Abstract 526: Leveraging germline genetics to predict the efficacy of repurposing approved cancer therapies for cancer prevention

Abstract Background: The high costs and failure rates of drug development have hindered progress in new cancer therapies, driving interest in systematic drug repurposing. Recent studies suggest that genetic evidence can improve the success rate of drug development. We aimed to investigate the yield of germline genetics in recapitulating the efficacy of approved cancer therapies for preventing the same cancer. Methods: We mapped 3, 081 plasma protein cis-protein quantitative trait loci (pQTLs) from UK Biobank Olink and Iceland SomaLogic GWAS data to approved and currently marketed cancer drug information on the Open Targets Platform. Using publicly available GWAS summary statistics for 39 cancer subtypes, we performed cis-pQTL Mendelian randomization (MR) analyses to estimate the association between drug-target protein concentrations and corresponding cancer risk. Odds ratios (ORs) were oriented to the drug’s action on the protein i.e., a drug inhibiting a risk-increasing protein would correspond to a reduced risk. Using our MR results, we evaluated the success of genetic evidence in repurposing cancer treatments for prevention, relative to the approval rate of new drugs progressing from phase 1 trials. We quantified this using the relative probability with 95% confidence intervals (CI) that a target-indication (T-I) pair approved for treatment will have genetic evidence for use in prevention (R(G)), using the ratio of the probability of a directionally consistent significant MR result (P(G)) to the probability of approval for pairs that completed phase 1 trials (P(D)). We tested the sensitivity of R(G) by cancer type and targeted pathways. Results: Of the 298 cancer T-I pairs available for MR analyses, we found 49 significant protein-cancer associations representing 28/72 targets across 22/39 cancer subtypes. Of these, 36 were directionally congruent with the drug’s action on the protein, representing 20/72 targets of 54/139 drugs across 19/39 cancer subtypes. MR-significance and directionality of T-I pairs did not differ significantly by the presence of a targeted somatic mutation in the protein. We estimated the relative probability (R(G)) that a T-I pair approved for treating cancer has genetic support for prevention to be 1.95 (95% CI: 1.46-2.60). R(G) was significantly higher for 8/18 cancers [pancreatic cancer (3.7, 1.1-12.1) to brain cancer (67.5, 18.6-244.6)], significantly lower for thyroid cancer, and not significantly different for 9/18 cancers. Furthermore, R(G) was significantly higher for T-I pairs mapped to 9/12 pathways, including extracellular matrix organization, transcription and signal transduction. Conclusion: Our findings indicate that certain targets of existing cancer therapies may be repurposed for cancer prevention and demonstrate a potentially high yield in the utility of genetics in improving target identification for drug repurposing in cancer. Citation Format: Chibuzor F. Ogamba, Elio Riboli, James Yarmolinsky, Philip C. Haycock, Ruth C. Travis, Karl Smith-Byrne. Leveraging germline genetics to predict the efficacy of repurposing approved cancer therapies for cancer prevention [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 526.