|
Maintenance paclitaxel after a complete response to initial therapy did not prolong survival of women with advanced ovarian cancer in a phase III trial.
No differences in progression-free survival or overall survival were observed between women randomized to observation and those given six courses of maintenance paclitaxel in the After-6 protocol 1 trial. The study enrolled women with advanced ovarian cancer that was in complete response to first-line platinum or paclitaxel chemotherapy.
"The results of this trial do not support the administration of maintenance paclitaxel to patients with carefully defined clinical or pathological complete response," said lead investigator Dr. Pier F. Conte of the University of Modena and Reggio Emilia, Modena, Italy.
This is the fifth randomized trial to evaluate maintenance or consolidation chemotherapy in ovarian cancer, and only one has shown a positive result (J. Clin. Oncol. 2003;21:2460-5). Markman et al. reported that progression-free survival was significantly improved with 12 versus 3 cycles of maintenance paclitaxel.
"There are now four of five negative trials in ovarian cancer, so I don't know if we really need a meta-analysis," commented discussant Dr. William P. McGuire, director of the Weinberg Cancer Institute, Franklin Square Hospital Center, Baltimore.
Dr. McGuire hailed the After-6 trial for defining the increase in neurotoxicity (28%) associated with six cycles of paclitaxel, something not reported in the longer Markman trial. "That was just with 6 cycles, so I can imagine with 12 cycles it would be significantly more," he said.
Between March 1999 and May 2006, 200 women with stages IIb-IV ovarian cancer in clinical or pathologic complete response after six courses of first-line platinum or paclitaxel chemotherapy were randomized to observation or to receive six cycles of intravenous paclitaxel 175 mg/m² every 3 weeks.
At admission, 47.5% of the women had a clinical response to first-line therapy as defined by a negative pelvic exam plus negative ultrasonography or CT scan and a CA-125 level less than 35 units since at least two courses. In addition, 52.5% of women had a pathologic response as defined by a clinical response plus negative peritoneal cytology, and at least eight negative random biopsies.
The median age was 59 years for 101 patients who received paclitaxel and 58 years for 99 patients in the control group. Overall, 63.5% of women had grade 3 tumors, 62.5% had stage IIIc disease, and 52.5% had no macroscopic disease after their first surgery.
As of March 30, 2007, 107 patients (53.5%) had relapsed and 48 patients (24%) had died.
Compliance with therapy was good, with 77% of patients completing the six cycles. Grade 3 sensory neuropathy was reported in 6.7% of treated patients and grade 2 in 21.3%, said Dr. Conte.
The study's primary outcome of median progression-free survival was 34 months in the paclitaxel group and 30 months in the control group. Two-year progression-free survival was not statistically different between groups (59% vs. 54%, P = .068).
Median overall survival was 77 months in the paclitaxel maintenance arm. It has not been reached in the control arm. Two-year overall survival was not statistically different between groups (87% vs. 90%, P = .13).
Because 14 patients randomized to observation received maintenance paclitaxel, the investigators also performed a per protocol analysis. Again, progression-free survival and overall survival were not significantly different between groups, Dr. Conte said.
In a Cox proportional hazard model, study arm, tumor grading, and clinical or pathologic response were not independent predictors of progression-free survival, while tumor stage and residual disease were. Women with stage IIIc-IV tumors were three times more likely to progress (hazard ratio 3.10) than were those with lower-stage tumors, while women with macroscopic residual disease were nearly twice as likely to progress (HR 1.91) compared to those with no residual disease.
The study was designed to enroll 250 patients, but accrual was stopped after an unplanned interim analysis showed the probability of observing a 15% improvement in 2-year progression-free survival was less than 5%. The small size of the study and poor compliance in the control group were weaknesses of the study, Dr. McGuire observed.
Conte P.F. et al. Final results of After-6 protocol 1: A phase III trial of observation versus 6 courses of paclitaxel (Pac) in advanced ovarian cancer patients in complete response (CR) after platinum-paclitaxel chemotherapy (CT). Abstract 5505.
Commentary |
Only one previous randomized trial has shown a benefit for the approach in this trial by the After 6 Italian Cooperative Group. An earlier intergroup study by the Southwest Oncology Group (SWOG) and the Gynecologic Oncology Group (GOG) demonstrated a highly significant difference in median progression-free survival (22 vs. 14 months, P = .0023) favoring 12 additional cycles of maintenance paclitaxel in advanced-disease patients in a clinical complete response at the end of initial paclitaxel/platinum therapy. That trial also reported a nonsignificant difference in overall survival (53 vs. 46 months, P = .30). The temptation here is to conclude that the After 6 trial refutes the earlier SWOG-GOG results. Several major differences and one particular similarity between the two studies should be considered before such a conclusion is reached. First, the populations are different. The After 6 trial included patients with a substantially better outlook, as is evidenced by the overall progression-free survival of 34 months, which is longer than that seen in GOG intraperitoneal studies of patients with small volume disease (28 months being the longest in those trials). Half of the patients in the After 6 study had no macroscopic disease at the conclusion of their initial laparotomy, and only 20% had macroscopic disease greater than 2 cm. Furthermore, over half of the patients in the After 6 trial were documented to be in a pathologic complete response at the end of initial therapy. Second, this study used only six additional cycles of paclitaxel, whereas the SWOG-GOG trial gave 12 additional cycles to the experimental arm. The After 6 trial may actually demonstrate only that six additional cycles are not enough. At the very least, these results show that the oft-heard criticism of the SWOG-GOG trial's not having a control arm is a hollow criticism. If there is no difference between zero and six cycles, then there certainly is no difference between zero and three cycles. The three-cycle arm of the SWOG-GOG trial is, thus, a valid control arm. Finally, both this trial and the SWOG-GOG study suggest a much greater effect of maintenance therapy in those with excellent initial responses. In the group with a pathologic complete response to initial therapy, the hazard ratio was 0.79 favoring the six cycles of maintenance paclitaxel. Similarly, in the SWOG-GOG trial, those patients with a drop in their initial CA-125 to less than 10 experienced a statistically significant improvement in overall survival: 57 months vs. 41 months at the median. No final conclusion about the value of maintenance paclitaxel in ovarian carcinoma can be reached on the basis of current evidence. The ongoing GOG study of maintenance taxane versus no further therapy in clinical complete responders should provide a basis for a final conclusion, but this study will need 4-6 more years of accrual and then further follow-up before results are available. For now, physicians should discuss the available studies openly and honestly with patients, so that the 75% of patients in a complete response after initial therapy can participate in the decision as to whether to take maintenance paclitaxel. — J. Tate Thigpen, M.D.
|