New culture-free sequencing speeds TB resistance and transmission detection

Drug-resistant tuberculosis (DR-TB) is hard to treat in part because clinicians lack fast tests that predict resistance to all the drugs used in modern regimens. To tackle that gap, a team led by corresponding author Andrea Maurizio Cabibbe evaluated next-generation sequencing approaches that work without the slow step of growing Mycobacterium tuberculosis in culture. They tested two technologies on 96 MTBC–positive decontaminated sputum samples collected during a vaccine trial: a targeted culture-free approach, Deeplex Myc-TB XL tNGS (Genoscreen, beta-testing), and an enrichment-based direct whole-genome approach, QIAseq xHYB MTB (Qiagen). DNA was prepared using a host-depletion protocol, quantified with an MTBC-specific real-time PCR, and sequenced on Illumina platforms. The study compared these culture-free tests to culture-based whole-genome sequencing (cWGS) and used tools such as the WHO mutation catalogue and Ridom SeqSphere+ to analyse resistance, lineage, and transmission. Samples covered a wide range of bacterial amounts, from <10 to >1,000 genome copy (gc)/µL, letting the researchers assess how bacillary load affects results.

The team measured analytical sensitivity, resistance prediction accuracy, and genotyping concordance. Deeplex Myc-TB XL tNGS produced interpretable resistance profiles in 96.6% of specimens, with a limit of detection (LoD) of about 10 gc and 100% sensitivity for all evaluated drugs; specificity was 100% for most drugs and ≥97% for isoniazid/ethionamide. QIAseq xHYB MTB dWGS provided data suitable for drug resistance (DR) analysis in 75% of samples and had a LoD near 100 gc. dWGS showed 100% sensitivity for most drugs, though one fluoroquinolone-resistant case was missed because the variant was below low-frequency thresholds; resistance to delamanid and clofazimine was misclassified in one case each due to interpretation rules. Specificity was 100% for most targets and ≥97% for rifampicin. Lineage assignment from both culture-free approaches matched cWGS, and dWGS allowed transmission analysis in 65% of samples, confirming the cluster seen by cWGS. Analyses linked genome copy number to sequencing coverage and used assay-specific pipelines for interpretation.

These results show complementary strengths: tNGS offers very sensitive and specific targeted drug resistance profiling at a LoD comparable to the most sensitive rapid tests, while dWGS, which needs higher DNA input, enables genome-wide analysis without culture, including transmission inference and searching for candidate DR loci. In practice, a bacillary load–guided strategy that uses tNGS when genome copy numbers are low and dWGS when they are higher could give clinicians reliable resistance results quickly and support public health genomic surveillance. The study demonstrates that culture-free sequencing can be applied directly to respiratory samples to inform precision diagnosis, surveillance, and trial applications, accelerating decisions about first-, second-line, newer, and repurposed drugs. Using tools such as the WHO mutation catalogue and Ridom SeqSphere+ to interpret results helps ensure genotyping and resistance calls align with existing standards, while preserving the ability to detect clusters and inform outbreak responses.