Effect of PCSK9 Inhibition on Plasma Levels of Small Dense Low Density Lipoprotein-Cholesterol and 7-Ketocholesterol

Abstract

Background:

Oxidized forms of cholesterol (oxysterols) are implicated in atherogenesis and can accumulate in the body via direct absorption from food or through oxidative reactions of endogenous cholesterol, inducing the formation of LDL particles loaded with oxidized cholesterol. It remains unknown whether drastic reductions in LDL-cholesterol (LDL-C) are associated with changes in circulating oxysterols and whether small dense LDL (sdLDL) are more likely to carry these oxysterols and susceptible to the effects of PCSK9 inhibition (PCSK9i).

Objective:

We investigate the effect of LDL-C reduction accomplished via PCSK9i on changes in plasma levels of sdLDL-cholesterol (sdLDL-C) and a common, stable oxysterol, 7-ketocholesterol (7-KC), among 134 patients referred to our Preventive Cardiology clinic.

Methods:

Plasma lipid panel, sdLDL-C, and 7-KC measurements were obtained from patients before and after initiation of PCSK9i.

Results:

The intervention caused a significant lowering of LDL-C (−55.4%). The changes in sdLDL-C levels (mean reduction 51.4%) were highly correlated with the reductions in LDL-C levels (R=0.829, p<0.001). Interestingly, whereas changes in plasma free 7-KC levels with PCSK9i treatment were much smaller than (−6.6%) and did not parallel those of LDL-C and sdLDL-C levels, they did significantly correlate with changes in triglycerides and very low-density lipoprotein-cholesterol (VLDL-C) levels (R=0.219, p=0.025).

Conclusion:

Our findings suggest a non-preferential clearance of LDL subparticles as a consequence of LDL receptor upregulation caused by PCSK9 inhibition. Moreover, the lack of significant reduction in 7-KC with PCSK9i suggests that 7-KC may be in part carried by VLDL and lost during lipoprotein processing leading to LDL formation.

Introduction

Elevated plasma low density lipoprotein (LDL)-cholesterol (LDL-C) concentrations are strongly associated with atherosclerotic cardiovascular disease (ASCVD) (1). Proprotein convertase subtilisin/kexin type 9 (PCSK9) is a low abundance plasma protein that has an outsized effect on regulating plasma LDL-C by facilitating lysosomal degradation of hepatic low-density lipoprotein receptors (LDLR) (2). Therapeutic monoclonal antibodies fully bind and disarm PCSK9 to drastically reduce plasma LDL-C levels (~60%) and incrementally improve cardiovascular outcomes in patients with ASCVD on background statin therapy (3, 4). In 2015, the FDA approved alirocumab and evolocumab, two fully human monoclonal antibody PCSK9 inhibitors (PCSK9i), for use in individuals with familial hypercholesterolemia and/or ASCVD who require LDL-C lowering beyond what can be achieved with lifestyle and conventional lipid-lowering therapy (5).

Due to the accelerated pace by which PCSK9i proceeded from bench to bedside, there remain important gaps in understanding PCSK9 physiology. In particular, little is known about the impact of PCSK9i on various lipoprotein subfractions, such as small dense LDL-C (sdLDL-C) and oxysterols, both of which are additional contributors to atherogenic risk. It is hypothesized that sdLDL exhibits greater atherogenicity compared to larger LDL subfractions for multiple reasons including smaller size to enter the arterial wall, increased affinity for arterial wall proteoglycans, longer resident time in circulation, and higher susceptibility to modification including oxidation (6–10). Numerous observational studies demonstrate the association of sdLDL with prevalent and incident ASCVD, independent of traditional cardiovascular risk factors, including LDL-C (6–10). Previous small studies suggest that PCSK9i reduce all LDL subfractions, though with mildly greater proportions of large LDL particles compared to small (11–14).

Oxysterols have long been implicated in atherosclerosis (15, 16). While PCSK9i dramatically decrease circulating LDL, it is unknown whether the therapeutic anti-PCSK9 monoclonals impact plasma oxysterol concentrations. Certain oxysterols, such as 7-ketocholesterol (7-KC) (17, 18), which is considered a stable oxysterol, accumulate primarily through non-enzymatic reactions between cholesterol and reactive oxygen species (ROS), and when abundantly present give rise to measurable levels of oxidized LDL (oxLDL). Additionally, 7-KC may also enter the circulation from dietary ingestion and intestinal absorption (19). In humans, 7-KC is the predominant oxysterol found on oxLDL (20–24) as well as in human plasma and carotid plaque (22, 25, 26). Moreover, elevated plasma 7-KC concentrations have been associated with ASCVD risk (27–29). Most of these studies have measured free 7-KC, although circulating 7-KC is predominantly present in its esterified form.

It is not known whether drastic reductions in LDL-C levels are associated with changes in levels of oxidized cholesterol and whether sdLDL particles are more likely to carry these oxidized cholesterol forms. In this study, we set out to investigate the effect of LDL-C reduction accomplished via inhibition of PCSK9 on changes in the levels of sdLDL-C and of a common form of oxidized cholesterol, free 7-KC, in a group of patients referred to our Preventive Cardiology clinic.

Material and Methods

Study cohort

The study cohort included patients enrolled in the Oregon Health & Science University (OHSU) Center for Preventive Cardiology Registry and Biorepository (IRB #00018643) between October 2016 and January 2019. Patients were referred from within OHSU or by outside providers to the PCSK9i clinic at the OHSU Center for Preventive Cardiology. Patient demographics and clinical characteristics were collected prior to the initiation of PCSK9i. Individuals evaluated at the PCSK9i clinic undergo structured clinical visits prior to and after PCSK9i initiation with plasma lipid measurements obtained at baseline and after at least 3 doses of PCSK9i therapy, as previously described (30, 31). To be included in the present analysis, participants needed to meet the following inclusion criteria: (a) evaluated in the OHSU PCSK9i clinic, (b) initiated on a PCSK9i, (c) have plasma lipid panel measurements before and at least 6 weeks (at least 3 doses of PCSK9i) after initiation of PCSK9i, and (d) have either plasma sdLDL-C or 7-KC measurements before and at least 6 weeks (at least 3 doses of PCSK9i) after initiation of PCSK9i.

Lipid measurements

Fasting plasma lipid panel measurements were completed by standard techniques in the core and clinical lipid laboratories at OHSU, as previously described (32). Plasma lipid measurements included total cholesterol, triglycerides (TG), high-density lipoprotein-cholesterol (HDL-C), very low-density lipoprotein-cholesterol (VLDL-C), and LDL-C. LDL-C values were calculated using the Friedwald formula (33) and VDL-C values were calculated from TG values (VLDL-C = TG / 5). LDL-C was directly measured in patients with TG greater than or equal to 300 mg/dL. Plasma Lp(a) was quantified using a turbidimetric assay where a latex-sensitized antibody to kringle 4 was added to serum to form an insoluble aggregate. This increases the turbidity in the solution, which was measured optically at 700 nm and quantified in comparison to standards (Polymedco, Cortland Manor, NY). The lower limit of detection of Lp(a) was 6 mg/dL and measurements <6 mg/dL were recorded as the median level of 3 mg/dL.

Plasma sdLDL-C measurements were performed using the automated homogenous Denka Seiken assay, as previously described (34). In brief, the enzyme sphingomyelinase is directly used to dissociate large buoyant LDL from the whole LDL spectrum, while a polyoxyethylene styrenephenyl ether derivative (surfactant B) protects sdLDL during the dissociation. A polyoxyethylene alkyl ether is then used to dissociate sdLDL-C, which is then measured through the use of commercial cholesterol oxidase/esterase.

Plasma free 7-KC measurements were determined by liquid chromatography-mass spectrometry (LC-MS) following protein precipitation and extraction with acetonitrile and derivatization with the quaternary aminooxy mass tag reagent (commercially available as Amplifex Keto Reagent, Sciex) used by Star-Weinstock et al. in the analysis of testosterone (35). Full details of plasma 7-KC quantification can be found in supplemental material. Use of the OHSU Bionalytical Shared Resource/PK Core Facility (Research Resource Identifier (RRID): SCR_009963) for analysis of 7-KC is acknowledged.

Statistical analysis

The baseline characteristics of study participants were described as count (proportion) for categorical variables and continuous variables were presented as mean (standard deviation) or median [interquartile range], depending on variable distribution. The differences in lipid values pre- and post-PCSK9i were assessed by paired t-test and the Wilcoxon signed rank test as appropriate. Pearson correlation and Spearman correlation coefficients, R, were calculated for percent changes in various lipid fractions following PCSK9i. One outlier was removed from the analysis based on z-score >3 (see results section). All statistical analyses were performed with R Programming Language (Version 4.2.2). A p-value threshold <0.05 was considered statistically significant.

Results

The final study cohort consisted of 134 patients. Baseline characteristics of the study cohort are detailed in Table 1. The mean age of the cohort was 62.8 (SD 10.7) years and 47.8% (n=64) were female. Participants exhibited a high prevalence of hypertension (67.9%; n=91) and ASCVD (91%; n=122). Specifically, 83.6% (n=122) had a history of coronary artery disease. Statin therapy was present at baseline in 44.8% (n=60) of participants.

Table 1 :

Baseline Characteristics of Study Cohort

	Study Cohort (n=134)
Age, years, mean (SD)	62.8 (10.7)
Sex, male, n (%)	70 (52.2%)
BMI, kg/m2, mean (SD)	29.4 (5.3)
Hypertension, n (%)	91 (67.9%)
Diabetes mellitus, n (%)	17 (12.7%)
Current tobacco use, n (%)	10 (7.5%)
Coronary artery disease, n (%)	112 (83.6%)
Cerebrovascular disease, n (%)	5 (3.7%)
Peripheral arterial disease, n (%)	9 (6.7%)
Carotid artery disease, n (%)	21 (15.7%)
Statin use at baseline, n (%)	60 (44.8%)

SD, standard deviation; BMI, body mass index

Baseline, post-PCSK9i, and percent change in lipid values after at least 6 weeks of therapy are detailed in Table 2. Of the 134 patients in the cohort, 123 had sdLDL-C measurements and 105 had 7-KC measurements before and after PCSK9i. A total of 94 patients had both sdLDL-C and 7-KC measurements before and after PCSK9i. In total, 130 patients in the study cohort had Lp(a) measurements before and after PCSK9i. Mean LDL-C at baseline was 138.3 (54.8) mg/dL and decreased by 55.4% to 63.2 (48.9) mg/dL after PCSK9i. Mean baseline sdLDL-C decreased from 46.5 (22.7) mg/dL to 22.7 (16.4) mg/dL after PCSK9i, representing a mean 51.4% reduction. The sdLDL-C measurements for one patient increased from 11.3 mg/dL to 75.2 mg/dL (63.9% increase) after PCSK9i; this clear outlier (z-score = 4.7) was excluded from the analysis. PCSK9i mildly increased HDL-C by 8.4% (14.5), while reducing TG and VLDL-C by 23.6% [−38.1–0] as well as Lp(a) by 15.8% [−33.1–0]. Median baseline plasma free 7-KC decreased modestly by 6.6% from 6 [4.3–10.3] ng/mL to 5.1 [3.7–9.7] ng/mL following PCSK9i. All lipid values measured except 7-KC (p = 0.117) had a statistically significant change after PCSK9i therapy.

Table 2 :

Lipid Values Pre- and Post-PCSK9 Inhibitor Therapy

	Pre-PCSK9i	Post-PCSK9i	Percent Change with PCSK9i
LDL-C, mg/dL, mean (SD)	138.3 (54.8)	63.2 (48.9)	−55.4% (22.9)**
HDL-C, mg/dL, mean (SD)	53.7 (20.5)	57.4 (20.4)	8.4% (14.5)**
TG, mg/dL, median [IQR]	152.5 [105–225]	125 [85–170]	−23.6% [−38.1–0]**
VLDL-C, mg/dL, median [IQR]	30.5 [21–45]	25 [17–34]	−23.6% [−38.1–0]**
Lp(a), mg/dL, median [IQR]	35 [12.8–94]	25 [7–81.8]	−15.8% [−33.1–0]**
sdLDL-C, mg/dL, mean (SD)	46.5 (22.7)	22.7 (16.4)	−51.4% (24.8)**
7-KC, ng/mL, median [IQR]	6 [4.3–10.3]	5.1 [3.7–9.7]	−6.6% [−42.1–32]

^*Change is statistically significant at p < 0.05 level;

^**Change is statistically significant at p < 0.01 level.

PCSK9i, proprotein convertase subtilisin-kexin type 9 inhibitor; LDL-C, low-density lipoprotein-cholesterol; SD, standard deviation; HDL-C, high-density lipoprotein-cholesterol; TG, triglycerides; VLDL-C, very low-density lipoprotein-cholesterol; interquartile range; Lp(a), lipoprotein (a); sdLDL-C, small dense LDL-C; 7-KC, 7-ketocholesterol.

Correlations between baseline and post-PCSK9i percent reductions in lipid values are presented in Table 3. Percent reduction in LDL-C demonstrated a strong correlation with percent reduction in sdLDL-C (R=0.829, p<0.001; Figure 1). The percent reduction in LDL-C demonstrated a statistically significant, though weak, correlation with percent reduction in Lp(a) (R=0.28, p=0.001), but did not otherwise significantly correlate with percent changes in other lipid parameters, including 7-KC. As depicted in Figure 2, no correlation was observed between percent change in LDL-C and percent change in 7-KC after PCSK9i (R=0.102, p=0.301). Percent reductions in TG and VLDL-C correlated modestly with both percent reductions in sdLDL-C (R=0.252, p=0.005) and percent reductions in 7-KC (R=0.219, p=0.025). The weak but statistically significant correlation between percent changes in VLDL-C and percent changes in 7-KC is shown in Figure 3. Percent reduction in Lp(a) did not significantly correlate with percent reductions in 7-KC (R=0.095, p=0.344).

Table 3:

Correlations Between Percent Reductions in Measured Lipid Fractions Following PCSK9 Inhibitor Therapy

		LDL-C	TG VLDL-C	VLDL-C	Lp(a)	sdLDL-C	7-KC
LDL-C		1	‐		‐	‐	‐
TG	VLDL-C	0.06	1		‐	‐	‐
Lp(a)		0.28**	0.10		1	‐	‐
sdLDL-C		0.83**	0.25**		0.26**	1	‐
7-KC		0.10	0.22*		0.09	0.11	1

^*Pearson or Spearman correlation is statistically significant to p < 0.05 level;

^**Pearson or Spearman correlation is statistically significant to p < 0.01 level.

TG and VLDL-C levels are placed together since VLDL-C levels were calculated from TG levels. LDL-C, Low-density lipoprotein-cholesterol; TG, triglycerides; VLDL-C, very low-density lipoprotein-cholesterol; Lp(a), lipoprotein (a); sdLDL-C, small dense LDL-C; 7-KC, 7-ketocholesterol.

Figure 1 : Relationship Between Percent Change in LDL-C and Percent Change in sdLDL-C Following PCSK9 Inhibitor Therapy.

Pearson correlation R and p-value are shown with linear regression line. LDL-C, low-density lipoprotein cholesterol; sdLDL-C, small dense LDL cholesterol.

Figure 2 : Relationship Between Percent Change in LDL-C and Percent Change in 7-KC Following PCSK9 Inhibitor Therapy.

Spearman correlation R and p-value are shown with linear regression line. 7-KC, 7-Ketocholesterol; LDL-C, low-density lipoprotein cholesterol.

Figure 3 : Relationship Between Percent Change in VLDL-C and Percent Change in 7-KC Following PCSK9 Inhibitor Therapy.

Spearman correlation R and p-value are shown with linear regression line. 7-KC, 7-Ketocholesterol; VLDL-C, very low-density lipoprotein cholesterol.

Discussion

The purpose of this study was to evaluate the effect of PCSK9i on plasma lipid parameters beyond the standard lipid panel, specifically sdLDL-C and 7-KC. Secondarily, we aimed to determine if changes in sdLDL-C and 7-KC were related to changes in LDL-C. This analysis has two principal findings. First, PCSK9i robustly reduced both LDL-C and sdLDL-C by 55.4% (22.9) and 51.4% (24.8), respectively. Reductions in sdLDL-C strongly correlated with reductions in LDL-C, suggesting that the sdLDL subfraction is reduced by PCSK9 to the same extent as larger LDL. Secondly, and surprisingly, PCSK9i did not significantly influence plasma free 7-KC concentrations. Interestingly, TG and VLDL-C were the only lipid parameters whose change in therapy significantly correlated with changes in 7-KC, suggesting that a substantial amount of plasma 7-KC may be transported by VLDL and subsequently lost during lipoprotein hydrolysis leading to LDL formation.

It has long been known that atherogenesis is dependent on not only the cholesterol cargo but also the number and size of circulating LDL. Specifically, while all LDL particles are atherogenic, sdLDL has emerged as the LDL subclass with the greatest atherogenic potential (6, 7). The increased atherogenicity of sdLDL is believed to be multifactorial, including smaller size to penetrate the subendothelial space, increased affinity for arterial intimal proteoglycans, decreased affinity for LDLR leading to prolonged circulation time, and increased propensity for atherogenic modification such as oxidation, desialylation, and glycation (6, 7). Accordingly, several large studies have shown that sdLDL-C is an independent risk factor for ASCVD (9, 10, 36–39). While there are multiple methodologies for measurement of sdLDL and no established ‘gold standard’ method to date (40), the relatively new automated homogenous Denka Seiken assay (34) provides rapid clinical results and has been consistently used in large observational studies independently associating sdLDL-C with ASCVD (9, 10, 36–39). No specific sdLDL lowering therapy currently exists. Statins reduce risk of ACSVD through LDL-lowering but are associated with greater proportional reductions in large LDL over sdLDL, though this seems to be dependent on which statin is used (41, 42). In this context, the finding that PCSK9i are associated with substantial reductions in sdLDL would have pragmatic clinical consequences.

Our study demonstrates that PCSK9i drastically reduces plasma sdLDL-C. Multiple previous studies suggest mildly increased clearance of large compared to small LDL particles through PCSK9i (12–14), which was attributed to sdLDL’s lower affinity for the LDLR (43). Koren et al. (13) found that large LDL particles were reduced more than small LDL particles (−71.3% vs −54%) in 26 patients taking alirocumab. In a post hoc analysis involving 619 patients from the DESCARTES (Durable Effect of PCSK9 Antibody Compared with Placebo Study) trial, Toth et al. (14) found similar results with a preferential reduction in large LDL particles (−73.7% vs −35.4%) under evolocumab therapy. On the other hand, another recent study found that small LDL particles were preferentially reduced (76.8% vs 38.3% large-size LDL particles) in patients with recent acute coronary syndrome who were on combination evolocumab and statin therapy versus statin monotherapy (44). The aforementioned studies used nuclear magnetic resonance spectroscopy to measure small and large LDL particles while our study measured sdLDL-C using a homogenous assay. Nevertheless, our study is consistent with prior work in that small LDL is profoundly reduced via PCSK9i.

Oxysterols have also been implicated in atherosclerosis (15, 16) with 7-KC thought to be a central player (17, 18). Inflammation and oxidation are well-established steps in atherosclerotic plaque formation, particularly as oxidized LDL is not subject to feedback regulation for uptake by arterial wall macrophages via scavenger receptors (45). Cholesterol itself is very vulnerable to oxidation, and detrimental oxysterols can form spontaneously when cholesterol-rich foods are heated or stored. Oxysterols are also formed in vivo under conditions of oxidative stress (15, 16, 19). Oxysterols are generally difficult to accurately measure in laboratory conditions since cholesterol is eminently susceptible to oxidation and oxysterols are inherently unstable. In this study, nonenzymic oxidation of 7-KC during isolation was minimized through analysis of free 7-KC (because hydrolysis can contribute to 7-KC oxidation) and through use of rapid sample preparation and LC-MS analysis methodology (provided in supplemental information). The levels of 7-KC measurement in our study (6 ng/mL to 5.1 ng/mL after PCSK9i) are congruent with prior studies. Mean concentration of free 7-KC for 314 healthy controls has been previously reported to be 5.2 ng/ml (range nondetectable to 23 ng/ml) (46). Mean concentration for total (hydrolyzed) 7-KC has been previously reported to be >10-times higher at 84 ng/ml (47). While 7-KC is one of the most common oxysterols found in food (15, 19), endogenous 7-KC is predominantly formed via non-enzymatic auto-oxidation when ROS react at the C7 position on cholesterol. It can also be formed to a lesser extent enzymatically through cytochrome P450 7A1 triggered by oxidative stress or from 7-dehydrocholesterol (48, 49). Once formed, 7-KC appears to be cytotoxic to a variety of cells, including macrophages, by disrupting plasma membranes and inducing oxidative stress and apoptosis (17, 18). There is now mounting evidence that links 7-KC with atherosclerosis (17, 18). Free plasma 7-KC concentrations are associated with ASCVD (27, 28). While data is inconsistent regarding the prevalence of oxysterols in plasma and atherosclerotic lesions, 7-KC was recently demonstrated to be the most common cytotoxic free oxysterol in human plasma and carotid plaque (25). Most importantly, 7-KC is considered to be the most abundant oxysterol in oxLDL (15, 16, 20–24). OxLDL, formed by various oxidative modifications to apoB and lipids on LDL, has long been considered to have an integral role in the inflammatory process leading to atherosclerotic lesions (50, 51), and circulating oxLDL has been associated with ASCVD events (52). Despite containing many oxidized lipids, it remains unknown if oxLDL is the primary source of circulating oxysterols such as 7-KC. There is also minimal data in general about lipoprotein trafficking of 7-KC.

Our study demonstrates that PCSK9i do not significantly change circulating free 7-KC levels despite producing striking reductions in LDL-C and sdLDL-C. Surprisingly, the only significant correlations to changes in free 7-KC were changes in TG and VLDL-C, suggesting a potential mechanism whereby dietary 7-KC is absorbed with intestinal chylomicrons and repackaged by the liver in the endogenous pathway but lost in the plasma before processing of VLDL to LDL. For every 6.6% reduction in 7-KC, we observed corresponding 23.6% reductions in VLDL-C (R=0.219, p=0.025). Despite 7-KC being one of the most abundant oxysterols found in diet (15, 19), the amount of dietary 7-KC that enters circulation is thought to be relatively low as previous animal studies suggest rapid hepatic metabolism (53–55). Nonetheless, circulating 7-KC levels are very low in proportion to cholesterol levels (~0.1%) and even minute levels of 7-KC entering circulation via dietary means are likely significant. Given the significant association between changes in VLDL-C and 7-KC, and the absence of association with other lipoproteins and 7-KC, our data suggests that a noteworthy portion of circulating 7-KC may reside on remnant lipoproteins. This would agree with prior animal studies which demonstrate that dietary 7-KC is more selectively transported by triglyceride-rich lipoproteins (including VLDL) rather than LDL (56, 57). The relative proportion of circulating 7-KC that has a dietary origin versus endogenous oxidation cannot be discerned in our analysis. Interestingly, despite drastic reductions in LDL-C via PCSK9i, serum free 7-KC levels were not significantly reduced and changes in circulating 7-KC and LDL-C concentrations demonstrated no significant association.

Our study has limitations that need to be acknowledged. Given the observational nature of our study, the findings should be considered exploratory and hypothesis-generating, and further mechanistic studies are required to corroborate our findings. Additionally, although reasonable steps were taken to reduce nonenzymatic oxidation of sterols during isolation, we cannot rule out that 7-KC measured here represents nonspecific oxidation product. Furthermore, we did not track changes in lifestyle, weight, or other medications that could potentially influence some of the lipid parameters. Our study also has several notable strengths. Serum samples were obtained prospectively in a relatively large real-world clinical cohort of patients under standard of care PCSK9i therapy. Additionally, there is a paucity of data regarding the impact of PCSK9i on non-standard lipid fractions. Specifically, to our knowledge, this is the first report that has evaluated the effect of PCSK9i on 7-KC.

In conclusion, we demonstrate that PCSK9i dramatically reduces sdLDL-C, effectively mirroring the reductions in LDL-C. On the other hand, these reductions in LDL-C and sdLDL-C under PCSK9i therapy do not result in changes in plasma levels of free 7-KC. However, changes in 7-KC did correlate with changes in VLDL-C, suggesting that a significant portion of 7-KC is carried by VLDL and lost during lipoprotein processing leading to LDL formation. Further mechanistic studies are needed to clarify the relationship between serum 7-KC and VLDL.

Highlights:

• PCSK9 inhibition dramatically reduces both LDL-C and sdLDL-C to a similar degree

• Plasma 7-ketocholesterol levels were not significantly modulated by PCSK9 inhibition

• Changes in 7-ketocholesterol and VLDL-C after PCKS9 inhibition significantly correlated

• 7-Ketocholesterol may be carried by VLDL and lost during LDL formation

Abbreviations:

ASCVD

Atherosclerotic cardiovascular disease

7-KC

7-ketocholesterol

Lp(a)

lipoprotein (a)

PCSK9i

proprotein convertase subtilisin/kexin type 9 inhibitor

sdLDL-C

small dense LDL-cholesterol

TG

triglyceride

Data availability statement:

All data are contained within the manuscript.