UCLA

Background: Tuberculosis (TB) is one of the top ten causes of death by a single infectious agent. Extensively Drug-Resistant TB (XDR-TB) is a type of MDR TB, defined as TB resistant to at least isoniazid and rifampicin, plus any fluoroquinolone and at least one of three injectable second-line drugs. XDR-TB is considered a significant public health threat in Thailand because of the rising of reported XDR-TB cases, the more financial burden for both patient and provider sites, and more adverse events in XDR-TB treatment. This dissertation examines the biopsychosocial factors, such as a) whether non-adherence to the MDR-TB treatment regimen is a leading cause of development to XDR-TB, b) which NAT2 diplotype or HLA class 2 alleles associated with XDR-TB cases, c) the pattern of initial culture conversion in XDR-TB group while on MDR-TB treatment compare to MDR-TB group.

Methods: We conducted a matched case-control study and recruited all secondary XDR-TB cases that were reported to the Bureau of Tuberculosis from October 2014 to June 2019. A total of 45 secondary XDR-TB cases were enrolled at 32 original hospital sites, across 29 provinces in Thailand. MDR-TB controls were randomly selected from the same hospital, where XDR-TB cases were treated. The proportion of cases to controls was one XDR-TB to two MDR-TB matched by geographical region. In the first two studies (Chapter 2-3), nine of 45 secondary XDR-TB cases died before we started. Then, only 36 XDR-TB cases and 75 MDR-TB controls were available for interviews and genetic tests. The conditional logistic regression was used to assess the magnitude of the non-adherence effect. Fisher exact test and conditional logistic regression were also used to explore the association between NAT2 diplotype, HLA class 2 alleles, and XDR-TB cases. In the third study (Chapter 4), we could not find the record of sputum culture in seven of 45 secondary XDR-TB cases. We used Cox proportional hazard regression for the difference in time to initial culture conversion in 38 XDR-TB cases and 76 MDR-TB controls. We also used the sensitivity and specificity of initial culture conversion at the beginning of MDR-TB treatment to 15-month in the prediction of the XDR-TB susceptibility and MDR-TB treatment outcome.

Results: We found that patients with non-adherence while on MDR-TB treatment were ten times more likely to develop XDR-TB (AOR 10.08, 95%CL 1.36, 74.87) while adjusted for type of Directly Observed Treatment, age, gender, socioeconomic status, and harmful alcohol use which was less than the magnitude from improper regimen (AOR = 26.89, 95%CL 7.39, 97.81) while adjusted for culture sending and region. For the genetic study, NAT2*4/5B (OR 3.9, 95%CI 1.2 13.0) and the heterozygosity of HLA-DRB1 16:02 were about four times associated with XDR-TB cases (OR 3.7, 95%CI 1.2, 11.9). None of the associations was found in HLA DQB1. For the third study, the improper regimen for MDR-TB treatment was found to significantly reduce the chance of initial sputum culture conversion (aHR 0.24, 95%CL 0.15, 0.38) adjusted for wrong drug administration and dosage, and follow-up sputum culture. Furthermore, positive culture at the end of four-month of treatment was a good predictor for failure outcome (sensitivity 67.7%, specificity 93.8%) and the development of XDR-TB (sensitivity 63.2%, specificity 94.7%).

Conclusion: The main factor that causes XDR-TB was not non-adherence; the more significant factor was improper regimens for MDR-TB treatment, which mainly came from the provider site. From the aspect of host genetic, the result of this study was found significant in NAT2 *4/5B and heterozygosity of HLA-DRB1 16:02. There was no clinically significant for NAT2. However, HLA-DRB1 16:02 is associated with acquired immunodeficiency that can make the host more likely to develop XDR-TB. For the follow-up of sputum culture conversion, the improper MDR-TB treatment regimen also played a significant role in reducing the chance of sputum culture conversion. Also, the lack of sputum culture conversion at month four would give us the best prediction for the development of XDR-TB. Overall, the finding from these three studies suggested the probable risk and protective factors of XDR-TB. From a policy perspective, further work is needed to examine the impact of the new DR-TB guidelines on the actual MDR-TB treatment regimen and assess the effectiveness of the treatment audit system when it starts.