Background:  Vascular endothelial dysfunction may contribute to the increase in cardiovascular events in HIV-1-infected patients. We hypothesize that HIV replication is a cardiovascular risk factor and aimed at assessing markers of endothelial activation in patients with and without HAART, and to correlate these factors with HIV RNA replication.

Method:  We analyzed ARV-naïve patients included in STACCATO trial. Patients were treated for at least 6 months until HIV RNA was undetectable and CD4 count >350 cells/mm³. They were randomized to CD4-guided treatment interruption, or to continuous treatment. Samples were available before treatment, at randomization, weeks 12 or 24, and again after 3 months' re-treatment in the treatment interruption group. P-selectin, leptin, adiponectin, granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin (IL) -10, IL-6, soluble vascular cell adhesion molecule (s-VCAM), MIP1-alpha, and monocyte chemotactic protein (MCP) -1 plasma levels were assessed by treatment arm (Student's t test). We estimated the differences of means in metabolic markers according to HIV RNA strata (one way ANOVA). Linear regression was used to define the correlation between HIV-RNA and markers of interest.

Results:  We included 145 patients (62% female): 48 in the continuous treatment , and 97 in the  treatment interruption arm. Median baseline CD4 count was 270 cells/mm³ (IQR 232 to 328) and median log₁₀ HIV RNA was 4.7 (IQR 4.2 to 5.15) before ART. Patients then received HAART including ritonavir-boosted saquinavir. At week 24, 59 of 98 patients (60%) were still on treatment interruption. There was a significant difference in means between the continuous treatment and the treatment interruption arm with regards to s-VCAM (1.8 ng/mL [0.7] and 2.2 ng/mL [0.8], p = 0.012), adiponectin (5.3 μg/mL [3.3] and 3.9 μg/mL [2.1], p = 0.007), MCP1 (27.5 pg/mL [39.4] and 95.8 pg/mL [116.5], p <0.001), and IL-10 (15.84 pg/mL [23.1] and 3.9 pg/mL [10.6], p = 0.002). We also found a strong correlation between HIV RNA and plasma levels for s-VCAM (b+0.17, p = 0.001), adiponectin (b–0.41, p = 0.017), MCP-1 (b+20.48, p = 0.005), and IL-10 (b-2.7, p = 0.018). After final re-treatment in the treatment interruption arm, we observed a 30% decrease for adiponectin (p <0.001), a 76% decrease for IL-10 (p = 0.03), and a 55% increase for s-VCAM (p = 0.002).

Conclusions: There are significant differences in endothelial activation markers between patients on HAART, and patients without HAART. Acute viral replication occurring after treatment interruption is related to a change in key markers of cardiovascular risk. These changes were at least partly reversible at time of re-treatment.