Background
Phase III RCT (ICON7). 1,528 patients with newly diagnosed stage I-IIa (clear cell or grade 3) or stage IIb-IV epithelial ovarian, fallopian tube, or primary peritoneal cancer. Randomized to carboplatin AUC 5/6 + paclitaxel 175 mg/m² × 6 cycles vs carboplatin + paclitaxel + bevacizumab 7.5 mg/kg q3wk (concurrent during chemo) then bevacizumab maintenance × 12 months (18 cycles total). This used a lower bevacizumab dose (7.5 mg/kg) than GOG 218 (15 mg/kg). Stratified by residual disease and stage. Final OS analysis reported here (Oza Lancet Oncol 2015).
Interventions and follow up
Arm A: Carboplatin + paclitaxel × 6 cycle
Arm B: Carboplatin + paclitaxel + bevacizumab 7.5 mg/kg q3wk (during chemo) then bevacizumab maintenance × 12 month
Primary endpoint: PFS; secondary: OS
mFollow up: 48.9 months (final analysis)
Arm B: Carboplatin + paclitaxel + bevacizumab 7.5 mg/kg q3wk (during chemo) then bevacizumab maintenance × 12 month
Primary endpoint: PFS; secondary: OS
mFollow up: 48.9 months (final analysis)
Results
Median PFS — all patients: 19.8 months (bev) vs 17.4 months (no bev), HR 0.87 (95% CI 0.77–0.99) — modest benefit
PFS — high-risk subgroup (suboptimally debulked stage III-IV): 16.0 vs 10.5 months, HR 0.73 (95% CI 0.60–0.93)
Median OS — all patients: 57.4 vs 54.3 months, HR 0.99 (95% CI 0.85–1.14) — not significant
OS — high-risk subgroup: 39.3 vs 34.5 months, HR 0.78 (95% CI 0.63–0.97) — exploratory
PFS — high-risk subgroup (suboptimally debulked stage III-IV): 16.0 vs 10.5 months, HR 0.73 (95% CI 0.60–0.93)
Median OS — all patients: 57.4 vs 54.3 months, HR 0.99 (95% CI 0.85–1.14) — not significant
OS — high-risk subgroup: 39.3 vs 34.5 months, HR 0.78 (95% CI 0.63–0.97) — exploratory
Adverse events
Main adverse events: Grade ≥3 hypertension: 18% (bev) vs 2% (control). Thromboembolic events: 7% vs 4%. GI perforation/fistula: low but higher with bev. Wound healing complications. QoL analysis showed worse scores during bev maintenance but resolved after stopping.
Conclusions
Bevacizumab provided a statistically significant but modest PFS benefit overall, with no OS benefit in all patients; an exploratory survival benefit was seen in the high-risk (suboptimally debulked stage III-IV) subgroup. Bevacizumab's role is most supported in high-risk patients.
Key Limitations
Key Limitations: Final OS analysis showed no survival benefit overall — the PFS improvement didn't translate to OS in all comers. The high-risk subgroup OS benefit is exploratory/post-hoc and requires cautious interpretation. Bevacizumab dose was 7.5 mg/kg (half GOG 218 dose), making cross-trial comparisons complex. PARP inhibitors have substantially greater efficacy than bevacizumab in BRCA-mutated/HRd patients. The early-stage (I-IIa) patients in ICON7 are not typical bevacizumab candidates.
Clinical Context
ICON7 and GOG 218 together established bevacizumab as an option in newly diagnosed ovarian cancer, but both failed to show OS benefit in unselected patients. Modern practice reserves bevacizumab primarily for HRp/BRCAm-wild-type patients (who gain less from PARP inhibitors) and as the backbone for PAOLA-1 (olaparib + bevacizumab) in HRd/BRCAm patients. ESMO-MCBS: 3 (high-risk subgroup) / 1 (overall population).
References