Rajarshi Sarkar
Background: Reporting directly measured homogeneous Direct Low-Density Lipoprotein Cholesterol (D-LDL) becomes imperative especially in tertiary level laboratories which have to frequently deal with dyslipidaemic specimens. In absence of credible studies involving dyslipidaemic specimens on comparability of test results between different platforms or between different platforms and reference methods, reporting of D-LDL becomes very uncertain.
Methods: The current study includes 328 subjects classified from Type I to Type V according to Fredrickson classification of dyslipidaemia. Standard lipid profile including D-LDL was tested on their serum specimens and D-LDL was repeat tested after saline dilution on three platforms viz AU5800, Alinity ci and Cobas Pure. Calculated LDL-Cholesterol for all the specimens were derived from the NIH equation proposed by Sampson, et al.
Results: Mean Absolute Percent Variation (MAPV) between D-LDL and C-LDL for each class interval was found to increase with increasing triglycerides concentrations of the specimen and at the two extremes of non-High-Density Lipoproteins (HDL) cholesterol concentrations. Passing-Bablok regression, Bland-Altman plot and Receiver Operating Characteristic curves constructed for each dyslipidaemia phenotype revealed that AU5800 outperformed the other two for Type II and III specimens, while Alinity ci and Cobas Pure outperformed AU5800 for Types I, IV and V specimens.
Conclusion: Variation in test results of D-LDL in dyslipidaemic specimens on most widely used platforms is a matter of concern as it might lead to misclassifications in diagnosis and treatment monitoring.