Quality of Life and Cancer Pain: Satisfaction and Side
`Effects With Transdermal Fentanyl Versus Oral Morphine
`
`By Richard Payne, Susan D. Mathias, David J. Pasta, Lee A. Wanke, Rhys Williams, and Ramy Mahmoud
`
`Purpose: To compare pain-related treatment satisfac-
`tion, patient-perceived side effects, functioning, and
`well-being in patients with advanced cancer who were
`receiving either transdermal fentanyl (Duragesic, Jans-
`sen Pharmaceuticals, Titusville, NJ) or sustained-release
`oral forms of morphine (MS Contin, Perdue Frederick
`Co, Norwalk, CT, or Oramorph SR, Roxanne Laborato-
`ries, Columbus, OH).
`Patients and Methods: Atotal of 504 assessable can-
`cer patients participated in this cross-sectional, quality-
`of-life study. Relevant elements of four validated scales
`were used-the Functional Assessment of Cancer
`Therapy-General (FACT-G) scale, the Brief Pain Inven-
`tory (BPI), the Medical Outcomes Study (MOS) question-
`naire, and the Memorial Symptom Assessment Scale
`(MSAS)-as well as original scales that were developed
`and validated for this study.
`Results: The majority of patients in both treatment
`groups had late-stage (IV/D) cancer. Patients who re-
`ceived transdermal fentanyl were more satisfied over-
`
`F OR MORE THAN 20 YEARS, oral morphine has been
`
`standard therapy for moderate to severe cancer pain,
`and it has been incorporated into World Health Organization
`and Agency for Health Care Policy and Research guidelines
`for pain management.' However, there has been a lack of
`data comparing the efficacy of morphine with that of other
`opioid analgesics, particularly with regard to possible side
`effect differences and overall patient satisfaction with treat-
`ment. Until recently, the only alternatives to oral delivery of
`opioids were the intravenous, subcutaneous, or spinal infu-
`sion routes of administration, which are labor-intensive and
`expensive, and often require hospitalization to initiate. The
`development of long-acting controlled-release oral and
`transdermal opioid analgesics has been a major advance that
`has had a substantial impact on clinical practice, notably in
`
`From the University of Texas M.D. Anderson Cancer Center,
`Houston, TX; Technology Assessment Group, San Francisco, CA;
`Immunex Corporation, Seattle, WA; and Janssen Pharmaceutica,
`Titusville, NJ.
`Submitted July 3, 1997; accepted November 20, 1997.
`Supported by an unrestricted educational grant from Janssen Phar-
`maceutica, Titusville, NJ.
`Address reprint requests to Richard Payne, MD, Chief Pain and
`Symptom Management, M.D. Anderson Cancer Center; 1515 Holcombe
`Blvd, Houston, TX 77030; Email rp@mdanderson.com.
`© 1998 byAAmerican Society of Clinical Oncology.
`0732-183X/98/1604-0018$3.00/0
`
`all with their pain medication than those who received
`sustained-release oral forms of morphine (P = .035).
`Fentanyl patients also experienced a significantly lower
`frequency (P < .002) and impact (P < .001) of pain
`medication side effects. These results occurred despite
`the fact that cancer patients who received fentanyl were
`significantly older (P < .001) and had significantly
`lower functioning and well-being scores (P = .001).
`Measures of pain intensity, sleep adequacy, and symp-
`toms demonstrated no significant differences between
`treatment groups.
`Conclusion: These data suggest that patients are
`more satisfied with transdermal fentanyl compared
`with sustained-release oral forms of morphine. A lower
`frequency and reduced impact of side effects with trans-
`dermal fentanyl may be one reason cancer patients
`who receive fentanyl are more satisfied with their pain
`management.
`J Clin Oncol 16:1588-1593. © 1998 by American
`Society of Clinical Oncology.
`
`improved compliance and in the ability of patients to sleep
`through the night or for several uninterrupted hours. 2,3
`Oral sustained-release formulations exist for morphine
`and oxycodone, and one is under development for hydromor-
`phone. Fentanyl can be delivered in a transdermal controlled-
`release formulation. The continuous mode of administration
`made possible by these formulations permits blood and CSF
`opioid concentrations to be sustained, so that peak-and-
`is important in the
`trough effects are minimized. This
`management of cancer or other chronic pain syndromes in
`which pain persists around the clock. Patients may not
`experience recurrent pain at the end of a dosing interval,
`which occurs when pain medications are given solely on an
`as-needed basis.2,3
`Numerous instruments have been developed that reliably
`measure specific quality-of-life concerns of cancer patients.4
`Pain management is essential to high-quality survival in
`patients.5 The objective of this study was
`to compare
`treatment satisfaction, patient-perceived side effects, func-
`tioning, and well-being of patients with advanced cancer
`who were receiving transdermal fentanyl versus sustained-
`release oral morphine. This study is one of the largest to date
`to compare these variables in patients who received transder-
`mal fentanyl or sustained-release oral forms of morphine. 6
`Patients who received Contin (Perdue Frederick Co, Nor-
`walk, CT), Oramorph SR, (Roxanne Laboratories, Colum-
`bus, OH) two sustained-release formulations of oral mor-
`phine sulfate, and Duragesic (fentanyl transdermal system,
`
`1588
`
`Journal of Clinical Oncology, Vol 16, No 4 (April), 1998: pp 1588-1593
`
`Downloaded from jco.ascopubs.org on November 25, 2014. For personal use only. No other uses without permission.
`Copyright © 1998 American Society of Clinical Oncology. All rights reserved.
`
`1
`
`

`

`QUALITY OF LIFE AND CANCER PAIN
`
`Janssen Pharmaceuticals, Titusville, NJ) were selected for
`this study because these medications are the three most com-
`monly used long-acting opioid analgesics in the United States.
`
`PATIENTS AND METHODS
`
`Participants
`
`A total of 513 outpatients receiving Duragesic, MS Contin, or
`Oramorph SR for cancer pain took part in the study. The patient group
`represented 68 sites in the United States, including the University of
`Texas M.D. Anderson Cancer Center, a large academic medical center
`(n = 135); 18 smaller cancer treatment centers (n = 151); and 49
`private oncology practices (n = 227). Participation
`in the study
`involved the completion of a health-related quality-of-life and treatment
`satisfaction questionnaire. For entry onto the study, patients were
`required to be at least 18 years of age; to have completed at least 2
`weeks of cancer pain treatment with current therapy; to be able to
`understand and sign an informed consent; and to be able to speak, read,
`and understand English.
`
`Outcomes Questionnaire
`The questionnaire was designed
`to address key quality-of-life
`domains known to be affected by cancer, cancer pain, and receipt of a
`pain medication. The questionnaire used scales recommended by
`quality-of-life researchers and a clinician panel. The selection of
`domains was refined
`through one-on-one interviews with cancer
`patients, as well as extensive literature reviews. Approximately two
`thirds of the questionnaire consisted of elements of previously existing,
`widely used, validated measures. These included the Functional Assess-
`ment of Cancer Therapy-General (FACT-G) 7 scale, the Wisconsin Brief
`Pain Inventory (BPI),'
`the Memorial Symptom Assessment Scale
`(MSAS),9 and the Medical Outcomes Study (MOS) questionnaire. 10
`The remainder of the questionnaire consisted of original measures
`developed for the study that addressed issues relevant to this population.
`Table I lists more details on the measures and scales included in the
`questionnaire.
`
`Questionnaire Administration
`The questionnaire was designed for self-administration and the
`majority (94.6%) of patients completed the questionnaire unaided. In
`most instances, when patients completed the questionnaire at the site,
`they did so while waiting for their previously scheduled appointments.
`On average, the questionnaire took approximately 35 minutes to
`complete. Twenty-seven questionnaires (5.4%) were administered by an
`interviewer because illness or poor eyesight precluded self-administra-
`
`Table 1. Quality-of-Life and Treatment Satisfaction Questionnaire:
`Domains and Sources
`
`Domain
`
`No. of Items
`
`Physical well-being
`Social/family well-being
`Emotional well-being
`Functional well-being
`Pain
`Symptom assessment
`Sleep
`Side effects
`Satisfaction
`
`7
`7
`5
`7
`3
`24
`9
`2
`8
`
`Abbreviation: TAG, Technology Assessment Group.
`
`Source
`
`FACT-G
`FACT-G
`FACT-G
`FACT-G
`BPI
`MSAS
`MOS/TAG
`TAG
`TAG
`
`1589
`
`tion. Approximately 30% of all surveys were administered by mail
`(rather than by self-administration at the clinic or hospice). The
`response rate for mailed questionnaires was equally poor across
`treatment groups: 23% for patients who received MS Contin, 24% for
`those who received Oramorph SR, and 24% for those who received
`Duragesic. Low response rates are a potential source of bias. Response
`rates for nonmailed questionnaires are not available.
`Study coordinators at each site were centrally trained either in person
`or by telephone to ensure that questionnaires were consistently adminis-
`tered to all patients at all sites. Coordinators were also responsible for
`monitoring the administration of questionnaires and reviewing question-
`naires for completeness.
`
`Sample Size
`The sample size of the study was originally calculated by examining
`the expected variability of two primary outcomes: treatment satisfaction
`and sleep adequacy. The target sample size of 400 patients, with a
`minimum of 200 patients per treatment arm, was based on an 80%
`power to detect a 10% difference at the .05 level of significance,
`assuming a standard deviation of approximately 35% of the mean.
`
`Statistical Methods
`
`The quality-of-life data were analyzed using all assessable patients.
`All of the analyses were conducted using SAS for Windows version
`6.10 (SAS Institute, Cary, NC).
`A variety of analyses were conducted to compare health-related
`quality of life and treatment satisfaction for transdermal fentanyl versus
`sustained-release oral morphine. The primary outcome measures identi-
`fied were treatment satisfaction and sleep adequacy, for which it was
`hypothesized that transdermal fentanyl patients would be more satisfied
`with their pain-related medication and sleep better than oral morphine
`patients, on the basis of previous findings.6 All other outcome measures
`were considered secondary and no differences were expected between
`the two groups.
`Basic validity analyses were performed as part of the overall analysis
`plan, for both previously validated and newly constructed scales, such
`as satisfaction. Convergent validity was assessed by examining correla-
`tions among the quality-of-life scales.
`Variable clustering was used to evaluate the validity of the prespeci-
`fled composite scales and to explore other possible scales. The approach
`uses an iterative splitting technique to divide a group of variables into
`nonoverlapping subgroups, each of which
`is approximately uni-
`dimensional.
`Exploratory evaluation of the outcome measures to determine which
`variables could be combined into scales showed a fairly simple
`structure. Variable cluster analysis showed a total of six underlying
`scales that explained a substantial proportion of the variability. Further-
`more, creating those scales as simple averages of the components was
`an adequate first approximation to the optimal principal component
`scale. Those scales are summarized as follows: the Functioning scale
`consisted of the social, emotional, and functional scores from the
`FACT-G. The Symptom scale consisted of the 24 MSAS items. Any of
`the symptoms could have been caused by the effects of the cancer itself,
`by other analgesic therapies for pain, or by antineoplastic or other
`treatments for cancer. The Pain scale consisted of the items from the
`BPI, plus the Pain Disruption During Sleep item. The Sleep scale
`consisted of all MOS sleep items, except the Sleep Somnolence item,
`"Have trouble staying awake during the day." The Satisfaction scale
`consisted of the items related to ease of use, satisfaction with delivery
`system (oral v patch), willingness to continue medication, likelihood of
`recommending medication to others, whether medication met expecta-
`
`Downloaded from jco.ascopubs.org on November 25, 2014. For personal use only. No other uses without permission.
`Copyright © 1998 American Society of Clinical Oncology. All rights reserved.
`
`2
`
`

`

`1590
`
`PAYNE ET AL
`
`tions, and whether medication offered relief similar to that of others.
`Finally, the Side Effects scale contained two items: one on the frequency
`and the other on the bothersomeness of any side effects related to
`analgesic therapies.
`Because of differences found in demographics and baseline func-
`tional status, an analysis of covariance (ANCOVA) was performed. This
`analysis controlled for patient demographics (age, race, education, sex,
`and marital status), physical well-being, cancer stage, and study site.
`Mean composite scores were compared across transdermal fentanyl
`versus oral morphine groups.
`A total of 24 patients were omitted from the statistical analyses for a
`variety of reasons. Fifteen were excluded because they did not meet
`entry criteria or for various administrative reasons and nine were
`excluded because they were being cared for at sites where all patients
`were receiving only one of the three treatment drugs.
`
`Demographics
`
`RESULTS
`
`At the end of data collection, this cross-sectional study
`had unbalanced treatment arms: 295 patients who received
`sustained-release oral forms of morphine and 209 who
`received transdermal fentanyl were enrolled onto the study,
`for a total of 504 assessable patients.
`Overall, patients in the two treatment groups were equiva-
`lent for most demographic measures, including sex, race,
`education, and marital status (Table 2). However, patients in
`the transdermal fentanyl group were significantly older than
`patients in the oral morphine group (P < .001).
`Distribution of cancer type was statistically equivalent in
`both groups. The largest groups of cancer types include
`breast, colorectal, lung, prostate, and pancreatic cancer.
`Patients in each treatment group were also not statistically
`different as regards cancer stage. The majority of patients
`had late-stage cancers: approximately 80% of transdermal
`fentanyl patients and 74% of oral morphine patients had
`stage IV/D disease.
`The mean dose of transdermal fentanyl was 84.35 pg/h
`(SD = 63.3 pg/h; range, 25 to 400 pg/h) and the median dose
`was 75 pg/h. A 3-day 25-pg/h fentanyl transdermal system
`patch is often reported in the literature to be equivalent to 60
`to 90 mg of oral morphine per 24 hours (90 mg is reported in
`the package insert). These conversion rates suggest that the
`mean daily dose of fentanyl of 84.35 pg/h is equivalent to
`roughly 200 to 300 mg of morphine per day. For those who
`received oral morphine, the mean 24-hour dose was 194.95
`mg (SD = 308.4 mg; range, 15 to 3,000 mg) and the median
`dose was 120 mg. The average daily retail costs of sustained-
`release oral morphine and transdermal fentanyl are similar.
`In fact, they are within 10% of one another, based on the
`retail cost per day of therapy.
`Differences in dose distribution for transdermal fentanyl
`and oral morphine were statistically significant (P = .001).
`We investigated the relationship of dose to outcome
`measures and found that dose was related to pain. As
`
`Table 2. Demographic Characteristics by Treatment
`
`Transdermal
`Fentanyl
`
`Sustained-Release
`Oral Morphine
`
`Characteristic
`
`No.
`
`%
`
`No.
`
`%
`
`X2
`
`df
`
`P
`
`Age, years
`(mean ± SD)
`
`213 + 41.85
`61 ± 13.48
`
`296 ± 55.15 T = 5.25 <.001
`55 + 12.91
`
`Sex
`Male
`Female
`Race
`Black/African
`American
`White/Caucasian
`Latino/Hispanic
`Native American
`Asian/Pacific Islander
`Other
`Education
`Some college or less
`College degree or
`more
`Family income ($)
`< 20,000
`20,000-39,999
`40,000-59,999
`60,000-79,999
`> 80,000
`Marital status
`Single, never married
`Married or with
`partner
`Separated
`Divorced
`Widowed
`Cancer stage
`I/A-Ill/C
`IV/D
`
`100 48.08
`108 51.92
`
`139
`155
`
`47.28
`52.72
`
`27 13.04
`
`34
`
`11.68
`
`163 78.74
`9
`4.35
`0.48
`1
`2
`0.97
`5
`2.42
`
`228
`15
`9
`3
`2
`
`78.35
`5.15
`3.09
`1.03
`0.69
`
`162 77.88
`46
`22.12
`
`240
`53
`
`81.91
`18.09
`
`82 43.62
`53 28.19
`31
`16.49
`8
`4.26
`7.45
`14
`
`130
`75
`31
`16
`10
`
`49.62
`28.63
`11.83
`6.11
`3.82
`
`7.73
`16
`132 63.77
`
`34
`169
`
`11.60
`57.68
`
`8
`31
`20
`
`3.86
`14.98
`9.66
`
`13
`51
`26
`
`4.44
`17.41
`8.87
`
`20.62
`40
`154 79.38
`
`66
`194
`
`25.38
`74.62
`
`0.03 1 <.86
`
`7.03 5
`
`.22
`
`1.24 1
`
`.27
`
`5.98 4
`
`.20
`
`3.18 4
`
`.53
`
`1.41
`
`1
`
`.235
`
`expected, higher doses were associated with greater levels of
`pain intensity. In addition, we found that the transdermal
`fentanyl patients showed a weaker relationship between
`dose and sleep disturbance due to pain than did oral
`morphine patients, although the interaction fell barely short
`of significance (P = .08). This suggests that the original
`hypothesis, that patients who receive transdermal fentanyl
`therapy sleep better, may be true at higher doses.
`
`Outcome Measures
`Results of construct validity for the quality-of-life scales
`were consistent with expectations, that is, classes of items
`such as sleep or pain were well correlated among them-
`selves. The appropriateness of combining individual items
`into an overall scale was assessed using measures of internal
`consistency such as Cronbach's alpha (values ranged from
`.75 to .91 for all scales except social well-being at .58). For
`all composite scales, the simple average explained at least
`93% as much variation as the first principal component. That
`
`Downloaded from jco.ascopubs.org on November 25, 2014. For personal use only. No other uses without permission.
`Copyright © 1998 American Society of Clinical Oncology. All rights reserved.
`
`3
`
`

`

`QUALITY OF LIFE AND CANCER PAIN
`
`is, at most, 7% of the explanatory power of a composite was
`lost by making the simplifying assumption of equal weights
`for the individual variables. (This was determined by
`comparing the variance explained by the first principal
`component with the variance explained by the first centroid
`component.)
`The results for comparisons of the two treatment groups
`for each of the six composite scores are described in the
`following sections and listed in Table 3.
`Functioning. Cancer patients who received transdermal
`fentanyl had significantly lower scores on the Functioning
`scale, which is a combined measure of social, emotional, and
`functional well-being as measured on the FACT-G (P <
`.001).
`Pain. Composite pain scores were comparable for both
`treatment groups. However, an interaction with cancer stage
`was evident. The transdermal fentanyl patients reported
`higher pain for stages I/A and II/B cancer, lower pain for
`stage III/C, and similar pain for stage IV/D. It should be
`noted that the actual number of patients with stage I or stage
`II cancers was small (-8%). The individual pain items from
`the BPI-pain in the past week, pain on average, and pain
`now (Table 4)-were consistent with overall results. Self-
`reported pain levels for all three items were statistically
`equivalent between the groups.
`Sleep, symptoms. No statistically significant differences
`between the two treatment groups were found with respect
`to sleep measures or MSAS symptom scores.
`Side effects. The most dramatic difference was found for
`the two items (frequency and impact) that assess a global
`measure of side effects. Transdermal fentanyl patients
`reported less frequent and bothersome side effects. For
`instance, a higher percentage of fentanyl patients (50%)
`reported never having side effects than did oral morphine
`patients (36%) (P = .002), and a higher percentage reported
`no side effects or not being bothered by their side effects
`(68% v 46%; P = .001). Because these items are global in
`nature, they do not attempt to distinguish between
`the
`
`Table 3. Composite Scores by Treatment
`
`Transdermal
`Fentanyl
`
`Sustained-Release
`Oral Morphine
`
`Domain*
`
`Functioning
`Pain
`Symptom assessment
`Sleep
`Side effects
`Satisfaction
`
`Mean
`
`54.35
`45.05
`98.58
`96.00
`22.01
`82.98
`
`SEM
`
`2.10
`4.03
`0.08
`0.17
`3.90
`2.22
`
`Mean
`
`59.21
`39.40
`98.63
`95.81
`31.06
`79.81
`
`SEM
`
`1.94
`3.41
`0.08
`0.16
`3.62
`2.07
`
`P
`
`< .001
`.26
`.22
`.30
`< .001
`.035
`
`NOTE. Scores adjusted for site, cancer stage, physical well-being, and
`demographics (age, sex, race, education, and marital status).
`*Scales of 0 to 100, where higher scores indicate more of an attribute (eg,
`higher pain score indicates more pain, higher satisfaction score indicates
`greater satisfaction).
`
`1591
`
`Table 4. Pain Relief by Treatment
`
`Transdermal
`Fentanyl
`
`Sustained-Release
`Oral Morphine
`
`Subscale*
`
`Mean _ SD
`
`Pain (pastweek)
`Pain (on average)
`Pain (right now)
`Pain relief
`
`55.75 + 21.39
`40.17 ± 24.92
`29.91 + 28.53
`72.55 ± 21.39
`
`No.
`
`212
`214
`212
`210
`
`Mean ± SD
`
`No.
`
`58.11 ± 27.94
`296
`41.33 + 23.09
`294
`32.93 ± 25.94 297
`70.53 -± 22.64
`294
`
`P
`
`.36
`.57
`.21
`.31
`
`*Range, 0 to 100; higher score indicates more pain or more pain relief.
`
`frequency and/or impact of individual, particular side ef-
`fects.
`Satisfaction. The overall results for satisfaction indi-
`cated that patients who received transdermal fentanyl were
`significantly more satisfied than those who received oral
`morphine (P = .035). Differences by sex for each of the six
`satisfaction composite items and for the side effect items by
`treatment are listed in Table 5. Men who received transder-
`mal fentanyl reported significantly more willingness to
`continue the medication, a greater likelihood of recommend-
`ing its use, and a greater likelihood that it met their
`expectations. Men also reported significantly more often that
`transdermal fentanyl provided better pain relief than other
`prior pain medications. No significant differences emerged
`for women.
`
`DISCUSSION
`
`This large, cross-sectional study identifies several areas
`for further investigation. The transdermal fentanyl patients
`had lower functioning scores than did the oral morphine
`patients; however, despite this lower functioning, and even
`though the two groups had equivalent degrees of pain relief,
`the transdermal fentanyl patients reported many fewer and
`less bothersome side effects. In addition, there was greater
`satisfaction among transdermal fentanyl patients, as mea-
`sured by significantly more willingness to continue the
`medication and a greater likelihood that the medication met
`their expectations.
`It is unexplained why the mean age of the transdermal
`fentanyl patients was different from the mean age of the oral
`morphine patients. However, it has been the clinical experi-
`ence that older patients generally experience more side
`effects than do younger patients. The direction of the effect
`is opposite to that which would have been expected had
`there been a significant interaction between age and the
`appearance of side effects.
`There are known influences of sex on pain perception and
`response to analgesics, 11,12 although sex-related differences
`in the occurrence of side effects with opioid therapy have not
`been previously described. Subgroup analysis by sex showed
`most of the increase in satisfaction to be attributable to men.
`If this finding is borne out in future studies, it may suggest
`sex differences in side effect-related satisfaction, in the
`
`Downloaded from jco.ascopubs.org on November 25, 2014. For personal use only. No other uses without permission.
`Copyright © 1998 American Society of Clinical Oncology. All rights reserved.
`
`4
`
`

`

`1592
`
`PAYNE ET AL
`
`Table 5. Satisfaction and Side Effects Related to Pain Medication by Sex
`
`Men
`
`Women
`
`Domain*
`
`Satisfaction
`Ease oF use
`Expectation met
`Similar relief
`Delivery system
`Willingness to continue
`Recommended to others
`Side Effects
`Frequency
`Impact (bothersomeness)
`
`Transdermal Fentanyl
`(mean - SD)
`
`Sustained-Release Oral
`Morphine (mean t SD)
`
`87.71 ± 18.68
`81.12 + 27.35
`84.64 ± 20.60
`88.50 t 18.94
`94.75 + 12.46
`96.94 ± 17.32
`
`13.67 ± 21.22
`28.66 ± 39.89
`
`85.85 ± 18.60
`67.99 ± 31.28
`83.70 ± 22.22
`87.05 ± 21.78
`t 23.11
`87.96
`89.78 ± 30.40
`
`27.39 ± 27.88
`39.27 ± 36.33
`
`P
`
`.45
`.001
`.76
`.59
`.008
`.037
`
`.001
`.036
`
`Transdermal Fentanyl
`(mean ± SD)
`
`Sustained-Release
`Oral Morphine (mean ± SD)
`
`87.58 + 19.53
`76.17 ± 28.60
`83.84 A
`22.11
`91.51 + 16.14
`91.75 ± 19.74
`94.39 + 23.11
`
`21.94 ± 27.66
`32.76 ± 38.44
`
`86.88 + 19.46
`75.82 ± 26.97
`84.06 ± 22.99
`90.16 ± 19.43
`92.53 ± 15.69
`96.08 ± 19.47
`
`27.01 ± 28.61
`38.97 ± 36.88
`
`P
`
`.78
`.92
`.94
`.56
`.72
`.53
`
`.16
`.19
`
`*Range, 0 to 100; higher scores indicate more of an attribute (ie, higher satisfaction score indicates greater satisfaction, higher side effects value indicates more
`frequent or bothersome side effects).
`
`perception of stigma related to use of morphine, or in other
`contributing elements of satisfaction. Other possible expla-
`nations include sex differences in pain management services
`such as titration or side effect management, although both
`men and women reported fewer, less bothersome side effects
`with transdermal fentanyl. We found that transdermal fen-
`tanyl patients of both sexes reported more side effects
`associated with their previous pain medication than did the
`oral morphine patients, and no other sex differences in
`previous pain treatment experience or pain intensity (prior or
`current) were found that might explain these sex differences
`in satisfaction with pain treatment.
`As expected, patients reported being titrated by their
`treating physicians to similar levels of pain relief regardless
`of which medication was being used. Interestingly, function-
`ally worse cancer patients in the transdermal fentanyl group
`reported reduced analgesia-associated side effects in the face
`of somewhat higher mean doses of opioid therapy (194.95
`mg/d in the oral morphine group v 200 to 300 mg/d
`morphine equivalent in the fentanyl group). This difference,
`if true, may be explained by several factors, including
`in binding to various subclasses of opioid
`differences
`receptor between morphine and fentanyl molecules.'13
`in sleep quality did not reach statistical
`Differences
`significance between groups (P = .08), although the data
`suggest that possible advantages among fentanyl patients
`who received higher-dose opiate therapy were obscured by
`those who received lower doses, where little difference
`appeared to exist. The inability to distinguish between
`therapies may be explained by less impact on sleep quality
`by lower doses of either drug irrespective of delivery
`system.
`This is the largest cross-sectional study published to date
`to compare two opioids, morphine and fentanyl. Classically,
`the prospective design is considered optimal, but the major-
`ity of patients in this study had end-stage disease and the risk
`of attrition in a prospective, long-term, follow-up study is
`
`high. The long-term follow-up evaluation inherent in a
`prospective trial increases the risk of missing data.' 4 Two
`factors contributed to the decision that a cross-sectional
`design would be the most feasible approach. One was the
`potential for enrolling sufficient numbers of patients to
`detect differences with statistical significance into a random-
`ized study that offered no novel or curative therapy, and the
`other was the availability of a wide range of sites whose
`staffs were interested in participating in the study.
`The results of this study are consistent with those of
`another recently published, randomized, open, parallel-
`group crossover study that compared transdermal fentanyl
`with sustained-release oral morphine in patients receiving
`palliative care for advanced cancer.15 The investigators of
`that study, which was conducted throughout the United
`Kingdom, concluded that transdermal fentanyl was associ-
`ated with significantly less constipation (P < .001), a higher
`incidence of normal stools (P = .002), and less daytime
`drowsiness (P = .015) than was oral morphine, and that the
`more favorable side effect profile of transdermal fentanyl
`may have contributed to the patients' preference for that
`medication. Of those who were able to express a preference
`(n = 136), significantly more patients preferred transdermal
`fentanyl (P = .037). In that study as well, there were no
`significant differences between the two drugs with respect to
`pain control.
`Satisfaction is a complex outcome. For example, in a
`review of several recent studies, it was observed that
`between 48% and 63% of patients were "very satisfied"
`with their pain therapies, despite reporting BPI scale "worst
`pain" scores of 6.6 to 7.7 on a scale of 0 to 10, which is in the
`severe pain range.16 Our finding that approximately 70% of
`patients were "very pleased" with their pain medicine when
`reporting mean "pain right now" scores of 29.91 and 32.93,
`respectively, with transdermal fentanyl and oral morphine
`(Table 4) is more consistent than the previous report. These
`
`Downloaded from jco.ascopubs.org on November 25, 2014. For personal use only. No other uses without permission.
`Copyright © 1998 American Society of Clinical Oncology. All rights reserved.
`
`5
`
`

`

`QUALITY OF LIFE AND CANCER PAIN
`
`scores indicate mild pain and are consistent with the high
`satisfaction scores.17
`It is likely that a global outcome measure such as
`"satisfaction" encompasses many aspects of a patient's
`experience with an analgesic, including pharmacologic and
`physical effects such as pain relief, side effects, ease of
`compliance, and ability to function, and psychologic effects
`such as worry about addiction, stigmatization about taking
`opioids, and fears about disease progression. To understand
`patient satisfaction truly, one must evaluate these related
`factors more specifically. It is also unclear whether the
`magnitude of the difference would achieve clinical rel-
`evance. The passive absorption of the 72-hour transdermal
`formulation of fentanyl may not only reduce constipation by
`avoiding the gastrointestinal tract as the primary route of
`absorption, but may also improve compliance with pre-
`scribed analgesic therapy, since patients do not have to
`remember to take medications as often as every 12 hours and
`have fewer options to lower the dose in misguided attempts
`to minimize their exposure to opioids because of unfounded
`fears of addiction.
`The limitations of the cross-sectional design, the un-
`known magnitude of nonresponse bias as a result of low
`response rates, and the nonblinded conditions of the study do
`not allow definitive conclusions regarding the two drugs.
`Nonetheless, the finding of differences in side effects and
`patient preference for transdermal fentanyl over oral mor-
`phine is intriguing. Although both of these opioids are mu
`agonists, the transdermal route of administration may mini-
`
`1593
`
`mize the emetic and constipating effects of opioids"8 in
`comparison with oral administration. This may account for
`the lower frequency of side effects in this group. One can
`also speculate that the transdermal route of administration
`may be less stigmatizing to patients than taking "pills," and
`fentanyl is less familiar to patients (and thus less stigma-
`tized) than is morphine. These factors may be important in
`contributing to the patient preference for continuing transder-
`mal fentanyl in comparison with oral morphine. Just as
`future research into receptor subclass binding or route of
`administration may elucidate side effect profile differences
`among strong mu agonists, patient preference studies should
`explicitly address the contribution of social stigma and
`dosing mechanism to differences between opioids.
`In conclusion, the results of this large cross-sectional
`study suggest that patients who received transdermal fen-
`tanyl were more satisfied with their medication than were
`patients who received sustained-release oral forms of mor-
`phine and that the transdermal fentanyl group reported less
`impact from and lower frequency of side effects. A random-
`ized, controlled trial with appropriate measures of blinded
`analgesic treatments is needed to evaluate the comparative
`advantages of transdermal fentanyl and sustained-release
`oral morphine definitively.
`
`ACKNOWLEDGMENT
`
`We thank Lori Potter for statistical help and Patty Mallery and Mike
`Burchmore for help in data collection.
`
`REFERENCES
`1. Jacox A, Carr DB, Payne R, et al: Management of Cancer Pain.
`10. Cleeland CS: Assessment of pain in cancer: Measurement issues,
`Clinical Practice Guideline No. 9. Agency for Health Care Policy and
`in Foley KM, Bonica JJ, Ventafridda V, et al (eds): Advances in Pain
`Research publication no. 94-0592. Rockville, MD, United States
`Research and Therapy (vol 16). New York, NY, Raven, 1990, p 52
`Department of Health and Human Services, Public Health Service,
`11. Gear RW, M