Description of participant characteristics

A total of 28 LTBI-DM (71.4% female) and 17 HC (58.8% female) participants were included, as shown in Table 2. The LTBI-DM participants were older (median age 48.0 [25.0–69.0] years) compared to HC participants (31.0 [26.0–42.0] years). Among the LTBI-DM group, the median number of children was 5.0 (2.0–17.0), and a higher proportion reported alcohol use (21.4% vs. none among HC). The median BMI was also higher in the LTBI-DM (27.7 [16.9–40.2] kg/m² vs. 25.3 [22.3–31.7] kg/m²), consistent with elevated waist and hip circumferences measured only in LTBI-DM participants. Reflecting their diabetic status, the LTBI-DM participants showed elevated fasting blood glucose (11.1 [4.0-31.2] mmol/L) and HbA1c (9.2% [4.0–15.0]) levels, and all received medications (primarily metformin alone or with glibenclamide). By month 6, the LTBI-DM median HbA1c had decreased to 7.9% (6.5–9.9), reflecting ongoing clinical management of DM. Most LTBI-DM participants (78.6%) and controls (94.1%) had a BCG scar, and while 89.3% of LTBI-DM participants had no prior TB history, all HC participants also reported no TB history.

IPT decreases Mtb-specific CD4⁺ and CD8⁺ T-cell effector memory phenotypes in LTBI-DM participants

We performed a phenotypic analysis of CD4⁺ and CD8⁺ T-cell subsets of participant PBMC samples with LTBI-DM and HC. Flow cytometry was used to classify Mtb-specific T-cell memory phenotypes into four categories based on the expression of CD45RA and CCR7 as a percentage of total CD4⁺ and CD8⁺ T cells. The T-cell memory phenotypes were defined as: central memory (CM; CD45RA⁻CCR7⁺), naïve (CD45RA⁺CCR7⁺), terminally differentiated effector memory (TEMRA; CD45RA⁺CCR7⁻), and effector memory (EM; CD45RA⁻CCR7⁻) (Fig. 1). For LTBI-DM participants, effector memory CD4⁺ T cells were significantly decreased from M0 to M6 [median (IQR); 11.55 (7.78–15.45): 11.30 (6.60-23.78); p = 0.043] (Fig. 2C). Similarly, effector memory CD8⁺ T cells were significantly decreased from M0 to M6 [30.85 (16.33–45.25): 20.15 (13.95–34.43); p = 0.031] (Fig. 2G). In comparison, the EM CD4⁺ and CD8⁺ T-cell frequencies in HC participants were 12.20 (8.30–14.2) and 25.00 (18.70–28.70), respectively, and non-significant from M0 and M6 (Fig. 2C and G). No significant trends were observed for the CD4⁺ and CD8⁺ Naïve, Central Memory and TEMRA T cells in the LTBI-DM and HC participants (Fig. 2A, B, D, E, F, H).

Mtb-specific CD4⁺ and CD8⁺ T-cell memory phenotypes before and after IPT for LTBI-DM; and HC participants.

PBMCs were stimulated and cultured for 18 h with ESAT-6 and CFP-10 peptide pools plus brefeldin A and stained for surface markers. (B-E and G-J) Representative combined dot-violin-box plots depicting the percentage expression of naïve, central memory, effector memory and TEMRA phenotypes in CD4⁺ and CD8⁺ T cells before and post-IPT in LTBI-DM; and HC participants. The grey lines depict individual participant phenotype trajectories from M0 to M6, while the red line represents the aggregate median trend for LTBI-DM participants over the same period. Size of group: LTBI-DM (n = 24), HC (n = 17). Wilcoxon matched-pairs signed rank and Mann-Whitney U tests were performed to determine the statistical significance between M0 and M6 for LTBI-DM, as well as between HC and M0/M6, respectively. *p < 0.05. Non-significant (ns) p-values were not shown.

IPT increases Mtb-specific CD4⁺ and CD8⁺ T-cell CXCR5 expression and decreases CXCL13 protein levels in LTBI-DM participants

Next, we assessed the Mtb-specific CXCR5 expression and CXCL13 levels to determine the homing capabilities of CD4⁺ and CD8⁺ T-cells (Fig. 3). For LTBI-DM participants, CD4⁺ and CD8⁺ CXCR5 expression was significantly upregulated post-IPT [16.60 (12.80-18.55): 18.05 (15.63–21.25); p = 0.038] (Fig. 3B) and [1.79 (1.35–2.40): 2.60 (2.00-3.45); p = 0.006] (Fig. 3D) respectively. The HC participants showed the CXCR5 expression of 18.05 (15.93–24.05) for CD4⁺ T cells and a significantly higher CXCR5 expression on CD8⁺ T cells [3.05 (2.28–3.59); p = 0.012] compared to LTBI-DM at baseline (Fig. 3B and D). Conversely, high levels of CXCL13 were measured in the plasma of LTBI-DM participants at baseline, and these decreased significantly post-IPT [35.55 (26.17–44.44): 27.77 (24.07–35.25); p = 0.007] (Fig. 3E). Additionally, HC participants had the CXCL13 levels significantly decreased compared to the baseline LTBI-DM participants [21.44 (17.25–35.56) vs. 35.55 (26.17–44.44); p = 0.013] (Fig. 3E).

Mtb-specific CD4⁺ and CD8⁺ T-cell CXCR5 and CXCL13 expression profiles before and after IPT for LTBI-DM; and HC participants.

Representative flow cytometry plots are shown for (A and C) CD4⁺ and CD8⁺ CXCR5 expression profiles. PBMCs were stimulated and cultured for 18 h with ESAT-6 and CFP-10 peptide pools plus brefeldin A and stained for surface markers. (B and D) Representative combined dot-violin-box plots depict the percentage expression of functional profiles in CD4⁺ T cells before and post-IPT in LTBI-DM; and HC participants. (E) CXCL13 protein expression levels were measured by ELISA in the plasma of LTBI-DM and DM participants before and post-IPT, as well as the HC. The grey lines depict individual participant phenotype trajectories from M0 to M6, while the red line represents the aggregate median trend for LTBI-DM participants over the same period. Size of group: LTBI-DM (n = 24) for CXCR5, LTBI-DM (n = 28) for CXCL13, HC (n = 17). Wilcoxon matched-pairs signed rank and Mann-Whitney U tests were performed to determine the statistical significance between M0 and M6 for LTBI-DM, as well as between HC and M0/M6, respectively. p < 0.05 (*), p < 0.01 (**). Non-significant (ns) p-values were not shown.

IPT decreases Mtb-specific CD4⁺ and CD8⁺ T-cell PD-1 but not HLA-DR expression in LTBI-DM participants

Next, to characterise the Mtb-specific exhaustion and activation profiles of CD4⁺ and CD8⁺ T-cells, we determined the expression profiles of PD-1 and HLA-DR (Fig. 4A and E). For LTBI-DM participants, CD4⁺ and CD8⁺ T-cell PD-1 expression was significantly downregulated post-IPT [32.75 (11.80–42.80): 24.70 (9.95–38.43); p = 0.042] and [26.00 (16.23–40.38): 20.05 (12.48–35.18); p = 0.031], respectively (Fig. 4B and F). The HC participants exhibited significantly lower CD4⁺ PD-1 expression compared to LTBI-DM participants at baseline [16.70 (13.60–20.40); p = 0.018] (Fig. 4B), indicating a trend toward normalisation post-IPT in LTBI-DM participants for CD4⁺ T cells. Conversely, CD4⁺ and CD8⁺ T-cell HLA-DR expression did not differ before and post-IPT for LTBI-DM participants [0.91 (0.56–1.91): 0.83 (0.65–2.57); p = 0.107] and [0.95 (0.68–1.30): 1.11 (0.66–1.98); p = 0.331)], respectively (Fig. 4C and G). Notably, HC participants showed significantly upregulated HLA-DR expression compared to LTBI-DM at both M0 [CD4⁺: 2.53 (1.85–3.31), p = 0.006; CD8⁺: 1.84 (1.58–2.30), p = 0.006] and M6 [CD4⁺: p = 0.013] (Fig. 4C and G). Additionally, for LTBI-DM participants, CD8⁺ T-cell CD107a production was significantly increased post-IPT [1.59 (0.94–2.40): 2.16 (1.36–3.14); p = 0.037] (Fig. 4H), while CD4⁺ T-cell CD107a was non-significant post-IPT [0.57 (0.52–0.98): 0.78 (0.55–1.29); p = 0.360] (Fig. 4D).

The HC participants showed significantly upregulated CD4⁺ CD107a production [1.30 (0.82–2.20)] compared to LTBI-DM at M0 (p = 0.003) and M6 (p = 0.024) (Fig. 4D).

Mtb-specific CD4⁺ and CD8⁺ T-cell PD-1, HLA-DR, and CD107a before and after IPT for LTBI-DM; and HC participants.

Representative flow cytometry plots are shown for (A and E) CD4⁺ and CD8⁺ PD-1, HLA-DR, and CD107a functional profiles. PBMCs were stimulated and cultured for 18 h with ESAT-6 and CFP-10 peptide pools plus brefeldin A and stained for surface markers. (B-D and F-H) Representative combined dot-violin-box plots that depict the percentage expression of PD-1, HLA-DR, and CD107a functional profiles in CD4⁺ and CD8⁺ T cells in LTBI-DM; and HC participants. The grey lines depict individual participant phenotype trajectories from M0 to M6, while the red line represents the aggregate median trend for LTBI-DM participants over the same period. Size of group: LTBI-DM (n = 24), HC (n = 17). Wilcoxon matched-pairs signed rank and Mann-Whitney U tests were performed to determine the statistical significance between M0 and M6 for LTBI-DM, as well as between HC and M0/M6, respectively. p < 0.05 (*), p < 0.01 (**). Non-significant (ns) p-values were not shown.

IPT upregulates Mtb-specific CD4⁺ T-cell IL-17 A and CD8⁺ T-cell IL-13 production but not IFN-γ in LTBI-DM participants

To further assess the quality of CD4⁺ and CD8⁺ T-cell responses, we determined the Mtb-specific production of cytokines (IFN-γ, IL-13 and IL-17 A) by the two subsets. Among LTBI-DM participants, CD4⁺ T-cell IL-17 A production was upregulated post-IPT [0.49 (0.20–0.69): 1.01 (0.63–1.31); p = 0.004] (Fig. 5C). This post-IPT IL-17 A production was comparable to the higher levels observed in HC participants [0.97 (0.71–1.47); p = 0.001] compared to LTBI-DM participants at baseline (Fig. 5C). Although CD4⁺ T-cell IFN-γ and IL-13 production in LTBI-DM participants was non-significant post-IPT [0.57 (0.24–0.85): 0.53 (0.32–0.90)] and [0.33 (0.14–0.61): 0.50 (0.27–0.86)], respectively, IFN-γ production in HC participants was significantly lower than in LTBI-DM participants at M6 [0.27 (0.17–0.43); p = 0.040] (Fig. 5A and B). For LTBI-DM participants, CD8⁺ T-cell IL-13 production was significantly upregulated post-IPT [0.45 (0.23–0.79): 0.80 (0.52–1.20); p = 0.033] (Fig. 5E). In addition, CD8⁺ T-cell production IFN-γ and IL-17 A production was not upregulated post-IPT (Fig. 5D and F). Conversely, CD8⁺ T-cell IFN-γ [0.52 (0.28–1.29)] production at baseline in LTBI-DM participants was significantly higher compared to HC participants [1.27 (0.75–2.05); p = 0.040], with no changes post-IPT (Fig. 5D).

Mtb-specific CD4⁺ and CD8⁺ T-cell cytokine profiles before and after IPT for LTBI-DM; and HC participants.

Representative flow cytometry plots are shown for (A and D) CD4⁺ and CD8⁺ T-cell IL-17 A and IL-13 cytokine profiles, respectively. PBMCs were stimulated and cultured for 18 h with ESAT-6 and CFP-10 peptide pools plus brefeldin A and stained for intracellular markers. (B-C and E-F) Representative combined dot-violin-box plots depicting the percentage expression of cytokine profiles in CD4⁺ and CD8⁺ T-cell IFN-γ, IL-13 and IL-17 A cytokine profiles in LTBI-DM; and HC participants. The grey lines depict individual participant phenotype trajectories from M0 to M6, while the red line represents the aggregate median trend for LTBI-DM participants over the same period. Size of group: LTBI-DM (n = 24), HC (n = 17). Wilcoxon matched-pairs signed rank and Mann-Whitney U tests were performed to determine the statistical significance between M0 and M6 for LTBI-DM, as well as between HC and M0/M6, respectively. p < 0.05 (*), p < 0.01 (**).