Background
GOG-240 was a phase III RCT evaluating addition of bevacizumab to doublet chemotherapy in persistent, recurrent, or metastatic cervical cancer. Using a 2×2 factorial design (cisplatin-paclitaxel vs topotecan-paclitaxel ± bevacizumab), it was the first positive phase III trial to add a targeted agent to first-line cervical cancer therapy.
Interventions and follow up
Arm A: Cisplatin 50mg/m² + paclitaxel 135–175mg/m² q21d (control backbone)
Arm B: Topotecan 0.75mg/m² days 1–3 + paclitaxel 175mg/m² q21d (experimental backbone)
Randomization: both backbones ± bevacizumab 15mg/kg q21d; primary comparison bevacizumab vs no bevacizumab
Primary endpoint: overall survival
mFollow-up: 20.8mo
Arm B: Topotecan 0.75mg/m² days 1–3 + paclitaxel 175mg/m² q21d (experimental backbone)
Randomization: both backbones ± bevacizumab 15mg/kg q21d; primary comparison bevacizumab vs no bevacizumab
Primary endpoint: overall survival
mFollow-up: 20.8mo
Results
mOS (bevacizumab vs no bevacizumab, n=452): 17.0 vs 13.3mo (HR 0.71, 95% CI 0.54–0.95, P=.004)
mPFS: 8.2 vs 5.9mo (HR 0.67, P<.001)
ORR: 48% vs 36% (P=.008)
Backbone: topotecan-paclitaxel not superior to cisplatin-paclitaxel (mOS HR 1.20, 95% CI 0.82–1.76)
mPFS: 8.2 vs 5.9mo (HR 0.67, P<.001)
ORR: 48% vs 36% (P=.008)
Backbone: topotecan-paclitaxel not superior to cisplatin-paclitaxel (mOS HR 1.20, 95% CI 0.82–1.76)
Adverse events
Bevacizumab-related (grade ≥3): hypertension 25% vs 2%, venous thromboembolism 8% vs 1%, GI fistula 3% vs 0%
Other: myelosuppression modestly increased; overall grade ≥3 AEs higher with bevacizumab.
Other: myelosuppression modestly increased; overall grade ≥3 AEs higher with bevacizumab.
Conclusions
Adding bevacizumab to cisplatin-paclitaxel significantly improved OS, PFS, and ORR in recurrent/metastatic cervical cancer, establishing the first molecularly targeted first-line regimen in this disease. Cisplatin-paclitaxel remained the preferred chemotherapy backbone.
Key Limitations
2×2 factorial design required cross-arm pooling for the bevacizumab comparison, limiting subgroup power; fistula risk particularly concerning in previously irradiated patients; absolute OS gain modest (3.7mo); no PD-L1 or biomarker selection; subsequently superseded by PD-1/PD-L1 inhibitor-containing regimens (KEYNOTE-826, BEATcc).
Clinical Context
GOG-240 established bevacizumab + cisplatin-paclitaxel as the first-line standard for recurrent/metastatic cervical cancer for nearly a decade. The paradigm has since evolved with KEYNOTE-826 (pembrolizumab + chemo ± bevacizumab, especially PD-L1 CPS ≥1) and BEATcc (atezolizumab). Bevacizumab-containing backbones remain relevant for PD-L1-unselected populations and resource-limited settings.