Table 1.

Implications of understanding evolutionary mechanisms of drug resistance for tuberculosis control

Experimental observation	Physiological consequence	Implications
Resistant strains gain compensatory mutations that change the basic physiology, e.g. RIF-rpoC	Increased transmissibility, increased propensity to acquire additional resistance	Focus surveillance on mutations that correlate with transmissible, highly resistant strains Use spent biosamples to establish wide catchment area for WGS-based surveillance
Continued exposure to RIF directs evolution towards the acquisition of compensatory mutations	More rapid evolution of compensatory mutations	Use molecular tools to probe the genotype of strains and stop administering ineffective drugs immediately
Epistasis exists between drug-resistance mutations for a single drug: e.g. EMB, FQ, INH, RIF	Multi-step acquisition of high level of drug resistance	Define clinical breakpoint concentrations based on specific susceptibility profiles for a drug
Positive epistasis between rpoB and gyrA mutations	Strains with specific combinations of resistance mutations (e.g. gyrA D94G and rpoB H445Y) are fitter than the wild type	Drug regimens containing both RIF and FQ may fail more quickly. Assess current clinical trials for evidence Evaluate if the order in which drugs are administered might enhance negative epistatic interactions
Some mutations conferring resistance to FQ and bedaquiline have no fitness costs attached	No need to compensate for specific resistance mutations	Enforce appropriate administration of drugs to avoid adding these antibiotics to failing regimens or use as monotherapy Explore the spectrum of resistance mutations, identify low cost and frequent mutations before releasing the drug into the market
Mutation rates vary between M. tuberculosis lineages	Beijing family strains acquire resistance to INH and RIF at a higher rate and are 22 times more likely to develop into MDR than the laboratory-adapted strain	Increase frequency of phylogeny-based surveillance and focus monitoring on areas with high rates of Beijing strains

EMB ethambutol, FQ fluoroquinolones, INH isoniazid, RIF rifampicin, WGS whole genome sequencing