The number of human tuberculosis (TB) cases that are due to transmission from animals, as opposed to human-to-human transmission, may be much higher than previously estimated, according to an international team of researchers. The results could have implications for epidemiological studies and public health interventions.
"Tuberculosis kills 1.4m people every year, making it the most-deadly disease arising from a single infectious agent," said Vivek Kapur, professor of microbiology and infectious diseases and Huck distinguished chair in global health, Penn State. "India has the largest burden of human tuberculosis globally, with more than 2.6m cases and 400,000 deaths reported in 2019. Additionally, the cattle population in India exceeds 300m, and nearly 22m of these were estimated to be infected with TB in 2017.
Kapur noted that the World Health Organisation, World Organisation for Animal Health and Food and Agriculture Organisation of the United Nations define zoonotic TB as human infection with Mycobacterium bovis, a member of the Mycobacterium tuberculosis complex (MTBC).
To evaluate the use of M. bovis as a proxy for zoonotic tuberculosis and to investigate the potential role of other MTBC subspecies, Kapur and his colleagues analysed 940 bacterial samples – both pulmonary (from lung fluid or tissue) and extra-pulmonary (from tissues other than the lungs) – collected from patients who were visiting a large reference hospital for TB in southern India. The researchers used PCR to speciate M. tuberculosis complex organisms and then sequenced all the non-M. tuberculosis samples. Next, they compared the sequences to 715 sequences from cattle and humans that had previously been collected in south Asia and submitted to public databases.
"Surprisingly, we did not find any evidence for the presence of M. bovis in any of the samples," said Sreenidhi Srinivasan, post-doctoral scholar in the Huck Institutes of the Life Sciences. "Instead, we found that seven of the patient samples contained M. orygis. Six of these came from patients with extrapulmonary TB."
As expected, most of the remainder of the sequences from the patients belonged to M. tuberculosis – the TB bacterium that is generally thought to be transmitted only among humans. "Our findings suggest that M. bovis might be uncommon in India, and that its detection may not be an adequate proxy for zoonotic TB infection in humans," said Srinivasan. "These data indicate that members of the TB complex other than M. bovis might be more prevalent in livestock in India."
Kapur added that the operational definition of zoonotic TB should be broadened to include other MTBC subspecies capable of causing human disease.
"By 2035, the World Health Organisation is aiming to reduce the incidence of tuberculosis by 90% as a part of its End TB Strategy," he said. "The increasing evidence supporting M. orygis endemicity in south Asia and the identification of M. tuberculosis in cattle highlight the importance of using a One Health approach, involving multisectoral collaboration across the veterinary and clinical sectors, to meet the WHO's goal in India."
Background: Zoonotic tuberculosis is defined as human infection with Mycobacterium bovis. Although globally, India has the largest number of human tuberculosis cases and the largest cattle population, in which bovine tuberculosis is endemic, the burden of zoonotic tuberculosis is unknown. The aim of this study was to obtain estimates of the human prevalence of animal-associated members of the Mycobacterium tuberculosis complex (MTBC) at a large referral hospital in India.
Methods: We did a molecular epidemiological surveillance study of 940 positive mycobacteria growth indicator tube (MGIT) cultures, collected from patients visiting the outpatient department at Christian Medical College (Vellore, India) with suspected tuberculosis between Oct 1, 2018, and March 31, 2019. A PCR-based approach was applied to subspeciate cultures. Isolates identified as MTBC other than M tuberculosis or as inconclusive on PCR were subject to whole-genome sequencing (WGS), and phylogenetically compared with publicly available MTBC sequences from south Asia. Sequences from WGS were deposited in the National Center for Biotechnology Information Sequence Read Archive, accession number SRP226525 (BioProject database number PRJNA575883).
Findings: The 940 MGIT cultures were from 548 pulmonary and 392 extrapulmonary samples. A conclusive identification was obtained for all 940 isolates; wild-type M bovis was not identified. The isolates consisted of M tuberculosis (913 [97·1%] isolates), Mycobacterium orygis (seven [0·7%]), M bovis BCG (five [0·5%]), and nontuberculous mycobacteria (15 [1·6%]). Subspecies were assigned for 25 isolates by WGS, which were analysed against 715 MTBC sequences from south Asia. Among the 715 genomes, no M bovis was identified. Four isolates of cattle origin were dispersed among human sequences within M tuberculosis lineage 1, and the seven M orygis isolates from human MGIT cultures were dispersed among sequences from cattle.
Interpretation: M bovis prevalence in humans is an inadequate proxy of zoonotic tuberculosis. The recovery of M orygis from humans highlights the need to use a broadened definition, including MTBC subspecies such as M orygis, to investigate zoonotic tuberculosis. The identification of M tuberculosis in cattle also reinforces the need for One Health investigations in countries with endemic bovine tuberculosis.
Funding: Bill & Melinda Gates Foundation, Canadian Institutes for Health Research.
Shannon C Duffy, Sreenidhi Srinivasan, Megan A Schilling, Tod Stuber, Sarah N Danchuk, Joy S Michael, Manigandan Venkatesan, Nitish Bansal, Sushila Maan, Naresh Jindal, Deepika Chaudhary, Premanshu Dandapat, Robab Katani, Shubhada Chothe, Maroudam Veerasami, Suelee Robbe-Austerman, Nicholas Juleff, Vivek Kapur, Marcel A Behr
Penn State material
The Lancet Microbe abstract