Simplifying the Approach to the Management of Dyslipidemia

In this issue of JAMA, Kuklina and colleagues¹ report trends over the last decade in cholesterol levels and treatment. While the decreasing prevalence of “high” low-density lipoprotein cholesterol (LDL-C) levels and increasing use of lipid-lowering medications in appropriate individuals are encouraging, the most sobering message in this article is the disappointing rates of screening, awareness, and treatment. Among those with high LDL-C levels, one-third had not been screened, and one-fourth were unaware that their levels were high. About two-thirds of those who were high risk were not receiving medication, although the vast majority of these individuals would likely benefit from lipid-lowering therapy.

With a wealth of data supporting the benefit of treatment with lipid-lowering medications, why are so few persons with elevated LDL-C levels being treated? The authors speculate that one reason for these disappointing rates may be the complexity of existing guidelines. There are several sets of guidelines that clinicians could use, but each guideline has become increasingly complex, with multiple steps to determine who should be treated and demanding procedures to evaluate the success of treatment.

In the wake of early recommendations for smoking cessation and blood pressure control and management, the Expert Panel of the National Cholesterol Education Program (NCEP) introduced the first guidelines² for cholesterol screening and treatment in 1988. These early guidelines played an important role in educating health care practitioners about the risks associated with high cholesterol and provided early guidance for its management. Since then, the cholesterol guidelines have been revised twice, and a fourth version is currently in progress. What have emerged are cumbersome guidelines that are a challenge to implement. The Third Report of NCEP Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (ATP III),³ released in 2002, was 280 pages. Clinicians must decide which patients should receive treatment and how successful that treatment has been. To determine who should be treated, the clinician needs to assess the LDL-C level in the context of the patient's risk to determine whether the patient exceeds 1 of 3 LDL-C thresholds. The patient receiving lipid-lowering therapy is then followed for the attainment of 1 of 3 or 4 fixed LDL-C targets.

In a thought-provoking Commentary in this issue of JAMA, Hingorani and Psaty⁴ propose simplifying the approach for primary prevention of cardiovascular disease (CVD) by extending lipid-lowering therapy with a generic statin (perhaps as a combination-drug “polypill”) to all adults, using age as a threshold for this approach. The authors suggest that 2 factors justify a careful reassessment and simplification of the current guidelines. First, the cost of statins, the most effective lipid therapies, has decreased substantially, and second, the wealth of recent trial data suggest benefits of lipid-lowering therapy among a broad segment of the population with a wide range of CVD risk and across the spectrum of initial LDL-C levels and favorable safety profiles. Careful interpretation of these data has led Hingorani and Psaty to propose a rather provocative approach that could certainly simplify the process. However, even though the available data call into question the current guidelines, there are advantages to a simple risk-based approach that uses multiple risk factors rather than age alone.

When deciding who should be treated, the main goal is to identify those who are at sufficient risk of future events to offset the risk of the intervention (which is quite low for statins) and justify the costs of the intervention. The ATP guidelines rely on the assessment of risk using several risk factors, or a score, and LDL-C levels to trigger initiation of therapy. An important issue is whether both risk levels and LDL-C levels are necessary, especially if the LDL-C levels are used in the risk assessment tool. The linear association between cholesterol levels and CVD risk⁵ provides no biological justification for current threshold levels. Furthermore, fixed thresholds for initiation of treatment create odd dynamics. For example, a 55-year-old man with an LDL-C level just over 150 mg/dL can have a 10-year risk of CVD that ranges from 1% to more than 30% depending on his other risk factors.

Even within an ATP risk category, using the LDL-C thresholds may mean favoring treatment for a patient whose LDL-C level is above a certain threshold but who has a relatively low overall risk, while forgoing treatment for a patient who is at higher overall risk but whose LDL-C level is just below the arbitrary threshold. It would be more cost-effective to treat the second patient. Thus, the current guidelines lead to potential costly overtreatment of those at low risk and potential undertreatment of those at higher risk for CVD. Since treatment with statins lowers risk at all levels of LDL-C, the absolute CVD risk rather than the LDL-C level should be the trigger for initiating treatment. For this reason, New Zealand⁶ and the United Kingdom⁷ have suggested using CVD risk alone as the trigger for initiation of treatment.

Although the individual's risk and not necessarily the LDL-C level should be the main reason for initiation of lipid therapy, using age as the sole determinant of risk may oversimplify the process because risk assessment can be done rather easily. The use of non–laboratory-based risk scores as an initial screen can facilitate this process and reduce costs.⁸

Similarly, once treatment has been initiated, the fixed targets now in place for treatment are not justified by the current literature. Suppose, for example, 2 patients in the cardiac care unit both had myocardial infarctions. One has an LDL-C level of 195 mg/dL and the other an LDL-C level of 101 mg/dL. The rationale for treating both patients to the same LDL-C target is not clear. Both will benefit from substantial LDL-C reduction. Furthermore, the trials of cholesterol reduction have generally not treated to a particular goal. In the Heart Protection Study,⁹ participants had the same relative risk reduction from a statin regardless of starting cholesterol level or history of coronary heart disease. This finding is consistent with meta-analyses of 90 000 patients in 14 primary and secondary prevention trials,¹⁰ with statins showing a 20% reduction in major CVD end points for a 38 mg/dL (1 mmol/L) reduction in LDL-C levels regardless of starting cholesterol level.

Since the reduction in CVD among those treated with statins is proportionate to the amount of lipid lowering, the LDL-C targets can be eliminated. The recently published Canadian guidelines¹¹ suggest a 50% reduction in LDL-C levels, an achievable goal for many patients, as an alternative to a fixed target. The UK guidelines⁷ for lipids suggest a fixed-dose statin without a target for cholesterol in primary prevention. In low-resource settings where monitoring of LDL-C levels is difficult, statins can be safely given in fixed doses.

The elimination of arbitrary, fixed LDL-C thresholds for initiation of therapy and fixed targets as goals of treatment can greatly simplify the identification of patients who warrant treatment and subsequent management. The guideline should begin with simple risk assessment with the goal of classifying patients into only 2 strata: those for whom lipid-lowering therapy should be considered and those for whom it is not warranted. Individuals with known CVD and diabetes mellitus are in the first, higher risk group. For those free of overt CVD or diabetes, the overall total CVD risk should be assessed using a simple score system. If this score is above a threshold deemed appropriate (such as the 20% proposed by the United Kingdom), then these individuals are included in the higher risk group. All patients should be advised of lifestyle interventions that could improve the lipid profile as well as other risk factors. For those in the higher risk category, treatment with lipid-lowering therapy should be considered. Once treatment has been initiated, the goal should be to lower the LDL-C as much as possible without causing adverse effects. The 50% reduction in LDL-C levels suggested by the Canadian guidelines¹¹ is a reasonable benchmark.

Even though there has been progress in identifying and treating patients with dyslipidemia, the current guidelines are overly complicated, and a simplified risk-based approach is supported by the current data. Abandoning the fixed LDL-C threshold and targets used in many guidelines is justified by the linear relationship of cholesterol lowering and the benefit of the intervention for preventing CVD. The use of a simplified risk-based approach could increase the ease of implementation of treatment and increase the number of patients receiving beneficial lipid-lowering therapy.