Abstract
Patients with end stage renal disease have a high prevalence of cardiovascular disease and coronary arteriography is often routinely performed prior to kidney transplantation. However, the value of the conventional risk factors and non-invasive markers of coronary artery disease (CAD) in triaging patients for coronary arteriography has not been fully examined. 116 patients with end stage renal disease were evaluated. Coronary arteriography was performed in all patients either for a suspicion of CAD or as part of a routine pre-transplant evaluation. Lesions causing ≥ 50% luminal diameter stenosis in any of the three major coronary artery systems were considered significant. The mean age was 53.3 ± 9.3 years. Significant CAD was present in 69 patients (60%). Increasing age, family history of premature ischemic heart disease, the presence of angina, abnormal Q waves on the ECG or abnormal ST segment depression and the presence of coronary calcification were significant markers of coronary artery disease. However male gender, diabetes mellitus and obesity did not correlate with coronary disease. Even though hypertension, hypercholesterolemia and smoking were also not useful predictors these could have been modified by the renal failure. In conclusion increasing age, a family history of premature ischemic heart disease and some non-invasive markers were useful predictors of coronary disease.
INTRODUCTION
Cardiovascular diseases currently account for about half the deaths among the dialysis population Citation[1-4]. The prevalence of angiographically documented significant CAD (CAD) varies between 24–85%, depending on the age, sex, and presence of angina or diabetes mellitus Citation[5-7]. Despite this high prevalence of CAD, the non-invasive approach to the diagnosis of CAD in these patients is often difficult and strewn with pitfalls. Patients with end stage renal disease (ESRD) are more prone to have silent ischemia partly due to the high prevalence of diabetes mellitus. On the other hand, they may have non-ischemic chest pain due to uremic pericarditis or neuropathy Citation[[8]]. In addition, angina without significant CAD has been reported in 25–30% of dialysis patients, probably due to a combination of anemia, hypertension and left ventricular hypertrophy Citation[9-10]. Abnormal baseline electrocardiograms (EKG), poor exercise tolerance and the presence of coexisting myopathy and neuropathy conspire to limit the usefulness of stress testing in these patients. While coronary calcification has been shown to correlate with the atheroma load in patients with CAD Citation[11-12], its value in patients with ESRD has not been established. Consequently, many centers consider coronary arteriography an essential part of the pre-transplant evaluation of coronary disease among patients with ESRD. However, the value of conventional coronary risk factors and non-invasive markers of CAD in the identification of high-risk patients has not been fully elucidated.
OBJECTIVES
To evaluate the relevance of conventional cardiovascular risk factors, and other markers as possible predictors of coronary artery disease in patients with end stage renal disease on renal replacement therapy.
METHODOLOGY
Patients
We studied 116 patients (81 males and 35 females) with ESRD undergoing coronary arteriography at the Chest Diseases Hospital, Kuwait. All patients were on a maintenance dialysis program or were being considered for kidney transplant. Coronary arteriography was performed either because of a clinical suspicion of coronary disease or as part of a routine pre transplant evaluation. Dialysis was performed either on the same day or the day prior to cardiac catheterization in the majority of the patients. All patients were interviewed and examined by a staff cardiologist to obtain information relating to the presence of angina, past myocardial infarction, diabetes mellitus, hypertension, smoking habits and family history of premature ischemic heart disease. Angina pectoris was diagnosed only if the patient described the typical features of the chest pain. The height and weight were recorded for estimation of body mass index.
Cardiovascular Risk Factors
Patients were classified as diabetics if this diagnosis had been made by a physician earlier, irrespective of whether they were on anti-diabetic agents or not. In most of the patients this diagnosis could be confirmed by the presence of elevated non-stressed fasting blood sugar value of >7.7 mmol. Systemic hypertension was diagnosed if the systolic or diastolic blood pressures exceeded 140 or 90 mmHg respectively, or if antihypertensive therapy was prescribed. Current smokers and those who had smoked in the previous ten years were classified as smokers. Patients who were on treatment for hyperlipidemia or whose fasting serum cholesterol levels were ≥ 5.2 mmol/L were considered to have hypercholesterolemia. A family history of premature ischemic heart disease was considered to be present only if it was diagnosed before the age of 55 years in a first-degree male relative (or 65 years in a first-degree female relative). The Body Mass Index (weight in kilograms/square of the height in meters) was used as an index of obesity.
Electrocardiogram
All EKGs were examined for the presence of abnormal rhythms, left ventricular hypertrophy, ST segment and T waves abnormalities and for evidence of past myocardial infarction. Left ventricular hypertrophy was diagnosed using the criteria of Sokolow and Lyon Citation[[13]]. An ST segment elevation or depression ≥1 mm was considered abnormal as also inversion of the T waves in leads where the net QRS deflection was positive. A Q wave with duration equal to or greater than 0.04 s or a depth exceeding one fourth of the height of the R wave in the same lead was considered abnormal.
Cardiac Catheterization
All patients underwent selective coronary arteriography and left ventriculography via the femoral route using standard techniques. Non-ionic contrast medium was used in all cases. Multiple projections were employed to assess the presence and severity of CAD. Two interventional cardiologists independently reviewed all coronary arteriography films. Stenosis severity was calculated using hand held calipers. Left ventricular end diastolic and end systolic volumes and the ejection fraction were calculated using the Siemens Ancor software (Siemens-Elema AB, Sweden) based on Chapman's modification of Simpson's rule.
Significant coronary artery stenosis was defined as the presence of ≥50% luminal diameter stenosis in any of the three major coronary artery systems. Patients with coronary artery stenosis of <50% were classified along with the patients with normal coronary arteries. Single vessel CAD was defined as significant stenosis affecting any one of the three major epicardial coronary artery systems, i.e. the left anterior descending coronary artery and its major branches, the left circumflex coronary artery and its major branches, and the right coronary artery and its major branches. Two-vessel CAD was defined as significant stenosis in two of the three major coronary artery systems and three-vessel CAD defined as significant stenosis in all three major coronary artery systems.
Statistical Analysis
Patients were divided into two groups based on the presence or absence of significant CAD. Various clinical and laboratory parameters were evaluated as possible markers for CAD. Mean and standard deviation were calculated for continuous variables, which were then compared using the Student's t-test. The chi square test was used to test the difference between dichotomous variables. A value of p<0.05 was considered statistically significant.
RESULTS
A total of 116 patients were studied. The mean age was 53.3 ± 9.3 years. Eighty-six patients were on maintenance hemodialysis and 22 patients on regular peritoneal dialysis. The remaining 8 patients had not yet entered a dialysis program but were being considered for the initiation of renal replacement therapy. There were no complications related to cardiac catheterization.
Significant CAD was present in 69 patients (60%). Single vessel CAD was seen in 30 of the 69 patients (43%) with coronary disease, while two and three vessel CAD were seen in 24 (35%) and 15 (22%) patients, respectively. The left anterior descending coronary artery, the circumflex coronary artery and the right coronary artery were involved in 51 (74%), 39 (57%) and 32 (46%) patients, respectively.
As in the non-renal population, CAD was found to be more common with increasing age. Only one of 8 patients (12%) under the age of 40 years had significant CAD, whereas among the 32 patients who were 60 years or older, 25 had significant CAD. Among the 11 patients who reported a family history of premature ischemic heart disease, 10 had significant CAD (positive predictive value 91%, 95% confidence interval 74% to 100%).
Although 69 patients (60%) had significant CAD, only 21 patients (18%) gave a history of angina. Coronary disease was found in 19 of the 21 patients with angina (positive predictive value 91%, 95% confidence interval 78% to 100%). ().
Table 2. Predictors of CAD
While abnormalities on the surface electrocardiogram were seen in 92 (79%) patients, significant Q waves or ST segment depression were observed in only 7 (7%) and 18 (16%) patients respectively. All 7 patients with abnormal Q waves and 15 of the 18 (positive predictive value 83%, confidence interval 66–100%) with ST segment depression on the surface ECG were found to have significant CAD. The presence of T wave abnormalities was not helpful in predicting CAD.
Fluoroscopically evident coronary calcification was noted in 37 patients, 31 of whom were found to have significant CAD (positive predictive value 84%, 95% confidence interval 72% to 96%).
The etiology of renal failure did not significantly influence the presence of coronary disease. (). Similarly, there were no significant differences in the left ventricular end-diastolic volumes (94 ± 25 versus 101 ± 31 ml/M2, p = 0.24), end-systolic volumes (37 ± 18 versus 38 ± 22 ml/M2, p = 0.94) or ejection fraction (61 ± 11% versus 63 ± 11%) between the patients with and without CAD.
Table 1. Demographic, Laboratory and EKGFootnote1 Characteristics: (n = 116)
Male gender, presence of diabetes mellitus and obesity were not significant predictors of coronary disease nor were hypertension, hypercholesterolemia and a history of smoking (see discussion).
DISCUSSION
The diagnosis of CAD among patients with renal failure has important implications in management and in many cases may preclude future kidney transplantation. In our study we found that 60% of our patients had significant CAD. This finding is similar to that of Joki et al Citation[[7]] who documented significant CAD in 15 of 24 patients (62.5%) with ESRD. Two other angiographic studies have noted a somewhat lower CAD prevalence of 31% Citation[[14]] and 43% Citation[[15]] respectively among patients on maintenance hemodialysis. A high prevalence of CAD such as ours would mandate screening of all patients with ESRD, especially those who are being considered for kidney transplantation.
Many reports have focused on the screening of diabetic patients with ESRD because they constitute a ‘high risk group' for the development of CAD Citation[[5]], Citation[16-17]. Manske et al Citation[[5]] found that among diabetic patients above the age of 45 years the prevalence of CAD might be as high as 85%. However, in our series we found that the prevalence of CAD among non-diabetics was not significantly different from that of diabetic subjects. This observation has important therapeutic implications since many centers recommend routine coronary arteriography only in patients with diabetes mellitus. Our findings indicate that the index of suspicion for CAD should be the same in diabetic and non-diabetic patients with ESRD.
We found that with increasing age, the prevalence of CAD also increased. In patients below 40 years of age the CAD prevalence was as low as 12%. However, among those above the age of 59 years, the CAD prevalence was so high (78%) that coronary arteriography could probably be indicated based on this criterion alone.
Among the variables tested, increasing age, a history of angina, presence of significant ST segment depression or Q waves on the ECG and presence of coronary calcification on fluoroscopy were found to be predictive of CAD. Despite the atypical symptomatology of patients with ESRD, we found that a history of typical angina had a positive predictive value of 91% for CAD. However a history of angina was obtained only in a minority of our patients (18%). Other investigators have documented symptoms suggestive of CAD in 50–77% of patients undergoing maintenance dialysis, but found significant CAD in only about 50% of the symptomatic patients Citation[14-15]. These differences may partly be due to the subjective nature of the symptom and its interpretation. Only patients with typical angina were included in our analysis.
ECG abnormalities are commonly observed among patients with ESRD, in part due to the frequent electrolyte disturbances and the high prevalence of left ventricular hypertrophy secondary to the volume and pressure overload conditions that often prevail in this population. Nearly 80% of our patients had an abnormal ECG but abnormal Q waves or ST segment depression were noted in just 7% and 16% respectively. All the patients with abnormal Q waves and 83% of those with ST segment depression were found to have significant CAD.
The association between coronary atherosclerosis and coronary calcification is well known Citation[[11]], Citation[[18]]. While metastatic calcification commonly occurs in patients with ESRD, the preferential cardiac sites of calcification are the mitral annulus and valve and the aortic valve Citation[[19]]. The significance of coronary calcification as a predictor of CAD among patients with ESRD has not been fully explored. We could demonstrate significant CAD in 84% of the 37 patients with coronary calcification. These data indicate that the presence of fluoroscopically evident coronary calcification may be a significant marker of CAD in patients with ESRD. Similar findings have been documented by Marwick et al Citation[[12]] who found coronary calcification by digital subtraction fluoroscopy to have a sensitivity and specificity of 78% and 66% respectively for the detection of CAD.
Contrary to what is observed in the general population, males were not at a higher risk for coronary artery disease and diabetes mellitus and obesity did not correlate with significant coronary disease. The lack of correlation observed with hypertension, hypercholesterolemia and a history of smoking may have other explanations. The use of antihypertensive drugs could have lowered the blood pressure. The ejection fraction >60%, makes left ventricular dysfunction unlikely as a cause of a lowered blood pressure. In chronic renal failure patients on dialysis, the total serum cholesterol may be normal while the HDL cholesterol is low Citation[[19]]. This may explain our lack of correlation with the total cholesterol and the HDL level was not measured in our patients. The incidence of smoking at 28% was low in our renal failure patients with coronary disease when compared to the 60% that we have found in those with coronary disease and normal renal function. This low incidence of smoking, presumably related to the “sick” status of these patients may explain the lack of correlation with coronary disease.
Limitations
Although we were able to identify several non-invasive markers of CAD, this study does suffer from some limitations. Since we followed a policy of routine coronary arteriography prior to kidney transplantation, stress testing and in particular radionuclide scans were not routinely performed in all our patients. This policy was influenced by the well-known limitations of stress testing in patients with ESRD.
CONCLUSION
Significant CAD is common among patients with ESRD, and was seen in 60% of our patients. Age, a family history of premature ischemic heart disease, history of typical angina, presence of abnormal Q waves or ST segment depression on the ECG and coronary calcification seen at fluoroscopy were significant predictors of coronary disease. Male gender, diabetes mellitus and obesity were not helpful in predicting coronary disease. Chronic renal failure and the influence of treatment could in part explain the lack of correlation with hypertension, hypercholesterolemia and smoking.
ACKNOWLEDGMENT
We thank Mr. Jaynold A. Collamar for his help in data collection and Mr. Shahid Alam Khan for his secretarial assistance.
REFERENCES
- Morbidity and mortality of Renal Dialysis: An NIH Consensus conference statement. Ann Intern Med. 1994; 121: 62–70
- Held P J, Brunner F, Odaka M, Garcia J R, Port F K, Gylin D S. Five-year survival for end-stage renal disease patients in the United States, Europe, and Japan, 1982–1987. Am J Kidney Dis 1990; 15: 451–457
- Geerlings W, Tufveson G, Ehrich J H, et al. Report on management of renal failure in Europe, XXIII. Nephrol Dial Transplant 1994; 9(Suppl 1)6–25
- Fassbinder W, Brunner F P, Brynger H. Combined Report on Regular Dialysis and Transplantation in Europe XX, 1989. Nephrol Dial Transplant 1991; 6(Suppl I)5–35
- Manske C, Thomas W, Wany Y, Wilson R. Screening diabetic transplant candidates for CAD: Identification of a low risk subgroup. Kidney Int 1993; 44: 617–627
- Kramer W, Wizemann V, Lammlein G, et al. Cardiac dysfunction in patients on maintenance hemodialysis. Contrib Nephrol. Basel, Karger 1986; 52: 110–124
- Joki N, Hase H, Nakamura R, Yamaguchi T. Onset of CAD prior to initiation of hemodialysis in patients with end-stage renal disease. Nephrol Dial Transplant 1997; 12: 718–723
- Kremastinos D, Paraskevaidis I, Voudiklari S, et al. Painless myocardial ischemia in chronic hemodialysed patients: A real event?. Nephron 1992; 60: 164
- Rostand S G, Gretes J C, Kirk K A, Rutsky E A. The epidemiology of CAD in patients on maintenance hemodialysis: Implications for management. Contrib Nephrol Karger Basel 1986; 52: 34–41
- Roig E, Betriu A, Castaner A, Magrina J, Sanz G, Navarra-Lopez F. Disabling angina pectoris with normal coronary arteries in patients undergoing hemodialysis. Am J Med 1981; 71: 437–444
- Budoff M J, Georgiou D, Brody A, et al. Ultrafast computed tomography as a diagnostic modality in the detection of CAD- a multicenter study. Circulation 1996; 93: 898–904
- Marwick T, Hobbs R, Vanderlaan R L, Steinmuller D, Braun W. Use of digital subtraction fluorography in screening for CAD in patients with chronic renal failure. Am J Kidney Dis 1989; 14(2)105–109
- Sokolow M, Lyon T P. The ventricular complex in left ventricular hypertrophy as obtained by unipolar precordial and limb leads. Am Heart J 1949; 37: 161–168
- Ikram H, Lynn K L, Bailey R R, Little P J. Cardiovascular changes in chronic hemodialysis patients. Kidney Int 1983; 24: 371–376
- Rostand S G, Kirk K A, Rutsky E A. Dialysis-associated ischemic heart disease: Insights from coronary angiography. Kidney Int 1984; 25: 653–659
- Bates J R, Sawada S G, Douglas S S, et al. Evaluation using dobutamine stress echocardiography in patients with insulin dependent diabetes mellitus before kidney and/or pancreas transplantation. Am J Cardiol 1996; 77: 175–179
- Vandenberg B F, Rossen J D, Grover-McKay M, Shammas N W, Burns T L, Rezai K. Evaluation of diabetic patients for renal and pancreas transplantation. Transplantation 1996; 62: 1230–1235
- Blankerhorn D H, Stern K. Calcification of the coronary arteries. Am J Roentgenol 1959; 81: 772–777
- Leier C V, Boudoulas H. Renal disorders and heart disease. Heart Disease, a textbook of cardiovascular medicine, E Braunwald, W B Saunders. (5th Edition), 1997; 1914–1938