Proprotein convertase subtisilin/kexin 9 levels decline with hepatitis C virus therapy in people with HIV/hepatitis C virus and correlate with inflammation

Abstract

Background: Proprotein convertase subtisilin/kexin 9 (PCSK9) raises low-density lipoprotein cholesterol (LDL-C) levels and is associated with inflammation, which is elevated in HIV and hepatitis C virus (HCV) infection. We compared PCSK9 levels in people with co-occurring HIV and HCV (HIV/HCV) vs. HIV alone, and evaluated the impact of HCV direct-acting antiviral (DAA) therapy on PCSK9.

Design: A prospective, observational cohort study.

Methods: Thirty-five adults with HIV/HCV and 37 with HIV alone were evaluated, all with HIV virologic suppression and without documented cardiovascular disease. Circulating PCSK9 and inflammatory biomarkers were measured at baseline and following HCV treatment or at week 52 (for HIV alone) and compared using Wilcoxon tests and Spearman correlations.

Results: At baseline, PCSK9 trended higher in HIV/HCV vs. HIV alone (307 vs. 284 ng/ml, P = 0.06). Twenty-nine participants with HIV/HCV completed DAA therapy with sustained virologic response. PCSK9 declined from baseline to posttreatment 1 (median 7.3 weeks after end of therapy [EOT]) and posttreatment 2 (median 43.5 weeks after EOT), reaching levels similar to HIV alone; median within-person reduction was -60.5 ng/ml ( P = 0.003) and -55.6 ng/ml ( P = 0.02), respectively. Decline in PCSK9 correlated with decline in soluble (s)E-selectin and sCD163 ( r = 0.64, P = 0.002; r = 0.58, P = 0.008, respectively), but not with changes in LDL-C or other biomarkers. No significant change in PCSK9 occurred in the HIV alone group over 52 weeks.

Conclusion: PCSK9 declined with DAA therapy in participants with HIV/HCV, correlating with declines in several inflammatory biomarkers but not LDL-C. Elevated PCSK9 with HCV may be linked to particular HCV-associated inflammatory pathways more so than cholesterol homeostasis.

Figure 1.

Flow chart of baseline and follow-up assessments for the HIV/HCV and HIV alone groups. For participants with HIV/HCV, HCV treatment with DAAs was intiated outside of the study (at any time after study entry) by participants’ healthcare providers or through clinical trials. For participants with HIV/HCV who initiated DAAs, the first follow-up visit (post-treatment 1) occurred during a window of 3 to 8 weeks post-treatment and the second follow-up visit (post-treatment 2) occurred during a window of 23 to 60 weeks post-treatment. For participants with HIV alone, follow-up visits occurred at week 52 +/− 28 days after entry. Blood collection for stored serum and plasma for testing of PCSK9 and other biomarker levels occurred at the designated time points. Abbreviations: DAA: direct-acting antiviral.

Figure 2.

(A) Spaghetti plot for within-person change in PCSK9 levels in individuals with HIV/HCV with HCV DAA therapy. Each dot represents an individual, and lines connect paired samples. P-values are based on the Wilcoxon signed-rank test for within-person change. (B) Boxplots representing PCSK9 levels in the HIV/HCV and HIV alone groups at baseline and follow-up time points. Median (interquartile range) PCSK9 levels in ng/mL at each time point are as follows: HIV/ HCV baseline: 306.9 (239.1, 418.7), HIV/HCV post-treatment 1: 237.1 (202.2, 307.7), HIV/HCV post-treatment 2: 251.3 (203.5, 300.7), HIV alone baseline: 284.1 (208.9-364.6), HIV alone week 52: 255.7 (246.5-282.2). There were no significant differences between PCSK9 levels in the HIV/HCV group at post-treatment 1 or 2 and the HIV alone group at baseline (p=0.57 and p=0.43, respectively) or week 52 (p=0.66 and p=0.66, respectively), based on the Wilcoxon rank sum test for comparison between groups. (C) Spaghetti plot for within-person change in PCSK9 levels in individuals with HIV alone from baseline to week 52 follow-up. Each dot represents an individual, and lines connect paired samples. P-value is based on the Wilcoxon signed-rank test for within-person change. (D) Boxplots representing within-person change in PCSK9 in the HIV/HCV vs. HIV alone groups. P-values are based on the Wilcoxon rank sum test for comparison between groups of within-person change in PCSK9. The length of the boxes represents the IQR, the horizontal lines within the boxes represent the group medians, the whisker length represents 1.5 * IQR, and the dots represent outliers.

Figure 3.

Correlations between change in PCSK9 and change in sCD163 with HCV therapy from baseline to post-treatment 1 (A) and from baseline to post-treatment 2 (B). Correlations between change in PCSK9 and change in sE-selectin with HCV therapy from baseline to post-treatment 1 (C) and from baseline to post-treatment 2 (D). All correlations are by nonparametric Spearman rank correlations. Ellipses represent 95% prediction estimates. Abbreviations: sE-selectin: soluble E-selectin; sCD163: soluble CD163.