Intrinsic vulnerability, extrinsic resistance: rethinking therapeutic targets in solid tumors

Abstract Despite decades of oncology drug development, patients with advanced solid tumors rarely experience survival gains beyond a few months, exposing a fundamental misalignment between research focus and clinical benefit. This persistent shortfall is most clearly illustrated by primary non-response, the failure of therapy to achieve meaningful tumor control from the outset, which is not fully explained by treatment-induced adaptation and points instead to barriers present before any therapeutic selection pressure. While intrinsic resistance is often invoked as the explanation, our analysis shows that cancer cell models enriched for an aggressive, clinically unfavourable transcriptional state retain broad pharmacologic vulnerabilities in vitro, suggesting that failure in patients cannot be fully explained by broad intrinsic insensitivity alone. We examined solid tumor cell lines enriched for epithelial-to-mesenchymal transition-associated transcriptional programs through high VIM, low CDH1, and high ZEB1 expression, operationally defining an EMT-enriched cohort (EMTe cohort; n = 48). Despite these aggressive programs, EMTe lines retained multiple pharmacologically relevant vulnerabilities in vitro across mechanistically diverse agents, including several FDA-approved compounds, at both 1 µM and 100 nM thresholds. Mean drug sensitivity across the EMTe cohort was comparable to that of the broader solid tumor cell line population, challenging the assumption that EMT-associated states confer broad intrinsic resistance. EMTe lines also showed elevated expression of genes implicated in extracellular matrix remodelling and immune-regulatory interactions, including LOXL2, LOX, POSTN, and TGFB1. These findings, together with prior clinical and preclinical studies, suggest that the persistent failure of increasingly refined intrinsic targets to resolve primary non-response points toward extrinsic microenvironmental constraints as limiting factors alongside intrinsic cellular determinants. Such constraints include stromal barriers, immune exclusion, and limited drug penetration. Jointly addressing cancer cell-intrinsic vulnerabilities and relevant microenvironmental constraints may therefore provide a strategy for unlocking latent vulnerabilities and improving the translation of preclinical efficacy into durable clinical benefit. This framework provides a rationale for prioritizing combination strategies and designing early-phase trials that directly address extrinsic barriers.