A diagnostic algorithm for tuberculous pleurisy using the ELISPOT assay on peripheral blood and pleural effusion.

BACKGROUND

Diagnosis of tuberculous (TB) pleurisy remains challenging due to the paucibacillary nature of the disease. We prospectively assessed the diagnostic usefulness of the T-cell based ELISPOT assay, and created a clinical algorithm for differentiating TB pleurisy from other diagnoses.

METHODS

All adult patients with suspicion for TB pleurisy were enrolled in a tertiary hospital in Seoul, South Korea, over a 7-year period. ELISPOT assays were performed using mononuclear cells from peripheral blood and pleural effusion.

RESULTS

Seventy-seven patients with suspected TB pleurisy were enrolled. Of these, 33 (43%) patients, comprising 27 confirmed and 6 probable TB pleurisy, were classified as TB pleurisy, and 36 (47%) were classified as not TB. The remaining 8 with possible TB pleurisy were excluded from the final analysis. The sensitivities and specificities, respectively, of the diagnostic methods were as follows: pleural fluid adenosine deaminase (ADA) level 32 U/L, 81% and 79%; peripheral blood mononuclear cells (PBMC) ELISPOT assay, 82% and 73%; pleural effusion-mononuclear cells (PE-MC) ELISPOT assay, 58% and 87%. When the diagnostic algorithm was applied, PBMC ELISPOT ≥6 spots or ADA ≥32 U/L' as a rule-out test safely excluded 46% (12/26) of the not TB patients, and 'PE-MC ≥6 spots' as a rule-in test accurately classified 23% (7/31) of the patients with TB pleurisy.

CONCLUSIONS

A diagnostic algorithm combining ELISPOT assays and ADA levels in pleural fluid appears to be a promising and non-invasive approach for patients with suspected TB pleurisy.