Comparative evaluation of molecular based detection methods for drug resistant tuberculosis

Abstract

Tuberculosis (TB) is a globally problematic infectious disease. According to the World Health Organization (WHO) multidrug resistant tuberculosis (MDR-TB) has recently increased by an annual rate of N20%. In 2015; the WHO noted 580,000 MDR-TB cases of which b65,000 were cured. This highlights the need for new and improved methods for TB drug resistance profiling. The aim of this study is to evaluate and compare three molecular based drug resistance profiling kits to determine which is the most effective in targeting mutations in the MTBC genome known to confer resistance to first- and second-line anti-tuberculosis drugs. The three kits are: A commercially available kit (HAIN Genotype MTBDR VER 2.0), a kit in the process of validation (Kit A) and a kit of our own design (Kit B). A total of 100 blinded and randomized Mycobacterium tuberculosis complex (MTBC) DNA isolates were collected from the Centre for Tuberculosis (CTB) at the National Institute for Communicable Diseases (NICD). Isolates will be subjected to analysis with each kit. The kits will be evaluated in terms of cost, turnaround time, ease of use and sensitivity and specificity using whole genome sequencing data available on the isolates at the CFB as reference. Five isolates have been analysed during pre-evaluations on Kit A. The kit was able to identify 4 gene mutations i.e. rpoB, katG, embB and rrs and did not present any false positive phylogenetic markers. The kit however failed to detect a gyrA mutation in one isolate. Mutations in the gyrB, pncA and inhA promotor region are not detectable with Kit A. The sensitivity and specificity for the selective gene mutations were 90% and 100% respectively. A further six isolates with known gyrA mutations were evaluated using Kit A to determine if the prior result was due to a low analytical sensitivity or suboptimal reagent design. The sensitivity for identifying the gyrA mutations was 50% and reflected that the kit was not able to detect the Asp94His mutated codon as confirmed by the supplier. In addition, the results concurred with RIF and INH resistance on available phenotypic data. Results of pre-evaluations on Kit A show high sensitivity and specificity for selective gene mutations. It was determined though that Kit A lacks in sensitivity for the gyrA mutations and will need optimization of the current reagent design. The kit however has a low coverage of the 11 different conferring mutations of interest. Kit A is robust and takes less than two hours but requires repetitive pipetting, which could lead to cross contamination and false positive results. Further evaluations are in progress