THU0622B PREVENTION OF ACTIVE TUBERCULOSIS INFECTION USING INTERFERON GAMMA RELEASING ASSAY (IGRA) IN PATIENTS WITH RHEUMATOID ARTHRITIS UNDERGOING TUMOR NECROSIS FACTOR-ALPHA INHIBITORS TREATMENT

Abstract

Background: Data on effectiveness reducing incidence of active tuberculosis (TB) by use of interferon gamma releasing assay (IGRA) have not yet been available. Objectives: To evaluate whether diagnosis of latent tuberculosis infection (LTBI) using IGRA is effective in reducing occurrence of active-TB among patients with RA who will use Tumor Necrosis Factor alpha inhibitors (TNFi). Methods: A retrospective cohort study was conducted using the National Health Information Database of Korea between 2008 and 2017. The enrollees had at least two claims with the RA diagnosis code, and has at least one prescription of TNFi since 2009 when national insurance of Korea started to cover the expense of IGRA. Exclusion criteria were organ transplantation, HIV infection, or active TB infection within 1 year prior to enrollment. The follow up was initiated on the day of the first prescription for TNFi, and ended on the date of the first claim on active TB, the date of death or December 31, 2017 whichever came first. The utilization of the diagnostic test method was confirmed by the codes for tuberculin skin test (TST) and IGRA claimed within 2 months prior to the initiation of the TNFi. Patients were categorized into 3 groups; no-test, TST only, IGRA (IGRA only or IGRA and TST). The incidence of active TB was defined as the presence of claims with diagnosis code for TB and the start of standard 3-drug or 4-drug first line anti-TB medication combination therapy. Descriptive statistics on baseline characteristics of participants in each group were presented before and after application of weighting method using standardized mortality rate weight (SMRW). The incidence of active TB in each group was presented as event per 1000 person-year (PY). We treated death as competing event to the occurrence of TB. Therefore a sub-distributional hazard model was used to calculate weighted hazard ratio (wHR) and 95% confidence intervals (CI). Results: Among the 27,021 patients who started TNFi, 15,538 RA patients were included in the analysis through inclusion and exclusion criteria. There were 6,287 patients in the no-test group, 4,677 patients in the TST group, and 4,574 patients in the IGRA group. Diabetes mellitus, chronic kidney disease, and congestive heart failure were more prevalent in IGRA group, while stroke was observed less frequently in IGRA group. After weighting using SMRW, all confounders were well balanced. During the follow up period, 4.24, 4.93, and 4.01 active TB events per 1,000 PY were observed for no-test, TST, and IGRA group, respectively. The death rate was 8.63, 9.75 and 7.96 per 1,000 PY for no-test, TST, and IGRA group, respectively. Considering death as competing event, IGRA significantly reduced the risk of active TB than no-test group (wHR 0.628, 95% CI: 0.459-0.859) and TST group (wHR 0.416, 95% CI: 0.311-0.556). Conclusion: The incidence of active TB in RA patients using TNFi could be reduced by use of IGRA.