INT J TUBERC LUNG DIS 19(2):185–190 Q 2015 The Union http://dx.doi.org/10.5588/ijtld.14.0651 Disputed rpoB mutations can frequently cause important rifampicin resistance among new tuberculosis patients A. Van Deun,* † K. J. M. Aung, ‡ Md. A. Hossain, ‡ P. de Rijk,* M. Gumusboga,* L. Rigouts,* § B. C. de Jong* ¶ *Mycobacteriology Unit, Institute of Tropical Medicine, Antwerp, Belgium; † International Union Against Tuberculosis and Lung Disease, Paris, France; ‡ Damien Foundation, Dhaka, Bangladesh; § Department of Biomedical Sciences, Antwerp, Belgium; ¶ New York University, New York, New York, USA SUMMARY SETTING: Greater Mymensingh area, Bangladesh. OBJECTIVES: To document among new tuberculosis (TB) patients the proportions and treatment outcomes of silent, non-disputed and disputed (generally missed by rapid drug susceptibility testing [DST]) rpoB mutations, and their detection by commercial molecular assays. DESIGN: Retrospective analysis of rpoB sequences from randomly selected ethanol-preserved diagnostic sputum samples; comparison of sequencing with conventional DST results and standard first-line treatment outcome; retesting of samples with mutations using the Xpert w MTB/RIF and GenoType w MTBDRplus assays. RESULTS: Of 1091 samples, 5.8% failed amplification, and six contained other mycobacteria. In 2005 and 2010, respectively 2/500 (0.4%) and 11/522 (2.1%) amplicons showed non-silent mutations. At least 7/13 of these belonged to the disputed group, with 5/7 patients suffering adverse treatment outcome. One silent muta- tion went undetected by commercial assays. Following routine DST indications, only three cases with a non- silent mutation were eventually detected. CONCLUSIONS: Disputed rpoB mutations may be responsible for the majority of rifampicin (RMP) resistance among new cases, and lead to adverse outcomes of first-line treatment. Silent mutations do not necessarily cause Xpert or line-probe assay false RMP-resistant results. Molecular RMP DST could greatly simplify resistance surveillance, in addition to offering the best prospects for early and accurate individual diagnosis. KEY WORDS: tuberculosis; drug resistance; surveil- lance; rifampicin; rpoB gene WE REPORTED previously that rpoB mutations with disputed significance are responsible for over 10% of rifampicin (RMP) resistance among first-line failure and relapse cases in Bangladesh and Kinshasa, and that the outcome of standardised retreatment of these patients is as poor as in those with well-known, undisputed mutations. 1 Although the resistance caused by these mutations, such as 511Pro, 516Tyr, 526Asn, 526Leu, 526Cys, 533Pro and 572Phe, is often missed with phenotypic drug susceptibility testing (DST), particularly using the automated MGITe 960 system (BD, Sparks, MD, USA), the minimal inhibitory concentration (MIC) of such isolates on egg-based medium with final reading at 6 weeks was clearly above the RMP resistance cut- off. 2 Due to missing information, we had pointed out that such isolates might not have the same importance among new, treatment-na¨ıve patients. Subsequent publications questioning the importance of these isolates and the failure of the MGIT 960 system to detect them were highlighted as a potential limitation of our previous studies, i.e., that these applied only to patients who were already failing treatment or relapsing. 3 Currently widely used genotypic DST methods, such as Xpert w MTB/RIF (Cepheid, Sunnyvale, CA, USA) and line-probe assays (LPA), have little problem detecting disputed mutations located in the core region of the rpoB gene (codons 507 to 533). Rapid, phenotypic DST regularly misses this resistance, resulting in false-resistant scores, when used as the gold standard to evaluate molecular test perfor- mance. These false results of an (imperfect) gold standard thus lead to an underestimation of the specificity of new molecular tests. 1 Particularly among new cases with a low prevalence of RMP resistance, this reliance on rapid phenotypic DST as an imperfect gold standard would then lead to the seemingly low accuracy of molecular resistant results and the need for confirmation using another method, as previously recommended by the World Health Organization (WHO). 4 In the context of low-income Correspondence to: Armand Van Deun, Mycobacteriology Unit, Institute of Tropical Medicine, Nationalestraat 155, 2000 Antwerp, Belgium, Tel: (þ32) 3 345 5548. Fax: (þ32) 3 247 6333. e-mail: avdeun@itg.be Article submitted 30 August 2014. Final version accepted 16 October 2014.