New test shows promise for tuberculosis detection on sputum and tongue swabs

Tuberculosis remains a global health challenge, and recent figures show that despite better tools, only 48% of newly diagnosed TB cases were confirmed using nucleic acid amplification tests (NAATs) in 2023. That diagnostic gap motivated a clinical evaluation led by Anura David of the BD MAX TM MDR-TB assay, a moderate complexity NAAT that identifies Mycobacterium tuberculosis complex (MTBC) and screens for resistance to rifampicin (RIF) and isoniazid (INH). The study took place in South Africa and set out to answer practical questions clinicians face: can the test be used on different kinds of sputum preparations, and could a less invasive sample — a tongue swab (TS) — also work? To do this, the team compared results from raw sputum, NALC/NaOH decontaminated sputum (a common processing step that produces a sputum pellet), and TS specimens under different transport and processing conditions. Liquid culture was used as the reference standard, so the study measured how often the BD MAX TM MDR-TB assay matched a culture diagnosis of TB. The work addresses both how well the test detects MTBC and how reliably it finds resistance to core TB drugs.

The evaluation focused on two key outcomes: MTBC detection and profiling for RIF and INH resistance using the BD MAX TM MDR-TB (MAX MDR-TB) assay, comparing results against liquid culture. For sputum, the assay showed similar sensitivity whether run on the sputum pellet (87%) or on raw sputum (89%), with raw sputum identifying one additional true positive case. Resistance detection had some limitations: two false RIF-resistant results occurred, and INH resistance was missed in two cases. Tongue swab (TS) specimens were tested under different transport and processing conditions and, while the number of TS samples was small, diagnostic accuracy appeared better when a diluted (66%) STR buffer was used. Across specimen types, a substantial portion of positives were reported at low levels: 15 out of 55, or 27%, were classified as "MTB Low POS." These detailed findings show where the MAX MDR-TB assay performs well for MTBC detection and where further work is needed on resistance calls and specimen handling.

The study’s results have practical implications for how the BD MAX TM MDR-TB assay might be used in clinics and laboratories. That the assay performed comparably on sputum pellet and raw sputum suggests laboratories could use either preparation without large losses in sensitivity, which could simplify workflows and reduce the need for extra processing steps. The promise shown by tongue swabs, especially when processed with a diluted (66%) STR buffer, points to a less invasive sampling option that would be easier to collect from patients, but the small sample size means this finding needs confirmation in larger studies. The observation of two false RIF-resistant results and two missed INH resistances highlights risks of incorrect resistance profiling, and the high rate of "MTB Low POS" calls (27%) suggests many results hover near the assay’s detection limit, potentially leading to repeat testing. Taken together, the work led by Anura David indicates the MAX MDR-TB assay could be a useful tool in diagnostic algorithms, but that assay optimisation, validation on larger TS cohorts, and careful integration are required to improve reliability and patient outcomes.