Mitigating unfavourable treatment outcomes and acquired rifampicin resistance in isoniazid-resistant tuberculosis: the role of fluoroquinolone

BACKGROUND: Rifampicin (RMP), ethambutol, and pyrazinamide (PZA) for 6-9 months were recommended for the management of isoniazid-resistant, rifampicin-susceptible tuberculosis (Hr-TB), but recommendations on fluoroquinolones (FQs) were inconsistent. We investigated treatment outcomes and acquired RMP resistance in Hr-TB compared to isoniazid-susceptible TB (Hs-TB). METHODS: We retrospectively enrolled TB patients notified from 2010 to 2018 in Taiwan. Logistic regression model was constructed to estimate the odds of favourable outcomes and acquired RMP resistance. Propensity score matching (PSM) was conducted to address selection bias. RESULTS: 6115 Hr-TB and 71,184 Hs-TB were included. 25.6 % of Hr-TB and 24.7 % of Hs-TB had unfavourable treatment outcomes (p = 0.149). 0.9 % of Hr-TB and 0.1 % of Hs-TB had acquired RMP resistance (p < 0.001). In Hr-TB treated with RMP and PZA throughout regimens and Hs-TB treated with RMP throughout regimens, unfavourable treatment outcomes (16.1 % vs 13.3 %, p < 0.001), and acquired RMP resistance (1.0 % vs 0.1 %, p < 0.001) was significantly higher in Hr-TB than that in Hs-TB. Among Hr-TB, treatment with FQs were significantly associated with favourable outcomes (adjOR: 3.18, 95 % CI: 2.45-4.15) and less acquired RMP resistance (adjOR: 0.16, 95 % CI: 0.05-0.55). FQs remain significantly associated with favourable outcomes (adjOR 3.44, 95 % CI 2.56-4.63) after PSM. Of the 747 Hr-TB patients treated with a FQ, one (0.13 %) had acquired FQ resistance. CONCLUSIONS: RMP and PZA throughout regimens did not completely remove the influence of isoniazid resistance. The use of FQs was associated with better treatment outcomes and a lower risk of acquired RMP resistance in Hr-TB but acquired FQ resistance may occur.