Diagnostic value of C-reactive protein in patients with angiographically documented coronary heart disease.

AIM
The aim of this study was to evaluate the diagnostic value of serum C-reactive protein (CRP) level measurement in predicting coronary artery disease (CAD) that can be shown angiographically.


METHODS
CRP levels were determined in the blood of 198 patients (patients group, PG) with angiographically documented coronary artery disease and compared with that of 85 patients (control group, CG) who had a clinical indication for coronary angiography but have no angiographically determined coronary artery stenosis, as well as with that of 41 healthy volunteers as a healthy control group (HG) who did not have any complaint and did not have coronary angiography. CRP levels were measured 24 hours prior to angiography in PG and CG patients, and in the morning after not having eaten for same time. Any coronary artery stenosis or plaque formation was defined as CAD. Severity of the disease was assessed by both the number of diseased vessels (0 to 3) and the degree of stenosis (<50% mild, 50-70% moderate and >70% severe).


RESULTS
Receiver Operating Characteristics (ROC) curves of CRP in angiographically documented CAD group showed a diagnostic value of 0.659 in female patients, followed by 0.542 in male patients, in predicting CAD. CRP levels were found to be significantly different between groups, higher in PG (6.2 +/- 0.86 mg/L) than those of CG (3.7 +/- 0.92 mg/L) and HG (0.854 +/- 0.2 mg/L) (p<0.05). CRP levels were not associated with the number of diseased vessels, neither with the degree of the occlusion (p>0.05). Multiple logistic regression analysis after adjustment for the established coronary risk factors showed CRP as an independent discriminating risk factor for CAD.


CONCLUSION
It is concluded that CRP measurement has a value in predicting the presence of angiographically documented CAD. However, CRP levels were not associated with the degree or severity of CAD.

Summary: Aim: The aim of this study was to evaluate the diagnostic value of serum C-reactive protein (CRP) level measurement in predicting coronary artery disease (CAD) that can be shown angiographically. Methods: CRP levels were determined in the blood of 198 patients (patients group, PG) with angiographically documented coronary artery disease and compared with that of 85 patients (control group, CG) who had a clinical indication for coronary angiography but have no angiographically determined coronary artery stenosis, as well as with that of 41 healthy volunteers as a healthy control group (HG) who did not have any complaint and did not have coronary angiography. CRP levels were measured 24 hours prior to angiography in PG and CG patients, and in the morning after not having eaten for same time. Any coronary artery stenosis or plaque formation was defined as CAD. Severity of the disease was assessed by both the number of diseased vessels (0 to 3) and the degree of stenosis (<50 % mild, 50-70 % moderate and >70 % severe). Results: Receiver Operating Characteristics (ROC) curves of CRP in angiographically documented CAD group showed a diagnostic value of 0.659 in female patients, followed by 0.542 in male patients, in predicting CAD. CRP levels were found to be significantly different between groups, higher in PG (6.2 ± 0.86 mg/L) than those of CG (3.7 ± 0.92 mg/L) and HG (0.854 ± 0.2 mg/L) (p<0.05). CRP levels were not associated with the number of diseased vessels, neither with the degree of the occlusion (p>0.05). Multiple logistic regression analysis after adjustment for the established coronary risk factors showed CRP as an independent discriminating risk factor for CAD. Conclusion: It is concluded that CRP measurement has a value in predicting the presence of angiographically documented CAD. However, CRP levels were not associated with the degree or severity of CAD. also the likelihood of clinical disease development (3,6,27). In normal healthy individuals, CRP is a trace protein with a range of up to 5 mg/L. 90 % of apparently healthy subjects have levels of less than 3 mg/L, and 99 % of them have that level less than 10 mg/L. There is no difference between levels in healthy men and healthy women, and no variation with age.
Many questions are still unanswered, including whether CRP is a culprit molecule or simply a marker, and the cutoff points and specificity of the methods are still debated. Drugs and other treatments have little or no affect on CRP production unless they also affect the disease process which is responsible for induction of CRP synthesis. It is thus of great interest that the serum level of CRP has been found to be higher in subjects having no complaint but at greater risk of developing MI, stroke and peripheral vascular disease (4,27).
The aim of this study was to compare CRP concentration in patients with CAD, documented by coronary angiography, with that of patients with normal angiograms but having complaints related to ischemic heart disease, as well as with that of normal healthy individuals, to determine the diagnostic value of CRP levels in the progression of CAD. We also aimed to study the association of this marker with the number of diseased vessels and the degree of stenosis.

Material and Methods
The study was designed to evaluate the serum CRP levels of individuals of three different groups. 198 patients were included in the Patient Group (PG) who had documented CAD proven by coronary angiography. The Control Group (CG) consisted of 85 patients with clinical indications for angiography, but normal coronary angiogram. Fortyone healthy subjects were included in the Healthy Group (HG).
Although the healthy subjects did not have coronary angiograms, they underwent a comprehensive physical examination by a physician, and completed the World Health Organisation standards Rose questionnaire on chest pain, and answered other questions about their past medical history. None of the individuals in the healthy group had angina or a prior history of CAD. All of them had normal electrocardiograms based on the Minnesota Coding Criteria.
All Coronary Angiographies were performed in the same centre. Patients diagnosed with acute coronary syndrome six months prior to the study were excluded.
All patients were monitored for somatic illness throughout the investigation period and excluded if symptoms of infection or systemic illness were present (rheumatic disease, chronic liver disease, renal disorder, cancer, etc). Additional exclusion criteria included the use of statins and other medications that could affect CRP concentrations.
Severity of the disease was assessed by counting both the number of diseased vessels (0 to 3) and the degree of the stenosis (<50 % mild, 50-70 % moderate and >70 % severe ) (13).
Prior medical history, personal characteristics and habits were obtained by the questionnaire.
The presence of obesity, diabetes mellitus, and hypertension were also assessed. Obesity was defined as body mass index (BMI) greater than 27.8 kg/m 2 as proposed by the National Institute's of Health consensus statement. Diabetes was defined as fasting blood glucose >126 mg/dL or diagnosis of diabetes needing diet or drug therapy. Hypertension was defined as resting systolic blood pressure >140 mm Hg and/or diastolic blood pressure>90 mm Hg (17). Smoking was defined as being a current smoker and smoking at least 10 cigarettes per day for at least ten years.

Standard Blood Sampling
After an eight-hour fast venous blood samples were collected into vacuum tubes in the morning time on the day before the angiographic procedure on the PG and the CG patients. Blood was centrifuged at 3000 g for 10 minutes at 4 °C. After separation, the aliquoits were frozen at -70 °C until analysis.

Measurements
Serum CRP concentrations were determined with BNA 100 Nephlometer (Dade Behring, Germany) according to the manufacturer's instructions. The assay was linear from 0.175 to 230 mg/L, calibrators and controls (R-CL-I) were supplied by the manufacturer. Specifications of intra assay and inter assay coefficients of variation of CRP assay were assessed from quality control data of the laboratory which were 3.4 % and 6.2 %, respectively. The levels of glucose, urea, creatinine, CK, CK-MB, LDH, cholesterol, trygliceride, high-density lipoprotein cholesterol (HDL-C) both in PG and CG were analyzed by BM/902 (ROCHE Diagnostic Systems, Germany). Low-Density lipoprotein cholesterol (LDL-C) was calculated using Friedewald formula if tryglicerides were <400 mg/d (9).

Statistical Analysis
Values are expressed as mean ± SD for normally distributed variables, and log mean CRP serum levels at 20 th and 80 th percentiles are given.
Because the distribution of CRP is rightward skewed, values derived from log-transformed means were used as means for this variable throughout the study; these values virtually coincided with median values. A general linear model was used to consider case control differences after adjustment for other factors and the area under the ROC curves were used for comparison of the diagnostic values of different analyses (including PG and CG groups using angiography as the highest accepted standard). Optimal cut-off levels, sensitivity and specificity of CRP were selected based on ROC curves.

Results
No statistically significant difference was noted in HDL-C levels among the three groups. However, age, diabetes, hypertension, BMI, total cholesterol and trygliceride levels were significantly different in three groups (Table 1).
Smokers consisted of 47 % in the PG, 16% in the CG, and 37 % in HG (Table 1).
CRP levels were different in the studied groups, and these differences remained statistically significant after correction (ANOVA) for other risk factors (Table 2). CRP levels were not related to the number of diseased vessels and degree of occlusion (p>0.05) ( Table 3).
CRP levels were significantly higher in the PG (6.2 ± 0.86 mg/L) than in the CG (3.7 ± 0.92 mg/L) and the HG (0.85 ± 0.2 mg/L) (p<0.05) (Figure 1, Table 2). The number of subjects with concentrations of CRP ≥10 mg/L comprised of HG, CG and PG groups were 0.2 %, 9 %, and 11 % in the HG, CG and PG respectively.
Optimal cut-off levels of CRP providing the maximum efficiency were found in female and male patients with CADs of 6.9 mg/L and 6.4 mg/L, respectively ( Figure 2). The area under the ROC curve was 0.659 in females and 0.542 in male patients (Figure 3 and 4 respectively). ROC curve based specificity of CRP levels (38 % for male, 49 % for female) was rather lower than the sensitivity of CRP levels (76 % male, 78 % for female) ( Table 4).
A significant difference was seen in the gender-related difference of CRP concentrations in the PG (Table 5), and these differences were statistically significant after correction with ANCOVA for age, blood pressure, BMI, smoking, glucose, lipids concentration, number of diseased vessels and degree of occlusion.
CRP did not correlate with other continuous variables such as total cholesterol, (r: 0.095, p: 0.15), triglycerides

Discussion
The data from our study demonstrated that the serum concentrations of CRP were increased during atherosclerotic process. However these increases failed to reflect the severity of the CAD. Our results also confirm the existing relationship between C-reactive protein levels and the presence of coronary artery disease documented by coronary angiography. The small area under the ROC curve suggest a possible presence of CAD in angiography. It should be pointed out that the differences seen between groups in CRP concentrations remained highly significant after the adjustment for age, smoking, hypertension, diabetes mellitus, BMI, total cholesterol, LDL-C and triglycerides.
Despite changes in lifestyle and the use of new pharmacological approaches to lower plasma cholesterol concentrations, cardiovascular disease continues to be the principal cause of death in Europe, USA and most of Asia. Half of the heart attacs occur in people with normal cholesterol levels (13). Although hyprecholesterolemia is important in approximately 50% of the patients with CAD, other risk factors need to be taken into consideration (21).
CRP is a sensitive marker for systemic inflammation. Recent work from Pasceri et.al. indicates that CRP itself might be involved in the pathogenic response. In their study they suggested that CRP could induce adhesion molecule expression in endothelial cells via its effect on monocyte chemoattracting protein 1. Regardless of the CRP's precise mechanism for inflammatory effects in the vasculature, perhaps the most relevant is the prospect of inflammation as a therapeutic target (18).
The peak plasma CRP values are strongly associated with post-infarct morbidity and mortality (17). In a metaanalysis of 7 prospective studies, elevated serum CRP concentration was shown to predict future risk of CAD (10). C-reactive protein levels below the conventional upper normal limit of 10 mg/L have been associated with a 2-to-3 fold increase in risk of myocardial infarction, ischemic stroke, peripheral artery diseases and CAD (18,21). Several prospective studies have shown that healthy people with higher serum CRP levels, though usually within the normal reference range, are at an increased risk of developing MI, stroke or peripheral vascular disease (1,5,14,22).
Rifai et al. demonstrated higher plasma concentration of CRP in men with angiographically documented CAD (20). Similarly, Haideri et.al. found an association of elevated CRP levels with angiographically documented coronary artery disease. They also reported that CRP concentration was positively related to the number of diseased vessels and associated with the degree of stenosis of coronary arteries (12). In this study, although CRP levels were found to be increased in patients with angiographically documented CAD, no correlation could be demonstrated between the CRP concentration and severity of CAD. These findings are in agreement with the results of Rifai and colleagues. We also found that, not with standart the normal coronary angiograms, the CG patients had elevated CRP levels, which has not been reported before (12,20).
It has been reported in previous studies that there was no sex-related difference nor association between the CRP levels and age in healthy individuals (7,15,16). However Onat et al. reported that gender is a significant independent determinant of CRP levels in a healthy population in western Turkey (17). In the NHLBI Family Heart Study, CRP values were associated with age, family type, and gender (8). Our study showed that CRP levels in male individuals were higher than that of females in HG individuals. ROC curves ( figure 3,4) showed a low level of sensitivity and specificity in the present study, and showed uselessness to diagnose angiographically significant CAD. The area under the ROC curve was 0.542 in males in comparison to 0.659 in females in PG (p<0.05). To best of our knowledge these data were not reported in the literature so far.
One possible explanation of the increased CRP levels in patients with a clinical indication for cardiac catheterisation but with normal angiograms is the association of CRP levels with ongoing atherosclerosis. The increase in CRP levels in these individuals may reflect the diffuse atherosclerotic process in the vascular system rather than the degree of the localised obstruction.
In one study a cut-off value of 15 mg/L was chosen on the basis of ROC curves in patients with unstable angina and myocardial infarction (7). Recommended reference limits for serum CRP levels appeared to be insufficient for discriminating subjects with and without CAD because of considerable overlaps. In contrast to these findings, we found optimal cut-off levels for ROC curves to be 6.4 mg/L for males, and 6.9 mg/L for females with CAD. However, the potential limitations of this study warrant consideration. Although CRP was an independent discriminator between the PG and the CG patients, there was considerable overlap in CRP concentrations between the two populations. The presence of occult or ongoing coronary artery disease in the CG could not be excluded and may have attenuated the differences among the two groups, whereby the diagnostic value of CRP in CAD may be underestimated.

Conclusions
In conclusion, serum CRP level measurement appears to have an important diagnostic value in predicting the presence of angiographically detectable coronary artery disease. However, the higher levels of serum CRP seem not to be associated with the degree or severity of coronary artery disease.