Background

Direct sputum-based whole genome sequencing (WGS) can detect drug resistance TB in 30-40% of samples. However, compared to culture-based WGS or drug sensitivity testing (DST), the performance of sputum-based WGS has not been assessed properly.

FLASH-Mtb could help improve the accuracy of resistance-associated variant calling in samples, even in samples with a very low Mtb load. Furthermore, the FLASH method is flexible for modification in terms of type and number of targets of interest. Hence, the success of this pilot study could make the method ready for trialling in our collaborating hospitals in Ha Noi and Ho Chi Minh City.

Objectives

• To demonstrate that in direct sputum samples from TB patients, targeted sequencing FLASH-Mtb can outperform WGS in terms of detection rate and accuracy in diagnosing drug susceptibility.
• To demonstrate that FLASH-Mtb of sputum can predict drug susceptibility as accurately as WGS and DST of their cultures but at a much earlier time point.
• To demonstrate that in sputum, FLASH-Mtb can outperform the commercial Deeplex assay in terms of detection rate and cost in diagnosing drug susceptibility.

Methods

Pre-treatment sputum samples from PTB patients will be used. The Mtb bacterial load will be identified by microscopy and quantified by 16s rRNA molecular bacterial load assay (MBLA). DNA from at least 1ml sputum will be extracted and sequenced by WGS and the two targeted sequencing methods, Deeplex and FLASH-Mtb. Mtb from sputum will also be cultured and positive cultures will be used to perform phenotypic DST and DNA culture WGS.