Rapid detection of non-rifampicin drug-resistant tuberculosis using GeneXpert MTB/XDR testing: findings from a multi-site drug-resistant tuberculosis case detection surge in Zambia

BackgroundZambia continues to face a high burden of tuberculosis with rising drug resistance, yet diagnosis of non-rifampicin resistance remains limited due to reliance on rifampicin as a proxy for multidrug-resistant tuberculosis. The Xpert MTB/XDR assay offers rapid detection of resistance to isoniazid, fluoroquinolone, Ethionamide and aminoglycosides, addressing a major diagnostic gap. This study assessed the feasibility and utility of reflex MTB/XDR testing using data from a drug-resistant tuberculosis active case finding surge. MethodsWe conducted an analysis of programmatic data from a four-week active case finding surge in May 2025, implemented across 39 high-TB burden facilities in five provinces of Zambia. We evaluated bacteriologically confirmed (Xpert MTB/RIF Ultra) tuberculosis patients using MTB/XDR regardless of Rifampicin resistance. We analyzed patient demographics, characteristics, and drug-resistance patterns descriptively. ResultsOf 509 TB patients identified, 70.9%(n=361) were bacteriologically confirmed by Xpert MTB/RIF Ultra among whom 92.8%(n=335) were eligible for reflex Xpert MTB/XDR testing. Of the eligible, 89.3% (n=299) underwent reflex MTB/XDR testing. Results were successfully retrieved for 282 (94.3%). Overall, 8.5%(n=24) of patients with XDR results had a form of drug-resistant tuberculosis. Mono-resistance to rifampicin (3.5%, n=10) or isoniazid (3.2%, n=9) was more prevalent than multi-drug-resistant tuberculosis (1.1%, n=3), while only 0.4% (n=1) had Pre-XDR TB. Non-rifampicin resistance, including isoniazid mono-resistance and isoniazid combined with fluoroquinolone or ethionamide resistance, were detected in 3.2% (n=9) of all patients with XDR TB results (37.5% of all DR-TB cases). ConclusionReflex Xpert MTB/XDR testing enabled rapid identification of non-rifampicin-resistant TB cases that would have been missed by rifampicin-centric diagnostics. Strengthening the integration of MTB/XDR into routine diagnostic algorithms can improve early drug-resistant tuberculosis detection, guide individualized treatment, and prevent amplification of resistance in high-burden settings. National scale-up, accompanied by laboratory strengthening, capacity building, and robust data systems, are essential to optimizing its use and advancing drug-resistant tuberculosis elimination.