Cancer therapy resistance converges on immune evasion: a unified framework from immunotherapy to targeted, endocrine, and DNA repair pathways

Abstract

Cancer immunotherapy, including immune checkpoint inhibitors, chimeric antigen receptor T-cell (CAR-T) therapy, and bispecific antibodies, has transformed the treatment of advanced malignancies, yet primary and acquired resistance remain major barriers to durable benefit. Across these therapies, resistance reflects the failure of one or more of three shared requirements for effective antitumor immunity: tumor immune visibility, lost through defective antigen presentation and interferon signaling; immune access to tumor nests, restricted by stromal and vascular barriers; and effector cell function, undermined by suppressive myeloid and regulatory populations, metabolic restriction, and loss of tertiary lymphoid structures. In this review, we argue that resistance to targeted therapies, endocrine therapies, and PARP inhibitors should not be treated as a problem separate from immunotherapy resistance. Rather, these resistant states often converge on the same three axes that govern immunotherapy resistance (immune visibility, access, and effector function), yielding a unified framework across therapeutic classes. Using canonical immunotherapy resistance as a reference model, we examine how osimertinib resistance, androgen receptor (AR)-pathway-resistant states, and BRCA reversion-mediated PARP inhibitor resistance may remodel these same immune-evasion axes. Targeted therapy resistance provides the most complete case, with epithelial-mesenchymal transition, PD-L1 upregulation, STING suppression, and immune exclusion extending across all three axes. AR pathway reactivation shows plausible links to immune remodeling through direct MHC-I repression and myeloid reprogramming, although the clinical evidence base remains less mature. BRCA reversion illustrates a mechanistically suggestive dual-valence state in which restored DNA repair may diminish innate immune sensing while also generating computationally predicted junctional neopeptides whose immunogenicity remains to be validated. By reframing resistance as a new immune state that converges on the same three immune-evasion axes rather than a terminal event, this review shifts attention from why a therapy failed to which immune-evasion axis now predominates and how it might be countered. This perspective supports mechanism-matched combination therapy, adaptive molecular monitoring, and cautious exploitation of vulnerabilities generated by the resistant state itself. As an organizing model rather than a claim of uniform causality, the framework now requires prospective validation.