Management of Screen Detected Cancer
3
Clinical Science Symposium Friday, 21 March 2014 S165 Methods: We included patients from the population-based FOCUS cohort, which comprises all incident breast cancer patients aged 65 years or older who were diagnosed in the geographically defined Comprehensive Cancer Centre Region West in The Netherlands between January 1997 and December 2004. The Adjuvant! Online predicted 10-year overall survival and cumulative recurrence rates were calculated. These were compared with observed 10-year overall survival and cumulative recurrence using one-sample T-tests for the whole cohort as well as for several subgroups of patient, tumour and treatment characteristics. Discriminatory accuracy was tested by composing ROC-curves and calculating corresponding C-indices; calibration was tested by plotting observed versus predicted outcomes. Results: Overall 2,012 patients were included. Predicted and observed 10-year overall survival strongly differed for all patients [predicted 48.8%, observed 39.0%, difference 9.8, 95% Confidence Interval (CI) 5.9 to 13.7, p < 0.001], as well as for most subgroups. Similarly, predicted and observed 10-year cumulative recurrence differed substantially (predicted 40.8%, observed 18.2%, difference −22.6, 95% CI −34.6 to −20.6, p < 0.001), and significant inequalities were observed in almost all subgroups. The discriminatory accuracy of Adjuvant! Online was moderate for overall survival (C-index 0.75) and poor for cumulative recurrence (C-index 0.56), as was the calibration (p < 0.001 for both overall survival as well as recurrence). Conclusion: Adjuvant! Online does not accurately predict overall survival and recurrence in older breast cancer patients and should be interpreted with caution. This is in sharp contrast with current guidelines and recommendations in which use of Adjuvant! Online is recommended in older patients. Taken into account the sharp rise of older breast cancer patients in our ageing society, a new prediction tool specifically for older breast cancer patients should be developed, thereby individualizing clinical decision making in older breast cancer patients and optimizing outcome. No conflicts of interest Friday, 21 March 2014 08:45–10:15 CLINICAL SCIENCE SYMPOSIUM Management of Screen Detected Cancer 402 Invited Screen-detected low grade DCIS: Take it or leave it J. Wesseling 1 , L. Elshof 2 , E. Bleiker 3 , N. Bijker 4 , E. Rutgers 5 . 1 The Netherlands Cancer Institute, Department of Pathology, Amsterdam, Netherlands; 2 The Netherlands Cancer Institute, Division of Molecular Pathology, Amsterdam, Netherlands; 3 The Netherlands Cancer Institute, Department of Psychosocial Research and Epidemiology, Amsterdam, Netherlands; 4 Academic Medical Center, Department of Radiotherapy, Amsterdam, Netherlands; 5 The Netherlands Cancer Institute, Department of Surgery, Amsterdam, Netherlands Background: The introduction of population-based breast cancer screen- ing and implementation of digital mammography has led to an increased incidence of ductal carcinoma in situ (DCIS) without a decrease in incidence of advanced breast cancer. This suggests DCIS overdiagnosis exists. Currently, most women with low-grade DCIS are being treated with surgery aiming for radical margins, and, if breast-conserving therapy is given, radiotherapy forms part of this treatment. We hypothesize that asymptomatic, low-grade DCIS can safely be managed by watchful waiting, because these lesions harbor a low risk of progression to invasive breast cancer, and if this progression occurs, this will be low-grade and hormone receptor positive with excellent survival rates. Purpose: We will compare watchful waiting with standard treatment for primary low-grade DCIS detected by microcalcifications only in a non-inferiority randomized clinical trial. Primary outcome is 5-year ipsilateral invasive breast cancer-free rate (iiBCF). Secondary outcomes are complementary survival analyses and evaluation on whether extra breasts are conserved. In addition, we will analyze the willingness of women diagnosed with DCIS to be randomized in a watchful waiting group vs. the standard treatment group, and the impact of watchful waiting on cancer worries and quality of life. We also aim to identify morphological and molecular features in low-grade DCIS patients predicting the risk of developing an invasive ipsilateral BC and analyze if subsequent invasive lesions originate from the primary DCIS lesion, i.e. whether the DCIS lesion was likely to be the precursor lesion. Plan of investigation: In a non-inferiority, prospective, randomized clinical trial, 1842 women (49 years) will be included with asymptomatic, pure, low-grade DCIS, based on vacuum-assisted biopsies of microcal- cifications as detected by population-based or opportunistic screening, without prior breast cancer. Agreement between imaging and pathology is required. Before randomization, all eligible women will be invited to complete a baseline-questionnaire including questions on the willingness to be randomized, perceived breast cancer risk, worries about cancer and quality of life. Secondary end-points are overall survival, breast cancer- specific survival, more extensive vs. recurrent DCIS in the ipsilateral breast in the watchful waiting arm vs. the standard treatment arm respectively, number of breasts conserved, and patient reported outcomes (e.g. cancer worries, quality of life, etc.). In this trial we consider watchful waiting to be non-inferior to the standard arm, if the observed 5-year iiBCF in the watchful waiting arm at 5 years does not exceed 6.5%. This is based on the comparable risk on iiBCF for LCIS, a risk considered to be sufficiently low to allow watchful waiting. An interim analysis for futility will be performed when half of the events have occurred. Outcome will be evaluated by Kaplan–Meier statistics, calculation of the size of treatment difference, and analysis of the influence of patient and disease characteristics on the hazard of an ipsilateral invasive breast cancer using Cox proportional hazards regression. Expected results and relevance for patient care: If low-grade DCIS patients can be managed by watchful waiting safely, this can be implemented into daily patient care, saving them from the potential physical and psychological harm of invasive treatment. In addition, if novel prognostic factors will be found, these may have additional value in the choice between watchful waiting and classic treatment. No conflicts of interest 403 Invited Screen detected invasive breast cancer: More favourable disease should lead to less adjuvant therapy L. Esserman 1 . 1 University of California San Francisco, Department of Surgery and Radiology, San Francisco, USA Screening will necessarily identify lesions that will not be life threatening. This includes both precancerous conditions and those that are called invasive cancer. Conditions currently labeled “cancer” include a spectrum from indolent to fast-growing lesions. Therefore our approach to cancer screening and care requires several critical shifts, including the use of new terminology for the indolent and precancerous conditions currently labeled as cancer. Conditions recognized that are precancerous or cancerous but unlikely to cause harm if left untreated should be identified and relabeled using terminology such as IDLE (indolent lesion of epithelial origin) conditions. The rationale for the shift is that indolent lesions with low malignant potential are more frequently brought to clinical attention, especially with screening. Identification of these conditions leads to overdiagnosis and, if unrecognized, possible overtreatment. To minimize that potential, we must adopt diagnostic strategies to better define and manage IDLE conditions. We also need to revise screening guidelines to lower the chance of detecting minimal risk IDLE lesions and “inconsequential cancers” with the same energy that is traditionally devoted to increasing the sensitivity of screening tests. We can use the differing approaches to cancer screening in the US and Europe to evaluate thresholds for biopsy and intervention. In the United States, changing terminology is particularly critical as it will allow physicians to shift medico- legal notions and perceived risk to reflect our evolving understanding of biology, be more judicious about what we biopsy and how we interpret the pathology, and organize studies and registries that offer observation or less invasive approaches for indolent disease. We have the opportunity, through the use of advanced diagnostic tests to change how we label lesions. We can take advantage of longitudinal tumor banks and trials, especially those with long term follow up and the absence of systemic therapy to better define ultralow conditions. As our diagnostic tools improve, we can better tailor our treatments and minimize interventions that cause harm and improve our approach to screening. Focusing on persons at higher risk for more lethal outcomes, in facilities that use tools for differentiating biologic behavior of tumors, that have screening expertise and access to tools for engaging patients in decisions, and that focus on quality improvement. Additional emphasis on avoiding harm, while assuring benefit, will improve screening and treatment of patients and will be equally effective in preventing death from cancer. No conflicts of interest
Transcript of Management of Screen Detected Cancer
Clinical Science Symposium Friday, 21 March 2014 S165
Methods: We included patients from the population-based FOCUScohort, which comprises all incident breast cancer patients aged 65 yearsor older who were diagnosed in the geographically defined ComprehensiveCancer Centre Region West in The Netherlands between January 1997 andDecember 2004. The Adjuvant! Online predicted 10-year overall survivaland cumulative recurrence rates were calculated. These were comparedwith observed 10-year overall survival and cumulative recurrence usingone-sample T-tests for the whole cohort as well as for several subgroups ofpatient, tumour and treatment characteristics. Discriminatory accuracy wastested by composing ROC-curves and calculating corresponding C-indices;calibration was tested by plotting observed versus predicted outcomes.
Results: Overall 2,012 patients were included. Predicted and observed10-year overall survival strongly differed for all patients [predicted 48.8%,observed 39.0%, difference 9.8, 95% Confidence Interval (CI) 5.9 to 13.7,p < 0.001], as well as for most subgroups. Similarly, predicted and observed10-year cumulative recurrence differed substantially (predicted 40.8%,observed 18.2%, difference −22.6, 95%CI −34.6 to −20.6, p < 0.001),and significant inequalities were observed in almost all subgroups. Thediscriminatory accuracy of Adjuvant! Online was moderate for overallsurvival (C-index 0.75) and poor for cumulative recurrence (C-index 0.56),as was the calibration (p < 0.001 for both overall survival as well asrecurrence).
Conclusion: Adjuvant! Online does not accurately predict overall survivaland recurrence in older breast cancer patients and should be interpretedwith caution. This is in sharp contrast with current guidelines andrecommendations in which use of Adjuvant! Online is recommended inolder patients. Taken into account the sharp rise of older breast cancerpatients in our ageing society, a new prediction tool specifically for olderbreast cancer patients should be developed, thereby individualizing clinicaldecision making in older breast cancer patients and optimizing outcome.
No conflicts of interest
Friday, 21 March 2014 08:45–10:15
CLINICAL SCIENCE SYMPOSIUM
Management of Screen Detected Cancer
402 InvitedScreen-detected low grade DCIS: Take it or leave it
J. Wesseling1, L. Elshof2, E. Bleiker3, N. Bijker4, E. Rutgers5. 1TheNetherlands Cancer Institute, Department of Pathology, Amsterdam,Netherlands; 2The Netherlands Cancer Institute, Division of MolecularPathology, Amsterdam, Netherlands; 3The Netherlands Cancer Institute,Department of Psychosocial Research and Epidemiology, Amsterdam,Netherlands; 4Academic Medical Center, Department of Radiotherapy,Amsterdam, Netherlands; 5The Netherlands Cancer Institute, Departmentof Surgery, Amsterdam, Netherlands
Background: The introduction of population-based breast cancer screen-ing and implementation of digital mammography has led to an increasedincidence of ductal carcinoma in situ (DCIS) without a decrease inincidence of advanced breast cancer. This suggests DCIS overdiagnosisexists. Currently, most women with low-grade DCIS are being treatedwith surgery aiming for radical margins, and, if breast-conserving therapyis given, radiotherapy forms part of this treatment. We hypothesize thatasymptomatic, low-grade DCIS can safely be managed by watchful waiting,because these lesions harbor a low risk of progression to invasive breastcancer, and if this progression occurs, this will be low-grade and hormonereceptor positive with excellent survival rates.
Purpose: We will compare watchful waiting with standard treatmentfor primary low-grade DCIS detected by microcalcifications only ina non-inferiority randomized clinical trial. Primary outcome is 5-yearipsilateral invasive breast cancer-free rate (iiBCF). Secondary outcomesare complementary survival analyses and evaluation on whether extrabreasts are conserved. In addition, we will analyze the willingness ofwomen diagnosed with DCIS to be randomized in a watchful waiting groupvs. the standard treatment group, and the impact of watchful waiting oncancer worries and quality of life. We also aim to identify morphologicaland molecular features in low-grade DCIS patients predicting the risk ofdeveloping an invasive ipsilateral BC and analyze if subsequent invasivelesions originate from the primary DCIS lesion, i.e. whether the DCIS lesionwas likely to be the precursor lesion.
Plan of investigation: In a non-inferiority, prospective, randomizedclinical trial, 1842 women (�49 years) will be included with asymptomatic,
pure, low-grade DCIS, based on vacuum-assisted biopsies of microcal-cifications as detected by population-based or opportunistic screening,without prior breast cancer. Agreement between imaging and pathologyis required. Before randomization, all eligible women will be invited tocomplete a baseline-questionnaire including questions on the willingnessto be randomized, perceived breast cancer risk, worries about cancer andquality of life. Secondary end-points are overall survival, breast cancer-specific survival, more extensive vs. recurrent DCIS in the ipsilateral breastin the watchful waiting arm vs. the standard treatment arm respectively,number of breasts conserved, and patient reported outcomes (e.g. cancerworries, quality of life, etc.). In this trial we consider watchful waiting tobe non-inferior to the standard arm, if the observed 5-year iiBCF in thewatchful waiting arm at 5 years does not exceed 6.5%. This is based onthe comparable risk on iiBCF for LCIS, a risk considered to be sufficientlylow to allow watchful waiting. An interim analysis for futility will be performedwhen half of the events have occurred. Outcome will be evaluated byKaplan–Meier statistics, calculation of the size of treatment difference,and analysis of the influence of patient and disease characteristics onthe hazard of an ipsilateral invasive breast cancer using Cox proportionalhazards regression.
Expected results and relevance for patient care: If low-gradeDCIS patients can be managed by watchful waiting safely, this canbe implemented into daily patient care, saving them from the potentialphysical and psychological harm of invasive treatment. In addition, if novelprognostic factors will be found, these may have additional value in thechoice between watchful waiting and classic treatment.
No conflicts of interest
403 InvitedScreen detected invasive breast cancer: More favourable disease
should lead to less adjuvant therapy
L. Esserman1. 1University of California San Francisco, Department ofSurgery and Radiology, San Francisco, USA
Screening will necessarily identify lesions that will not be life threatening.This includes both precancerous conditions and those that are calledinvasive cancer. Conditions currently labeled “cancer” include a spectrumfrom indolent to fast-growing lesions. Therefore our approach to cancerscreening and care requires several critical shifts, including the use of newterminology for the indolent and precancerous conditions currently labeledas cancer. Conditions recognized that are precancerous or cancerousbut unlikely to cause harm if left untreated should be identified andrelabeled using terminology such as IDLE (indolent lesion of epithelialorigin) conditions. The rationale for the shift is that indolent lesionswith low malignant potential are more frequently brought to clinicalattention, especially with screening. Identification of these conditionsleads to overdiagnosis and, if unrecognized, possible overtreatment. Tominimize that potential, we must adopt diagnostic strategies to betterdefine and manage IDLE conditions. We also need to revise screeningguidelines to lower the chance of detecting minimal risk IDLE lesionsand “inconsequential cancers” with the same energy that is traditionallydevoted to increasing the sensitivity of screening tests. We can use thediffering approaches to cancer screening in the US and Europe to evaluatethresholds for biopsy and intervention. In the United States, changingterminology is particularly critical as it will allow physicians to shift medico-legal notions and perceived risk to reflect our evolving understanding ofbiology, be more judicious about what we biopsy and how we interpret thepathology, and organize studies and registries that offer observation or lessinvasive approaches for indolent disease.We have the opportunity, through the use of advanced diagnostic tests
to change how we label lesions. We can take advantage of longitudinaltumor banks and trials, especially those with long term follow up and theabsence of systemic therapy to better define ultralow conditions. As ourdiagnostic tools improve, we can better tailor our treatments and minimizeinterventions that cause harm and improve our approach to screening.Focusing on persons at higher risk for more lethal outcomes, in facilities thatuse tools for differentiating biologic behavior of tumors, that have screeningexpertise and access to tools for engaging patients in decisions, and thatfocus on quality improvement. Additional emphasis on avoiding harm, whileassuring benefit, will improve screening and treatment of patients and willbe equally effective in preventing death from cancer.
No conflicts of interest
S166 Friday, 21 March 2014 Clinical Science Symposium
404 InvitedDoes every screen detected cancer patient need radiotherapy?
B. Chua1. 1Peter MacCallum Cancer Institute, Radiation Oncology,Melbourne, Australia
Radiotherapy (RT) after breast conserving surgery (BCS) for earlybreast cancer is considered standard of care to optimise local controland decrease breast cancer mortality. However, in conjunction withearly detection enabled by mammographic screening, the falling localrecurrence (LR) rates due to advances in multidisciplinary breast cancermanagement underpin the significance of tailoring RT utilisation accordingto individual risks of LR.Randomised trials have been conducted to evaluate the outcomes of
selected patients with node-negative breast cancer treated with tamoxifenplus RT versus tamoxifen only after BCS. Hughes et al. addressed thisquestion in women aged �70 years with hormone receptor (HR)-positivetumour �2 cm [1,2].
Number Age Median follow-up Local recurrence rate (%)
(years) Tamoxifen Tamoxifen+RT
Hughes et al, [1,2] 636 �70 10.5 8.0 2.0Blamey et al. [3] 1172 <70 10.2 4.8 1.1Fyles et al. [4] 769 >50 5.6 7.7 0.6Fisher et al. [5] 1009 >18 8.0 16.5 2.8Valagussa et al. [6] 749 55−75 4.4 2.5 0.7
A meta-analysis showed that a 20% absolute reduction in 5-year LRrate was associated with a 5% absolute reduction in 15-year breast cancermortality [7]. It reported that a decrease in 10-year recurrence risk of 9%was associated with a 0.1% reduction in 15-year breast cancer mortality.The implication is that RT confers no appreciable benefit in breast cancermortality in patients for whom RT decreases the absolute LR risk by <10%.The challenge lies in optimising the prognostic precision for LR risk
to inform selective omission of RT to avoid over-treatment of low-riskpatients. Breast cancer can no longer be considered and treated asa single disease entity [8,9]. Classification based on gene expressionprofiles has shown that each subtype is associated with distinct clinicaloutcomes. Luminal-A tumours (HR-positive, HER2-negative, Ki-67 <14%)are shown to be associated with the lowest LR rate compared tothe other subtypes [10,11]. A study of gene expression profiling usingOncotype DX Recurrence Score (RS) in patients with HR-positive, node-negative breast cancer demonstrated a significant association between RSand LR rates [12]. However, these data are limited by the retrospectivestudy design and should be considered exploratory, warranting prospectivestudies to address these important limitations before implementation inclinical practice.
References
[1] Hughes KS, Schnaper LA, Berry D, et al. Lumpectomy plus tamoxifenwith or without irradiation in women 70 years of age or older with earlybreast cancer. N Engl J Med 2004; 351: 971−7.
[2] Hughes KS, Schnaper LA, Cirrincione C, et al. Lumpectomy plustamoxifen with or without irradiation in women 70 years of age or olderwith early breast cancer. J Clin Oncol 2010; 28: 69s.
[3] Blamey RW, Chetty U, Bagtes T, et al. Radiotherapy and/or tamoxifenafter conserving surgery for breast cancers of excellent prognosis:BASO II trial. Eur J Cancer 2008; S6: 55.
[4] Fyles AW, McCready DR, Manchul DA, et al. Tamoxifen with or withoutbreast irradiation in women 50 years of age or older with early breastcancer. N Engl J Med 2004; 351: 1021−3.
[5] Fisher B, Bryant J, Dignam JJ, et al. Tamoxifen, radiation therapy,or both for prevention of ipsilateral breast tumour recurrence afterlumpectomy in women with invasive breast cancers of one centimetreor less. J Clin Oncol 2002; 20: 4141−9.
[6] Valagussa P, Costa A, Tinterri C, et al. Breast conservative surgerywith and without radiotherapy in patients aged 55−75 with early stagebreast cancer. Eur J Cancer 2008; S6: 47.
[7] Early Breast Cancer Trialists Collaborative Group. Effects of radiother-apy and of differences in the extent of surgery for early breast cancer onlocal recurrence and 15-year survival. Lancet 2005; 366: 2087–2106.
[8] Perou CM, Sorlie T, Elsen MB, et al. Molecular portraits of human breasttumours. Nature 2000; 17: 747−52.
[9] Sorlie T, Perou CM, Tibshirani R, et al. Gene expression patternsof breast carcinomas distinguish tumour subclasses with clinicalimplications. Proc Natl Acad Sci USA 2001; 98: 10869−74.
[10] Voduc D, Cheang MCU, Nielsen TO, et al. Breast cancer subtypes andthe risk of local and regional relapse. J Clin Oncol 2010; 28: 1684−91.
[11] Nguyen P, Taghian A, Katz M, et al. Breast cancer subtypeapproximated by ER, PR and HER-2 is associated with local anddistant recurrence after breast-conserving therapy. J Clin Oncol 2008;26: 2373−8.
[12] Mamounas EP, Tang G, Fisher B, et al. Association between the 21-gene recurrence score assay and risk of locoregional recurrence innode-negative, oestrogen receptor-positive breast cancer: Results fromNSABP B-14 and NSABP B-20. J Clin Oncol 2010; 28: 1677−83.
No conflicts of interest
405 Proffered paper oralAdverse outcomes in the treatment of ductal carcinoma in situ in the
UK NHS Breast Screening Programme 2003−2012
J. Thomas1. 1Lothian University Hospitals Trust on behalf of The SloaneProject Steering Group, Pathology, Edinburgh, United Kingdom
Background: Concerns have been raised about the efficacy of breastscreening programmes particularly the overdiagnosis of malignancy andpotential overtreatment of patients. The Sloane Project, with over 12,000registered UK patients between 2003 and 2012, is the largest prospectiveaudit of ductal carcinoma in situ (DCIS) worldwide. We have analysedsurgical outcomes in relation to pathological and radiological features andhospital of treatment.
Methods: Multi-disciplinary data from 8,313 screen-detected patientswith DCIS were analysed with a focus on operation type, success ofbreast conservation surgery (BCS) and mastectomy rate for small lesions.Particular attention was paid to the agreement between radiological andpathological lesion-size estimation and key pathological parameters.
Results: Of 6,633 patients embarking on BCS, 799 (12.0%) requiredmastectomy. Failed BCS accounted for one third of all mastectomy patientsand was associated with significant radiological under-estimation of diseaseextent. 510 (21%) mastectomies were carried out for lesions <20mm andthese patients had significant radiological over-estimation of disease extent.Specimen slice radiography to assist tissue sampling had no significantimpact on the radiological-pathological size mismatch. These two operativesubgroups accounted for 49% of all mastectomies with considerableand significant inter-hospital variation in the failed BCS outcome (range3−32% of a hospital’s BCS workload) and a non-significant variation butwide range (0−60%) in the mastectomy rate for small lesions in differenthospitals. Failed BCS had a significant impact on a hospital’s overall DCISmastectomy rate. There was no consistent association between theseoutcomes and a range of pathological variables.
Conclusions: This study shows that large numbers of patientsundergoing mastectomy for DCIS do so either as a result of failed BCSor for small lesions and further identifies substantial differences in these
Table (abstract 405).
BCS Mastectomy
1 Op 2 Ops followed bymastectomy
Primary For tumours�20mm
For tumours<20mm
Number a 3946 1435 799 1680 1969 510Median pathology size (mm) 12 17 40 35 42 1295%CI (mm) 12−12 16−18 36−40 32−37 40−45 12−13Inter Quartile Range (mm) 13.0 15.00 30.00 30.00 30.00 8.00Altman Bland Bias (mm) b 2.95 (2.90) −0.32 (0.01) −13.49 (−11.43) 9.9 (9.82) −2.29 (−3.97) 19.86 (18.8)% High grade 58.5 59.3 70.2 77.5 78.1 63.9% ER Positive 83.2 79.9 72.8 69.1 69.3 73.7Slice radiography (%) 21.0 20.3 21.5 15.1 16.9 18.4
a Excludes therapeutic mammoplasty − 47 patients, diagnostic biopsy 349 patients, successful BCS after more than two operations 57 patients.b Data in parentheses refer to specimens with slice radiography.
Clinical Science Symposium Friday, 21 March 2014 S167
outcomes in different hospitals. Improved pre-operative multidisciplinaryassessment of disease extent can ensure that patients are offered themost appropriate surgical treatment thereby reducing rates of failed BCSand mastectomy for small lesions. We provide an evidence base forbenchmarking practice in the future.
No conflicts of interest
406 Proffered paper oralBreast cancer screening in older women
N.A. de Glas1, E. Bastiaannet1, M. Kiderlen1, W. van de Water1,A.J.M. de Craen2, S. Siesling3, H.M. Schuttevaer4, G.H. de Bock5,C.J.H. van de Velde1, G.J. Liefers1. 1Leiden University Medical Center,Surgery, Leiden, Netherlands; 2Leiden University Medical Center,Gerontology & Geriatrics, Leiden, Netherlands; 3Comprehensive CancerCentre, Research, Utrecht, Netherlands; 4Rijnland Hospital, Radiology,Leiderdorp, Netherlands; 5University of Groningen University MedicalCenter Groningen, Epidemiology, Groningen, Netherlands
Background: In upcoming decades, an increasing proportion of breastcancer patients will be elderly. It has been assumed that diagnosis atan earlier stage through screening programs could improve prognosis.However, elderly may be at risk for over diagnosis due to screeningprograms, and consequently unnecessarily at risk for possible harmfuleffects of cancer treatment. In The Netherlands, the upper age limit of thescreening program was extended from 69 to 75 years in 1998. However,it remains unclear whether the mass screening program has a beneficialeffect in women aged 70 years and older.If a screening program is effective, it can be expected that the incidence
of early stage breast cancer increases, while the incidence of advancedstage cancers decreases. According to several studies, this is the mostappropriate method to investigate the efficacy of a screening program inpopulation-based data, as studying mortality rates as an indicator for theeffect of screening programs can lead to several forms of bias. Therefore,we investigated the effect of the implementation of the screening programon the stage distribution of incident breast cancer in women aged 70−75years in the Netherlands.
Materials and Methods: The Netherlands Cancer Registry was used toinclude all patients aged 70−75 years who were diagnosed between 1995and 2011 with invasive or in situ breast cancer. Time trends of incidencerates of different tumor stages were analyzed in linear regression analyseswith the incidence rate of both early stage (0, I and II) and advanced stage(III and IV) breast cancer as the outcome, and year of diagnosis as theindependent variable.
Results: Overall, we included 25,414 patients aged 70−75 years atdiagnosis. The incidence of early stage tumors significantly increased afterextension of the upper age limit to 75 years in 1998 (260 cases per100,000 women in 1995 up to 382 cases per 100,000 women in 2011,p for trend = 0.03), while the number of advanced stage breast cancersdid not significantly change (59 cases per 100,000 women in 1995 to53 cases per 100,000 women in 2011, p for trend = 0.2).
Conclusions: The extension of the upper age limit to 75 years hasnot led to a decrease of advanced stage breast cancer, while the numberof early stage tumors strongly increased. This implies that the effect ofscreening in elderly women is limited and leads to a large proportion of overdiagnosis. Until new studies in this specific group have been performed, wepropose that the decision to participate in the screening program shouldbe personalized based on remaining life expectancy, functional status andpatients’ preferences.
No conflicts of interest
Friday, 21 March 2014 08:45–10:00
CLINICAL SCIENCE SYMPOSIUM
New Insights into Epidemiology and
Outcomes
407 InvitedEpidemiology of breast cancer 2013
P. Boyle1. 1International Prevention Research Institute and University ofStrathclyde Institute of Global Public Health at iPRI, President, Lyon, France
Breast cancer remains a major and growing threat to global Public Health:the global burden currently exceeds 1.6 million new cases annually andhas been increasing at an annual rate of 3.1 per cent.
The World Breast Cancer Report captures the global nature of breastcancer; the rapid increase in incidence; disparities in awareness andresponse to symptoms; disparities in the availability of diagnostic facilities,treatment options and availability; and consequent major disparities inbreast cancer outcome. Reports from lower income countries highlightthe late presentation of women with breast cancer to treatment providers,frequently when palliation is the only treatment option. Lower resourcecountries lack adequate treatment facilities and when treatment is possible,the range of options is limited and, frequently, not state of the art. Thelarge proportion of women presenting with metastatic disease is muchhigher in lower resource countries than in higher resource countries. Manylower resource countries lack significant palliation resources, includingan absence of radiotherapy and no availability of opioid medication.Paracetamol is not an effective treatment for the pain frequently associatedwith terminal cancer.Prospects for primary prevention of breast cancer remain limited. Many
risk factors are not easily modifiable. Moderating alcohol consumption,avoiding overweight and increasing physical activity would lead todecreases in risk as would an increase in duration of breast feeding. Withthe global tendency to breast feed each child for a shorter period and tohave fewer children, the major benefit of this lifestyle choice seems limited.Despite more new cases being diagnosed in the developed countries,
the 450,000 annual breast cancer deaths are equally divided betweenthe developing and developed world. Increasing population size andlife expectancy, accompanied by a decrease in stigma, an increase inawareness and the introduction of effective early detection programmesin lower resource countries will lead to an increase in the numbers of newcases diagnosed.Global priorities must be to reduce the number of women developing
and dying from breast cancer and to eliminate the major disparities whichcurrently exist. Strategies for primary prevention remain in the researchdomain while continual progress in the treatment of breast cancer isadvancing slowly but surely. Epidemiology must adapt quickly to this newsituation.
No conflicts of interest
408 InvitedEvolving risk factors for breast cancer occurrence and death
P. Autier1. 1International Prevention Research Institute and Universityof Strathclyde Institute for Global Public Health at iPRI, Research, Lyon,France
Background: Until 2000, most studies or reviews on risk factors for breastcancer used to lump together all cancers without distinguishing their highlyvariable aggressiveness, and rarely addressed risk factors for breast cancerdeath.
Methods: We systematically searched for publications done after 2000on risk factors for breast cancer and compared factors associated withoccurrence, with death, poor survival and with occurrence of aggressivecancer (e.g., ER−/PR− or TNBC).
Results: Reproductive factors are the main risk factors for breast canceroccurrence, but they have little influence on the risk to die from breastcancer. Adiposity and diabetes are risk factors for the occurrence of moreaggressive breast cancer, and are also risk factors for breast cancer death.In premenopausal women, if adiposity seems associated with reducedrisk of breast cancer occurrence, it is associated with higher risk ofbreast cancer death. Physical activity reduces breast cancer occurrence,especially of aggressive phenotypes, and improves survival of breastcancer patients. Women giving birth in their 40s have become increasinglycommon, and breast cancer occurring in the first 2 years after childbirthis known to be more lethal. Obesity and diabetes would influence breastcancer occurrence and phenotype through low grade chronic systemicinflammation. This hypothesis is supported by other observations: physicalactivity reduces systemic inflammation; inflammation induces low vitaminD concentrations that are also associated with more aggressive disease;long-term aspirin use reduces the risk of breast cancer. Breast canceraggressiveness after giving birth would be due to angiogenic factorsreleased by the placenta. Breast cancer is associated with use of hormonereplacement therapy but this use would have no impact on the risk ofbreast cancer death. High breast density is associated with breast canceroccurrence, especially of more aggressive phenotype, but is not a predictorof survival once usual histological factors are taken into account. Obesity,diabetes and giving birth at older age are increasing in most populations,which may explain recent increases in the incidence of late stage breastcancer in women less than 50.
Conclusion: The success of primary, secondary (screening) or tertiary(treatment) prevention may depend on different characteristics of women,and personalized counseling needs to target variable factors depending onthe outcome to be prevented.
No conflicts of interest
