A comparative study of the real-world safety and effectiveness of metronomic cyclophosphamide and bevacizumab with or without pembrolizumab for recurrent ovarian cancer.

5590 Background: Metronomic cyclophosphamide and bevacizumab (CB) alone or in combination with pembrolizumab (CBP) are active regimens in heavily pretreated patients with recurrent ovarian cancer. However, it is unknown to what extent the addition of pembrolizumab augments the effectiveness or increases the toxicity compared to CB alone. Methods: We conducted a multi-institutional retrospective cohort study utilizing electronic medical records of patients treated for recurrent ovarian cancer in oncology practices affiliated with four New England hospitals from 2012 through 2024. We included patients who received either CB or CBP and abstracted baseline characteristics, outcomes, and adverse events. The overall response rate (ORR) was defined as the proportion of patients with a complete or partial response. Net benefit (NB) was defined as the proportion of patients with a response or stable disease based on clinician assessment. Kaplan Meier curves were used to summarize progression-free survival (PFS) and overall survival (OS). Cox regression models were used to calculate and estimate the relative hazard of death (OS) and death or progression (PFS). Multivariable models adjusted for age, number of prior lines of therapy, time since diagnosis, platinum sensitivity, BRCA status, and Charlson comorbidity index. Results: We identified 163 patients, of whom 126 (77.3%) received CB and 37 (22.7%) received CBP. The median age at enrollment was 65.4 in the CB arm and 62.9 in the CBP arm (p=0.21). Most patients were non-Hispanic White (85.9%), with high grade (96.6%) serous (87.1%), platinum-resistant (76.7%) ovarian cancer. There were no statistically significant differences in BRCA status (21.4% vs. 18.9%, p=0.04), the median number of prior lines (3 vs. 3, p=0.25), or the proportion of platinum-resistant patients (77.8 % vs. 73.0%, p=0.41) between the CB and CBP arms. The ORR was 19.8 vs 21.6 % (p=0.81), and NB was observed in 44.8% in the CB arm and 41.2% in the CBP arm (p=0.71). The median PFS was 5.2 versus 4.8 months in the CB and CBP groups (HR 1.02; CI 0.67 - 1.55, p=0.37). Median OS was 17.3 vs 15.2 months, respectively (HR 1.51; CI 0.93 – 2.46, p=0.96). Adjustment for potential confounders produced similar results for PFS (adjusted HR=1.25, CI 0.77 – 2.02, p=0.37) and OS (adjusted HR=1.56, 95% CI 0.91 – 2.70 p=0.16). Seventeen (45.9%) patients in the CBP arm developed an immune reaction during treatment, but hospitalization rates were similar in both groups (25.4% vs 21.6%, p = 0.64). Conclusions: The addition of pembrolizumab to metronomic cyclophosphamide and bevacizumab was not associated with an improved response rate, PFS, or OS among patients with heavily pretreated recurrent ovarian cancer.