Introduction/Background*In the randomized phase 3 JAVELIN Ovarian 200 trial (N=566), avelumab alone or combined with pegylated liposomal doxorubicin (PLD) did not significantly prolong progression-free survival (PFS; blinded independent central review) or overall survival (OS) vs PLD alone in patients with platinum-resistant/refractory ovarian cancer (PRROC). Here, we report exploratory biomarker analyses associated with outcomes in the avelumab alone arm.

Methodology Biomarkers analyzed in tumour tissue or blood were somatic and germline mutations by whole-exome sequencing and whole-transcriptomic profiling by RNAseq. Associations were assessed using the Cox proportional hazards model. For continuous biomarkers, hazard ratios (HRs) and 95% CIs were reported using a median cutoff.

Result(s)*Within the avelumab alone arm, higher expression of CD8+ T-cell signature correlated with longer PFS (HR, 0.61 [95% CI 0.43-0.87]), and higher tumour mutational burden (>0.6 mut/Mb) and presence of the high-affinity FCGR2A allele correlated with longer OS (HR [95% CI], 0.63 [0.42-0.96] and 0.53 [0.34-0.83], respectively), while presence of APOBEC and homologous recombination deficiency (HRD) mutational signatures were associated with shorter OS (HR [95% CI], 2.17 [1.22-3.89] and 1.68 [1.07-2.62], respectively). In patients with PFS of <6 months, a lower score for several signatures, including P53 and angiogenesis, predicted better outcomes within the avelumab alone arm (HR [95% CI], 0.37 [0.23-0.60] and 0.52 [0.33-0.82], respectively). BRCA1/2 mutations (germline and/or somatic) were not associated with improved outcomes. Within the avelumab alone arm, higher expression of TGF-β signaling, estrogen receptor early expression signature, and epithelial-to-mesenchymal transition gene signatures correlated with shorter OS (HR [95% CI], 1.64 [1.10-2.47], 1.63 [1.09-2.45], and 1.57 [1.05-2.35], respectively). Higher expression of signatures for mesenchymal cells, cancer-associated fibroblasts, and fat cells were marginally associated with shorter OS (HR [95% CI], 1.47 [0.97-2.23], 1.45 [0.95-2.21], and 1.50 [0.99-2.27], respectively). Multivariable analyses are ongoing.

Conclusion*Although JAVELIN Ovarian 200 did not meet its primary endpoints, these analyses indicate potential subgroups of patients who could have improved outcomes with immunotherapy alone and provide insights into the biology of PRROC that may inform future trials.