|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 18 (June 20), 2006: pp. 2828-2835 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.04.8207 Quality of Life and Pain in Advanced Stage Prostate Cancer: Results of a Southwest Oncology Group Randomized Trial Comparing Docetaxel and Estramustine to Mitoxantrone and Prednisone
From the University of Washington; Southwest Oncology Group Statistical Center/Fred Hutchinson Cancer Research Center, Seattle, WA; Columbia Presbyterian Medical Center, New York, NY; Central Arkansas Veterans’ Affairs Healthcare System, Little Rock, AR; University of California Davis Cancer Center, Sacramento, CA; Southeastern Medical Oncology Center, Goldsboro, NC; Dana-Farber Cancer Institute, Boston, MA; Mayo Clinic, Rochester, MN; University of Michigan, Ann Arbor, MI; and the University of Colorado Cancer Center, Denver, CO Address reprint requests to Donna L. Berry, PhD, Box 357266, University of Washington, Seattle, WA 98195-7266; e-mail: donnalb{at}u.washington.edu
PURPOSE: Palliation of bone pain can be achieved in men with androgen-independent prostate cancer treated with docetaxel and estramustine (DE) or mitoxantrone and prednisone (MP). While Southwest Oncology Group trial 99-16 demonstrated a survival improvement of DE over MP, the study also was designed to compare the palliation of disease-related symptoms. METHODS: Pain palliation and global quality of life (QOL) were the two primary patient-reported outcomes. Pain was measured with the Present Pain Intensity scale of the McGill Pain Questionnaire-Short Form. The European Organisation for Research and Treatment of Cancer Core Quality of Life Questionnaire C30 (QLQ-C30) and its Prostate Cancer Module (PR25) measured QOL and symptom status. Pain and analgesic use were measured at random assignment, every cycle for eight cycles, and 1 year from random assignment; the QLQ-C30 and the PR25 were administered at random assignment, before cycle four (week 10) and cycle eight (month 6) and at 1 year. In addition to the primary intent-to-treat, missing at random analysis, sensitivity analyses were performed to assess robustness of global QOL conclusions under alternative informative missing data assumptions. RESULTS: Six hundred seventy four eligible patients received DE (n = 338) or MP (n = 336). In an intention-to-treat analysis, median overall survival was 17.5 months for the DE arm and 15.6 months for the MP arm (P = .02). There were no statistically significant differences in pain palliation between the treatment arms. The sensitivity analyses showed a consistent lack of statistically significant global QOL differences for the two arms. CONCLUSION: DE had superior clinical efficacy (overall survival, time-to-progression, and prostate-specific antigen declines) with similar global QOL and pain palliation in the MP arm.
The standard of care for metastatic prostate cancer is androgen blockade, which delays progression and relieves pain for an average of 18 months to 24 months.1,2 Although the first manifestation of tumor resistance to androgen blockade is often an asymptomatic rise in serum prostate-specific antigen (PSA), nearly all patients eventually develop significant disease related morbidity including fatigue, bone pain, and weight loss. In 1996, the US Food and Drug Administration approved the use of mitoxantrone combined with corticosteroids for palliation of bone pain without requiring concomitant improvement in survival for men with androgen-independent prostate cancer. Two recently published randomized trials, Southwest Oncology Group (SWOG) trial 99-163 and TAX 327,4 demonstrated a 20% to 24% reduction in risk of death in favor of docetaxel and estramustine (DE) or docetaxel and prednisone over mitoxantrone-based therapy in patients with androgen-independent prostate cancer. Improved quality of life (QOL), as measured by the Functional Assessment of Cancer Therapy Prostate Cancer Subscale (FACT-P) score, as well as reductions in bone pain, were demonstrated in both the weekly and every 3-week doses of docetaxel when compared with doses of mitoxantrone in patients treated on TAX 327.4 A secondary end point of SWOG 99-16 was to evaluate and compare the impact of DE or mitoxantrone and prednisone (MP) treatment on aspects of QOL including the palliation of metastatic bone pain and global QOL (GQOL). Given the previously reported gastrointestinal, cardiovascular, and hematologic toxicities with DE,5,6 a QOL analysis is particularly germane, balancing the risks and toxicities against the survival benefits of chemotherapy.
Patient Selection Patients with pathologically confirmed adenocarcinoma of the prostate, stage D1 or D2, and progressive disease despite surgical or medical castration were eligible for entry on SWOG 99-16. Disease progression was defined by at least one of the following criteria: progression of a bi-dimensionally measurable lesion assessed within 28 days of study registration; progression of assessable but not measurable disease (eg, bone scan) assessed within 42 days of registration; or rising serum PSA, with at least two consecutive increasing PSA measurements over baseline obtained over at least 7-day intervals. Study enrollment commenced in October 1999 and was completed in January 2003. All enrolled patients signed informed consent, and the protocol was approved by institutional review boards of each participating institution (SWOG, Cancer and Leukemia Group B, North Central Cancer Treatment Group, the Clinical Trials Support Unit, and the Extended Participation Project program through the National Cancer Institute in Bethesda, MD.)
Treatment and Evaluation
QOL Assessment MPQ-SF. The MPQ-SF9,10 is a validated measure of pain including three sections: pain words, pain degree, and pain intensity. The pain words total score (0 to 45) is obtained for the rating of 15 words describing pain with a 0 to 3 response scale (none to severe). This score reflects the quality of experienced pain (eg, throbbing, aching, tender). A second score, adapted from the original visual analog scale, reflects the degree of pain on a 0 to 10 numerical scale.11 The present pain intensity (PPI) pain item from the MPQ-SF, one of two primary QOL end points, lists responses ranging from 0 (no pain) to 5 (excruciating). A pain response based on the definition by Tannock et al12 is a two-point reduction on this scale for two consecutive measurements, 3 weeks apart, without an increase in pain medications. (See Statistical Considerations section for rules on creating this response variable). EORTC QLQ-C30, PR25. The second primary QOL end point is change in the EORTC QLQ-C30, two-item GQOL scores from baseline to the beginning of cycle eight (6 months) for the two treatment arms. In designing the QOL assessment strategy, we replicated (as much as possible) the scales used by Tannock and his colleagues12 in earlier trials with prednisone, mitoxantrone, and the combination. GQOL offered an overall perception of the patient's well-being, it correlated highly with the pain measure in an earlier phase II trial,13 and showed significant improvement in the combined regimen change.14 Version 3.0 incorporates four-point scales for the five physical functioning items (addendum to QLQ-C30 scoring manual).8 Analgesic use. A pain medication log, used in a previous SWOG trial,15 was implemented to monitor analgesic use in this trial. This log was modified for ease of data entry to include the following information: an analgesic code (no analgesic, non-narcotic agent, weak opioid, strong opioid), the dose, and the number of doses taken. The patient recorded all analgesic use for the 24-hour period before each clinic treatment visit. Quality control. Several efforts enhanced quality control for the patient-reported measures in this trial. The QOL assessment schedule was to be followed even if the patient went off treatment/study early (eg, in the event of progression or treatment-related toxicities). In addition, a cover sheet was required with submission of questionnaires at each assessment point. The cover sheet included the following information: whether the assessment occurred; if assistance was required; and if the assessment did not occur, documentation of the reason was required. Special instructions for administering the QOL questionnaires and the pain medication log were included in the study protocol. Finally, institutions were reimbursed for submission of the patient-completed forms to compensate for the number of forms and the different assessment schedules for the forms.
Statistical Considerations
The primary time point for comparison and statistical testing was at cycle eight (month 6). Statistical comparison of pain response (responder v nonresponder v stable) by treatment arm was conducted using the
Patient Characteristics Between October 1999 and January 2003, a total of 770 patients were enrolled on the trial. The baseline characteristics of the 629 eligible patients with pain response data are presented in Table 1. None of the patient characteristics listed in Table 1 were significantly different at the .05 level between the primary trial participants who did and did not report QOL data.
QOL Submission Rates The submission rates of the MPQ-SF, QLQ-C30, and QLQ-PR25 for patients qualifying for assessments are summarized in Table 2. The QLQ-C30 and QLQ-PR25 questionnaires were only administered at baseline, cycle four, cycle eight, and 1 year. Reasons for nonsubmission of 6-month (cycle eight) QOL assessment for DE versus MP, respectively, include: illness (7% v 5%), unrelated to illness (6% v 5%), institutional error (30% v 32%), death (21% v 24%), other (11% v 13%), and unknown (25% v 20%). A total of 88 patients died by the 6-month assessment, 39 receiving DE and 49 receiving MP. The DE arm had better submission rates than the MP arm did, consistent with DEs longer survival rate (data not shown).
Primary QOL End Points: Pain Response and GQOL A summary of the PPI score, numerical pain intensity, and total score collected on the MPQ-SF is presented in Table 3. The primary end point of pain response was met in 68 of 318 patients (21%) on the DE arm and in 75 of 311 patients (24%) on the MP arm. Stable response was achieved in 25 patients (8%) versus 37 patients (12%) and no response was recorded in 225 patients (71%) versus 199 patients (64%) on DE and MP, respectively. The pain response was not significantly different between the two treatment arms (P = .12).
The mean GQOL scores by assessment time are displayed in Table 4; higher scores reflect better GQOL. The DE arm experienced a greater increase in the GQOL score from baseline to 6 months as estimated by the MAR model and the four pattern mixture models. However, this difference was only statistically significant in the complete case model (P = .02), the most extreme and unrealistic mechanism of the informative missing models, and the difference achieved borderline significance in the all available model (P = .05). These analyses examined a model assuming data are MAR as well as a set of models varying potential mechanisms for informative missing data. The value of including the sensitivity analyses is the opportunity to look for consistency in results, given problematic levels of missing data (ie, any single analysis would be suspect in this context). Since significance was not achieved in either the MAR or the majority of the four sensitivity models, there was no conclusive evidence for a significant difference in the GQOL end point.
Secondary QOL End Points: Other QLQ-C30 and PR25 Scales A summary of the QLQ-C30 physical functioning and symptom scores is presented in Table 4. The patients in the DE arm reported a significant increase in nausea and vomiting symptoms from baseline to cycle four (P = .02) and cycle eight at 6 months (P < .001) compared with the MP arm. The number of participants with an increase from baseline of  10 points on the GQOL score, indicating improved QOL, was significantly greater (P = .008) in the DE group at 1 year (Table 5) . These secondary results should be qualified by the considerable missing data at the 1-year time point.
The mean QLQ-PR25 scores by assessment time are presented in Table 6. The absence of notable trends or differences on this prostate symptom specific measure did not warrant additional testing.
Patients treated with DE in SWOG 99-16 did not demonstrate statistically or clinically significant differences for pain palliation or GQOL when compared with patients treated with MP. A modest improvement in favor of DE was measured with the GQOL scales. Consistent with the previously reported toxicity data,3 significantly more nausea and vomiting were reported by patients receiving DE. These findings contrast with those reported by Tannock et al,4 where pain palliation was significantly greater in those patients treated with either weekly or every-3-week docetaxel than in those patients treated with MP. Since the patients treated in SWOG 99-16 and TAX 327 had similar baseline characteristics, the reason for this difference in palliation may be attributed to the continuous administration of prednisone in both docetaxel arms in the TAX 327, as well as the addition of estramustine to docetaxel in SWOG 99-16. Prednisone, a corticosteroid, has well-documented analgesic properties but its effects on well-being are less supported.18 Estramustine has known deleterious adverse effects that were experienced by many participants, notably nausea and vomiting, fatigue, and deep venous thrombosis.3 Despite these considerations, there are no statistically significantly different quality of life and pain outcomes reported for DE compared with MP, the previous palliative standard of care. Of note, Oudard et al19 reported similar pain palliation results of no significant differences using a modified McGill Pain Index and analgesic consumption in all three arms of a phase II study administering oral prednisone 5 mg twice daily and randomly assigning patients to a weekly or every-3-week docetaxel regimen combined with estramustine or mitoxantrone. Using a pain response index that incorporates both self-reported pain score as well as concomitant analgesic use is one way to address potential limitations of competing explanations for pain response other than the primary intervention (in this case, the new chemotherapy regimen, DE). Despite the labor intensity of documenting and calculating analgesic type and dose, we found the index a useful tool for such clinical trial analyses. We were able to show that, the chemotherapy regimens palliated cancer pain in 21% of the participants receiving DE and in 24% of the patients receiving MP without the need for an increase in analgesic use. Similarly, both regimens maintained stability in another 8% and 12%, respectively. The mean PPI scores were never higher than 1.62 (on the 0 to 5 scale) at any data collection point, indicating a sample in which pain was being fairly well managed. This is consistent with previous reports of PPI scores in trials with cancer pain patients.20,21 Missed QOL evaluations during or after treatment often is cited as a limitation of longitudinal clinical trials.22-24 Participants who submitted fewer questionnaires did indeed report lower QOL at baseline and at their last data collection time point compared with those who submitted more questionnaires. Fishman et al25 reported results from an observational study of patients with metastatic hormone-refractory prostate cancer over a 9-month period. They noted a statistically significant rapid decrease in QOL over the study period. This similar type of deterioration has been evident for longitudinal studies using the QLQ-C30, FACT-P, and EuroQoL EQ-5D scores.7,26,27 Fishman et al25 interpreted these findings as supporting the need for palliative treatment for patients with metastatic hormone-refractory prostate cancer. Under a concern that data may not be MAR, we conducted a sensitivity analysis corresponding to different informative missing data mechanisms. The consistent finding of no significant differences in GQOL scores (one exception occurred for the method with the least plausible assumption) gives us confidence in our conclusion that DE did not result in statistically significantly better or worse GQOL compared with treatment with MP; this is important for clinicians selecting regimens for their patients. Our trial is limited by the duration of time between questionnaire completions. We may have missed important incidents or pain data that occurred during the 21-day period but were not reported at the beginning of the new cycle. For example, pain indices may initially worsen in the week after treatment, due to flare, and can be observed 7 days to 14 days after docetaxel. Moreover, the improvement in pain and fatigue from dexamethasone administration may not last more than a few days after docetaxel treatment. Likewise, measuring QOL only after 10 weeks, 6 months, and 1 year may not always have documented important, time-limited participant perceptions. Tannock et al,4 comparing MP versus DP, reported results for the FACT-P QOL questionnaire measured every-3-weeks during treatment. Median change scores indicated no change in physical well being, social/family functioning, and functional well being, but did indicate improvement in emotional well being for all study regimens and significantly better improvement in prostate-specific issues for the docetaxel group. The investigators did not report attrition rates for QOL measures nor any evaluation of missing data effects. In summary, pain palliation and GQOL did not differ statistically with respect to improvement or deterioration for patients receiving either the DE regimen or the MP regimen, despite significantly worse nausea at 10 weeks and 6 months in the DE group. Future work in advanced prostate cancer and patient-reported outcomes should document and perform sensitivity analyses for potential informative missing data mechanisms. Use of a composite variable that accounts for both pain report and analgesic use is also recommended.
Although all authors completed the disclosure declaration, the following authors or their immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Dollar Amount Codes (A) < $10,000 (B) $10,000-99,999 (C)
We thank the patients who participated in this trial and nurses and research associates who facilitated submission of patient-completed forms. We acknowledge the Hope Foundation for supporting the pain medication log data analysis, plus the efforts of Mary Galligan, Shirley Livingston, and Gary Donaldson, PhD.
Presented in part in poster format at the 40th Annual Meeting of the American Society of Clinical Oncology, New Orleans, LA, June 5-8, 2004. Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Crawford ED, Eisenberger MA, McLeod DG, et al: A controlled trial of leuprolide with and without flutamide in prostatic carcinoma. N Engl J Med 321:419-424, 1989[Abstract] 2. Eisenberger MA, Blumenstein BA, Crawford ED, et al: Bilateral orchiectomy with or without flutamide for metastatic prostate cancer. N Engl J Med 339:1036-1042, 1998 3. Petrylak D, Tangen C, Hussian M, et al: Docetaxel and estramustine compared with mitoxantrone and prednisone for advanced refractory prostate cancer. N Engl J Med 351:1513-1520, 2004 4. Tannock IF, de Wit R, Berry WR, et al: Docetaxel plus prednisone or mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med 351:1502-1512, 2004 5. Sinibaldi VJ, Carducci MA, Moore-Cooper S, et al: Phase II evaluation of docetaxel plus one-day oral estramustine phosphate in the treatment of patients with androgen independent prostate carcinoma. Cancer 94:1457-1465, 2002[CrossRef][Medline] 6. Petrylak DP: Future directions in the treatment of androgen-independent prostate cancer. Urology 65:8-12, 2005[CrossRef][Medline] 7. Aaronson NK, Ahmedzai S, Bergman B, et al: The European Organization for Research and Treatment of Cancer QLQ-C30: A quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 85:365-376, 1993 8. Fayers P, Aaronson N, Bjordal K, et al: EORTC QLQ-C30 Scoring Manual (3rd Ed). Brussels, Belgium, EORTC, 2001 9. Melzack R: The short-form McGill Pain Questionnaire. Pain 30:191-197, 1987[CrossRef][Medline] 10. Melzack R: The McGill pain questionnaire: From description to measurement. Anesthesiology 103:199-202, 2005[CrossRef][Medline] 11. Wilkie DJ, Lovejoy N, Dodd M, et al: Cancer pain intensity measurement: Concurrent validity of three instruments-finger dynamometer, pain intensity number scale, visual analogue scale. The Hospice Journal 6:1-13, 1990[CrossRef][Medline] 12. Tannock IF, Osoba D, Stockler MR, et al: Chemotherapy with mitoxantrone plus prednisone or prednisone alone for symptomatic hormone-resistant prostate cancer: A Canadian randomized trial with palliative end points. J Clin Oncol 14:1756-1764, 1996 13. Tannock I, Gospodarowicz M, Meakin W, et al: Treatment of metastatic prostatic cancer with low-dose prednisone: Evaluation of pain and quality of life as pragmatic indices of response. J Clin Oncol 7:590-597, 1989[Abstract] 14. Osoba D, Tannock IF, Ernst DS, et al: Health-related quality of life in men with metastatic prostate cancer treated with prednisone alone or mitoxantrone and prednisone. J Clin Oncol 17:1654-1663, 1999 15. Geyer CE Jr, Green SJ, Moinpour CM, et al: Expanded phase II trial of paclitaxel in metastatic breast cancer: A Southwest Oncology Group study. Breast Cancer Res Treat 51:169-181, 1998[CrossRef][Medline] 16. Little RJ, Wang Y: Pattern-mixture models for multivariate incomplete data with covariates. Biometrics 52:98-111, 1996[CrossRef][Medline] 17. Osoba D, Rodrigues G, Myles J, et al: Interpreting the significance of changes in health-related quality-of-life scores. J Clin Oncol 16:139-144, 1998 18. Zoorob RJ, Cender D: A different look at corticosteroids. Am Fam Physician 58:443-450, 1998[Medline] 19. Oudard S, Banu E, Beuzeboc P, et al: Multicenter randomized phase II study of two schedules of docetaxel, estramustine, and prednisone versus mitoxantrone plus prednisone in patients with metastatic hormone-refractory prostate cancer. J Clin Oncol 23:3343-3351, 2005 20. Berry DL, Wilkie DJ, Huang HY, et al: Cancer pain and common pain: A comparison of patient-reported intensities. Oncol Nurs Forum 26:721-726, 1999[Medline] 21. Beer TM, Eilers KM, Garzotto M, et al: Quality of life and pain relief during treatment with calcitriol and docetaxel in symptomatic metastatic androgen-independent prostate carcinoma. Cancer 100:758-763, 2004[CrossRef][Medline] 22. Bernhard J, Cella DF, Coates AS, et al: Missing quality of life data in cancer clinical trials: Serious problems and challenges. Stat Med 17:517-532, 1998[CrossRef][Medline] 23. Pauler DK, McCoy S, Moinpour C: Pattern mixture models for longitudinal quality of life studies in advanced stage disease. Stat Med 22:795-809, 2003[CrossRef][Medline] 24. Fairclough DL: Design and Analysis of Quality of Life Studies in Clinical Trials. Boca Raton, FL, Chapman & Hall, 2002 25. Fishman M, Jacobsen P, Chen N, et al: Quality of life findings from a multicenter, multinational, observational study of patients with metastatic hormone-refractory prostate cancer. J Clin Oncol 23:755S, 2005 (suppl; abstr 8106) 26. Esper P, Mo F, Chodak G, et al: Measuring quality of life in men with prostate cancer using the functional assessment of cancer therapy-prostate instrument. Urology, 50:920-928, 1997[CrossRef][Medline] 27. The EuroQol Group: EuroQol: A new facility for the measurement of health-related quality of life. Health Policy 16: 199-208, 1990[CrossRef][Medline] Submitted November 15, 2005; accepted March 22, 2006. This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 test. The change in the GQOL score from baseline to 6 months (eight cycles) was estimated from a time factor mean model, where each assessment time is treated as an independent factor. For the primary pain and GQOL analyses, missing data were assumed to be missing at random (MAR). A sensitivity analysis was performed for GQOL comprising a series of four pattern mixture models corresponding to four alternative potential informative missing mechanisms. A sensitivity analysis is performed because there is no empirical test for the type of missing data mechanism. A result is considered significant and robust if it is statistically significant under the MAR analysis and all models in the sensitivity analysis. The pattern mixture models estimate different QOL patterns for different patterns of attrition.
