Diagnostic performance of an enzyme-linked immunospot assay for interferon-γ in skeletal tuberculosis.

The aim of this study was to evaluate the diagnostic performance of an enzyme-linked immunospot (ELISPOT) assay for interferon-γ in patients with suspected skeletal tuberculosis (TB). From March 2007 to June 2010, a total of 36 patients with suspected skeletal TB in a tertiary care hospital in Taiwan were enrolled. Twelve patients (35.3%) had culture-confirmed TB, three (8.8%) patients had probable TB, and the remaining 21 (58.3%) patients did not have TB. Fourteen patients with mycobacterial infection had available biopsy or surgical specimens for histopathological examination and 12 (85.7%) specimens had pathological features consistent with mycobacterial infection. Among the 12 patients with positive findings indicating mycobacterial infection, all seven patients with spinal TB and three of five patients with TB arthritis had positive ELISPOT assays. All nine patients with spinal TB had positive ELISPOT assays, but only four of six patients with TB arthritis had positive ELISPOT assays. The sensitivity, specificity, positive predictive value, and negative predictive value for skeletal TB diagnosis by the ELISPOT assay were 86.7%, 61.9%, 61.9%, and 86.7%, respectively. In conclusion, the ELISPOT assay can provide useful support in diagnosing skeletal TB, and spinal TB can be excluded based on a negative ELISPOT assay.