Prognostic Value of a Joint K-PD Model With Tumor Size Dynamics and CA-125 Kinetics in Recurrent Ovarian Cancer Patients: BOLD Phase II GINECO Study.

Abstract

In patients with recurrent advanced ovarian cancer, there is a need for companion tests to guide the development of innovative chemotherapy-free treatments. The modeled longitudinal CA-125 ELIMination rate constant K KELIM-B was a major prognostic factor for progression-free survival (PFS) and overall survival (OS) in recurrent advanced ovarian cancer patients treated with bevacizumab, olaparib, and durvalumab in the BOLD trial. The objective was to determine if a joint semi-mechanistic model with tumor size and CA-125 kinetics would increase KELIM-B accuracy/prognostic value. The BOLD phase II trial (NCT04015739) investigated the triplet regimen in 74 patients with recurrent platinum-sensitive/resistant advanced ovarian cancer. Two kinetic-pharmacodynamic models were developed to fit the data collected during the first 100 treatment days: (1) a CA-125 longitudinal kinetics model, and (2) a joint model integrating both CA-125 kinetics and tumor size. The prognostic value of KELIM-B and KELIM-joint was assessed using univariate/multivariate analyses (PFS/OS). The modeling of CA-125 and tumor size dynamics was feasible with adequate quality checks. The prognostic value of the categorical KELIM-joint, binarized by the median (PFS, HR = 0.29, 95% CI [0.12-0.72]; OS, HR = 0.24, 95% CI [0.08-0.74]), was not clinically different from that of KELIM-B (PFS, HR = 0.35, 95% CI [0.14-0.84]; OS, HR = 0.34, 95% CI [0.12-0.99]). Interactions between tumor size changes and CA-125 kinetics could be assessed in the joint model. However, the improvement in prognostic value was not sufficient to justify the higher complexity of the joint model. Assessing early longitudinal CA-125 kinetics alone remains the best pragmatic strategy for future development.