Circulating Proprotein Convertase Subtilisin/Kexin type 9 level independently predicts incident cardiovascular events and all-cause mortality in hemodialysis black Africans patients

In chronic kidney disease (CKD) patients not on dialysis, cardiovascular mortality is largely justified by the high level of cholesterol-rich in low-density lipoproteins (LDL-c) [1]. In addition, cardiovascular (CV) events remain the leading cause of death in hemodialysis (HD) patients [2, 3]. However, in those patients, the association between mortality and LDL-c is not yet clearly established. This weakly established relationship is one of the arguments that motivated the international guidelines, the Kidney Disease Improving Global Outcomes (KDIGO), which advise against the initiation of lipid-lowering drugs (statins) in HD patients [4, 5]. These molecules do not effectively reduce the LDL-c level or the resulting mortality [6, 7]. Proprotein Convertase Subtilisin/Kexin type 9 (PCSK9) by inhibiting the recycling of LDL-c receptors and promoting their degradation [8, 9] appears to be the main regulator of plasma levels of LDL-c. PCSK9 is a new marker potentially playing a causative role on CV mortality in maintenance HD patients. However, this association has been rarely evaluated [10]. Therefore, we conducted a cohort study in the HD centers located in Kinshasa, the Democratic Republic of the Congo (DRC) to assess the association between plasma level of PCSK9 and incident CV events as, well all-cause mortality, in Congolese HD patients.

This prospective observational study showed for the first time that the incidence of CV events increases with plasma PCSK9 level in black African HD patients. We found also that Kt/V, serum albumin, total cholesterol, triglyceride, and HDL-C, were independent determinants of PCSK9 in HD patients without statin treatment. These parameters alone accounted for more than half of the fluctuations in PCSK9 levels. Patients with plasma PCSK9 levels at tertile 3 had a high incidence of CV events and mortality. This risk of CV event and subsequent mortality increased linearly with the PCSK9 level. In addition, high PCSK9 levels negatively influenced the survival of the patients in the study. These observations suggest that PCSK9 may be a predicting biomarker for CV event occurrence in the hemodialysis patient population. While total cholesterol, triglyceride, HDLc were independent determinants of PCSK9 level in this study, Hwang HS et al. [10] studying HD patients on statin treatment identified no association between PCSK9 and lipid fractions (total cholesterol, LDL, HDL-c, triglyceride). The discrepancy in results between the two studies may be due in part to the fact that the HD patients in the present study were not on statin therapy and the hemodialysis treatment was not adequate given the fact that Kt/V is an independent determinant of PCSK9 (giving the preponderance of HD patients having Kt/V < 1.21, 51%). Indeed, both statin treatment and adequate HD-induced dysregulation in lipid parameters and the effect of HD on circulating PCSK9 should normally decrease the strength of this relationship [6, 10, 13, 14]. Hypoalbuminemia is associated both with elevated PCSK9 levels and all-cause mortality in the present study. The exact mechanism of hypoalbuminemia–associated with elevated PCSK9 levels remains unknown but might be due to increased secretion of PCSK9 from hepatocytes coupled with decreased clearance encountered in nephrotic syndrome [15]. It is likely that most HD patients in this study exhibit proteinuria given the preponderance of primitive or secondary glomerulonephritis (diabetes, HVC, HIV, and HVB infection). Yet, this study did not address the specific causes of albuminuria in individual subjects.

HD patients at tertiles 2 and 3 of PCSK9 had a higher risk of death than those in tertile 1. This association remained significant after adjusting for all risk factors for death in the patients included in this cohort. A similar association was observed at the beginning of HD in Korean patients with a history of cardiovascular disease and under lipid-lowering treatment [16]. Yet, our patients were not on lipid-lowering therapy; both family and personal history of cardiovascular risk did not show any significant difference between surviving and deceased patients. The association of PCSK9 and mortality in HD patients appeared to be explained by the fact that the gene encoding PCSK9 in CKD patients can undergo mutations [17]. Indeed, in animal models, a gain in the function of PCSK9 is associated with an increase in levels of fasting plasma triglycerides in mice [18]. Conversely, a loss of PCSK9 function corresponded to a marked reduction in postprandial triglyceride levels in mice [19]. Furthermore, PCSK9 loss-of-function mutations have been associated with lower LDL-c levels in patients with ESKD and reduced cardiovascular risk [20, 21]. Therefore, it is likely that our patients may have a gain in the function of PCSK9. As the gain of function is crucial for the catabolism of LDL-c receptors, LDL-c levels increase. LDL-c contains ApoB particles rich in VLDL-c and triglycerides, are very sensitive to oxidation, and are very atherogenic [22]. Indeed, an interaction between lipoproteins and the immune system creates atheromatous plaques in the vascular wall [23, 24]. A positive correlation between plasma PCSK9 levels and carotid calcifications has also been demonstrated by CT scans in patients with familial hypercholesterolemia [25]. Thus, the high plasma level of PCSK9 and the subsequent atherosclerosis process may explain, at least partly, the high cardiovascular mortality rate observed in HD patients studied [26, 27].

Patients in the present study were not on statin treatment. Indeed, statins often prescribed in CKD increase the circulating level of PCSK9 by activating a binding protein Sterol Regulatory element-binding Protein-2 [SREBP-2] [14]. The high level of PCSK9 observed in our study implies a higher CV event [28] which in turn may justify the higher mortality in these patients.

In addition, PCSK9 is synthesized as a precursor of about 72 kDa which must undergo autocatalytic cleavage between the catalytic pro-domain and domain within the endoplasmic reticulum of the hepatocyte before being secreted into the bloodstream [29]. This molecular weight makes PCSK9 a large molecule that theoretically is not filtered by conventional HD high flux. Conventional filters are effective for molecules with a molecular weight less than 65 kDa. Therefore, high levels of circulating PCSK9 might be due to a lack of dialyzability with high-flux, conventional HD. This finding suggests that hemofiltration or the use of High Flux filters could be beneficial in this setting in order to reduce C-V deaths in dialysis patients. Alternatively, these high PCSK9 levels may also be due to the inflammatory nature of kidney failure often observed in maintenance HD patients [15, 30].

The present study showed a high incidence of CV events and mortality in haemodialysis patients with PCSK9 in tertile 3. The risk of CV events and mortality increased linearly with PCSK9 level while the levels of plasma lipid fractions remained without significant differences between the survivors and the deceased patients. The current finding is consistent with previous studies showing that PCSK9 contributes to the development of CV events independently of traditional CV risk factors [10]. The level of PCSK9 could therefore be a new biomarker of CV risk in HD patients.

Strengths of this study include the concomitant measurement of PCSK9 and lipid fractions in HD patients attending multicenter of HD in Kinshasa and the investigation of the implications of their association on CV events as well as the survival of these patients. Finally, several variables were included in the analysis including laboratory parameters and HD session associated variables.

Limitations are related to the methodological approach. Firstly, measurement of PCSK9 and lipids was only performed once at the beginning of the study. Secondly, PCSK9 activity would be more accurate if PCSK9 has been measured bound to LDL-c receptors. Thirdly, the type of vascular access was not considered in the analysis, the variable which could have some impact on survival. Finally, our study is also based on relatively small sample size.

In this study, elevated plasma PCSK9 levels independently predicts incident CV events and all-cause mortality in HD black African patients. Future studies are needed to determine the genetic mutations of PCSK9 in the black African hemodialysis patient population.