Abstract LB098: Deep plasma proteome characterization in two independent clinical cohorts identifies clusters of biomarkers separating benign and malign tumors in women with suspicion of ovarian cancer

Abstract

Abstract Early detection is essential for ovarian cancer survival but today’s biomarkers are not specific enough to clear benign from malignant tumors nor sensitive enough to be used in screening. Here we have used the proximity extension assay (PEA) Olink Explore HT to characterize 5416 plasma proteins. For a subset of the women, we have also characterized the transcriptional landscape in corresponding tumor tissue using total RNA-sequencing on the Illumina NovaSeq 6000 instrument. The plasma samples were from two independent Swedish clinical cohorts (N1 = 171, N2 = 233) and tumor RNA was extracted from 112 of the samples from the second cohort. Each cohort consisted of women who after suspicion of ovarian cancer were surgically diagnosed with either benign or malign tumors. All plasma samples were collected at time of diagnosis, before commencement of treatment. In the plasma protein analyses, the first cohort was used as a discovery and the second as replication. In the discovery cohort, we found a total of 328 plasma proteins differing significantly between women with benign tumors and; early stage (I-II), late stage (III-IV) or any stage(I-IV) malignant tumors. These associations were examined in the replication cohort, and were all found to have similar foldchange between women with benign and malignant tumors. 99.7% (327/328) differed significantly after adjustment for multiple hypothesis testing. The 328 associations corresponded to a total of 192 proteins out of which 26 (13.5%) were found to have a significant correlation with RNA-expression in the corresponding tumor. A correlation network analysis based on the 192 proteins identified 5 major clusters, consisting of 18 to 59 protein-protein couplings. 80% of these clusters contained proteins that are annotated with up to five major KEGG cancer pathways and several of these contained proteins correlating significantly with the tumor RNA-SEQ data. All except one cluster contained one or two proteins that are either currently approved FDA drug targets or listed as druggable in DrugBank by specific compounds. Ovarian cancer has a low 5-year survival and eludes the current treatment regimens if discovered late. By applying high-throughput proteomics in two independent cohorts, we identified a large number of affected proteins out of which only a few have corresponding changes in tumor RNA-expression. We also identified large highly co-regulated networks of proteins, suggesting large downstream systemic effects which could be a contributing factor to the difficulties in finding efficient treatments. Among these proteins, there are several druggable targets, however, not all the associated compounds are currently primarily used as anti-cancer drugs, highlighting the possibility of drug re-purposing for future use against ovarian cancer. Citation Format: Stefan Enroth, Mikaela Moskov, Julia Lindberg Hedlund, Svetlana Popova, Simn K. Forsberg, Klev Diamanti, Maria Lycke, Emma Ivansson, Anna Tolf, Karin Sundfeldt, Karin Stålberg, Ulf Gyllensten. Deep plasma proteome characterization in two independent clinical cohorts identifies clusters of biomarkers separating benign and malign tumors in women with suspicion of ovarian cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB098.