Drug resistant tuberculosis (DR-TB) is a serious public health issue both globally and nationally, with South Africa and Kwazulu-Natal, in particular, being among the regions with the highest burden of DR-TB. Detecting drug resistance and initiating patients onto the appropriate therapy, in the shortest possible time, is of utmost importance to the effective management of DR-TB. The development of molecular diagnostic techniques allows for more rapid diagnosis of TB, as well as drug resistance, leading to earlier diagnosis and subsequent initiation onto appropriate treatment.
For phenotypic drug susceptibility testing (DST), the laboratory turnaround time is 4 – 6 weeks, thus patients are either initiated onto empiric and sometimes inappropriate treatment or have to wait to be initiated onto appropriate therapy, remaining untreated and infectious for extended periods of time. The introduction of GeneXpert testing revolutionised TB diagnostics as it allowed for diagnosis of TB whilst also providing susceptibility results for rifampicin within a few hours. Direct 1st and 2nd line LPA testing was included in the DR-TB management algorithm to further reduce the time to treatment initiation of multidrug resistant TB (MDR-TB) and extensively drug resistant TB (XDR-TB). This also ostensibly reduces the amount of time a patient is transmissible for and improves treatment outcomes. This study was undertaken to assess the impact of the introduction of the direct 1st and 2nd LPA reflex testing on the management of DR-TB in the Umzinyathi District of Kwazulu-Natal.
The cohorts before and after the roll-out of direct 1st and 2nd line LPA testing were analysed for patient characteristics, diagnostic information, time to appropriate treatment initiation and treatment outcomes. Furthermore, the diagnostic tests were compared to ascertain if 1st and 2nd line LPA is comparable to phenotypic DST for drug susceptibility testing.
There were 141 patients included in the 2015/2016 cohort before direct 1st and 2nd LPA was included in the algorithm, and 102 patients in the 2017/2018 cohort after its implementation. There was a significant decrease between cohort 1 and cohort 2, in the laboratory turnaround time for both 1st line LPA, which decreased from 36 days (IQR 23 – 60) to 17 days (IQR 11 – 30), respectively, and 2nd line LPA, which compared to phenotypic DST, decreased from 45 (IQR 23 – 67) to 21 days (IQR 12 – 50). Time to appropriate treatment initiation was similar across both cohorts for RR- and MDR-TB, from 8 days (IQR 5 – 13) to 9 days (IQR 7 – 29) in the second for RR-TB, and from 8 days in cohort 1 (IQR 6 – 20) to 12 days (IQR 6 – 50) in cohort 2 for MTB-TB. The time to appropriate treatment was significantly reduced in XDR-TB patients from 267 (IQR 145 – 796) to 62 days (IQR 45 – 182)
(p=0.018). Moreover, the treatment outcomes in XDR-TB improved after the roll-out of direct 1st and 2nd line LPA. Xpert, 1st line and 2nd line LPA performed well compared to phenotypic DST for antibiotic resistance detection.
The laboratory turnaround time and time to appropriate treatment initiation improved after the implementation of direct 1st and 2nd line LPA. Despite a delay in initiating therapy after laboratory diagnosis, there were positive impacts found regarding treatment outcomes of XDR-TB. Patients were initiated on the appropriate treatment, in response to 2nd line LPA results, in the first instance, which improved treatment success rates in XDR-TB patients.
Keywords: Mycobacterium tuberculosis, line probe assay, MDR-TB, XDR-TB, phenotypic DST||en_US