Impact of Change in HDL Measurement Technology on the Prediction Atherothrombotic Disease Risk

Science occasionally achieves knowledge by a giant leap into the unknown, but it more often does so by the continuous accumulation of knowledge in a layeron- layer effect, similar to the layering of sediments in a lake bed. As Isaac Newton once remarked, "If I have seen a little farther, it is because I have stood on the shoulders of giants." So it is with the fields of lipidology and preventive cardiology. The Framingham Heart Study (FHS) led the way in identifying risk factors for atherothrombotic disease (ATD), showing that lipids were a prime cause of ATD and that the spectrum of lipids typical of the ATD population was different from the lipid spectrum of the population without clinical ATD. The FHS eventually developed the concept of the CT (Total Cholesterol):HDL ratio, or Framingham Fraction (FF), as the best lipid predictor. The FF predicts the population at risk of ATD better than LDL by itself.^1–5

In or around the year 2000, the methodology for measuring HDL changed, and manufacturers of the autoanalyzers switched from the precipitation (indirect) method of measuring HDL to the enzymatic (direct) method. This method change was not much publicized, and disagreements between the two measurements may exist. Since lowdensity lipoprotein (LDL) is not usually measured but rather calculated by the Friedewald equation, this means that, of necessity, the LDL calculated when the direct measurement of HDL is utilized may differ from the result that would have been obtained had the indirect method been utilized.

This is important because many clinical trials have been done using the indirect measurement of HDL.^6,7,8,9 which in turn may have impacted results and subsequent interpretations. In other words, the lipids on paper, as determined by the direct method of HDL measurement, may not reflect the atherogenic potential of the lipids, based on the indirect method of HDL measurement, at the level of the arterial wall.

These points are not trivial. I previously reported a case of a 55-year-old patient who sustained an acute myocardial infarction (AMI).¹⁰ There was no obvious reason to measure his lipids - no obesity, cigarette smoking, hypertension, diabetes or even a family history of ATD. He did have recurrent bouts of job-related depression. At the time of his AMI, a lipid panel done at the hospital on admission showed an LDL-C of 143 mg/dl, HDL-C of 43 mg/dl and TG of 66 mg/dl. Many practitioners may have elected not to treat such a lipid panel in a primary-prevention scenario. However, changes in HDL measurement technology might have led to different HDL-C and LDL-C values that would have caused his risk to be interpreted differently and perhaps earlier and more aggressive lipid management. With the introduction of new technology, I requested that our hospital laboratory take this into account. Other authors^11,12 have commented on the effects of this change in technology and the interpretation of current lipid values but have not suggested a remedy. I believe my suggestion constitutes such a remedy, and I urge lipidologists start a dialogue with their laboratories regarding these issues.

NB: Readers of the LipidSpin will know that my preferred lipid predictor is the cholesterol retention fraction ([LDL-HDL]/ LDL) because it predicts the population at risk of ATD better than the LDL:HDL ratio and avoids the pitfalls of FF.¹³

Disclosure statement: Dr. Feeman has no disclosures to report.