In the Strategies for Management of AntiRetroviral Therapy (SMART) study, higher baseline levels of the inflammatory cytokine interleukin-6 (IL-6) and the thrombotic marker d-dimer were strongly associated with mortality risk [cardiovascular disease (CVD)- and non-CVD-related] . In that study, the authors speculated that activation of tissue factors and endothelial surfaces may be driving elevations in IL-6 and d-dimer levels. In this pilot study of persons with untreated HIV infection, we explore the relationship between IL-6 and d-dimer with the following measures of vascular dysfunction: large and small artery elasticity (LAE and SAE, respectively) and plasma markers of endothelial function [E-selectin and soluble intercellular adhesion molecule-1 (sICAM-1)].
HIV-infected participants had not received antiretroviral therapy in the previous year, and had no known CVD. Arterial elasticity was assessed via pulse waveform analysis of the diastolic decay curve (model HDI/PulseWave CR-2000; Hypertension Diagnostics Inc., Eagan, MN, USA). Biomarker levels were measured at the Laboratory for Clinical Biochemistry Research at the University of Vermont. Details of these methods and the study protocol have been reported previously [2,3].
To consider the combined influence of IL-6 and d-dimer, a joint mortality risk score was also generated according to the weighted contribution of each biomarker for risk of death in SMART . The joint mortality risk score was obtained from SMART data using a conditional logistic regression model that considered both IL-6 and d-dimer (each log10-transformed) for the outcome of all-cause mortality. The joint mortality risk score was calculated by solving for the logit formed with the estimated parameters from SMART and the log10-transformed values of IL-6 and d-dimer from the current study. Higher values of this score were associated with a higher risk of death in SMART. Data were analysed using R statistical software (version 2.8.1; http://www.cran.r-project.org).
Characteristics of the 32 HIV-infected participants who were enrolled have been previously reported [2,3]. Mean (standard deviation) age was 40 (9.6) years and body mass index was 26 (5.1) kg/m2. Twenty-eight participants (88%) were male, 19 (59%) were current smokers, 11 (34%) had hepatitis C virus coinfection, two (6%) had diabetes mellitus, and two (7%) had a prior AIDS clinical event. Mean CD4 count was 391 (182) cells/μL and mean HIV RNA level was 4.15 (0.73) log10 HIV-1 RNA copies/mL.
The median (interquartile range) values for IL-6, d-dimer, and the joint mortality risk score were 1.79 (1.34–4.88) pg/mL, 0.39 (0.19–0.60) μg/mL, and 0.47 (0.33–0.74), respectively. Mean values for each surrogate measure of vessel function (untransformed) and HIV RNA level (log10-transformed) are reported by quartile of IL-6 and d-dimer (Table 1). Higher levels of IL-6 (fourth vs. first quartile, and as a continuous variable in Spearman rank correlations) tended to be associated with impaired SAE and higher levels of sICAM-1 and E-selectin. A similar pattern was seen when comparing markers of vascular dysfunction with d-dimer levels. LAE and CD4 cell count (data not shown) did not vary by IL-6 or d-dimer level. For comparisons using the joint (IL-6/d-dimer) mortality risk score, the associations with markers of vascular dysfunction (SAE, sICAM-1 and E-selectin) became more pronounced.
|Surrogate marker||Quartile of IL-6||Comparison of quartiles (4th vs. 1st)||Spearman's rank correlation|
|1st||2nd||3rd||4th||Per cent difference*||P-value||R||P-value|
|LAE (mmHg/mL × 10)||16.5||15.7||18.2||15.1||−9.90||0.50||−0.14||0.45|
|SAE (mmHg/mL × 100)||8.2||6.3||6.7||5.6||−37.8||0.03||−0.33||0.06|
|HIV RNA (log10 copies/mL)||4.0||4.1||4.1||4.4||11.3||0.08||0.26||0.15|
|Quartile of d-dimer|
|LAE (mmHg/mL × 10)||15.6||16.1||17.6||16.2||2.1||0.90||0.18||0.34|
|SAE (mmHg/mL × 100)||8.0||6.7||6.5||5.7||−33.9||0.12||−0.27||0.13|
|HIV RNA (log10 copies/mL)||4.1||4.5||3.5||4.4||5.9||0.48||0.06||0.73|
|Quartile of joint (IL-6/d-dimer) mortality risk score|
|LAE (mmHg/mL × 10)||15.6||17.6||16.9||15.5||−2.1||0.89||−0.06||0.74|
|SAE (mmHg/mL × 100)||8.5||7.1||5.6||5.5||−41.5||0.02||−0.42||0.02|
|HIV RNA (log10 copies/mL)||4.0||4.1||4.1||4.3||7.1||0.33||0.25||0.17|
In summary, we have shown that higher IL-6 and d-dimer levels among persons with untreated HIV infection are associated with vascular dysfunction, indicated by higher endothelial biomarkers and impaired SAE – a marker of early vascular disease and future clinical risk. Findings from SMART suggest that non-AIDS-related mortality may be a consequence of greater inflammation (IL-6 levels) and thrombotic activity (d-dimer levels) in persons with HIV infection . Levels of IL-6 and d-dimer and estimates of artery elasticity (LAE and SAE) are being ascertained in a subset of participants in the ongoing Strategic Timing of Antiretroviral Therapy trial, and will provide valuable insight into the mechanisms contributing to early vascular disease in persons with HIV infection. Future research should consider the role of HIV-mediated endothelial injury as a contributor to both CVD- and non-CVD-related mortality in the current era.