Beyond pyroptosis: Non-lytic functions of GSDMD in tumor immunity and cancer immunotherapy

Gasdermin D (GSDMD) is classically defined as the pore-forming executioner of pyroptosis, but GSDMD activation does not invariably culminate in terminal lysis. Direct mechanistic studies show that limited GSDMD pores can support interleukin-1β (IL-1β), ATP, or interferon-stimulated gene 15 (ISG15) release before overt plasma-membrane rupture, whereas Ca²⁺-triggered endosomal sorting complex required for transport (ESCRT)-III recruitment and ALG-2-interacting protein X (ALIX)-dependent endosomal clearance can limit or reverse membrane injury. Nerve injury-induced protein 1 (NINJ1) acts downstream of early pore-mediated permeability to execute terminal plasma-membrane rupture and macromolecular damage-associated molecular pattern (DAMP) release. Emerging evidence extends GSDMD activity to nuclear-envelope transport and extracellular-vesicle-mediated transfer of pore structures, although these mechanisms are model-specific and are not uniformly non-lytic. In tumor immunity, GSDMD has opposing compartment-dependent effects. In a B16F10 anti-programmed death-ligand 1 (PD-L1) model, GSDMD in Cx3cr1⁺ myeloid/antigen-presenting cell (APC) populations restricted a cyclic GMP-AMP synthase (cGAS)-dependent ISG program. In MC38 and KPC models, non-lytic GSDMD pores in CD4⁺ T cells amplified Ca²⁺-nuclear factor of activated T cells (NFAT)-IL-2 helper signaling. In selected colorectal cancer (CRC) and oral squamous cell carcinoma (OSCC) models, distinct full-length, nuclear, or cleavage-associated GSDMD states regulated DNA repair, growth, metastasis, and immune exclusion. In selected pancreatic and ovarian cancer models, treatment-induced gasdermin activation contributed to immunogenic lytic death. This review distinguishes established basic mechanisms, model-defined causal evidence, retrospective associations, and hypothesis-generating translation. We propose cell-type- and state-resolved biomarkers and compartment-selective interventions as prospective research frameworks rather than clinically validated decision tools. Standardized functional assays, unified multicompartment models, selective delivery, and prospective human validation are required before GSDMD can be considered clinically actionable.