Assessing the effect of BCG revaccination on long-term mortality

Vaccination is one of the most effective and efficient public health interventions of the past two centuries, with organisation of national vaccination programmes around the world being associated with significant decreases in mortality, especially mortality due to infections.¹ Vaccines are designed to induce heightened specific humoral (memory B cells producing specific antibodies) and cellular (memory T cells) responses against specific pathogens that upon reinfection ensure a rapid elimination of an infection. However, increasing evidence accumulated since the mass vaccination programmes of the 20th century suggest that some vaccines, especially those derived from live attenuated microorganisms, also have important beneficial heterologous effects outside the target disease. Probably the vaccine that has been most studied in relation to its protective heterologous effects is BCG, which was developed 100 years ago against tuberculosis.²

In The Lancet Infectious Diseases, Judith Glynn and colleagues³ report findings on the long-term effects of BCG revaccination on mortality in adults of various ages, providing an important piece of the complex puzzle of the broad effects of BCG vaccination. The strength of the study is its use of well balanced randomisation, with BCG revaccination (or placebo) administered in a large population of older children and adults living in Malawi. Despite long-term follow-up of 8 and 30 years in the two cohorts investigated, no effect of early BCG revaccination was shown. The study seems thus to exclude a long-term effect of BCG on mortality in this setting, which is important information about the limitation of BCG heterologous effects.

However, it is also important to highlight the conclusions that cannot be extracted from this study. First, the data collected by Glynn and colleagues do not permit clear distinction of effects in the initial 1–2 year interval after revaccination due to the very low number of deaths in this period and thus the lack of power. Indeed, the effects of BCG on mortality in infants have been shown to be strongest in the immediate period of weeks or months after BCG vaccination.⁴ In line with this, the putative immunological mechanisms responsible for these protective effects of BCG are the induction of heterologous T-cell immunity and innate immune memory (also termed trained immunity), both of which have a short duration of 1–2 years.2, 5 Second, as the authors also acknowledge, there was only a small number of deaths due to communicable diseases in the studied cohorts. Therefore, non-infectious mortality could have obscured any effect of BCG vaccination on infectious mortality, which would be the most likely expected potential effect. Further studies are therefore needed to assess the potential long-term effects of BCG revaccination on mortality due to infections in adults and older people. Additionally, although a long-term effect on mortality seems to be excluded, no information is available in this study on the effects of BCG vaccination on morbidity.

The absence of an effect on overall mortality by BCG revaccination observed by Glynn and colleagues is in line with other studies in the literature, as reviewed by the authors.³ However, not all previous studies indicate a lack of effect, as Glynn and colleagues also acknowledge and discuss. In addition, some studies were not reviewed by the authors because they assessed long-term effects of first BCG vaccination, rather than revaccination: these studies do seem to suggest potential decreased mortality in older children and adults aged up to 45 years, if previously vaccinated with BCG.⁶ In line with this, some large epidemiological studies suggest reduction of cancer prevalence in individuals vaccinated with BCG,⁷ which could suggest an important protective effect of BCG, in addition to prevention of infections, that might explain a potential effect on mortality in some populations. Indeed, BCG-induced activation of immune responses is routinely used in bladder cancer treatment, and has been suggested for other types of malignancies.⁸

Glynn and colleagues initiated their study to provide arguments for the usefulness (or not) of BCG vaccination or revaccination in adults as a preventive measure against COVID-19. This approach has been proposed on the basis of earlier studies showing a decrease of viraemia⁹ and prevalence of respiratory viral infections¹⁰ in adults vaccinated or revaccinated with BCG, with more than 15 ongoing randomised trials. Glynn and colleagues suggest that such an approach is not supported by the current study, yet acknowledge that no conclusions can be made about the short-term (1–2 years) effects of BCG on infection morbidity and mortality. Indeed, the answer to the question regarding the short-term effects of BCG vaccination on COVID-19 can be given only by the ongoing trials that are awaited with interest.

In conclusion, the study by Glynn and colleagues is an important investigation showing that BCG revaccination in an adult African population does not have an important effect on long-term overall mortality. Because other studies suggest beneficial effects of BCG vaccination in other populations, this should stimulate large epidemiological investigations that could give a definitive answer regarding the effects of BCG vaccination on mortality in adults of different ethnicities, demographics, and geographical locations. Finally, as new data accumulate on short-term protective effects of BCG vaccination on infections,¹⁰ randomised trials are warranted to establish the validity and strength of such effects.

© 2021 iStock - Matthew de Lange

We declare no competing interests.