Immune activation and HIV-specific T cell responses are modulated by a cyclooxygenase-2 inhibitor in untreated HIV-infected individuals: An exploratory clinical trial

Abstract

Pathologically elevated immune activation and inflammation contribute to HIV disease progression and immunodeficiency, potentially mediated by elevated levels of prostaglandin E2, which suppress HIV-specific T cell responses. We have previously shown that a high dose of the cyclooxygenase-2 inhibitor celecoxib can reduce HIV-associated immune activation and improve IgG responses to T cell-dependent vaccines. In this follow-up study, we included 56 HIV-infected adults, 28 antiretroviral therapy (ART)-naïve and 28 on ART with undetectable plasma viremia but CD4 counts below 500 cells/μL. Patients in each of the two study groups were randomized to receive 90 mg qd of the cyclooxygenase-2 inhibitor etoricoxib for six months, two weeks or to a control arm, respectively. T cell activation status, HIV Gag-specific T cell responses and plasma inflammatory markers, tryptophan metabolism and thrombin generation were analyzed at baseline and after four months. In addition, patients received tetanus toxoid, conjugated pneumococcal and seasonal influenza vaccines, to which IgG responses were determined after four weeks. In ART-naïve patients, etoricoxib reduced the density of the activation marker CD38 in multiple CD8+ T cell subsets, improved Gag-specific T cell responses, and reduced in vitro plasma thrombin generation, while no effects were seen on plasma markers of inflammation or tryptophan metabolism. No significant immunological effects of etoricoxib were observed in ART-treated patients. Patients receiving long-term etoricoxib treatment had poorer tetanus toxoid and conjugated pneumococcal vaccine responses than those receiving short-course etoricoxib. Cyclooxygenase-2 inhibitors may attenuate harmful immune activation in HIV-infected patients without access to ART.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1. CONSORT flow chart.

Fig 2. Study overview.

COX-2i treatment duration indicated in grey. S, blood sampling; V, vaccination.

Fig 3. Baseline CD4+ T cell count and CD4/CD8 ratio.

P-values derived from comparison of all included patients. For analysis of immunological effects of COX-2i after four months, arms 2 and 3 were combined (denoted Long-term COX-2i -), and patients with a CD4/CD8 ratio >0.6 (light grey color) were excluded. Median and lower/upper quartile indicated.

Fig 4. Change (Δ) in immunological variables from baseline to four months.

ART use and long-term COX-2i treatment indicated. Data given as median and lower/upper quartile. A: T cell activation, B: Gag-induced CD8 T cell responses, C: Plasma inflammatory markers, D: Tryptophan metabolism, E: Coagulation.

Fig 5. Polyfunctional responses of CD8+ T cells after stimulation with an HIV Gag peptide pool.

A: Two prevalent polyfunctional subsets (CD107a+MIP1β+IFN-γ+ and CD107a+MIP1β+) increased after four months of etoricoxib in ART-naïve patients. ART and COX-2i use indicated, coloured bar pairs represent response at baseline and four months, respectively. B: The CD8+ T cell response tended to be more polyfunctional after four months of etoricoxib in ART-naïve patients (2 leftmost pies). P-values derived from unpaired permutation tests.

Fig 6. Vaccine IgG antibody responses.

Fold change (FC) of tetanus toxoid and conjugated pneumococcal vaccine IgG levels in serum from baseline to 4 weeks post-vaccination, according to ART use and COX-2i treatment arm. All patients vaccinated in the study were included the analysis. Median and upper quartile indicated.