Comparison of Three Cellular Immunoassays to Detect Tuberculosis Infection in 876 Healthy Recruits

The currently purified protein derivative (PPD) skin test and 2 interferon (IFN)-γ release assays (IGRAs) were usually used to detect Mycobacterium tuberculosis infection. We try to evaluate the performance of these methods to detect latent tuberculosis infection (LTBI) in this study. Each subject of the 876 recruits (19.05 ± 1.55, 17-24) underwent the PPD test, enzyme-linked immunospot (ELISPOT) assay, and chemiluminescent enzyme immunoassay (CLEIA). The prevalence of LTBI among the participants, as estimated by PPD, ELISPOT, and CLEIA, was 49.89% (437/876), 25.34% (222/876), and 28.77% (252/876), respectively. Of the participants, positive results were noted in 12.79% (112/876) for both ELISPOT and PPD, 19.52% (171/876) for both CLEIA and PPD; 9.82% (86/876) for 2 IGRAs; and 6.62% (58/876) for all 3 methods. Overall, the consistency among the 3 tests was 36.99% (324/876). ELISPOT-positive rate (41.38%) in the recruits with a PPD result ≥20 mm was higher than PPD <20 mm (24.76%; P < 0.05). Increased PPD skin reactions were associated with significantly increased CLEIA-positive rates and IFN-γ levels. Of 307 recruits without the bacillus Calmette-Guérin (BCG) vaccination, 2 IGRA (42.19%)-positive rates in the PPD-positive group were significantly higher than those in the PPD-negative group (28.40% and 23.05%; P < 0.05 and P < 0.01, respectively).There was low correlation and poor consistency among 2 IGRAs and PPD in healthy recruits, but IGRAs may be more accurate screening methods for TB infection in the countries with BCG vaccination.