P29

Validation and implementation of thin-layer agar for direct Mycobacterium tuberculosis testing for bedaquiline resistance: a promising technique to increase access in low-resource settings.

I Cuella Martin(1) D Runyambo(2) B E Niyigena(2) S De Bock(1) M Diels(1) E Ardizzoni(1) B C de Jong(1) J C.S Ngabonziza(1,3,4) L Rigouts(1,5)

1:Unit of Mycobacteriology, Institute of Tropical Medicine, Antwerp, Belgium; 2:National Reference Laboratory Division, Department of Biomedical Services, Rwanda Biomedical Center, Kigali, Rwanda; 3:Research Innovation and Data Science Division, Rwanda Biomedical Centre, Kigali, Rwanda; 4:Department of Clinical Biology, University of Rwanda, Kigali, Rwanda; 5:Department of Biomedical Sciences, Antwerp University, Antwerp, Belgium

Since 2018, WHO recommends bedaquiline (BDQ) for treatment of rifampicin-resistant tuberculosis (RR-TB), yet the correlation between mutations and phenotypic resistance is incompletely understood, hindering molecular testing. Moreover, the great variety in level of BDQ resistance mutations seen complicates design of a rapid molecular assay. Thus, phenotypic tests (pDST) remain indispensable, but require BSL3 laboratories and have a long turnaround time, so few patients who start BDQ access such tests.

We validated thin-layer agar (TLA), for determining the minimal inhibitory concentration (MIC) of BDQ in isolates (indirect DST, 0.008-2.0 μg/ml), as well as its use for primary isolation of M.tuberculosis with direct DST to BDQ, rifampicin, isoniazid and levofloxacin at the National Reference Laboratory in Rwanda.

Results of indirect DST on 21 clinical isolates and 10 in-vitro resistant strains, compared with parallel 7H11 standard testing, showed a higher accuracy when interpreted on day 7 vs day 14 (94.0% vs 84.7%), and with plate incubation in standard incubator compared to incubation at 5-10% C02 (96.4% vs 81.9%). When used as direct method on 104 RR-TB decontaminated fresh samples, primary isolation was successful in 47 patients vs 57 in solid culture. However, only 3 (2.9%) TLA plates vs 7 (6.7%) solid culture were contaminated. DST results for RIF and IHN were 75% concordant with solid culture,yet TLA had a quicker turnaround time: median 19 days (IQR,14-21). On 41 patients with valid BDQ direct DST results, all were BDQ susceptible. These results confirm that TLA is a fast and reliable technique for the pDST of BDQ.