Apolipoproteins A and B and PCSK9: Nontraditional Cardiovascular Risk Factors in Chronic Kidney Disease and in End-Stage Renal Disease

Purpose Nontraditional cardiovascular risk factors as apolipoprotein A (ApoA), apolipoprotein B (ApoB), and the proprotein convertase subtilisin/kexin type 9 (PCSK9) increase the prevalence of cardiovascular mortality in chronic kidney disease (CKD) or in end-stage renal disease (ESRD) through quantitative alterations. This review is aimed at establishing the biomarker (ApoA, ApoB, and PCSK9) level variations in uremic patients, to identify the studies showing the association between these biomarkers and the development of cardiovascular events and to depict the therapeutic options to reduce cardiovascular risk in CKD and ESRD patients. Methods We searched the electronic database of PubMed, Scopus, EBSCO, and Cochrane CENTRAL for studies evaluating apolipoproteins and PCSK9 in CKD and ESRD. Randomized controlled trials, observational studies (including case-control, prospective or retrospective cohort), and reviews/meta-analysis were included if reference was made to those keys and cardiovascular outcomes in CKD/ESRD. Results 18 studies met inclusion criteria. Serum ApoA-I has been significantly associated with the development of new cardiovascular event and with cardiovascular mortality in ESRD patients. ApoA-IV level was independently associated with maximum carotid intima-media thickness (cIMT) and was a predictor for sudden cardiac death. The ApoB/ApoA-I ratio represents a strong predictor for coronary artery calcifications, cardiovascular mortality, and myocardial infarction in CKD/ESRD. Plasma levels of PCSK9 were not associated with cardiovascular events in CKD patients. Conclusions Although the “dyslipidemic status” in CKD/ESRD is not clearly depicted, due to different research findings, ApoA-I, ApoA-IV, and ApoB/ApoA-I ratio could be predictors of cardiovascular risk. Serum PCSK9 levels were not associated with the cardiovascular events in patients with CKD/ESRD. Probably in the future, the treatment of dyslipidemia in CKD/ESRD will be aimed at discovering new effective therapies on the action of these biomarkers.


Introduction
Worldwide, chronic kidney disease (CKD) represents a high public health priority [1]. Worldwide, over 2 million people require renal replacement therapy (hemodialysis (HD), peritoneal dialysis (PD), or kidney transplantation) to increase their survival rates [1,2]. The prevalence of CKD has had an upward trend both in Europe and around the world, ESRD being merely the top of the iceberg [3]. CKD is an important cause of global mortality [1,4]. The number of deaths caused by CKD has increased by 82.3% over the past two decades, being the third cause of the top 25 causes of deaths, after HIV/AIDS and diabetes [4].
Dyslipidemia in patients with impaired renal function is characterized by both qualitative changes in the cholesterol homeostasis and quantitative changes regarding the lipid

Method: Search Strategy
We searched the electronic database of PubMed, Scopus, EBSCO, and the Register of Controlled Trials (Cochrane CENTRAL) from 3 January 2018 to 30 December 2018 for studies that evaluated the apolipoprotein profile in patients with CKD and ESRD and its cardiovascular outcomes. The terms used for searching were "apolipoprotein A-I", "apolipoprotein A-IV", "apolipoprotein B", "apolipoprotein B/apolipoprotein A-I ratio", "PCSK9", "end-stage renal disease", "ESRD", "chronic kidney failure", "CKD", "advanced CKD", "hemodialysis", and "peritoneal dialysis". Relevant references in these articles were searched manually to identify possible additional studies [10]. Randomized controlled trials, observational studies, including case-control studies, prospective or retrospective cohort studies, reviews, and meta-analyses were included if reference was made to apolipoproteins and their cardiovascular outcomes in CKD/ESRD [10]. Case reports were excluded, and studies were selected by two independent reviewers by screening the title and abstract. In a second phase, the full articles which conformed to the selection criteria were obtained, the essential data was extracted independently, and the results were analysed [10]. Discrepancies were resolved by discussion and consensus, and duplicates were excluded both manually and through a reference manager software [10]. Of these, only 18 met the inclusion criteria (Table 1). For the selected studies, we reviewed the full-text article and additional relevant publications were added after screening the reference section.

Results and Discussion
Apolipoproteins A and B and the ApoB/ApoA-I ratio are predictors of cardiovascular outcomes and potential biomarkers for cardiovascular mortality in both the general population and CKD/ESRD patients [11,12]. It is not currently clear whether these biomarkers represent cardiovascular risk factors or could help in CVD diagnosis and the setting of the therapeutic targets in CKD/ESRD patients. In a comprehensive review, Vlad et al. showed that lipoprotein(a) (Lp(a)), the genetic polymorphisms of apolipoprotein(a), apolipoprotein E (ApoE), and apolipoprotein B (ApoB) undergo modifications in uremic patients, being correlated with cardiovascular events [10]. Furthermore, it was pointed out that in ESRD patients, Lp(a) levels were independent risk factors for atherothrombosis and cardiovascular mortality, LMW apo(a) phenotype was the best predictor for coronary events, single nucleotide polymorphisms in ApoE gene increased the risk of cardiovascular events, and ApoB had a significant correlation with the value of carotid intima-media thickness and vascular stiffness [10].
Our search has led to several studies with different results and conclusions (Table 1), which has created confusion regarding the roles ApoA-I, ApoA-IV, ApoB/ApoA-I ratio, and PCSK9 within the CKD/ESRD framework. This lack of consistency could be caused by the methodology of different types of studies (cross-sectional/case-control studies), small numbers of patients in the study groups, different clinical and laboratory outcomes, lack of homogeneous criteria for inclusion/exclusion, different definitions of endpoints, various periods of follow-up, or different statistical approaches.

Apolipoprotein A-I
5.1. Background. Apolipoprotein A-I (ApoA-I) is secreted predominantly by the liver and intestine as lipid-free ApoA-I and constitutes approximately 70% of HDL protein, being required for the normal HDL biosynthesis [13]. ApoA-I levels are strongly associated with those of HDL-C [11]. ApoA-I binds to circulating phospholipids and forms pre-β HDL (lipid-poor nascent discoid HDL particles) [7]. ApoA-I is involved in the elimination of excess cholesterol in tissues, which it incorporates into HDL for direct, indirect, or reverse transport via LDL to the liver [11]. ApoA-I inhibits the expression of endothelial adhesion molecules such as intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1), while it prevents the production of monocyte chemoattractant protein-1 (MCP-1), which are critical steps for the production of reactive oxygen species (ROS) in the arterial wall [7]. Likewise, ApoA-I has anti-inflammatory properties, which may contribute to its cardioprotective role [14]. In addition, ApoA-I displays strong antioxidant properties [7] as well as numerous functions ( Figure 2) [15][16][17].

Apolipoprotein A-I in CKD and ESRD.
ESRD is associated with a significant decrease in plasma ApoA-I and HDL-C [7,18]. ApoA-I values in relation to ApoB values are used in estimating cardiovascular risk in patients with CKD/ESRD and CVD [19].
Thus, ESRD is associated with decreased levels of HDL-C and ApoA-I and may contribute to the atherogenic pathology [7]. In patients with CKD or ESRD, ApoA-I can efficiently evaluate the risk for cardiovascular disease [20], while its elevated level has been associated with a good survival rate [21].

Study Data.
In a prospective cohort study conducted by Honda et al., the serum levels of ApoA-I were significantly decreased in patients with CKD 5D as compared to those with CKD stages 2-3 [22], in consent with the data reported by the ARIC study (Atherosclerosis Risk in Communities), in which CKD patients in stages 3-4 without coronary heart disease (CHD) had a low ApoA-I concentration [23].

Pro Studies.
In a cross-sectional multicenter study enrolling 995 HD patients, Hung et al. found that ApoA-I was positively correlated with total cholesterol, HDL-C, blood urea (BUN), and serum albumin [20], but ApoA-I had a negative correlation with pulse wave velocity (PWV) [24].
ApoA-I was negatively associated with cardiovascular morbidity in both predialyzed CKD patients (area under the curve ðAUCÞ = 0:372; p < 0:0001) [21] and HD patients [11]. Moreover, ApoA-I has been significantly associated with the development of a new cardiovascular event [21] and has had the strongest independent correlation for CHD, the cut-off value for ApoA-I being 216.2 mg/dl [20]. Therefore, Zhan et al. also identified an association between ApoA-I and cardiovascular events in PD patients [25]. In addition, the low level of ApoA-I and serum creatinine constituted significant predictors of coronary pathology [20].
ApoA-I was significantly associated with all-cause mortality and cardiovascular mortality in HD patients [26] and PD patients [25]. In CKD patients, Cerezo et al. revealed that high ApoA-I concentrations have been significantly associated with the development of new cardiovascular episodes and were negatively associated with mortality (but with a lower level of significance) [21].

Con
Studies. Despite these findings, in an observational study that enrolled 412 HD patients, Honda et al. established that ApoA-I was not a risk factor for cardiovascular events [27]. Likewise, ApoA-I was correlated with cardiovascular events, but without any predictive strength in CKD patients (stages 2-5D) [22]. In an observational study with 91 HD patients, Bevc et al. revealed that ApoA-I did not correlate with carotid intima-media thickness (cIMT) [28]. Furthermore, in cohort CARE FOR HOMe (Cardiovascular and Renal Outcome in CKD 2-4 Patients-The Forth Homburg evaluation), which enrolled 443 patients, Rogacev et al. have revealed in a multivariate analysis after adjusting for confounders that ApoA-I levels were not associated with CV events (p = 0:483) [6].        [29][30][31][32]. In fasting plasma, ApoA-IV circulates as part of a lipid-poor, small HDL-like particle and ApoA-I free particles [29,33]. ApoA-IV activates lecithin-cholesterol acyltransferase (LCAT) and modulates lipoprotein lipase (LPL) activation, favoring cholesteryl ester transfer from HDL to LDL, hence suggesting that ApoA-IV may behave like an antiatherogenic factor [29,32]. ApoA-IV has antioxidant and antiatherogenic properties, and therefore, low levels of ApoA-IV increase the risk of CHD [31,32] (Figure 3).

Apolipoprotein A-IV in CKD and ESRD.
ApoA-IV also plays a relevant role in the reverse cholesterol transport [30][31][32][33][34], which is affected in patients with CKD [34]. Few studies have investigated serum ApoA-IV in patients with CKD and have shown that the kidney plays a crucial part in its metabolism [30,32]. Renal function parameters (GFR, creatinine, and BUN) were the most important determinants of serum ApoA-IV levels in patients with CKD [29]. Immunohistochemical studies have indicated that ApoA-IV is filtered in the glomerulus and is mostly reabsorbed by proximal tubular cells [33,34]. ApoA-IV is significantly raised in HD and DP patients [29,30,32]. ApoA-IV begins to grow from the initial stages of CKD, becoming thus an early marker of renal failure [29]. Serum ApoA-IV is associated with the development of atherosclerotic lesions in HD patients and can be useful for estimating the cardiovascular risk [30]. Moreover, low levels of ApoA-IV have been validated as a risk predictor for all-cause mortality and sudden cardiac death, being adjusted by the nutritional status [31].

Study Data.
ApoA-IV is a key link between the decreased GFR and the presence of cardiovascular events [29]. Kronenberg et al. revealed that ApoA-IV significantly increased with the decreasing glomerular filtration rate (GFR), especially in dialyzed patients [29]. Subjects with mild and moderate CKD [29] and ESRD (HD patients) [30], who developed atherosclerotic complications (carotid artery plaques and a low ankle-brachial index), had a decrease in ApoA-IV plasma concentrations [30] as compared to control participants (24:9 ± 8:7 versus 22:3 ± 7:7 mg/dl, p < 0:15 for mildmoderate CKD) [29]. Also, ApoA-IV had a 60% sensitivity, a 69% specificity, and a cut-off value of 321.92 μg/ml [30]. In addition, in a post hoc analysis performed in the Die Deutsche Diabetes Dialyse Studie (Study 4D), the scientists found the average ApoA-IV concentration of 49:8 ± 14:2 mg/dl about three times higher in HD patients than in the general population, which could not be influenced by statin administration [31].

Pro Studies.
Patients with elevated ApoA-IV levels displayed a lower risk factor on coronary atherosclerosis and cardiovascular disease as compared to low ApoA-IV subjects [30]. In the post hoc 4D study, at baseline, ApoA-IV concentrations were closely associated with the congestive heart failure and with ECG changes (arrhythmia, atrial fibrillation, atrial flutter, left bundle branch block, or right bundle branch block) [31]. Kronenberg  ApoA-IV increase of 1 mg/dl decreased odds ratio by 8% for these complications (p < 0:011) [29]. Likewise, Omori et al. indicated that the ApoA-IV level was independently associated with maximum carotid intima-media thickness (cIMT) and cardiovascular disease [30].
Furthermore, Luczak et al. have carried out a comparative proteomic analysis of plasma proteins isolated from 75 patients in different stages of renal disease, 25 patients with advanced cardiovascular disease, and 25 healthy volunteers [35]. A direct comparison between CKD and CVD groups revealed significant differences in the accumulation of 2 proteins: ApoA-IV and α-1-microglobulin [35]. These proteins individually contributed to the formation of atheroma plaque, yet the inflammatory process was more powerful in patients with CKD [35]. On the other hand, the stimulation or inhibition of ApoA-IV in CVD and CKD groups suggested differences in the cholesterol transport efficacy [35].
Following the body mass index (BMI) adjustment, in ESRD patients with BMI > 23 kg m 2 , ApoA-IV was associated with all-cause mortality, heart rate mortality, sudden cardiac death, cerebrovascular events, and cardiovascular disorders [31]. Also, Kollerits et al. showed that the increase of 10 mg/dl was associated with a reduced risk of death of 11% (p = 0:001) [31].  [11,25,26] and reflects the atherogenic particles from the body [11,25,26]. Other roles of ApoB are displayed in Figure 4 [36]. ApoB has two forms: ApoB-48 which is exclusively intestine secreted in chylomicrons and ApoB-100 which is only secreted by the liver in VLDL [37].
In the general population, several clinical studies (e.g., AMORIS (Apolipoprotein-related MOrtality RISk) [38] or INTERHEART [39]) have shown that the ApoB/ApoA-I ratio is strongly correlated with cardiovascular events such as myocardial infarction and stroke [40].

Apolipoprotein B and ApoB/ApoA-I Ratio in CKD and ESRD.
Patients with CKD (G1-G4) have elevated ApoB values [23,24]. In HD patients, ApoB concentration is within the normal range, in contrast with PD patients, who have elevated ApoB levels (due to overproduction) [41]. HD patients display advanced atherosclerosis that is associated with nontraditional risk factors (ApoB) [11,28]. The reduction of ApoB could be a critical risk marker for eccentric left ventricular remodeling and could be useful for cardiovascular risk stratification in PD patients [42].
In CKD patients, the ApoB/ApoA-I ratio reflects the cholesterol balance between atherogenic and antiatherogenic lipoprotein particles [12], and an increased ratio highlights the progression of atherosclerosis [25]. Moreover, the Apo-B/ApoA-I ratio represents a strong predictor for coronary artery calcifications [12] and myocardial infarction in CKD or ESRD [43]. Also, ApoB/ApoA-I ratio measurement was significantly associated with all-cause mortality and cardiovascular mortality in HD and PD patients [25,26].

Study Data.
In the ARIC study (Atherosclerosis Risk in Communities), patients with CKD stages 3-4 recorded an important ApoB/ApoA-I ratio increment as compared to those without CKD [23].

Promotes the formation of chylomicrons
Acts as a ligand for the LDL Promotes the formation of nascent VLDL ApoB Figure 4: The roles of ApoB. In the liver, ApoB promotes the formation of nascent VLDL and also is essential for the linking of LDL particles to LDLR for cellular absorption and degradation of LDL particles. In the intestine, ApoB stimulates the formation of chylomicrons. LDL: lowdensity lipoprotein; VLDL: very low-density lipoprotein.
7.3.1. Pro Studies. ApoB/ApoA-I ratio had a 100% sensitivity, a 77% specificity, and a cut-off value of 1.13 [11]. Besides, in the same case-control study conducted by Kirmizis et al., the ApoB/ApoA-I ratio was positively correlated with cardiovascular morbidity [11]. Moreover, in patients with stages 3-4 of CKD, the incidence of acute myocardial infarction was strongly associated with this ratio [43].
After adjustment for confounders, the ApoB/ApoA-I ratio was associated with cardiovascular events when serum levels exceeded the cut-off value, in patients with mild renal impairment [11], HD patients [27], and PD subjects [25]. Although the ApoB/ApoA-I ratio was associated with cardiovascular events, it displayed differences between the quartile models and the 1-SD increases, being influenced by interleukin-6 (IL-6) [27].
However, Sato et al. have found a substantial association of ApoB/ApoA-I ratio with all-cause and cardiovascular mortality in HD patients [26]. It appears that this ratio was significantly associated with all-cause mortality in PD patients [25]. 7.3.2. Con Studies. In contrast, in an observational study, Kim et al. reported that ApoB/ApoA-I ratio was not associated with coronary artery calcification in patients with mild renal impairment [12]. Moreover, the ARIC study did not detect an association between the ApoB/ApoA-I ratio and the risk for coronary disease [23].
It triggers the reduction of LDLR levels without affecting the LDLR mRNA (messenger ribonucleic acid). PCSK9 determines the degradation of LDLR [50] and inhibits receptor recycling in the hepatocyte membrane [6]. The human PCSK9 protein (hPCSK9) is synthesized mainly in the liver, the kidney, the small intestine [46], and the brain [44,45] via sterol regulatory element-binding protein 2 (SREBP-2) regulation [46]. Sterol regulatory element-binding proteins (SREBPs) coordinate the synthesis and cellular uptake for cholesterol and fatty acids [51]. SREBP-2 is primarily responsible for the activation of genes involved in the cholesterol synthesis, as opposed to fatty acid synthesis [51]. Thus, by activating the SREBP-2 pathway, statins increase the level of PCSK9, limiting the efficiency of these drugs for lowering LDL cholesterol [46].
In the kidney, PCSK9 modulates the sodium absorption by degrading the epithelial sodium channel, and it also plays a part in the regulation of blood pressure [52]. PCSK9 has been identified in human pancreatic cells and does not modify endocrine pancreatic function [53]. Although adipocytes do not express PCSK9, they are rich in LDL and VLDL receptors, which play an important part in the hydrolysis of triglyceride-rich lipoproteins, and are helpful for fat storage in these cells [54]. In addition, PCSK9 is observed in carotid atherosclerotic lesions, especially in vascular smooth muscle cells [55]. Additional roles [56] are shown in Figure 5.
The human PCSK9 gene is found on the human chromosome 1 and encodes the PCSK9 protein. The "gain-offunction" mutations of PCSK9 are associated with a rare form of autosomal dominant hypercholesterolemia (ADH), whereas "loss-of-function" mutations result in lowering cholesterol levels by reducing the CHD rate [46]. These genetic variants of PCSK9 affect both the plasma concentrations of PCSK9 and the serum level of LDL-C [49], thus becoming new targets for the treatment of hypercholesterolemia [57].

PCSK9 and CKD/ESRD
In patients with CKD, the available evidence for PCSK9 is insufficient, with very few observational studies and with a small number of patients. At the same time, there are no data on the use of PCSK9 inhibitors to them [57]. In HD patients, PCSK9 levels were close to control group levels [58], or a differentiation of serum PCSK9 values before and after HD was not identified [59]. In PD patients, PCSK9 levels were close to those measured in patients with nephrotic syndrome [60].  [6], and they observed that plasma PCSK9 was poorly correlated with the total cholesterol, ApoB, and triglycerides [6].

Con Studies.
Conflicting results were further reported by Rogacev et al. who found no significant correlation between PCSK9 and GRF in nonstatin users of the LURIC cohort (p = 0:733) [6]. In addition, in the same two independent cohorts, Rogacev et al. observed that plasma PCSK9 values were correlated neither with baseline GFR values nor with LDL-C [6]. Likewise, plasma levels of PCSK9 were not associated with cardiovascular events in patients with low renal function [6].
Although PCSK9 is a potential determinant of serum cholesterol, no relationship with early or current cardiovascular disease has been identified, and thus, it cannot be considered a cardiovascular risk factor in CKD or ESRD patients [61]. Furthermore, Kaplan-Meier analysis revealed that serum PCSK9 levels did not predict cardiovascular events in any cohort (CARE FOR HOMe p = 0:622; LURIC p = 0:729) [6].

Therapy Options in CKD/ESRD for Apolipoprotein and PCSK9
Treating dyslipidemia with statins and ezetimibe results in favorable effects for cardiovascular disease (CVD) prevention, in patients with moderate CKD [57]. This therapeutic strategy has not proven effective in HD patients as indicated in the cardiovascular outcomes of 4D, AURORA, and SHARP studies [7]. Furthermore, in CKD/ESRD patients treated with statin, PCSK9 concentration was higher compared to nonstatin subjects [6,59,61].
In patients with CKD/ESRD, the use of PPARα agonists is still controversial, and long-term safety and efficacy remain open to research [62]. By using niacin, besides having no additional benefit in patients with satisfactory control of LDL-C concentrations, tolerability was reduced [62]. ApoA-I mimetic is a new challenge for improving the lipid profile (in animal models), but clinical trials are still needed to confirm widespread use [62]. Although cholesteryl ester transfer protein (CETP) and cholesterol acyltransferase (ACAT) inhibitors have significantly improved HDL levels, results from major clinical trials have identified increased cardiovascular events [62]. Therefore, the need for an improvement of the lipid panel has emerged; new biomarkers (PCSK9) and new therapeutic strategies (monoclonal antibodies-evolocumab, alirocumab) are being further identified [57].

Key Points and Future Directions
The results of the studies focusing upon the association between apolipoproteins and CKD, after adjusting for cardiovascular events, suggest that supplementary mechanisms may be involved in addition to large vessel atherosclerosis. These include glomerulosclerosis, small vessel atherosclerosis, and direct toxic effects of apolipoproteins and lipids on the podocytes as well. The harmful effects are mediated by numerous mechanisms: expanded oxidative stress due to low levels of the apolipoprotein A1-enriched HDL fraction; the oxidized LDL-ApoB rich, local foam cell development; and the activation of inflammation [63].
The use of PCSK9 as a potential biomarker to identify patients with diabetic nephropathy who could benefit from anti-PCSK9 strategies and inhibition of PCSK9 could become an important treatment target in patients with CKD.
Patients with CKD should be enrolled in multicenter, randomized, double-blind trials (e.g., FOURIER, ODYSSEY) and closely monitored for the treatment efficacy against major cardiovascular events.

Conclusions
ESRD is associated with decreased levels of ApoA-I, with the increased levels ApoA-IV and ApoB, and with a high Apo-B/ApoA-I ratio, but plasma PCSK9 levels are not associated with GFR decrease.
In patients with CKD/ESRD, even if there are the controversial results in the relationship between biomarkers and major cardiovascular events, ApoA-I, ApoA-IV, and ApoB/ApoA-I ratio are predictors of cardiovascular events. Nevertheless, these biomarkers could be useful for monitoring therapies with an impact upon cardiovascular morbimortality. Plasma PCSK9 levels are not associated with cardiovascular events in CKD or ESRD patients. However, circulating PSCK9 stands out as a promising biomarker for the diagnosis of dyslipidemia in patients with CVD and those affected by familial hypercholesterolemia.
In CKD patients, statins and ezetimibe can contribute to the prevention of CVD. The CETP, ACAT, and PCSK9 inhibitors have significantly improved HDL levels and reduced LDL-C. Also, ApoA-I mimetic is a new challenge for improving the lipid profile, but clinical trials are still needed to confirm widespread use.