CELL INTELLIGENCE · 肿瘤细胞治疗研究

PURPOSE: Adoptive tumor-infiltrating lymphocyte (TIL) therapy is an established personalized cellular immunotherapy with demonstrated activity in selected solid tumors, particularly metastatic melanoma. However, clinical outcomes, safety, and manufacturing feasibility vary across tumor types and treatment strategies. This systematic review evaluates the efficacy, safety, and operational characteristics of TIL therapy across solid malignancies. METHODS: This systematic review was conducted in accordance with PRISMA guidelines. PubMed, Scopus, the Cochrane Central Register of Controlled Trials, and the WHO International Clinical Trials Registry Platform were searched from inception to 31 December 2025. Eligible studies included clinical trials and observational studies evaluating autologous TIL therapy in solid tumors. Due to substantial clinical and methodological heterogeneity, quantitative synthesis was restricted to clinically comparable cohorts, predominantly melanoma studies reporting objective response rates (ORR). A single-arm random-effects meta-analysis using the Freeman-Tukey transformation was performed. All other outcomes, including survival, safety, manufacturing success, and resection-to-infusion time, were synthesized narratively. RESULTS: Thirty-eight studies were included: 5 randomized controlled trials (RCTs), 22 prospective non-randomized studies, and 11 retrospective analyses, spanning melanoma and 8 other solid tumour types. Meta-analysis of 18 melanoma single-arm cohorts demonstrated a pooled objective response rate (ORR) of 42% (95% CI 37%-47%; I² = 33.1%; prediction interval 29%-56%) under a random-effects model. In the phase III RCT (Rohaan et al.), TIL therapy produced superior ORR (49% vs. 21%) and progression-free survival (median 7.2 vs. 3.1 months; HR 0.50, 95% CI 0.35-0.72) compared with ipilimumab. Subgroup analysis by TIL product type revealed a statistically significant difference (χ² = 7.25, p = 0.0266): tumor-reactive TIL products pre-screened ex vivo for antigen-specific reactivity showed the highest pooled ORR at 50% (95% CI 35%-64%), followed by young TIL at 43% (95% CI 29%-58%) and bulk TIL at 36% (95% CI 31%-42%); this observation is based on only 4 cohorts with a limited aggregate patient number and should be regarded as hypothesis-generating. No significant difference in ORR was observed by lymphodepletion status (p = 0.9544). Evidence in non-melanoma solid tumours was limited and heterogeneous, with generally lower response rates. Safety profiles were consistent across studies and primarily attributable to lymphodepleting chemotherapy and interleukin-2 administration, including haematologic and cytokine-related toxicities; treatment-related mortality was uncommon. Manufacturing success rates were high across contemporary cohorts, with a resection-to-infusion time typically spanning 4-6 weeks. CONCLUSION: TIL therapy demonstrates consistent and clinically meaningful antitumor activity in melanoma, while evidence in non-melanoma tumors remains limited and heterogeneous. Future studies should prioritize biomarker-driven patient selection, optimization of manufacturing and conditioning strategies, and rational combination approaches to expand its applicability. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/PROSPERO/, identifier CRD420261291389.