OR28
Drivers of levofloxacin resistance in a high MDR-TB incidence country
O E Solomon(1,7,8) V N Nguyen(2) B N Cam(3) T A Nguyen(3,4) B Marais(4) S Graham(5,9) D Menzies(1,7) G Marks(6) G Fox(3,4) M A Behr(1,7,8)
1:McGill University; 2:National Lung Hospital, Ha Noi, Vietnam; 3:Woolcock Institute; 4:University of Sydney; 5:University of Melbourne; 6:University of New South Wales; 7:McGill International TB Centre; 8:Research Institute, McGill University Health Centre; 9:Burnet Institute
The VQUIN trial evaluated the safety and efficacy of levofloxacin-based tuberculosis preventative treatment (TPT) in household contacts of patients with RR/MDR-TB in Vietnam. Per trial protocol, contacts of Xpert-positive RR/MDR-TB index cases were screened for TB, prior to offering TPT. Individuals with culture-confirmed TB (co-prevalent cases) were excluded while Tuberculin Skin Test (TST) positive individuals were randomized to placebo or levofloxacin arm and followed for 24 months for culture conversion (incident cases). We investigated whether primary or acquired resistance is driving levofloxacin-resistance amongst MDR-TB strains. Isolates were whole genome sequenced and reads mapped to the H37Rv reference genome using Snippy for variant calling. TB-profiler was used for antibiotic resistance prediction and lineage annotation. A variant threshold was set at 5 as an indicator of transmission in 32 household pairs. We found no cases of acquired resistance amongst levofloxacin naïve strains or in isolates from patients who completed a 6-month levofloxacin-based TPT. Based on the absence of de novo resistance in household contacts, we set out to see the geographic and phylogenetic distribution of levofloxacin-resistant strains in Vietnam. Amongst 115 isolates from lab-confirmed RR/MDR-TB patients sampled across Vietnam, 19.1% (n=22) had levofloxacin resistance mutations. Although these strains were primarily found in sublineage 2.2.1 (21/22), genotypes were not phylogenetically or geographically linked. Further, different mutations were observed pointing to independent resistance acquisition events. In conclusion, we found primary resistance to drive levofloxacin-resistance amongst households in the VQUIN trial cohort and acquired resistance to be prominent in levofloxacin-resistant strains in the community.