Pyrazinamide (PZA) is a key first-line antituberculosis drug that plays an important role in eradicating persister Mycobacterium tuberculosis (TB) bacilli and shortening the duration of tuberculosis treatment. However, PZA-resistance is on the rise, particularly among persons with multidrug-resistant (MDR) tuberculosis. This nationwide study was conducted to explore the prevalence of mutations conferring PZA resistance, catalogue mutation diversity, investigate the associations of PZA resistance with specific lineages, examine co-resistance to 13 first- and second-line drugs, and evaluate the diagnostic accuracy of sequencing pncA and panD genes for predicting PZA resistance. Whole genome sequencing was performed on 2,207 M. tuberculosis isolates from 25 States and 4 Union Territories of India. The majority of phenotypically PZA-resistant isolates (77%) harbored 171 distinct mutations in pncA; however, a small number of mutations in panD, rpsA and clpC1 were also observed. A set of novel mutations associated PZA resistance was uncovered, along with an additional 143 PZA resistance-conferring mutations in pncA based on application of WHO-endorsed grading rules. PZA resistance was predominately observed in Lineage 2 and eight lineage-specific resistance markers were identified. Mutations distributed across pncA correlate to 94% of PZA resistance and were the predominant drivers of phenotypic resistance; evidence generated herein substantiates sequencing the entire gene and promoter for comprehensive genotypic-based prediction of PZA resistance. This work provides key insights into the scope of PZA-resistance in India, a high drug-resistant TB burden country, and can support the effectiveness of TB prevention and control efforts.

India has the largest number of multidrug-resistant TB | (MDR-TB) cases, defined as Mycobacterium tuberculosis | resistant to at least isoniazid (INH) and rifampicin (RIF).1 | However, less than half of all persons with MDR-TB in | India successfully complete treatment.1 Although initial | end-of-treatment outcomes offer a standardised time point | to assess the effect of treatment, these tend to | underestimate the overall burden of unfavourable longterm outcomes among persons treated for TB.2,3 The longterm outcomes of persons diagnosed with MDR-TB in | India, including the proportion with recurrent TB disease | or mortality, are unknown. This analysis was conducted | under programmatic conditions in a high-burden setting, | with no regular check-ups after treatment. The results can | be used to show the burden of recurrent illness and death | following treatment, and can be used as a benchmark to | measure improvement.