Methods: We included TB patients during 2006–2013, whose specimens tested RIF susceptible phenotypically but had a missense mutation in rpoB determined by pyrosequencing. Demographic data were abstracted from the California TB registry. Clinical information including medical history, treatment and outcomes were abstracted from medical records. Treatment outcomes defined by World Health Organization were used.
Results: Of 3011 specimens tested, 368 specimens had a rpoB mutation (12.2%). Of these, 7 different mutations were identified in 16 specimens (4.3%) with phenotypic-molecular discordant RIF results: 511Pro (n = 3), 516Phe (n = 1), 526Asn (n = 6), 526Ser [AGC (n = 2) and TCC (n = 1)], 526Cys (n = 1), and 533Pro (n = 2). Fourteen (88%) were resistant to isoniazid (INH), of which 6 were also resistant to ethambutol (EMB) and/or pyrazinamide (PZA). Four case-patients (25%), 1 with 511Pro (treated with RIF, PZA and EMB) and 3 with 526Asn (1 treated with RIF, EMB and fluoroquinolone (FQ) and 2 treated with RIF, PZA and EMB) relapsed or failed. Upon retreatment with an expanded regimen, 3 (75%) had good outcomes and 1 moved prior to treatment completion. The remaining 12 case-patients either completed treatment or were cured; their regimens included a FQ for 10 patients, of which 6 had at least 1 other second line medication.
Conclusion: IF phenotypic-genotypic discordance was rare and most isolates also had INH resistance. This case series suggests an association of poor outcomes among patients not treated with expanded regimen and rpoB mutations not conferring phenotypic resistance. These mutations may have clinical importance and expanded treatment regimen should be considered. Small numbers, and variable drug resistance profiles and treatment regimens in our cohort are limitations. Additional studies of these rpoB mutations are needed to confirm their role in poor treatment outcomes.
P. Barry, None
G. Schecter, None
E. Desmond, None