Diabetes-associated Phenotypic Masking of Fluoroquinolone Resistance in Multidrug-resistant Tuberculosis: Evidence from Genotypic-Phenotypic Discordance in Surabaya, Indonesia.

Abstract

Background: Emerging fluoroquinolone (FQ) resistance threatens multidrug-resistant tuberculosis (MDR-TB) management, complicating access to shorter regimens. This study characterizes FQ-resistant TB (FQ-R-TB) patients in Surabaya, Indonesia, and assesses diagnostic performance and therapeutic outcomes associated with this resistance.

Methods: A retrospective cohort analysis was conducted using national TB data from 1037 MDR-TB patients (2020-2025). Among them, 32 cases with genotypic FQ resistance (Xpert MTB/XDR) were stratified into concordant (genotypic mutation with phenotypic resistance, n = 21) and discordant (genotypic mutation with phenotypic susceptibility, n = 11) groups. Demographics, comorbidities, diagnostics, and treatment outcomes were evaluated.

Results: The proportion of FQ resistance among MDR-TB patients rose from 0% in 2020 to 1.37% in 2025. Discordant cases were significantly older (mean: 54 ± 14; P = 0.203), predominantly male (10:1; P = 0.037), had higher body mass index (BMI; 24.44 ± 6.04; P = 0.197), and had a markedly higher prevalence of diabetes mellitus (90.9% versus 9.5%; P = 0.028). Diagnostic discordance occurred in 34.4% of FQ-R-TB cases, causing 1-3 month delays in regimen adjustment and initial ineffective FQ exposure. While subsequent FQ-free BPaL regimens improved outcomes, initial diagnostic uncertainty complicated decisions and increased adverse effects.

Conclusion: FQ resistance is an emerging threat in Surabaya. Diagnostic discordance and high comorbidity burden complicate treatment. Advanced age and higher BMI are linked to increased diabetes probability. Diabetes is a powerful predictor of genotypic-phenotypic discordance, likely via host-mediated mechanisms affecting bacterial fitness and resistance expression. A strong clinical consensus is needed to recommend genotypic test results as the primary guide for treating TB patients with diabetes. However, for diabetic TB patients whose initial phenotypic results indicate FQ susceptibility, close monitoring of their clinical and microbiological response during therapy remains crucial to confirm effectiveness until more definitive evidence is available.