Dose and Route of rBCG-LTAK63 Immunization Determine Protective Efficacy and Pathological Outcome in Tuberculosis

BCG shows variable protection against adult pulmonary tuberculosis. Recombinant BCG platforms, such as rBCG-LTAK63, aim to improve efficacy, but how dose and route influence protection versus immunopathology remains to be determined. We investigated how these factors shape bacterial control, lung injury, inflammation, and CD4⁺ memory.

We established a dose-response by subcutaneous (SC) immunization and then fixed a high dose (1 × 10⁷ CFU) to compare SC, intranasal (IN), and intravenous (IV) administration. Lung CFU, histopathology, lung cytokines, and CD4⁺ memory phenotypes in lung and spleen were quantified.

rBCG-LTAK63 showed a clear SC dose-response, with higher doses improving bacterial control and immunogenicity. At the fixed high dose, outcomes diverged by route. IN vaccination produced the greatest reduction in lung CFU but also induced marked lung injury, with increased inflammation, fibrosis, necrosis, together with elevated TNF, IL-6, and IL-17. On the other hand, IV delivery achieved robust bacterial control with minimal pathology and promoted IFN-γ-skewed CD4⁺ responses enriched for lung central and tissue-resident memory. SC resulted in the lowest bacterial control but still a ~2 log reduction, and predominantly systemic responses. These patterns reveal distinct correlates: IFN-γ with lung TCM/TRM associated with protection, and TNF/IL-6/IL-17 with inflammatory/TEM bias associated with pathology.

rBCG-LTAK63 safety and efficacy are strongly route- and dose-dependent. The same high antigenic dose is beneficial via IV yet crosses an inflammatory threshold via IN. These findings support route-specific dose optimization, including defining minimal effective IV doses and maximal safe IN doses, and motivate evaluation of fractional, heterologous, or mixed-route regimens. Understanding how protection and pathology uncouple by route is critical for advancing rBCG-LTAK63 toward clinical development.