HIV and diabetes impair M. tuberculosis-specific interferon-gamma responses on QuantiFERON-TB Gold Plus testing

declared


HIV and diabetes impair M. tuberculosis-specific interferon-gamma responses on QuantiFERON-TB Gold Plus testing
Dear Editor, TB is the leading cause of death among people with HIV (PWH). 1 One-quarter of the global population is infected with TB (TBI), of whom 5-10% will develop active TB, a risk increased 20-fold by HIV co-infection. 1,2Interferon-gamma (IFN-c) release assays (IGRAs) are the most predictive tests for TBI progression. 35][6] The QuantiFERON-Plus (QFT-Plus) IGRA, which measures IFN-c production to TB antigen stimulation in CD4þ and CD4þ/CD8þ specific tubes, was developed to improve sensitivity including among PWH. 7We evaluated IFN-c response to TB antigen stimulation measured by QFT-Plus among Kenyan adults with and without HIV.We hypothesized PWH would have reduced Mtb-specific IFN-c responses to QFT-Plus and that CD4þ count and viral load would negatively influence IFN-c response.This is a cross-sectional sub-study in a previously described cohort of adults evaluated for cardiovascular and pulmonary disease in Kisumu, Kenya. 8lood was collected for QFT-Plus testing for Mycobacterium tuberculosis (MTb) between December 2018 to December 2019.We enrolled a convenience sample of PWH (�30-years old) from an HIV clinic and sex-matched people without HIV (PWoH) from HIV testing centers.PWH were on antiretroviral therapy (ART) for at least 6 months.Exclusion criteria included pregnancy and self-reported acute respiratory infections (including active TB symptoms).Questionnaires were performed at enrollment.Diabetes mellitus (DM) diagnosis was self-reported.For PWH, HIV RNA viral load (VL) and CD4 counts were obtained in the parent study or abstracted from medical records.Prior TB diagnosis and isoniazid prevention therapy (IPT) were self-reported and confirmed by chart review where possible.QFT-Plus testing was performed at enrollment per the manufacturer's (Qiagen; Hilden, Germany) protocol. 7The QFT-Plus measures IFN-c response (IU/ml) in a nil tube (negative control, measuring endogenous IFN-c), a mitogen tube (positive control, measuring IFN-c response to a non-specific T-cell stimulator), and two TB-specific tubes (TB1 and TB2) which elicit CD4þ and CD8þ T-cell responses to Mtb-specific antigens.The test is positive if the TB1 and/or TB2 IFN-c response minus nil is �0.35IU/ml.Indeterminate results are due to high nil or low mitogen response. 7We defined borderline results as 0.20-0.70IU/ml. 9We compared prevalence of TBI (positive QFT-Plus), absolute IFN-c levels (TB1 and TB2 minus nil), borderline, and indeterminate results between PWH and PWoH.Wilcoxon rank-sum test was used to compare medians.Univariate and multivariable linear regression analyses were used to evaluate factors (selected a priori) associated with quantitative TB-specific IFN-c response.Ethical approval was obtained from the University of Washington Institutional Review Board and the Kenyatta National Hospital/University of Nairobi Ethical and Scientific Review Committee.Written informed consent was obtained from all study participants.
Our analyses demonstrate diminished Mtb-specific IFN-c response among PWH using QFT-Plus testing.Lower CD4 count was associated with lower IFN-c response to both TB antigens.Overall test positivity was lower among PWH, suggesting reduced sensitivity.In this cross-sectional study we cannot definitively establish this difference in sensitivity.TBI prevalence was high among PWoH; this high prevalence was similar to prior surveys among healthcare workers and school workers in Kisumu. 10Prior IPT use did not modulate IFN-c response among PWH.For participants with CD4,200 cells/mm 3 , we did not find increased test sensitivity with a lower cut-off value at 0.2 IU/ml, however analyses were limited by the small number of participants with severe immunosuppression (n ¼ 18).The present study supports the hypothesis that among individuals with controlled HIV on ART, HIV infection is associated with diminished Mtb-specific IFN-c response with QFT-Plus testing, as with prior iterations of IGRA, and greater immunosuppression is associated with further reduction in this response.Interestingly, we also found lower IFN-c response among PWoH with DM.Interestingly, this effect was not observed among PWH, potentially due to the already muted IFN-c response associated with positive HIV status and the small number of participants with DM in this cohort.2][13] The large proportion of participants with borderline results reiterate previous concerns on test specificity. 9,14We did not evaluate immunosuppression other than with HIV and DM.Further study is needed to determine the optimal operating parameters for IGRA testing among patients with HIV, DM, and other immunologically significant comorbidities.
Our analyses demonstrate substantially diminished IFN-c response among PWH in response to QFT-Plus testing, particularly those with lower CD4 counts, as well as PWoH with DM.Additionally, many participants had borderline results.Our study brings into question the appropriateness of current cut-offs for the QFT-Plus, particularly in PWH and those with comorbidities impacting immune response.These findings suggest that TBI is likely underdiagnosed in PWH and people with DM despite these groups being  Letter

*
Entire cohort TB1-nil (95% CI)P-valueEntire cohort TB2-nil (95% CI) P-value PWH TB1-nil (95% CI) P-value PWH TB2-nil (95% CI) Linear regression of cofactors associated with TB1 or TB2 minus nil; adjusting for listed covariates; limited to participants with valid QFT-Plus tests, excluded 3 PWH (2 high nil and 1 low mitogen) and 6 people without HIV (all high nil).† Per 100 cells/mm 3 CD4þ T-cell count increase.‡ Detectable viral load as a yes/no variable, detectable is .20 copies/mL.§ In sensitivity analyses, we included IPT in analyses for the entire cohort and it was not significantly associated with interferon-c response.Mtb ¼ Mycobacterium tuberculosis; CI ¼ confidence interval; PWH ¼ people with HIV; IPT ¼ isoniazid prevention therapy.

Table .
Multivariable* analyses of factors associated with Mtb-specific interferon-c response levels.