RESUMO
INTRODUCTION: Bacille Calmette-Guérin (BCG) vaccine has beneficial non-specific effects on overall survival. After BCG vaccination, positive PPD response and scar formation are associated with increased survival. During a trial randomising low-birth-weight neonates to BCG at birth or the usual delayed BCG, the manufacturer of the BCG vaccine experienced a period with relatively slow growth rate of the BCG. We investigated the association between growth rate of BCG when manufacturing the vaccine and its capability to induce immune responses in vivo and in vitro. METHODS: 1633 neonates were randomised to BCG at birth and examined for scar at 12 months; a subgroup was tested for PPD response at 2 and 6 months. The BCG batches from the Slow growth period were compared with the precedent and subsequent Normal growth batches with regard to prevalence and size of BCG scar and PPD response. We also tested the effect of batches on in vitro cytokine responses. RESULTS: At 12 months, the Slow growth batches were associated with higher BCG scar prevalence (98.2%) than the precedent batches (92.3%, p=0.01) but the prevalence remained high after return to normal growth (98.8%, p=0.52). The Slow growth batches were associated with larger scar size (5.0mm) than precedent (4.4mm, p<0.01) and subsequent batches (4.8mm, p=0.03). Compared with Normal growth batches, the Slow growth batches were associated with a higher prevalence of positive PPD responses, and among PPD positive children, a larger PPD reaction (geometric mean ratio: 1.40 (1.20-1.63)) at 2 months. In response to secondary heterologous stimulation, monocytes primed with Slow growth batches induced higher IL-6 (p=0.03) and TNF-α responses (p=0.03) compared with Normal growth batches. CONCLUSION: The study indicates that variations in the production of BCG vaccine may influence important immunological effects of the vaccine. TRIAL REGISTRATION: clinicaltrials.gov (NCT00625482).