Tumor genomic analysis for biomarker identification in a phase I trial of the Wee 1 inhibitor adavosertib (AZD1775).

3624 Background: Adavosertib, a first-in-class Wee1 kinase inhibitor, abrogates G2/M cell cycle arrest causing premature mitosis and DNA replication stress, yielding enhanced DNA damage. Here we report on potential biomarkers of response from tumor genomic analysis in patients (pts) with solid tumors treated with adavosertib. Methods: Adavosertib was administered once daily on days 1-5 and 8-12 of a 21-day cycle. RECIST 1.1 was used to evaluate clinical response. Paired tumor biopsies were obtained for RNASeq gene expression profiling (GEP) and for whole-exome sequencing (WES) to evaluate gene mutation and copy number amplification (CNA). Fold change (FC) was calculated to define gene overexpression. To identify the frequency of CNA and mRNA overexpression for the genomic biomarkers of interest, cBioPortal analysis using TCGA and MSK-IMPACT datasets was performed. Differential GEP analysis of tumor and paired normal tissue was performed using the gene expression profiling interactive analysis (GEPIA) interface (Tang et al. 2017). Results: Out of 35 pts evaluable for response, 6 (17%) had partial response (PR; 4 ovarian carcinoma [OVC], 2 endometrial carcinoma [EC]). The median duration of response was 5.2 months (range 4.0-23.1). Eighteen pts (51.4%) had stable disease. Genomic analysis of tumor biopsies was available for 9 pts; 7 of these pts were evaluable for response, and 3 had PR (2 OVC, 1 EC). WES revealed TP53 mutations in 6 pts (66.6%; 3 pts with PR, 2 with progressive disease,1 not evaluable). On WES, tumor Cyclin E1 ( CCNE1) CNA was present in 1 of 3 PR pts while tumors from all 3 PR samples showed relatively high CCNE1 expression by RNAseq (FC = 4.07). In the MSK-IMPACT 2017 dataset, CCNE1 CNA was identified in 1.8% of pts (194 of 10336); of which, OVC (10.3%) and EC (8.7%) had the highest incidence of CCNE1 CNAs. In separate tumor-specific (OVC, EC) TCGA datasets having CCNE1 overexpression and/or CNA, overlap in CCNE1 overexpression with CCNE1 CNA was 35.5% (OVC) and 25.2% (EC). Compared to normal ovarian/ endometrial tissues, GEPIA analysis revealed significantly higher CCNE1 mRNA expression in OVC (FC = 3.5) and EC (FC = 3.8). Conclusions: CCNE1alterations (overexpression and/or CNA) tend to be enriched in OVC and EC with a limited fraction showing both overexpression and CNA. Tumor genomic analysis of additional OVC and EC pts treated with adavosertib is required to determine whether CCNE1 mRNA overexpression, regardless of CCNE1 CNA, is a potential biomarker of response to this drug in these tumor types. Funded by NCI contract No. HHSN261200800001E. Clinical trial information: NCT01748825 .