Abstract
Mixed infection refers to the presence of multiple Mycobacterium tuberculosis strains within one host, while heteroresistance denotes the coexistence of drug-susceptible and drug-resistant strains or genotypes. Mixed infections and heteroresistance with Mycobacterium tuberculosis can complicate drug resistance diagnosis, treatment options, and transmission inference. We conducted a population-based genomic epidemiological study of multidrug-resistant tuberculosis (MDR-TB) in Shanghai, China, between January 1, 2005, and December 31, 2018, to evaluate the prevalence and impact of mixed infection and heteroresistance on MDR-TB diagnosis and treatment outcomes. Demographic, clinical, and laboratory data were collected, and factors associated with mixed infections and heteroresistance were identified with multivariable logistic regression analysis. Among the 936 MDR-TB patients in our study, 10.8% (101/936) had mixed infections and 16.5% (154/936) exhibited heteroresistance, which was more frequent with second-line anti-TB drugs (P < 0.01). There was a higher risk of heteroresistance in older patients (≥60 years: aOR 1.91, 95% CI 1.02-3.57), patients with diabetes (2.59, 1.36-4.91), and mixed infections (2.85, 1.67-4.88). Mixed infections and heteroresistance accounted for 22.6% (58/257) of the strains with discrepancies between phenotypic and genotypic drug susceptibility testing (DST). Strains with heteroresistance to EMB had a higher discordance rate than those without (29.1% VS 17.2%, P < 0.05). Isolates that were phenotypically susceptible but genotypically resistant harboured minority or low-frequency resistance mutations and were more common in patients with mixed infections and heteroresistance. In summary, mixed infections are significantly associated with heteroresistance, and both mixed infections and heteroresistance can lead to discrepancies between phenotypic and genotypic DST.