The study herein is a single-center and retrospectively designed study, consisting of 156 consecutive patients eligible for selective (CAG) between May 2013 and July 2014 in our clinic. The study was approved by the local ethics committee, and was implemented in complete concordance with the Declaration of Helsinki on human research. We selected only subjects with systolic HF that persisted with an unknown etiology after an initial non-invasive clinical assessment. Inclusion criteria were: left ventricular ejection fraction (LVEF) <40% with global left ventricular hypokinesia and age of ≥18 years.

Information including gender, age, smoking status, hyperlipidemia, hypertension (HT), and diabetes mellitus (DM) were gathered. The definition of HT appeared as a systolic blood pressure (BP) value of ≥140 mmHg and/or a diastolic BP value of ≥90 mmHg at least on >2 office BP measurements or being on an antihypertensive therapy. The definition of DM comprised a blood sugar value of ≥126 mg/dL (7.0 mmol/L) in the fasting state or being on an antidiabetic therapy whereas the status of hyperlipidemia was based on the presence of a blood cholesterol level of ≥200 mg/dL or a triglyceride level of ≥150 mg/dL in the fasting state. Smoking was defined as currently smoking or ex-smokers who forwent smoking within the past 6months.

Patients with moderate or advanced cardiac valvulopathy, malignancy, clinically decompensated congestive HF, haematological disorder, severe kidney or liver dysfunction, active infection or systemic inflammatory conditions, or autoimmune disorders and patients using steroids were excluded from the study.

After an overnight fasting, peripheral venous blood samples were drawn from patients. Dry tubes were used for biochemical analysis, and EDTA tubes were used for hematological tests. White blood cells and erythrocyte counts, hemoglobin along with hematocrit levels were analysed by an automated hematology device Coulter Counter LH Series (Beckman Coulter Inc, Hialeah, FL, USA). The biochemical measurements were determined by an automated analyzer of biochemistry (Abbott Laboratories, Abbott Park, IL, USA).

All patients enrolled in the study underwent selective CAG by Judkins technique through the femoral artery. The Gensini scoring system was harnessed to identify the severity of coronary artery disease.10) According to this score, 1% to 25% of stenosis in the lumen of coronary arteries is assigned a score of 1 whereas the stenoses of 26% to 50%, 51% to 75%, 76% to 90%, 91% to 99% and 100% (totally occluded) represent the scores of 2, 4, 8, 16 and 32. The score based on the stenosis degree is thereafter multiplied by a constant number in accordance with the location of lesions along the coronary arteries. A multiplication constant of 5 points is generally assigned for left main coronary artery (LMCA) lesions whereas 2.5 points are assigned for left circumflex (LCX) or proximal left anterior descending (LAD) artery lesions. Regarding LCX and LAD, lesions involving mid segments and distal segments are given the points of 1.5 and 1, respectively. The lesions involving the first obtuse marginal (OM) and diagonal branches, intermediate and posterior descending arteries (PDA) are generally multiplied with a constant of 1 point while the lesions in second OM and diagonal branches are assigned the multiplication constant of 0.5. Assessment of Gensini score was made by 2 interventional cardiologists totally blinded to the laboratory and clinical data and the coronary evaluation of the other cardiologist. In the case of disagreement, a third blinded interventionalist assessed the coronary angiograms.

According to their coronary angiograms, subjects were categorized into three groups. The first group comprised 60 patients with a normal CAG (control group). The rest was categorized into two based on their Gensini score: mildly atherosclerotic group (n=64; score of <25 points) and severely atherosclerotic group (n=32; score of ≥25 points).

All echocardiographic examinations were performed by a certified cardiologist experienced in this field using a Vıvıd-7 (GE Healthcare, Horten, Norway) device in compliance with American Society of Echocardiography (ASE) guidelines.11) LVEF was calculated from apical 4-chamber views, according to a modified Simpson's rule.

Continuous variables were defined as means and standard deviation; categorical variables were presented as percentages. The normality of distribution for continuous variables was confirmed with the Kolmogorov-Smirnov test. According to the distribution pattern of continuous variables, Mann-Whitney U test or independent-sample t-test was used for continuous variables, and the chi-square test was used for categorical variables. One-way analysis of variance (ANOVA) or Kruskal-Wallis test was used to compare 3 groups. When there was a significant difference between the three study groups, the comparison of two groups in terms of the relevant parameter was performed with post hoc tests in a one-way ANOVA and with Mann-Whitney U test after Bonferroni correction in Kruskal-Wallis test. Correlations were assessed using either Pearson's correlation test or Spearman's rank test according to the distribution pattern of the variable. Independent associations between Gensini score and independent variables were evaluated through a backward stepwise multiple linear regression analysis by comprehending all parameters showing p value of less than 0.1 on a univariate analysis. Coefficients of standardized β regression along with their significance from the multiple linear regression analysis were also reported. A receiver operating characteristic (ROC) curve analysis was used to determine the optimum cut-off level of preprocedural PLR values to predict severe coronary atherosclerosis. Statistical analyses were performed using SPSS 16.0 (SPSS Inc., Chicago, IL, USA). A two-tailed p value of <0.05 was regarded as statistically significant.

There are several limitations regarding the present study. Sample size was small, enrolment was retrospective and it was a single center design. The amount of coronary atherosclerosis was evaluated only by a coronary angiography. We know it does not provide sufficient data about the coronary plaque burden, and shows only the lumen of coronary arteries. An intravascular ultrasound and/or coronary CT may provide necessary data about the amount of coronary atherosclerosis. We had no data on inflammation markers including C-reactive protein, tumour necrosis factor-alpha, etc. However, assessment of these substances may strengthen the study. We did not analyze other platelet indices including mean platelet volume, platelet distribution width, and inflammatory indices including neutrophil lymphocyte ratio (NLR) or correlation of these parameters with the PLR. Notwithstanding these limitations, this is reportedly the first study focusing on PLR with regard to its predictive value in asymptomatic low EF subjects, and more comprehensive studies are still warranted to corroborate our findings.

This is reportedly the first trial primarily focusing on the potential association between PLR and the severity of atherosclerotic heart disease in asymptomatic low EF patients. Our results indicate that harnessing the index of PLR as well as a variety of clinical variables including advanced age, DM and high LDL levels may work as valuable predictors of atherosclerotic burden, hence severity of atherosclerosis. However, the current information on this issue is still at its infancy suggesting the need for future clinical studies to validate absolute implications of this index in the setting of atherosclerotic heart disease.