Improving Tuberculosis Vaccine Trial Efficiency: A Tough Nut to Crack

The only licenced vaccine against tuberculosis (TB), Bacille Calmette Guerin (BCG), was developed in 1921.More than 100 years later, the need for a more effective vaccine is more pressing than ever.Rates of drug-resistant TB are rising and the COVID-19 pandemic had a devastating effect on TB control programmes, reversing improvements in mortality and morbidity seen in recent years (1).TB vaccine development is extremely challenging, in part because Mycobacterium tuberculosis (M.tb) is a highly complex organism, adept at evading and subverting the host protective immune response.We do not fully understand what constitutes a protective immune response.Validated immune correlates of protection with which to guide vaccine R&D are lacking.Modelling the real-world environment of exposure to common co-infections such as HIV, helminths, nontuberculous mycobacteria and CMV, which may influence vaccine efficacy, is not feasible in preclinical animal models and the predictive value of animal models for human efficacy is unclear.There is currently no substitute for human efficacy testing -but efficacy trials with new TB vaccine candidates are costly and take years to enrol, even in the highest burden settings.We need to explore more efficient ways to conduct these trials and demonstrate human efficacy.Using M.tb infection, rather than TB disease, as an endpoint, may be one way to achieve this (2).However, whilst infection is a prerequisite for disease, the relationship between immunity against infection and immunity against disease is unclear.Using a Prevention of Infection efficacy endpoint as a stage gate for a Prevention of Disease trial may lead to the premature rejection of an effective TB vaccine.