Soluble Adhesion Molecules in Patients Coinfected with HIV and HCV: A Predictor of Outcome

Abstract

Background: Higher serum levels of adhesion molecules (sICAM-1 and sVCAM-1) are associated with advanced liver fibrosis in patients coinfected with human immunodeficiency virus and hepatitis C virus. We assessed the relationship between serum levels of adhesion molecules and liver-related events (LRE) or death, in coinfected patients.

Methods: We studied clinical characteristics and outcomes of 182 coinfected patients with a baseline liver biopsy (58 with advanced fibrosis) and simultaneous plasma samples who were followed for median of 9 years. We used receiver-operating characteristic (ROC) curves to calculate optimized cutoff values (OCV) of sICAM-1 and sVCAM-1, defined as the values with the highest combination of sensitivity and specificity for LRE. We used multivariate regression analysis to test the association between OCVs of sICAM-1 and sVCAM-1 and outcomes. The variables for adjustment were age, HIV transmission category, liver fibrosis, baseline CD4+ T-cell counts, antiretroviral therapy, and sustained virologic response (SVR).

Results: During the study period 51 patients had SVR, 19 had LRE, and 16 died. The OCVs for LRE were 5.68 Log pg/mL for sICAM-1 and 6.25 Log pg/mL for sVCAM-1, respectively. The adjusted subhazard ratio (aSHR) (95% confidence interval [CI]) of death or LRE, whichever occurred first, for sICAM-1 and sVCAM-1 > OCV were 3.98 ([1.14; 13.89], P = 0.030) and 2.81 ([1.10; 7.19], respectively (P = 0.030).

Conclusions: Serum levels of sICAM-1 and sVCAM-1 can serve as markers of outcome in HIV/HCV-coinfected patients. Therapies targeting necroinflammatory damage and fibrogenesis may have a role in the management chronic hepatitis C.

Fig 1. Accuracy of sICAM-1 and sVCAM-1 serum levels for the prediction of liver-related events.

The P value refers to the comparison of ROC curves according to the method of Hanley and McNeil. Abbreviations: AUROC, area under the receiver operating characteristic curves.

Fig 2. Scatter plot showing the distribution of baseline serum concentrations (Log pg/mL) of sICAM-1 and sVCAM-1 in 182 HIV/HCV coinfected patients included in the study.

Each symbol represents a single patient. Triangles represent patients who developed liver-related events during follow-up; circles represent patients who remained free from liver-related events during follow-up. Horizontal bars represent the optimized cutoff values for prediction of liver-related events of sICAM-1 (5.68 Log pg/mL) and sVCAM-1 (6.25 Log pg/mL). Abbreviations: LRE, liver-related events.

Fig 3

Kaplan-Meier plots showing cumulative incidence of 1) Liver-related events taking into account death as a competitive risk according to sICAM-1 serum levels higher than or lower than or equal to the optimized cutoff values (A) and sVCAM-1 serum levels higher than or lower than or equal to the optimized cutoff values (B). 2) Liver-related events or death, whichever occurred first, according to sICAM-1 serum levels higher than or lower than or equal to the optimized cutoff values (C) and sVCAM-1 serum levels higher than or lower than or equal to the optimized cutoff values (D). Abbreviations: LRE, liver-related events; OD, overall death; LRE/OD, liver-related events or death, whichever occurred first.