Although inflammation and immune dysfunction are implicated in pathogenesis of comorbidities in treated people with HIV (PWH), whether an immune risk profile can predict PWH at higher risk of comorbidities is unclear.
In the UCD Infectious Diseases cohort study of PWH on anti-retroviral therapy, we measured 24 biomarkers using bead based quantitative ELISA, covering pathways of systemic inflammation (hsCRP, IL6, TNFR1,2, TNFa), innate immune activation (sCD14, sCD163, MCP1, MIP1, sCD40), endothelial function (Pselectin, Eselectin, sVCAM, sICAM), coagulation (Ddimer, vWF) and intestinal permeability (IL18, LBP). Principal component analysis was performed followed by unsupervised hierarchical clustering to partition subjects into biomarker derived clusters. Logistic regression assessed association between clusters and prevalent comorbidities (CVD, kidney, liver, hypertension, dyslipidemia). Data are median[IQR] or odds ratio (OR) [95%CI].
We included 99 PWH, age 41 (36,48) years; 44% male; 54% African; 93% with HIVVL<40cps/ml, duration of ART 7.1 (2.3,10.8) years. We observed 3 distinct clusters, two characterized by higher inflammation; cluster 2 (19% subjects) reflecting platelet/macrophage pathways and cluster 3 (34% subjects), systemic, vascular and endothelial pathways (see Table 1). PWH in cluster 3 were older, more likely male and Caucasian. Although prevalence of comorbidities was higher in cluster 2 (42%) and 3 (48.9%) versus cluster 1 (20%), only cluster 3 was associated with prevalent comorbidites in regression analysis (OR 3.1 [1.1, 8.3] p= 0.03). This association remained significant after adjustment for CMV seropositivity, smoking and CD4:CD8 ratio, (OR 3.3 [1.1, 9.8] p=0.03). Further adjustment for age, gender and ethnicity attenuated the relationship (OR 2.4 [0.7, 7.9] p=0.16). Conversely Cluster 1, characterized by lower levels of inflammation, was associated with reduced risk of comorbidities (OR 0.28 [0.10, 0.74] p=0.01), an association which persisted after adjustment for baseline demographics, smoking and T cell ratio (OR 0.29 [0.10, 0.85] p=0.02).
We have identified distinct inflammatory patterns in treated PWH that predict prevalent co-morbidities. That these patterns, characterized by pathways including systemic and vascular inflammation remain associated with clinical outcomes even after correction for CMV and CD4:CD8 ratio suggest a number of distinct pathways contributing to co-morbidities in PWH.