A Comparison Of Low Density Lipoprotein Cholesterol Concentration Estimation By The Roche Diagnostics Direct Measurement : Comparison With Other Lepidic Cardiovascular Risk Estimation Biomark

ABSTRACT

The primary target for atherosclerosis associated cardiovascular (AACV) risk prediction and treatment assessment is the lipid profile fraction of low-density lipoprotein cholesterol (LDLC). The routinely used Friedewald method for LDLC estimation has an unacceptably high analytical error. In this study, patient sera LDLC was estimated using the Friedewald method (LLC) and the Roche Diagnostics direct homogeneous method (dLDLC). Serum lipid calculations of non-HDLC, HDLC/LDLC ratios, and sdLDLC were made. A comparison of the cLDLC and the dLDLC methods in the estimation of the LDLC and the classification of the sera into CV risk categories using the cLDLC, dLDLC, non-HDLC, HDLC/LDLC ratios, and sdLDLC was done. Upon using dLDLC in place of cLDLC, 41 out of 121 (34%) serum samples had their CV risk categories changed. This corresponds to a clinical discordance of 34%. Forty of the 41 CV risk category changes were in the form of a decrease by one CV risk category. Also, many other samples showed a decrease in LDLC estimation which did not result in change in CV risk category. The paired t test calculations showed that the means of the cLDLC method and the dLDLC method were significantly different at p < 0.01. Upon using non-HDLC in place of cLDLC, 39 out of 121 (32%) serum samples had their CV risk categories changed. This corresponds to a clinical discordance of 32%. Twenty seven of the 39 CV risk category changes were in the form of a decrease by one CV risk category. Upon using non-HDLC in place of dLDLC, 43 out of 121 (36%) serum samples had their CV risk categories changed. This corresponds to a clinical discordance of 36%. Thirty of the 43 CV risk category changes were in the form of an increase by one CV risk category. On all the lipid pairs (cLDLC and dLDLC, cLDLC and non-HDLC, and dLDLC and non-HDLC), the CV risk categories and the paired t test showed that there were significant CV risk category changes and statistically significant result differences between methods although the correlations were very high and in a linear manner. The calculated CV ii risk category changes and the statistical outcomes observed in this study show that the cLDLC significantly overestimates the LDLC levels compared to dLDLC, in patient sera with normal or slightly elevated triglycerides (TGs) levels, classifying the patients into higher CV risk categories. On the Low Risk/Risk category analysis, all the methods gave significantly different results, with sdLDLC giving the highest number of sera samples classified into the Risk category. The CV risk classification varies with the method used. Further studies using large samples taken from different ethnic and geographic populations and preferably compared with the Beta Quantification reference method are needed to verify these findings. Institutions in the same geographic location must use the same LDLC method (preferably the dLDLC) for CV risk estimation and treatment assessment. The dLDLC must replace the cLDLC for AACVD risk prediction and treatment assessment. Also, more population based studies must be conducted to establish the best lipid biomarker and lipid panel for CV risk estimation and treatment assessment.