels, elevated C-reactive protein, elevated soluble P-selectin, body mass index ≥35 kg/m2, antiphospholipid antibody and several biomarkers such as pre-chemotherapy platelet count over 350 × 103/µL or leukocyte count over 11 × 103/µL [14, 22-26]. Also it is noted that risk factors and markers involved in cancer-associated thrombosis, inflammation and atheroscle-rosis share similarities. Thrombogenic risk factors such as tissue factor reported as a promotion factor for a human

Cancer-Associated Thrombosis The Open Cardiovascular Medicine Journal, 2010, Volume 4 79

coronary atherosclerosis plague. In addition, platelets can release pro-inflammatory substances relevant to athero- thrombosis. Beside the cardiovascular risk factors including; family history of coronary heart disease, smoking, high cholesterol level, hypertension, age and diabetic mellitus several markers including fibrinogen level, markers of fibrinolytic systems and inflammation markers such as C-reactive protein is evaluated as risk factors of cardiovascular disease [27, 28]. The aim of this review was to overview evidence on pathophysiology, prophylaxis and treatment of cancer-associated thrombosis.

PATHOPHYSIOLOGY OF CANCER-ASSOCIATED THROMBOSIS

The pathophysiology of thrombosis formation and blood coagulation in cancer is complex and reflects different mechanisms that are generally related to the host response to cancer. These mechanisms include activation of the coagula-tion and fibrinolytic systems, acute phase reaction, inflam-mation, necrosis and cytokine production [5, 8, 29, 30]. Malignant cells can directly activate blood coagulation by producing tissue factor, cancer procoagulant activity, inflammatory reactions and cytokines [9, 12, 29, 31]. Tissue factor, an important coagulation factor that has been reported in many types of cancers [32], is constitutively expressed on solid malignant cells and acute myelogenous leukemia cells. It has been reported to promote thrombotic state [30, 31, 33]. Cancer procoagulant is a 68-kDa cysteine endopeptidase that can directly activate coagulation factor X. It is released by many tumor cells and its activity promotes thrombosis. It

is also has been shown that cancer procoagulant can cause platelet activation [8, 10, 34, 35].

Malignant cells release various types of cytokines includ-ing interleukin (IL)-1β, tumor necrosis factor-α and vascular endothelial growth factor (VEGF), which have important effects on coagulation. These cytokines can induce tissue factor production by vascular endothelial cells, downregulate thrombomodulin expression, produce plasminogen activator inhibitor-1 and increase endothelial cell adhesion molecule expression. This last effect increases the capacity of the ves-sel wall to attach leukocytes and platelets, promoting local-ized clotting factor activation and thrombosis formation [8-10, 29, 36]. Angiogenesis is an important process in patho-physiology of cancer. VEGF and angiopoietins with their receptors, Flt-1 for VEGF and Tie-2 for angiopoietin, that are the most potent proangiogenic factors are involved in the pathogenesis of cancer. Abnormal levels of VEGF, angio-poietins, IL-6 (an inflammatory cytokine) and soluble P selectin and their receptors are established in breast and other cancers that treatment is effective in reduce levels of these markers [37-39].

Interactions between malignant and endothelial host cells, platelets and leukocytes are another mechanism by which tumor cells promote thrombosis. The attachment of malignant cells can promote thrombosis by activating local-ized clotting factor, favoring platelet aggregation and activat-ing leukocytes which then release their cytokines [8, 10, 29, 40]. It has also been shown that inflammation induced by cancer can increase acute-phase proteins including fibrino-gen, coagulation factor VIII and von Willebrand factor, which can promote thrombosis [9, 41]. Fig. (1) illustrates

Fig. (1). Factors involved in cancer-associated thrombosis.

Tumor cells

CP, TF Inflammation and cytokinergic reactions

involving TNF-α, IL-1β, VEGF, fibrinogen,

FVIII, vWF

Cells adhesion and interaction: endothelial cells,

monocyte/macrophage, T and B lymphocytes interaction, platelet

hyperactivation with the release of ADP, thrombin, adhesion molecules

Activation of coagulation

Clot formation

Thrombotic vascular events

80 The Open Cardiovascular Medicine Journal, 2010, Volume 4 Karimi and Cohan

the mechanism of thrombosis formation induced by cancer cells.

EPIDEMIOLOGY OF CANCER-ASSOCIATED THROMBOSIS

It is estimated that about 4-20% of patients with cancer experience venous thrombosis [7, 10, 25, 30, 42, 43] with the annual incidence of 0.5% in cancer patients compared to 0.1% in the general population [25]. In patients with cancer, the risk of VTE is estimated to be 4-fold to 7-fold higher than in patients without cancer [44]. Venous thromboem-bolism and thrombotic complications are the second most frequent cause of mortality in patients with cancer [10]. Several studies have showed that the incidence of VTE is associated with the duration of the underlying illness. The highest rate of VTE is seen in the initial period after diagno-sis [14, 29, 45] and mortality from VTE is highest 1 year after diagnosis [14]. Alcalay et al. showed that the incidence of VTE in patients with colorectal cancer is about 5.0% dur-ing the first 6 months after diagnosis, 1.4% during the subse-quent 7-12 months and 0.6% 13-24 months after diagnosis [46]. The appearance of venous thrombosis has been clearly associated with metastatic cancer and the stage of cancer [14, 29, 42, 44]. About 10% of patients with idiopathic throm-botic complications are diagnosed with malignancy within a few years after thrombotic events and approximately 40% of them have metastatic cancer at diagnosis [47, 48]. As a result VTE is sometimes the first manifestation of occult cancer [10, 29, 42, 47]. In a study by Monreal et al. the rate of oc-cult cancer in patients with idiopathic VTE was 2.2%-12% [48]. More advanced stages of cancer on initial diagnosis are also related with a higher incidence of VTE [14, 45, 46]. The presence of medical comorbodity has an adverse effect on prognosis and survival in patients with cancer who also have thrombosis [47]. Khorana et al. proposed a simple risk scoring system based on clinical and laboratory vari-ables to predict chemotherapy-associated VTE in patients with cancer [25, 49]. They identified five variables based on the site of cancer, pre-chemotherapy platelet and leukocyte count, hemoglobin level and body mass index. This model predicts a risk of chemotherapy-associated VTE in of about 7% in patients with cancer [25].

PREVENTION AND TREATMENT OF VENOUS THROMBOEMBOLISM IN CANCER

Effective prophylaxis and treatment of VTE can reduce morbidity and mortality due to thrombosis. Low-molecular-weight-heparin (LMWH) is the first choice for prophylaxis and treatment of acute VTE, having largely replaced unfrac-tionated heparin and oral vitamin K antagonists such as warfarin [50, 51]. Warfarin is a long-term anticoagulant for the prevention and treatment of VTE that is given after initial therapy with unfractionated heparin or LMWH to maintain an interna-tional normalized ratio of 2-3 [42, 52]. But warfarin treat-ment in patients with cancer has several problems that have limited its use. Long-term treatment with warfarin increases the risk of bleeding and recurrent VTE in patients with cancer [47, 53, 54]. Difficulties in dose adjustment also limit its use because the anticoagulant effect may reach to its peak

after 3-4 days and its clearance from plasma is slow. Also warfarin may interact with chemotherapeutic agents and foods—interactions that also make this drug difficult to manage [12]. The CLOT trial (Comparison of Low-Molecular-Weight Heparin Versus Oral Anticoagulant Therapy for the Preven-tion of Recurrent Venous Thromboembolism in Patients with Cancer) compared the efficacy of LMWH plus dalteparin and oral warfarin in preventing recurrent VTE in patients with cancer. Recurrent VTE was found in 27 of 336 patients in the LMWH group compared to 53 of 336 patients in the warfarin group (hazard ratio 0.48, P=0.002) [55]. Low-molecular-weight heparin and dalteparin reduced the risk of recurrent VTE by 52% compared to warfarin therapy [56]. Unfractionated heparin is also widely used for the treat-ment and prophylaxis of thrombophylactic events and its efficacy is similar to that of LMWH [56]. A metaanalysis revealed that unfractionated heparin decreased the incidence of deep vein thrombosis and pulmonary embolism by 56% and 58% respectively compared to the control group [57]. The most important limitation of unfractionated heparin is the appearance of heparin-induced thrombocytopenia, which is significantly less frequent with LMWH [9, 29]. Treatment with LMWH led to a significant proportion of improvements in thrombosis and higher survival compared to unfraction-ated heparin [9]. These benefits make subcutaneous (sc) LMWH the first-line choice for the treatment and prophy-laxis of thrombosis in patients with cancer.

Randomized trials to test different types of prophylaxis for VTE in patients with cancer showed significantly lower rates of VTE with LMWH compared to unfractionated hepa-rin or a placebo [58-60]. New treatment guidelines from the National Comprehensive Cancer Network (NCCN) and American Society of Clinical Oncology (ASCO) accept prophylaxis for VTE in hospitalized cancer patients in the absence of major bleeding or other contraindications to anti-coagulants [50, 61]. Contraindications to anticoagulation according to ASCO guidelines include uncontrollable bleed-ing, active cerebrovascular hemorrhage, dissecting or cere-bral aneurysm, bacterial endocarditis, active peptic or other gastrointestinal ulceration, severe uncontrolled or malignant hypertension, severe head trauma, pregnancy (warfarin), heparin-induced thrombocytopenia and epidural catheter placement. It is recommended that patients undergoing major surgery for cancer should be considered candidates for thrombophylaxis, although routine prophylaxis is not recommended in ambulatory cancer patients without VTE except in patients with myeloma receiving thalidomide or lenalidomide treatment [50]. Low-molecular-weight heparin, dalteparin 5000 units s.c daily, enoxaparin 40 mg s.c daily or fondaparinux 2.5 mg s.c daily is recommended for prophy-laxis of VTE in patients with cancer [50, 61, 62]. A random-ized clinical trial of combination chemotherapy with LMWH with dalteparin 5000 units once daily for 18 weeks compared to chemotherapy alone in patients with small-cell lung cancer revealed that combination therapy with dalteparin increased disease-free survival (10 months) compared chemotherapy alone (6 months, P=0.01) [63].

Cancer-Associated Thrombosis The Open Cardiovascular Medicine Journal, 2010, Volume 4 81

CONCLUSIONS

Venous thromboembolism is a serious complication and the second most frequent cause of death in patients with cancer. The appearance of VTE reduces survival in cancer patients compared to those without VTE and adversely affects quality of life [14]. It is estimated that thrombosis in patients with cancer increases the risk of death 4-fold to 8-fold compared to patients without cancer [47]. Landmark studies showed that anticoagulant therapy and thrombopro-phylaxis are efficacious and can protect patients from VTE. Based on clinical trial findings, subcutaneous LMWH is the first line therapy for VTE in patients with cancer and has largely replaced unfractionated heparin and vitamin K antagonists. New treatment and prophylaxis guidelines are now available for the management of thrombotic events and these guidelines can decrease complications and morbidity and increase survival and quality of life in patients with cancer.

ACKNOWLEDGEMENTS

We thank Shirin Parand (Hematology Research Center, Shiraz University of Medical Sciences) and K. Shashok (AuthorAID in the Eastern Mediterranean) for improving the use of English in the manuscript.

ABBREVIATIONS

CP = Cancer procoagulant TF = Tissue factor TNF-α = Tumor necrosis factor-α IL-1β = Interleukin-1β VEGF = Vascular endothelial growth factor FVIII = Factor VIII vWF = Von Willebrand factor ADP = Adenosine diphosphate

REFERENCES [1] Lee AY, Levine MN. Venous thromboembolism and cancer: risks

and outcomes. Circulation 2003; 107: I17-21. [2] White RH, Chew HK, Zhou H, et al. Incidence of venous throm-

boembolism in the year before the diagnosis of cancer in 528,693 adults. Arch Intern Med 2005; 165: 1782-7.

[3] Prandoni P, Piccioli A, Girolami A. Cancer and venous throm-boembolism: an overview. Haematologica 1999; 84: 437-45.

[4] Caine GJ, Stonelake PS, Lip GY, Kehoe ST. The hypercoagulable state of malignancy: pathogenesis and current debate. Neoplasia 2002; 4: 465-73.

[5] Bick RL. Cancer-associated thrombosis. N Engl J Med 2003; 349: 109-11.

[6] Sato T, Tsujino I, Ikeda D, Ieko M, Nishimura M. Trousseau's syndrome associated with tissue factor produced by pulmonary adenocarcinoma. Thorax 2006; 61: 1009-10.

[7] Khorana AA, Francis CW, Culakova E, Kuderer NM, Lyman GH. Thromboembolism is a leading cause of death in cancer patients re-ceiving outpatient chemotherapy. J Thromb Haemost 2007; 5: 632-4.

[8] Rickles FR, Falanga A. Molecular basis for the relationship be-tween thrombosis and cancer. Thromb Res 2001; 102: V215-24.

[9] Deitcher SR. Cancer and thrombosis: mechanisms and treatment. J Thromb Thrombolysis 2003; 16: 21-31.

[10] Haddad TC, Greeno EW. Chemotherapy-induced thrombosis. Thromb Res 2006; 118: 555-68.

[11] Gomes MP, Deitcher SR. Diagnosis of venous thromboembolic disease in cancer patients. Oncology (Williston Park) 2003; 17: 126-35.

[12] Falanga A, Zacharski L. Deep vein thrombosis in cancer: the scale of the problem and approaches to management. Ann Oncol 2005; 16: 696-701.

[13] Pritchard KI, Paterson AH, Paul NA, Zee B, Fine S, Pater J. In-creased thromboembolic complications with concurrent tamoxifen and chemotherapy in a randomized trial of adjuvant therapy for women with breast cancer. National Cancer Institute of Canada Clinical Trials Group Breast Cancer Site Group. J Clin Oncol 1996; 14: 2731-7.

[14] Wun T, White RH. Venous thromboembolism (VTE) in patients with cancer: epidemiology and risk factors. Cancer Invest 2009; 27 (Suppl 1): 63-74.

[15] Blom JW, Doggen CJ, Osanto S, Rosendaal FR. Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA 2005; 293: 715-22.

[16] Saphner T, Tormey DC, Gray R. Venous and arterial thrombosis in patients who received adjuvant therapy for breast cancer. J Clin Oncol 1991; 9: 286-94.

[17] Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004; 350: 2335-42.

[18] Rubbia-Brandt L, Audard V, Sartoretti P, et al. Severe hepatic sinusoidal obstruction associated with oxaliplatin-based chemo-therapy in patients with metastatic colorectal cancer. Ann Oncol 2004; 15: 460-6.

[19] Donati MB. Cancer and thrombosis. Haemostasis 1994; 24: 128-31.

[20] Prandoni P. Antithrombotic strategies in patients with cancer. Thromb Haemost 1997; 78: 141-4.

[21] Levitan N, Dowlati A, Remick SC, et al. Rates of initial and recur-rent thromboembolic disease among patients with malignancy ver-sus those without malignancy. Risk analysis using Medicare claims data. Medicine (Baltimore) 1999; 78: 285-91.

[22] Heit JA, Silverstein MD, Mohr DN, Petterson TM, O'Fallon WM, Melton LJ 3rd. Risk factors for deep vein thrombosis and pulmo-nary embolism: a population-based case-control study. Arch Intern Med 2000; 160: 809-15.

[23] Goldhaber SZ, Tapson VF; DVT FREE Steering Committee. A prospective registry of 5,451 patients with ultrasound-confirmed deep vein thrombosis. Am J Cardiol 2004; 93: 259-62.

[24] Chasan-Taber L, Stampfer MJ. Epidemiology of oral contracep-tives and cardiovascular disease. Ann Intern Med 1998; 128: 467-77.

[25] Sud R, Khorana AA. Cancer-associated thrombosis: risk factors, candidate biomarkers and a risk model. Thromb Res 2009; 123 (Suppl 4): S18-21.

[26] Horowitz N, Brenner B. Thrombophilia and cancer. Pathophysiol Haemost Thromb 2008; 36: 131-6.

[27] Packard RR, Libby P. Inflammation in atherosclerosis: from vascu-lar biology to biomarker discovery and risk prediction. Clin Chem 2008; 54: 24-38.

[28] Libby P, Ridker PM, Maseri A. Inflammation and atherosclerosis. Circulation 2002; 105: 1135-43.

[29] Prandoni P, Falanga A, Piccioli A. Cancer, thrombosis and heparin-induced thrombocytopenia. Thromb Res 2007; 120 (Suppl 2): S137-40.

[30] López JA, Chen J. Pathophysiology of venous thrombosis. Thromb Res 2009; 123 (Suppl 4): S30-4.

[31] Falanga A, Donati MB. Pathogenesis of thrombosis in patients with malignancy. Int J Hematol 2001; 73: 137-44.

[32] Zwicker JI. Tissue factor-bearing microparticles and cancer. Semin Thromb Hemost 2008; 34: 195-8.

[33] Davila M, Amirkhosravi A, Coll E, et al. Tissue factor-bearing microparticles derived from tumor cells: impact on coagulation ac-tivation. J Thromb Haemost 2008; 6: 1517-24.

[34] Lee AY. Cancer and thromboembolic disease: pathogenic mecha-nisms. Cancer Treat Rev 2002; 28: 137-40.

[35] Rickles FR, Hair GA, Zeff RA, Lee E, Bona RD. Tissue factor expression in human leukocytes and tumor cells. Thromb Haemost 1995; 74: 391-5.

[36] Grignani G, Maiolo A. Cytokines and hemostasis. Haematologica 2000; 85: 967-72.

82 The Open Cardiovascular Medicine Journal, 2010, Volume 4 Karimi and Cohan [37] Caine GJ, Lip GY, Zanetto U, Maheshwari M, Stonelake PS, Blann

AD. A comparison of plasma versus histologic indices of angio-genic markers in breast cancer. Appl Immunohistochem Mol Mor-phol 2007; 15: 382-8.

[38] Caine GJ, Stonelake PS, Lip GY, Blann AD. Changes in plasma vascular endothelial growth factor, angiopoietins, and their recep-tors following surgery for breast cancer. Cancer Lett 2007; 248: 131-6.

[39] Caine GJ, Lip GY, Blann AD. Platelet-derived VEGF, Flt-1, an-giopoietin-1 and P-selectin in breast and prostate cancer: further evidence for a role of platelets in tumour angiogenesis. Ann Med 2004; 36: 273-7.

[40] Lo SK, Cheung A, Zheng Q, Silverstein RL. Induction of tissue factor on monocytes by adhesion to endothelial cells. J Immunol 1995; 154: 4768-77.

[41] Deitcher SR, Carman TL, Sheikh MA, Gomes M. Hypercoagulable syndromes: evaluation and management strategies for acute limb ischemia. Semin Vasc Surg 2001; 14: 74-85.

[42] Dotsenko O, Kakkar AK. Thrombosis and cancer. Ann Oncol 2006; 17 (Suppl 10): x81-4.

[43] Heit JA, O'Fallon WM, Petterson TM, et al. Relative impact of risk factors for deep vein thrombosis and pulmonary embolism: a popu-lation-based study. Arch Intern Med 2002; 162: 1245-8.

[44] Stein PD, Beemath A, Meyers FA, Skaf E, Sanchez J, Olson RE. Incidence of venous thromboembolism in patients hospitalized with cancer. Am J Med 2006; 119: 60-8.

[45] Chew HK, Wun T, Harvey DJ, Zhou H, White RH. Incidence of venous thromboembolism and the impact on survival in breast can-cer patients. J Clin Oncol 2007; 25: 70-6.

[46] Alcalay A, Wun T, Khatri V, et al. Venous thromboembolism in patients with colorectal cancer: incidence and effect on survival. J Clin Oncol 2006; 24: 1112-8.

[47] Piccirillo JF, Tierney RM, Costas I, Grove L, Spitznagel EL Jr. Prognostic importance of comorbidity in a hospital-based cancer registry. JAMA 2004; 291: 2441-7.

[48] Monreal M, Trujillo-Santos J. Screening for occult cancer in pa-tients with acute venous thromboembolism. Curr Opin Pulm Med 2007; 13: 368-71.

[49] Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Development and validation of a predictive model for chemother-apy-associated thrombosis. Blood 2008; 111: 4902-7.

[50] Lyman GH, Khorana AA, Falanga A, et al. American Society of Clinical Oncology guideline: recommendations for venous throm-boembolism prophylaxis and treatment in patients with cancer. J Clin Oncol 2007; 25: 5490-505.

[51] Büller HR, Agnelli G, Hull RD, Hyers TM, Prins MH, Raskob GE. Antithrombotic therapy for venous thromboembolic disease: the

Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126 (Suppl 3): 401S-28S.

[52] Lee AY. Treatment of venous thromboembolism in cancer patients. Thromb Res 2001; 102: V195-208.

[53] Murchison JT, Wylie L, Stockton DL. Excess risk of cancer in patients with primary venous thromboembolism: a national, popu-lation-based cohort study. Br J Cancer 2004; 91: 92-5.

[54] Hutten BA, Prins MH, Gent M, Ginsberg J, Tijssen JG, Büller HR. Incidence of recurrent thromboembolic and bleeding complications among patients with venous thromboembolism in relation to both malignancy and achieved international normalized ratio: a retro-spective analysis. J Clin Oncol 2000; 18: 3078-83.

[55] Lee AY, Levine MN, Baker RI, et al. Low-molecular-weight hepa-rin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer. N Engl J Med 2003; 349: 146-53.

[56] Lee A. VTE in patients with cancer--diagnosis, prevention, and treatment. Thromb Res 2008; 123 (Suppl 1): S50-4.

[57] Hirsh J, Warkentin TE, Raschke R, Granger C, Ohman EM, Dalen JE. Heparin and low-molecular-weight heparin: mechanisms of action, pharmacokinetics, dosing considerations, monitoring, efficacy, and safety. Chest 1998; 114 (5 Suppl): 489S-510S.

[58] McLeod RS, Geerts WH, Sniderman KW, et al. Subcutaneous heparin versus low-molecular-weight heparin as thromboprophy-laxis in patients undergoing colorectal surgery: results of the cana-dian colorectal DVT prophylaxis trial: a randomized, double-blind trial. Ann Surg 2001; 233: 438-44.

[59] Cohen AT, Davidson BL, Gallus AS, et al. Efficacy and safety of fondaparinux for the prevention of venous thromboembolism in older acute medical patients: randomised placebo controlled trial. BMJ 2006; 332: 325-9.

[60] Alikhan R, Cohen AT, Combe S, et al. Prevention of venous thromboembolism in medical patients with enoxaparin: a subgroup analysis of the MEDENOX study. Blood Coagul Fibrinolysis 2003; 14: 341-6.

[61] Khorana AA. The NCCN Clinical Practice Guidelines on Venous Thromboembolic Disease: strategies for improving VTE prophylaxis in hospitalized cancer patients. Oncologist 2007; 12: 1361-70.

[62] Geerts WH, Pineo GF, Heit JA, et al. Prevention of venous throm-boembolism: the Seventh ACCP Conference on Antithrom- botic and Thrombolytic Therapy. Chest 2004; 126 (3 Suppl): 338S-400S.

[63] Altinbas M, Coskun HS, Er O, et al. A randomized clinical trial of combination chemotherapy with and without low-molecular-weight heparin in small cell lung cancer. J Thromb Haemost 2004; 2: 1266-71.

Received: November 10, 2009 Revised: November 24, 2009 Accepted: December 11, 2009 © Karimi and Cohan; Licensee Bentham Open.

This is an open access article licensed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/ by-nc/3.0/) which permits unrestricted, non-commercial use, distribution and reproduction in any medium, provided the work is properly cited.

C

CE: Swati; MBC/200692; Total nos of Pages: 7;

MBC 200692

Original article 1

Weight-adjusted dalteparin fo
r prevention of vascularthromboembolism in advanced pancreatic cancer patientsdecreases serum tissue factor and serum-mediated inductionof cancer cell invasionAnthony Maraveyasa, Camille Ettelaieb, Hussein Echrisha, Chao Lib,Eric Gardinerc, John Greenmana and Leigh A. Maddena
The aim of the present study was to assess the role of tissue

factor and serum-induced cell invasion in patients with

advanced pancreatic cancer (APC). A cohort of 39 patients

with APC, without thrombosis, receiving chemotherapy,

were entered in a randomized controlled trial

(ISRCTN U 76464767) of thromboprevention with weight-

adjusted dalteparin (WAD). A total of 19 patients received

WAD, the remaining 20 acting as a control group. Serum

from baseline and week 8 was analysed for circulating-

tissue factor antigen using ELISA. Circulating-tissue factor

antigen rose from 324 pg/ml, [interquartile range (IQR)

282–347 pg/ml] to 356 pg/ml, (IQR 319–431 pg/ml) in

controls (C), and decreased in the dalteparin-treated group

(D) from 336 pg/ml (IQR 281–346 pg/ml) to 303 pg/ml (IQR

274–339 pg/ml). The difference in median percentage

change between D and C was statistically significant [S4.0

(D) vs. 4.7 (C); P U 0.005, n U 39]. Serum-induced cellular

invasion of MIA-Paca-2 cells in response to patient serum

was studied using a Boyden chamber assay in 30 patients

(14 WAD and 16 C). The median percentage change

between C and D was significant [R54.9 (C) vs. S21.9 (D)

P U 0.025, n U 30]. There was a weak correlation between

opyright © Lippincott Williams & Wilkins. Unauth

0957-5235 � 2010 Wolters Kluwer Health | Lippincott Williams & Wilkins

BB-tissue factor reduction and cellular invasion reduction

(Spearman) [0.384 (P U 0.037, n U 30)]. APC patients treated

with WAD have lower tissue factor antigen levels and

attenuated induction of cellular invasion in their blood. These

assays may provide useful markers to guide appropriate

dalteparin (and other low-molecular weight heparin) dosing

schedules to optimize anticancer effects of dalteparin in APC.

Blood Coagul Fibrinolysis 21:000–000 � 2010 Wolters

Kluwer Health | Lippincott Williams & Wilkins.

Blood Coagulation and Fibrinolysis 2010, 21:000–000

Keywords: blood-borne tissue factor, dalteparin, low-molecular weightheparin, pancreatic cancer

aDivision of Cancer, Postgraduate Medical Institute in Association with Hull YorkMedical School, bBiomedical Section, Department of Biological Sciences,University of Hull, Hull, UK and cFreelance Statistical Consultant

Correspondence to Dr Leigh A. Madden, Postgraduate Medical Institute,University of Hull, Cottingham Road, Hull HU6 7RX, UKTel/fax: +44 1482 466031; e-mail: [email protected]

Received 7 December 2009 Revised 15 January 2010Accepted 14 February 2010

IntroductionConventionally, the state of aberrant coagulation in can-

cer relates to the morbidity and mortality that accom-

panies vascular thromboembolic events (VTEs) and dis-

seminated intravascular coagulation (DIC). One of the

malignancies with the greatest burden of VTE is

advanced pancreatic cancer (APC) [1,2]. Compelling

data, however, have accrued to show that mechanisms

we associate with the metastatic process in cancer (e.g.

angiogenesis, immune evasion and tissue invasion) are

also promoted by coagulation factors [3]. A pivotal mol-

ecule would appear to be tissue factor, which facilitates

the malignant process at all the abovementioned levels.

Tissue factor is a 47 kDa transmembrane glycoprotein

that forms a complex with circulating factor VIIa to

initiate the extrinsic pathway of coagulation [4]. Direct

correlations between elevated tissue factor expression

and advanced stages of malignancy have been reported

in several types of cancer, including nonsmall cell lung

[5], breast [6], pancreatic [7], prostate [8] and colorectal

cancer [9] and glioma [10]. Many of these studies have

suggested that tissue factor may also play a major role in

growth, invasion and dissemination of the tumour cells.

Previous research has shown that the expression of tissue

factor can promote tumour metastasis in a mouse model

of melanoma [11] and enhance primary tumour growth in

a pancreatic adenocarcinoma cell line [12]. Clinical stu-

dies on APC have shown that tissue factor expression

shows a significant positive correlation with histological

grade of the tumour, with 77% of poorly differentiated

pancreatic tumours expressing tissue factor compared

with only 20% of well differentiated tumours [13].

Furthermore, recently, a correlation of pancreatic cancer

tissue factor levels, as assessed immunohistochemically,

was found with clinical thrombosis in APC [14].

In contrast to tissue-associated tissue factor, the role

of circulating-tissue factor in cancer propagation is

unknown. Correlative studies of the circulating com-

ponent of tissue factor are few. An early report has shown

that patients with ovarian cancer have higher levels

of serum circulating-tissue factor antigen than normal

orized reproduction of this article is prohibited.

DOI:10.1097/MBC.0b013e328338dc49


http://dx.doi.org/10.1097/MBC.0b013e328338dc49

C


MBC 200692

2 Blood Coagulation and Fibrinolysis 2010, Vol 21 No 00

Fig. 1

FRAGEM study

(a)

(b)

Inoperable advanced pancreatic cancerStratify for PS (90−100 vs 60−80) and metastatic vs locally

advanced*

Randomization1:1

FRAGEM: Patient and Sample flow-chart

Single agentGemcitabine

(Burris schedule)And **Dalteparin(CLOT schedule)

Single agentGemcitabine

(Burris schedule)

Gemcitabine Gem+Dalteparin

*Maraveyas et al., ASCO $ ISTH 2007**All patients studied for BB-TF***Comprehensively studied for BB-TF and C-INV#Inadequate sample amount to complete duplicate studies for 9 patients (baseline week 8 or both)

(78%)

25 patients

19 patients

14 patients

100%

76%

59%

*Clinical report

*BB-TF (W 0+8)

***BB-TF & C-INV#

26 patients

20 patients

16 patients16 patients

Histology (cytology)-proven**Dalteparing is scheduled for 12 weeks scoddaily CLOT investigators” dose ‘prophylactically’ 200 U/kg week 1−4 and 150 U/kg week 5−12

51 patients randomized

(a) Randomization protocol and treatment schedule. ��Dalteparin isscheduled for 12 weeks daily per CLOT investigators’ dose’prophylactically’ 200 U/kg week 1–4 and 150 U/kg week 5–12. (b)Patient and sample flow chart. �Maraveyas et al. [19]. ��All patientsstudied for BB-tissue factor (BB-TF). ��Comprehensively studied forBB-TF and cell invasion (C-INV). #Inadequate sample amount tocomplete duplicate studies for nine patients (baseline or week 8 orboth).

Table 1 Demographics of patients randomized to the trial

Characteristics Arm A (n¼26) Arm B (n¼25)

Median age (years) 67 60Male sex (%) 50 60KPS 80–100% (%) 73 68Stage IV disease (%) 61 44

The demographics of 51 patients who had been randomized to the trial when theanalysis of the samples was undertaken. KPS, Karnofsky’s performance status.

nontumour controls and this was correlated with a worse

prognosis [15]. A single case report on a patient with APC

and death from DIC demonstrated very high levels of

tissue factor, both in the tumour and in the circulation

[7].

The central position of tissue factor in promoting

the malignant phenotype suggests that inhibition of

tissue factor might not only be beneficial through

reduction of risk of VTE, but also there may be further

benefit through interference of the cancer-promoting

properties of tissue factor. Putative anticancer effects

of anticoagulants have been demonstrated in preclinical

models [16]. Prominent agents in this respect are

the low-molecular weight heparins (LMWHs) [17]. A

major mechanism involved in metastasis, and shown in

these models to be inhibited by LMWHs, is the ability

of cancer cells to migrate and invade through stroma.

The use of LMWHs in treating patients with various

cancers with a VTE has already been shown to reduce

circulating levels of circulating-tissue factor antigen

[18].

Here, in patients with pancreatic cancer treated with

weight-adjusted dalteparin (WAD) in a primary preven-

tion setting of thrombosis, we present data that, for the

first time, demonstrate reduction in circulating tissue

factor-antigen levels and reduction in the induction of

in-vitro invasion of pancreatic cancer cells in response to

dalteparin-treated patient’s serum.

Patients and methodsPatientsPatients in this report are all involved in a prospectively

randomized controlled multicentre national (UK) trial of

primary VTE prophylaxis in APC called FRAGEM (Phase

II randomized study of chemo-anticoagulation (Gemcita-

bine_LMWH) vs chemotherapy alone (Gemcitabine) for

locally advanced and metastatic pancreatic adenocarci-

noma; ISRCTN¼ 76464767; UKCRN¼ 1290).

Among other common trial selection criteria, patients

with a preexisting VTE or, on active treatment for

VTE or for risk of VTE, are excluded. Patients are

stratified according to stage (local or metastatic) and

Karnofsky performance status (90–100 vs. 60–80) and

then are randomized to either conventional single agent

gemcitabine chemotherapy or gemcitabine and thera-

peutic dose dalteparin (Fig. 1a) [19]. Gemcitabine in

both arms is given on a weekly schedule at a dose of

1000 mg/m2 for weeks 1–7 and weeks 9–11 [20]. Patients

receiving dalteparin commenced treatment 3 days before

chemotherapy is initiated and received 200 U/kg of dal-

teparin daily as a single subcutaneous injection for 4

weeks. Then, for the remainder of the 12-week trial

period, the dose was reduced to 150 U/kg [21]. The

primary endpoint was to demonstrate a reduction in

VTE while on dalteparin.

opyright © Lippincott Williams & Wilkins. Unautho

A scientific substudy, the results of which are the subject

of this report, was designed to probe the potential effects

of WAD on circulating-tissue factor antigen and the

potential effects on markers of tissue invasion that may

become apparent in WAD-treated patients as opposed to

pretreatment baseline and the control group. Patient

demographics are shown in Table 1.

rized reproduction of this article is prohibited.

C


MBC 200692

Dalteparin for VTE in pancreatic cancer Maraveyas et al. 3

SamplingBlood samples were collected according to a standard

validated protocol to collect and store serum and plasma

that is followed by the Hull Trials Unit and conforms to

Good Clinical Laboratory Practice (GCLP) requirements

for commercial national and international research stu-

dies. Blood samples were collected in a research clinic at

baseline, weeks 4, 8, 12 (end of study treatment) and 17.

Dalteparin had been administered between 18 and 24 h

prior to attending clinic and never on the day of the

sampling. Samples were coded, aliquoted and stored at

�708C until analysis. The laboratory study was under-

taken in a blinded fashion and staff were not aware of the

treatment allocation of coded samples. From April 2003

to April 2006, 51 patients, subsequent to written informed

consent for research serum and plasma samples to be

drawn, were randomized into the FRAGEM trial [19].

This report presents data relating to the changes

observed between baseline and week 8 in patients

for whom appropriate clinical material was collected

(Fig. 1b). Week 8 was chosen as the preferred comparator

date after a longitudinal study on a limited number of

samples showed that beyond this time point, no further

increase in the observed effects in serum values was

apparent [22].

Baseline samples from four patients with metastatic

pancreatic cancer from patients screened for the trial

were the biological samples used for the tissue factor

inhibition and dalteparin-‘spiking’ experiments.

Enzyme-linked immunosorbent assay measurement ofserum circulating-tissue factor antigenThe concentration of circulating-tissue factor antigen in

the patients’ sera was measured using tissue factor-specific

paired antibody ELISA method (Affinity Biologicals,

Ancaster, Canada). The capture antibody was diluted

1/100 in coating buffer (15 mmol/l Na2CO3 buffer, pH

9.6), added (100 ml) to each microtitre plate well and

incubated overnight at 48C. The wells were then blocked

with 150 mmol/l phosphate-buffered saline (PBS), pH

7.4, containing 1% (w/v) bovine serum albumin (BSA)

for 90 min followed by four washes with PBS. Patient

sera (100 ml) were added in triplicate to the wells and

incubated for 60 min at room temperature to allow cap-

ture and then washed carefully four times with PBS.

Recombinant tissue factors of known concentrations

(American Diagnostica, Stamford, Connecticut, USA)

were used to construct a standard curve. The detect-

ing antibody [antitissue factor-horse radish peroxidase

(HRP)] was diluted 1/100 in conjugate buffer [100 mmol/l

4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

(HEPES) buffer pH 7.4, 10 mmol/l NaCl, 1% (w/v) BSA

and 1% (w/v) Tween 20)] and added (100 ml) to each well.

The plate was incubated for a further 60 min at room

temperature, washed four times with PBS and developed

by adding 100 ml of 3,30,5,50 tetramethyl benzidine (TMB)


substrate (3,30,5,5 tetramethylbenzidine; Vector Labora-

tories Ltd., Peterborough, UK) for 15 min at room

temperature. The reaction was quenched by the addition

of 2.5 mol/l H2SO4 (100 ml) to the wells and absorbance

measured at 450 nm on an Anthos 2010 microplate reader

(Anthos Labtec Instruments, Wals, Austria).

Prothrombin time assayA modified one-stage prothrombin time assay, using

dilute recombinant human Innovin tissue factor (Dade

Behring, Milton Keynes, Buckinghamshire, UK), was

undertaken. This test relies on the activity of the recom-

binant tissue factor to initiate the coagulation mechan-

ism, whereas the low activity of the 1 in 500 diluted tissue

factor allows the values to be significantly altered by the

presence of any endogenous tissue factor. To carry out

the assays, 25 mmol/l CaCl2 (100 ml) and the patient

plasma (100 ml) were incubated for 30 s before the

addition of diluted recombinant human tissue factor

(100 ml). The time taken for the clot formation was

recorded using a Cascade-M coagulometer (Helena

Laboratories, Sunderland, UK). The assay was calibrated

and data compared to a NormTrol control plasma

(Helena Laboratories). The concentration of supple-

mented tissue factor was adjusted to produce a clotting

time of 60 s� 10 using NormTrol control plasma and

used as a standard against which, the clotting times from

the patients’ plasma were measured.

Cell cultureMIA-Paca-2, a human-derived pancreatic cancer cell line,

was obtained from the LGC Promochem (Teddington,

UK) and maintained in supplemented Dulbecco’s modi-

fied eagle medium [DMEM; 90% (v/v)], fetal calf serum

[10% (v/v)], horse serum [1% (v/v)], containing 1% (v/v)

antibiotic/antimycotic solution (Sigma Chemical Com-

pany Ltd., Poole, UK). Cells were maintained at 378Cunder 5% CO2. MIA-Paca-2 cells have low endogenous

tissue factor (undetectable in our prothrombin assay) and

a low inherent invasiveness.

Investigation of induction of cellular invasion by patientseraMIA-PaCa-2 cells were used to measure the ability of

the patient sera to induce locomotion and cellular inva-

siveness. Boyden chambers (8 mm pore size) were coated

with collagen by adding 50 ml collagen type IV [1 mg/ml

(w/v)] to each chamber. The chambers were placed in a

24-well plate and incubated at 378C for 12 h. The excess

collagen was then discarded, and the cells were seeded

(2� 105 cells/well) into the upper compartment of each

chamber, on to the collagen, in 250 ml serum-free media.

Medium in the base of the well was supplemented with

patients’ serum in duplicate (10% v/v; 250 ml total

volume). The plates were incubated at 378C under 5%

CO2 for 24 h. Following incubation, the medium was

removed from the upper compartment of each chamber


C


MBC 200692


Fig. 2

Control Dalteparin250

200

150

100

50

−50

0

Boxplot of the percentage change of tissue factor antigen in the sera ofpancreatic cancer patients in both the control and dalteparin groups.

and the cells on the upper side of the filter chamber were

scraped off using a sterile cotton swap. The number of

cells that had migrated through the collagen-coated pore

was quantified by adding CellTitre Aqueous One reagent

(40 ml; Promega Inc., Southampton, UK) to the media, in

the lower chamber and the plates, which also included a

control containing media but no cells were incubated at

378C for 3 h until a colour change was clearly observed.

Finally, 240 ml of each sample was diluted with 360 ml of

distilled water in a plastic cuvettes and the absorption

values measured against the control sample at 490 nm.

The values of absorption were converted to number of

cells from a previously prepared standard curve.

Dalteparin-spiking assay and circulating-tissue factorantigen inhibition assayFor these experiments, the baseline serum cancer cell-

invasion induction (cellular invasion) was assayed from

four patients with APC entered into the trial. Patients

with the highest cellular invasion were used to assess

any direct influence of dalteparin on cellular invasion.

Samples of sera from patients with increased cellular

invasion were placed in the bottom chamber (10% v/v)

and supplemented with dalteparin (Fragmin; Pharmacia

Ltd., Milton Keynes, UK) over a range of 0–100 units/ml

(final concentration). Experiments were in duplicate and

repeated twice.

To confirm the function of tissue factor in the induction

of cell migration, aliquots of patients’ sera (n¼ 4) with

increased cellular invasion were incubated with a range of

concentrations (0–48 mg/ml final concentrations) of a

neutralizing polyclonal antihuman tissue factor antibody

(Santa Cruz Biotechnology, Heidelberg, Germany), prior

to diluting with base media. The samples were placed in

the bottom chambers and assayed using the procedure

described above. The rate of cell migration was deter-

mined following 24 h incubation.

Statistical analysisStatistical analysis was carried out using the Statistical

Package for the Social Sciences (SPSS Inc., Chicago,

USA). Percentage changes from baseline to week 8 were

calculated for circulating-tissue factor antigen levels and

proportion of cells migrated. The Mann–Whitney tests

were performed to test for differences in the median

percentage change between arms for circulating-tissue

factor antigen levels and proportion of cells migrated.

The Spearman’s rank correlation was used to correlate

percentage changes in circulating-tissue factor antigen

and proportion of cells migrated. Results are reported

on an intention-to-treat analysis. Thus, patients in the

control arm who may have been anticoagulated due to

development of VTE during the first 8 weeks of treat-

ment or patients who may have discontinued dalteparin

for clinical or personal reasons have not been excluded.


ResultsSerum tissue factor antigenMeasurement of circulating-tissue factor antigen in serum

obtained from the two cohorts showed comparable values

at the start of the investigation with significant increase

from baseline for the control group [n¼ 20; baseline

median, 324, interquartile range (IQR) 282–347 pg/ml;

8-week treatment, 356, IQR 319–431 pg/ml], and a ‘tigh-

ter’ distribution for the dalteparin-treated group when

assayed at week 8 [n¼ 19; baseline median, 336 (IQR

281–346) pg/ml to a median of 303 (IQR 274–339)

pg/ml]. The difference in median percentage change

from baseline to week 8 was significant between the groups

(�4.0 vs. 4.7; P¼ 0.005; Fig. 2). Thirty-five of the APC

patients had high levels of circulating-tissue factor antigen,

greater than the approximate normal range for this assay

[8].

Prothrombin time assayA one-step plasma prothrombin time was also done as a

measure of the integrity of the extrinsic coagulation

pathway of these patients. Measurement of prothrombin

time in plasma obtained from the two cohorts showed a

slightly lower baseline value for the dalteparin-treated

arm at the start of the investigation [median baseline

clotting times of 168 (IQR 113–192) s and 177 (IQR 128–

255) s in the dalteparin-treated and untreated groups,

respectively]. At week 8, the dalteparin-treated cohort

showed a similar median, though the IQR had increased

[median 165 (139–222) s], whereas the control group

showed no change [median 179 (IQR 127–250) s]. The

difference in median percentage change between groups

was not statistically significant (P¼ 0.075).

Investigation of induction of cellular invasion bypatients’ seraMeasurement of number of MIA-Paca-2 cells migrated at

24 h showed an increase from baseline to week 8 for the


C


MBC 200692


Fig. 3

Control Dalteparin

400

300

200

100

−100

0

Cellular invasion (migration proportion percentage change) of MIA-PaCa-2 pancreatic cells in response to patients’ sera in both the controland dalteparin groups.

Fig. 4

12000

10000

8000

6000

4000

2000

0

0 0.01 0.1

Heparin (U/ml)

Num

ber

of in

vasi

ve M

IA-P

aca-

2 ce

lls

1 10

Effect of increasing doses of extraneous dalteparin on cellular invasionof MIA-Paca-2 cells in response to patients’ sera.

Fig. 5

12000

10000

8000

6000

4000

2000

0

0 4 8

Anti TF (µg/ml)

Num

ber

of in

vasi

ve M

IA-P

aca-

2 ce

lls

16 24 48

Incubation of the patient’s serum with polyclonal antitissue factorantibody demonstrated a dose-dependent inhibitory effect on cellinvasiveness of MIA-Paca-2. TF, tissue factor.

control group and a decrease for the dalteparin-treated

group. The median value of the proportion of cells

migrated was significantly different (54.9 vs.�21.9

P¼ 0.025; Fig. 3) between groups. Spearman’s rank cor-

relation of the percentage change in circulating-tissue

factor antigen and the proportion of cells migrated in this

cohort (n¼ 30) demonstrated a weak correlation 0.384

(P¼ 0.037, n¼ 30)

Dalteparin-spiking and effect on cellular invasionThere was no effect of extraneous dalteparin on the cell

locomotion invasion of MIA-Paca-2 cells exposed to

serum (n¼ 4) from APC patients (Fig. 4). A test for linear

trend in the number of cells migrated with dalteparin

level of exposure (U/ml) was made. This was not statis-

tically significant (P¼ 0.194). However, the data pre-

sented in Fig. 4 suggest that there may be a trend for

the range of pharmacologically relevant dalteparin con-

centrations studied (0.01–1.0 U/ml) and so it is possible

that a relationship might be observed with inclusion of

more data.

Circulating-tissue factor antigen inhibition and effect oncellular invasionFurthermore, the role of tissue factor in inducing cell

invasion was demonstrated by preincubation of patient’s

sera with the antitissue factor polyclonal antibody that

showed a clear reduction in MIA-Paca-2 cellular invasion,

paralleling the increase in neutralizing antibody (Fig. 5).

DiscussionAmong the factors that confound the interpretation

of the results of clinical trials in cancer patients receiving

LMWHs is the lack of dose and schedule optimization

and differentiation of a LMWH-dose schedule that

will achieve the maximum thromboprevention benefit,

compared to one that will deliver the postulated


antineoplastic effects. This is further complicated by

the lack of surrogate markers of the antineoplastic effects

of the LMWHs or other candidate anticoagulants. The

work we report may provide useful methodology in these

two areas.

Baseline circulating-tissue factor antigen levels were

raised above the normal range for this assay in 90% of

the study patients. Circulating-tissue factor antigen levels

in the control group (nondalteparin) of APC patients rose

over the 8-week study; in contrast, the concentrations in

the dalteparin-treated patients were even slightly

reduced. The percentage change at week 8 compared

with baseline was statistically significant. Moreover, the


C


MBC 200692


distribution of the tissue factor range in the untreated

patients became much wider (outliers in boxplot), whereas

that of the patients receiving dalteparin remained rela-

tively narrow (Fig. 2).

The proportion of MIA-Paca 2 cells migrating in response

to the dalteparin-treated patient’s serum was reduced

compared with the untreated patients. The difference in

percentage change at week 8 compared with baseline was

statistically significant. A Spearman’s correlation analysis

between cellular invasion and BB-tissue factor levels

showed a weak but significant correlation. A number of

reasons for this are possible, for example, too small a

number of patients studied, variability in in-vivo effect of

WAD dosing, the underlying effect of the chemotherapy

on the tumour over the 8 weeks, the intention-to-treat

type of analysis or any combination of these factors. Also

possible is that the net in-vivo effect is mediated by

molecular pathways not affected by dalteparin or not

related to tissue factor or that the dalteparin dose is

not high enough. We show that the in-vitro incubation

of patient’s sera with polyclonal antitissue factor antibody

inhibits circulating-tissue factor antigen and results in

reduction of cellular invasion properties of these sera.

This is a reproducible, dose-dependent, in-vitro effect,

which supports the hypothesis that the reduction of

circulating-tissue factor antigen in the serum of these

patients could be one of the mechanisms that could lead

to reduction of cellular invasion. We have recently

demonstrated that soluble tissue factor is capable of

interacting with the cell surface [23,24] and second that

this interaction and subsequent activation of different

sets of associated coagulation enzymes results in a diverse

set of signalling pathways, which result in proliferation,

apoptosis [24] and cell migration [25]. In the present

study, tissue factor antigen concentration was determined

as a measure of procoagulant activity and prothrombin

time was used to assess the functional activity of the

extrinsic coagulation pathway. Here, we show that tissue

factor, as found in patients’ sera, has a measurable che-

motactic effect on cancer-cell invasiveness that can be

modulated by the use of dalteparin. Previously, a signifi-

cant correlation (P< 0.0001) has been shown to exist in

the concentration of tissue factor antigen and tissue factor

activity in atherosclerotic plaques [26]. Furthermore,

procoagulant activity has been linked to the amount of

tissue factor antigen [27].

The fact that we have used uncharacteristically high

doses of dalteparin (weight adjusted) in a primary throm-

boprophylaxis sense in these patients raises the concern

that ‘excess’ free dalteparin could be responsible for the

differences shown between the groups by this assay. The

dalteparin-spiking experiment showed no statistically

significant trend effect at the concentration range studied

excluding a ‘pharmacological’ dose-related effect on

the assay. However, statistical power was limited and a

decrease was observed at 0.01 U/ml, in comparison to the


control (Fig. 4). We have recently shown in cancer cell

lines that there may be a direct effect of dalteparin on

tissue factor expression through mechanisms as yet not

clear (manuscript in preparation). We cannot exclude that

the reduction of cellular invasion in vivo is the result of

the effect of other molecules induced or reduced by the

in-vivo effect of dalteparin such as tissue factor pathway

inhibitor (TFPI), which is known to have direct inhibi-

tory effects on cancer-promoting mechanisms [28]. Nor

can we say from our work whether lower prophylactic

dalteparin doses or different heparin-derived molecules

would have similar effects.

Collectively, our results demonstrate for the first time that

the level of circulating tissue factor antigen in pancreatic

cancer patients, as compared with a contemporaneous

unbiased (randomized) control group, and as compared

to pretreatment baseline values, can be significantly con-

trolled through WAD treatment for 8 weeks. We also show

that WAD is strongly associated with the reduction in the

capacity of patient’s sera to stimulate cancer cell loco-

motion and invasion in vitro, a keystone phenotypic feature

of cancer progression and metastasis. We suggest that

circulating-tissue factor antigen and cellular-invasion-type

assays may find use as surrogate markers of the antimalig-

nant effect of dalteparin and other LMWHs and may

inform the appropriate dosing of these agents in the use

of LMWH for ‘oncological’ effects.

AcknowledgementEli Lilly provided funding for consumables but had no

further role in study design, collection or interpretation of

data, preparation or submission of the article.

References1 Khorana AA, Fine RL. Pancreatic cancer and thromboembolic disease.

Lancet Oncol 2004; 5:655–663.2 Sgouros J, Maraveyas A. Excess premature (3-month) mortality in advanced

pancreatic cancer could be related to fatal vascular thromboembolicevents. A hypothesis based on a systematic review of phase IIIchemotherapy studies in advanced pancreatic cancer. Acta Oncol 2008;47:337–346.

3 Rickles FR, Patierno S, Fernandez PM. Tissue factor, thrombin, and cancer.Chest 2003; 124:58S–68S.

4 Bromberg ME, Sundaram R, Homer RJ, Garen A, Konigsberg WH.Role of tissue factor in metastasis: functions of the cytoplasmaticand extracellular domains of the molecule. Thromb Haem 1999;82:88–92.

5 Koomagi R, Volm M. Tissue-factor expression in human nonsmall-cell lungcarcinoma measured by immunohistochemistry: correlation between tissuefactor and angiogenesis. Int J Cancer 1998; 79:19–22.

6 Ueno T, Toi M, Koike M, Nakamura S, Tominaga T. Tissue factor expressionin breast cancer tissues: its correlation with prognosis and plasmaconcentration. Br J Cancer 2000; 83:164–170.

7 Ueda C, Hirohata Y, Kihara Y, Nakamura H, Abe S, Akahane K, et al.Pancreatic cancer complicated by disseminated intravascular coagulationassociated with production of tissue factor. J Gastroenterol 2001;36:848–850.

8 Forster Y, Meye A, Albrecht S, Kotzsch M, Fussel S, Wirth MP,Schwenzer B. Tissue specific expression and serum levels of humantissue factor in patients with urological cancer. Cancer Lett 2003; 193:65–73.

9 Yu JL, May L, Lhotak V, Shahrzad S, Shirasawa S, Weitz JI, et al. Oncogenicevents regulate tissue factor expression in colorectal cancer cells:implications for tumor progression and angiogenesis. Blood 2005;105:1734–1741.


C


MBC 200692


10 Gerlach R, Scheuer T, Bohm M, Beck J, Woszczyk A, Raabe A, et al.Increased levels of plasma tissue factor pathway inhibitor in patients withglioblastoma and intracerebral metastases. Neurol Res 2003; 25:335–338.

11 Bromberg ME, Konigsberg WH, Madison JF, Pawashe A, Garen A. Tissuefactor promotes melanoma metastasis by a pathway independent of bloodcoagulation. Proc Natl Acad Sci U S A 1995; 92:8205–8209.

12 Kakkar AK, Chinswangwatanakul V, Lemoine NR, Tebbutt S, WilliamsonRC. Role of tissue factor expression on tumour cell invasion and growth ofexperimental pancreatic adenocarcinoma. Br J Surg 1999; 86:890–894.

13 Kakkar AK, Lemoine NR, Scully MF, Tebbutt S, Williamson RC. Tissuefactor expression correlates with histological grade in human pancreaticcancer. Br J Surg 1995; 82:1101–1104.

14 Khorana AA, Ahrendt SA, Ryan CK, Taubman MB, Hu YC, Ahrendt SA.Tissue factor expression, angiogenesis, and thrombosis in pancreaticcancer. Clin Cancer Res 2007; 13:2870–2875.

15 Han LY, Landen CN Jr, Kamat AA, Bender DP, Mueller P, Schmandt R,et al. Preoperative serum tissue factor levels are an independent prognosticfactor in patients with ovarian carcinoma. J Clin Oncol 2006; 24:755–761.

16 Hejna M, Raderer M, Zielinski CC. Inhibition of metastases byanticoagulants. J Natl Cancer Inst 1999; 91:22–36.

17 Smorenburg SM, Van Noorden CJF. The complex effects of heparins oncancer progression and metastasis in experimental studies. Pharmacol Rev2001; 53:93–105.

18 Fareed J, Hoppensteadt D, Cort S, Iqbal O, Bacher P, Fareed D, et al.Enoxaparin (E) and warfarin (W) differentially regulate tissue factor (TF),tissue factor pathway inhibitor (TFPI) and thrombin activatable fibrinolyticinhibitor (TAFI) in cancer patients with thrombosis. J Clin Oncol 2005;23:2057.

19 Maraveyas A, Holmes M, Lofts F, Garadi KK, Gardiner E, Sgouros J.Chemoanticoagulation versus chemotherapy in advanced pancreaticcancer (APC): results of the interim analysis of the FRAGEM trial. J ClinOncol 2007; 25:4583.


20 Burris HA, Moore MJ, Andersen J, Green MR, Rothenberg ML, MadianoMR, et al. Improvements in survival and clinical benefit with gemcitabine asfirst-line therapy for patients with advanced pancreas cancer: a randomizedtrial. J Clin Oncol 1997; 15:2403–2413.

21 Lee AYY, Levine MN, Baker RI, Bowden C, Kakkar AK, Prins M, et al. Low-molecular-weight heparin versus a coumarin for the prevention of recurrentvenous thromboembolism in patients with cancer. N Engl J Med 2003;349:146–153.

22 Maraveyas A, Li C, Greenman J, Ettelaie C. Serum tissue factor levels inpancreatic cancer patients treated with therapeutic dose dalteparin andgemcitabine, in a randomised phase IIb trial FRAGEM. ‘Genes & Cancer2004’; Warwick, UK.

23 Ettelaie C, Li C, Collier MEW, Pradier A, Frentzou GA, Wood CG, et al.Differential functions of tissue factor in the trans-activation of cellularsignalling pathways. Atherosclerosis 2007; 194:88–101.

24 Pradier A, Ettelaie C. The influence of exogenous tissue factor on theregulators of proliferation and apoptosis in endothelial cells. J Vasc Res2007; 45:19–32.

25 Sato Y, Asada Y, Marutsuka K, Hatakeyama K, Sumiyoshi A. Tissue factorinduces migration of cultured aortic smooth muscle cells. Thromb Haem1996; 75:389–392.

26 Ardissino D, Merlini PA, Ariens R, Coppola R, Bramucci E, Mannucci PM.Tissue-factor antigen and activity in human coronary atheroscleroticplaques. Lancet 1997; 349:769–771.

27 Marmur JD, Thiruvikraman SV, Fyfe BS, Guhu A, Sharma SK, Ambrose JA,et al. Identification of active tissue factor in human coronary atheroma.Circulation 1996; 94:1226–1232.

28 Hembrough TA, Swartz GM, Papathanassiu A, Vlasuk GP, Rote WE, GreenSJ, Pribluda VS. Tissue factor/factorVIIa inhibitors block angiogenesis andtumour growth through nonhemostatic mechanism. Cancer Res 2003;63:2997–3000.


as opposed to immunosuppressive in itsmechanism of action.

To summarize, Flowers et al report the re-sults of the largest prospective and randomizedstudy of ECP in corticosteroid refractory or de-pendent chronic GVHD. While the primary endpoint was not demonstrated, the study had nu-merous strengths, as described above. It remainsto be seen whether either earlier institution orlonger therapy with ECP improves the outcomesof patients with chronic GVHD.

Conflict-of-interest disclosure: The authordeclares no competing financial interests. ■

REFERENCES1. Couriel DR, Hosing C, Saliba R, et al. Extra-corporeal photochemotherapy for the treatment ofsteroid-resistant chronic GVHD. Blood. 2006;107:3074-3080.

2. Greinix HT, Pohlreich D, Maalouf J, et al. A single-center pilot validation study of a new chronic GVHD skinscoring system. Biol Blood Marrow Transplant. 2007;13:715-723.

3. Pavletic SZ, Martin P, Lee SJ, et al. Measuring thera-peutic response in chronic graft-versus-host disease: Na-tional Institutes of Health Consensus DevelopmentProject on Criteria for Clinical Trials in Chronic Graft-versus-Host Disease: IV. Response Criteria WorkingGroup report. Biol Blood Marrow Transplant. 2006;12:252-266.

● ● ● CLINICAL OBSERVATIONS

Comment on Ay et al, page 2703

Cancer and thrombosis is a stickybusiness----------------------------------------------------------------------------------------------------------------

Frederick R. Rickles THE GEORGE WASHINGTON UNIVERSITY

In this issue of Blood, Ay and colleagues demonstrate that elevated levels of solubleP-selectin are a risk factor for a first episode of venous thromboembolism in pa-tients with all types of cancer.

I t is a well-known fact that cancer cells of vari-ous types exhibit abnormal membrane surface

properties, often an exaggeration of normal func-tions, such as vesiculation (or blebbing), shed-ding of microparticles (MPs), and excessivephospholipid turnover, resulting in further am-

plification of the adhesive characteristics of thecells. Several lines of evidence have implicatedthese aberrant properties in the propensity ofcancer cells to induce hypercoagulability andpredispose patients with cancer to venousthromboembolism (VTE). Indeed, one of those

hyperactive adhesive mecha-nisms, selectin-ligand bind-ing, has been postulated to beresponsible for the associa-tion of Trousseau syndrome(migratory thrombophlebi-tis) with mucin-secretingadenocarcinomas1 (see fig-ure). P-selectin, a member ofthe selectin family ofcell-adhesion receptors, islocalized principally in themembranes of platelet�-granules andendothelial-cell Weibel-Palade bodies. In response tovarious agonists, P-selectin istranslocated to the cell sur-face, where it can function asa receptor and mediate celladhesion via binding to sev-eral ligands; the principal

ligand, or counterreceptor, is P-selectin gly-coprotein ligand-1 (PSGL-1), which is aheterodimeric mucin expressed on the cellsurface of the majority of leukocytes. Thisinteraction, particularly with soluble P-selectin (sP-selectin), may play a critical rolein the complex cell-cell adhesive processthat results in release of procoagulant-rich(ie, tissue factor [TF] and prothrombinase)MPs from leukocytes, endothelial cells,platelets, and cancer cells.2-4 Leukocyte MPs(and presumably MPs from other sources,eg, cancer cells) can further aggravate theprocoagulant state by inducing endothelial-cell TF expression.5 Elevated circulatinglevels of sP-selectin are clearly prothrom-botic in the experimental animal4,6 and havebeen implicated as a risk factor for thrombo-sis in a number of diseases, including recur-rence in patients with a first episode of un-provoked VTE—in the absence of cancer orother known risk factors,7 suggesting per-haps that patients with high levels of sP-selectin might benefit from longer durationof anticoagulation therapy.

Ay and colleagues report a similar abilityto predict a first episode of VTE among pa-tients enrolled in a large, prospective cohortstudy, the Vienna Cancer and ThrombosisStudy. Over a 6-month period of time afterthe initial diagnosis of cancer, or of progres-sion or recurrence after partial or completeremission, the cumulative probability ofsymptomatic VTE (documented by objec-tive criteria), was 11.9%, or 3-fold higher inpatients with an initial level of sP-selectingreater than the 75th percentile, comparedwith patients whose levels were below the75th percentile (3.7%; P � .002). Ideally,the investigators would have provided theresults of sequential sampling over time, inorder to increase the confidence in their ob-servations and, perhaps, the relationship ofsP-selectin levels to the natural history ofVTE. Similarly, a more quantitative, step-wise relationship between levels of sP-selectin and risk for VTE would have beenreassuring, as would a higher degree of asso-ciation between known high-risk tumors(eg, pancreatic or brain tumors, etc) andsP-selectin levels. Also missing from thisstudy was the ability to determine the pos-sible relationship of sP-selectin levels with otherknown markers for hypercoagulability, such as

Proposed model for L- and P-selectin–mediated, mucin-induced activation andaggregation of platelets. Yellow circle labeled with L indicates L-selectin ligandson mucins. Blue boxes labeled with P indicate P-selectin ligands, includingcarcinoma mucins and PSGL-1 expressed on leukocytes. L-sel indicatesL-selectin; P-sel, P-selectin. Reproduced from Wahrenbrock et al1 with permis-sion of the publishers.

2594 1 O C T O B E R 2 0 0 8 I V O L U M E 1 1 2 , N U M B E R 7 blood

fibrin d-dimer levels. Regardless, this study pro-vides the impetus to test the utility of using sP-selectin levels for stratification in trials of antico-agulant prophylaxis in cancer patients. Furtherin the future might come trials of inhibitors ofsP-selectin function8 in the prevention of VTEin patients with cancer, selected by virtue of highlevels of the circulating protein.


REFERENCES1. Wahrenbrock M, Borsig L, Le D, Varki N, Varki A.Selectin-mucin interactions as a probable molecular ex-planation for the association of Trousseau syndrome withmucinous adenocarcinomas. J Clin Invest. 2003;112:853-862.

2. Celi A, Pellegrini G, Lorenzet R, et al. P-selectin in-duces the expression of tissue factor on monocytes. ProcNat Acad Sci U S A. 1994;91:8767-8771.

3. Polgar J, Matuskova J, Wagner DD. The P-selectin,tissue factor, coagulation triad. J Thromb Haemost. 2005;3:1590-1596.

4. Hrachovinova I, Cambien B, Hafezi-Moghadam A, etal. Interaction of P-selectin and PSGL-1 generates micro-particles that correct hemostasis in a mouse model of hemo-philia A. Nat Med. 2003;9:1020-1025.

5. Mesri M, Altieri DC. Leukocyte microparticles stimu-late endothelial cell cytokine release and tissue factor induc-tion in a JNK1 signaling pathway. J Biol Chem. 1999;274:23111-23118.

6. Andre P, Hartwell D, Hrachovinaova I, Saffaripour S,Wagner DD. Pro-coagulant state resulting from high levelsof soluble P-selectin in blood. Proc Nat Acad Sci U S A.2000;97:13835-13840.

7. Kyrle PA, Hron G, Eichinger S, Wagner O. CirculatingP-selectin and the risk of recurrent venous thrombosis.Thromb Haemost. 2007;97:880-883.

8. Meier TR, Myers DD, Wrobleski SK, et al. Prophylac-tic p-selectin inhibition with PSI-421 promotes resolutionof venous thrombosis without anticoagulation. ThrombHaemost. 2008;99:343-351.

● ● ● GENE THERAPY

Comment on Shi et al, page 2713

Killing 2 birds with 1 stone----------------------------------------------------------------------------------------------------------------

Edward G. D. Tuddenham ROYAL FREE HOSPITAL

Bone marrow transplantation with stem cells transgenically modified to expressfactor VIII in platelets corrects the bleeding tendency in hemophilic mice with neu-tralizing anti–factor VIII antibodies. Or, “Why the blind mice of Milwaukee didn’tbleed to death after their tails were cut off.”

I t has been sufficiently clear for some yearsthat only 2 problems remain in the field of

inherited bleeding disorders. These prob-lems are easy to state but hard to solve. Thefirst problem is delivery of the clotting fac-tor in the right amount at the right time; thatis to say, a normal level all the time. All thatis required to correct the bleeding tendencyin single factor deficiency is to deliver a nor-mal level of factor, be it VIII, IX, Von Wille-brand, or one of the 7 other Roman numeral–designated enzymes or cofactors that takepart in fibrin formation. This might seemlike a simple proposition given that safe ef-fective preparations of nearly all those pro-teins are available. Sadly, most of theworld’s hemophiliacs have no access to fac-tor replacement and solving this depends onnew approaches that have a mainly politicaldimension. In economically advanced coun-tries, there is now no shortage of most of thefactors (factor V is the exception) but getting

patients to adhere to a challenging routine ofintravenous injections poses a separate set ofproblems with largely psychological/socialaspects.

Many, myself included, consider that allthese issues could be solved by gene therapy.Indeed, one of the earliest targets identifiedas being a sensible target for cure by DNAtransfer was hemophilia. Although mice anddogs have now been cured of hemophilia bygene transfer, safe effective procedures forcuring human hemophilia have yet to bedemonstrated.

The second outstanding problem in

clinical hemophilia care is that of inhibitory

antibodies arising after replacement

therapy. Patients who have developed such

antibodies become resistant to standard

treatment. They either undergo immune

tolerance induction therapy, which is ex-

tremely expensive and sometimes fails, or

are treated with bypassing agents that arealso ruinously costly and not always effec-tive. Patients with persistent inhibitors haveshortened life expectancy.

In this issue of Blood, Shi et al, in a paperwhose title says it all, now present proof of aprinciple that can solve the inhibitor problemand the delivery problem in one fell swoop forthe majority of problem cases, that is to say,those with hemophilia A and antibodies tofactor VIII (FVIII).

An ingenious transgenic mouse model isused in which a line of mice has been estab-lished where FVIII is ectopically expressedin megakaryocytes under the control of aplatelet membrane receptor promoter se-quence. Platelets formed by the megakaryo-cytes derived from hematopoietic stem cellsof such mice contain FVIII stored in alphagranules along with Von Willebrand factor,the natural carrier for FVIII. This FVIII isboth protected from plasma antibodies andreleased upon platelet activation at sites ofvascular injury where haemostatic plugs areforming. With this line of mice on hand, asecond line of mice that have had their factorVIII gene knocked out are stimulated toform high titer antibody to FVIII by infus-ing them with recombinant human B do-main deleted FVIII. Subsequently, thesemice that mimic human hemophilia A pa-tients with inhibitors are rescued by trans-planting them, after lethal irradiation withhematopoietic stem cells, from the mice thatmake platelets with releasable FVIII storedin platelet alpha granules. After successfulengraftment of this modified bone marrow,the mice are no longer apt to bleed to deathafter minor injury. One can already see theoutlines of a feasible clinical approach: Au-tologous hematopoietic stem cells from apatient with antibodies to FVIII are modi-fied ex vivo with a lentivirus-based FVIIIexpression cassette under control of theGPIIb alpha promoter and then reinfusedafter irradiation to a level sufficient to pro-duce partial engraftment. If the level of ex-pression is high enough, it might be worthconsidering this as a general approach tohemophilia gene therapy.


blood 1 O C T O B E R 2 0 0 8 I V O L U M E 1 1 2 , N U M B E R 7 2595

ISSN: 1524-4636 Copyright © 2009 American Heart Association. All rights reserved. Print ISSN: 1079-5642. Online

7272 Greenville Avenue, Dallas, TX 72514Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart Association.

DOI: 10.1161/ATVBAHA.108.182196 2009;29;316-320 Arterioscler. Thromb. Vasc. Biol.

Tarek Sousou and Alok A. Khorana New Insights Into Cancer-Associated Thrombosis

http://atvb.ahajournals.org/cgi/content/full/29/3/316located on the World Wide Web at:

The online version of this article, along with updated information and services, is

http://www.lww.com/reprintsReprints: Information about reprints can be found online at

[email protected]. E-mail:

Fax:Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters

http://atvb.ahajournals.org/subscriptions/Biology is online at Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular

by guest on April 9, 2009 atvb.ahajournals.orgDownloaded from

http://atvb.ahajournals.org/cgi/content/full/29/3/316

http://atvb.ahajournals.org/subscriptions/


http://www.lww.com/reprints

http://atvb.ahajournals.org

Venous Thromboembolism: Mechanisms, Treatment,and Public Awareness

New Insights Into Cancer-Associated ThrombosisTarek Sousou, Alok A. Khorana

Abstract—Venous thromboembolism (VTE) is an increasingly frequent complication of anticancer therapy. Theunderlying mechanisms are not completely understood, but are related in part to oncogene activation and tissue factor(TF) expression. Several risk factors have been identified including site and stage of cancer, patient comorbidities, andspecific therapeutic agents. Candidate biomarkers such as blood counts, TF, and P-selectin have recently been identified.A risk model predictive of chemotherapy-associated VTE has been validated. Thromboprophylaxis with low molecularweight heparin (LMWH), unfractionated heparin (UFH), or fondaparinux is recommended for hospitalized medical andsurgical cancer patients. Long-term anticoagulation with LMWH is safe and effective in reducing recurrent VTE incancer. The role of thromboprophylaxis in ambulatory cancer patients receiving chemotherapy is an area of activeinvestigation. (Arterioscler Thromb Vasc Biol. 2009;29:316-320.)

Key Words: thrombosis � risk factors � cancer

Cancer is a prothrombotic state, and cancer treatments areoften complicated by thromboembolism. Venous events

are the most common, presenting as either deep venousthrombosis (DVT) or pulmonary embolism (PE), togetherdescribed as venous thromboembolism (VTE). Indeed, cancerpatients account for as much as 20% of the total burden ofVTE.1 Arterial events, including stroke and myocardial in-farction, are also more prevalent in cancer patients.2

The prothrombotic state of cancer is driven by specificoncogenic events.3,4 Activation of the coagulation cascadeappears integrally linked to the processes of tumor growth,metastasis, and angiogenesis. Elegant preclinical studies haveshown, for instance, that defects in fibrinogen and plateletactivation can decrease metastatic potential.5–7 This has led toa renewed interest in studying the anticancer effects ofinterrupting the coagulation cascade.

Several other factors have contributed to an increasingawareness of the impact of VTE in cancer. The incidence ofVTE in cancer is on the rise.8 Novel anticancer drugs,particularly antiangiogenic agents, may be contributing to thisincrease.9,10 VTE is the second leading cause of death incancer patients11 and the most common cause of death in thepostoperative period.12 VTE in cancer is associated with a21% annual risk of recurrent VTE, a 12% annual risk ofbleeding complications, requirement for long-term anticoag-ulation, and interruption of chemotherapy.13,14

This brief review will focus on new insights into thepathophysiology of cancer-associated thrombosis, risk factorsand candidate predictive biomarkers for VTE, as well asappropriate strategies for the prevention and treatment ofVTE in cancer.

Mechanisms of ThrombosisThe pathophysiology of cancer-associated thrombosis is notentirely understood. Rather than one unifying mechanism, theetiology is likely multifactorial with different factors assum-ing lesser or greater degrees of importance depending on theclinical setting.

Much of the research in this area has focused on theintrinsic properties of tumor cells that lead to a prothromboticstate. The role of tissue factor (TF) has gathered the mostattention. TF, a transmembrane glycoprotein, is the primephysiological initiator of coagulation and is expressed in avariety of human cancers, induced by activation of oncogenesor inactivation of tumor suppressor genes.4 Overexpression ofTF in tumor cells or elevated TF levels in association withmicroparticles in the systemic circulation may contribute tosystemic hypercoagulability.15–19 Much of this work hasfocused on selected cancers, particularly pancreas, andwhether TF is equally important in other cancers remains tobe seen. Activation of the MET oncogene has been shown ina mouse model of hepatocarcinogenesis to result in a throm-bohemorragic state mediated by upregulation of plasminogenactivator inhibitor type 1 (PAI-1) and cyclooxygenase-2 geneactivity.3 However, the applicability of this model to othercancers and to the clinical setting is not known. Carcinomamucins, glycosylated molecules that act as ligands for theselectin family, may also play a role in thrombosis.20 Finally,the role of tumor hypoxia and inflammatory cytokines hasalso been speculated to contribute to the prothrombotic statein cancer, but firm experimental evidence is awaited.21–23

Extrinsic factors are also important but, unfortunately, arenot accounted for by the various experimental models dis-

Received January 7, 2009; revision accepted January 15, 2009.From the James P. Wilmot Cancer Center and the Department of Medicine (T.S., A.A.K.), University of Rochester, NY.Correspondence to Alok A. Khorana, MD, 601 Elmwood Ave, Box 704, Rochester, NY 14642. E-mail [email protected]© 2009 American Heart Association, Inc.

Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org DOI: 10.1161/ATVBAHA.108.182196

316 by guest on April 9, 2009 atvb.ahajournals.orgDownloaded from


cussed above. Chemotherapy can result in activation ofhemostasis within a few hours of administration.24 Thisoccurs via a variety of mechanisms, including induction of TFin tumor cells25 and monocytes,26 downregulation of antico-agulant proteins such as protein C and S,27,28 damage tovascular endothelium,29 and platelet activation.30 Antiangio-genic agents also contribute to thrombosis, perhaps throughendothelial cell and platelet activation.31

Risk FactorsMultiple recent studies have evaluated risk factors for VTE incancer patients in the general population, in hospitalizedpatients, and in registries of outpatients receiving chemother-apy. Overall, these risk factors for VTE can be categorizedaccording to patient characteristics and comorbidities,malignancy-related characteristics, and therapeutic interven-

tions for cancer (Table 1). Comorbid conditions such asinfection, obesity, anemia, and pulmonary and renal diseaseparticularly add to the risk of VTE.8 The primary site ofcancer is an important risk factor, with highest rates observedin patients with brain, pancreas, gastric, kidney, ovary, andlung cancers and hematologic malignancies, particularly lym-phomas.8,32–34 In a population-based study, the risk of VTEwas greatest in the first 3 months after the diagnosis of cancer(OR 53.5, 95% CI 8.6 to 334.3).32 Hospitalization increasesthe risk of VTE in cancer patients.35 Major surgery has longbeen known to be associated with an increased risk of VTE;more recent data indicate that this risk extends for a pro-longed period after the procedure, with 40% of all VTEevents in one registry occurring later than 21 days fromsurgery.12 Chemotherapy is associated with a 2- to 6-foldincreased risk of VTE as compared to the general popula-tion.36,37 VTE is also associated with the use of central venouscatheters.38 Erythropoiesis-stimulating agents (ESAs) havebeen found to increase the risk of VTE; unfortunately, redblood cell transfusions may have a similar association.39,40

Novel therapeutics such as the antiangiogenic class of agentsare also associated with VTE. Thalidomide and lenalidomide-containing regimens increase the risk several-fold in patientswith myeloma.41 Regimens containing bevacizumab, a mono-clonal antibody directed against the proangiogenic vascularendothelial growth factor, are associated with high rates ofboth arterial and venous events.10,42

Candidate BiomarkersResearch conducted primarily in cancer outpatients has re-sulted in the identification of novel candidate biomarkers thatmay be predictive of cancer-associated VTE. In an observa-tional study, VTE occurred in 4% of patients with a preche-motherapy platelet count �350 000/mm3 as compared to1.25% for those with counts �200 000/mm3 34. An elevatedprechemotherapy leukocyte count (defined as �11 000/mm3)was also significantly and independently associated with anincreased risk of VTE.43 High grades of TF expression intumor cells and elevated levels of circulating TF have beenassociated with the risk of VTE in pancreatic and ovariancancers.18,19 In a prospective cohort study, elevated levels ofsoluble P-selectin levels (�53.1 ng/mL, representing the 75th

percentile) were predictive of VTE (HR 2.6, CI 1.4 to 4.9).44

Markers of hemostatic activation, particularly D-dimer, havebeen observed to be elevated in cancer patients and predic-tive of recurrent VTE in cancer patients.45 In an observa-tional study of 507 cancer patients, an elevated C-reactiveprotein (�400 mg/dL) was associated in multivariateanalysis with VTE.35

A Predictive Risk ModelVTE in cancer is a multifactorial disease that involves variousrisk factors, as is evident from the preceding discussion. Arisk model for chemotherapy-associated VTE has recentlybeen published and is based on scores assigned to 5 predictivevariables identified in a development cohort of 2701 ambu-latory cancer patients initiating chemotherapy (Table 2).43

The score was then validated in an independent cohort of1365 patients from the same study. Rates of VTE in the

Table 1. Risk Factors and Candidate Biomarkers for VTE

Patient-related factors

Older age

Female gender

Race

Higher in blacks

Lower in Asians/Pacific Islanders

Comorbidities

Infection, renal disease, pulmonary disease, obesity

Inherited prothrombotic mutations

Prior history of VTE

Cancer-related factors

Primary site of cancer

Brain, pancreas, kidney, stomach, lung, gynecologic, lymphoma, myeloma

Advanced stage of cancer

Initial period after diagnosis of cancer

Treatment-related factors

Major surgery

Hospitalization

Cancer therapy

Chemotherapy

Hormonal therapy

Antiangiogenic agents

Thalidomide, lenalidomide, bevacizumab

Erythropoiesis-stimulating agents

Transfusions

Central venous catheter

Candidate biomarkers

Prechemotherapy platelet count �350 000/mm3

Prechemotherapy leukocyte count �11 000/mm3

Tissue factor (TF)

High grade of TF expression by tumor cells

Elevated TF plasma levels

Soluble P-selectin

D-dimer

C-reactive protein

Sousou and Khorana New Insights Into Cancer-Associated Thrombosis 317



development and validation cohorts, respectively, were 0.8%and 0.3% in the low-risk category (score�0), 1.8% and 2% inthe intermediate-risk category (score�1 to 2), and 7.1% and6.7% in the high-risk category (score �3) over a medianperiod of 2.5 months.43 Rates of VTE in this high-risksubgroup are comparable to hospitalized patients for whomprophylaxis is safe and effective. The National Heart, Lung,and Blood Institute has recently funded a study of outpatientprophylaxis in cancer patients identified as high-risk based onthis model.

Prevention of VTEHospitalized Medical Cancer PatientsThree large randomized controlled trials studied either enox-aparin, dalteparin, or fondaparinux for thromboprophylaxis inacutely ill hospitalized medical patients and reported relativerisk reductions in VTE ranging from 45% to 63% withanticoagulation.46 – 48 Unfortunately, none of these werecancer-specific study populations and cancer patients formedonly a minority (ranging from 5% to 15%) of study patients.Furthermore, bleeding complications, a major concern withanticoagulation in cancer, were not separately reported forcancer patients. Unfractionated heparin (UFH) is an accept-able alternative to low-molecular-weight heparins (LMWHs)as thromboprophylaxis in this setting.49 Despite the lack ofcancer-specific data, the American Society of Clinical Oncol-ogy (ASCO) guidelines recommend that hospitalized cancerpatients should be considered for VTE prophylaxis withanticoagulants in the absence of bleeding or other contrain-dications to anticoagulation.50

Surgical Cancer PatientsMultiple clinical trials have established the safety and effi-cacy of thromboprophylaxis in the perioperative period forcancer patients undergoing major surgical procedures. Morerecently, two studies (including one cancer-specific study)have suggested that extending the duration of postoperativeLMWH prophylaxis for 2 to 4 weeks after hospital dischargereduces the incidence of late venographic VTE.51,52 Both theASCO and the National Comprehensive Cancer Network(NCCN) guidelines support either UFH, LMWH, or fondapa-rinux in the surgical cancer patient for VTE prophylaxis andsuggest using prolonged prophylaxis in high-risk patients.50,53

Ambulatory Cancer PatientsThe treatment of cancer has now primarily moved to theoutpatient setting. Several clinical trials have been conducted

to evaluate the benefit of thromboprophylaxis for canceroutpatients, with varying inclusion criteria and contradictoryresults.54–56

Data from the most recent and largest study found fewerthromboembolic events (venous and arterial combined) oc-curred in the nadroparin arm than in the placebo arm (2.0%versus 3.9% P�0.033, NNT�53.8).57 Current guidelines donot recommend prophylaxis for cancer outpatients, althoughdata from more targeted approaches such as the risk modeldescribed above are awaited. One exception to this is patientswith multiple myeloma receiving thalidomide/lenalidomide-based regimens for whom prophylaxis is recommended witheither LMWH or warfarin based on data from nonrandomizedstudies.50

Treatment of VTE in Cancer PatientsWarfarin has previously been the standard for chronic anti-coagulation but in the cancer population is associated withincreased rates of bleeding, recurrent VTE, and dietary anddrug-related interactions. In the CLOT trial, 672 cancerpatients with documented VTE were randomized to receiveeither dalteparin or dalteparin followed by a vitamin Kantagonist (control group) for a total of 6 months.58 RecurrentVTE at 6 months occurred in 9% of patients in the dalteparingroup compared to 17% in the control group. These findingsare consistent with data from multiple other smaller studiesand a meta-analysis.59 These data have established LMWHfor at least 3 to 6 months as the standard of care for treatmentof VTE in cancer, as recommended by the ASCO, NCCN,and other guidelines.50,53 The optimal duration of anticoagu-lation in cancer patients with VTE remains unknown. Giventhat cancer patients remain at risk for VTE, it is recom-mended that patients with active cancer be considered forindefinite anticoagulation. It is important to note in thiscontext preclinical and clinical data (albeit conflicting) sug-gesting that anticoagulants, particularly LMWHs, may impactcancer processes such as angiogenesis and tumor cell adhe-sion and, therefore, clinical outcomes.60–62

ConclusionsOngoing areas of investigation include understanding thepathophysiology of cancer-associated thrombosis in waysthat can impact tumor biology, targeted prophylaxis in canceroutpatients, and studying the impact of anticoagulation onsurvival in cancer. Although many new beginnings have beenmade in the field of cancer-associated thrombosis in the pastdecade, much learning awaits.

Sources of FundingDr Khorana is supported by grants from the National Cancer InstituteK23 CA120587, the National Heart, Lung, and Blood Institute1R01HL095109-01, and the V Foundation.

DisclosuresA.K. has received research funding from sanofi-aventis. Eisai, andBristal Myers Squibb as well as speakers fees and consulting feesfrom sanofi-aventis and Eisai.

Table 2. A Validated Predictive Model forChemotherapy-Associated VTE43

Patient Characteristic Risk Score

Site of cancer

Very high risk (stomach, pancreas) 2

High risk (lung, lymphoma, gynecologic, bladder, testicular) 1

Prechemotherapy platelet count �350 000/mm3 1

Hemoglobin �10g/dL or use of red cell growth factors 1

Pre-chemotherapy leukocyte count �11 000/mm3 1

Body mass index �35 kg/m2 1

318 Arterioscler Thromb Vasc Biol March 2009



References1. Heit JA, O’Fallon WM, Petterson TM, Lohse CM, Silverstein MD, Mohr

DN, Melton LJ III. Relative impact of risk factors for deep veinthrombosis and pulmonary embolism: a population-based study. ArchIntern Med. 2002;162:1245–1248.

2. Khorana AA, Francis CW, Culakova E, Fisher RI, Kuderer NM, LymanGH. Thromboembolism in hospitalized neutropenic cancer patients.J Clin Oncol. 2006;24:484–490.

3. Boccaccio C, Sabatino G, Medico E, Girolami F, Follenzi A, Reato G,Sottile A, Naldini L, Comoglio PM. The MET oncogene drives a geneticprogramme linking cancer to haemostasis. Nature. 2005;434:396–400.

4. Yu JL, May L, Lhotak V, Shahrzad S, Shirasawa S, Weitz JI, CoomberBL, Mackman N, Rak JW. Oncogenic events regulate tissue factorexpression in colorectal cancer cells: implications for tumor progressionand angiogenesis. Blood. 2005;105:1734–1741.

5. Palumbo JS, Talmage KE, Massari JV, La Jeunesse CM, Flick MJ,Kombrinck KW, Jirouskova M, Degen JL. Platelets and fibrin(ogen)increase metastatic potential by impeding natural killer cell-mediatedelimination of tumor cells. Blood. 2005;105:178–185.

6. Palumbo JS, Potter JM, Kaplan LS, Talmage K, Jackson DG, Degen JL.Spontaneous hematogenous and lymphatic metastasis, but not primarytumor growth or angiogenesis, is diminished in fibrinogen-deficient mice.Cancer Res. 2002;62:6966–6972.

7. Camerer E, Qazi AA, Duong DN, Cornelissen I, Advincula R, CoughlinSR. Platelets, protease-activated receptors, and fibrinogen in hema-togenous metastasis. Blood. 2004;104:397–401.

8. Khorana AA, Francis CW, Culakova E, Kuderer NM, Lyman GH. Fre-quency, risk factors, and trends for venous thromboembolism amonghospitalized cancer patients. Cancer. 2007;110:2339–2346.

9. Zangari M, Barlogie B, Anaissie E, Saghafifar F, Eddlemon P, JacobsonJ, Lee CK, Thertulien R, Talamo G, Thomas T, Van Rhee F, Fassas A,Fink L, Tricot G. Deep vein thrombosis in patients with multiplemyeloma treated with thalidomide and chemotherapy: effects of prophy-lactic and therapeutic anticoagulation. Br J Haematol. 2004;126:715–721.

10. Nalluri SR, Chu D, Keresztes R, Zhu X, Wu S. Risk of venous throm-boembolism with the angiogenesis inhibitor bevacizumab in cancerpatients: a meta-analysis. JAMA. 2008;300:2277–2285.

11. Khorana AA, Francis CW, Culakova E, Kuderer NM, Lyman GH.Thromboembolism is a leading cause of death in cancer patients receivingoutpatient chemotherapy. J Thromb Haemost. 2007;5:632–634.

12. Agnelli G, Bolis G, Capussotti L, Scarpa RM, Tonelli F, Bonizzoni E,Moia M, Parazzini F, Rossi R, Sonaglia F, Valarani B, Bianchini C,Gussoni G. A clinical outcome-based prospective study on venous throm-boembolism after cancer surgery: the @RISTOS project. Ann Surg. 2006;243:89–95.

13. Prandoni P, Lensing AW, Piccioli A, Bernardi E, Simioni P, Girolami B,Marchiori A, Sabbion P, Prins MH, Noventa F, Girolami A. Recurrentvenous thromboembolism and bleeding complications during antico-agulant treatment in patients with cancer and venous thrombosis. Blood.2002;100:3484–3488.

14. Elting LS, Escalante CP, Cooksley C, Avritscher EB, Kurtin D, HamblinL, Khosla SG, Rivera E. Outcomes and cost of deep venous thrombosisamong patients with cancer. Arch Intern Med. 2004;164:1653–1661.

15. Dvorak HF, Quay SC, Orenstein NS, Dvorak AM, Hahn P, Bitzer AM,Carvalho AC. Tumor shedding and coagulation. Science. 1981;212:923–924.

16. Khorana AA, Ahrendt SA, Ryan CK, Francis CW, Hruban RH, Hu YC,Hostetter G, Harvey J, Taubman MB. Tissue factor expression, angio-genesis, and thrombosis in pancreatic cancer. Clin Cancer Res. 2007;13:2870–2875.

17. Tesselaar ME, Romijn FP, Van Der Linden IK, Prins FA, Bertina RM,Osanto S. Microparticle-associated tissue factor activity: a link betweencancer and thrombosis? J Thromb Haemost. 2007;5:520–527.

18. Uno K, Homma S, Satoh T, Nakanishi K, Abe D, Matsumoto K, Oki A,Tsunoda H, Yamaguchi I, Nagasawa T, Yoshikawa H, Aonuma K. Tissuefactor expression as a possible determinant of thromboembolism in ovar-ian cancer. Br J Cancer. 2007;96:290–295.

19. Khorana AA, Francis CW, Menzies KE, Wang JG, Hyrien O, HathcockJ, Mackman N, Taubman MB. Plasma tissue factor may be predictive ofvenous thromboembolism in pancreatic cancer. J Thromb Haemost. 2008;6:1983–1985.

20. Wahrenbrock M, Borsig L, Le D, Varki N, Varki A. Selectin-mucininteractions as a probable molecular explanation for the association ofTrousseau syndrome with mucinous adenocarcinomas. J Clin Invest.2003;112:853–862.

21. Varki A. Trousseau’s syndrome: multiple definitions and multiple mech-anisms. Blood. 2007;110:1723.

22. Denko NC, Giaccia AJ. Tumor hypoxia, the physiological link betweenTrousseau’s syndrome (carcinoma-induced coagulopathy) and metastasis.Cancer Res. 2001;61:795–798.

23. Rickles FR, Falanga A. Molecular basis for the relationship betweenthrombosis and cancer. Thromb Res. 2001;102:V215–V224.

24. Weitz IC, Israel VK, Waisman JR, Presant CA, Rochanda L, LiebmanHA. Chemotherapy-induced activation of hemostasis: effect of a lowmolecular weight heparin (dalteparin sodium) on plasma markers ofhemostatic activation. Thromb Haemost. 2002;88:213–220.

25. Paredes N, Xu L, Berry LR, Chan AK. The effects of chemotherapeuticagents on the regulation of thrombin on cell surfaces. Br J Haematol.2003;120:315–324.

26. Walsh J, Wheeler HR, Geczy CL. Modulation of tissue factor on humanmonocytes by cisplatin and adriamycin. Br J Haematol. 1992;81:480–488.

27. Rogers JS II, Murgo AJ, Fontana JA, Raich PC. Chemotherapy for breastcancer decreases plasma protein C and protein S. J Clin Oncol. 1988;6:276–281.

28. Woodley-Cook J, Shin LY, Swystun L, Caruso S, Beaudin S, Liaw PC.Effects of the chemotherapeutic agent doxorubicin on the protein Canticoagulant pathway. Mol Cancer Ther. 2006;5:3303–3311.

29. Cwikiel M, Eskilsson J, Albertsson M, Stavenow L. The influence of5-fluorouracil and methotrexate on vascular endothelium. An experi-mental study using endothelial cells in the culture. Ann Oncol. 1996;7:731–737.

30. Togna GI, Togna AR, Franconi M, Caprino L. Cisplatin triggers plateletactivation. Thromb Res. 2000;99:503–509.

31. Kuenen BC, Levi M, Meijers JC, Kakkar AK, van Hinsbergh VW,Kostense PJ, Pinedo HM, Hoekman K. Analysis of coagulation cascadeand endothelial cell activation during inhibition of vascular endothelialgrowth factor/vascular endothelial growth factor receptor pathway incancer patients. Arterioscler Thromb Vasc Biol. 2002;22:1500–1505.

32. Blom JW, Doggen CJ, Osanto S, Rosendaal FR. Malignancies, pro-thrombotic mutations, and the risk of venous thrombosis. JAMA. 2005;293:715–722.

33. Sallah S, Wan JY, Nguyen NP. Venous thrombosis in patients with solidtumors: determination of frequency and characteristics. Thromb Haemost.2002;87:575–579.

34. Khorana AA, Francis CW, Culakova E, Lyman GH. Risk factors forchemotherapy-associated venous thromboembolism in a prospectiveobservational study. Cancer. 2005;104:2822–2829.

35. Kroger K, Weiland D, Ose C, Neumann N, Weiss S, Hirsch C, UrbanskiK, Seeber S, Scheulen ME. Risk factors for venous thromboembolicevents in cancer patients. Ann Oncol. 2006;17:297–303.

36. Heit JA, Silverstein MD, Mohr DN, Petterson TM, O’Fallon WM, MeltonLJ III. Risk factors for deep vein thrombosis and pulmonary embolism: apopulation-based case-control study. Arch Intern Med. 2000;160:809–815.

37. Blom JW, Vanderschoot JP, Oostindier MJ, Osanto S, van der Meer FJ,Rosendaal FR. Incidence of venous thrombosis in a large cohort of 66,329cancer patients: results of a record linkage study. J Thromb Haemost.2006;4:529–535.

38. Lee AY, Levine MN, Butler G, Webb C, Costantini L, Gu C, Julian JA.Incidence, risk factors, and outcomes of catheter-related thrombosis inadult patients with cancer. J Clin Oncol. 2006;24:1404–1408.

39. Bohlius J, Wilson J, Seidenfeld J, Piper M, Schwarzer G, Sandercock J,Trelle S, Weingart O, Bayliss S, Djulbegovic B, Bennett CL, Langen-siepen S, Hyde C, Engert A. Recombinant human erythropoietins andcancer patients: updated meta-analysis of 57 studies including 9353patients. J Natl Cancer Inst. 2006;98:708–714.

40. Khorana AA, Francis CW, Blumberg N, Culakova E, Refaai MA, LymanGH. Blood transfusions, thrombosis, and mortality in hospitalizedpatients with cancer. Arch Intern Med. 2008;168:2377–2381.

41. Zangari M, Anaissie E, Barlogie B, Badros A, Desikan R, Gopal AV,Morris C, Toor A, Siegel E, Fink L, Tricot G. Increased risk of deep-veinthrombosis in patients with multiple myeloma receiving thalidomide andchemotherapy. Blood. 2001;98:1614–1615.

42. Scappaticci FA, Skillings JR, Holden SN, Gerber HP, Miller K, Kab-binavar F, Bergsland E, Ngai J, Holmgren E, Wang J, Hurwitz H. Arterialthromboembolic events in patients with metastatic carcinoma treated withchemotherapy and bevacizumab. J Natl Cancer Inst. 2007;99:1232–1239.

Sousou and Khorana New Insights Into Cancer-Associated Thrombosis 319



43. Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Devel-opment and validation of a predictive model for chemotherapy-associatedthrombosis [see comment]. Blood. 2008;111:4902–4907.

44. Ay C, Simanek R, Vormittag R, Dunkler D, Alguel G, Koder S, KornekG, Marosi C, Wagner O, Zielinski C, Pabinger I. High plasma levels ofsoluble P-selectin are predictive of venous thromboembolism in cancerpatients: results from the Vienna Cancer and Thrombosis Study (CATS).[see comment]. Blood. 2008;112:2703–2708.

45. Cosmi B, Legnani C, Cini M, Guazzaloca G, Palareti G. The role ofD-dimer and residual venous obstruction in recurrence of venousthromboembolism after anticoagulation withdrawal in cancer patients.Haematologica. 2005;90:713–715.

46. Samama MM, Cohen AT, Darmon J-Y, Desjardins L, Eldor A, Janbon C,Leizorovicz A, Nguyen H, Olsson CG, Turpie A, Graham W, Nadine.A Comparison of Enoxaparin with Placebo for the Prevention of VenousThromboembolism in Acutely ILL Medical Patients. The N Engl J Med.1999;341:793–800.

47. Leizorovicz A, Cohen AT, Turpie AG, Olsson CG, Vaitkus PT,Goldhaber SZ. Randomized, placebo-controlled trial of dalteparin for theprevention of venous thromboembolism in acutely ill medical patients.Circulation. 2004;110:874–879.

48. Cohen AT, Davidson BL, Gallus AS, Lassen MR, Prins MH, TomkowskiW, Turpie AG, Egberts JF, Lensing AW. Efficacy and safety offondaparinux for the prevention of venous thromboembolism in olderacute medical patients: randomised placebo controlled trial. BMJ. 2006;332:325–329.

49. Mismetti P, Laporte-Simitsidis S, Tardy B, Cucherat M, Buchmuller A,Juillard-Delsart D, Decousus H. Prevention of venous thromboembolismin internal medicine with unfractionated or low-molecular-weightheparins: a meta-analysis of randomised clinical trials. Thromb Haemost.2000;83:14–19.

50. Lyman GH, Khorana AA, Falanga A, Clarke-Pearson D, Flowers C,Jahanzeb M, Kakkar A, Kuderer NM, Levine MN, Liebman H, Men-delson D, Raskob G, Somerfield MR, Thodiyil P, Trent D, Francis CW.American Society of Clinical Oncology Guideline: Recommendations forvenous thromboembolism prophylaxis and treatment in patients withcancer. J Clin Oncol. 2007;25:5490–5505.

51. Bergqvist D, Agnelli G, Cohen AT, Eldor A, Nilsson PE, Le Moigne-Amrani A, Dietrich-Neto F. Duration of prophylaxis against venousthromboembolism with enoxaparin after surgery for cancer. N EnglJ Med. 2002;346:975–980.

52. Rasmussen MS, Jorgensen LN, Wille-Jorgensen P, Nielsen JD, Horn A,Mohn AC, Somod L, Olsen B. Prolonged prophylaxis with dalteparin to

prevent late thromboembolic complications in patients undergoing majorabdominal surgery: a multicenter randomized open-label study. J ThrombHaemost. 2006;4:2384–2390.

53. Wagman LD, Baird MF, Bennett CL, Bockenstedt PL, Cataland SR,Fanikos J, Fogarty PF, Goldhaber SZ, Grover TS, Haire W, Hassoun H,Jahanzeb M, Leung LL, Linenberger ML, Millenson MM, Ortel TL,Salem R, Smith JL, Streiff MB. Venous thromboembolic disease. Clinicalpractice guidelines in oncology. J Natl Compr Canc Netw. 2006;4:838–869.

54. Levine M, Hirsh J, Gent M, Arnold A, Warr D, Falanga A, Samosh M,Bramwell V, Pritchard KI, Stewart D, et al. Double-blind randomised trialof a very-low-dose warfarin for prevention of thromboembolism in stageIV breast cancer. Lancet. 1994;343:886–889.

55. Haas S, Kakkar AK, B K-M. Prevention of venous thromboembolismwith low molecular weight heparin in patients with metastatic breast orlung cancer-results of the TOPIC studies. ISTH. 2005.

56. Perry JR, Rogers L, J L. A phase III randomized placebo-controlled trialof thromboprophylaxis using dalteparin LMWH in patients with newlydiagnosed malignant glioma. J Clin Oncol. 2007.

57. Agnelli G, Gussoni G, Bianchini C, Verso M, Tonato M. A randomizeddouble-blind placebo-controlled study on nadroparin for prophylaxis ofthromboembolic events in cancer patients receiving chemotherapy: ThePROTECHT Study. American Society of Hematology Annual Meeting.San Francisco, California: Blood. 2008.

58. Lee AY, Levine MN, Baker RI, Bowden C, Kakkar AK, Prins M, RicklesFR, Julian JA, Haley S, Kovacs MJ, Gent M. Low-molecular-weightheparin versus a coumarin for the prevention of recurrent venous throm-boembolism in patients with cancer. N Engl J Med. 2003;349:146–153.

59. Akl EA, Barba M, Rohilla S, Terrenato I, Sperati F, Muti P, SchunemannHJ. Low-Molecular-Weight Heparins are superior to Vitamin K Antag-onists for the long term treatment of venous thromboembolism in patientswith cancer: a Cochrane systematic review. J Exp Clin Cancer Res.2008;27:21.

60. Khorana AA, Sahni A, Altland OD, Francis CW. Heparin inhibition ofendothelial cell proliferation and organization is dependent on molecularweight. Arterioscler Thromb Vasc Biol. 2003;23:2110–2115.

61. Kuderer NM, Khorana AA, Lyman GH, Francis CW. A meta-analysisand systematic review of the efficacy and safety of anticoagulants ascancer treatment: impact on survival and bleeding complications. Cancer.2007;110:1149–1161.

62. Fritzsche J, Simonis D, Bendas G. Melanoma cell adhesion can beblocked by heparin in vitro: Suggestion of VLA-4 as a novel target forantimetastatic approaches. Thromb Haemost. 2008;100:1166–1175.

320 Arterioscler Thromb Vasc Biol March 2009



www.myGCRonline.orgNovember/December 2008 267

REVIEW ARTICLE

Thromboembolism in Gastrointestinal CancersEric D. Tetzlaff, Jonathan D. Cheng, Jaffer A. Ajani

ABSTRACT

The link between thromboembolism and cancer has been recognized forover 100 years. Venous thromboembolism (VTE) is associated with consider-able morbidity in patients with cancer, with emerging research also indi-cating a detrimental effect on survival. Investigations aimed at improvingoutcomes for patients with cancer have focused on the role of low molec-ular weight heparin in primary and secondary prevention of VTE and inimproving patient survival. Important fundamental questions remain un-answered, however, and a significant line of research needs to be dedi-cated to investigating VTE in GI cancers. The effect of VTE on survival needsto be clarified, as does the role of anticoagulation in this patient population.Opportunities for additional research include investigating methods toidentify patients at risk of developing VTE and developing new strategiesand therapeutic interventions to reduce the morbidity and mortality associ-ated with VTE. This review focuses on the current understanding of VTErelated to gastrointestinal cancers and directions of interest in researchspecific to GI cancers and VTE.

Gastrointest Cancer Res 2:267–272. ©2008 by International Society of Gastrointestinal Oncology

E.D. Tetzlaff, MHS, PA-C; J.D. Cheng, MD:Department of Medical Oncology,

Fox Chase Cancer Center,Philadelphia, Pennsylvania

J.A. Ajani, MD: Department of GastrointestinalMedical Oncology, University of Texas

M. D. Anderson Cancer Center,Houston, Texas

Armand Trousseau was the first todescribe the association between

cancer and thrombosis when he reportedcases of migratory thrombophlebitis inpatients with cancer in 1867.1 Over 125years later, the association between cancerand thrombosis has been well estab-lished.2–4 The prothrombotic state gener-ated by malignancy is multifactorial.Several molecular and cellular etiologieshave been proposed, with some of themost intriguing hypotheses includingexpression of tissue factor by tumor cells,factor X-activating cysteine protease, andprocoagulant microparticles derived fromplatelets, endothelial cells, and leukocytes.5,6

Research on thromboembolism in pa-tients with cancer has focused on improvingthe primary and secondary prevention ofvenous thromboembolism (VTE) andassessing the detrimental effect of VTE onsurvival and quality of life. This reviewsummarizes the current understanding ofVTE in gastrointestinal (GI) cancers and im-plications for future research in this area.

INCIDENCEGastrointestinal (GI) cancers are associ-ated with a high incidence of thromboem-bolic events. In a review of over 1 million

Medicare patients hospitalized with adiagnosis of malignancy, GI cancers (pan-creas, stomach, liver, colon, rectum) wereamong the top 10 cancers with the highestrate of deep vein thrombosis (DVT) or pul-monary embolism (PE) out of 18 cancersreported (Table 1).2 In addition, the risk ofcancer within 1 year after idiopathic DVT orPE is elevated for pancreatic cancer, gas-tric cancer, esophageal cancer, colorectalcancer, and primary liver cancer.4,7 Suchretrospective data do not, however, providesufficient information on the effects of treat-ment, disease stage, and use of anticoagu-lation for primary prophylaxis on throm-boembolism risk. In a prospective series of2,482 patients treated in clinical trials atMemorial Sloan-Kettering Cancer Centerduring a 2-year period, Asmis et al found arate of VTE of 8.6% among GI cancerpatients, compared with 3.3% in patientswith non-GI cancers (P < .0001).8

The reported incidence of VTE varieswidely based on the location of the primaryGI tumor, disease stage, and other factors.Although there appears to be increasedawareness of the relationship betweenthrombosis and cancer, the reporting ofVTE in prospective series and clinical trialsremains inconsistent.

Advanced DiseaseAlthough the general perception is thatVTE is common in the palliative chemo-therapy setting, rates have not been welldefined in prospective clinical trials. Thereported overall rate of VTE has been ashigh as 71% in patients with advanced pan-creatic cancer treated with chemotherapy,9

although rates are infrequently reported inphase III trials.10–14 In gastric cancer, thereported rate of VTE ranges from 5.3% to25.5% in phase II clinical trials, with fewphase III data being available.15–19 Reportedrates for VTE during palliative therapy forother GI cancers include colorectal cancer,0.4% to 19.4%; hepatocellular carcinoma,1.4% to 4%; and bile duct and gallbladdercancer, 2% to 16.7%.20–24

It is likely that the true rate of VTE is inthe range of 10% to 20% for patients withGI cancers in advanced stages, with thehighest rates in cancers of the pancreas,stomach, and colon. In the future, withmore advanced imaging technology, it islikely that reported rates of VTE will

Address correspondence to: Eric Tetzlaff, MHS, PA-C, Department of Medical Oncology, Fox ChaseCancer Center, 333 Cottman Avenue, Philadelphia,PA 19111. Phone: (215) 214-4228; Fax: (215)728-3639; E-mail: [email protected].

Submitted: September 22, 2008Accepted: November 14, 2008

268 Gastrointestinal Cancer Research Volume 2 • Issue 6

E. D. Tetzlaff, J. D. Cheng, J. A. Ajani

increase—albeit in the setting of in-creased reporting of asymptomatic VTEwith unknown clinical significance. Never-theless, it is important that diligent report-ing of thromboembolic events be under-taken and maintained so that a betterunderstanding of the natural history of thisphenomenon can be gained.

Localized DiseaseThe reporting of VTE has been even lessconsistent in clinical trials in patients withlocalized GI cancers. Recent adjuvant andpreoperative phase III trials for gastro-esophageal cancer have not reported therate of VTE during treatment or during theperioperative period.25,26 In a small phase IItrial, a retrospective analysis, and a briefcommunication, reported rates of VTE forlocalized gastroesophageal cancer were6.1% to 16% in patients treated with chemo-radiotherapy and 30% in patients treatedwith preoperative biochemotherapy.18,27,28

The rates of VTE in adjuvant and pre-operative therapy trials for most other GIcancers are equally under-reported, with theexception of colon cancer. In the MOSAICtrial, which investigated two different chemo-therapy regimens for adjuvant therapy instage II and III colorectal cancer, the rate ofVTE among patients who received at leastone cycle of chemotherapy was 6.1%.29

The effort at reporting VTE has been dupli-cated in other recent phase III trials in ad-juvant therapy for colorectal cancer.30,31

Thus, the risk of VTE during adjuvanttherapy for GI cancers currently is not welldefined. Trial investigators should be en-couraged to report VTE as part of theiroriginal publication or in publications ofcorrelative studies. It would be beneficialfor younger investigators to seize an oppor-tunity to fill the gaps in knowledge aboutVTE in GI cancers by adopting such re-porting as a routine part of the reporting oftrial results.

RISK FACTORSThe risk factors for the development of VTEcan be divided into intrinsic and extrinsicfactors. More data on risk factors are avail-able for pancreatic cancer than for other GIcancers. In a study by Blom et al in 202patients with pancreatic cancer, patientswith tumors of the corpus and cauda of thepancreas had a 2- to 3-fold increased risk

of VTE compared with tumors located inthe caput.32 In addition, presence of distantmetastasis increased the risk of VTE 2-fold(hazard ratio [HR] 1.9, 95% confidence in-terval [CI] 0.7–5.1). Other intrinsic tumor-related risk factors may include high levelsof tissue factor expression, platelet aggre-gation induced by tumor cells, high levelsof plasma plasminogen activator inhibitor-

1, and elevated levels of other proteinsactivated in the coagulation cascade.32–35

With regard to extrinsic factors, Blom etal also determined that receiving chemo-therapy (adjusted HR 4.8) and the 30-daypostoperative period (adjusted HR 4.5)were associated with increased risk for VTE.Patients receiving radiotherapy and thosewith mucinous adenocarcinomas of thepancreas did not appear to be at greaterrisk for VTE. Other series have confirmedthat the location of the primary pancreatictumor affects the risk of VTE, while chal-lenging the notion that mucinous adeno-carcinomas are not associated with in-creased risk.36,37

It is expected that surgery and thepresence of metastatic disease wouldincrease the risk of VTE in patients withpancreatic cancer, since both conditionscontribute to Virchow’s classic triad ofvenous stasis, vessel wall damage, and anincreased hypercoaguable state. Surgicaltrauma can lead to vessel wall damagewhile the presence of metastatic diseasecan lead to venous stasis due to vesselobstruction from tumor or adenopathy ordecreased patient mobility. However, it isnot clear why radiation therapy does notappear to increase the risk of VTE to thesame degree as chemotherapy. One expla-nation could be the selection bias ofpatients for radiation therapy; patients withlower disease burden and better perform-ance status are selected for aggressive localtherapy, and these factors might be associ-ated with lower overall risk for VTE.

Fewer data are available on intrinsicand extrinsic risk factors for VTE in otherGI cancers. However, it is likely that riskfactors reported for pancreatic tumors andother solid tumors are applicable to otherGI cancers, as well. These variables includetreatment with cytotoxic chemotherapy, angi-ogenesis inhibitors, erythropoietin stimulat-ing agents and endocrine/metabolic agents,presence of metastatic disease, number ofmedical comorbidities, tumor histology,presence of central venous catheters, andcompression or obstruction of vessels bytumors or adenopathy (Table 2).38–45

MORBIDITY AND MORTALITYVTE is associated with considerable mor-bidity. Patients who develop clinicallyrelevant DVT may suffer from many direct

Table 2. Reported risk factors fordevelopment of VTE in cancer38-45

Clinical StageAdvanced > localizedLocation of Primary Tumor

HistologyAdenocarcinoma >squamous-cell carcinomaMucinous > nonmucinous

Chemotherapy AgentsCisplatin > oxaliplatinIrinotecan

Vascular Epithelial Growth Factor InhibitorsBevacizumab(arterial thrombotic events)

Erythropoiesis-Stimulating AgentsDarbepoetinErythropoetin

Radiotherapy

Surgery

Medical ComorbiditiesNumber of conditions

Procoagulant FactorsTissue factorPlasminogen activator inhibitor-1Cancer procoagulant

Hereditary Risk FactorsFactor V LeidenProthrombin 2021A mutations

Table 1. Risk analysis using Medicareclaims data.

Rate of DVT/PE Rank outPer 10,000 of 18

Cancer patients malignancies

Pancreas 110 3

Stomach 85 5

Colon 76 8

Liver 69 9

Rectal 62 10

Esophagus 43 15

Abbreviations: DVT = deep vein thrombosis;PE = pulmonary embolism.From Levitan et al.2


Thromboembolism in Gastrointestinal Cancers

adverse effects of thromboembolism, suchas peripheral edema, pain, chronic venousinsufficiency, and pulmonary hyperten-sion. In addition, anticoagulation therapyfor thromboembolism can be a significantburden. The development of VTE in pa-tients with advanced cancer complicatessymptom management and the alreadydaunting challenge of maintaining a pa-tient’s quality of life. The morbidity of VTEfurther underscores the necessity to dedi-cate additional research not only towardprevention and management of VTE, butalso toward maintaining or improvingquality of life in patients with concurrentmalignancy and thromboembolism.

Recognition of the effect of VTE onmortality in patients with GI cancers is justnow emerging. Tetzlaff et al reported thatpatients with advanced gastric cancertreated in clinical trials had a significantlyshorter median survival if they developedVTE before or during protocol chemo-therapy compared with patients who neverdeveloped VTE (3.9 months vs. 8.7 months,P = .007).46 In a study in 198 patients withlocalized gastroesophageal cancer, theseinvestigators also found that patientsdeveloping VTE before or during chemora-diotherapy had a significantly decreasedmedian survival compared with patientswithout VTE (17.7 months vs. 32 months,P = .014). The development of VTE was anindependent prognostic factor for overallsurvival in multivariate analysis.18

The negative effect of VTE on survivalhas also been demonstrated in colorectalcancer and pancreatic cancer. In a retro-spective analysis of two merged databasesin California including 68,142 colorectalcancer patients, it was shown that VTE wasa significant predictor of death within 1year of cancer diagnosis for patients withlocal and regional stage colorectal cancer(HR 1.5 for both).38 In what appears to bethe first randomized trial to report theeffect of VTE on survival in colorectalcancer, Mandalà et al found that the devel-opment of VTE in 203 patients withmetastatic disease had a negative effect onsurvival.47 This finding remained significantafter adjusting for age, disease site, andtreatment schedule in multivariate analysis(HR 1.6, 95% CI 1.0–2.5).

In 227 patients with unresectable pan-creatic cancer, Mandalà et al found that

VTE during treatment was associated witha decrease in progression-free survival(HR 2.59, 95% CI 1.69–3.97) and overallsurvival (HR 1.64, 95% CI 1.04–2.58).48

VTE during chemotherapy remained a sig-nificant prognostic factor after multivariateanalysis. A second series reported by Zawinet al supports the observation that VTE is apoor prognostic factor.9 In this series, ahigh rate of VTE (71% overall) was foundamong 21 chemotherapy-naïve patients withadvanced pancreatic cancer enrolled in aphase II trial; median survival was 8 monthsfor patients with VTE compared with 21months for patients without VTE.

Two series in gastroesophageal cancerdid not show a detrimental effect of VTE onsurvival.17,49 It should be noted that onestudy, in which 12 (25%) of 47 patients de-veloped VTE, was not powered to detect adifference in survival.17 The second series49

included patients with catheter-relatedthrombosis in the overall rate of VTE, whichmay have limited the ability of the investi-gators to detect a detrimental effect onsurvival. It could be the case that VTE thatdevelops as a result of malignancy is amore accurate manifestation of aggressivetumor biology and thus associated with in-creased mortality, whereas VTE related tocentral venous catheters might be amanifestation of vascular trauma andendothelial reactivity that does not reflecttumor biology.

The role that VTE plays in patient sur-vival merits further attention. Available datasuggests that VTE is generally a poorprognostic factor for patients with cancer.It is unclear, however, whether the effect ofVTE on survival is a direct reflection ofaggressive tumor biology or rather relatedto the development and complications ofVTE. If the latter is true, one strategy thatmight improve outcomes is selective anti-coagulation for patients thought to be atincreased risk for the development of VTE.Alternatively, if the development of VTErepresents aggressive tumor biology, it maybe possible to target anticancer therapy atthe up-regulated pathway that leads tohypercoagulation and thrombosis— eg,with heparins. In animal models, heparinshave been shown to inhibit cancer cell ad-hesion, proliferation, migration, and invasion,providing a rationale for further investiga-tion of their use in preventing VTE.50

ANTICOAGULATION ANDSURVIVALAnticoagulation is recommended for pa-tients with cancer and thromboembolismfor the secondary prevention of recurrentthrombosis. In patients with cancer, antico-agulation with low molecular weight hepa-rin (LMWH), unfractionated heparin, orvitamin K antagonists is often prescribed.Initial studies examining anticoagulation inpatients with cancer focused on recurrentthrombosis and bleeding risks as endpoints. However, post hoc and subgroupanalyses and meta-analyses from thesetrials have suggested that anticoagulationmay have a beneficial effect on survival inpatients with cancer and a history ofthrombosis.51–55 These findings remaincontroversial, though, since the reports ofbenefit involve heterogeneous cohorts ofpatients with differing primary tumors,stage of disease, and current treatment.

Several prospective trials have exam-ined the effect of anticoagulation on sur-vival in patients without a history of throm-bosis. In a study by Klerk et al, 302patients with advanced solid tumors wererandomized to 6 weeks of treatment withdaily nadroparin or placebo.56 Patients wereallowed to receive concomitant anticancertherapy at the discretion of the treatingphysician. The median survival was signif-icantly longer in the nadroparin arm (8.0months) compared with the placebo arm(6.6 months) (HR 0.75, 95% CI 0.59–0.96, P = .02). The survival difference re-mained significant after adjusting forperformance status, histology, concomitanttreatment, and life expectancy.

In contrast, the Fragmin AdvancedMalignancy Outcome Study (FAMOUStrial), a randomized, double-blind, multi-center trial with the primary end point ofmortality at 1 year, found no survivalbenefit with anticoagulation treatment.57 Inthe FAMOUS trial, 385 patients with ad-vanced solid tumors and no history of throm-bosis received daily injections of dalteparin(5,000 IU) or placebo. There was no differ-ence in survival at 1 year from randomiza-tion between the dalteparin group and theplacebo group (10.8 months vs. 9.14months, P = .19). In a post-hoc analysis,dalteparin improved median survival by19.2 months in patients with a goodprognosis and who had survived greater

270 Gastrointestinal Cancer Research Volume 2 • Issue 6


than 17 months from randomization com-pared with placebo patients in the samecategory. However, this analysis was notdefined a priori and represented only 26%of the intention-to-treat population.

In a smaller phase III trial, 141 patientswith advanced cancer were randomized tostandard clinical care plus LMWH or stan-dard care alone.58 Median survival did notdiffer between the two groups in the inten-tion-to-treat population, in the subgroup ofpatients with a favorable prognosis (patientsthat survived longer than 6 months), or insubgroups based on disease site. LMWHdid not increase the rate of toxic effects forpatients nor was it able to improve patients’quality of life. A meta-analysis was per-formed to determine the potential effect ofLMWH on survival in patients with ad-vanced cancer, including the three trialsdiscussed above plus a trial reported byAltinbas et al.59,60 The meta-analysis sug-gested that LMWH plus standard therapyimproved survival at 1 and 2 years withodds ratios of 0.70 (95% CI 0.49–1.00, P= .05) and 0.57 (95% CI 0.42–0.84, P =.03), respectively.60 In addition, LMWH wasnot associated with an increased risk ofbleeding.

The effect of anticoagulation on survi-val has been examined according tospecific disease sites, as well. In a smallrandomized study in small-cell lung cancer,84 patients received chemotherapy alone(cyclophosphamide/epirubicin/vincristine)or the same chemotherapy regimen withdalteparin (5,000 IU daily).59 The durationof anticoagulation was limited to the 18weeks of chemotherapy. At a medianfollow up of 10 months, there was a signif-icant difference in median survival in favor

of the patients treated with chemotherapyplus dalteparin vs. chemotherapy alone(10 months vs. 6 months, P = .01).Progression-free survival was also signifi-cantly longer in patients treated withdalteparin, and a nonsignificant trend in im-provement in response rate was reported.The findings of this study are similar toother small studies in lung cancer, sug-gesting a benefit for treatment with heparinor coumarin derivatives.61–63

At this time, insufficient evidence existsto support initiation of anticoagulation thera-py in the absence of thromboembolism.Still, continued research is warranted inthe subgroups of patients in which a sug-gestion of benefit has emerged. Initiatingwell-designed trials in locally advancedcancers of specific sites would be arational approach to defining the potentialutility of anticoagulation therapy, and suchtrials should be of interest to both patientsand researchers.

The potential benefit of anticoagulationon survival in GI cancers cannot be de-fined from the data discussed above, sinceonly a minority of the patients included inthose trials had GI cancers. It is thus fortu-nate that a number of studies are activelyrecruiting patients to help answer this im-portant clinical question (Table 3). Prelim-inary safety results have been presentedfrom randomized phase II and III trialsexamining nadroparin, enoxaparin, or dal-teparin in combination with chemotherapyin GI cancers. Three trials in pancreaticcancer have reported that LMWH incombination with gemcitabine was safe,but efficacy data are not yet mature.64–66

Results from these trials and others in solidtumors are pending.

CONCLUSIONA significant line of research needs to bededicated to investigation of thromboem-bolic phenomena in GI cancers. The effectof VTE on survival continues to be clarified.As well, the role of anticoagulation on thesurvival of patients with GI cancers alsoremains to be defined. Opportunities foradditional research include investigation ofmethods to identify patients at risk for thedevelopment of VTE, as well as investiga-tion of strategies and therapeutic interven-tions to reduce the morbidity and mortalityof VTE. In addition, more convenient anti-coagulation therapy schedules (daily/weekly) and routes (oral) are needed to im-prove the management of VTE while mini-mizing toxicity and reducing monitoringrequirements.

REFERENCES1. Trousseau A, Bazire PV, Cormack JR: Lectures

on Clinical Medicine. London, R. Hardwicke, 1867

2. Levitan N, Dowlati A, Remick SC, et al: Rates ofinitial and recurrent thromboembolic diseaseamong patients with malignancy versus thosewithout malignancy. Risk analysis using Medi-care claims data. Medicine (Baltimore), 78:285–291, 1999

3. Sorensen, HT, Mellemkjaer L, Olsen JH, et al:Prognosis of cancers associated with venousthromboembolism. N Engl J Med 343:1846–1850, 2000

4. Sorensen HT, Mellemkjaer L, Steffensen FH, etal: The risk of a diagnosis of cancer after pri-mary deep venous thrombosis or pulmonaryembolism. N Engl J Med 338:1169–1173, 1998

5. Furie B, Furie BC: Cancer-associated thrombo-sis. Blood Cells Mol Dis 36:177–181, 2006

6. Furie B, Furie BC: Mechanisms of ThrombusFormation. N Engl J Med 359:938–949, 2008

7. Carrier M, Le Gal G, Wells PS, et al: Systematicreview: the Trousseau syndrome revisited:should we screen extensively for cancer inpatients with venous thromboembolism? AnnIntern Med 149:323–333, 2008

Table 3. Ongoing and completed trials of low molecular weight heparin in gastrointestinal cancers

Study Identifier* Patient population Treatment Trial status Phase Primary end point

NCT00718354 Gastric cancer Standard CT Recruiting IIIb Overall survival± enoxaparin

NCT00312013 Pancreatic cancer Standard CT Recruiting III Death due to allProstate cancer ± nadroparin causes at study endLung cancer

NCT00003674 Colorectal cancer Standard CT Completed III Overall survivalBreast cancer ± dalteparinLung cancerProstate cancer

Abbreviations: CT=chemotherapy.*www.clinicaltrials.gov


Thromboembolism in Gastrointestinal Cancers

8. Asmis TR, Templeton M, Trocola R, et al: Inci-dence and significance of thromboembolicevents (TE) in patients with gastrointestinal (GI)and non-GI malignancies on systemic cytotoxictherapy. J Clin Oncol 25:18S, 2007 (abstr9049)

9. Zawin M, Camputaro C, Rahman Z, et al: Multi-row detector CT detection of subclinical visceraland pulmonary thrombosis in advanced pancre-atic cancer is an adverse prognostic indicator.Proc Am Soc Clin Oncol 22: 2003 (abstr 1419)

10. Burris HA 3rd, Moore MJ, Andersen J, et al: Im-provements in survival and clinical benefit withgemcitabine as first-line therapy for patientswith advanced pancreas cancer: a randomizedtrial. J Clin Oncol 15:2403–2413, 1997

11. Kindler HL, Niedzwiecki D, Hollis D, et al: Adouble-blind, placebo-controlled, randomizedphase III trial of gemcitabine (G) plus beva-cizumab (B) versus gemcitabine plus placebo(P) in patients (pts) with advanced pancreaticcancer (PC): a preliminary analysis of Cancerand Leukemia Group B (CALGB). J Clin Oncol25:18S, 2007 (abstr 4508)

12. Louvet C, Labianca R, Hammel P, et al: Gemcita-bine in combination with oxaliplatin comparedwith gemcitabine alone in locally advanced ormetastatic pancreatic cancer: results of a GER-COR and GISCAD phase III trial. J Clin Oncol23:3509–3516, 2005

13. Moore MJ, Goldstein D, Hamm J, et al: Erlotinibplus gemcitabine compared with gemcitabinealone in patients with advanced pancreaticcancer: a phase III trial of the National CancerInstitute of Canada Clinical Trials Group. J ClinOncol 25:1960–1966, 2007

14. Philip PA, Benedetti J, Fenoglio-Preiser C, et al:Phase III study of gemcitabine [G] plus cetuxi-mab [C] versus gemcitabine in patients [pts]with locally advanced or metastatic pancreaticadenocarcinoma [PC]: SWOG S0205 study. JClin Oncol 25:18S, 2007 (abstr 4509)

15. Cunningham D, Starling N, Rao S, et al: Cape-citabine and oxaliplatin for advanced esophago-gastric cancer. N Engl J Med 358:36–46, 2008

16. Dank M, Zaluski J, Barone C, et al: Random-ized phase 3 trial of irinotecan (CPT-11) +5FU/folinic acid (FA) vs CDDP + 5FU in 1st-lineadvanced gastric cancer patients. J Clin Oncol23:16S, 2005 (abstr 4003)

17. Shah MA, Ramanathan RK, Ilson DH, et al: Multi-center phase II study of irinotecan, cisplatin,and bevacizumab in patients with metastaticgastric or gastroesophageal junction adenocar-cinoma. J Clin Oncol 24:5201–5206, 2006

18. Tetzlaff ED, Correa AM, Komaki R, et al: Signifi-cance of thromboembolic phenomena occur-ring before and during chemoradiotherapy forlocalized carcinoma of the esophagus and gastro-esophageal junction. Dis Esophagus 2008 May2 [Epub ahead of print]

19. Van Cutsem E, Moiseyenko VM, Tjulandin S, etal: Phase III study of docetaxel and cisplatinplus fluorouracil compared with cisplatin andfluorouracil as first-line therapy for advancedgastric cancer: a report of the V325 StudyGroup. J Clin Oncol 24:4991–4997, 2006

20. Barbare JC, Bouche O, Bonnetain F, et al: Ran-domized controlled trial of tamoxifen in ad-vanced hepatocellular carcinoma. J Clin Oncol23:4338–4346, 2005

21. Giantonio BJ, Catalano PJ, Meropol NJ, et al:Bevacizumab in combination with oxaliplatin,fluorouracil, and leucovorin (FOLFOX4) for pre-viously treated metastatic colorectal cancer:results from the Eastern Cooperative OncologyGroup Study E3200. J Clin Oncol 25:1539–1544, 2007

22. Hurwitz H, Fehrenbacher L, Novotny W, et al:Bevacizumab plus irinotecan, fluorouracil, andleucovorin for metastatic colorectal cancer. NEngl J Med 350:2335–2342, 2004

23. Iyer RV, Gibbs J, Kuvshinoff B, et al: A phase IIstudy of gemcitabine and capecitabine inadvanced cholangiocarcinoma and carcinomaof the gallbladder: a single-institution prospec-tive study. Ann Surg Oncol. 14:3202–3209 2007

24. Knox JJ, Hedley D, Oza A, et al: Combininggemcitabine and capecitabine in patients withadvanced biliary cancer: a phase II trial. J ClinOncol 23:2332–2338, 2005

25. Cunningham D, Allum WH, Stenning SP, et al:Perioperative chemotherapy versus surgeryalone for resectable gastroesophageal cancer.N Engl J Med. 355:11–20, 2006

26. Macdonald JS, Smalley SR, Benedetti J, et al:Chemoradiotherapy after surgery comparedwith surgery alone for adenocarcinoma of thestomach or gastroesophageal junction. N Engl JMed 345:725–730, 2001

27. Shah MA, Ilson D, Kelsen DP: Thromboembolicevents in gastric cancer: high incidence inpatients receiving irinotecan- and bevacizumab-based therapy. J Clin Oncol 23:2574–2576,2005

28. Ilson DH, Bains M, Kelsen DP, et al: Phase Itrial of escalating-dose irinotecan given weeklywith cisplatin and concurrent radiotherapy inlocally advanced esophageal cancer. J ClinOncol 21:2926–2932, 2003

29. Andre T, Boni C, Mounedji-Boudiaf L, et al:Oxaliplatin, fluorouracil, and leucovorin as adju-vant treatment for colon cancer. N Engl J Med350:2343–2351, 2004

30. Andre T, Colin P, Louvet C, et al: Semimonthlyversus monthly regimen of fluorouracil and leu-covorin administered for 24 or 36 weeks asadjuvant therapy in stage II and III colon can-cer: results of a randomized trial. J Clin Oncol21:2896–2903, 2003

31. Saltz LB, Niedzwiecki D, Hollis D, et al: Irino-tecan fluorouracil plus leucovorin is not superiorto fluorouracil plus leucovorin alone as adjuvanttreatment for stage III colon cancer: results ofCALGB 89803. J Clin Oncol 25:3456–3461,2007

32. Blom JW, Osanto S, Rosendaal FR: High risk ofvenous thrombosis in patients with pancreaticcancer: a cohort study of 202 patients. Eur JCancer 42:410–414, 2006

33. Andren-Sandberg A, Lecander I, Martinsson G,et al: Peaks in plasma plasminogen activatorinhibitor-1 concentration may explain throm-botic events in cases of pancreatic carcinoma.Cancer 69:2884–2887, 1992

34. Khorana AA, Ahrendt SA, Ryan CK, et al: Tissuefactor expression, angiogenesis, and thrombo-sis in pancreatic cancer. Clin Cancer Re13:2870–2875, 2007

35. Khorana AA, Fine RL: Pancreatic cancer andthromboembolic disease. Lancet Oncol 5:655–663, 2004

36. Mao C, Domenico DR, Kim K, et al: Obser-vations on the developmental patterns and theconsequences of pancreatic exocrine adeno-carcinoma. Findings of 154 autopsies. ArchSurg 130:125–134, 1995

37. Pinzon R, Drewinko B, Trujillo JM, et al: Pan-creatic carcinoma and Trousseau's syndrome:experience at a large cancer center. J ClinOncol 4:509–514, 1986

38. Alcalay A, Wun T, Khatri V, et al: Venous throm-boembolism in patients with colorectal cancer:incidence and effect on survival. J Clin Oncol24:1112–1118, 2006

39. Bennett CL, Silver SM, Djulbegovic B, et al:Venous thromboembolism and mortality associ-ated with recombinant erythropoietin and dar-bepoetin administration for the treatment ofcancer-associated anemia. JAMA 299:914–924, 2008

40. De Cicco M: The prothrombotic state in cancer:pathogenic mechanisms. Crit Rev OncolHematol 50:187–196, 2004

41. Rooden CJ, Tesselaar ME, Osanto S, et al: Deepvein thrombosis associated with central venouscatheters— a review. J Thromb Haemost3:2409–2419, 2005

42. Scappaticci FA, Skillings JR, Holden SN, et al:Arterial thromboembolic events in patients withmetastatic carcinoma treated with chemothera-py and bevacizumab. J Natl Cancer Inst99:1232–1239, 2007

43. Tesselaar M, Steeghs N, Rosendaal F, et al:Incidence of thrombosis in gastro-esophagealcancer: a cohort study of 761 patients. J ClinOncol 22:14S, 2004 (abstr 8218)

44. Hussain S, Singh M, Shi R, et al: Megestrol ace-tate increases the incidence of deep venousthrombosis in patients with non small cell lungcancer. J Clin Oncol. 24:18S, 2006 (abstr18510)

45. Blom JW, Doggen CJ, Osanto S, et al: Malig-nancies, prothrombotic mutations, and the riskof venous thrombosis. JAMA 293:715–722, 2005

46. Tetzlaff ED, Correa AM, Baker J, et al: Theimpact on survival of thromboembolic phenom-ena occurring before and during protocolchemotherapy in patients with advanced gas-troesophageal adenocarcinoma. Cancer 109:1989–1995, 2007

47. Mandala M, Barni S, Isa L, et al: Incidence andclinical implications of venous thromboem-bolism in advanced colorectal cancer patients:findings from the ‘GISCAD-Alternating sched-ule’ study. J Clin Oncol 26:15S, 2008 (abstr20502)

48. Mandala M, Reni M, Cascinu S, et al: Venousthromboembolism predicts poor prognosis inirresectable pancreatic cancer patients. AnnOncol 18:1660–1665, 2007

49. Starling N, Rao S, Norman AR, et al: Analysis ofthromboembolic events (TEs) in the REAL-2randomised controlled (RCT) study of fourchemotherapy regimens for the treatment ofoesophagogastric (OG) cancers. 2007 Gastro-intestinal Cancers Symposium, Orlando, FL;l2007 (abstr 74)

50. Smorenburg SM, Van Noorden CJF: The com-plex effects of heparins on cancer progressionand metastasis in experimental studies.Pharmacol Rev 53:93–106, 2001

51. Lyman GH, Khorana AA, Falanga A, et al:

Volume 2 • Issue 6Gastrointestinal Cancer Research272

Disclosures of Potential Conflicts of Interest

The authors indicated no potential conflicts of interest.

American Society of Clinical Oncology guide-line: recommendations for venous thromboem-bolism prophylaxis and treatment in patientswith cancer. J Clin Oncol 25:5490–5505, 2007

52. von Tempelhoff GF, Harenberg J, Niemann F, etal: Effect of low molecular weight heparin (Cer-toparin) versus unfractionated heparin on cancersurvival following breast and pelvic cancer surgery:a prospective randomized double-blind trial. IntJ Oncol 16:815-824, 2000

53. Lee AYY, Rickles FR, Julian JA, et al: Ran-domized comparison of low molecular weightheparin and coumarin derivatives on the sur-vival of patients with cancer and venous throm-boembolism. J Clin Oncol 23:2123–2129, 2005

54. Lee AY, Levine MN, Baker RI, et al: Low-molec-ular-weight heparin versus a coumarin for theprevention of recurrent venous thromboem-bolism in patients with cancer. N Engl J Med349:146–153, 2003

55. Ferretti G, Bria E, Giannarelli D, et al: Low-molecular-weight heparin versus oral antico-agulant therapy for the long-term treatment ofsymptomatic venous thromboembolism: isthere any difference in cancer-related mortality?J Clin Oncol 23:7248–7250, 2005

56. Klerk CPW, Smorenburg SM, Otten H-M, et al:

The effect of low molecular weight heparin onsurvival in patients with advanced malignancy.J Clin Oncol 23:2130–2135, 2005

57. Kakkar AK, Levine MN, Kadziola Z, et al: Lowmolecular weight heparin, therapy with dal-teparin, and survival in advanced cancer: theFragmin Advanced Malignancy Outcome Study(FAMOUS). J Clin Oncol 22:1944–1948, 2004

58. Sideras K, Schaefer PL, Okuno SH, et al: Low-molecular-weight heparin in patients withadvanced cancer: a phase 3 clinical trial. MayoClin Proc 81:758–767, 2006

59. Altinbas M, Coskun HS, Er O, et al: A random-ized clinical trial of combination chemotherapywith and without low-molecular-weight heparinin small cell lung cancer. J Thromb Haemost2:1266–1271, 2004

60. Lazo-Langner A, Goss GD, Spaans JN, et al:The effect of low-molecular-weight heparin oncancer survival. A systematic review and meta-analysis of randomized trials. J ThrombHaemost 5:729–737, 2007

61. Chahinian AP, Propert KJ, Ware JH, et al: A ran-domized trial of anticoagulation with warfarinand of alternating chemotherapy in extensivesmall-cell lung cancer by the Cancer and Leu-kemia Group B. J Clin Oncol 7:993–1002, 1989

62. Lebeau B, Chastang C, Brechot JM, et al: Sub-cutaneous heparin treatment increases survivalin small cell lung cancer. Cancer 74:38–45,1994

63. Zacharski LR, Henderson WG, Rickles FR, etal: Effect of warfarin anticoagulation on survivalin carcinoma of the lung, colon, head andneck, and prostate: Final Report of VA cooper-ative study # 75. Cancer 53:2046–2052, 1984

64. Voorthuizen TV, Vervenne WL, Van Daalen EH,et al: A randomized phase II study comparinggemcitabine plus nadroparine versus gemcita-bine in patients with locally advanced or meta-static pancreatic carcinoma: The GEMFRAXtrial. J Clin Oncol. 24:18S, 2006 (abstr 4112)

65. Pelzer U, Hilbig A, Stieler J, et al: A prospective,randomized trial of simultaneous pancreaticcancer treatment with enoxaparin andchemotherapy (PROSPECT—CONKO 004). JClin Oncol 24:18S, 2006 (abstr 4110)

66. Maraveyas A, Holmes M, Lofts F, et al: Chemo-anticoagulation versus chemotherapy in ad-vanced pancreatic cancer (APC): results of theinterim analysis of the FRAGEM trial. J ClinOncol. 25:18S, 2007 (abstr 4583)


Cost-effectiveness of Dalteparin Versus

Unfractionated Heparin as Venous

Thromboembolism Prophylaxis in

Malignant Gynecologic Surgery

William E. Wade, Pharm D, FASHP, FCCP* andWilliam J. Spruill, Pharm D, FASHP, FCCP

Venous thromboembolic events (VTEs) are serious complications that may occur in the patientundergoing surgery for gynecologic malignancies. The American College of Chest Physiciansrecommends unfractionated heparin or low–molecular weight heparin as prophylaxis for deep veinthrombosis and pulmonary embolism in this patient population. Cost-effectiveness analysescomparing unfractionated heparin 3 times a day versus once daily dalteparin using publishedefficacy and safety data demonstrate cost savings if dalteparin were routinely utilized as VTEprophylaxis. Sensitivity analyses support this finding at the upper end of the range of reportedproximal DVT, nonfatal pulmonary embolism, and major bleeding incidences. These findings shouldbe viewed as preliminary, and institutions are encouraged to perform their own cost-effectivenessstudies in this patient population.

Keywords: venous thromboembolic events, dalteparin, heparin, cost-effectiveness, gynecologicmalignancy

INTRODUCTION

Venous thromboembolic events (VTEs) are major causesof morbidity and mortality in patients undergoing sur-gery for gynecologic malignancies. Deep vein thrombosis(DVT) occurs in approximately 38% of this population,and pulmonary embolism (PE) is a leading cause ofpostoperative death in patients at highest risk.1

Additionally, intraoperative bleeding is a majorconcern in these patients. The American College ofChest Physicians recommends low–molecular weightheparin (LMWH) .3400 units/d or unfractionatedheparin (UFH) 5000 units every 8 hours as prophylaxisin these patients.2 The purpose of the current study isto perform a cost analysis on the use of the LMWH,

dalteparin, versus UFH in patients undergoing surgicalintervention for gynecologic malignancy using pub-lished efficacy and safety data.

MATERIALS AND METHODS

A PubMed literature search was conducted to identifyall published English language studies evaluatingeither UFH or LMWH as DVT and/or PE prophylaxisin patients undergoing surgical intervention for gyne-cologic malignancies. Preferred studies included pro-spective, randomized trials. Studies reported as abstractswere excluded due to limited information provided,and no attempt was made to identify unpublishedstudies regarding this subject matter. To be includedin the current study, efficacy trials had to report thefollowing data: types and numbers of patients evalu-ated; prophylaxis regimen used; duration of prophy-laxis; and incidences of total DVT, proximal DVT,and nonfatal PE and major bleeding (MB) episodesobserved. Data were pooled and averaged when

Department of Clinical and Administrative Pharmacy, College ofPharmacy, University of Georgia, Athens, GA.*Address for correspondence: Department of Clinical and Admin-istrative Pharmacy, College of Pharmacy, University of Georgia,Athens, GA 30602. E-mail: [email protected]

American Journal of Therapeutics 15, 512–515 (2008)

1075-2765 � 2008 Lippincott Williams & Wilkins

more than 1 study was identified evaluating a specificprophylactic regimen.

Cost-effectiveness3,4 analyses were performed tocompare the relevant costs and consequences of theidentified prophylaxis regimens from an institutionalperspective in 1000 patients over a 5-day period. Asocietal perspective was utilized for costs. A therapy isconsidered more cost effective if it has a favorable cor-responding cost relative to competing alternatives. Tomake this determination, incremental cost-effectivenessratios (ICERs) were calculated.4 Incremental costs arethe additional costs one program or treatment optionimposes over another and compares them with anyadditional effect, benefit, or outcome produced bya selected treatment. These calculations were made bydividing the difference in all costs associated with theidentified prophylactic regimens by the difference inthromboembolic events avoided with each regimen.The outcome of this computation represents the num-ber of dollars saved per thromboembolic event avoidedif 1 regimen were routinely used as DVT prophylaxisover the other in this patient population. Literature-reported values of costs discounted at 3% per year toreflect 2007 US dollars utilized in these calculationsare found in Table 1.5 The cost for heparin 5000 unitssubcutaneously every 8 hours for 5 days plus anappropriate dispensing fee is US $135 per patient.6

The cost of dalteparin 5000 units subcutaneously dailyfor 5 days plus an appropriate dispensing fee is US$195 per patient.6

Simple sensitivity analyses combined with analysisof extremes were performed to determine the robust-ness of the ICER calculations.7 Sensitivity analyses arecommonly performed on economic evaluations forthe purpose of determining the soundness of studyconclusions. These analyses consist of mathematicalmanipulations of variables used in performing thecost-effectiveness calculations over a plausible range. Ifstudy results do not vary significantly over the range ofvalues of the study variables, confidence in the studyresults can be expressed. Simple sensitivity analyses areextremely effective in determining the generalizabilityof study results. Analysis of extremes considers upper

and lower cost estimates and the upper and lowereffectiveness of treatment options being compared todetermine the best case and worst case therapeuticmodalities. Variables evaluated in these analyses in thecurrent study include drug costs and the range ofliterature-reported incidences of proximal DVT, non-fatal PE, and MB.

A final cost calculation performed in this study wascost per death averted with UFH or dalteparinprophylaxis versus no prophylaxis. The literaturereports a fatal PE rate of 1% to 5% in the absence ofprophylaxis in patients at very high risk for this com-plication.8 We assumed an average rate of 2.5% forstudy purposes. The cost per death averted was calcu-lated by dividing the cost of providing prophylaxis to1000 patients by the number of patients expected to dieif they did not receive prophylaxis.

RESULTS

A total of 11 studies evaluating either UFH or LMWHas DVT prophylaxis following surgery for gynecologicmalignancies were identified. Eight of these studies didnot report results in a format that could be used in ouranalysis. Fricker et al9 compared calcium heparin 5000units 2 hours before surgery followed by 5000 unitsevery 8 hours with Fragmin (dalteparin) 2500 anti-Xaunits 2 hours before surgery followed by an additionaldose 12 hours later and then 5000 anti-Xa units daily.Both regimens were continued for 10 days. SuspectedDVT was confirmed by venography, whereas clinicallysuspected PE was confirmed by lung scintigraphy andarterial gazometry. Thirty-two patients received hep-arin, and 28 patients received Fragmin. No patients ineither group developed a DVT, whereas no patients inthe Fragmin group and 2 patients in the calciumheparin group developed nonfatal PE (without evi-dence of DVT). Two patients in the Fragmin group and1 in the UFH group experienced MB.

Clarke-Pearson et al10 compared low-dose heparinwith intermittent pneumatic calf compression (IPC) asDVT prophylaxis. Patients were randomly assigned to

Table 1. Literature-reported costs averaged and discounted to 2007 US dollars.

Maxwell et al5

(2000), in $Etchells et al12

(1999), in $Gould et al13

(1999), in $

Discounted valueused in currentstudy (2007), in $

Dx and Tx PDVT 3869 1990 NR 3640Dx and Tx PE 5278 3600 6187 6296Tx MB 1500 2000 1245 1985

NR = not reported; Dx and Tx = diagnosis and treatment; PDVT = proximal DVT.

American Journal of Therapeutics (2008) 15(6)

Dalteparin Versus Unfractionated Heparin 513

receive heparin (n = 107) 5000 units subcutaneouslyevery 8 hours 1 day before surgery followed by 5000units every 8 hours for 7 postoperative days or untilambulating or discharged, or IPC (n = 101) initiated atinduction of anesthesia and continued for 5 days oruntil ambulating or discharged. Suspected DVT wasconfirmed by ascending venography while signs andsymptoms of PE were further evaluated by ventilation–perfusion lung scanning and pulmonary arteriography.Three patients in the UFH group and 1 in the IPC groupdeveloped a proximal DVT, whereas no patients ineither group developed a nonfatal PE. MB was experi-enced by 3 patients in the UFH group with no patientsreceiving IPC developing this complication.

Maxwell et al11 studied IPC versus dalteparin 2500units subcutaneously beginning 1–2 hours beforesurgery, 2500 units 12 hours later, followed by 5000units daily for 5 days or until discharged. Real-timeduplex and color Doppler ultrasound imaging wereused for DVT assessment. Of the 211 study subjects, 2experienced proximal DVT in the dalteparin group and1 in the IPC group. No patients developed nonfatal PE.Four dalteparin subjects experienced MB in the post-operative period.

In the current study, we performed a cost analysisbased on the efficacy and safety data obtained from the3 above-mentioned trials. Percentage occurrence valuesfor complications used for calculations in this studyfrom the dalteparin trials included proximal DVT1.9%, nonfatal PE 0%, and MB 5.6%. Values used forthe UFH arm included proximal DVT 2.7%, nonfatal PE1.8%, and MB 3.5%. Total costs in the analyses includedcost of the drugs; cost of diagnosing and managingproximal DVT and nonfatal PE; and costs of treatingMB events for the 2 groups. Cost calculations werebased on a cohort of 1000 patients. Table 2 lists totalcosts for each regimen evaluated.

Incremental cost-effectiveness calculations demon-strate that if dalteparin 5000 units daily were routinelyused as prophylaxis in this patient population, $6961.60would be saved for each thromboembolic event avertedover UFH 5000 units every 8 hours. Sensitivity analysesdemonstrate that UFH would produce a cost savingsof $5142.89 over dalteparin at the lower range of VTEsand MB events, whereas dalteparin would producea $7630.20 per patient savings over UFH at the upperrange of reported VTE and MB events, supporting thefindings of the original ICER calculation. At the lowerrange of reported incidences of complications, sensi-tivity analysis demonstrates that if the cost of dalte-parin was reduced to $97 per day, dalteparin wouldproduce cost savings over UFH. Cost per deathavoided was $5400 for UFH and $7800 for dalteparin.

DISCUSSION

By applying the rates of VTEs and complications notedin the Fricker et al, Clarke-Pearson et al, and Maxwellet al studies to our population of 1000 patients, we haveexamined the financial impact of UFH versus daltepar-in as VTE prophylaxis in patients undergoing surgeryfor gynecologic malignancy. In our costing analysis, weelected to include only VTEs that would be routinelydetected and treated in a clinical setting. Subsequently,only costs associated with proximal DVT and non-fatal PE were evaluated. Distal vein thrombosis wasexcluded, as these tend to be asymptomatic and inclinical practice would not commonly be treated. Thisresulted in cost savings when daily dalteparin iscompared with 3-times a day UFH.

It is of interest to note that the dalteparin patients inthese studies had a slightly higher incidence of bleed-ing complications than those receiving UFH. Consider-able extra costs may incur because of this complication.

The results of the sensitivity analysis show therobustness of our incremental cost ratio as dalteparinproduces cost savings at the upper ends of the extremestested when compared with UFH. However, this wasnot the case at the lower end of VTE and MB episodesevaluated.

Limitations to the current study are recognized. Fewstudies have been published in which VTE prophylaxisin this patient population has been reported in theformat allowing for economic assessment of therapy.Additionally, the sample sizes in these publicationswere small. Although the preferred study type forevaluating efficacy and safety of a product is a ran-domized trial, not all studies fit this model. Clearly, theresults of our study must be viewed as preliminary, andlarge-scale randomized trials are needed to evaluate

Table 2. Costs utilized in the ICER calculations(1000 patients).

UFH 5000 unitsevery 8 hours

Dalteparin5000 units

daily

Drug costs (includesdispensing fee) $135,000 $195,000

Dx and Tx DVT 98,280 69,160Dx and Tx PE 113,328 0Treatment MB 63,520 111,160Total costs $410,128 $375,320Cost per patient $410.13 $375.32

Dx and Tx = diagnosis and treatment.


514 Wade and Spruill

efficacy, safety, and costs of various LMWH formula-tions as DVT and PE prophylaxis in this population.Individual institutions may wish to conduct their ownstudies using institution-specific costs of drug therapyand treatment of complications to determine the mostcost-effective agent for DVT prophylaxis in patientsexperiencing gynecologic malignancies.

REFERENCES

1. Crandon AJ, Koutts J. Incidence of post-operative deepvein thrombosis in gynaecological oncology. Aust N Z JObstet Gynaecol. 1983;23:216–219.

2. Geerts WH, Pineo GF, Heit JA, et al. Prevention of venousthromboembolism: the seventh ACCP conference onantithrombotic and thrombolytic therapy. Chest. 2004;126(Suppl 3):338S–400S.

3. Eisenberg JM. Clinical economics. A guide to the economicanalysis of clinical practices. JAMA. 1989;262:2879–2886.

4. Lee J, McLaughlin-Miley C, Chatteron ML. Cost-effec-tiveness analysis. In: Grauer D, Odom T, Osterhaus J, et al,eds. Pharmacoeconomics and Outcomes: Applications forPatient Care. Kansas City, MO: American College ofClinical Pharmacy; 2003:133–145.

5. Maxwell GL, Myers ER, Clarke-Pearson DL. Cost-effectiveness of deep venous thrombosis prophylaxis ingynecologic oncology surgery. Obstet Gynecol. 2000;95:206–214.

6. Anonymous. Red Book. Montvale, NJ: NJ ThomsonHealthcare; 2007:434, 461.

7. Armstrong E. Sensitivity analysis. In: Grauer D, Odom T,Osterhaus J, et al, eds. Pharmacoeconomics and Outcomes:Applications for Patient Care. Kansas City, MO: AmericanCollege of Clinical Pharmacy; 2003:231–242.

8. Clagett GP, Anderson FA Jr, Heit J, et al. Prevention ofvenous thromboembolism. Chest. 1995;108(Suppl 4):312S–334S.

9. Fricker JP, Vergnes Y, Schach R, et al. Low-dose heparinversus low-molecular weight heparin (Kabi 2165, Frag-min) in the prophylaxis of thromboembolic complicationsof abdominal oncological surgery. Eur J Clin Investig. 1988;18:561–567.

10. Clarke-Pearson DL, Synan IS, Dodge R, et al. A ran-domized trial of low-dose heparin and intermittentpneumatic calf compression for the prevention of deepvenous thrombosis after gynecologic oncology surgery.Am J Obstet Gynecol. 1993;168:1146–1154.

11. Maxwell GL, Synan I, Dodge R, et al. Pneumaticcompression versus low molecular weight heparin ingynecologic oncology surgery: a randomized trial. ObstetGynecol. 2001;98:989–995.

12. Etchells E, McLeod RS, Geerts W, et al. Economic analysisof low-dose heparin vs the low-molecular-weight heparinenoxaparin for prevention of venous thromboembolismafter colorectal surgery. Arch Intern Med. 1999;159:1221–1228.

13. Gould MK, Dembitzer AD, Doyle RL, et al. Low-molecular-weight heparins compared with unfractionatedheparin for treatment of acute deep venous thrombosis. Ameta-analysis of randomized, controlled trials. Ann InternMed. 1999;130:789–799.


Dalteparin Versus Unfractionated Heparin 515

Ash Syllabus Cast 137

Documents

Transcript of Ash Syllabus Cast 137