IP-10 distinguished with high accuracy active TB from LTBI irrespective of HIV infection and declined during anti-TB chemotherapy. Plasma IP-10 may serve as a diagnostic biomarker to differentiate between the stages of TB infection and for monitoring therapy efficacy.

BackgroundInterferon-γ release assays (IGRA) serve as immunodiagnostics of tuberculosis (TB) infection to identify individuals with latent TB infection (LTBI) eligible for preventive anti-TB therapy. In this longitudinal study of HIV-infected LTBI patients we have observed for possible progression to active TB as well as evaluated repeated IGRA testing in a TB low-endemic setting.MethodsQuantiFERON TB-Gold In-tube® assay (QFT), TB-SPOT.TB® (TSPOT) and tuberculin skin test (TST) were performed on 298 HIV-patients recruited from seven out-patient clinics in Norway. Patients with active TB, LTBI and negative IGRA were followed with repeat QFTs and clinical evaluation over a period of 24 months.ResultsSeven HIV-patients (median CD4 count 270; IQR 50–340) were diagnosed with active TB at inclusion, all IGRA positive. Sixty-four (21%) HIV-patients (median CD4 count 471; IQR 342–638) were diagnosed with LTBI and of these 39 (61%) received TB preventive treatment. Neither treated nor untreated HIV-infected LTBI patients developed active TB during the 24 months. At baseline, the median interferon-γ (INF-γ) level measured by QFT was 3.48 IU/ml (IQR 0.94 – 8.91 IU/ml) for treated LTBI compared to 1.13 IU/ml (IQR 0.47 – 4.25 IU/ml) for untreated LTBI patients (p = 0.029). The QFT reversion rates were 75% for active TB, 23% for treated LTBI and 44% for untreated LTBI, whereas the conversion rate for the non-TB group was 7% despite no new TB exposure. There was no significant difference in the trend of INF-γ levels over time between treated and untreated LTBI patients.ConclusionThe prevalence of LTBI is high among HIV-patients, but the risk of developing active TB seems to be low in patients with high CD4 counts in this TB low-endemic setting. In several patients, especially with baseline IFN-γ levels close to cut-offs, the QFT tests reverted to negative independent of preventive anti-TB treatment indicating possibly false positive tests. This highlights the importance of defining reliable cut-offs for immunodiagnostic tests and deferring preventive therapy in selected patients. Randomized studies with longer follow-up time are needed to identify HIV-patients that would benefit from LTBI treatment in a TB low-endemic setting.Electronic supplementary materialThe online version of this article (doi:10.1186/s12879-014-0667-0) contains supplementary material, which is available to authorized users.

The prevalence of positive IGRAs in HIV-infected patients was high in this low TB endemic setting. Lower QFT-GIT positivity in long-term residents from TB-endemic countries may reflect a waning of TB-specific immune responses.