The aim of this study was to evaluate the diagnostic performance of an
enzyme-linked immunospot (ELISPOT) assay (T-SPOT.TB; Oxford Immunotec, Oxford, UK) for
interferon-γ in patients with suspected
cutaneous tuberculosis (TB). From March 2007 to June 2010, a total of 45 patients with suspected cutaneous TB were enrolled. Data on clinical characteristics of the patients and conventional laboratory results were collected, and blood samples were obtained for ELISPOT assay. Ten subjects (22.2%) had culture-confirmed TB, 2 (4.4%) subjects had probable TB, and the remaining 33 (73.3%) subjects did not have TB. Twenty-one patients with mycobacterial
infection had available biopsy or surgical specimens for histopathologic examination and 16 (76.2%) specimens had pathologic features (granulomatous
inflammation, caseating
necrosis, or
acid-fast bacilli) consistent with mycobacterial
infection. Among these 16 patients, all 6 patients with TB and 3 patients with
Mycobacterium marinum infection had positive ELISPOT results, but none of the patients with Mycobacterium abscessus, Mycobacterium chelonae, or Mycobacterium avium
infection had positive ELISPOT results. Overall, 9 of 10 patients with culture-confirmed TB and both patients with probable TB had positive ELISPOT assay. For patients with nontuberculous mycobacteria
infections, ELISPOT assay was positive only in patients with M. marinum and M. kansasii
infection. The sensitivity, specificity, positive predictive value, and negative predictive value for cutaneous TB diagnosis by the ELISPOT assay were 91.6% (95% CI, 64.6-98.5%), 75.8% (95% CI, 59.0-87.2%), 57.9% (95% CI, 34.0-78.9%), and 96.2% (95% CI, 59.0-87.2%), respectively. In conclusion, ELISPOT assay can provide useful support in the diagnosis of cutaneous TB.