Back

Association between the rifampicin resistance mutations and rifabutin susceptibility in Mycobacterium tuberculosis: a meta-analysis

Wang, W.; Zhou, H.; Cai, L.; Yang, T.

2023-07-14 microbiology
10.1101/2023.07.13.548878 bioRxiv
Show abstract

Some rifampicin-resistant Mycobacterium tuberculosis (MTB) strains were susceptible to rifabutin (RFB) and may be amenable to treatment with RFB. We performed a meta-analysis of available cross-sectional studies to determine which RIF-resistance mutations were associated with rifabutin susceptibility. We identified studies through PubMed, Web of Science, Embase, and Cochrane Library up to June 1, 2023. Studies that met our criteria were those that investigated rpoB mutations and reported phenotypic drug susceptibility for RIF and RFB. The relationship between RIF-resistance mutations to RFB-susceptibility was evaluated using odds ratio (OR). Twenty-five studies comprised 4,333 clinical RIF-resistant MTB isolates from 21 different countries met our criteria for inclusion. Of these isolates, 21.00% (910/4333) were susceptible to RFB. We found seven RIF-resistance mutations were high confidence (OR>10) in predicting RFB-susceptibility, which were D435V, D435Y, D435F, H445L, L430R, S441L, and S441Q. Among strains carrying these mutations, 83.01% (435/524) were susceptible to RFB. The minimum inhibition concentrations (MICs) of these strains revealed that they had low MIC (D435V, D435F, H445L, and D435Y) or were susceptible (S441L) for RFB and exhibited a significant lack of correlation between MICs to RIF and RFB. Mutations such as H445C, H445G, H445N, L430P, and L452P showed a moderate confidence (5<OR[&le;]10) in prediction of RFB-susceptibility. Of these mutants, 62.16% (69/111) were susceptible to RFB. The most common RIF-resistance mutations S450L, as well as S450W, were associated with RFB-resistance (OR<1). These results provide a theoretical basis for molecular detection of RFB-susceptible TB and alternative treatment with RFB in MDR/RR-TB patients.

Matching journals

The top 1 journal accounts for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.