332    

Background: Since interleukin(IL)-18 is critical for generating the Th1 response in vivo and IL-18 inhibits HIV production in infected peripheral blood mononuclear cells in vitro, IL-18 may participate in HIV control.  However, retrospective cross sectional studies demonstrated a direct association between serum IL-18 levels and clinical stage of HIV disease, raising the possibility that IL-18 enhances HIV production in infected persons.  Thus the role of IL-18 in HIV production in vivo has not been fully delineated.

Methods: We conducted a prospective phase II clinical trial in 28 men with acute or early HIV infection with a median age of 32.5 (21-59) yrs and a median follow-up of 104 weeks.  Enrollees had HIV RNA or p24 antigen in blood without anti-HIV antibodies (negative ELISA or ≤2 bands on Western blot), or had a clinical syndrome of acute HIV infection and a negative HIV antibody test in the past 12 weeks (acute infection), or had seroconverted to HIV in the prior year (early infection). Twenty-one of the 28 patients were on ART (7 acute, 14 early) and 7 were not on ART (all early).  At weeks 0, 4, 12, 24, 48, 96, and 104, heparinized whole blood was collected and lysed using Triton-X-100.  Whole blood IL-18 concentrations were measured in blinded samples using an ELISA that measures biologically active IL-18.  IL-18 concentrations were assessed for longitudinal association with the prospectively defined outcomes of viral load, CD4 count, and CXCR4 expression on CD4+ T cells determined by flow cytometry. 

Results:  Using a longitudinal mixed-effects model that controlled for ART use and week and duration of ART, for every log increase in IL-18 we observed a ‑0.48 (95% CI: ‑0.96, 0, p=0.052) change in log viral load, and a ‑659 (CI: ‑888, ‑430, P<0.0001) change in CXCR4 molecules per CD4+ cell.

Conclusions:  This is the first prospective study assessing whole blood IL-18 in persons infected with HIV.  After controlling for ART use and week and duration of ART, increased circulating IL-18 predicted a reduced plasma HIV RNA level, a finding consistent with an antiretroviral IL-18 effect.  Increased IL-18 also predicted reduced CXCR4 co-receptor expression on CD4+ T cells, suggesting that IL-18 may block HIV entry into cells by constraining receptor availability.  Further studies of the role of IL-18 in the pathogenesis of HIV disease and of possible therapeutic use of IL-18 are warranted.

Keywords: IL-18; viral load; CXCR4