A real-world analysis of the effectiveness of pembrolizumab by type of associated treatment in patients with metastatic cervical cancer: Added value of bevacizumab.

5531 Background: Pembrolizumab (pembro) showed a statistical significant survival benefit in persistent, recurrent, or metastatic cervical cancer (CC) when added to chemotherapy. In this real-world study we aimed to evaluate if the effectiveness of pembro may vary according to the type of platinum used or by the incorporation of bevacizumab (bev). Methods: The analysis was conducted with the TriNetX Global Collaborative Network. In our study, we defined different cohort of patients by type of platinum (cisplatin or carboplatin) and bev use (yes or no). First, we compared 597 CC patients treated cisplatin with 1.080 CC patients treated with carboplatin, in addition to paclitaxel and pembro. After this analyses, we evaluated the addition of bev (701 patients) to a platinum-based regimen compared to a cohort of patients treated without bev (562 patients). A propensity score matching (PSM) was used to balance the cohorts by age, race, previous radiotherapy, body mass index and treatment (bev for first analyses and cisplatin for the second). Subsequently, a COX analyses, adjusted for the same factors, was used in unmatched cohorts to validate our results. Hazard ratio (HR) was used to compare the overall survival (OS) and the development of fistulae, bowel perforation or pulmonary embolism (PE) in the matched cohorts. Results: PSM generated 583 pairs of CC patients treated with cisplatin (mean age of 50.6, +/-12.6 standard deviation, SD) or carboplatin (mean age of 50.2, +/-12.7 SD). Median OS was not statistical different between the groups (HR 0.92, 95% CI 0.75-1.28), confirmed also by the Cox model (HR 0.95, 95% CI 0.79-1.15). Among the 386 matched pairs of CC patients treated with (mean age of 53.8, +/-13.1 SD) or without bev (mean age of 53.8, +/-13.4 SD), median OS was 35.7 months versus 20.1 months (HR 0.64, 95% CI 0.49–0.84, p = 0.001), without differences in the rate of fistulae (HR 0.77, 95% CI 0.50-1.18), but an increased risk in bowel perforation (HR 3.08, 95% CI 1.01-9.38, p = 0.037) and a lower risk of PE (HR 0.57, 95% CI 0.35-0.93, p = 0.022). In multivariate analyses, bev significantly reduced the risk of death (HR 0.75, 95% CI 0.61-0.94, p = 0.01). Conclusions: The survival rates associated with immunotherapy treatment in our real-world study are consistent with those reported in previous studies. Specifically, we showed that the effectiveness of pembro is not influenced by the type of platinum used. However, the addition of bev may extend OS in patients with CC, without increasing the risk of fistulae or PE.