Data from the phase 3 AGO-OVAR 17/BOOST/GINECO OV118/ENGOT Ov-15 trial of patients with primary epithelial ovarian, fallopian tube, or peritoneal cancer show that even though prolonged treatment with bevacizumab is feasible and safe, it does not improve survival.
Data from the phase 3 AGO-OVAR 17/BOOST/GINECO OV118/ENGOT Ov-15 trial (NCT01462890) of patients with primary epithelial ovarian, fallopian tube, or peritoneal cancer show that even though prolonged treatment with bevacizumab (Avastin) for up to 30 months is feasible and shows consistent safety, no improvement in survival over the standard 15 month treatment regimen was observed, according to findings presented during the 2021 ASCO Annual Meeting.1
Results from the trial indicated that the median progression-free survival (PFS) for patients treated with bevacizumab for 30 months was 26.0 months (95% CI, 23.7-29.7) vs 24.2 months (95% CI, 22.2-26.5) for those treated for 15 months (Hazard ratio [HR], 0.99; 95% CI, 0.85-1.15; P = .90), which was not significantly different. Additionally, due to evidence suggesting non-proportional distribution of events, a restricted mean analysis was performed that demonstrated a mean PFS of 39.3 months (95% CI, 36.2-42.4) in the experimental arm compared with 39.5 months (95% CI, 36.3-42.7) in the standard of care arm, which was also not statistically significantly different (P = .92).
“The duration of treatment for bevacizumab with 15 months as part of the first-line treatment in advanced ovarian cancer remains the standard of care,” lead study author Jacobus Pfisterer, MD, director of the Gynecologic Oncology Center in Kiel, Germany, said during a presentation on the data.
Previously, the phase 3 GOG-218 (NCT00262847) and ICON7/AGO-AVAR 11 (NCT00483782) trials, which aimed to assess the effects of prolonging treatment with bevacizumab, demonstrated that the early and continuous addition of 12 and 15 months of treatment with bevacizumab to standard carboplatin- and paclitaxel-based chemotherapy significantly improved PFS. In both trials, the maximum PFS benefit was reached at the time point of highest cumulative bevacizumab exposure, which was immediately after the last cycle of treatment had been completed.2,3
As such, investigators launched the AGO-OVAR 17 trial in order to further examine if prolonging treatment with bevacizumab for up to 30 months would improve efficacy.
The trial enrolled patients with histologically confirmed epithelial ovarian, fallopian tube or peritoneal cancer, but excluded patients with non-epithelial and borderline tumors. Additionally, patients needed to have International Federation of Gynecology and Obstetrics (FIGO) stage IIB to stage IV disease of any grade or histologic subtype. Treatment on study had to be started within 8 weeks or less following primary debulking surgery or more than 4 than weeks before the first dose of bevacizumab. Moreover, adequate coagulation parameters, as well as bone marrow, liver, and renal function were necessary for enrollment. Patients needed to have an ECOG performance status of 0 to 2. Additionally, standard bevacizumab exclusion criteria, such as uncontrolled hypertension, were utilized.
From November 2011 to August 2013, 927 patients were enrolled on the study and randomized 1:1 to receive either bevacizumab for 30 months (n = 463) or 15 months (n = 464). Patient were stratified based on FIGO stage IIB to stage IIIC disease with no residual tumor and FIGO stage IIB or stage IIIC disease with residual tumor or FIGO stage IV disease.
In both study arms, patients received 15 mg/kg of bevacizumab every 3 weeks for 22 cycles in the control arm and 44 cycles in the experimental arm, as well as 6 cycles of paclitaxel at a dose of 175 mg/m2 and 6 cycles of carboplatin AUC5 every 3 weeks.
The primary end point of the study was PFS, while key secondary end points included overall survival (OS), safety and tolerability, objective response rate, and health-related quality of life. Additionally, statistical assumptions included:
Additionally, due to a low event rate, the study was closed after 97% of the planned events (n = 673/697) had been observed. Moreover, at closure, the conditional probability of reaching a significant result with the full event number was 7.5%.
As of January 2021, of the total intent-to-treat population (n = 927). In total, 7% of patients were lost to follow-up (n = 60) and 9% withdrew consent (n = 80). Moreover, 58% of patients had died (n = 534) and the trial had median follow-up was 85 months.
Patients on the study had a median age of 61 years (range, 21-89), and the majority had an ECOG performance status of 0 (54%), while others had a status of 1 (41%) and 2 (4%). In total, 58% of patients had no residual tumor following surgery and 79% had high-grade serous disease.
Additionally, 84% of patients on the study had ovarian cancer, while 8% had fallopian tube cancer and 8% peritoneal cancer. Moreover, there was an even split of patients who had FIGO stage IIB to stage IIIC disease with no residual tumor, and FIGO stage IIB to stage IIIC disease with residual tumor or FIGO stage IV disease, respectively. Patient characteristics were well-balanced across study arms.
“The relative dose intensity of carboplatin and paclitaxel was approximately 95% in both treatment arms,” Pfisterer said. “For bevacizumab, it was approximately 97% in total, 97% during chemotherapy, and 97% for the maintenance phase—not different between treatment arms.”
In the subgroup analysis that included patients with FIGO stage IIB to stage IIIC disease with no residual tumor, the median PFS was 38.8 months (95% CI, 35.2-44.6) in the experimental arm vs 38.4 months (95% CI, 28.6-46.6) in the control arm (HR, 0.93; 95% CI, 0.74-1.18; P = .93). Moreover, the restricted mean PFS was 53.2 months (95% CI, 48.5-57.9) and 51.0 months (95% CI, 46.2-55.8) for the experimental and control arms, respectively (P = .53).
Additionally, in the subgroup analysis that included patients with FIGO stage IIB to stage IIIC disease with residual tumor or FIGO stage IV disease, the median PFS in the experimental arm was 18.2 months (95% CI, 16.7-19.7), compared with 19.3 months (95% CI, 16.4-20.0) in the standard arm (HR, 1.06 95% CI, 0.87-1.29; P = .58). The restricted mean PFS was 25.5 months (95% CI, 22.4-28.6) in the experimental arm and 27.8 months (95% CI, 24.1-31.4) in the standard arm (P = 0.35).
Additional data indicated that the median OS for the experimental arm was 60.0 months (95% CI, 54.0-68.6) vs 54.3 months (95% CI, 51.0-64.6) for the standard arm (HR, 1.04; 95% CI, 0.87-1.23; P = .68). Moreover, the restricted mean OS for the experimental and standard arms was 60.8 months (95% CI, 57.8-63.8) and 60.4 months (95% CI, 57.2-63.6), respectively (P = .97).
In terms of safety, adverse effects (AEs) of any grade occurred in 99% of all patients on the study, while serious AEs occurred in 45%, grade 3 to 5 AEs were reported in 65%, and grade 5 AEs occurred in 1%. Additionally, AEs of special interest were reported in 35% of patients.
The most common serious AEs and AEs of special interest that were grade 3 or higher in the total patient population were hypertension (22%), any grade of intestinal perforation/fistula (4%), grade 3 or higher thromboembolic events (3%), and grade 3 or higher proteinuria (3%).
In the experimental cohort, the most common serious AEs or AEs of special interest were grade 3 of higher hypertension (25%), all grades of intestinal perforation/fistula (4%), and grade 3 or higher proteinuria (4%), while the experimental cohort included grade 3 or higher hypertension (20%), any grade intestinal perforation/fistula (5%), and grade 3 or high thromboembolic events (4%).