Tuberculin diagnostics in pulmonary tuberculosis combined with HIV-infection

Abstract

Aim. To evaluate the sensitivity to tuberculin in patients with infiltrative pulmonary tuberculosis in combination with an infection, caused by the human immunodeficiency virus (HIV), using various different tests. Methods. Examined were patients with infiltrative pulmonary tuberculosis with HIV infection (134 patients, the main group) and without HIV infection (100 patients, control group). Tuberculin diagnostics included an intradermal Mantoux test with 2 tuberculin units, the skin prick test with a 100% solution of PPD-L and the intradermal test «Diaskintest». Results. In the main group, a positive reaction to the Mantoux test with 2 tuberculin units was reported significantly less frequently (56.0%) than in the comparison group (89.0%), the diagnostic value of the test decreased with a decrease in the number of CD4+ lymphocytes. The prick test was characterized by a complete correlation with the results of the Mantoux test, although having a number of advantages. The sensitivity of the «Diaskintest» in the main group was 48.3%, and in the control group - 84.6%, at the same time the sensitivity of the «Diaskintest» decreased to 17% in cases of pronounced immunodeficiency. Conclusion. The skin prick test can be used as a screening method for early detection of tuberculosis in HIV-infected individuals with CD4+ lymphocyte count of 300 cells in 1 ml or more with the aim of forming risk groups, followed by screening for tuberculosis.

The diagnostic performance of novel skin-based in-vivo tests for tuberculosis infection compared with purified protein derivative tuberculin skin tests and blood-based in vitro interferon-γ release assays: a systematic review and meta-analysis

“…Four studies 25,27,32,40 performed head-to-head comparisons of Diaskintest vs TST 5mm and IGRA in the same study. Study populations enrolled in studies included HIVinfected individuals 33,38,44,49 (346 adults, 23 children)) and children under 18 (N=3,803) (Table 1). 21,[25][26][27][28][29][30][31][32][34][35][36]41,48 Approximately half of the HIV-infected cohorts had a CD4 count lower than 200 cells/mm 3 .…”

“…33,38 Proportion BCG vaccinated was reported in 4 studies, and ranged from 93-100%. 25,27,28,44 Diaskintest threshold for positivity varied and included any skin induration (Diaskintest AI ) according to national guidance 50 or 5mm (Diaskintest 5mm ). Studies used PPD-L TST previously shown bioequivalent to PPD-RT23; 51,52 5mm (TST 5mm ) induration denoted positivity, and 1-5 mm indeterminate.…”

“…Of the 14 studies evaluating sensitivity of Diaskintest, risk of bias was high in 5 (35.7%) studies where test assessors were not blinded to TB culture results, 22,33,37,38,44 and unclear in at least one of the four risk of bias criteria in 12 (85.7%) studies as information on patient selection or blinding was not presented. 21,22,32,33,35,37,38,40,[43][44][45][46] Of those evaluating Diaskintest concordance, 11/13 (84.6%) had high risk of bias in the reference standard criterion as assessors of reference standard (TST) were not blinded to index test results, 25,[27][28][29][30][31]34,42,44,47,49 whereas for the index test criterion one had high risk of bias as index test assessors were not blinded to reference standard results 44 and the remaining 12 (92.3%) were classed as unclear as this information was not provided. 21,[25][26][27][28][29][30][31]34,42,47,49 Of all Diaskintest studies, patient selection bias was unclear for 23 out of 29 (79.3%) studies as reporting of patient selection was incomplete.…”

“…21,[25][26][27][28][29][30][31]34,42,47,49 Of all Diaskintest studies, patient selection bias was unclear for 23 out of 29 (79.3%) studies as reporting of patient selection was incomplete. [21][22][23][24][25][26][27][28]30,33,[37][38][39][40]42,44,45,[47][48][49] One C-Tb study scored high on a risk of bias criterion because not all participants received the same reference standard (IGRA or TST). 54 Four out of five (80.0%) C-Tb studies and one EC-skintest study had conflict of interest concerns, as studies either did not report disclosures or were directly affiliated with the test manufacturer.…”

“…Pooled estimates of Diaskintest sensitivity in two-way comparisons were 67% and 88% in HIV-uninfected adults for Diaskintest AI and Diaskintest 5mm , respectively. 22,32,36,37,40,[43][44][45][46] Highly variable methods and sub-populations precluded meaningful meta-analysis for most risk groups; sensitivity estimates from individual studies ranged from 40%-71% in HIV-infected adults 33,38 and from 92% to 100% in uninfected children, 21,32,36 (Supplement Table S7-S9).…”