Immunogenic cell death (ICD) links tumor cell demise to the activation of anti-tumor immunity, but its adoption has also generated inconsistent definitions and frequent overinterpretation of surrogate biomarkers. Here, we synthesize mechanistic and methodological evidence showing that danger-associated molecular patterns (DAMPs), cytokine release, and endoplasmic reticulum stress report immunogenic potential rather than ICD itself. We propose that ICD should be defined by its functional immunological endpoint, namely efficient antigen presentation and antigen-specific adaptive immunity, ideally culminating in protective immunological memory. To operationalize this principle, we introduce a hierarchy of experimental validation ranging from correlative hallmarks (Level 0) to innate immune integration (Level 1), antigen-specific T-cell priming (Level 2), definitive vaccination-rechallenge protection with immune-dependence testing (Level 3), and translational relevance supported by convergent human data (Level 4). We also discuss common pitfalls, equating inflammation, necrosis-associated DAMP release, or therapeutic benefit with ICD, and outline minimal immune-context controls (e.g., MHC-I, CD8+ T cells, Batf3-dependent dendritic cells, and innate sensing pathways) required to support robust claims. Finally, we highlight why ICD remains strongly context-dependent, shaped by dendritic-cell competence, innate licensing, purinergic metabolism, and microenvironmental constraints. Evidence-graded standards should improve reproducibility, strengthen peer review, and accelerate clinically meaningful ICD-based strategies.
Boaru et al. (2026) studied this question.