Optimizing macrolide resistance detection for Mycobacterium abscessus: a potential low-cost, time-saving alternative

Mycobacterium abscessus complex (MABC) is notoriously difficult to treat. Current guidelines suggest a 14-day-long incubation and/or sequencing of erm41 to detect inducible macrolide resistance. We assessed whether the evolution of minimum inhibitory concentrations (MICs) can reliably predict inducible macrolide resistance and clinical outcomes of extrapulmonary MABC infections. Between 2013 and 2015, M. abscessus isolates were identified from extrapulmonary sites at a medical center. Drug susceptibility testing (DST) was conducted with results read on D3, D7, and D14. A fourfold or more increase in clarithromycin MIC between D3 and D7 was used to predict inducible macrolide resistance. The results were compared with the genotypic prediction of DST using sequencing of erm41 and rrl. Clinical data were reviewed. Thirty-five unique M. abscessus isolates comprising 16 subsp. abscessus, 18 subsp. massiliense, and 1 subsp. bolletii were identified. The overall inducible macrolide resistance was 40%. Using the proposed prediction rule, the sensitivity and specificity were 81% and 100%, while using genotypic prediction, the sensitivity and specificity were 75% and 100%, respectively. From the chart review, 28 patients received treatment, among whom 18 had evaluable outcomes. Most (92.9%) of the patients received macrolide-based combinatorial treatment; 71.4% of them received surgery. Of the outcome-evaluable patients, 77.8% had favorable outcomes. Our preliminary study highlighted that susceptibility testing read within one week holds promising potential for detecting inducible macrolide resistance. Further validation with larger cohorts of M. abscessus pulmonary and extrapulmonary disease is warranted.IMPORTANCEDrug susceptibility testing for difficult-to-treat microorganisms like Mycobacterium abscessus is frequently unavailable due to costs and technical demands. We propose a potential low-cost, time-saving alternative, using interval minimum inhibitory concentration evolution to predict the final phenotypic results. Our preliminary findings suggest comparable performance of the proposed early prediction method to the gold standard and the genotypic prediction in our small, extrapulmonary M. abscessus complex collections. Given its promising potential, validation in a larger cohort with pulmonary disease is needed for wider clinical application.