Cisplatin and carboplatin may offer comparable survival when combined with first-line pembrolizumab and paclitaxel for advanced cervical cancer, according to a large multinational real-world analysis.
The comparative effectiveness study, published in JAMA Network Open, found no difference in overall survival between cisplatin and carboplatin, while bevacizumab was associated with an early survival benefit that diminished with longer follow-up.
Researchers used multinational real-world data to examine outcomes among patients receiving first-line pembrolizumab plus platinum-based chemotherapy, with or without bevacizumab.
The study was designed to address two practical treatment questions that remain at the discretion of clinicians: whether the choice of platinum agent affects survival and whether adding bevacizumab provides additional benefit in the immunotherapy era.
Similar survival with cisplatin and carboplatin
The platinum analysis included 1,931 patients, of whom 719 received cisplatin and 1,212 received carboplatin. After propensity score matching, the analysis included 623 matched pairs.
Median overall survival was 25.1 months in both groups, with no significant difference between the two platinum agents (HR 0.98; 95% CI 0.81–1.18; P=0.35).
The findings suggest that cisplatin and carboplatin may have comparable effectiveness when used alongside pembrolizumab and paclitaxel, potentially providing clinicians with greater flexibility to select a platinum agent based on individual patient factors.
No significant differences were observed between the cisplatin and carboplatin groups in the adverse events of special interest assessed, which included fistula, bowel perforation and pulmonary embolism.
Early survival signal with bevacizumab
The researchers separately compared outcomes among 803 patients treated with bevacizumab and 667 who did not receive bevacizumab, generating 455 propensity-matched pairs.
At 12 months, overall survival was 74.8% among patients receiving bevacizumab compared with 64.5% among those who did not.
However, the difference narrowed substantially over time. At 24 months, overall survival was 56.2% with bevacizumab and 54.0% without bevacizumab.
In the initial matched analysis, median overall survival was 33.2 months with bevacizumab and 33.4 months without it, with an overall hazard ratio for mortality of 0.77 (95% CI 0.61–0.97; P=0.01).
The researchers described the findings as suggesting an early survival benefit from bevacizumab that attenuated with longer follow-up.
Rates of fistula and bowel perforation were not significantly different between the groups. Pulmonary embolism occurred in 6.4% of patients receiving bevacizumab and 9.9% of those who did not, although the difference was not statistically significant.
Observational findings require caution
The authors stressed that the findings are hypothesis-generating and cannot establish that bevacizumab itself was responsible for the observed early survival difference.
Despite propensity score matching, there was residual imbalance in prior chemoradiotherapy between the bevacizumab groups. In addition, the database did not capture several potentially important confounders, including ECOG performance status, tumour burden, extent of pelvic invasion, tumour histology and PD-L1 combined positive score.
The authors noted that bevacizumab may be withheld from patients at greater risk of complications, including those with extensive pelvic disease, high fistula risk or significant comorbidities. This creates the possibility that patients selected to receive bevacizumab had a more favourable prognosis at baseline.
A post-hoc rematched analysis addressing the imbalance in prior chemoradiotherapy found 12-month overall survival rates of 72.5% with bevacizumab and 67.5% without it. By 24 months, survival rates were 57.3% and 58.7%, respectively.
Overall, the findings support flexibility in platinum selection, while the potential additional benefit of bevacizumab remains less certain and requires prospective validation.
The researchers concluded that cisplatin and carboplatin were associated with comparable effectiveness when combined with pembrolizumab, while the early survival signal observed with bevacizumab warrants prospective validation.
Paper: Farolfi A, Casadei C, Scarpi E, et al. Bevacizumab and Platinum Choice in Pembrolizumab-Treated Advanced Cervical Cancer. JAMA Netw Open. 2026;9(7):e2623166. Access online here.
