This was hospital based observational case control study conducted in the Department of Biochemistry, tertiary care teaching hospital in Karnataka. The study protocol was approved by Institutional Ethics Committee (IEC) with reference number SNMC/IECHSR/2014-15/A-18-1.1. Study was conducted over a period of one year from January to December 2016.

The study comprises of 120 subjects. Sample size was calculated using open-epi version 2.3.1 software. Taking Mohan V et al., reference, at 95% confidence interval and 80% power, sample size was 57 in each group [11]. Sixty patients who were clinically diagnosed of CAD coming to Medicine, Cardiology OPD and admitted patients were chosen for the study. The CAD patients were selected on the basis of coronary angiography i.e., presence of atleast 50% stenosis in major coronary arteries, ECG, Echocardiograph and previous medical records. Age and sex matched 60 healthy controls in the age group of 40-80 years, who were non-diabetic, non-hypertensive and nonsmokers were enrolled in the study. Presence of CAD in controls was ruled out by detailed clinical history, ECG pattern, echocardiography and family history. Patients with thyroid disorders, nephrotic syndrome, stroke, chronic liver and renal diseases, cancer, pregnant women and women on oral contraceptives, hormone replacement therapy were excluded from the study group.

Informed consent was taken from all participants before the study. Care was taken to maintain the confidentiality of the patients as per Helsinki declaration. For each patient detailed history was taken on smoking, family history, medication and medical records. General physical examination and anthropometric measurements like height and weight were recorded. Body Mass Index (BMI) was calculated using the formula Wt (Kg)/Ht (m²).

Under aseptic precaution 6 mL of fasting venous blood was drawn and serum was separated by centrifugation. The serum was used for estimation of biochemical parameters like blood glucose (FBS), lipoprotein (a) and lipid profile fasting. Two mL blood was collected in sodium floride bulbs for glucose estimation. Glucose was determined by glucose oxidase and peroxidase (GOD –POD) method. Total cholesterol by cholesterol oxidase and peroxidase method, TG’s by glycerol phosphate method. HDL-C estimated by direct method and LDL-C was calculated by Friedewald’s formula i.e., LDL=TC-TG/5-HDL. Lp(a) was estimated by immunoturbidemetric method [12]. All parameters were measured in fully automated analyser, Biosystem A25 using Biosystem kits on the same day. Lipid ratios were calculated by using formulae TC/HDL, LDL/HDL, and TG/HDL.

All statistical analysis was done using software Statistical Package for the Social Sciences (SPSS) version 13. Quantitative data has been expressed in terms of mean±SD. Student’s t-test was used for comparison of groups and Pearson’s correlation test was applied for correlation of Lp(a) with other parameters. Receiver Operating Curve (ROC) was put to find the best cut off value for Lp(a) in CAD patients. A p<0.05 was considered as statistically significant.

The 60 CAD patients were in the age group of 40 to 80 years. Among them 38 were male and 22 were female patients and there was no significant difference in age and sex ratio between cases and controls. Males are affected more in our study, this may be due to (29) 48% of CAD patients were smokers. The BMI, systolic and diastolic blood pressure and fasting blood glucose were significantly (p<0.05) higher in CAD patients as compared to controls as depicted in [Table/Fig-1].

There was statistically significant increase (p<0.001) in TGs and VLDL and significant decrease (p<0.001) in HDL in CAD patients compared to control group. Even though, there was difference in TC and LDL but it was not statistical significant. There was significant increase (p<0.001) in Lp(a) in CAD patients compared to controls. Mean Lp(a) levels in CAD patients were 52.34±24.87 which was highly significant (p<0.001) compared to controls 16.81±3.91. There was significant increase in lipid ratios TC/HDL, TG/HDL and LDL/HDL in cases compared to controls as shown in [Table/Fig-2]. Distribution of Lp(a) among the CAD patients was shown in [Table/Fig-3]. Majority of the CAD patients i.e., 43.3% had Lp(a) levels in the range of 41-60 mg/dL followed by 31.6 % in the range of 21-40 mg/dL. The cut-off value for Lp(a) was 23.7 mg/dL in CAD patients in our study. Most of CAD patients had Lp(a) levels were above this value.

[Table/Fig-3]:

Percentage of CAD patients depending on Lp(a) levels.

[Table/Fig-4] shows the correlation of Lp(a) with other lipid parameters. Lp(a) shows significant positive correlation with TC and LDL where as negative correlation with HDL. There was also positive correlation with TC/HDL and LDL/HDL. ROC for Lp(a) in CAD patients is shown [Table/Fig-5]. Area under the curve is >0.8 which is highly significant (p<0.001). In our study we found cut-off value was 23.7 mg/dL with 78% sensitivity and 86% specificity shown in [Table/Fig-6]. This suggests that Lp(a) above this cut off value can be used to evaluate risk of CAD in our region.

[Table/Fig-5]:

ROC curve of Lp(a) to detect the best cut-off value in CAD patients.

In the present study we evaluated the significance of novel biomarker i.e., Lp(a) in CAD patients. Recently Lp(a) is emerging as a strong biomarker of CAD. Our result shows high levels of TG and low levels of HDL and is strongly associated with CAD. A large cross-sectional study conducted in North India by Ashfaq F et al., found increased TG and very low HDL levels in diseased vessel patients as compared to those with normal coronary status [13]. They also indicated that usefulness of estimation of TG to identify the presence of CAD even when the patients are receiving treatment i.e., hypolipidemic drugs like statins. Austin MA et al., reported TG levels more than 130 mg/dL are strongly associated with the extent of patient’s coronary atherosclerosis and states that hypertriglyceridemia is commonly associated with CAD patients [14]. Low levels of HDL are reported to increase the risk of cardiovascular disease even when cholesterol is not elevated [15,16]. Low HDL level is an independent risk factor for CAD [17]. In our study, we did not find any statistical difference in TC and LDL levels in CAD patients compared to controls. Atherogenicity of Lp(a) is more marked with concomitant decrease in HDL levels which was also seen in our study [18].

In this study we estimated Lp(a) levels in CAD patients and controls. We also made an attempt to find out cut-off value for Lp(a) levels using ROC. We found higher levels of Lp(a) in CAD patients, our results are in agreement with other studies [19-21]. Ashfaq F et al., reported that Lp(a) levels correlated positively with severity of atherosclerosis [19]. Lp(a) levels of 21mg/dL are associated with presence of CAD even when cardiovascular risk factors and specific treatments were taken. In north Indian studies by Yusuf J et al., and Gupta R et al., showed highest Lp(a) levels were observed in triple vessel disease, followed by double and single vessel disease [20,21]. CAD patients showed almost five fold higher Lp(a) levels as compared to controls [22]. Zampoulkis JD et al., studied the relationship of excess Lp(a) with extent and severity of atherosclerosis in CAD patients. They revealed that Lp(a) is related to diffuse lesions covering large part of coronary vasculature [23]. Lp(a) competitively inhibit plasminogen activity, leading to impaired fibrinolysis. The smaller the apo (a), the higher are the Lp(a) levels and the risk for CAD. It enhances the oxidation and foam cell formation.

In our study, cut-off value was 23.7 mg/dL with 78% sensitivity and 86% specificity. A study on 151 south Indian patients, by Rajashekhar D et al., shown Lp(a) >25 mg/dL is associated independently with around two fold risk of CAD [10]. A north Indian study by Hoogeveen RC et al., had proposed a cut off value was >19 mg/dL in 103 subjects [24]. Another largest north Indian study by Yusuf J et al., showed Lp(a) above 40 mg/dL assessed by an isoform insensitive assay is an independent risk factor for CAD [20]. According to Saini V et al., Lp(a) showed independent predictor of CAD irrespective of traditional risk factor with good sensitivity and specificity. The best cut-off value of Lp(a) came out to be 45 mg/dL [9]. Lp(a) levels above 20mg/dL are reported to be associated with a high risk of CAD [25]. Studies found that patients with Lp(a) levels more than 30 mg/dL had risk of CAD increased by about three fold [26,27].

We noted there was significant positive correlation of Lp(a) with TC, LDL, TC/HDL and LDL/HDL and negative correlation with HDL which are in agreement with other studies [28,29]. A large cohort study from five European countries comprising of 56,804 individuals reported that elevated Lp(a) was robustly associated with an increased risk of major coronary events. Total cholesterol correlated positively with Lp(a) levels [28]. A study by Pedreno J et al., reported Lp(a) levels associated with LDL cholesterol [29].

Small sample size, genetic data is not provided. It is recommended to conduct study in larger sample size.

In this study we were able to show Lp(a) as one of the reliable marker, as Lp(a) was found to be raised significantly in CAD patients compared to controls. Since Lp(a) levels show wide ethnic variation, an attempt has been made to derive the best cut-off value for Lp(a) levels in CAD patients in North Karnataka population.