Abstract Body

People with HIV (PWH) on antiretroviral therapy (ART) are at increased risk of comorbidities, including cardiovascular disease (CVD), with underlying inflammation thought to contribute to excess, unmeasured risk.

The UCD Infectious Diseases Cohort Study recruited PWH on effective ART and analysed 27 biomarkers of systemic, innate and vascular inflammation by chemiluminescence immunoassays and 7 T-cell markers by flow cytometry (CD4+ and CD8+ T-cell senescence, activation, exhaustion and T-regulatory cells). We used principal component analyses and unsupervised hierarchical clustering to partition participants into biomarker-derived clusters and explored associations between clusters and 1) multimorbidity (2 or more of CVD, hypertension, dyslipidaemia, malignancy, osteoporosis, diabetes, kidney, liver, respiratory or psychiatric disease); and 2) CVD, with multivariate models including variables associated on univariate analysis (p<0.1).

277 PWH were included in the analysis [median (IQR) age 44 (39, 50) years, 57.4% male, 45% caucasian, 25.1% current smoker]. Three biomarker derived clusters were identified: cluster 1 (n=148, 53.4%) characterized by lower systemic (TNFa, TNFR1, TNFR2, IFN-gamma, IL-6, IL-1b), vascular (P-selectin, E-selectin, VCAM, ICAM-1) and innate (sCD14, MCP-1) inflammation, cluster 2 (n=100, 36.1%) characterized by higher values of vascular and innate inflammation and cluster 3 (n=29, 10.5%) characterised by high values of markers of T-cell activation and proliferation (IL-1b, TNFa, IL-6, IFN-gamma, IL-10, IL-12, IL-2, IL-4) and microbial translocation (IFABP). There was no significant differences in demographics, HIV or immunological indices between clusters. 77 (27.8%) individuals reported a history of multimorbidity, 13 (5%) with CVD. On univariate analysis both cluster 2 and 3 were associated with multimorbidity compared to non-inflamed cluster 1 (see figure 1b). Cluster 3 was significantly associated with CVD (figure 1c), an association that persisted after adjustment for age, gender, cholesterol and smoking [adjusted odds ratio 7.07 (95% confidence interval 1.28, 38.9), p=0.025].

In PWH on ART, individuals can be partitioned into clusters characterised by inflammatory patterns, with clusters of higher inflammation associated with multimorbidity and specifically CVD. Further studies are warranted to examine the utility of bio-profiling populations to identify individuals at high risk of developing CVD for targeted interventions.