Mildly Decreased Glomerular Filtration Rate Is Associated With Poor Coronary Collateral Circulation in Patients With Coronary Artery Disease

Abstract

Background:

The aim of this study was to evaluate the association between mildly decreased glomerular filtration rate (GFR) and coronary collateral circulation (CCC).

Hypothesis:

There would be an association between mildly decreased GFR and CCC.

Methods:

Patients who had an occlusion in at least 1 major coronary artery were included in this study. Patients with severely and moderately decreased GFR were excluded. Patient data were obtained from their files. To classify CCC, we used the Rentrop classification. Patients were classified as having poor CCC (Rentrop grades 0 to 1) or good CCC (Rentrop grades 2 to 3). We used the Modification of Diet in Renal Disease (MDRD) equation to calculate GFR. Mildly decreased GFR was defined as 60 mL/min per 1.73 m² ≥ eGFR ≤89 mL/min per 1.73 m² according to the MDRD definition. Multivariate logistic regression analysis was performed to determine independent variables.

Results:

The study group consisted of 299 patients. Ninety‐three patients had poor CCC and 206 patients had good CCC. The frequency of mildly decreased GFR was higher in the poor CCC group than in the good CCC group (P<0.001). Also, the frequency of diabetes and dyslipidemia, and the plasma high sensitive C‐reactive protein levels, were higher in the poor CCC group (P = 0.003, P = 0.018, P<0.001, respectively). Logistic regression analysis revealed that eGFR is an independent predictor of CCC (B = 1.68; odds ratio = 5.4; P<0.001; 95% confidence interval, 3.1–9.4).

Conclusions:

We found that CCC was worse in patients with mildly decreased GFR compared to patients with normal GFR in patients with coronary artery disease. © 2011 Wiley Periodicals, Inc.

The authors have no funding, financial relationships, or conflicts of interest to disclose.

Introduction

In the normal human heart, there are many collateral vessels linking major coronary arteries.1 Because most of the collateral channels in normal people or patients with mild coronary artery disease are small and carry only minimal flow, they are not seen on coronary angiography. For collateral vessels to show up on coronary angiography, there has to be 99% to 100% stenosis of the coronary artery.2 Coronary collateral circulation (CCC) is an alternative mode of blood supply to an ischemic region in coronary artery disease. It has been shown that well‐developed collateral circulation limits the area affected by an infarction,3 decreases both the frequency of cardiogenic shock after myocardial infarction4 and the development of a left ventricular aneurysm.5 Developments in molecular biology and findings from genetic studies have provided some insight into how CCC develops, but the exact mechanism is unknown. Although currently, the only known predictor of the development of CCC is the degree of coronary artery stenosis,6 and there are significant differences seen even among patients with a similar degree of coronary artery disease.

There is considerable evidence that chronic renal disease is independently associated with coronary artery disease. Several studies suggest that mild to moderate elevations in serum creatinine (Cr) levels are associated with increased rates of death from cardiovascular diseases.7, 8, 9 Sezer et al have shown that coronary collateral development is impaired in patients who have a serum Cr level of 1.5 mg/dL or more.10 Another study showed that coronary collateral development is impaired in patients with mild to moderate chronic kidney disease.11 However, there have been no studies investigating the association between mildly decreased glomerular filtration rate (GFR) and CCC. The aim of this study was to evaluate the association between mildly decreased GFR and coronary collateral development in patients with coronary artery disease. We hypothesized that there would be an association between mildly decreased GFR and CCC.

Methods

Patients

This was a retrospective, cross‐sectional study. Patients who underwent coronary angiography between December 2007 and March 2011 were evaluated for possible inclusion in the study. All study participants had at least 1 occluded major coronary artery as seen on angiography. Demographic, clinic, and laboratory data were obtained from the files of the patients.

Inclusion Criteria

The exclusion criteria consisted of patients (1) who underwent coronary angiography at our hospital and who had total occlusion in at least 1 major epicardial coronary artery, and (2) patients with normal or mildly decreased eGFR according to the Modification of Diet in Renal Disease (MDRD) definition.12

Exclusion Criteria

The exclusion criteria consisted of patients (1) who had previously received a percutaneous coronary intervention or coronary artery surgery, (2) who were categorized in classes III‐IV of the New York Heart Association functional capacity, (3) who had been diagnosed with severe valve stenosis and regurgitation, (4) who had been diagnosed with cardiomyopathy, (5) with acute coronary syndrome (within the previous month), (6) with systemic inflammatory disease (such as collagen tissue disease) or severe disease (malignancy or liver disease), (7) who were younger than 18 or older than 75 years, and (8) with moderate and severely decreased eGFR based on the MDRD definition.

Coronary Angiography and Coronary Collateral Scoring

Coronary angiography was performed on all patients using the Judkins technique. Evaluation of CCC and examination of the angiograms were performed by 2 experienced cardiologists who were blinded to the clinical information and/or laboratory data of the patients. If there was an inconsistency between the 2 cardiologists, the angiograms were evaluated by another cardiologist and a consensus was reached. A patient was considered to have significant narrowing when the diameter of any major coronary artery was <50% of its original value. The Rentrop classification was used to categorize CCC.13 Collateral filling was classified according to the following criteria: 1 = filling of side branches via collateral channels without visualization of the epicardial segment; 2 = partial filling of the epicardial major coronary artery via collateral channels; and 3 = complete filling of the epicardial major coronary artery via collateral channels. The vessel that had the highest Rentrop grade was selected for analysis when there was more than 1 occluded vessel. When there was more than 1 collateral vessel filling the occluded vessel, the collateral vessel with the highest Rentrop grade was used for analysis. Patients were then classified as having poor CCC (Rentrop grades 0–1) or good CCC (Rentrop grades 2–3).

Biochemical Parameters

Biochemical parameters used in the present study were measurements including serum creatinine, which were performed 1 to 7 days prior to the diagnostic coronary angiography.

Glomerular Filtration Rate Calculation

We used the MDRD equation to estimate the GFR (GFR [mL/min per 1.73 m²] = 186 × [serum creatinine]^−1.154 × [age]^−0.203× [0.742 if female] × [1.210 if African‐American]). According to MDRD criteria, normal GFR was defined as eGFR ≥90 mL/min per 1.73 m², and mildly decreased GFR was defined as 60 mL/min per 1.73 m² ≥ eGFR ≤89 mL/min per 1.73 m².

Statistical Analyses

SPSS for Windows version 15.0 (SPSS Inc., Chicago, IL), a statistical package program, was used for statistical analysis. Normally distributed, continuous data are expressed as mean ± standard deviation; non‐normally distributed continuous variables are presented as median (minimum‐maximum). Categorical data are expressed as numbers with percentages. We used the Student t test for normally distributed continuous variables and the Mann‐Whitney U test for non‐normally distributed continuous variables. Categorical data were compared using the χ ² test. To identify factors that independently affected CCC development, the Rentrop score (poor vs good CCC) was used as a dichotomous dependent variable, and multivariate logistic regression analysis was performed to identify independent variables. Potential predictors of CCC were first identified using univariate analyses, with a threshold of P<0.05 and were then included as independent variables in regression models. A P value of <0.05 was considered statistically significant.

The study protocol was approved by the ethics committee at our institution.

Results

Six hundred eighty out of the 3541 consecutive patients who underwent a coronary angiography at our institution between December 2007 and March 2011 had a total occlusion in at least 1 major coronary artery. Two hundred ninety‐nine of these patients fit with the study criteria. Ninety‐three patients (31.1%) had grade 1, 92 patients (30.8%) had grade 2 and, 114 patients (38.1%) had grade 3 collateral circulation. Thus, there were 93 patients in the poor CCC group and 206 patients in the good CCC group. Baseline characteristics of the patients in the good and the poor CCC group are shown in Table 1. The median age (minimum‐maximum) was 62 years (range, 38–74 years) in the good CCC group, and 65 years (range, 39–74 years) in the poor CCC group (P = 0.122). One hundred sixty‐two patients (78.6%) in the good CCC group were male, and 67 (72%) were male in the poor CCC group (P = 0.213). Diabetes was more frequent in the poor CCC group than the good CCC group (41.9% vs 24.8%, P = 0.003). The frequency of dyslipidemia was also higher in the poor CCC group than the good CCC group (49.5% vs 35%, p = 0.018). The frequency of mildly decreased eGFR was 76.3% in the poor CCC group and 37.4% in the good CCCgroup (P < .001). The plasma levels of high sensitivity C‐reactive protein (hs‐CRP) level (mg/dL) ‐median (minimum‐maximum)‐ were higher in the poor CCC group than the good CCC group [4.125 (2.08–6.7) vs 3.27 (1.17–6.7), P<0.001]. Coronary angiographic findings in the 2 groups are summarized in Table 2. The location of the occluded vessel and the number of diseased vessels were similar in the 2 groups. A multivariate logistic regression analysis was conducted using collateral circulation as the dependent variable and hs‐CRP, diabetes, dyslipidemia, and eGFR as independent variables (covariates). Using multivariate logistic regression analysis, we found that eGFR was an independent predictor of CCC (B = 1.68; odds ratio = 5.4; P<0.001; 95% confidence interval, 3.1–9.4). After adjusting for hs‐CRP, frequency of diabetes, and dyslipidemia, eGFR was still independently associated with CCC. Results of multivariate logistic regression analysis are presented in Table 3.

Table 1.

Baseline Characteristics of the Groups

Variable	Poor CCC Group, n = 93	Good CC Group, n = 206	P
Age (y), median (min‐max)	65 (39–74)	62 (38–74)	0.122
Male, n (%)	67 (72)	162 (78.6)	0.213
Hypertension, n (%)	55 (59.1)	118 (57.3)	0.763
Diabetes, n (%)	39 (41.9)	61 (24.8)	0.003
Dyslipidemia, n (%)	46 (49.5)	72 (35)	0.018
Smoking, n (%)	35 (37.6)	96 (46.6)	0.149
Mildly decreased eGFR, n (%)	71 (76.3)	77 (37.4)	<0.001
hs‐CRP (mg/dL), median (min‐max)	4.125 (2.08–6.7)	3.27 (1.17–6.7)	<0.001
Cardiovascular medications
ACEIs, n (%)	52 (55.9)	116 (56.3)	0.949
Angiotensin receptor blockers, n (%)	15 (16.1)	22 (10.7)	0.186
β‐blockers, n (%)	68 (73.1)	151 (73.3)	0.829
Calcium channel blockers, n (%)	17 (18.3)	29 (14.1)	0.352
Statins, n (%)	30 (32.2)	62 (30.1)	0.708

Abbreviations: ACEIs, angiotensin converting enzyme inhibitors; CCC, coronary collateral circulation; eGFR, estimated glomerular filtration rate; hs‐CRP, high sensitive C‐reactive protein; min‐max, minimum‐maximum. hs‐CRPs were available for 74 patients in the poor CCC group and 124 patients in the good CCC group.

Table 2.

Coronary Angiographic Findings of the Groups

Variable	Poor CCC Group, n = 93	Good CCC Group, n = 206	P
1‐vessel disease, n (%)	29 (31.2)	48 (23.3)	0.150
2‐vessel disease, n (%)	28 (30.1)	76 (36.9)	0.255
3‐vessel disease, n (%)	36 (38.7)	82 (39.8)	0.858
Occluded LAD, n (%)	43 (46.2)	88 (42.7)	0.571
Occluded CX, n (%)	22 (23.7)	49 (23.8)	0.980
Occluded RCA, n (%)	42 (45.2)	107 (51.9)	0.279
No. of diseased vessels, mean ± SD	2.06 ± 0.832	2.17 ± 0.785	0.326

Abbreviations: CCC, coronary collateral circulation; CX, circumflex artery; LAD, left anterior descending artery; RCA, right coronary artery; SD, standard deviation.

Table 3.

Results of Multivariate Logistic Regression Analysis

Variable	B	P	Odds Ratio	95% CI
Dyslipidemia	−0.323	0.346	0.72	0.37–1.41
hs‐CRP	−0.426	0.002	1.53	1.16–2.01
Diabetes	−0.546	0.122	0.579	0.29–1.15
Mildly decreased eGFR	1.535	0.000	4.64	2.33–9.22

Abbreviations: CI, confidence interval; eGFR, estimated glomerular filtration rate; hs‐CRP, high sensitive C‐reactive protein.

Discussion

There were 2 main findings in this study. First, poor coronary collateral circulation was associated with mildly decreased eGFR in patients with coronary artery disease. Second, mildly decreased eGFR was an independent predictor of CCC in multivariate models.

The MDRD equation has been developed using a large database and does not require height or weight. In addition, it has been validated in various kidney diseases, African Americans with nephrosclerosis, and kidney transplant recipients.14 The MDRD equation is more accurate and precise than the Cockcroft‐Gault equation for persons with a GFR less than approximately 90 mL/min per 1.73 m².12, 15 For these reasons, the MDRD equation was used for GFR calculation.

There is considerable evidence that chronic renal disease is independently associated with coronary artery disease. Moreover, in many studies it was demonstrated that individuals who have even mildly impaired renal function are at high risk for cardiovascular morbidity and mortality, independent of traditional risk factors for cardiovascular disease.16, 17, 18, 19, 20 Although chronic renal disease is associated with several risk factors for cardiovascular disease, such as male gender, old age, hypertension, smoking, diabetes, obesity, and preexisting cardiovascular disease, the excess cardiovascular mortality in chronic renal disease is not fully explained by traditional risk factors. Our findings suggest that the excess cardiovascular mortality in chronic renal disease could be partially attributed to impaired coronary collateral development.

The relationship between impaired renal function and poor CCC may be explained by increased inflammatory activity and impaired endothelial functions. There are a couple of possible explanations.

First, the endothelium is 1 of the critical factors in collateral development. Many studies showed that nitric oxide (NO) released by endothelium plays a role in collateral development. In the study by Ziche et al, it was shown that when the endogenous NO production was increased, there was also increased proliferation and migration of endothelium cells, and that NO inhibitors inhibited capillary endothelial cell proliferation and migration.21 Additionally, in vivo studies showed that endogenous NO inhibition negatively affects collateral development.22 Evidence from these studies illustrates that the endothelium plays a significant role in coronary collateral development. It has been proposed that endothelial dysfunction may be an important mechanism linking impaired renal function to cardiovascular disease.23 Perticone et alreported that the cause of mild to moderate renal disease in patients with uncomplicated essential hypertension with normal serum CRP levels is endothelial dysfunction, independent of blood pressure.24 Another study by the same authors demonstrated that endothelial dysfunction is associated with renal function decline in hypertensive patients with normal or mildly impaired baseline renal function.25 In a study where the endothelial function in a patient group with chronic renal insufficiency was evaluated using flow‐mediated dilatation, it was shown that the endothelial function was impaired even in patients with mild renal insufficiency.26

Second, renal insufficiency, a condition whereby endothelial dysfunction is pervasive, is very frequently accompanied by an elevated blood CRP level.23, 27, 28 It has been shown that the CRP levels in patients with mild and moderate renal impairment are associated with both renal and endothelial functions.29 CRP, an acute phase reactant, has been shown to impair endothelial functions in many studies. In 1 study, the response of brachial artery blood flow to acetylcholine has been shown to be negatively associated with blood CRP concentrations in patients with coronary artery disease.30 In a study by Verma et al, it was shown that CRP directly represses NO synthesis, and as a result it was proposed that CRP could inhibit angiogenesis.31 Another study showed that CRP decreases the expression and bioactivity of nitric oxide synthase enzyme in human aortic endothelial cells.32 Vascular endothelial growth factor (VEGF) is a strong mitogen of endothelial cells. Recently, Schneeweis et alshowed that CRP inhibits endothelial cell migration induced by VEGF.33 In addition, it was clearly proven in clinical studies that increased plasma hs‐CRP levels decrease coronary collateral development.34, 35, 36 In the present study, we found that plasma hs‐CRP levels were higher in the poor CCC group than the good CCC group. The results of this study and other previous studies suggest that in patients with slight renal dysfunction, increases in CRP levels could be an underlying cause of impaired coronary collateral development.

Study Limitations

Several limitations should be considered. The major limitation was the retrospective nature of the study. Another limitation was that endothelial function, which has a significant role in collateral development, was not studied. Also, the Rentrop scoring system was used for collateral grading even though the Rentrop scoring system shows weak correlation with invasive parameters of collateral function. However, the effect of this problem on CCC is the same in both groups and thus should not change the interpretation of our results.

Conclusion

We found that coronary collateral circulation was impaired in patients with mildly decreased eGFR, and that eGFR was an independent predictor of collateral circulation. Impaired coronary collateral development in patients with mildly decreased GFR could be associated with endothelial dysfunction and elevated inflammatory activity observed in these patients.