|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2001 American Society for Clinical Oncology
Medical Research Council Adjuvant Trial in High-Grade Gliomas
University of Southern California, Los Angeles, CA To the Editor:The Medical Research Council (MRC) Brain Tumor Working Party has reported a phase III comparative trial of radiotherapy with adjuvant versus no adjuvant chemotherapy with procarbazine, lomustine, and vincristine (PCV) in patients with high-grade gliomas.1 At the time of study design, PCV was considered to be potentially the best standard of care, although the superiority of this regimen over single-agent nitrosourea has recently been questioned.2 The results of the MRC trial indicated no significant differences in survival between the two arms. The authors conclude that ". . . no-chemotherapy control arms remain ethical in randomized trials in high-grade astrocytoma and that it is reasonable and appropriate not to offer routine, adjuvant chemotherapy to individual patients outside trials." We, however, suggest that caution be observed before one makes generalizations from these results to the treatment of individual patients, or even necessarily incorporate these data in the design of future trials of high-grade glioma. We therefore offer the following comments regarding the MRC trial. First, conflicting results from earlier trials are not necessarily negated by observations on a newer trial. Of particular interest in the MRC trial is the observation that there was no statistical correlation of survival with the accepted prognostic variables of age, performance status, histologic grade, or extent of resection by proportional hazards regressional analysis. In prior prospective trials, these specific prognostic factors have consistently been shown to represent independent variables in multivariate analyses that correlate significantly with survival.3-6 The authors conclude that in their trial ". . . the results. . . do not directly contradict the results of any of the 14 adequately randomized trials evaluating nitrosourea-based chemotherapy that have been published to date." However, the lack of correlation in the MRC trial of survival with these known prognostic variables is in contrast to the prior studies mentioned. At a minimum, this raises the possibility that the MRC patient population seemed to have varied from those in prior reported trials, based on unknown biologic factors or the unintentional introduction of additional bias.7 A second significant difference in this trial is the relatively poor outcome of the anaplastic astrocytoma subset, which in the MRC trial had a median survival of 13 months. This is unusually low in comparison with prior reported trials and actually is similar to that reported for glioblastoma. In contrast, a recent prospective randomized Radiation Therapy Oncology Group trial, with a larger cohort of anaplastic astrocytoma patients than in the MRC trial, involved a similar therapy (radiotherapy with PCV ± bromodeoxyuridine).8 In this latter trial, treatment produced a median overall survival of 45 months. Because the anaplastic astrocytoma subset in the MRC study constituted nearly 20% of the study population, this finding also raises the question of whether the MRC population was in some way different from other populations reported previously. Third, there are additional differences in the MRC trial as compared with prior reported trials, which may have had an effect on the outcome in the MRC trial. Eligibility parameters were vague and often defaulted to the standard of practice and, as such, contain a greater degree of subjectivity.9 Comorbid medical conditions are not reported but, if present in a large number of patients, could have influenced survival. The interval from surgery to initial treatment, a parameter that has been shown to be an important prognostic indicator, was not given. Surgery was comparatively nonaggressive, as reflected by the majority of patients undergoing biopsy or partial resections not otherwise radiographically quantified. It is mentioned that "radiotherapy was to commence within 6 weeks of surgery," but it is not clear how many patients received radiotherapy within that time frame. Radiotherapy schedules were not standardized, and total prescribed dose varied between 45 and 60 Gy. Brain imaging was not routinely performed either to assess extent of surgical resection or response to therapy. Nineteen (6%) of the 335 patients randomized to the chemotherapy PCV arm did not receive any chemotherapy, which potentially could influence outcome on that arm, and the dose-intensity of PCV in the remaining PCV patients is not detailed. Most patients were not retreated at recurrence, which may be one explanation for why survival seemed shorter in this trial. Furthermore, twice as many patients randomized initially to the radiotherapy-only arm received subsequent chemotherapy at recurrence, which potentially may have influenced ultimate survival differences between the two groups. Finally, survival is defined in the MRC study as "to death date or date last known to be alive." It is not stated how many patients fit into the latter category, which, if weighted heavily on one of the two arms, could influence survival. There is no argument that adjuvant chemotherapy has had a very modest impact on survival of high-grade glioma patients, and such marginal benefit has to be considered in view of other chemotherapy-related factors such as toxicity, effects on quality of life, and cost.10 However, it is also important to consider the results of earlier trials that show benefit of adjuvant chemotherapy, albeit modest and perhaps limited to selected subgroups (eg, younger patients with anaplastic astrocytoma). The differences observed in the MRC trial population compared with previously reported trials does not indicate that the MRC results are wrong, but instead may be reflective of differences in the treatment and design of the treated study population. This outcome may be influenced by multiple factors, some which may not be readily apparent. However, the results of the MRC trial, in our opinion, do not justify the conclusion that it is reasonable to not offer chemotherapy to these patients or to justify inclusion of a no-chemotherapy arm in clinical trials. REFERENCES
1.
The Medical Research Council Brain Tumour Working Party: Randomized trial of procarbazine, lomustine, and vincristine in the adjuvant treatment of high-grade astrocytoma: A Medical Research Council trial. J Clin Oncol 19: 509-518, 2001
2.
Prados MD, Scott C, Curran WJ, et al: Procarbazine, lomustine, and vincristine (PCV) chemotherapy for anaplastic astrocytoma: A retrospective review of Radiation Therapy Oncology Group protocols comparing survival with carmustine or PCV adjuvant chemotherapy. J Clin Oncol 17: 3389-3395, 1999
3.
Curran WJ, Scott CB, Horton J, et al: Recursive partitioning analysis in three Radiation Oncology Therapy Group malignant gliomas trials. J Natl Cancer Inst 85: 704-710, 1993 4. Wood JR, Green SB, Shapiro WR, et al: The prognostic importance of tumor size in malignant gliomas: A computed tomographic scan study by the Brain Tumor Cooperative Group. J Clin Oncol 6: 338-343, 1988[Abstract] 5. Prados MD, Russo C: Chemotherapy of brain tumors. Semin Surg Oncol 14: 88-95, 1998[Medline] 6. Galanis E, Buckner JC: Chemotherapy of brain tumors. Curr Opin Neurol 13: 619-625, 2000[Medline] 7. Wong TW, Hess KR, Gleason MJ, et al: Outcomes and prognostic factors in recurrent glioma patients enrolled onto phase II clinical trials. J Clin Oncol 71: 2572-2578, 1999 8. Prados MD, Gutin PH, Phillips TL, et al: Highly anaplastic astrocytoma: A review of 357 patients treated between 1977 and 1989. Int J Radiat Oncol Biol Phys 23: 3-8, 1992[Medline]
9.
Perry JR, DeAngelis LM, Schold SC Jr, et al: Challenges in the design and conduct of phase III brain tumor therapy trials. Neurology 49: 912-917, 1997 10. Levin VA, Silver P, Hannigan J, et al: Superiority of post-radiotherapy adjuvant chemotherapy with CCNU, procarbazine, and vincristine (PCV) over BCNU for anaplastic gliomas: NCOG 6G61 final report. Int J Radiat Oncol Biol Phys 18: 321-324, 1990[Medline]
Response
Medical Research Council Clinical Trials Unit, London, United Kingdom In Reply:Drs Chamberlain and Jaeckle raise a number of points concerning the Medical Research Council BR5 trial to which we are happy to respond. Their primary concern was that the BR5 trial population differed from previous trials with respect to the prognostic value of factors, such as age and performance status, and the median survival of the anaplastic astrocytoma subgroup. With respect to the first of these, the authors state "there was no statistical correlation of survival with the accepted prognostic variables of age, performance status, histologic grade, or extent of resection by proportional hazards regression analysis." In fact we did not present any such analyses and believe that this statement comes from a misunderstanding of the hazards ratio plots. These represented the results of the subgroup analyses, which we reported as showing "no evidence of a differential treatment effect across any of these factors." This does not mean that the factors themselves are not prognostic; certainly tumor grade, age, and performance status are prognostic factors in our trial population. It does mean they are not predictive of response to chemotherapy; ie, they do not define subgroups in which the effect of chemotherapy, as judged by the treatment hazards ratio, is greater or less than in other subgroups defined by the same factor. Thus, for example, younger glioma patients live longer than older patients regardless of whether or not they receive chemotherapy (ie, age is prognostic); however, we have found no evidence that any additional benefit that younger patients might gain through receiving chemotherapy is greater or less than any additional benefit that older patients might gain (ie, age is not predictive). Concerning the outcome of the anaplastic astrocytoma (AA) patients, it is very difficult to compare outcomes across trials without data on the distribution of prognostic factors such as age and performance status within the AA populations. In addition, a more aggressive approach to surgery for both low- and high-grade tumors may well lead to apparently improved survival in each grade subgroup, without any overall change in survival, simply through stage/grade migration. For example, centers with a policy of resection rather than biopsy for tumors that are considered low-grade radiologically are likely to identify a higher proportion of patients with anaplastic elements. These patients are then classified as having AA (and entered onto appropriate trials) but are likely to have a better prognosis than those with a high-grade appearance radiologically, who may comprise the bulk of AA patients in other trials. Such variation is inherent in any cross-trial comparison. One further point we would make is that median survival is an isolated point on the survival curve, and just as one would not compare treatment results by focusing on a single time point but rather by comparing the whole survival curves, so one should apply the same principle here. Thus although our AA and glioblastoma multiforme populations differed by only a few months with respect to median survival, survival rates in the two groups beyond this point (ie, after 1 year) are systematically substantially different as illustrated in Fig 4 of the study report. Regarding trial entry criteria, these were indeed quite broad to reflect practice outside of the trial setting and included all patients in whom there was the possibility that chemotherapy may be beneficial (thus those with comorbid conditions that would have made treating or assessing response to treatment impossible were excluded). Although potentially this may mean that a higher proportion of poorer (or indeed better) prognosis patients entered the trial than might otherwise have been the case, these patients would have been equally distributed across the trial arms by randomization. In addition, the subgroup analyses referred to above indicate that the overall negative result of the trial is not simply explained by poor chemotherapy response in the poorer prognosis patients; we were unable to identify any subgroups in which chemotherapy substantially improved survival. Regarding time from surgery to start of radiotherapy, the median interval was 3 weeks; 90% started within 6 weeks and 95% within 7 weeks; the remaining patients incurred additional short delays for a variety of practical and personal reasons. We found no evidence that chemotherapy was more beneficial in patients with a short interval between surgery and the start of radiotherapy. As reported in the article, the median number of cycles received appears comparable with previous trials, but most trials report insufficient data to compare dose-intensity. With regard to treatment at recurrence, the re-treatment rate was low, although the extent to which overall survival may have been extended had the rate been higher is perhaps debatable. Certainly the consistency of the trial results with respect to progression-free and overall survival suggest that cross-over in the radiotherapy arm did not have a major influence on the overall survival results. Finally, with regard to the number of patients followed-up not to death but to date last known to be alive, these comprised less than 10% of the total randomized population, and the numbers according to treatment arm are given in Fig 1 of the trial report; 29 of the radiotherapy patients and 28 of the patients who received radiotherapy plus procarbazine, lomustine, and vincristine were still alive at the time of the final analysis, with a median follow-up for survivors of approximately 4 years in each group. It is clear that any benefit to chemotherapy is indeed marginal, a statement supported not just by BR5 but by the most recent quantitative summary of the worldwide evidence.1 Equally, it is clearly wholly appropriate that patients and their physicians should make decisions regarding the need for chemotherapy on an individual patient basis and that some will choose to have chemotherapy even in the absence of unequivocal evidence of survival gain. This is entirely compatible with our conclusion that it is currently reasonable and appropriate not to offer routine adjuvant chemotherapy to individual patients outside of clinical trials. REFERENCES 1. Stewart LA, Burdett S, Souhami R: Chemotherapy in high-grade glioma: A meta-analysis using individual patient data from randomized clinical trials (RCTs)—The Glioma Meta-Analysis Trialist Group. Proc Am Soc Clin Oncol 19: 168a, 2000 (abstr 650)
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2001 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
