|
|||||
|
|
||||||
© 2003 American Society for Clinical Oncology Phase III Randomized Trial of 12 Versus 3 Months of Maintenance Paclitaxel in Patients With Advanced Ovarian Cancer After Complete Response to Platinum and Paclitaxel-Based Chemotherapy: A Southwest Oncology Group and Gynecologic Oncology Group TrialFrom the Cleveland Clinic Foundation, Cleveland; Ohio State University Health Center, Columbus, OH; Southwest Oncology Group Statistical Center, Seattle, WA; City of Hope National Medical Center, Duarte; University of California Irvine Medical Center, Orange, CA; University of Alabama Medical Center at Birmingham, Birmingham, AL; and University of Arizona Cancer Center, Tucson, AZ. Address reprint requests to Maurie Markman, MD, The Cleveland Clinic Foundation (R35), 9500 Euclid Ave, Cleveland, OH 44195; email: markmam{at}ccf.org.
Purpose: To determine whether continuing paclitaxel for an extended time period in women with advanced ovarian cancer who had achieved a clinically defined complete response to a platinum/paclitaxel-based chemotherapy could prolong subsequent progression-free survival (PFS) and affect ultimate survival. Patients and Methods: Patients were randomly assigned to either three or 12 cycles of single-agent paclitaxel administered every 28 days and were then followed up for progression-free and overall survival. Results: As of September 6, 2001, 277 patients (262 assessable) had entered the trial, with a total of 54 PFS events having developed among 222 patients with follow-up data. With the exception of peripheral neuropathy, there were no major differences in toxicity between the regimens. The median PFS was 21 and 28 months in the three-cycle and 12-cycle paclitaxel arms, respectively. One-sided P values from an unadjusted log-rank test and an adjusted Cox model analysis (for stratification factors) were .0035 and .0023, respectively, both in favor of the 12-cycle arm. The Cox model-adjusted three-cycle versus the 12-cycle progression hazard ratio was estimated to be 2.31 (99% confidence interval, 1.08 to 4.94). With a protocol-specified early termination boundary of P = .005, these findings led the Southwest Oncology Group Data Safety Monitoring Committee to discontinue the trial. As of the date of study closure, there was no difference in overall survival between the treatment arms. Conclusion: Twelve cycles of single-agent paclitaxel administered to women with advanced ovarian cancer who attain a clinically defined complete response to initial platinum/paclitaxel-based chemotherapy significantly prolongs the duration of PFS.
STANDARD INITIAL therapy for ovarian cancer is generally considered to include five to six courses of a platinum/taxane regimen,1,2 with limited randomized trial data failing to reveal any advantage for additional treatments or a consolidation strategy either in ovarian cancer35 or other malignancies.6,7 However, nonrandomized experience in ovarian cancer has suggested the possible clinical utility for a more protracted treatment program when paclitaxel, a cycle-specific antineoplastic agent, is used.810 In one evaluation, a subset of patients with heavily pretreated disease survived for more than 3 years after the initiation, and continuation until disease progression, of single-agent paclitaxel.9 It might be hypothesized that more prolonged treatment with this drug in patients who had previously demonstrated major inherent chemosensitivity to a front-line treatment program, which included paclitaxel, could favorably affect both time to subsequent disease progression and overall survival. To test this hypothesis, the Southwest Oncology Group (SWOG) and the Gynecologic Oncology Group (GOG) initiated a trial of single-agent paclitaxel in women with advanced ovarian cancer who had achieved a clinically defined complete response to front-line platinum/paclitaxel chemotherapy.
Eligibility Criteria included the following: (1) histologically confirmed stages III/IV epithelial ovarian cancer, fallopian-tube cancer, or primary carcinoma of the peritoneum; (2) treatment with five to six cycles of a platinum/paclitaxel regimen; and (3) attainment of a clinically defined complete response (ie, no cancer-related symptoms; normal physical examination, computed tomography scan of the abdomen/pelvis and chest x-ray; CA-125 level ≤ 35 U/mL). Patients who had received prior treatment on a SWOG or GOG front-line chemotherapy regimen where either progression-free or overall survival were study end points or had ≥ grade 2 persistent neuropathy from prior therapy were ineligible.
Treatment Plan
Treatment Modifications Recovery from all toxicities to ≤ grade 1 was required before a subsequent cycle was delivered. The maximum permitted treatment delay for toxicity was 2 weeks. Persistent episodes of grade 4 neutropenia that occurred despite mandated dose reductions and prophylactic use of granulocyte-colonystimulating factor or ≥ grade 3 neuropathy resulted in treatment discontinuation.
Evaluation
Statistical Considerations Interim analysis was to be conduced when 225 eligible patients had been entered and again when 50% and 75% of the anticipated events had been observed. The interim analysis was conducted to guard against extreme findingseither excessive toxicity or a substantial improvement in efficacy. Evidence suggesting early termination of the study would be if the null hypothesis of no difference in PFS, or the alternative hypothesis that the progression hazard ratio of the control arm over the prolonged paclitaxel arm is 1.33 or greater, were rejected at the .005 level. The actual decision to terminate the study early would be made by the SWOG Data and Safety Monitoring Committee and would take into consideration other factors, such as overall survival, toxicities, and complications. Stratification factors used for randomization in this trial included (1) prior treatment with paclitaxel for greater than ≥ 24 hours versus less than 24 hours; (2) initial presentation with optimal stage III versus suboptimal stage III versus stage IV ovarian cancer (for purposes of this study, suboptimal disease was defined as residual disease after exploratory laparotomy with at least one lesion ≥ 1 cm in its greatest dimension); and (3) age ≤ 65 versus more than 65 years.
Criteria for Evaluation and End Point Definitions
Patient Characteristics As of September 6, 2001, a total of 277 patients (15 ineligible) had entered into this trial (Table 1
Protocol Modification for Neuropathy In March 2001, a review of toxicity revealed that more patients were electing to discontinue treatment early on the 12-cycle arm (n = 10) compared with the three-cycle arm (n = 3). The reason for early treatment discontinuation was the development of both grade 2 and grade 3 peripheral neuropathy. Because the aim of the study was to compare treatment duration and not the total dose administered, the paclitaxel dose was reduced as of this date in both treatment arms to 135 mg/m2. For patients experiencing excessive toxicity (as previously defined), the paclitaxel dose was to be reduced to 100 mg/m2. (However, because of the early closure of the study by the SWOG Data and Safety Monitoring Committee, all patients reported in this article had treatment initiated at the 175 mg/m2 dose level.)
Toxicity
Survival The protocol stipulated an interim analysis would be performed at 50% enrollment; this was done in September 2001. At that time, there were 54 PFS events among 222 patients with follow-up data34 on the three-cycle arm and 20 on the 12-cycle arm.
An unadjusted log-rank test and a Cox model analysis adjusting for randomization stratification factors revealed the one-sided P values from the unadjusted test and the adjusted Cox model analysis were .0035 and .0023, respectively, in favor of the 12-cycle arm (Fig 1
As of September 2001, there were no significant differences in overall survival, with only 17 deaths in the study. The one-sided P values were .63 (unadjusted) and .70 (adjusted). The median follow-up of surviving patients is currently 8.5 months (range, 0 to 37 months). A natural question that arises based on these data is whether the rate of subsequent disease progression increased shortly after treatment was discontinued. An exploratory analysis was performed to assess whether the current data provide any support for this hypothesis. A treatment-stop indicator was added to the Cox model, which also included treatment and stratification factors, as before. Unlike the baseline stratification factors, the values of which do not change over time within a patient, the value of the treatment-stop indicator is time dependent. Specifically, this indicator starts out as 0 for each patient but changes to 1 at 28 days after treatment stopped. The change point of 28 days posttreatment was chosen because that is when the next cycle of paclitaxel would have been given had treatment continued. For those patients whose disease progressed during treatment (three patients on the three-cycle arm and seven patients on the 12-cycle arm), the value of this indicator remains as 0 throughout. The results showed the treatment-stop indicator to be highly significant (one-sided P = .0004), suggesting a higher hazard for disease progression after treatment stopped. The estimated hazard ratio posttreatment versus during treatment was 5.80, with a 95% confidence interval of 2.09 to 16.10. Treatment with 12-cycle versus 3-cycle paclitaxel became nonsignificant (hazard ratio = 1.07, one-sided P = .43). This is partially due to the fact that the treatment-stop indicator is inherently confounded with treatment assignment, which differed in duration. The analysis was then repeated in the 12-cycle arm only, and the treatment-stop indicator was equally significant (one-sided P = .0005) with an estimated post versus during treatment progression hazard ratio of 9.17, and a 95% confidence interval of 2.47 to 34.0. In both cases, the wide confidence intervals reflect the relatively limited number of events in the data set.
For a crude description of the observed progression risk, the actual event rates by three-month intervals during the first 2 years are presented in Table 3
A number of issues are raised by the provocative results of this multicenter randomized phase III trial. First, it might be asked how we determined that the control arm of this study would be three monthly cycles of single-agent paclitaxel, because before the conduct of this trial, there was no convincing evidence to demonstrate the clinical utility of any maintenance or consolidation strategy in ovarian cancer. The decision to use a short-course (three cycles) paclitaxel control arm, rather than a no-further-treatment control, was made for both practical and theoretical considerations. Despite the absence of definitive randomized controlled trial data demonstrating any value for consolidation or maintenance therapy in women with ovarian cancer who achieved a clinically defined complete response, there was concern that when women were presented with the justification for the conduct of this trial, few would accept randomization to a no-treatment control. This assessment resulted from the fact that patients considered for entry onto this trial would need to be told the following: (1) approximately 75% of patients in this clinical setting ultimately experience recurrence of their cancer1,2,12; (2) if disease recurs, there is no evidence for the curative potential of any second-line regimen; (3) the drug they would receive in the trial (paclitaxel) was a component of the regimen to which they had already attained a clinical complete response and tolerated (otherwise, they would not be a candidate for the trial); (4) there is currently evidence that paclitaxel can be given safely for extended periods9; and (5) a reasonable rationale could be presented to support the potential clinical utility of extending the duration of paclitaxel administration (eg, exposure of tumor during active cell cycling13; antiangiogenetic effect associated with continued intermittent drug exposure14). Finally, because paclitaxel was commercially available when this trial was initiated, a patient presented with this information could simply elect to ask her oncologist to administer additional cycles of the agent. However, as the study was ultimately designed, it could be rationally (and ethically) argued that even patients randomized to a three-cycle control arm would be receiving some additional treatment beyond that known to be associated with an extremely high recurrence rate. Despite the absence of a pure evidenced-based control arm in this randomized trial, the investigators recognized that the truly important question being addressed in the study was whether an extended paclitaxel treatment regimen (eg, 12 monthly cycles) would produce a favorable clinical effect. We reasoned that if 12 courses of monthly paclitaxel were not shown to be superior to three courses, it would be most appropriate to conclude that we had failed to demonstrate the benefit of any extension of the duration of paclitaxel treatment in this clinical setting. A second question to be addressed is whether it was appropriate to consider PFS to be the primary end point in this trial, rather than overall survival. Although the ultimate goal of any antineoplastic treatment should be its impact on survival, the investigators believed that prolongation of PFS, resulting from a treatment program that does not cause a major impairment in overall quality of life, is a highly clinically relevant goal, particularly in a disease such as ovarian cancer, where multiple second-line treatment programs may affect ultimate survival independent of what has occurred during initial treatment.15 While this study, initiated in 1997, did not include a formal quality-of-life assessment, it would be difficult to argue this extended paclitaxel program is associated with an unacceptable side-effect profile. For example, the incidence of significant neurotoxicity (grade 2/3) observed was comparable to that noted when a platinum agent (cisplatin or carboplatin) plus paclitaxel are used as initial chemotherapy for this malignancy.1,12,16,17 This degree of neurotoxicity is generally believed to be acceptable in routine clinical practice. Lowering the dose of paclitaxel in any future study and in standard clinical practice should make the issue of peripheral neuropathy less of a concern. Moreover, because it is far more likely that the observed improvement in PFS resulted from the extended treatment schedule, rather than the specific dose of paclitaxel used in the trial, it is likely that reducing the dose of paclitaxel should enhance the therapeutic ratio associated with this management strategy (ie, no change in efficacy, but reduced toxicity). However, it must be noted that the current study provides no data on either the efficacy or toxicity associated with initiating maintenance paclitaxel at a lower dose level (eg, 135 mg/m2). Third, it might be asked if the SWOG Data and Safety Monitoring Committee might have closed this study prematurely, particularly considering the likely negative impact on our ultimate ability to evaluate the effect of this treatment program on overall survival. This is due to the fact that patients currently on the three-cycle arm, when informed of the results of the study, may elect to be treated with a total of 12 cycles. Furthermore, patients who were randomized to the three-cycle arm, have recently completed therapy and have not as yet experienced disease progression may also request additional courses of paclitaxel. Despite this concern, it would be inappropriate for the investigators to comment on the decision of this federally mandated review panel, whose sole purpose is to independently evaluate the results of ongoing studies and determine whether a trial should be discontinued based on excessive toxicity or extremely favorable or unfavorable results.18 We must accept this decision, regardless of its potential impact on the ultimate study results or other ongoing or future clinical trials. Fourth, it is important to question the potential implications of the provocative results of the exploratory analysis examining the rate of disease progression after discontinuation of consolidation chemotherapy. Acknowledging the limitations associated with the small sample size and the observational nature of the analysis, this evaluation revealed a highly statistically significant acceleration in the risk of disease progression shortly after the completion of treatment. Although it might be argued that this result is obvious and may be explained simply by the fact patients will ultimately progress with the passage of time independent of the continuation (or discontinuation) of therapy, the strikingly similar findings observed in both the three-cycle and 12-cycle study arm and the contrast in progression rates between the two arms during the first year are consistent with the role of the maintenance program in at least delaying disease progression. To address the concern that patients discontinuing therapy before the planned completion of their treatment program (eg, toxicity, refusal) may have actually done so because of unrecognized symptoms of disease progression, we examined the ultimate date of documented recurrence for all such patients. Although several patients removed from study early did experience clinical progression, these events occurred many months after discontinuation of treatment (data not shown). This evaluation leads us to doubt the conclusion that the results of this exploratory analysis were substantially influenced by patients experiencing unrecognized disease progression while on active therapy. It is therefore interesting to speculate that continuation of single-agent paclitaxel for more prolonged time periods (eg, 2 years, 3 years) may exert an even greater impact on time to disease progression and possibly ultimate survival. A mechanism for the activity of paclitaxel, when delivered in this clinical setting, remains to be elucidated and might include the persistent chemosensitivity of slowly dividing tumor cells to the cycle-specific agent13 or an antiangiogenic effect associated with intermittent drug exposure.14 The clinical utility of such a prolonged maintenance strategy using single-agent paclitaxel is worthy of investigation in a randomized controlled phase III trial. However, it should be noted that in addition to the issue of the impact of such a management approach on quality of life (eg, neurotoxicity, continued alopecia), there is the concern for the development of additional adverse effects that have been observed when other cytotoxic agents are delivered over extended time periods (eg, secondary acute leukemia).19 Finally, it must be asked whether the standard of care for all patients treated for advanced ovarian cancer has been changed as a result of this trial. For several reasons it would be inappropriate to provide a definitive affirmative answer to this question. First, as of this date, there is no evidence that the 12 additional monthly cycles of single-agent paclitaxel favorably influences overall survival. In addition to the concern noted above regarding the effect of early closure on the studys final results, a variety of second-line treatment regimens may impact overall survival in this malignancy independent of the initial chemotherapy strategy.15 Second, because entry into this trial required patients to be in a clinically defined complete response and without significant preexisting peripheral neuropathy, it would be inappropriate to generalize regarding the potential efficacy and toxicity of this management approach in those who did not meet these precise clinical characteristics (eg, responsive but documented persistent disease or the presence of treatment-related grade 2 neuropathy). In other patient populations, the continuation of paclitaxel, as used in this trial, may be ineffective or result in excessive toxicity. However, despite these concerns and recognized study limitations, it is most reasonable to conclude that oncologists should discuss with appropriate patients the results of this trial and the possible implications for their subsequent management. When presented with the evidence of a favorable impact on PFS, some patients may elect to continue paclitaxel, with its associated treatment-related toxicity, whereas others may decide to stop therapy when the initial front-line platinum-paclitaxel regimen is complete. Finally, the highly provocative results of this study should stimulate further investigation of the role of a variety of consolidation/maintenance antineoplastic strategies (including both cytotoxic and cytostatic agents) in advanced ovarian cancer. This trial, although not establishing a new standard of care in the management of ovarian cancer, has demonstrated the potential for a maintenance strategy to have a considerable impact on progression of the disease process. Thus it would perhaps be most appropriate to consider the trial to serve as a critically important proof of principle that should be built on in future studies. In ovarian cancer, a malignancy increasingly characterized by prolonged survival, it is appropriate to argue that substantial lengthening of the time to subsequent disease progression is a worthy goal of antineoplastic therapy if it can be convincingly shown that such treatment is not associated with toxicity of such severity that it negates the benefits of the therapeutic program.
1. McGuire WP, Hoskins WJ, Brady MF, et al: Cyclophosphamide compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med 334:16, 1996
2. Piccart MJ, Bertelsen K, James K, et al: Randomized intergroup trial of cisplatin-paclitaxel versus cisplatin-cyclophosphamide in women with advanced epithelial ovarian cancer: Three-year results. J Natl Cancer Inst 92:699708, 2000 3. Hakes TB, Chalas E, Hoskins WJ, et al: Randomized prospective trial of 5 versus 10 cycles of cyclophosphamide, doxorubicin, and cisplatin in advanced ovarian carcinoma. Gynecol Oncol 45:284289, 1992[CrossRef][Medline] 4. Bertelsen K, Jakobsen A, Stroyer I, et al: A prospective randomized comparison of 6 and 12 courses of cyclophosphamide, Adriamycin, and cisplatin in advanced epithelial ovarian cancer: A Danish Ovarian Study Group trial (DACOVA). Gynecol Oncol 49:3036, 1993[CrossRef][Medline]
5. Lambert HE, Rustin GJS, Gregory WM, et al: A randomized trial of five versus eight courses of cisplatin or carboplatin in advanced epithelial ovarian carcinoma. Ann Oncol 8:327333, 1997 6. Einhorn LH, Williams SD, Loehrer PJ, et al: Evaluation of optimal duration of chemotherapy in favorable-prognosis disseminated germ cell tumors: A Southeastern Cancer Study Group protocol. J Clin Oncol 7:387391, 1989[Abstract] 7. Early Breast Cancer Trialists Collaborative Group: Polychemotherapy for early breast cancer: An overview of the randomized trials. Lancet 352:930942, 1998[CrossRef][Medline]
8. Einzig AI, Wiernik PH, Sasloff J, et al: Phase II study and long-term follow-up of patients treated with Taxol for advanced ovarian adenocarcinoma. J Clin Oncol 10:17481753, 1992
9. Markman M, Hakes T, Barakat R, et al: Follow-up of Memorial Sloan-Kettering Cancer Center patients treated on National Cancer Institute Treatment Referral Center Protocol 9103: Paclitaxel in refractory ovarian cancer. J Clin Oncol 14:796799, 1996 10. Zanotti KM, Belinson JL, Kennedy AW, et al: Treatment of relapsed carcinoma of the ovary with single-agent paclitaxel following exposure to paclitaxel and platinum employed as initial therapy. Gynecol Oncol 79:211215, 2000[CrossRef][Medline]
11. Alberts DS, Liu PY, Hannigan EV, et al: Intraperitoneal cisplatin plus intravenous cyclophosphamide versus intravenous cisplatin plus intravenous cyclophosphamide for stage III ovarian cancer. N Engl J Med 335:19501955, 1996
12. Muggia FM, Braly PS, Brady MF, et al: Phase III randomized study of cisplatin versus paclitaxel versus cisplatin and paclitaxel in patients with suboptimal stage III and IV ovarian cancer: A Gynecologic Oncology Group study. J Clin Oncol 18:106115, 2000
13. Rowinsky EK, Donehower RC: Paclitaxel (Taxol). N Engl J Med 332:10041014, 1995
14. Browder T, Butterfield CE, Kraling BM, et al: Antiangiogenic scheduling of chemotherapy improves efficacy against experimental drug-resistant cancer. Cancer Res 60:18781886, 2000
15. Gordon AN, Fleagle JT, Guthrie D, et al: Recurrent epithelial ovarian carcinoma: A randomized phase III study of pegylated liposomal doxorubicin versus topotecan. J Clin Oncol 19:33123322, 2001
16. Neijt JP, Engelholm SA, Tuxen MK, et al: Exploratory phase III study of paclitaxel and cisplatin versus paclitaxel and carboplatin in advanced ovarian cancer. J Clin Oncol 18:30843092, 2000 17. Ozols RF, Bundy BN, Fowler J, et al: Randomized phase III study of cisplatin/paclitaxel versus carboplatin/paclitaxel in stage III epithelial cancer: A Gynecologic Oncology Group trial. Proc Am Soc Clin Oncol 18:356a, 1999 (abstr 1373)
18. Smith MA, Ungerleider RS, Korn EL, et al: Role of independent data-monitoring committees in randomized clinical trials sponsored by the National Cancer Institute. J Clin Oncol 15:27362743, 1997 19. Greene MH, Boice JD, Greer BE, et al: Acute nonlymphocytic leukemia after therapy with alkylating agent for ovarian cancer: A study of five randomized clinical trials. N Engl J Med 307:14161421, 1982[Abstract] Submitted July 2, 2002; accepted January 24, 2003.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|