'Geographical and temporal analysis of M. tb whole genome sequences' is now published in Lancet Microbe!

Its improved thanks to feedback from reviewers and tweeter-scientists.

Below is a thread to explain the studies main methods and findings. 1/12 https://www.thelancet.com/journals/lanmic/article/PIIS2666-5247(20)30195-6/fulltext

Among 8550 M.tb curated isolates with geographic tags, resistance phenotype and WGS, we performed molecular dating assuming a relaxed molecular clock rate with mean 0.5 SNPs/year (results held varying this between 0.3-0.6 in sensitivity analysis) 2/12

We dated 4869 resistance instances in more than 15 countries, grouping isolates into lineage-drug-country based phylogenies. (this took a lot of compute time!) 3/12

Looking across all lineages and countries we find multi-drug resistance to be acquired in consistent order in M.tb with the oldest being Isoniazid > Rifampicin and/or Streptomycin > Ethambutol > PZA and other drugs. (note we had no data on novel TB agents) 4/12

A novel finding was that resistance to streptomycin, the first TB agent -used as mono-therapy in the 1940s-early 50s, was dated after Isoniazid resistance (introduced ~1952), and around rifampicin resistance (introduced ~1965). 5/12

The best explanation of this is that a substantial proportion of Strep resistance was acquired during the now defunct Category II therapy i.e. a regimen when Streptomycin is added to a failing TB regimen. 6/12

Now comparing resistance ages across countries, we found a consistent signal of older resistance age in wealthier countries. This is not just related to TB importation in western countries but also observed in China Russia and other countries with lower disease importation. 7/12

Assuming that wealthier countries invest more in TB care, treatment monitoring, and diagnostics etc.. this wealth correlation with resistance age suggests that increased investment may prevent resistance acquisition in TB. To support this further we ... 8/12

.. focused on MDR to XDR/preXDR resistance amplification, i.e. to fluoroquinolones and second line injectables, we found a substantial proportion of MDR isolates amplified resistance to FLQ in the past 5 years from sample collection (MRCA median was 4.7 years) 9/12

So clearly resistance amplification is ongoing in many parts of the world. We finally looked at the expected performance of molecular resistance diagnostics across geographies and we found 10% of the important resistance determinants to have very high geographic variance. 10/12

We computed the expected sensitivity of molecular diagnostics across the five countries for which we had the most isolate data, and found molecular sensitivity to vary substantially in comparison with phenotypic tests- snap shot of table with data from 3 countries below. 11/12

A lot packed into this paper! We also discuss resistance transmission and fitness in the manuscript. If there any questions, please let us know. 12/12