Rapid assays for fluoroquinolone resistance in Mycobacterium tuberculosis: a systematic review and meta-analysis

Multidrug-resistant tuberculosis has emerged as a global health threat. Given poor treatment outcomes of fluoroquinolone-resistant multidrug-resistant tuberculosis, there is a pressing need for rapid drug susceptibility testing of multidrug-resistant Mycobacterium tuberculosis against fluoroquinolones. This review aims at evaluating these rapid assays. PubMed and OvidSP were used to search MEDLINE and EMBASE for publications in English regarding rapid assays that tested ofloxacin, levofloxacin or moxifloxacin. Studies were included only in the concurrent presence of sensitivity and specificity data. Summary estimates of sensitivity and specificity were generated by the bivariate random effects model when there were at least three sets of data under the same assay category that tested the same fluoroquinolone with reference to a standard test. Of 108 articles identified, 24 articles were included in a meta-analysis of rapid assays that tested ofloxacin in culture isolates. Overall, rapid genotypic assays targeting gyrA only are significantly less specific (96% versus 99%) and non-significantly less sensitive (88% versus 94%) than rapid phenotypic assays. To test for the presence or absence of ofloxacin resistance to a certainty threshold of 90%, the required pre-test prevalence ranges of ofloxacin resistance for genotypic assays targeting gyrA only are 29%-47% overall, 36%-55% for PCR-DNA sequencing and 23%-44% for others. Corresponding ranges are 7%-65% for phenotypic assays overall and 3%-75% for Mycobacteria Growth Indicator Tube (MGIT). Assuming that the mean pre-test prevalence of fluoroquinolone resistance in culture isolates of multidrug-resistant M. tuberculosis is similar to 20%, rapid genotypic assays other than PCR-DNA sequencing, targeting gyrA only, can reliably screen for ofloxacin resistance.