Predictive Value of Interferon Gamma Release Assay for Incident Active Tuberculosis: 2-year Follow-up Results of TBfreeCorea = Predictive Value of Interferon Gamma Release Assay for Incident Active Tuberculosis: 2-year Follow-up Results of TBfreeCorea

다국어 초록 (Multilingual Abstract)

Background
The aim of this study was to investigate the predictive value of interferon gamma release assay (IGRA) for incident active tuberculosis (TB).
Methods
Among the cohort of TBfreeCorea, participants who underwent IGRA in 2017 were followed up until January 2nd, 2020. To investigate the natural history of LTBI, those with negative IGRA Results and those with positive Results but who did not initiate LTBI treatment were included in this study. Sensitivity, specificity and predictive values were investigated with 2-year cumulative TB incidence. Effects of interferon (IFN) level and other risk factors on incident TB cases was analyzed in a Cox proportional hazard model with time varying covariates.
Results
Among the enrolled 810,294 participants, a total of 570 cases of active TB were identified. Two-year cumulative TB incidence was 62.7 cases per 100,000 population. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were 63.39%, 89.44%, 0.38% and 99.97% respectively, with current cutoff value of 0.35 IU/mL. PPV was higher among the young aged group - 0.90% in age group < 20 and 0.96% in age group 20-35. When cutoff value was raised to 0.67, 1.41 and 3.33 IU/ml, PPV were improved to 0.44%, 0.51% and 0.62%, respectively, but sensitivity decreased to 58.46%, 48.62% and 36.81%. In a Cox proportional hazard model, risk of active TB increased with higher IFN level - when compared with IGRA negative group, hazard ratio among the group with IFN level of 0.35-0.67, 0.67-1.41, 1.41-3.33, 3.33-10.0 and > 10 IU/ml was 6.56 (4.29-10.03), 13.38 (9.56-18.73), 18.15 (13.19-24.99), 31.81 (24.01-42.16), 43.51 (32.21-58.79), respectively. In a subgroup analysis stratified by age, such tendency was more prominent in younger age group, whereas obscured in elderly group.
Conclusions
IGRA showed quantitative predictive value for TB incidence, especially among young age group. However, low PPV remains a limitation.