Different effects of BCG strains – A natural experiment evaluating the impact of the Danish and the Russian BCG strains on morbidity and scar formation in Guinea-Bissau. Issue 38 (31st August 2016)

• Record Type: Journal Article
• Authors: Frankel, H.; Byberg, S.; Bjerregaard-Andersen, M.; Martins, C.L.; Aaby, P.; Benn, C.S.; Fisker, A.B.
• Is Part Of: Vaccine. Volume 34:Issue 38(2016)
• Page Start: 4586
• Page End: 4593
• Publication Date: 2016-08-31
• Subjects: Non-specific/heterologous effects of vaccines -- BCG vaccine -- BCG strain -- BCG scar -- Child morbidity; Vaccines -- Periodicals; 615.372
• DOI: 10.1016/j.vaccine.2016.07.022
• Languages: English
• ISSNs: 0264-410X

Abstract

Highlights: BCG has beneficial non-specific effects protecting against non-tuberculosis deaths. Different BCG strains are used interchangeably, and may have different effects. In a natural experiment Danish strain provided higher scar prevalence and size. Danish BCG strain tended to lower morbidity for girls. Altering BCG strain production may lower infant morbidity and mortality at low cost. Abstract: Background: Different Bacillus Calmette-Guerin (BCG) vaccine strains may have different non-specific effects. We assessed the effect of two BCG strains (Danish and Russian) on childhood morbidity and BCG scarification in Guinea-Bissau. Methods: During 2011–2013, infants in the Bandim Health Project's urban study area received the Danish or Russian BCG in a natural experiment. Health center consultations were registered at point of care and scar status and size at age 4½ months. We assessed the effect of strain on consultation rates between vaccination and age 45 days in Cox proportional hazards models. Scar prevalence and size were compared using binomial regression and ranksum tests. Results: Among 1206 children, 18% received Danish BCG ( n = 215) and 82% Russian BCG ( n = 991). The adjusted hazard ratio (aHR) for consultations was 0.94 (95% CI 0.60–1.46) for Danish BCG compared with Russian BCG. Girls vaccinated with Danish BCG tended to have lower consultation rates compared with girls vaccinated with Russian BCG (aHR 0.56 (0.25–1.24)), whereas the effect was opposite for boys (aHR 1.24 (0.74–2.11)), p = 0.09. Children vaccinated with Danish BCG were more likely to develop a scar (97%) than children vaccinated with Russian BCG (87%), the relative risk (RR) being 1.11 (1.06–1.16). The effect was stronger in girls, and BCG scar size was larger among infants vaccinated with the Danish strain. Conclusion: BCG strain influences scar prevalence and scar size, and may have sex differential effects on morbidity. BCG strains are currently used interchangeably, but BCG scarring has been linked to subsequent survival. Hence, more research into the health effects of different BCG strains is warranted. Small adjustments of BCG production could potentially lower childhood morbidity and mortality at low cost.