COMPARISON OF XPERTMTB/RIF AND GENOTYPE MTBDRPLUS ASSAYS IN THE DETECTION OF TUBERCULOSIS AND RIFAMPICIN RESISTANCE AMONG RETREATMENT SAMPLES IN KENYA

Jeremiah O 1 , Kiiyukia C 1 and Kiiru J 2 . 1. Jomo Kenyatta University of Agriculture and Technology, College of Health Sciences. 2. Kenya Medical Research Institute, Centre for Microbiology Research, Kenya. ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History

Tuberculosis (TB) continues to cause high morbidity and mortality in developing as well as the developed countries. In the African region, the pandemic is mainly propelled by TB/HIV co-infection since HIV increases the reactivation of latent TB as well as the rapid progression to active infection (Corbett et al., 2003). Kenya remains in the list of high TB burden countries (WHO, 2017). According to a prevalence survey carried out in 2016, it was reported that the TB burden in the country is higher than previously thought and currently stands at 558 cases per 100,000 population with approximately 40% of cases missed annually (Ministry of Health, 2016). In order to determine emergence of resistance to first line medicines. Sputum samples from previously treated patients are usually sent for culture and drug susceptibility testing in designated laboratories in the country. Kenya has adopted the new WHO approved molecular technologies which have greatly improved case detection as well as turnaround time in the diagnosis and management of TB (Ministry of Health, 2016). The XpertMTB/RIF introduced in Kenya in 2011 and the GenoTypeMTBDRplus have all been incorporated in the in-country's TB testing algorithm. The two tests have a great ability to detect TB as well drug resistance (Bablishvili et al., 2015). Genotype MDRTBplus technology also known as Line Probe Assay (LPA) is a qualitative in vitro diagnostic test based on DNA-STRIP technology for the identification of M. tuberculosis complex and its resistance to rifampicin (RMP) and/or isoniazid (INH), while the Xpert MTBRIF assay is based on real-time PCR. However LPA requires several manual steps to prepare DNA template, which could result in the loss of DNA during processing; the technique also involves manual hybridization steps which could decrease the sensitivity of the assay. Nevertheless, the XpertMTBRIF assay only detects RIF resistance while GenoTypeMTBDRplus detects both rifampicin and isoniazid (INH) (Rufai et al., 2014). Since their introduction in the country, there are no elaborate studies that have been done to compare their performance in our setting. Initially, surveillance was performed using culture and later GenoTypeMTBDRplus was introduced for DNA. Rollout of XpertMTRIF assay across the country is ongoing with plans to make it the first diagnostic test for all. Presumptive and previousely treated TB patients need quick responses of laboratory results for management. In the Kenyan context surveillance using culture and then later introduction of LineProbeAssay/ GenoTypeMTBDRplus has been in place. Further, the current testing algorithm aims at expanding XpertMTBRIF assay services which can readily be available and accessible to the peripheral facilities to ensure that there is access to universal drug susceptibility testing to all bacteriologically confirmed cases of Tuberculosis.
The current study assessed the performance of XpertMTB/Rif and GenoTypeMTBDRplus in the diagnosis of tuberculosis and detection of rifampicin resistance among retreatment cases considered as a high risk for developing drug resistant tuberculosis.

Materials and Methods:-
The study was carried out at Kenya Medical Research Institute, Centers for Disease Control Tuberculosis culture laboratory (KEMRI-CDC) Kisumu. The study was reviewed approved by Kenyatta National Hospital-University of Nairobi Ethical Review Committee (KNH-UoN ERC). Sputum samples collected from TB retreatment patients were subjected to, XpertMTB/Rif, GenoTypeMTBDRplus assays as well as liquid MGIT culture. Processed specimens were inoculated onto BACTECTMGIT TM 960 broth culture system (BD diagnostics) as described by (Lu, Heeren, & Dunne, 2002). All sputum samples received through courier delivery were processed using the N-acetyl-l-cysteinesodium citrate-NaOH (NALC-NaOH) method. Samples were decanted following centrifugation, and the sediments were resuspended in 3 ml of phosphate buffer solution. Aliquots were prepared from the processed samples to perform, MGIT960 culture, GenoTypeMTBDRplus assay as previously described by (Ombura et al., 2016). All the processed sputum samples were equally subjected to XpertMTBRIF assay according to the manufacturer's instructions. Data generated by XpertMTRIF, GenoTypeMTBDRplus and MGIT culture were recorded on laboratory log books as well as Excel worksheets. The data was coded and analysis was done using STATA version 13. The sensitivity and specificity of the tests were determined with reference to MGIT culture as a gold standard. Receiver operating characteristic (ROC) analysis was performed and the output of the areas under the curve (AUC) for the tests were compared.

Results:-
A total of 561 sputum samples were examined in this study. The mean age of the study participants was 38.7 years with a standard deviation of 15.64. The minimum and maximum ages were 1 and 95 respectively. Figure 1 gives an insight on the frequency of MTB detection by XpertMTBRIF assay and GenoTypeMTBDRplus (LPA) for the 561 samples tested. Twenty six percent (26.74%, n=150) of the samples tested were positive for MTB by both LPA and XpertMTBRIF assay while 54.9% (n=308) were negative for MTB by both tests. Among the samples that were positive for MTB by XpertMTBRIF assay, 11.76% (n=66) were negative for MTB by LPA. On the other hand, 4.28% (n=24) of samples that were positive by LPA were detected as negative for MTB by XpertMTBRIF assay. Among the samples that were detected as an error 2.3% (n=13) by XpertMTBRIF assay, 1.6% (n=9) and 0.71% (n=4) were detected as negative and positive for MTB by GenoTypeMTBDRplus (LPA) respectively. The performance of XpertMTBRIF assay in the detection of MTB was compared to MGIT culture results as presented in Figure 2. There was a concurrence of 18.9% (n=106) between XpertMTBRIF assay and MGIT on samples that were detected as positive for MTB. The XpertMTBRIF assay detected MTB in 17.5% (n=98) of samples that were negative for MTB by MGIT culture. Forty nine percent 49.9%, (n=280) of the samples were negative for both XpertMTBRIF and MGIT culture. Among the culture positive samples, 2.3% (n=13) were negative for MTB by XpertMTBRIF assay. Culture contamination rate was generally low but within the acceptable limit (5.17%). However, MTB was detected by XpertMTBRIF assay in (1.4%) of samples which had growth of non-Tuberculous mycobacterium (MOTT) on culture.        Table 3 illustrates the frequency of Rifampicin resistance detection by both diagnostic tests. A total of 14 samples were flagged as Rifampici resistant by LPA while 12 were detected by MGIT. Eight samples (57%) that were Rifampicin resistant by LPA, were similarly detected as Rifampicin resistant by MGIT culture DST. However, 4 (29%) of the samples that were detected as Rifampicin resistant by LPA were identified as Rifampicin susceptible by MGIT culture DST. Ninetythree MTB positive samples were negative for Rifampicin resistance by both diagnostic tools.     Figure 5 illustrates the corresponding ROC curves and the reported AUC. Test for equality between the AUC generated a p=0.9548 (p>0.05).

Discussion:-
The rapid detection of TB and Rifampicin resistance by XpertMTBRIF facilitates timely initiation of treatment for patients hence reducing the TB transmission cycle. GenoTypeMTBDRplus can detect resistance to Isoniazid (INH) which is also known to be significant in terms of patient treatment outcomes (Naidoo, Du Toit, et al., 2014 According to a study carried out to compare XpertMTB/RIF assay and GenoTypeMTBDRplus DNA probes for detection of mutations linked to Rifampicin resistance, the agreement of XpertMTB/RIF and GenoTypeMTBDRplus with LJ-DST for detection of Rifampicin susceptibility was found to be 93.5% and 92.4%, respectively (Rahman et al., 2016). The study also reported a 92.4% overall agreement of the two molecular methods for the detection of Rifampicin susceptibility. Results from another study demonstrated that XpertMTBRIF had an excellent ability to detect Rifampicin resistance (Sharma et al., 2014). A study that compared the utility of XpertMTB/RIF & GenotypeMDRTBplus in the diagnosis of bone and joint tuberculosis reported the sensitivity of XpertMTB/RIF for detecting Rifampicin resistance at 100%, and the sensitivities of GenotypeMDRTBplus in the detection of Rifampicin and Isoniazid (INH) resistance were 83.3% and 85.7%, respectively (Gu et al., 2015).
Findings from this study also point to the possibility of co-infection of some patients with MTB and MOTT. XpertMTB/RIF assay detected MTB in 1.43% (n=8) of MOTT positive samples while LPA detected 1.25% (n=7) of the samples. This results are critical for the management of these patients in order achieve the desired cure rates. Coinfection with MTB and MOTT has also been documented in other studies (Sekadde et al., 2013).
The introduction of new Tuberculosis diagnostic technologies has greatly improved the detection of Tuberculosis and drug resistance. The turnaround time towards initiation on treatment for patients has also improved contributing positively to the overall quality of TB services in the country. Evidence also show that the XpertMTBRIF assay detects with high specificity the extra-pulmonary TB (EPTB) cases with smear-positive non-respiratory samples such as cerebrospinal fluid and tissues (Maynard-Smith, Larke, Peters, & Lawn, 2014). Rapid and accurate results from such techniques has made it possible to reduce TB associated mortality (Naidoo, et al., 2014).

Conclusion:-
Overall, findings from this study indicates that XpertMTBRIF assay and GenoTypeMTBDRplus are excellent TB molecular diagnostic tools. There is no significant difference in the detection of MTB as well as Rifampicin resistance. These platforms have enabled the rapid confirmation of cases and increased access to universal Drug Sensitivity Testing. On the other hand, the cost of implementing and sustaining these technologies is high and requires more funds which may be prohibitive in resource limited settings.