Background: Diagnosis and treatment of latent tuberculous infection decrease the incidence of tuberculosis (TB) in HIV-infected patients.
Objectives: To evaluate the diagnostic yield of two IFN-γ release assays and tuberculin skin testing for the screening of latent infection in HIV-infected patients.
Methods: We performed a prospective study in 29 referral centers for HIV care in France. Asymptomatic, antiretroviral-naive patients infected with HIV-1 who consented to participate underwent two commercial tests (T-SPOT.TB and QuantiFERON-TB Gold In-Tube ELISA test [QFT]) and skin test at enrollment and were followed up for clinical events during 24 months.
Results: Between March 2009 and 2011, 506 patients were included, of whom 415 performed the three tests. Median age was 38 years (interquartile range, 31-45 yr), with median CD4 cell count of 466/μL (337-615 μL), and HIV viral load of 4.5 log10 copies/ml (3.6-4.9 log10 copies/ml). At least one IFN-γ release assay was positive for 55 (13.5%) patients: QFT (n = 43), T-SPOT.TB (n = 34), both (n = 22). Skin test was positive (>5 mm) in 66 (15.9%) patients, with intertest agreement at 81 to 86%. On multivariate analysis, positive IFN-γ release assay was only correlated with country of birth (8.4% for France vs. 17.9% for high-prevalence countries, P = 0.004). Of the 55 patients with positive IFN-γ release assay, 8 (14.5%) developed active TB, all within 120 days. No other case of active TB was diagnosed. Once active TB was excluded, IFN-γ release assay-based latent infection prevalence was 11.8%.
Conclusions: Systematic screening for latent TB infection by IFN-γ release assay identifies a population at high risk of active TB over the next months. An extensive diagnostic work-up for active TB must follow positive IFN-γ release assay, before considering treatment of latent infection. Clinical trial registered with www.clinicaltrials.gov (NCT00805272).
Keywords: HIV; IFN-γ release assay; latent tuberculosis infection; tuberculin skin test; tuberculosis.