Abstract
Although we partially assume the scrutiny of data provided by Whayne, 1 we do not share some of his conclusions. We understand the reluctance to abandon a treat-to-target low-density lipoprotein-cholesterol (LDL-C) approach. This is probably because this has contributed in the past to simplify the task of clinicians and to introduce the concept of “global” cardiovascular (CV) risk factor. These reasons do not necessarily help in the clinical decision task and are no longer sustainable.
Clinicians do not need other guidance than a thorough knowledge of the disease and scientifically based data regarding the risk of mortality and morbidity, including adverse effects, derived from their decisions. In this context, the only data with scientific basis are that lipoprotein disturbances play an important and causative role in the pathogenesis of atherosclerosis and that statins provide a rapid and sustained benefit in reducing the CV risk. 2 There is also evidence that mechanisms providing these benefits are not limited to improvement in the lipid profile. 3
In complex diseases, decisions based on surrogates are not always innocuous and may contribute to relax the necessary holistic approach. It is of no value to decrease blood pressure, lower blood lipids, or modify any CV risk if the procedure may lead to deleterious effects or fail the main objective. 4
In atherosclerosis, we do not support a treat-to-target LDL-C approach, among other reasons, because LDL-C is neither a target nor an adequate surrogate as detailed subsequently. We acknowledge, however, that undertreatment of lipid alterations even in patients with major CV events and/or advanced peripheral artery disease is a frequent condition. Whether the lack of support to this approach may increase the amount of patients not receiving adequate management cannot be inferred from available data.
One of our main concerns is methodological. To trust a surrogate marker, accuracy is a necessary condition. Our data conclusively support that this is not the case for LDL-C. First, we should not consider LDL as a simple variable. Lipoproteins are heterogeneous particles with different and variable size and composition that circulate interacting among them. Moreover, LDL does not contribute much to estimate CV risk as compared with other lipid measurements and further contributing factors, such as high-density lipoprotein cholesterol (HDL-C), triglycerides (TG), C-reactive protein, and others. 5,6 Second, there is no knowledge on the differential effect of available treatments in the long and short term and the effect of cumulative errors inherent to laboratory measurement and intraindividual variations or degree of adherence to prescribed treatments. Third, the heterogeneous nature of LDL is also accompanied by heterogeneous functional properties of each lipoprotein class and consequently both the relative amount and the qualitative composition should be considered. 7–9
Finally, there is no available method to accurately measure LDL-C. Ultracentrifugation and liquid chromatography are the gold standard methods to measure LDL-C but are extremely expensive, poorly suited for massive screening, and certainly not free of inaccuracies associated with the necessary technical skills and the accumulation of remnants and lipoprotein (a) in the LDL fraction as previously discussed. 10 The Friedewald equation and the consequent calculations currently enjoy a wide acceptance in most clinical laboratories, but it is only totally reliable in healthy population in which it is obviously unnecessary. The formula and their variations overestimate LDL-C, apply poorly to hyperlipemic individuals, it is absolutely inaccurate to monitor changes in patients under dietary interventions or modifications in lifestyle and should not be used in conditions with a high CV risk (eg, obesity, diabetes, and hepatic, renal and other metabolic conditions). 11,12 It is therefore unsuitable for the desired purpose. In addition to these limitations, the likelihood to achieve the standards of the National Cholesterol Education Program (coefficient of variation <4%) is low, as it aggregates the measurement errors of total cholesterol (TC), HDL-C, and circulating TGs. 13,14
The propaganda involved in the use of LDL-C as a clinical target has provided the necessary stimuli to develop direct methods of LDL-C measurement based on antibodies and/or combination of detergents. However, these methods are expensive, show a limited performance, they are not free of limitations in the presence of diseases that alter LDL structure and/or composition, and consequently are seldom used in clinical laboratories. 15–17
In conclusion, the treat-to-target LDL-C approach is not only debatable but also inaccurate and probably deleterious. Clinicians should use a combination of well-established laboratory measurements (TC, HDL-C, and TG) for the assessment of the effectiveness of statin treatment when prescription is considered. The importance of carefully considering that arteriosclerosis mortality may be delayed and decreased wisely using the available knowledge can hardly be overstated.