Genome study links a drug‑resistant TB lineage to severe extrapulmonary cases

Extrapulmonary tuberculosis (EPTB) — tuberculosis that spreads outside the lungs — can be harder to diagnose and treat because drugs sometimes do not reach infected sites well. Researchers worried that certain strains of Mycobacterium tuberculosis might prefer these sites and carry more drug resistance, potentially causing treatment failure even when standard tests appear reassuring. To investigate, a team led by corresponding author Renu Verma analyzed 117 M. tuberculosis culture isolates collected at PGIMER, India from seven different extrapulmonary sites: cerebrospinal fluid (CSF, n = 40), pus (n = 35), fine-needle aspiration cytology (FNAC, n = 22), tissue (n = 8), ileocaecal biopsy (n = 6), synovial fluid (n = 5), and vitreous fluid (n = 1). Their goal was to see which bacterial lineages were present at these sites, how often resistance occurred, and whether genomic tools could give faster, more complete information to guide care. The team emphasized that faster access to resistance profiles and lineage information might support more effective, personalized treatment for people with EPTB.

The study combined phenotypic and genomic testing. Phenotypic drug susceptibility testing (pDST) against 12 drugs was performed using the MYCOTBI sensititre assay. Whole genome sequencing (WGS) was done on the Illumina XTen platform and analyzed with in-house pipelines. By pDST, drug resistance was identified in 23.9% (28/117) of isolates; WGS identified resistance in 29.9% (35/117), with 93.0% overall concordance between methods. WGS detected additional resistant cases that the MYCOTBI assay missed (p = 0.039). Resistance varied by sample type (p = 0.0446) and was highest in CSF, where 16 of 40 isolates (40.0%) were resistant. Looking at bacterial families, lineage 2 showed the highest level of resistance (13/24, 54.1%), and most lineage 2 resistant isolates were from CSF (9/13, 69.2%). The team also found mixed infections in 6.8% (8/117) of isolates, mostly involving lineages 2 and 3, and observed that heteroresistance was more common in mixed infections (p = 0.0139).

The findings show that whole genome sequencing can reliably detect drug resistance while also revealing which M. tuberculosis lineages and mixed infections are present in extrapulmonary samples. In this collection, lineage 2 stood out as enriched and more often drug resistant, particularly in cerebrospinal fluid, suggesting certain clinical phenotypes of EPTB may carry greater risk of resistance. Because WGS picked up resistant cases that the MYCOTBI sensititre assay missed, genomic approaches can provide a more complete picture than phenotypic tests alone. The authors propose that the same WGS approach could be adapted into culture-free targeted sequencing workflows to get rapid resistance and lineage data directly from clinical samples. That combination — fast detection of resistance, lineage assignment, and recognition of mixed infections and heteroresistance — could enable more personalized regimens and, the authors suggest, improved outcomes for people with extrapulmonary tuberculosis.