Background:  HIV-infected patients may be at higher risk for cardiovascular events when not receiving ART, perhaps as a result of increased systemic inflammation. In a previous pilot study, we demonstrated that the NF-κβ inhibitor salsalate significantly improved endothelial function, a precursor to atherosclerosis, but was associated with treatment-limiting hepatotoxicity. Therefore, we hypothesized that inhibition of a different anti-inflammatory pathway, namely reducing tumor necrosis factor (TNF) production using the phosphodiesterase inhibitor pentoxifylline (PTX), would also improve endothelial dysfunction in HIV-infected subjects not receiving ART.

Methods:  An open-label, prospective single-arm study of PTX 400 mg 3 times daily was performed in 9 HIV-infected adults who had not received ART for ≥6 months and had CD4 ≥50/µL. Exclusion criteria included known vascular or pro-inflammatory conditions, diabetes, hypertension, or receipt of lipid-lowering, anti-inflammatory, or ARV drugs during the study. The Wilcoxon Signed-Rank test was used to evaluate changes in flow-mediated dilation (FMD) of the brachial artery at 4 and 8 weeks (primary objectives) and changes in metabolic, inflammatory, and endothelial activation markers at 8 weeks (secondary objectives).

Results:  Median age, CD4 count, and viral load of the subjects were 40 years, 552 cells/µL, and 35,300 copies/mL, respectively; 33% were black and 67% were men. Because of initiation of an inhaled steroid and ART, 2 subjects were removed from study at weeks 2 and 5, respectively. Median baseline FMD was low at 2.1% (n = 9; range 1.3 to 5.1%). FMD significantly improved after 4 weeks (n = 8; median absolute increase 1.5%, range –0.7 to 4.9%; p = 0.04] and improved further at 8 weeks (n = 7; median absolute increase 4.4%, range 2.8 to 6.5%; p = 0.02). There were no notable drug-related toxicities. At week 8, median triglycerides (p = 0.09), vascular cell adhesion molecule (VCAM) (p = 0.05), and CXCL10 (IP-10) (p = 0.02) decreased from baseline. No significant changes occurred in viral load, homeostasis model assessment insulin resistance (HOMA-IR), high-sensitivity C-reactive protein (hsCRP), interleukin-6 (IL-6), monocyte chemotactic protein-1 (MCP-1), anti-intercellular adhesion molecule (ICAM), or von Willenbrand factor (vWF).

Conclusions:  In this pilot study, PTX safely improved endothelial function in HIV-infected patients not receiving ART, supporting the concept that systemic inflammation may contribute to endothelial dysfunction and cardiovascular events in HIV-infected patients. Further study with PTX is warranted.