Difficulty in the interpretation of laboratory pyrazinamide (PZA) resistance results in clinical cases of multidrug-resistant tuberculosis in the UK and USA

Background. Multidrug-resistant tuberculosis is an increasing problem worldwide. Recently several nosocomial outbreaks of MDRTb have been reported throughout Europe and the USA. Contributory factors to these outbreaks have been delays in the recognition of drug resistance leading to inappropriate treatment and prolonged periods of infectivity. We examine the reliability of different methods of determining PZA resistance in the UK compared with those from a laboratory in the USA. Methods. BACTEC radiometric and conventional resistance ratio methods of sensitivity testing were performed on 12 isolates of PZA resistant MDRTb recently isolated in London. Each isolate was further digested and the pncA gene amplified and sequenced. These tests were carried out in both laboratories independently. SSCP was performed on each PCR product. Production of PZAase was also assessed using Wayne's method and ¹⁴C labelled High Performance Liquid Chromatography. Results. Twelve strains reported as resistant by the conventional resistance ratio method exhibited MICs from <100μg/ml to >400μg/ml by the BACTEC system. Single point mutations were detected in 10 out of 12 isolates. Low levels of PZAase activity were detected in four isolates with a point mutation and borderline MICs. Two isolates had no mutation in the pncA gene. Conclusion. In this study discordant laboratory results between conventional and BACTEC radiometric methods led to difficulties in the diagnosis of PZA resistance and clinical treatment of patients. More reliable routine methods of determining PZA resistance are required.