Background: Individuals with end-stage renal disease (ESRD) are 10- to 25-fold more likely than immunocompetent people to develop active tuberculosis (TB) and are candidates for being treated for latent TB infection (LTBI). However, diagnosis using the tuberculin skin test (TST) is doubly difficult due to cutaneous anergy and cross-reactions with Bacille-Calmette-Guérin (BCG) vaccination.
Materials and methods: This was a prospective, doublematched, cohort study in which 32 ESRD patients and 32 age-matched, healthy controls were enrolled. The TST and two new interferon-gamma blood tests, QuantiFERON-TB Gold (QFT-G) and T-SPOT.TB (ELISPOT), were performed. The subjects were followed up 2 years for active TB disease. ELISPOT was done in ESRD patients only.
Results: Compared to the healthy controls, a high prevalence of LTBI was found in the ESRD patients by TST (62.5%, 95% confidence interval [CI] 43.7-78.9), QFT-G (40.0%, 95% CI 22.7-59.4), and ELISPOT (46.9%, 95% CI 29.1-65.3). Agreement was moderate (kappa [kappa] = 0.53) for QFT-G and ELISPOT but only slight between TST and QFT-G (kappa = 0.25) and fair between TST and ELISPOT (kappa = 0.32). ESRD (p = 0.03) and diabetes mellitus (p = 0.04) were significant risk factors for QFT-G positivity on the multivariable analysis. The overall rate of active TB was 1.66 cases per 100 person-years (pys), with the rate higher in patients with ESRD (3.53 per 100 pys) and those with positive (3.40 per 100 pys) and indeterminate QFT results (30.16 per 100 pys), although the difference was not statistically significant. Sensitivity, specificity, and positive and negative predictive values of QFT-G for active TB was 100%, 62.1%, 8.3% and 100%.
Conclusion: This pilot study is the first to compare QFT-G, ELISPOT, and TST in ESRD patients on hemodialysis and demonstrates a high prevalence of LTBI in this population. In our study, the QFT-G was the more accurate method for identifying those truly infected with Mycobacterium tuberculosis, even in BCG-vaccinated individuals.