OR06

Growing Mycobacterium tuberculosis complex (MTBC) in culture remains the initial step in most diagnostic techniques, including whole-genome sequencing (WGS), to obtain higher amounts of DNA at the expense of a delay in results due to the slow-growth of mycobacteria. Moreover, the effect of the bottleneck-size in the loss of bacterial diversity still remains unknown, particularly relevant for drug-resistance detection. To advance towards faster and more accurate genome-based diagnostics, the most logical step would be culture-free sequencing (cfWGS). Our group has been developing approaches to enrich sputum samples in MTBC DNA (Goig et al. 2020). The main objectives are: 1) to perform cfWGS in samples from different settings; 2) to compare the genetic diversity in sputum-culture paired samples; and 3) to determine the factors associated with differences in diversity. We sequenced and analysed 42 and 19 sputum-culture paired samples from Mozambique and Georgia. Drug resistance prediction from cfWGS and culture-sequencing was 100% concordant. For samples with >90% of initial MTBC we found a median of 10 discrepant variants. However, for samples with <15% MTBC we found a median of 104 discrepant variants. In agreement, we estimated the bottleneck size imposed by culturing. Our results show a higher culturing bottleneck for bacterial diversity when the percentage of MTBC is low. We thus demonstrate that there is an influence of the initial amount of MTBC DNA in the bottleneck effect when culturing sputum samples that affects the loss of population diversity in culture and consequently the presence of unique variants in sputum.