Predictors of long-term progression-free survival in patients with ovarian cancer treated with niraparib in the PRIMA/ENGOT-OV26/GOG-3012 study

OBJECTIVE: To identify characteristics associated with long-term progression-free survival (≥2 years) in patients with advanced ovarian cancer treated with niraparib first-line maintenance therapy in the phase III PRIMA/ENGOT-OV26/GOG-3012 study. METHODS: In this post hoc analysis of PRIMA, patients randomized to niraparib were grouped based on investigator-assessed progression-free survival (progressive disease/censoring <2 years or ≥2 years after randomization). Variables assessed for predictive value were Eastern Cooperative Oncology Group performance status, International Federation of Gynecology and Obstetrics (FIGO) stage at diagnosis, clinical response to platinum-based chemotherapy, number of prior chemotherapy cycles, primary tumor location, body mass index, categorical age, debulking surgery type, number of baseline target lesions, number of baseline non-target lesions, BRCA/homologous recombination-deficiency status, residual disease status, and duration from end of chemotherapy to randomization. Logistic regression modeling using backward elimination (significance level=0.15) identified covariates associated with long-term progression-free survival (clinical cut-off date November 17, 2021). RESULTS: Of 487 patients randomized to niraparib, 152 (31%) had progressive disease/censoring ≥2 years after randomization. Multivariable logistic regression modeling using backward elimination identified BRCA1/2 mutation/homologous recombination deficiency status (p<0.0001), FIGO stage (p=0.041), primary tumor location (p=0.095), and number of baseline non-target lesions (p=0.0001) to be associated with long-term progression-free survival. Patients significantly more likely to achieve progression-free survival of ≥2 years in the final model were those with BRCA1- and BRCA2-mutated/homologous recombination-deficient tumors or BRCA wild-type/not determined/homologous recombination-deficient tumors (vs BRCA wild-type/homologous recombination-proficient/not determined tumors), FIGO stage III (vs IV), and 0 or 1 baseline non-target lesions (vs ≥2 baseline non-target lesions). CONCLUSIONS: The hypothesis-generating results of this analysis suggest that BRCA1/2 mutation/homologous recombination-deficiency status, FIGO stage, and number of baseline non-target lesions may predict progression-free survival of ≥2 years in patients with advanced ovarian cancer receiving niraparib first-line maintenance therapy. TRIAL REGISTRATION NUMBER: NCT02655016.