Genes and patients both shape TB outcomes in Georgia

Tuberculosis (TB) remains a major public health concern, and improving treatment success depends on understanding why some patients do poorly. To tackle this, Galo A. Goig and colleagues combined clinical records with DNA data from the Mycobacterium tuberculosis complex (MTBC) collected over 13 years in the country of Georgia. The study examined 4,536 TB patients to see how basic patient characteristics and the bacteria that cause TB interact to shape disease presentation and final treatment results. The researchers wanted to test whether factors we already know—like age, sex, body mass index (BMI) and other illnesses—explain most of the variation in outcomes, or whether the bacterial genome itself adds important information. By pairing large-scale bacterial genome sequencing with detailed clinical information and statistical models, the team sought to identify which demographic, clinical and microbial features predict unfavorable treatment outcomes. The goal was practical: to inform TB control programmes and improve treatment success rates by clarifying which patient and bacterial determinants matter most in real-world care.

The authors analyzed MTBC genomes alongside patient data from 4,536 cases and used multivariable modelling and GWAS analyses to search for links between bacterial genetics, clinical features and outcomes. Multivariable models confirmed that known demographic and clinical factors—sex, age, BMI and comorbidities—are important predictors of treatment outcomes. The analysis also supported the efficacy of novel TB treatments containing bedaquiline. On the bacterial side, specific variables were tied to worse outcomes: MTBC lineage, mutations conferring resistance to rifampicin and to fluoroquinolones, and a high bacterial burden were all associated with unfavorable treatment results. GWAS analyses did not identify any bacterial genetic mutations associated with outcomes beyond the established drug resistance-conferring mutations. However, the team did find that mutations in the bacterial gene sufD were linked to cavitary disease, and that mutations in sufD, mutations conferring resistance to rifampicin and fitness compensatory mutations were associated with the bacterial burden within patients.

These findings show that both who the patient is and the genetics of the infecting bacterium shape how TB presents and how well treatment works. For clinicians and public health teams, the results underline that demographic and clinical risk factors remain central to predicting outcomes, but that bacterial markers—especially resistance mutations to rifampicin and fluoroquinolones, MTBC lineage, and measures of bacterial burden—provide additional, actionable information. The lack of new bacterial predictors from GWAS beyond known resistance mutations suggests that routine surveillance of drug resistance and bacterial load remains a priority, while the association of sufD mutations with cavitary disease points to specific bacterial features that merit further study. Overall, integrating bacterial genomic data with patient information could sharpen risk prediction, guide use of bedaquiline-containing regimens, and help target more intensive care or follow-up to patients at higher risk of unfavorable outcomes.