Mycobacterium bovis bacillus Calmette-Guérin (BCG) is a live attenuated vaccine which can result in local or disseminated infection, most commonly in immunocompromised individuals. Differentiation of BCG from other members of the Mycobacterium tuberculosis complex (MTBC) is required to diagnose BCG disease, which requires specific management. Current methods for BCG diagnosis are based on mycobacterial culture and conventional PCR; the former is time-consuming and the latter often unavailable. Further, there are reports that certain BCG strains may be associated with a higher rate of adverse events. This study describes the development of a two-step multiplex real-time PCR assay which uses single nucleotide polymorphisms to detect BCG and identify early or late BCG strains. The assay has a limit of detection of 1 pg BCG boiled lysate DNA and was shown to detect BCG in both pure cultures and experimentally infected tissue. Its performance was assessed on 19 suspected BCG clinical isolates at Christian Medical College in Vellore, India, taken from January 2018 to August 2020. Of these 19 isolates, 10 were identified as BCG (6 early and 4 late strains), and 9 were identified as other MTBC members. Taken together, the results demonstrate the ability of this assay to identify and characterize BCG disease from cultures and infected tissue. The capacity to identify BCG may improve patient management, and the ability to discriminate between BCG strains may enable BCG vaccine pharmacovigilance. IMPORTANCE Vaccination against tuberculosis with bacillus Calmette-Guérin (BCG) can lead to adverse events, including a rare but life-threatening complication of disseminated BCG. This complication often occurs in young children with immunodeficiencies and is associated with an ∼60% mortality rate. A rapid method of reliably identifying BCG infection is important because BCG requires treatment unique to tuberculosis. BCG is resistant to the first-line antituberculosis drug pyrazinamide. Additionally, diagnosis of BCG disease would lead to further investigation of a possible underlying immune condition. We have developed a diagnostic assay to identify BCG which improves upon previously published methods and can reliably identify BCG from bacterial culture or directly from infected tissue. This assay can also differentiate between strains of BCG, which have been suggested to be associated with different rates of adverse events. This assay was validated on 19 clinical isolates collected at Christian Medical College in Vellore, India.
Keywords: BCG disease; Mycobacterium bovis BCG; Mycobacterium tuberculosis; disseminated BCG; real-time PCR identification.