Autologous cytokine-induced NK cells as candidate cellular senolytics: evidence, obstacles, and the experiments still needed

Cellular senescence and chronic low-grade inflammation occupy a central, mechanistically coupled position among the hallmarks of aging. Natural killer (NK) cells decline in per-cell cytotoxicity with age, and lower NK cell cytotoxicity has been associated with higher cancer incidence and infection-related mortality in older adults. NK cells are also endogenous effectors of senescent-cell clearance, engaging stress-induced NKG2D and DNAM-1 ligands and depending on perforin-mediated cytolysis. These literatures meet at a hypothesis: that autologous cytokine-induced NK cells, manufactured from a patient’s own peripheral blood mononuclear cells (PBMCs), might be developed as candidate cellular senolytics. Cytokine-induced killer (CIK) and cytokine-induced memory-like (CIML) manufacturing platforms are mature, with registry safety data across more than 2700 predominantly oncology patients and a first autologous NK cell trial in a non-cancer, aging-related indication (Alzheimer’s disease). Yet no trial has tested such a product against a senescent-cell-burden or biological-age endpoint. This review maps the evidence across NK cell aging, NK cell-mediated senolysis, and autologous NK cell clinical experience; separates what is established in humans from what remains preclinical or hypothetical; and confronts three obstacles: the HLA-E/NKG2A evasion axis, the heterogeneity of senescent-cell surface phenotypes across tissue and inducing stimulus, and the physiological roles of senescent cells in wound healing, tissue remodeling, development, and tumor suppression that make indiscriminate clearance non-trivial. We conclude that the rationale is defensible but not yet trial-ready, and specify the precursor experiments (autologous NK cell cytotoxicity against autologous senescent cells from older donors, and a preclinical infusion model) that should gate any first-in-human program.