Background/Purpose: Tuberculosis (TB) has been reported to occur at a higher rate in SLE patients than in the general population.  As a result, most clinical trials of new therapeutic agents in SLE seek to exclude patients with untreated latent or active TB. Unfortunately, there is no gold standard test for the detection of latent TB infection (LTBI).  Interferon gamma release assays (IGRAs) are potentially confounded by underlying immunologic impairments, e.g., lymphopenia, immunosuppressant use and disease activity, which can lead to indeterminate (IND) results even in the presence of LTBI. There can also be technical error impacting the results.  A common resolution to resolve IND results involves repeat testing. In this post hoc analysis, we sought to evaluate the utility of the QuantiFERON-TB GOLD test (QFT) and repeat testing for IND results in the screening of patients entering an SLE clinical trial.

Methods: Patients with active SLE who had signed informed consent to enter the Phase IIb ADDRESS II study (NCT01972568) with no prior history of TB infection were evaluated for LTBI either locally via IGRA or TB skin test; or QFT centrally.  Data for those who received centralized QFT were evaluated. Patients with a positive (POS) QFT result were excluded from trial participation.  Patients with an IND QFT result were to be retested: only if negative (NEG) on retest was a patient eligible for inclusion. 

Results: Ninety-nine patients from sites in Bulgaria, Chile, Germany, Poland, Russia and the United Kingdom received QFT centrally.  Of these, 84 (84.8%) had a NEG QFT, 8 (8.1%) were POS and 7 (7.1%) were IND on the initial test (Table). Nine patients with an initial NEG result underwent repeat testing for various reasons, and all remained NEG. Of the 7 patients who were initially IND: 2 were IND on repeat testing; 1 had no repeat result since screen failed; 1 was NEG based on a repeat QFT performed locally, and 3 were NEG on repeat QFT centrally.  Three of the 4 patients NEG on repeat QFT were enrolled. Even if meeting exclusion criterion for LTBI, patients were often screen failed for another reason, minimizing the impact of the exclusion criterion on overall enrollment.  No patients developed TB during the study. In those patients who had IND results, the absolute lymphocyte count was numerically lower with a higher percentage of patients having lymphocyte counts below the lower limit of normal, than in those with POS or NEG results (p-values by t-test > 0.05).

Conclusion: In our study, repeat IND QFT results for TB occurred infrequently. Consistent with previous reports, low lymphocyte counts may contribute to generation of IND results. Limitations of evaluating only screening data include not being able to evaluate the impact of disease activity or immunosuppressants on the QFT results. As only 2 of 99 patients had IND results twice, excluding patients on this basis should be considered a viable approach to confirm eligibility for a clinical trial in SLE.