Аңдатпа
Relevance: Prostate cancer (PCa) is one of the most common malignancies in men. Despite advances in immunotherapy, its efficacy in PCa remains limited, attributed to the immunosuppressive tumor microenvironment and mechanisms of tumor immune evasion. The study aimed to systematize current data on the immunological landscape and mechanisms of immune evasion in prostate cancer, assess the quality of the evidence base, and propose an integrated model explaining prostate cancer resistance to immunotherapy.Materials and Methods: Conducted per PRISMA 2020; PubMed, Scopus, and Web of Science were searched for 2016-2026. Of 613 identified records, 34 publications met the eligibility criteria: 27 original studies on prostate cancer material entered the qualitative synthesis, and 7 were used as additional (contextual) sources. The included papers mainly describe preclinical studies (12 of 27); only 3 had a low risk of bias.Results: Mechanisms of immune escape in PCa were systematized across three interrelated levels: cellular-structural (FAP+ fibroblasts and SPP1+ macrophages form a stromal barrier limiting CD8+ T lymphocyte penetration into the tumor core), molecular-signaling (stabilization of PD-L1 by the NDR1/USP10 axis and exosomal transfer of miR-100-5p and PD-1; contextual sources point to VISTA and reduced MHC class I expression), and metabolic (lipid accumulation in tumor-associated macrophages and competition for nutrients, leading to T lymphocyte exhaustion). Conclusion: The data reviewed are consistent with the view that PCa resistance to immunotherapy is sustained by the mutual reinforcement of cellular–structural, molecular–signaling and metabolic barriers; targeting a single level is therefore unlikely to restore the antitumor response. These conclusions are generalized hypotheses: quantitative synthesis proved unfeasible, GRADE was not used to assess evidence certainty, and the model has not been prospectively validated. Comprehensive “immunome” assessment, rather than isolated PD-L1 determination, appears a promising direction for research on patient selection; no clinical validation of such models exists.