CLINICAL PREDICTION MODELS FOR IMMUNOTHERAPY BENEFIT IN ADVANCED SOLID TUMORS: CURRENT EVIDENCE, CHALLENGES, AND FUTURE DIRECTIONS IN LOW-RESOURCE SETTINGS

Dr. Swarnendu Biswas*

Background: Immune checkpoint inhibitors (ICIs) have revolutionized the treatment of advanced solid tumours by producing durable responses and improving survival across multiple malignancies. However, only a subset of patients derives sustained clinical benefit, while many experience primary or acquired resistance. Reliable predictive biomarkers are therefore essential to optimize patient selection, minimize unnecessary toxicity, and improve the cost-effectiveness of immunotherapy. Although several biomarkers have been investigated, their predictive performance varies considerably across tumour types, and no single biomarker has demonstrated universal clinical applicability. Aim: To systematically review and critically appraise the evidence published between January 2016 and June 2026 regarding clinical, pathological, molecular, immunological, blood-based, and imaging biomarkers associated with response to immune checkpoint inhibitor therapy in patients with advanced solid tumours. Methods: A systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines. Electronic databases including PubMed/MEDLINE, Embase, Scopus, Web of Science, Cochrane Library, and ClinicalTrials.gov were searched for studies published between January 2016 and June 2026. Eligible studies included randomized controlled trials, prospective and retrospective cohort studies, and observational studies evaluating predictive biomarkers of immunotherapy response in adults with advanced solid tumours. Two reviewers independently screened studies, extracted data, and assessed methodological quality using validated risk-of-bias tools appropriate to study design. Results: A total of 8,200 records were identified through database searching. After removal of duplicates and application of predefined eligibility criteria, 211 studies were included in the qualitative synthesis. Evidence demonstrated that programmed death-ligand 1 (PD-L1) expression remains the most widely used predictive biomarker in routine clinical practice, although its predictive accuracy is limited by biological heterogeneity, assay variability, and inconsistent scoring methods. Microsatellite instability-high (MSI-H) and deficient mismatch repair (dMMR) emerged as the most robust validated biomarkers, consistently predicting durable responses across multiple tumour types. Tumour mutational burden (TMB) provided complementary predictive information but was limited by lack of assay standardization and variable cut-off values. Among emerging biomarkers, circulating tumour DNA (ctDNA) kinetics showed the strongest evidence as a dynamic predictor of treatment response, with early ctDNA clearance correlating with improved objective response rate, progression-free survival, and overall survival. Tumour-infiltrating lymphocytes, interferon-γ-related gene-expression signatures, inflammatory biomarkers including neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lactate dehydrogenase (LDH), and the Lung Immune Prognostic Index (LIPI), together with radiomics and artificial intelligence (AI)-based prediction models, demonstrated promising but variable predictive value. Overall, composite biomarker strategies integrating clinical, molecular, and immunological variables consistently outperformed single biomarkers. Discussion: The findings indicate that immunotherapy response is determined by complex interactions between tumour biology, the immune microenvironment, host inflammatory status, and dynamic treatment-related changes. Although PD-L1, MSI-H/dMMR, and TMB remain important components of current clinical decision-making, emerging biomarkers—particularly ctDNA—offer significant opportunities to improve treatment prediction. Integration of multiple complementary biomarkers is likely to represent the future of precision immuno-oncology. In low- and middle-income countries (LMICs), incorporation of affordable biomarkers such as NLR, LDH, PLR, and LIPI alongside selected molecular testing may provide a pragmatic and cost-effective approach to patient selection. Conclusion: No single biomarker accurately predicts response to immune checkpoint inhibitor therapy across all advanced solid tumours. Current evidence supports the use of integrated, multimodal biomarker strategies that combine clinical, pathological, molecular, blood-based, and imaging parameters to optimize patient selection. Future research should focus on prospective validation of composite prediction models, standardization of emerging biomarker assays, and development of resource-stratified approaches applicable across diverse healthcare settings.