ABC | Volume 113, Nº2, August 2019

Original Article Yan et al. The effect of LP(a) on chronic heart failure Arq Bras Cardiol. 2019; 113(2):197-204 enrolled over a continuous period between January 2014 and December 2016. Chronic HF was diagnosed by two cardiologists based on 2016 European Society of Cardiology guidelines. 18 Patients were enrolled based on the following criteria: 1) The etiology of chronic HF is CHD; 2) patients with HF in New York Heart Association functional class II to IV. Patients were excluded according to the following criteria: 1) chronic HF secondary to other heart diseases, such as valvular heart disease, obstructive hypertrophic cardiomyopathy, and myocarditis and pericardial disease; 2) complicated with infectious diseases, autoimmune diseases, malignant tumors, severe liver and end-stage kidney disease with dialysis and systemic disease such as hyperthyroidism; 3) removal of patients who lack clinical data; 4) administration of medications that affect Lp(a) levels (nicotinic acid including niceritrol, tocopherol nicotinate, and nicomol). Data at the first admission were collected for patients with multiple hospitalizations. Hypertension was defined as systolic blood pressure (BP) ≥ 140 mmHg or diastolic BP ≥ 90 mmHg on repeated measurements, or the use of antihypertensive medication. Diabetes mellitus (DM) was defined according to the World Health Organization criteria. 19 We assessed the estimated glomerular filtration rate (eGFR) according to the Chinese Modification of Diet in Renal Diseases equation based on serum creatinine, age, and gender. 20 This study was approved by the ethics committee of the First Affiliated Hospital of Jinan University and is in accordance with the Declaration of Helsinki. Written informed consent was obtained from the participants involved in the study. Laboratory measurements The venous blood samples were usually obtained on the 2 nd morning of admission after an 8-hour fasting. Serum Lp(a) levels were measured by latex agglutination immunoassays and apolipoproteins were determined by fixed-rate immunonephelometric assay using a HITACHI 7600 chemistry autoanalyzer (Hitachi High-Technologies Corporation, Tokyo, Japan). Echocardiography Transthoracic echocardiographic examination was performed on each enrolled patient by an experienced ultrasonographist using a Philips IE33 (Philips Healthcare, the Netherlands) cardiac ultrasound systemmachine within 24 to 48 hours after admission and within 24 hours after the primary PCI. Left ventricular ejection fraction (LVEF) was estimated with the modified Simpson method. Clinical outcome and follow-up The primary outcome evaluated in the present study was recurrent HF. The patients included in this study were followed for 1, 3, 6, 9, and 12 months after discharge with 1) access to medical records, outpatient electronic workstations, and medical insurance system; 2) telephone or face-to-face visits. For follow-up failures, we contacted the patients' family or neighbors, or even their workplaces. Statistical analysis First, continuous variables with normal distribution were expressed as mean (standard deviation [SD]); non-normal variables were reported as median (interquartile range [IQR]). Categorical variables were described as numbers and/or percentages. Means of 2 continuous normally distributed variables were compared by independent samples Student’s t test. Mann-Whitney U test was employed to compare means of 2 groups of variables not normally distributed. The frequencies of categorical variables were compared using Pearson χ 2 test. Second, patients were divided into 2 groups according to whether Lp(a) levels were above or below the median level for the entire cohort (20.6 mg/dL): the high Lp(a) group (n = 155) and the low Lp(a) group (n = 154). The event-free rate for recurrent HF was plotted using Kaplan−Meier method with the log-rank test. Third, we analyzed the association of plasma Lp(a) levels as a continuous variable and as categorical variables with recurrent HF. Cox proportional hazards models were used to evaluate these associations, both with and without adjustment for confounding variables. In the adjusted regression model I, number of stents, multiple lesions, aldosterone antagonists, LN-NT-proBNP, SBP, and NYHA class were included. Model II was further adjusted for the same variables as Model I plus the following risk factors: gender, DM, atrial fibrillation (AF), hypertension, LAD lesion, prior PCI, two lesions, diuretics, ACEI/ARBs, digoxin, beta‑blockers, anti‑diabetic drugs, heart rate, total cholesterol (TC), potassium, high‑density lipoprotein‑cholesterol (HDL-C), low-density lipoprotein‑cholesterol (LDL-C), triglycerides, hemoglobin, LVEF, age, body mass index (BMI), and eGFR. TheMedCalc software, version 15.2.2, was used to calculate the clinical outcomes with relative risk and 95% confidence interval (CI). The Cox proportional hazards models analyses was performed using the EmpowerStats statistical software (http:// www.empowerstats.com, X&Y Solutions, Inc. Boston, MA) and the statistical package R (http://www.R-project.org ). A 2-sided p < 0.05 was considered statistically significant. Results Baseline characteristics A total of 309 patients with chronic HF caused by CHD were enrolled in this study. 31 patients (10.03%) could not be reached during follow-up. The mean age of the patients was 68.6 ± 11.6 years, and 174 (56.3%) were males. The medians (IQR) of two groups of Lp(a) levels were 12.0 (7.6-16.6) mg/dL and 35.3 (25.4-52.0) mg/dL, respectively (p < 0.001). Baseline characteristics and laboratory results, past medical history, and medications at discharge are shown in Table 1. There were differences in NYHA class, LVEF, NT-proBNP and Lp(a) levels, prior MI, prior PCI, multiple lesions, and number of stents between 2 groups. Meanwhile, there were no differences regarding medication at discharge, age, male gender, current smoker, BMI, heart rate, eGFR, conventional lipid profile, DM, AF, hypertension, and prior CABG. 198