P079

The emergence of drug resistance in M. tuberculosis, seems to follow an order. In most cases resistance to isoniazid (INH) emerges first. Prevention of emergence of INH resistance can potentially prevent the emergence of resistance to other drugs. Here we present the prognostic value of specific mutations in predicting emergence of INH resistance through canonical mutations (katG315, inhA-15, and inhA-8).

In our approach, we used genomic and phenotypic data from over 16,000 samples collected by two projects, the TB Portals and the CRyPTIC consortium. We evaluated the potential of prediction canonical INH resistance using each mutation and combination of two mutations associated with phenotypic resistance.

Our best performing model used a combination of two genomic markers with an estimated prognostic accuracy of 73%, sensitivity of 55%, specificity of 84%, PPV of 69%, and an NPV of 75% for correctly predicting the emergence of the three canonical INH resistance mutations. These results are high enough to be demonstrate the potential of the approach. The relatively lower sensitivity is partly a reflection of different evolutionary paths to the eventual emergence of the three canonical mutations. This exercise only evaluates the predictability of the evolutionary paths that involve the prognostic mutation(s) used for building the model. This work also presents evidence that resistance to INH follows a stepwise evolutionary trajectory which can be exploited for prediction and avoidance of INH resistance.

Additional time course samples and analysis will uncover prognostic markers for other evolutionary trajectories that lead to resistance.