Background/Purpose: We aimed to evaluate the performance of QFT-GIT and T-SPOT.TB assays to predict the risk of latent tuberculosis infection (LTBI) in Japanese rheumatoid arthritis (RA) patients, measured simultaneously total and CD4-positive lymphocytes. When total or CD4-positive lymphocytes are suspected to be low due to aggressive treatment of RA, it remains unclear whether interferon-γ release assays (IGRAs) can be used reliably without increasing indeterminate results, and are affected by peripheral lymphocyte and CD4-positive cell counts

Methods: Because of no gold standard for LTBI, we compared with or without previous pulmonary tuberculosis infection as an alternative diagnosis using chest CT. We defined a history of TB infection (past TB infection group) as chest CT findings with old pulmonary TB infection. As a comparable control (non-TB infection group), we included a group of patients without prior contact with active TB, or old TB infection on chest CT. Sixty-eight patients treated with MTX and/or biologics were divided into two groups: 33 with and 35 without a past TB infection. MTX dose was 9.2 mg in those with a past TB infection and 11.0 mg in those without. Biologics was received 41% in the past TB and 57% in non-TB infection group. Mean total and CD4-positive lymphocyte in the past TB infection group decreased to 1,091/μL and 491/μL, respectively.

Results: The sensitivity and specificity of QFT-GIT for discriminating past TB infection were 21.2% and 100%, respectively. With a lower cutoff of 0.1 IU/mL (gray zone), the sensitivity and specificity of QFT-GIT were 30.3% and 96.9%, respectively. This gray zone of QFT-GIT requires further investigation to estimate the risk of TB. Positive T-SPOT.TB (SFC ≥6) were found in 21.9% in the past TB infection group, and 15.6% had SFC ≥8. With the positive cutoff (SCF ≥6) of T-SPOT.TB, the sensitivity and specificity were 21.9% and 100%. In the past TB infection group, 5 patients had positive QFT-GIT (≥0.35 IU/mL) and T-SPOT.TB (≥6 spots). The results for 4 patients with past TB were negative with T-SPOT.TB, but of these 4 patients, 3 in positive and 1 in the gray zone for QFT-GIT. The overall agreement of two IGRAs was high. Results were indeterminate in 4 (5.9%) of 68 patients, due to 3 patients decrease positive control of QFT-GIT and 1 increased negative control of T-SPOT.TB. QFT-GIT yielded results in 43% with low lymphocyte (<1000/µL) and 47% with low CD4-positive lymphocyte (<500/µL). In both IGRAs, PHA mitogen responses were not different between those treated with or without biologics. The positive rates of QFT-GIT and T-SPOT.TB decreased upon stimulation with TB antigens according to total and CD4-positive lymphocyte counts: this effect was more notable in QFT-GIT than T-SPOT.TB. When total lymphocytes (<1,000/μL) and CD4-positive lymphocytes (<500/μL) were low, the positive rates of QFT-GIT and T-SPOT.TB were low. Even where total and CD4-positive lymphocyte counts were low, the positive rate of QFT-GIT was increased when the gray zone range was included.

Conclusion: Two IGRAs had high specificities, but may falsely identify past TB infection owing to low sensitivities. Despite low total and CD4-positive lymphocyte counts, IGRAs could be utilized without high indeterminate rates.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/detection-of-previous-tuberculosis-infections-in-japanese-rheumatoid-arthritis-patients-comparison-of-two-interferon-g-releasing-assays-and-the-impact-of-cd4-positive-lymphocytes/