88P Evaluation of trending drug targets and technologies in current drug development

Novel investigational agents (IAs) aim to exploit new mechanisms and cancer vulnerabilities, expanding our arsenal of anticancer drugs. Recently, highly selective biomarker-directed IAs have allowed advancements in the precision treatment of oncogene-driven cancers by decreasing toxicity and increasing activity. In phase I drug development, a global problem of unknown magnitude is the multiplicity of similar drugs being investigated against the same target, colloquially known as the ‘me too’ phenomenon. The Clinicaltrials.gov online database was queried for phase I clinical trials active during a two-year timeframe from December 2020 to December 2022. Natural language processing was employed to extract IAs from experimental interventions. Autoritative sources (Cancer.gov, Pubmed, et similia) were queried to classify IAs based on target and drug class/structure. Whenever data was unavailable from these websites, lower tier sources were employed. Overall target frequency and co-occurrence analyses were subsequently evaluated. By surveying the current landscape of phase I clinical trials for a two-year timeframe (December 2020 to December 2022), we identified 4480 distinct IAs. Excluding agents without direct mechanistic targets, non-human genome encoded targets (such as viral proteins, metabolites, carbohydrates et similia) or undisclosed data, 1016 gene targets and 3495 investigational agents were identified. While only a minority of gene targets had 4 or more IAs directed against them (26.7%) the majority of IA, 83.3% of single-target and 92.5% of multi-targeted, were directed against this minority of gene targets. The most frequent IA classes were Inhibitors(1105), Cell products(933), Monospecific-monoclonal antibodies(488) and Vaccines(252). The most frequent gene targets by number of IA were: CD19(316), EGFR(131), CD3(126), ERBB2(125), BCMA(108) and PD-1(106), CTLA4-PD1 and PDL1-CTLA4 were the most frequent cooccurring target pairs. Only a minority (9.3%) of phase I IAs were explored against a target without a competitive agent; Unfortunately, the majority of IAs (86%) shared targets with at least 3 other agents. We argue that these duplicative efforts could be redirected toward unmet needs instead.