The diagnostic performance of an interferon-gamma assay in HIV- and tuberculosis endemic population

Abstract number: P1946

Aabye M., Ravn P., PrayGod G., Jeremiah K., Mugomela A., Jepsen M., Faurholt D., Range N., Friis H., Changalucha J., Andersen A.B.

Objectives: Pulmonary tuberculosis (PTB) in HIV positive patients is often microscopy negative and the Mycobacterium tuberculosis specific Interferon Gamma Release Assays (IGRAs) may be useful in this situation. The performance of IGRAs however, has not been sufficiently documented in tuberculosis- and HIV-endemic settings. We conducted a cross-sectional study in Mwanza, northern Tanzania in order to evaluate the sensitivity of the QuantiFERON TB-Gold In-Tube (QFT-G IT) test for diagnosing pulmonary tuberculosis disease and evaluate the degree to which the test is affected by HIV-status and CD4 cell count.

Methods: 287 patients diagnosed with pulmonary tuberculosis were included in the study and were subjected to control sputum microscopy and culture, HIV- and QFT-G IT testing and measurement of CD4 cell count.

Results: Among 193 patients with culture or microscopy positive TB, sensitivity of the QFT-G IT was 71% (95% CI: 65–77%) and the test was positive in a larger proportion of sputum positive than sputum negative subjects (72% vs. 48%, p < 0.0001). Sensitivity of QFT-IT was higher in HIV uninfected than in infected subjects (78% vs. 62%, p < 0.01).

56 out of 287 subjects (19.5%) had an indeterminate response due to low mitogen response. The proportion of subjects with an indeterminate QFT-G IT result was larger in those with a CD4 cell count below compared to above 200×10⁶/l (38.1% vs. 14.3%, p < 0.0001). No other differences were found between patients with determinate and indeterminate responses.

In the 157 subjects with a CD4 cell count above 200×10⁶/l, no difference in sensitivity was found between HIV-uninfected and -infected subjects (76% vs. 64%, p < 0.11), and the overall sensitivity was not higher (71% vs. 72%, p < 0.84). Conversely, the sensitivity was higher when correcting for QFT-G IT indeterminate results (71% vs. 84%, p < 0.005).

Conclusion: The sensitivity of the QFT-G IT in this study was similar to previous reports of sputum microscopy sensitivity. The test was affected by low CD4 cell count however this did not explain all the indeterminate results, nor did it affect the overall sensitivity. Other factors than the absolute number of circulating CD4 positive lymphocytes is suggested to influence the performance of the test.