Serum cytokines and inflammatory proteins in individuals with heroin use disorder: potential mechanistically based biomarkers for diagnosis.

Abstract

Opioid use disorders cause major morbidity and mortality, and there is a pressing need for novel mechanistic targets and biomarkers for diagnosis and prognosis. Exposure to mu-opioid receptor (MOR) agonists causes changes in cytokine and inflammatory protein networks in peripheral blood, and also in brain glia and neurons. Individuals with heroin use disorder (iHUD) show dysregulated levels of several cytokines in the blood. However, there is limited data on a comprehensive panel of such markers in iHUD versus healthy controls (HC), especially considered as a multi-target biomarker. We used a validated proximity extension assay for the relative quantification of 92 cytokines and inflammatory proteins in the serum of iHUD on medication-assisted therapy (MAT; n = 21), compared to HC (n = 24). Twenty-nine targets showed significant group differences (primarily iHUD>HC), surviving multiple comparison corrections (p = 0.05). These targets included 19 members of canonical cytokine families, including specific chemokines, interleukins, growth factors, and tumor necrosis factor (TNF)-related proteins. For dimensionality reduction, data from these 19 cytokines were entered into a principal component (PC) analysis, with PC1 scores showing significant group differences (iHUD > HC; p < 0.0001). A receiver-operating characteristic (ROC) curve analysis yielded an AUROC = 91.7% (p < 0.0001). This PC1 score remained a positive predictor of being in the HUD group in a multivariable logistic regression, that included select demographic/clinical variables. Overall, this study shows a panel of cytokines that differ significantly between iHUD and HC, providing a multi-target "cytokine biomarker score" for potential diagnostic purposes, and future examination of disease severity.