Background: Lack of rapid and reliable susceptibility testing for second-line drugs used in the treatment of multidrug-resistant tuberculosis (MDR-TB) may limit treatment success.

Methods: Mycobacterium tuberculosis isolates from patients referred to Kibong’oto National TB Hospital in Tanzania for second-line TB treatment underwent confirmatory speciation and susceptibility testing. Minimum inhibitory concentration (MIC) testing on MYCOTB Sensititre plates was performed for all drugs available in the second-line formulary. We chose to categorize isolates as borderline susceptible if the MIC was at or one dilution lower than the resistance breakpoint. M. tuberculosis DNA was sequenced for resistance mutations in rpoB (rifampin), inhA (isoniazid, ethionamide), katG (isoniazid), embB (ethambutol), gyrA (fluoroquinolones), rrs (amikacin, kanamycin, capreomycin), eis (kanamycin) and pncA (pyrazinamide).

Results: Of 22 isolates from patients referred for second-line TB treatment, 13 (59%) were MDR-TB and the remainder had other resistance patterns. MIC testing identified 3 (14%) isolates resistant to ethionamide and another 8 (36%) with borderline susceptibility. No isolate had ofloxacin resistance, but 10 (45%) were borderline susceptible. Amikacin was fully susceptible in 15 (68%) compared to only 11 (50%) for kanamycin. Resistance mutations were absent in gyrA, rrs or eis for all 13 isolates available for sequencing, but pncA mutation resultant in amino acid change or stop codon was present in 6 (46%). Ten (77%) of MDR-TB patients had at least one medication that could have logically been modified based on these results (median 2; maximum 4). The most common modifications were a change from ethioniamide to para-aminosalicylic acid, and the use of higher dose levofloxacin.

Conclusions: In Tanzania, quantitative second-line susceptibility testing could inform and alter MDR-TB management independent of drug-resistance mutations. Further operational studies are warranted.