A massive new analysis of Mycobacterium tuberculosis isolates from around the world provides an "unparalleled view" of the mechanisms behind drug-resistant tuberculosis (TB).
The analysis, described in two studies published last week in PLOS Biology, was conducted by scientists working with CRyPTIC (Comprehensive Resistance Prediction for Tuberculosis: an International Consortium), a large international effort employing whole-genome sequence (WGS) to create a comprehensive map of the genetic changes that are responsible for making TB drug-resistant.
Of the 10 million TB cases reported annually, the World Health Organization (WHO) estimates nearly 500,000 are resistant to the first-line TB drug rifampicin, and about three quarters of these are multidrug-resistant TB (MDR-TB).
Drug-resistant TB is treatable, and recently developed antibiotics have made the treatment regimen more effective and much shorter. What was once an 18- to 24-month regimen with roughly 55% effectiveness is now 6 months long and 85% effective. And molecular-based rapid diagnostic tests that look for genes to predict TB drug resistance have begun to replace slower, culture-bases drug susceptibility tests, shrinking the time to diagnosis and proper treatment to 2 hours.
But there are concerns that those tests look for only a limited number of resistance mechanisms, and that drug-resistant TB remains underdiagnosed and undertreated.
A member of CRyPTIC says he hopes this extensive global effort to uncover the genetic changes responsible for drug-resistant TB will ultimately lead to better molecular-based diagnostics and more targeted treatment.
"A sample from a patient can then be genetically sequenced and a list of antibiotics that the infection is susceptible to returned to the clinician," explained University of Oxford researcher and CRyPTIC member Philip Fowler, PhD, who was involved with both studies. "Several high-income countries are already using whole-genome sequencing for tuberculosis diagnostics in this way, and CRyPTIC will improve this yet further."
Studies enhance understanding of drug-resistant TB
The first study describes the collection and analysis of 15,211 M tuberculosis isolates from CRyPTIC partners in Asia, Europe, Africa, and South America. From this collection, 12,289 isolates belonging to the four main M tuberculosis lineages were sequenced and had susceptibility to 13 different antitubercular drugs measured to capture their genotypic and phenotypic profile.
Of the 12,289 isolates tested, 6,814 (55.4%) were resistant to at least one drug, including 4,685 that were resistant to the first-line drug rifampicin or were multidrug-resistant (RR/MDR). Of the RR/MDR isolates, 3% were extensively drug-resistant (XDR), and 38.8% were pre-XDR.
The drugs with the highest percentage of resistant isolates were the first-line drugs isoniazid (49%) and rifampicin (38.7%). Resistance was low for the newer and repurposed TB drugs, like bedaquiline (0.9%), clofazimine (4.4%), delaminid (1.6%), and linezolid (1.3%).
WGS results showed that the top mutations that were found among isolates were phenotypically resistant to first- and second-line drugs, as well as those that were resistant to newer and repurposed drugs.
"Through its sheer size and by oversampling for resistance, the compendium gives an unparalleled view of resistance and resistance patterns among the panel of 13 antitubercular compounds studied," the study authors wrote.
In the second study, CRyPTIC researchers analyzed the genomes of 10,228 M tuberculosis isolates from the collection to dig a little deeper into the many unknown resistance mechanisms for TB drugs. While many TB-resistance mutations are well documented, the authors note, very large sample sizes are needed to identify these previously uncatalogued resistance mechanisms.
But instead of looking at whether isolates were resistant or susceptible to the 13 drugs, they investigated the association between resistance mechanisms and the minimum inhibitory concentration (MIC) for each drug. MIC is a measurement of the lowest concentration needed of a drug to inhibit growth of bacteria. The purpose of using MIC, rather than "binary resistance phenotypes," was to find genetic mutations that may cause only subtle changes that reduce the effectiveness of a drug but could be overcome by using higher doses.
This approach enabled the team to uncover a host of novel resistance determinants associated with both resistance and subtle changes in MIC, including those for newer and repurposed TB drugs. They then selected the 20 most significant resistance genes identified for each of the 13 drugs, including catalogued and previously uncatalogued genes, for further analysis.
"Although there are several existing catalogues which relate mutations in specific genes to antibiotic resistance, these focus mainly on the first-line treatments for TB [isoniazid, pyrazinamide, rifampicin, and ethambutol]," Fowler said. "CRyPTIC has improved the catalogue of mutations that confer resistance to other treatments like fluoroquinolones, ethionamide, and clofazimine."
Fowler, whose research focuses on developing ways to predict antibiotic resistance, says TB was a good place to start because all resistance mechanisms occur in the chromosome and can be detected with short-read sequencing. Other bacterial pathogens, like the ESKAPE pathogens, will require long-read sequencing to sequence chromosomes and the plasmids that can carry resistance genes. But he hopes that the methods used for the CRyPTIC project can be used to predict antibiotic resistance in other pathogens.
"We also feel we've learnt a lot about how to coordinate a large international consortium like CRyPTIC and how to collect and accurately characterize thousands of clinical isolates," he said.