TL;DR
This study uses whole-genome sequencing and molecular dating to reconstruct the history of drug resistance in Mycobacterium tuberculosis (TB) across the globe. It reveals a consistent order of resistance acquisition and highlights that while transmission is a major driver of drug-resistant TB, the age of resistance acquisition varies significantly by country, correlating with wealth and suggesting a role for public health interventions.
What Problem Does This Paper Address?
Tuberculosis (TB) remains a significant global health challenge, exacerbated by the emergence and spread of drug-resistant strains (MDR-TB and XDR-TB). While we know that drug resistance varies geographically, the specific local factors driving its evolution and persistence have been poorly understood. This paper aims to fill this gap by investigating when and where drug resistance emerged and spread across diverse TB populations worldwide, providing insights into the dynamics of resistance acquisition and transmission.
Methods
The researchers employed a multi-pronged approach:
- Data Curation: They compiled a large dataset of over 10,000 M. tuberculosis whole-genome sequences (WGS) with associated drug susceptibility testing (DST) data from numerous countries.
- Phylogenetic Analysis: Using WGS data, they constructed phylogenetic trees to understand the evolutionary relationships between different TB isolates.
- Bayesian Molecular Dating: This is the core of their temporal analysis. They used Bayesian methods to estimate the "Most Recent Common Susceptible Ancestor" (MRSCA) age for resistance to various drugs. Essentially, this tells them how long ago a particular drug resistance mutation likely emerged in the TB lineage.
- Correlation Analysis: They investigated the relationship between the age of resistance acquisition and socio-economic factors, specifically using Gross Domestic Product (GDP) per capita as a proxy for TB control program funding and resources.
- Diagnostic Sensitivity Assessment: They compared the performance of commercial diagnostic tests for second-line drug resistance across different countries.
Datasets
The study utilized a dataset of 10,299 M. tuberculosis clinical isolates with WGS and culture-based DST data. These isolates were sourced from over 20 countries, with 6,099 meeting the criteria for molecular dating analysis.
Key Insights and Ideas
- A Consistent "Resistance Ladder": The study found a remarkably consistent order in which TB strains acquire resistance globally: isoniazid resistance typically emerges first, followed by rifampicin and streptomycin, and then other drugs. This suggests a predictable evolutionary pathway for resistance.
- Old Resistance, High Fitness: The presence of "old" resistance (dating back over 20 years) in many countries indicates that these resistant strains are not only surviving but are also fit for transmission and persistence in human hosts.
- Wealth Matters for Resistance Age: A significant finding is the correlation between country wealth (GDP per capita) and the age of resistance. Wealthier nations tend to have older resistance, suggesting that better healthcare infrastructure and TB control programs might slow down the emergence of new resistance, even if older strains persist.
- Diagnostic Disparities: Commercial diagnostic tests show significant variability in their sensitivity for detecting second-line drug resistance across different countries, highlighting a need for context-specific diagnostic strategies.
Key Results
- Molecular Dating: They estimated MRSCA dates for 4,869 instances of resistance to 10 different drugs.
- Transmission Burden: The analysis estimated that 14-52% of resistance could be attributed to ongoing transmission, with INH and RIF resistance being the most frequently transmitted.
- Order of Resistance: The median MRSCA for INH resistance was 11.4 years prior to sample collection, significantly older than for RIF (7.6 years). EMB resistance followed RIF, and then PZA, ETH, FLQ, SLIs, or CYS.
- Recent Amplification: Resistance to fluoroquinolones (FLQ) and second-line injectables (SLIs) among MDR isolates was estimated to have occurred very recently, with a median ancestral age of only 4.7 years.
- GDP Correlation: A direct correlation was found between country wealth and resistance age (R²=0.47, P=0.014), with wealthier nations exhibiting older resistance.
- Diagnostic Sensitivity: Commercial diagnostic sensitivity for second-line resistance varied significantly by country (P<0.0003).
Limitations and Future Work
- Distinguishing Importation vs. Acquisition: The study acknowledges that its approach cannot definitively distinguish between resistance imported through human migration and new resistance acquired within a country. This means their estimates of transmission burden might be underestimates.
- Oversampling of DR Isolates: The study's sample may have overrepresented drug-resistant isolates compared to the general TB population, as indicated by comparisons with WHO MDR rates.
- Phylogenetic and Molecular Clock Assumptions: The accuracy of molecular dating relies on accurate phylogenetic tree estimation and molecular clock assumptions, which, while rigorously applied, are inherent to the method.
- DST Accuracy: The analysis assumes the accuracy of culture-based phenotypic DST, acknowledging that test-to-test variability can exist.
Future work could focus on refining transmission estimates, developing more geographically tailored diagnostic tools, and further investigating the factors that contribute to the fitness and longevity of resistant TB strains. The authors also emphasize the importance of sharing systematically collected isolate data to enable future reassessments of these patterns.
Summary updated