Background:  HIV/TB-associated immune reconstitution inflammatory syndrome (TB-IRIS) is an immunopathological reaction to mycobacterial antigens induced by ART. We previously demonstrated in a randomized placebo-controlled trial that a 4-week course of prednisone (1.5 mg/kg/day for 2 weeks followed by 0.75 mg/kg/day for 2 weeks) reduces morbidity and results in more rapid symptom improvement in TB-IRIS, but the mechanisms are unclear. We aimed to determine the effect of prednisone on inflammatory response (antigen-specific T cells, cytokines and chemokines) in TB-IRIS.

Methods:  Samples from participants in the randomized placebo-controlled trial of prednisone were taken at 0, 2, and 4 weeks and were analyzed using:  ELISpot (antigen stimuli PPD, ESAT-6, Acr1, Acr2, 38kDa, and heat killed H37Rv Mycobacterium tuberculosis), rt-PCR of supernatants following heat-killed H37Rv M. tuberculosis (hkH37RvMTB) stimulation (for genes of selected cytokines and chemokines) and Luminex multi-analyte profiling of the same supernatants and unstimulated serum (for concentrations of selected cytokines and chemokines). Flow cytometry on hkH37RvMTB-stimulated whole blood was performed on samples from a separate cohort (n = 12) of TB-IRIS cases and controls to establish the cellular source of interleukin (IL)-6, IL-12p40, and tumor necrosis factor (TNF)-a. Student’s unpaired and paired t-tests and the paired Wilcoxon signed-rank test were used as appropriate. To factor multiple comparisons, the Bonferroni correction was applied.

Results:  We included 58 trial participants with TB-IRIS (31 prednisone, 27 placebo). In serum, significant decreases in IL-6, IL-10, IL-12p40, TNF-a, and IP-10 concentrations during prednisone, but not placebo, treatment were observed at weeks 2 and 4. A significant decrease in interferon-g (IFN-g) concentrations only occurred in prednisone-treated participants at week 4. No differences in ELISpot responses comparing prednisone and placebo groups were shown in response to ESAT-6, Acr1, Acr2, 38kDa, or heat-killed H37RvMTB, but PPD responses increased over 4 weeks in the prednisone group and decreased in placebo group (p = 0.007). In further experiments, neutrophils were found to be an important source of IL-12p40 and TNF-a with increased production in TB-IRIS.

Conclusions:  The beneficial effects of prednisone in TB-IRIS appear mediated via suppression of predominantly innate pro-inflammatory cytokine responses, not via a reduction of the numbers of antigen-specific T cells.