56

JOURNAL OF PALLIATIVE MEDICINEVolume 10, Number 1, 2007© Mary Ann Liebert, Inc.DOI: 10.1089/jpm.2006.0117

Incidence of Weak Opioids Adverse Events in the Management of Cancer Pain: A Double-Blind Comparative Trial

RENE FERNANDO RODRIGUEZ, M.D., M.Sc.,1 LUIS EDUARDO BRAVO, M.D., M.Sc.,2FERNANDO CASTRO, M.D.,3 OLGA MONTOYA, M.D.,4

JAVIER MAURICIO CASTILLO, M.D.,1 MARÍA PILAR CASTILLO, M.D.,1PAOLA DAZA, M.D.,1 JOSÉ MANUEL RESTREPO, M.D.,1

and MARIO FERNANDO RODRIGUEZ, M.D.1

ABSTRACT

With the objective of comparing incidence of adverse events of the opioids codeine, hy-drocodone, and tramadol in the relief of cancer pain, we conducted a randomized controlledtrial in which patients with cancer were randomly assigned according to a computer-gener-ated schedule to receive one of the three opioids. Of the 177 patients who participated, 62 pa-tients received hydrocodone, 59 patients received codeine, and 56 patients received tramadol.The pain experienced by the participants originated most frequently from the stomach, breast,or prostate gland and was classified as either somatic (33%), visceral (52%), mixed (6%), orneuropathic (9%). At the first visit, 60% of the patients described their pain intensity as mod-erate (4–6/10), with the remaining 40% of the patients describing their pain as severe (7–10/10).The symptoms most associated with pain were weakness, insomnia. and anorexia. In 77% ofthe total number of cases, the patient was aware of his/her diagnosis prior to admittance tothe palliative care unit. Of the total number of cases, 57% fell in the age range of 60–89 yearsold and 50% of the participants were female. No significant statistical difference in the anal-gesic efficacy of the three opioids was found (p: 0.69; �2: 0.73). Use of tramadol producedhigher rates of adverse events than codeine and hydrocodone: vomiting, dizziness, loss of ap-petite, and weakness (p � 0.05).

INTRODUCTION

PAIN IS ONE OF THE MOST FEARED CONSEQUENCES

of malignancy and often adversely affects thequality of life of patients and their families.1 Dur-ing their illness, approximately 60%–90% of pa-tients with cancer will suffer pain; the goal of

analgesic therapy in patients with cancer is toachieve continuous relief of pain.

In 1982 the World Health Organization(WHO) endorsed the WHO Cancer Pain ReliefProgram. According to this program, the strate-gies for the treatment of cancer pain are basedon a three-step analgesic ladder. WHO guide-

1Department of Palliative Medicine, Universidad Libre, Seccional Cali, Cali, Colombia.2Department of Pathology, Universidad del Valle, Cali, Colombia.3Department of Palliative Medicine, Instituto del Seguro Social, Cali, Colombia.4Department of Palliative Medicine, Hospital General de Medellín, Medellín, Colombia.

lines for the treatment of cancer pain recom-mend weak opioids analgesics combined withnonopioid analgesics as second-line therapy formoderate pain.2

The analgesic effects of codeine are mediatedthrough inhibition of opioid receptors and can besynergistically enhanced by the addition of ananalgesic having a different mode of action.3–5

One possible substance is acetaminophen. Itsanalgesic effects are directly related to its abilityto inhibit lipooxygenase and cyclooxygenase, re-sulting in decreased prostaglandin and inter-leukin-1 synthesis in the hypothalamus (factorsboth important for the transmission of pain).Acetaminophen also acts by increasing the re-lease of endogenous opioids in the periaqueduc-tal grey matter, thereby inhibiting descendingpain pathways.6,7 Hydrocodone is an analgesicagent chemically related to codeine. Tramadolhas a double mechanism of action: it not onlybinds the parent compound and its M1 metabo-lite to �-opioid receptors, but it also weakly in-hibits the reuptake of norepinephrine and sero-tonin.8–13

The incidence and severity of side effects fromthe administration of opioids can play an impor-tant role in the success or failure of pain man-agement in patients with cancer pain. In clinicalpractice, opioid pharmacotherapy for cancer painhinges on finding “a satisfactory balance betweenanalgesia and side effects.” Side effects per se re-duce quality of life and further impair quality oflife by limiting the dosage of opioid analgesicsthat patients can tolerate. The most frequently re-ported adverse events in several studies duringthe treatment of cancer pain with weak opioidsanalgesics were constipation, nausea, vomiting,dizziness, dry mouth, headache, somnolence, as-thenia, pruritus, sweating, and periods of mentalconfusion.14–18

Successful opioid therapy requires that thebenefits of analgesia outweigh the risk of side ef-fects. Understanding the incidence, severity, andmechanisms of side effects in all forms of paintreatment helps the clinician frame an optimalmanagement plan.

The aim of this study was to evaluate the inci-dence of side effects related to therapy withcodeine plus acetaminophen, hydrocodone plusacetaminophen, and tramadol clorhydrate in pa-tients with cancer pain, using an evaluationmodel with multidose administration in a dou-ble-blinded comparative trial.

MATERIALS AND METHODS

Patients

One hundred seventy-seven patients in need ofanalgesic medication because of persistent mod-erate or severe cancer-associated pain partici-pated in this double-blinded, randomized, com-parative trial. The study followed a commonprotocol approved by the investigative reviewboard and ethics committee at each site, and in-formed written consent to participate was ob-tained from all patients.

Exclusion criteria included: lack of ability tocommunicate; history of alcohol and/or drugabuse; use of any study analgesic during the 3weeks leading up to the commencement of thestudy; any known allergy, sensitivity or con-traindication to a study medication; kidney fail-ure (creatinine � 2 mg/dL); bilirubin � 2 mg; patients having surgery, chemotherapy, or ra-diotherapy within 23 days of initiation of thestudy; patients currently having chemotherapy orradiotherapy; Karnofsky � 50/100; and the pres-ence of a second pain more severe than and un-related to the cancer pain.

Study design

The study was conducted at three Colombiancenters: Universidad Libre Seccional Cali, Insti-tuto del Seguro Social de Cali, and Hospital Gen-eral de MedellÌn. The patients were randomly assigned to one of three groups, according to acomputer-generated schedule, to receive codeine/acetaminophen, hydrocodone/acetaminophen,or tramadol clorhydrate. Thereafter, the patientsrandomly received 150/2500 mg/d of codeine/acetaminophen; 25/2500 mg/d of hydrocodone/acetaminophen; or 200 mg/d of tramadol clorhy-drate. Drug distribution was blinded, with thedrugs packaged in identical containers. Each con-tainer had the medicinal agent in the appropriatenumber of tablets for the first 2 days of the studyand for the following 21 days (three 7-day peri-ods). If there was no pain relief in any control (vi-sual analogue score [VAS] �4), the doses weredoubled. If any overdose side effects were noticedupon doubling the dose, dosage was reduced by25%.

Patients discontinued all other analgesics for atleast 8 hours prior to study entry, but were al-lowed to continue taking medication for neuro-pathic pain, such as antidepressants or anticon-

OPIOIDS ADVERSE EVENTS IN CANCER PAIN 57

vulsants, if such treatment had already beenstarted. However, new treatment was not al-lowed to be initiated during the course of thestudy.

Because constipation is a very common collat-eral effect of opioids, every patient received thesame prophylactic laxatives for the duration ofthe study (bysacodil 5 mg/d). Prescriptions forother medications (e.g., antiemetics), however,were given only after the appearance of symp-toms.

During the course of the study, the patientswere given the telephone number of the princi-pal investigator in case any misunderstanding ofthe study procedure should arise, should any se-rious adverse reactions to the medications occur,or should the intensity of pain increase in inten-sity or fail to diminish.

Evaluation of adverse effects

The patients were asked to rate their adverseeffects 2 days after commencement of the studyand then weekly for 3 weeks. Pain intensity wasmarked on a 100-mm horizontal VAS on whichthe end points were defined as “no pain” and“very severe pain.” Patients were instructed tonote all adverse events believed to be drug re-lated. Adverse events that emerged or worsenedafter initiation of drug therapy were judged to betreatment related. Adverse events were reportedon a verbal rating scale (VRS) (0 � absence; 1 �mild; 2 � moderate; 3 � severe). Additional vari-ables obtained were Karnofsky performance sta-tus and demographic variables.

Patients would be withdrawn from the studyin the following situations. Severe adverse reac-tions to treatment, treatment failure, significantprotocol violation, or the beginning of radiother-apy, chemotherapy or surgery. Additionally, pa-tients were free to withdraw at any time at theirown request.

Medical history and physical examinationwere performed at the participant’s screeningvisit and at his/her final visit of the double-blinded phase or at the time of premature dis-continuation.

Statistics

This study was designed with 80% power atthe 5% significance level. All p values are two-tailed. Statistical comparisons were made by useof either the �2 statistic, analysis of variance(ANOVA), or the Student’s t test for grouped or

paired data. Categorical variables, incidence ofadverse effects and proportion showing at leastsome improvement were compared betweengroups using the �2 statistic.

Data are presented as mean � standard devia-tion (SD) with the 95% confidence interval given.

Power analysis indicate that a sample size of atleast 56 individuals per group would yield an80% chance (at � � 0.05) of detecting a differencebetween groups. Statistics were performed by us-ing commercially available software (Stata 8.0,Stata Corp., College Station, TX)

RESULTS

Recruitment of participants was conducted fromJanuary 2000 through October 2005. A total of 177patients were randomly assigned to one of threetreatment groups (62 to hydrocodone, 59 to codeine,and 56 to tramadol). There were no meaningful dif-ferences among the groups upon entry.

The principle cancer diagnoses were gastric(16%), breast (14%), prostate (11%), and lung(10%).

The pain intensity in the first appointment wasmoderate (averager 4–6/10) in 60% of the patientsand severe (average 7–10/10) in 40%, with 91 pa-tients (51%) reporting visceral pain, 59 patients(33%) reporting somatic pain, 11 patients (6%) re-porting both visceral and somatic pain, and 16 pa-tients (9%) reporting neuropathic pain.

In 75% of the total number of cases the patientsand their relatives knew the diagnosis when theyarrived to the palliative care unit. The symptomsmost associated with pain were weakness, in-somnia and loss of appetite. Table 1 presents asocio-demographic and health status profile ofstudy participants.

Sixty-six percent of patients who receivedcodeine, 73% using tramadol, and 71% using hy-drocodone experienced pain relief. The differ-ences in pain relief were not significant amongthe three groups.

The most frequently observed adverse reac-tions are presented in Table 2.

DISCUSSION

One of the aims of palliative care is to improvethe quality of life of the patient and to decreasehis/her symptom intensity. In Colombia, careand support shown by family members plays an

RF RODRIGUEZ ET AL.58

essential role in palliation, with a great numberof patients choosing to die in their homes sur-rounded by their relatives rather than in a hos-pital. Patients in Colombia also have the right toknow their medical diagnosis, decreasing the in-clination of family members to keep the patientin the dark regarding his/her health.

This study showed that the analgesic effects ofthe analyzed drugs are similar, with some im-portant differences in their collateral effects: vom-iting, dizziness, loss of appetite, and weakness.

We consider therefore that in choosing whichdrug to use (providing all three drugs are avail-able), the physician must consider two importantfactors: firstly, the antecedents of efficacy and ad-verse effects of the drugs and secondly, the costof the drugs.

Rates of fatigue, nausea, vomiting, and consti-pation were unlike those found in young healthyindividuals in the United States and Germany,three decades apart: fatigue, 37% to 65%; nausea,1% to 3%; vomiting, 0% to 2%; and constipation,

OPIOIDS ADVERSE EVENTS IN CANCER PAIN 59

TABLE 1. BASELINE SOCIODEMOGRAPHIC AND CLINICAL CHARACTERISTICS

Codeine Tramadol HydrocodoneBaseline characteristics (n � 59) (n � 56) (n � 62)

Age, mean � SD, yr 61.5 � 12.0 57.1 � 14.3 62.0 � 13.0Gender (%)

Female 54.2 62.5 35.5Site (%)

Ambulatory 91.5 91.1 87.1Numerical rating scale pain intensity 7.2 � 1.3 7.5 � 1.6 7.3 � 1.4

(0–10) mean � SD,Pain intensity (%)

Moderate: 4–6/10 61 54 66Severe: 7–10/10 39 46 34

Duration of pain (days) mean � SD 102.6–144.9 117.3–203.2 107.1–117.3Type of pain (%)

Somatic 33.9 35.7 30.7Visceral 52.5 48.2 53.2Both 5.1 3.6 9.7Neuropathic 8.5 12.5 6.5

Symptoms (%)Nausea 40.7 39.3 29.0Vomiting 23.7 16.1 9.7Constipation 44.1 42.9 35.5Anorexia 69.5 73.2 58.1Weakness 89.9 76.8 75.8Dry mouth 66.1 64.3 62.9

SD, standard deviation.

TABLE 2. PROPORTION OF SUBJECTS WITH SIDE EFFECTS THROUGHOUT FOLLOW-UP

Codeine Tramadol Hydrocodone Total(%) (%) (%) (%) p

Nausea 28.81 46.43 27.42 33.90 0.056Vomiting 23.73 35.71 16.13 24.86 0.047Constipation 35.6 25.0 29.03 29.94 0.45Diarrhea 0.00 1.79 4.84 2.26 0.19Loss of appetite 1.69 21.43 6.56 9.66 0.001Weakness 0.00 12.50 1.61 4.52 0.002Dry mouth 15.25 21.43 17.74 18.08 0.68Somnolence 35.59 46.43 37.10 39.55 0.438Dizziness 23.73 41.07 19.35 27.68 0.022Insomnia 0.00 3.57 0.00 1.13 0.112Urinary retention 0.00 3.57 1.61 1.69 0.332Pruritus 11.86 19.64 19.35 16.95 0.443Sweaty 8.47 12.50 14.5 11.86 0.58Hallucinations 6.78 7.14 11.29 8.47 0.613

3% to 4%17,18; rates of dry mouth (3% to 5%), weresimlar to our rates.

A large systematic review of constipationprevalence studies in North America19 foundrates between approximately 2% and 27%, butmostly approximately 15%, considerably higherthan the 5% found for placebo in the clinical tri-als, and the same as for patients receiving opioidsfor chronic nonmalignant pain.

Future trials should consider a number of cru-cial variables, including dose. The higher fre-quency of adverse events with opioids used totreat severe pain20 than with opioids more oftenused to treat moderate pain found by McQuay etal.21 but with similar rank order, may be ex-plained by higher dose equivalents with the for-mer. Duration is also important. Patients startingon opioids are usually told to expect initial ad-verse events, such as nausea and drowsiness, butthat these will improve rapidly. Judging the ad-verse event profile of long-term opioide use fromshort-term trials is unlikely to be satisfactory forextrapolation to longer-term use in clinical prac-tice.

ACKNOWLEDGMENTS

The study was supported by a research grantfrom the Universidad Libre Seccional Cali. Drugswere supplied free of cost from Grunenthal, Li-brapharma and Sanofi Synthelabo.

REFERENCES

1. Levy MH: Pain management in advanced cancer.Semin Oncol 1985;19:394–410.

2. Foley IGM: The treatment of cancer pain. N Engl JMed 1985;313:84-95.

3. Stein C: Peripheral mechanisms of opioid analgesia.Anestth Analg 1993;76:182.

4. Stein C, Yassouridis A: Peripheral morphine analge-sia. Pain 1997;71:119–121.

5. Miralles FP, González JM, Vaca JM: Investigación enel dolor experimental, medición y valoración del do-lor. In: Torres LM (ed): Medicina del Dolor. Barceona:Editorial Masson SA, 1997, pp. 19–33.

6. Murillo EC: Calidad de vida del enfermo neoplásico.In: González MB (ed): Tratado de medicina paliativa ytratamiento de soporte en el enfermo con cáncer. Madrid:Panamericana, 1996, pp. 75–76.

7. Rodríguez RF: Manejo farmacológico del dolor en pa-cientes con cáncer. In: Medicina del dolor y cuidados

paliativos. Edit. Universidad Libre de Cali. 1998, pp.173–182.

8. Acetaminophen. In: Physicians’ Desk Reference, 55th ed.Montvale, NJ: Medical Economics Co; 2001, pp.1832–1833

9. Ruoff GE: Slowing the initial titration rate of tramadolimproves tolerability. Pharmacoterapy 1999;24:115–123

10. Schnitzer TJ, Gray WL, Paster RZ, Kamin M: Efficacyof tramadol in treatment of chronic pain. J Rheuma-tol 2000;6:250–257

11. Fricke JR, Karim R, Jordan D, Rosenthal N: A doubleblind, single-dose comparison of the analgesic efficacyof tramadol/acetamonophen combination tablets, hy-drocodone/acetaminophen combination tablets andplacebo after oral surgery. Clin Ther 2002;24:953–968.

12. Nicholson B: Responsible prescribing of opioids for themanagement of chronic pain. Drugs 2003;63:17–32.

13. Przewlocki R, Przewlocka B: Opioids in chronic pain.Eur J Pharmacol 2001;429:79–91.

14. Jacox AlL Carr DB, Payne R: Management of cancerpain: clinical practice guideline no. 9. Agency forHealth Care Policy and Research (AHCPR), publica-tion no. 94-0592. Rockville, MD: AI-ICPR, PublicHealth Service, 1994.

15. Zenz M, Strumpf M, Tryba M: Long-term oral opioidtherapy in patients with clironic nonnmlignant pain.J Pain Symptom Manage 1992;7:69–77.

16. Ventafridda V, Olivieri E, Caraceni A, Spoldi E, DeConno F, Saita L, Ripamonti G: A retrospective studyon the use of oral morphine in cancer pain. J PainSymptom Manage 1987;2:77–81.

17. Reidenberg MM, Lowenthal DT: Adverse nondrug reactions. N Engl J Med 1968;279:678–679.

18. Meyer FP, Troger U, Rohl FW: Adverse nondrug re-actions: An update. Clin Pharmacol Ther 1996;60:347–352.

19. Higgins PD, Johanson JF: Epidemiology of constipa-tion in North America: A systematic review. Am JGastroenterol 2004;99:750–759.

20. Kalso E, Edwards JE, Moore RA, McQuay HJ: Opi-oids in chronic non-cancer pain: Systematic review ofefficacy and safety. Pain 2004;112:372–380.

21. Moore RA, McQuay H: Prevalence of opioid adverseevents in chronic non-malignant pain: Systematic re-view of randomised trials of oral opioids. Arthitis ResTher 2005;7:1046–1051.

Address reprint requests to:Rene Fernando Rodriguez, M.D.

Department of Palliative MedicineUniversidad Libre

Calle 59 Norte No 3E63Paseo de la Flora, Las Acacias Casa 3

Cali, ValleColombia

E-mail: renerodriguezmd@hotmail.com

RF RODRIGUEZ ET AL.60