The Development Of Predictive Model For Coronary Artery Calcification In Maintenance Hemodialysis Patients

Objective:As the golden criteria of coronary artery calcification evaluation was multi-slice spiral computed tomography. To explore the predictive value of various evaluative methods for vascular calcification in the assessment of coronary calcification in maintenance hemodialysis patients, analyse the risk factors for coronary artery calcifition, and develop the predictive model for coronary artery calcification.Methods:We selected thirty-six patients from our hemodialysis center. For each patient, we collected lab test results and did multi-slice spiral computed tomography, echocardiography, X ray, pulse wave velocity examination, and questionnaires. The lab tests we have done were as follows: white blood cell count, hemoglobin, platelet count, BUN, creatinine, albumin, C-reactive protein, total cholesterol, triglyceride, calcium, phosphorus, intact parathormone, ferritin. Serum sample was collected within one week for 1,84-PTH and 25-(OH) vitamin D3 tests. Group comparision was done using t-test for normal distributed continuous variables, Fisher's exact test for discrete variable, and Mann-Whitney test for non-normal distributed continuous variables. Logistic regression and ROC were used to evaluate the predictive power of various X-ray assessment methods. Ultimately, regression analysis was performed to develop a model which can be used to predict the presence of coronary artery calcification. Receiver operating curve was performed to examine the predictive power of the model.Results: 1. The proportion of patients who had evidence of CAC was 88.9%. The mean CACs was 1010.472 (range:0-10761.2). Four patients had CACs=0,nine patients had 00). Compared with patients without calcification(CACs=0), patients with calcifition(CACs>0) had higher age, BMI and PP, and lower 25(OH) vitamin D3 levlels. Compared to CACs<100 group, patients in CACs≥100 group had higher age, serum urea nitrogen and lower pre- dialysis/ post-dialysis diastolic blood pressure, and often had a history of coronary heart disease. Although serum P and CRP were higher and serum 25-(OH) vitamin D3 was lower in the group of CACs≥100, there were no significant differences between these two groups.6. Regression analysis demonstrated that the presence of coronary artery calcification was correlated with age, BMI, pulse pressure and 25-(OH) vitamin D3(P<0.05), and the extent of coronary artery calcification was correlated with age, the duration of hypertension, and pre-dialysis/ post- dialysis diastolic blood pressure(P<0.05).7. Multiple regression analysis demonstrated that 25-(OH) vitamin D3 and PP were two independent risk factors for coronary artery calcification, and age was an independent risk factor for the extent of coronary artery calcification.8. Predictive model for coronary artery calcification in maintenance hemodialysis patients with score higher than 100 was: Y= 0.280*age+0.027* duration of HBP+1.406* the score of four score pattern of iliac artery and femoral artery calcification-0.578*mPWV. ROC Curve analysis showed that this formula had a good prediction of coronary artery calcification score, with an area under ROC Curve of 0.958 (95% confidence interval was 0.000-1.000), the cut-off value best predicting coronary artery calcification was 7.9, whose specificity was equal to 92.3% and sensitivity was equal to 86.4%.Conclusion1. Coronary artery calcification is common in MHD patients, calcification is often involved in left anterior descending branch coronary artery and right coronary artery. Elder and patients who have coronary heart disease or diabetes mellitus have higher CACs.2. Plain X-rays can be first preformed to assess vascular calcification in MHD patients. Compared with signal site X-rays and multi-site X-rays and other methods, pelvis AP X-rays is the best assessment method for vascular calcification assessment. Using four score pattern of iliac artery and femoral artery calcification (namely, left and right iliac artery calcification is 2, left or right iliac artery calcification is 1, left and right femoral artery calcification is 2, left or right femoral artery calcification is 1, total 4 score) is as well as categorical bivariable model of iliac artery or/and femoral artery (namely, patient had iliac artery calcification or femoral artery calcification is 1, otherwise is 0).3. PP and 25-(OH) vitamin D3 are two independent risk factors for coronary artery calcification, and age is an independent risk factor for the extent of coronary artery calcification.4. The model which allows us to predict the extent of coronary artery calcification in MHD patients is: Y=0.280*age+0.027*duration of HBP+1.406* the score of four score pattern of iliac artery and femoral artery calcification-0.578*mPWV. Patient who gets Y > 7.9 is predicted the presence of CACs≥100.