Abstract
BACKGROUND: Tuberculosis (TB) remains a leading global cause of infectious mortality. Accelerating vaccine development requires validated immune correlates of protection (CoPs). Mechanistic studies have long highlighted Th1 cytokines (IFN-γ, IL-2, TNF-α) as crucial for anti-mycobacterial immunity, leading to the hypothesis that antigen-specific Th1 responses, particularly polyfunctional T cells, may serve as a CoP. However, clinical evidence linking these responses to protection has been inconsistent. METHODS: We conducted a systematic review and meta-analysis to evaluate antigen-specific IFN-γ, IL-2, and TNF-α as correlates of protection or risk. We searched PubMed/MEDLINE, Embase, Web of Science, and Cochrane Central up to June 30, 2025. We included human studies with longitudinal follow-up that measured these cytokines and reported progression to active TB disease (primary analysis) or sustained IGRA conversion (secondary analysis). Study selection, data extraction, and risk-of-bias assessment were performed in duplicate. Random-effects meta-analyses were conducted where feasible, pooling across different antigen classes (e.g., PPD, BCG, ESAT-6/CFP-10, vaccine antigens) and assay platforms (e.g., ICS, ELISpot, whole-blood). Although these assays and antigens differ in their ability to activate distinct immune responses, the pooled estimates reflect general trends in immune markers across various immunological contexts. This pooling approach was taken to maximize available data and evaluate broad immune responses. Where possible, we conducted subgroup and sensitivity analyses to explore the robustness of the findings across antigen and assay categories. RESULTS: From 1, 268 records, 10 studies were included. In the primary analysis of active TB disease (n=6 studies), pooled odds ratios for IFN-γ, IL-2, TNF-α, and polyfunctional responses were all close to 1.0 (range: 0.97-1.11) with confidence intervals spanning the null and low heterogeneity (I² = 0%). In the secondary analysis of sustained IGRA conversion (n=5 studies), continuous measures of IFN-γ and IL-2 were marginally higher in converters (pooled MDs: 0.07 and 0.06, respectively). Binary analyses showed a consistent but modest trend toward positive association (pooled ORs: 1.13 for IFN-γ, 1.07 for IL-2), though confidence intervals included 1.0. Sensitivity and subgroup analyses were conducted, including stratification by antigen type (e.g., PPD, BCG, ESAT-6/CFP-10, vaccine antigens) and assay platform (e.g., ICS, ELISpot, whole-blood). While no significant effect modification was observed, this stratification underscores the importance of considering heterogeneity across different antigenic stimuli and assay methodologies when interpreting the pooled results. A Baujat plot identified specific studies (e.g., Kagina et al., Nemes et al.) as primary contributors to heterogeneity. CONCLUSION: Available prospective evidence does not support antigen-specific Th1 cytokine magnitudes-individually or as polyfunctional profiles-as reliable, standalone correlates of protection against progression to active TB disease. These responses appear more strongly associated with immune activation states linked to recent antigen exposure or infection risk. The findings underscore the need to look beyond peripheral Th1 cytokine levels, recognizing that the pooled estimates across different antigen types and assay platforms reflect general directional trends rather than directly comparable quantitative effects. This highlights the need for a more nuanced approach, considering the diversity of immune responses induced by varying antigens and assay techniques. Future studies should aim for standardized, harmonized assays and endpoint definitions to allow for more accurate comparisons across different study designs and populations. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/prospero/, identifier INPLASY202610094.