Randomized Trial of Intravenous Versus Intraperitoneal Chemotherapy Plus Bevacizumab in Advanced Ovarian Carcinoma: An NRG Oncology/Gynecologic Oncology Group Study

Abstract

Purpose: To evaluate the impact of two different intraperitoneal (IP) chemotherapy regimens on progression-free survival (PFS) among women with newly diagnosed advanced ovarian carcinoma.

Methods: Eligible patients were randomly assigned to six cycles of IV paclitaxel 80 mg/m2 once per week with intravenous (IV) carboplatin area under the curve 6 (IV carboplatin) versus IV paclitaxel 80 mg/m2 once per week with IP carboplatin area under the curve 6 (IP carboplatin) versus once every 3 weeks IV paclitaxel 135 mg/m2 over 3 hours day 1, IP cisplatin 75 mg/m2 day 2, and IP paclitaxel 60 mg/m2 day 8 (IP cisplatin). All participants received bevacizumab 15 mg/kg IV every 3 weeks in cycles 2 to 22.

Results: A total of 1,560 participants were enrolled and had 84.8 months of follow-up. The median PFS duration was 24.9 months in the IV carboplatin arm, 27.4 months in the IP carboplatin arm, and 26.2 months in the IP cisplatin arm. For the subgroup of 1,380 patients with stage II/III and residual disease of 1 cm or less, median PFS was 26.9 (IV-carboplatin), 28.7 (IP-carboplatin), and 27.8 months (IP cisplatin), respectively. Median PFS for patients with stage II/III and no residual disease was 35.9, 38.8, and 35.5 months, respectively. Median overall survival for all enrolled was 75.5, 78.9, and 72.9 months, respectively, and median overall survival for stage II/III with no gross residual disease was 98.8 months, 104.8 months, and not reached. Mean patient-reported Functional Assessment of Cancer Therapy neurotoxicity scores (Gynecologic Oncology Group) were similar for all arms, but the mean Trial Outcome Index of the Functional Assessment of Cancer Therapy-Ovary scores during chemotherapy were statistically worse in the IP cisplatin arm.

Conclusion: Compared with the IV carboplatin reference arm, the duration of PFS was not significantly increased with either IP regimen when combined with bevacizumab and was better tolerated than IP cisplatin.

Trial registration: ClinicalTrials.gov NCT00951496.

Figures

FIG 1.

CONSORT diagram that accounts for all participants and the chemotherapy regimen arms to which they were randomly assigned. Crossover during treatment to the IV arm occurred in 16% of those randomly assigned to the IP carboplatin arm and 28% of those randomly assigned to the IP cisplatin arm. Taxotere was substituted for paclitaxel in 7% of those in the IV and IP carboplatin arms and 5% of those in the IP cisplatin arm. There was discontinuation of bevacizumab during the concurrent chemotherapy administration of the first six cycles in 15% of participants in the IV and IP carboplatin arms and 30% of participants in the IP cisplatin arm. The median number of cycles of bevacizumab was 20 for those in the IV carboplatin arm, 19 in the IP carboplatin arm, and 17 in the IP cisplatin arm. Reason for discontinuation of study regimen was completion of therapy in 51% of participants in the IV carboplatin arm, 49% in the IP carboplatin arm, and 45% in the IP cisplatin arm. Toxicity caused discontinuation in 24%, 28%, and 29%, respectively, and refusal was 6%, 6%, and 9%, respectively. Disease progression or death was cause for discontinuation in 16%, 13%, and 13%, respectively. The delivery of six cycles of any platinum agent was 90% in the IV and IP carboplatin arms and 84% in the IP cisplatin arm. Day 8 paclitaxel was delivered to 66% of participants in the IV or IP carboplatin arms for all six cycles and only 59% in the IP cisplatin arm. Four cycles of day 8 paclitaxel was received by 85% of participants in the IV and IP carboplatin arms and 80% of those in the IP cisplatin arm. In the IV and IP carboplatin arms, only 40% of participants received day 15 paclitaxel through six cycles, and 60% received it through four cycles, whereas 80% received two cycles. AUC, area under the curve; OS, overall survival; PFS, progression-free survival.

FIG 2.

Progression-free survival (PFS) of all randomly assigned participants. In the overall intention-to-treat population of 1,560 patients, 1,139 (73%) experienced progression. Compared with intravenous (IV) carboplatin, the hazard of first progression or death was 7.5% lower (hazard ratio, 0.925; 95% CI, 0.802 to 1.07) for the intraperitoneal (IP) carboplatin arm and 2.3% lower (hazard ratio, 0.977; 95% CI, 0.847 to 1.13) for the IP cisplatin arm. There was no statistically significant difference in PFS between the IV regimen and either of the IP regimens.

FIG 3.

Progression-free survival (PFS) of participants with optimally resected stage II/III with 1 cm or less residual disease per surgeon. In the 1,380 participants evaluated, 973 (71%) had a median PFS of 26.9 months in the intravenous (IV) carboplatin arm, 28.7 months in the intraperitoneal (IP) carboplatin arm (hazard ratio, 0.921; 95% CI, 0.789 to 1.07), and 27.8 months in the IP cisplatin arm (hazard ratio, 0.966; 95% CI, 0.828 to 1.13). There was no statistically significant difference in PFS between the IV regimen and either of the IP regimens in this subgroup of patients.

FIG 4.

Progression-free survival (PFS) of participants with stage II/III with no gross residual disease. There were 870 participants with stage II or III and no gross residual disease documented at the completion of surgery of whom 534 (61%) experienced disease progression or death. Adjusted HR (95% CI) of PFS for intraperitoneal (IP) carboplatin vs. intravenous (IV) carboplatin = 0.92 (0.75 – 1.13). Adjusted HR (95% CI) of PFS for IP cisplatin vs. IV carboplatin = 0.97 (0.79 – 1.19). There was no statistically significant difference in PFS between the IV regimen and either of the IP regimens in this subgroup of patients.

FIG 5.

Overall survival (OS) by randomized treatment of all enrolled patients. Forty-nine percent of participants were alive after a median follow-up of 84.8 months. Relative to the intravenous (IV) carboplatin arm, the hazard of death was similar in the intraperitoneal (IP) carboplatin arm (hazard ratio, 0.949; 95% CI, 0.799 to 1.128) and in the IP cisplatin arm (hazard ratio, 1.05; 95% CI, 0.884 to 1.24). OS for stage II/III with no gross residual disease was 98.8 months for IV carboplatin; 104.8 months for IP carboplatin; and yet to be determined for IP cisplatin, which needs a longer follow-up.