Nothing Lasts Forever, Including Infant BCG Vaccine Protection

Infants who received BCG [bacille Calmette-Guérin] vaccination at birth had protection against tuberculosis into early childhood, but that protection largely disappeared in only a few years, according to a meta-analysis.

Effectiveness of the BCG vaccination against tuberculosis in all age groups was 18%. However, when the results were stratified by age, only in children younger than 5 years did the infant vaccination show significantly effectiveness, reported Leonardo Martinez, PhD MPH, of Boston University, and colleagues in The Lancet Global Health.

Subgroup analyses, with data stratified by tuberculin skin test and IFNγ release assay results and by tuberculosis type, also showed trends toward declining efficacy with age: by the the time children reached adolescence, little to no efficacy from the infant vaccination was still evident. (However, the subgroup findings came with very broad 95% confidence intervals that made firm conclusions difficult to reach.)

Overall, the analysis suggested that booster doses may be necessary as children at risk for tuberculosis grow older.

“[T]here is widespread debate on the BCG vaccine’s effectiveness and duration of protection, as well as whether the vaccine only works in selective settings,” Martinez told MedPage Today. “Our findings indicate that BCG vaccination is effective at preventing tuberculosis in young children. Since tuberculosis in children can be a highly severe, debilitating disease in infants, it’s important for BCG vaccination to continue to be used.”

“However, since the results show that the vaccine was ineffective in adolescents and adults, boosting immunoprotection is needed for older populations,” Martinez continued. “We need novel vaccines urgently to supplement BCG vaccination in settings with a high-burden of tuberculosis.”

This meta-analysis drew on case-cohort studies of tuberculosis contacts published from January 1998 to April 2018. Participant-level data were included from 26 cohort studies performed in 17 countries, resulting in 68,552 total participants. In order to be included, an article’s dataset needed to have a minimum follow-up period of six months, individuals with close exposure information on age and sex of tuberculosis contact, and study start and follow-up dates.

Studies were excluded if information on BCG vaccination was not provided or if performed in a country that did not recommend BCG vaccination at birth (the U.S. does not). Individual-level participant data considered a list of variables, including characteristics of the exposed participant (contact), the index case, and the environment.

Exposure was defined as being in close contact (either living in the same household or having substantial interaction outside of household) with a person with microbiologically or radiologically diagnosed pulmonary tuberculosis.

The primary outcome assessed a composite of prevalent (diagnosed at or within 90 days of baseline evaluation) and incident (new tuberculosis case diagnosed more than 90 days after baseline) tuberculosis cases. Secondary outcomes considered the occurrence of pulmonary tuberculosis, extrapulmonary tuberculosis, and mortality.

Data were adjusted for variable of interest, baseline age, sex, previous tuberculosis, and whether data were collected prospectively or retrospectively. Results were stratified by contact age and Mycobacterium tuberculosis infection status.

Martinez and colleagues expressed vaccine efficacy as odds ratios for infection, comparing vaccinated and exposed individuals to those not vaccinated. In many of the analyses, odds ratios were close to 1 for individuals 10 and older, and often greater than 1. But in no case was there a statistically significant result indicating that BCG increased the risk of contracting tuberculosis.

Only for mortality did infant vaccination provide significant protection in adolescence, with an adjusted odds ratio of 0.13 (95% CI 0.04-0.41) for ages 10-14, which was similar to the efficacy in younger groups. But for those 15 and older, there was no apparent benefit at all (aOR 1.13, 0.35-3.68).

The studies used in this meta-analysis were observational in nature and, therefore, not randomized. Despite adjustments, residual or unmeasured confounding variables are possible. BCG vaccines were classified by vaccine scar, which may lead to misclassification if a scar did not occur. In addition, mortality analysis must be interpreted carefully, as vaccinated children may be from higher socioeconomic status and may have greater access to healthcare. Finally, extrapulmonary tuberculosis is often underreported.

“BCG vaccination was first invented over 100 years ago,” Martinez told MedPage Today. “Our findings suggest its effectiveness overall is lacking and only really works among young children <5 years of age. Despite the fact that over 10 million people develop tuberculosis every year leading to over 1 million deaths, a new vaccine has not become available.”

“We desperately need new research on novel vaccines that can prevent development of tuberculosis — especially among adolescents and adults,” Martinez added. “Several novel vaccines have shown some promise however much more funding and work is needed in this area.”