Assessing Prognostic and Predictive Biomarkers of Regorafenib Response in Patients with Advanced Soft Tissue Sarcoma: REGOSARC Study

SIMPLE SUMMARY: In a double-blind, placebo-controlled, randomized phase 2 trial, Regorafenib provided a clinical benefit to patients with advanced and anthracycline-pretreated soft tissue sarcoma. However, extensive mutational analysis of tumor genes could not identify predictive markers of regorafenib efficacy, including among genes involving in angiogenesis (FLT1, FLT2, FLT3, FLT4, KDR, TEK (TIE2), and VHL). The identification of the precise mechanism of action of multikinase inhibitor in sarcoma and the identification of responding patients requires further clinical studies. ABSTRACT: Regorafenib significantly prolonged progression-free survival (PFS) in pretreated patients with advanced non-adipocytic sarcoma (HR = 0.46; p < 0.001) in a placebo-controlled, randomized, phase-II trial (NCT01900743). Thus, here, we assessed the prevalence of 57 biomarkers and their prognostic and predictive values for PFS and overall survival (OS). We analyzed 134/182 patients included in this trial, treated with regorafenib (n = 71, 53%) or placebo (n = 63, 47%). Mutational analyses were performed via full coding sequence analysis for 10 genes, and mutation hotspot panel for 50 genes (four genes in common). H19 was studied with RNA in-situ hybridization. The prognostic and predictive biomarkers’ values were studied only for biomarkers found positive/mutated in at least 10 patients. Overall, 25 out of 57 studied biomarkers, including five out of seven genes involved in angiogenesis, were found mutated/positive in at least one patient, of which 23 biomarkers had low prevalence (fewer than eight out of 134 patients), contrasting with H19 (n = 24, 18%), and TP53 (n = 35, 26%). However, in multivariable models of PFS and OS, including treatment effects and interactions, no significant prognostic or predictive values of the tested biomarkers were observed. Though several promising biomarkers were found to be positive/mutated, none of them were identified as viable predictive and prognostic biomarkers.