This study investigated the relationship between coronary collateral circulation (CCC) and non–high-density lipoprotein cholesterol (non–HDL-C) in patients with stable coronary artery disease (CAD). Coronary collateral circulation plays a critical role in supporting blood flow, particularly in the ischemic myocardium. Previous studies show that non–HDL-C plays a more important role in the formation and progression of atherosclerosis than do standard lipid parameters.
A total of 226 patients with stable CAD and stenosis of more than 95% in at least 1 epicardial coronary artery were included in the study. Rentrop classification was used to assign patients into group 1 (n = 85; poor collateral) or 2 (n = 141; good collateral). To adjust for the observed imbalance in baseline covariates between study groups, propensity-score matching was used. Covariates were diabetes, Gensini score, and angiotensin-converting enzyme inhibitor use.
In the propensity-matched population, the plasma non–HDL-C level (mean [SD], 177.86 [44.0] mg/dL vs 155.6 [46.21] mg/dL; P = .001) was statistically higher in the poor-collateral group. LDL-C (odds ratio [OR], 1.23; 95% CI, 1.11–1.30; P = .01), non–HDL-C (OR, 1.34; 95% CI, 1.20–1.51; P = .01), C-reactive protein (OR, 1.21; 95% CI, 1.11–1.32; P = .03), systemic immune-inflammation index (OR, 1.14; 95% CI, 1.05–1.21; P = .01), and C-reactive protein to albumin ratio (OR, 1.11; 95% CI, 1.06–1.17; P = .01) remained independent predictors of CCC in multivariate logistic regression analysis.
Non–HDL-C was an independent risk factor for developing poor CCC in stable CAD.
Abbreviations and Acronyms
coronary artery disease
CRP to serum albumin ratio
coronary collateral circulation
high-density lipoprotein cholesterol
low-density lipoprotein cholesterol
- SII index
systemic immune-inflammatory index
Coronary collateral circulation (CCC) is an adaptive mechanism that occurs in the presence of chronic myocardial ischemia and stress after severe stenosis or obstruction.1 Previous studies have shown that a well-developed collateral circulation reduces the infarct area, protects cardiac functions, reduces the development of cardiogenic shock, decreases the development of aneurysm of the left ventricle, and plays a very important role in short- and long-term prognoses for coronary artery disease (CAD).2
Despite many studies, the factors affecting coronary collateral development cannot be fully explained; however, the normal vascular endothelial layer and its functions are accepted as the basis for collateral development.3,4 Previous studies on dyslipidemia, metabolic syndrome, diabetes mellitus (DM), and aging have shown the adverse effects of vascular endothelial dysfunction that result from decreasing collateral development.5–7
Non–high-density lipoprotein cholesterol (non–HDL-C) contains multiple atherogenic cholesterols, including low-density lipoprotein cholesterol (LDL-C), very low-density lipoprotein, intermediate-density lipoprotein cholesterol, and lipoprotein a [Lp(a)] cholesterol.8 It shows the total atherogenic burden better than does LDL-C, which is the primary cholesterol target.9,10 In addition to providing short- and long-term information for many diseases, such as CAD, various studies have shown that it is an early indicator of vascular endothelial dysfunction and is correlated with various inflammatory markers.11,12
This study aimed to investigate the relationship between CCC and non–HDL-C levels in patients with stable CAD.
Patients and Methods
This retrospective cross-sectional study was conducted in patients who underwent coronary angiography between December 2016 and December 2020 at a single center after ethics committee approval (Adnan Menderes University Non-Interventional Clinical Research Ethics Committee permission dated December 10, 2021, and numbered 2021/65). Patients who were diagnosed with stable CAD according to the criteria recommended by the European Society of Cardiology and who underwent coronary angiography and were determined to have coronary stenosis of 95% or more in at least 1 major coronary vessel were included in the study. All patients included in the study had angina or angina-equivalent symptoms, and coronary angiography indications were established with positive noninvasive tests (exercise stress test, stress echocardiography, and myocardial perfusion scintigraphy). The study was conducted in accordance with the Declaration of Helsinki.
Exclusion criteria for the study were as follows: anemia, kidney failure or kidney disease affecting kidney clearance, history of coronary bypass surgery, malignant disease, collagenous connective tissue disease, congestive heart failure, acute cerebrovascular disease, percutaneous coronary intervention after previous acute coronary syndrome, fever, active infection, and the use of statins or other antihyperlipidemic drugs. After applying the exclusion criteria, 226 patients were enrolled in the study.
The demographic, clinical, and angiographic characteristics of all patients included in the study were recorded. All patients were interviewed in detail about hypertension, DM, smoking, history of myocardial infarction, family health history, and medication use. Chronic kidney failure was defined as a glomerular filtration rate of less than 60 mL/min for over 3 months. A diagnosis of hypertension was accepted if antihypertensive treatment was given or if there were at least 3 measurements of higher than 140 mm Hg systolic and 90 mm Hg diastolic. A diagnosis of DM was made if patients were taking antidiabetic medication or had at least 2 fasting blood glucose measurements above 126 mg/dL.
Echocardiographic examination of all patients included in the study was performed using an iE33 cardiac ultrasound system (Phillips Healthcare) and a 2.5- to 5-MHz probe system. Ejection fraction was measured using the modified Simpson method.
Evaluation of Coronary Angiography
Coronary angiography was performed using the Allura Xper FD 10 (Phillips Healthcare). Angiography was performed by puncture of the femoral artery with 6F Judkins standard right and left catheters. Iodixanol was used as the radiological contrast agent. At least 4 projections for the left coronary system and at least 2 projections for the right coronary system were recorded in digital and analog formats. Two independent cardiologists examined the coronary angiography results without knowing patient characteristics. The culprit vessel (stenosis of ≥95%) areas were divided and recorded according to 3 groups: right coronary artery, left coronary artery, and circumflex coronary artery. Rentrop classification was used as a basis for the evaluation of CCC. Rentrop grade 0 is accepted as the absence of collateral flow, grade 1 is defined as the presence of side branches without an occluded main coronary artery, grade 2 is defined as partial visibility of an occluded main coronary artery, and grade 3 is defined as complete visibility of an occluded main coronary artery. Consistent with previous studies, Rentrop grades 0 and 1 were accepted as poor collateral flow, whereas Rentrop grades 2 and 3 were accepted as good collateral flow.13
Evaluation of CAD Severity
Gensini scoring was used to grade the extent and severity of atherosclerosis in the coronary arteries for all patients. According to the degree of angiographic stenosis, 1 point was given for 1% to 25% stenosis, 2 points for 26% to 50% stenosis, 4 points for 51% to 75% stenosis, 8 points for 76% to 90% stenosis, 16 points for 91% to 99% stenosis, and 32 points for 100% total lesions. The Gensini score was calculated by multiplying these scores with the coefficient defined for each segment of the coronary arteries and summing the results. The segments and their coefficients were multiplied by 5 for the left main coronary artery, 2.5 for the proximal segment of the left coronary artery, 1.5 for the middle, 1 for the apical, 1 for diagonal 1 and 0.5 for diagonal 2, 2.5 for the proximal segment of the circumflex coronary artery, 1 for the distal, 1 for the obtuse marginal, 1 for the posterior descending artery if left dominant, 0.5 for the posterolateral artery, 1 for the right coronary artery proximal, 1 for the middle, 1 for the distal, and 1 for the posterior descending artery.14
Analysis of Lipid Profile and Other Biochemical Values
Blood samples from all patients who participated in the study were collected after overnight fasting. The LDL-C, HDL-C, non–HDL-C, and total cholesterol (TC) levels of the patients were measured. The non–HDL-C level was measured by subtracting the HDL-C level from the TC level. In addition, the patients' other hematological and biochemical values were measured and recorded.
The systemic immune-inflammatory index (SII index), platelet to lymphocyte ratio, neutrophil to lymphocyte ratio, and C-reactive protein (CRP) to serum albumin ratio (CAR) were determined from patients' blood counts and biochemical values. The SII index was calculated according to the formula platelet count × neutrophil count / lymphocyte count.
SPSS Statistics for Windows (version 25.0; IBM) and Amos (version 24.0; IBM) statistical packages were used to analyze the data. Descriptive statistics (mean [SD], median [IQR], No. [%]) for categorical and continuous variables were reported. Homogeneity of variances, one of the assumptions of parametric tests, was tested using the Levene test. The normality assumption was tested using the Shapiro–Wilk test. To assess differences between the 2 groups, the independent t test was used when the assumptions of the parametric tests were met, whereas the Mann-Whitney U test was used when they were not. The relationship between the 2 continuous variables was assessed using the Pearson correlation coefficient, and when the conditions for the parametric test were not met, the Spearman correlation coefficient was calculated. P < .05 was considered statistically significant. Univariate analysis was used to calculate the association of different variables with CCC. Variables for which the unadjusted P value in the logistic regression model was <.05 were identified as potential risk markers and included in the full multivariate model.
To attenuate the observed imbalance in baseline covariates between the study groups, a propensity-score (PS) matching technique was used. A 1-to-1 pair-matching method with a specified tolerance distance was used to identify matched cohorts. This method resulted in a PS-matched population with 84 patients in the poor-collateral group and 84 patients in the good-collateral group.
A total of 226 patients were included in this study. According to Rentrop classification, groups were divided into those that had good (Rentrop grades 2–3) or poor (Rentrop grades 0–1) collateral. In the original study population, the median age of the 85 patients (68 males, 17 females) in the poor-collateral-circulation group was 67 (65.1–74.50) years. The mean age of the 141 patients (101 males, 40 females) in the good-collateral-circulation group was 69 (63.50–78) years.
Baseline clinical, demographic, and angiographic features are compared in Table I. Diabetes status, Gensini score, and use of angiotensin-converting enzyme inhibitors differed substantially between the 2 groups. Because of the differences in key baseline characteristics, the PS-matching method was used. In the 1-to-1 PS-matched data set, 84 patients remained in the good-collateral group and 84 patients remained in the poor-collateral group. All the variables were well balanced between the groups after the PS match (Table I).
In the PS-matched population, there was no difference between groups in terms of hematological and kidney function values. The CRP levels, SII index, and CAR were increased in the poor-collateral group (Table II). In a comparison of the two groups in terms of lipid parameters, there were no differences in the triglyceride and HDL-C levels. Low-density lipoprotein cholesterol, TC, and non–HDL-C levels were increased in the poor-collateral group in the PS-matched population (Table III).
Multivariate logistic regression analysis showed that LDL-C, non–HDL-C, CRP, SII index, and CAR remained independent predictors of poor-collateral development (Table IV).
This study is the first to examine the relationship between CCC and non–HDL-C in patients with stable CAD. In this study, high non–HDL-C was an independent risk factor for developing poor collaterals in patients with stable CAD.
Similar to previous studies, this study also supports the findings that LDL-C, CRP, SII index, and CAR are independent risk factors for poor-collateral development. A high LDL-C level, which is one of the most critical risk factors for developing endothelial dysfunction and atherosclerosis, is a risk factor for the development of poor collateral in many clinical and demographic studies. You et al15 showed that high LDL-C levels are an independent risk factor for developing poor collateral. In in vitro studies, CRP, one of the most noteworthy markers of inflammation, has been shown to reduce nitric oxide (NO) levels, which is a very important element of collateral development and angiogenesis inhibition. Gulec et al16 showed that high CRP levels are an independent risk factor for poor-collateral development. According to Fan et al17 in their study of 1,158 patients, high CRP levels were associated with poor-collateral development in patients with stable CAD. The SII index and CAR are inflammatory markers that have emerged in recent years, and various studies18,19 have shown them to be more predictive markers for inflammation than the platelet to lymphocyte ratio, the neutrophil to lymphocyte ratio, and CRP. According to Keleşoğlu et al,20,21 high SII index and CAR levels in patients with stable CAD are risk factors for developing poor collateral.
Although most studies do not fully explain the mechanism of collateral development, most agree that it is a complex process of cell organization.1,2 In previous studies, angiogenesis and arteriogenesis have been shown to be the 2 primary mechanisms for collateral development. Angiogenesis involves the coordinated migration, proliferation, and differentiation of endothelial cells and pericytes from existing vascular beds. On the other hand, arteriogenesis is the growth of muscular arteries requiring similar events regulated by endothelial cells and smooth muscle cells from preexisting arteries.22 A solid and functional vascular endothelial layer is essential for the healthy continuation of angiogenesis and arteriogenesis processes.23 The presence of vascular endothelial dysfunction is accepted as one of the main reasons for poor-collateral development. Numerous studies have shown that hyperlipidemia causes vascular endothelial dysfunction and adversely affects collateral development.24,25 Aras et al26 showed that at high Lp(a) levels, vascular endothelial growth factor secretion is negatively affected by Lp(a) and collateral development is poor as a result of endothelial dysfunction. Morishita et al27 showed that transforming growth factor-β release is decreased because of endothelial dysfunction resulting from high Lp(a), and collateral development is adversely affected. Duan et al28 demonstrated that collateral vessel formation and angiogenesis in response to hindlimb ischemia were significantly attenuated in rats with dietary hypercholesterolemia, which is related to endothelial cell dysfunction and decreased endothelium-derived NO.28 Shen et al29 found that in patients with DM, it was observed that Lp(a) levels were correlated with LDL-C and non–HDL-C; in addition, poor-collateral development was 4 times higher with high Lp(a) levels.
Many recent studies have shown that non–HDL-C is a good indicator for cardiovascular diseases compared with LDL-C, which is the primary target.8–10 It is also the primary indicator for endothelial dysfunction and is associated with various inflammatory markers. The mechanism of poor-collateral development with high non–HDL-C is thought to occur from more than 1 factor. Hyperlipidemia, especially with high non–HDL-C, may adversely affect collateral development by causing vascular endothelial dysfunction. Wang and Chang12 stated that non–HDL-C was an early marker of vascular endothelial dysfunction in patients with type 2 diabetes and was correlated with CRP. In a study that investigated serum lipid levels and the risk of microangiopathy in patients with type 2 diabetes, Toth et al30 showed that each 1-mg/dl increase in non–HDL-C increased the risk of microangiopathy by 0.3%. In addition, the risk level was observed to be 17.3% in patients who were above the non–HDL-C target. Karasek et al31 found a positive correlation between non–HDL-C and high-sensitivity CRP, which is an inflammatory marker; C-peptide and homeostatic model assessment for insulin resistance, which is a marker of insulin resistance; and plasminogen activation inhibitor. In addition, many studies have shown that the presence of high non–HDL-C is an independent risk factor for increased arterial stiffness.32 de Oliveira Alvim et al33 showed that high non–HDL-C is associated with increased arterial stiffness; as a result of this increased stiffness, high systolic blood pressure, low diastolic blood pressure, and increased pulse pressure develop. Baykan et al34 showed that coronary perfusion and shear stress decrease as a result of increased arterial stiffness and adversely affect coronary collateral arteriogenesis.
This study has some limitations. The study's main limitations are that it is based on a single center with a small sample and it is a retrospective study. Apolipoprotein B levels are a better indicator than non–HDL-C of atherogenic burden, but were not measured in this study. Vascular endothelial growth factor, fibroblast growth factor, NO, and interleukin levels, which are very important in pathophysiology, are also not measured. Another limitation is the lack of short- and long-term follow-up of the patients.
This study demonstrates that high non–HDL-C levels negatively affect the development of stable CAD collateral. With this result, one can understand the pathophysiology of collateral development and new treatment strategies can be developed. Large and long-term prospective studies may help us better understand the diagnostic and therapeutic value of non–HDL-C in collateral development.
Conflict of Interest/Disclosure: None