Within-subject variability and boosting of T-cell interferon-gamma responses after tuberculin skin testing

Rationale: The optimal strategy for the diagnosis of latent tuberculosis infection is controversial. Adoption of a two-step strategy (tuberculin skin test [TST] followed by an IFN-g release assay [IGRA], compared with an IGRA alone), may be limited by TST-mediated boosting of subsequent IGRA responses. Assessment of withinsubject IGRA variability will aid in establishing thresholds for conversions and reversions, and interpretation of serial testing results. Objectives: To determine short-term IGRA variability and the impact of TST on subsequent IGRA results. Methods: Within-subject variability and TST-mediated boosting of IGRA responses were evaluated in 26 South African participants with varying exposure risk. IGRAs (T-SPOT.TB, QuantiFERON-TB Gold In- Tube [QuantiFERON-TB-GIT], PPD, and heparin-binding hemagglutinin) were repeated four times over 21 days pre-TST, and on Days 3, 7, 28, and 84 post-TST administration. Measurements and Main Results: All participants showed withinsubject IGRA variability. Changes of 63 spots (T-SPOT.TB) or 680% from the mean IFN-g response (QuantiFERON-TB-GIT) over 3 weeks explained 95% of the variability. Spontaneous conversions/ reversions occurred in 7 of 26 subjects (27%) (6 for T-SPOT.TB and 1 for QuantiFERON-TB-GIT [P 5 0.049]) during the within-patient variability studies (pre-TST). After the TST eight subjects (33%) boosted above the defined baseline variability. By Day 7 post-TST, but not Day 3, 2 (12.5%) initially IGRA-negative test subjects converted. By contrast, boosting of PPD and heparin-binding hemagglutinin occurred by Day 3 post-TST. Conclusions: When using a two-step screening strategy it appears safe to perform a QuantiFERON-TB-GIT or T-SPOT.TB IGRA within 3 days of performing the TST. A 3-spot or 80% IFN-g response variation,oneither side of baseline values, explains95%of the shortterm variability and may be useful for interpreting conversions and reversions, and values close to the cut-point.