Long-term efficacy of BCG vaccination in goat herds with a high prevalence of tuberculosis

Caprine tuberculosis (TB), primarily caused by Mycobacterium caprae and Mycobacterium bovis, affects animal health, poses zoonotic risks, and causes economic losses from reduced milk production and trade restrictions. In high-prevalence settings, test-and-slaughter alone may be insufficient or not cost-effective. Bacillus Calmette-Guérin (BCG) vaccination is a safe, relatively inexpensive tool with variable efficacy reported across livestock and wildlife. A key challenge is interference with standard diagnostics (SIT/SICCT and IGRA), addressed by DIVA antigens developed to differentiate infected from vaccinated animals. Risk factors such as herd size, housing, and feeding practices influence TB persistence. The purpose of this study was to assess the long-term (three-year) efficacy of BCG vaccination at herd level and among vaccination batches in five goat farms with high TB prevalence in Catalonia, Spain, and to identify risk factors affecting vaccine program efficacy.

Prior studies demonstrate that BCG can reduce TB lesion severity and transmission in livestock and wildlife, though protection is typically incomplete. Field trials in cattle and goats have shown encouraging outcomes, including reductions in lesions and bacterial load with both subcutaneous and oral BCG, but with variability linked to exposure intensity and management. Diagnostic interference has been mitigated by DIVA reagents (e.g., ESAT-6/CFP-10), validated in goats. Herd-level risk factors for TB persistence include herd size, intensive housing, and management practices. Experimental and field data indicate BCG confers protection in goat kids, but efficacy may wane over time and can be influenced by prior sensitization to environmental mycobacteria and co-infections such as helminths that skew immune responses, potentially reducing vaccine effectiveness. Revaccination strategies in cattle have shown improved protection when immunity wanes.

Study design: A 36-month field trial in five TB-infected goat farms in Catalonia, Spain (confirmed M. caprae by culture) selected for high initial SICCT positivity (>30%). Farms varied in size, production (meat or milk), management (intensive/extensive), and replacement origin; predominant spoligotypes were SB0415 and SB0157. Baseline (Month 0): All adult and young goats (>6 months) tested with SICCT and IGRA (PPD-B/PPD-A) and a DIVA IGRA (ESAT-6/CFP-10). Animals positive to any test were classified TB-positive. Animals negative to all tests were vaccinated subcutaneously with BCG Danish 1331 (approx. 2–8 × 10^5 CFU in 0.5 ml) behind the axilla. Vaccination protocol: All replacement goat kids entering during the trial were vaccinated at 2–3 months of age. Batches were defined by first sampling entry: B-06, B-12, B-24, B-36; adult goats vaccinated at M0 comprised batch B-0. Follow-up testing: All goats >6 months were tested at Months 6, 12, 24, 36 by IGRA using tuberculins and the DIVA reagent. Post-vaccination, vaccinated animals were considered TB-positive only by DIVA IGRA. Diagnostics: - SICCT: Intradermal PPD-B and PPD-A (25,000 IU/ml), positivity as per Spanish program criteria (SFT thresholds and clinical signs). - IGRA: Whole blood stimulated with PPD-A, PPD-B, and ESAT-6/CFP-10 (20 µg/ml). ELISA readout (BOVIGAM). Positivity: PPD-B OD − PBS OD ≥ 0.05 and PPD-B > PPD-A; DIVA positive if ESAT-6/CFP-10 OD − PBS OD ≥ 0.05. Outcomes and calculations: - Prevalence (P): % TB-positive animals at each sampling. - Attributable fraction (AF) as prevalence reduction: AF(i+1) = [P(i) − P(i+1)] / P(i). - Incidence risk (IR) in vaccinated animals per interval: New cases / [N(i) + replacements − losses], adjusting for mid-interval losses (deaths/slaughter assumed present half-period). Risk factor analysis: - Variables considered: management (intensive/extensive), ventilation (high/low), initial prevalence, initial census, number of unvaccinated positives remaining at end, number of vaccinated TB-positive adults (B-0) at end, and vaccinated TB-positive kids at end. - Statistical approach: Generalized mixed-effect models (Poisson) to screen variables (P<0.05 or P<0.3), followed by multilevel Poisson model including farm as random effect; model selection via AIC with stepwise reduction. Post hoc Tukey comparisons for batch incidence risk across exposure times. Chi-square tests compared proportions of positives at M0 vs M36 per farm (significance P<0.05). Ethics: Approved by Generalitat de Catalunya (procedure 8697); procedures followed EU and Spanish legislation.

- Herd-level prevalence and AF: • Four of five farms (1, 2, 3, 5) showed favorable outcomes by Month 36. Overall AFs (M0 to M36): Farm 1: 93.5%; Farm 2: 28.5%; Farm 3: 23.2%; Farm 5: 14.3%. Farm 4 showed no prevalence reduction (increase from ~33% to 46.5% by M36). • Across all five farms, prevalence decreased from 41% at M0 to 33.6% at M36 (overall AF 18.1%). • In the four favorable farms combined, TB positives fell from 546/1255 (43.5%) at M0 to 451/1506 at M36 (303 vaccinated positives; 148 unvaccinated positives), with 31% prevalence reduction (P<0.001). By M36, unvaccinated animals still present comprised 32.8% of total TB positives; 90.2% of goats were vaccinated and 77.7% of those remained TB-negative. - Incidence: • In favorable farms (1, 2, 3, 5), incidence in vaccinated animals generally declined progressively, with an overall 50% reduction from M6 to M36. Interval IR examples (vaccinated): Farm 1 IRs: 6%, 4%, 2%, 3%; Farm 2: 15%, 5%, 11%, 9%; Farm 3: 19%, 29%, 28%, 11%; Farm 5: 31%, 15%, 21%, 7%. Farm 4 showed increasing incidence (1%, 10%, 39%, 26%). - Age at vaccination effects: • At M36, adult goats vaccinated at M0 (B-0) had higher TB prevalence than vaccinated kids: Adults 142/321 (44%) vs kids 161/1037 (16%). • During the first year of exposure, batch B-0 had significantly higher TB risk than kid batches B-06, B-24, B-36 (generalized mixed-effect model with Tukey comparisons; significant contrasts with P≤0.0001). No significant differences among kid-vaccinated batches, and differences among batches diminished after one year, indicating waning protection. - Risk factors (multilevel model): • Significant predictors increasing TB incidence in vaccinated animals: (1) Number of positive unvaccinated animals remaining in herds (P<0.05). (2) Number of positive adult vaccinated goats (B-0) (P<0.05). • Initial herd prevalence showed a trend but was not significant (P<0.1). Ventilation (high/low), management type, initial census, and number of vaccinated TB-positive kids were not significant in the final model. - Farm-specific notes: • Farm 5 had an early prevalence rise (46.0% to 60.7% at M6) amid census drop (174→135) and poor airflow, then improved by study end. • Farm 4 likely included kids infected before vaccination (early exposure, unpasteurized colostrum), contributing to vaccine failure.

The study addressed whether systematic BCG vaccination reduces TB prevalence and incidence in high-prevalence goat herds and what factors influence outcomes. Results show that, despite heterogeneity across farms, sustained vaccination of TB-negative adults and replacement kids, paired with removal of unvaccinated positive animals, can reduce transmission (50% incidence reduction in favorable farms) and lower prevalence over three years. Protective effects were more pronounced in kids than in adults, especially during the first year of exposure, aligning with evidence that prior sensitization to environmental mycobacteria and helminth co-infections can diminish vaccine efficacy in adults. Intensive management and poor ventilation likely increased transmission pressure, as illustrated by Farm 5’s early setback. Protection appeared to wane beyond one year, with incidence risks converging among batches after extended exposure, echoing cattle studies observing diminished protection after 24 months. Nonetheless, high vaccination coverage and progressive culling of unvaccinated positives reduced intra-herd transmission, supporting vaccination as a complementary control measure in endemic settings. The identification of unvaccinated positives and adult vaccinated positives as key risk amplifiers underscores the importance of removing infectious sources and possibly reconsidering adult vaccination strategies or boosting schedules to maintain immunity. Overall, findings support integrating BCG vaccination with targeted management to control caprine TB under field conditions.

Systematic BCG vaccination of replacement goat kids in herds with high TB prevalence can progressively reduce transmission of Mycobacterium caprae and lower overall TB prevalence, even as individual protection wanes after approximately one year. The concurrent removal of unvaccinated TB-positive goats decreases the risk of new cases and enhances program success. High coverage vaccination and culling strategies are recommended as part of an integrated control program. Future research should assess optimized schedules including revaccination/boosting to counter waning immunity, refine DIVA-based diagnostics to improve sensitivity in vaccinated animals, and evaluate cost-benefit and operational feasibility across diverse management systems.

- Diagnostic constraints: After vaccination, only DIVA IGRA was used in vaccinated animals; prior studies indicate ~70% sensitivity (100% specificity), implying up to 30% false negatives among vaccinated, potentially attenuating interval incidence estimates. - Heterogeneity and external factors: Farm-level intrinsic factors (management intensity, ventilation, early-life exposures, feeding practices such as unpasteurized colostrum) varied and likely influenced outcomes; small number of farms (n=5) limits generalizability. - Potential pre-vaccination infection/anergy: Some adults vaccinated at M0 may have been infected but tested negative (IGRA/SICCT), including possible anergy in chronic infections, reducing apparent vaccine effectiveness and contributing to transmission. - Waning immunity: Protection diminished after one year, complicating long-term impact assessment without boosting. - Population fluctuations: Changes in herd census (e.g., Farm 5) influenced prevalence metrics; weather and co-infections (e.g., helminths) may have affected transmission and immune responses but were not controlled.