How Low Should We Go With the LDL

Introduction These days, even most laypersons know that low-density lipoprotein (LDL) cholesterol is the "bad" cholesterol and that the class of drugs known as statins (3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors) is a proven treatment for patients with elevated levels of serum LDL cholesterol. For the physician, this guideline-mandated, evidence-based treatment strategy is supported by the results of numerous large, randomized, placebo-controlled clinical trials,[1-6] which have demonstrated highly significant reductions in the risk of death or cardiovascular events in populations with (secondary prevention) or without (primary prevention) a history of coronary artery disease (CAD). The 2003 National Cholesterol Education Program guidelines recommended that patients with established CAD or diabetes (sometimes called a "CAD equivalent") should be treated to reach a target LDL cholesterol level < 100 mg/dL.[7,8] In demonstrating their protection against future events, these first placebo-controlled statin trials employed standard doses to reduce LDL cholesterol levels by 25% to 35%. In other words, large reductions in LDL cholesterol can be achieved, and the guidelines set a target level -- do we have a final answer on how to treat our CAD (and diabetic) patients? While statins have been proven over the last decade to be among the safest and most efficacious drugs in the physician's armamentarium, these historic results have still left some remaining questions: What LDL level should be the threshold for initiating statin treatment? What kind of reductions, with what kind of benefit, have subsequent trials with more potent ("second generation") statins and/or higher doses of the first-generation statins demonstrated? How low can/should LDL cholesterol be lowered? Is there a "J" curve, below which the risk outweighs the benefits with the more intensive statin dosing? How High to Start Treatment? The question of the threshold for initiation of statin treatment was investigated in the Heart Protection Study (HPS), the largest study at that time to investigate the benefits of cholesterol lowering, and the results were presented at the American Heart Association meeting in 2001. While reinforcing the results of the earlier trials, HPS also reported that, regardless of the baseline LDL cholesterol level (above or even starting below 100 mg/dL), patients received the benefit of protection against cardiovascular events with the standard dose of simvastatin used in the study.[4] One simple way to interpret this result was to say that since all patients seemed to benefit from the statin therapy, it might not be necessary to measure for baseline cholesterol levels in patients - instead, simply treat all high-risk patients with statin therapy. (People often joke that we should just "put statins in the drinking water.") Further support for this concept can be found in the data from the Cholesterol Treatment Trialists (CTT) meta-analysis, which inspected the results in over 90,000 patients enrolled in the large placebo-controlled trials of standard-dose statin therapy. This analysis found that risk of CAD death or myocardial infarction (MI) was reduced by 23% (P < .0001), and the risk of any major vascular event by 21% (P < .0001), with consistency of benefit in every subgroup examined.[9] Thus, the conclusion appears as simple as the results of HPS suggest: For all patients at risk (another subject, but primarily meaning blood pressure, overweight, smoking, diabetes, family history), whatever their baseline LDL level, statin therapy will lower their risk. However, this begs the question of whether the benefit observed is based on just receiving statin therapy, or whether the degree of lipid lowering mattered. In other words, is it a qualitative factor (the drug, ie, statins) or a quantitative factor (the LDL level achieved) that matters? How Low Have We Achieved So Far? Once HPS and the initial placebo-controlled trials had established the undoubted efficacy of statin therapy, it became unethical to assess cholesterol lowering against a placebo. Recently, however, 4 large outcomes trials involving almost 28,000 patients have assessed more intensive statin therapy vs "standard" statin doses as used in the earlier trials, to reach target LDL cholesterol levels of < 130 mg/dL in stable CAD patients and <100 mg/dL in patients with additional risk (eg, diabetes). PROVE IT-TIMI 22 The Pravastatin or Atorvastatin Evaluation and Infection Treatment - Thrombolysis in Myocardial Infarction (PROVE IT-TIMI 22) trial compared standard therapy with a proven agent (pravastatin 40 mg, which achieved a median LDL level of 95 mg/dL, thus meeting the current National Cholesterol Education Program [NCEP] III Guideline target of LDL < 100 mg/dL[7]) vs intensive therapy with a new agent (atorvastatin 80 mg, which achieved a median LDL level of 62 mg/dL) in patients hospitalized for an acute coronary syndrome (ACS) event.[10] This 33-mg/dL difference resulted in a very clear and highly significant (P = .005) benefit of the more intensive statin treatment, with a 16% reduction in the risk of death and major cardiovascular events that emerged rapidly, and was seen over the subsequent 2 years following an acute coronary syndrome.[10] The interesting and very important point was that this large and highly significant benefit was achieved on top of active therapy with a proven standard-dose statin (since a placebo control would have been unethical). Moreover, this benefit emerged as early as 15 days after randomization and achieved statistical significance by 30 days for the composite endpoint of death, MI, or recurrent ACS.[11] A second trial in ACS patients, A to Z, showed similar trend toward benefit, but had limited power due to a smaller-than-anticipated number of events.[12] Treat to New Targets (TNT) The Treat to New Targets (TNT) trial was similar to PROVE IT except that it enrolled patients with stable CAD and tested treatment with high-dose vs standard-dose atorvastatin. Again, in this trial there was a highly significant reduction (22%) in major cardiac events,[13] and the benefit was seen with reductions in cardiovascular death, MI, the need for revascularization, and stroke. IDEAL The Incremental Decrease in End Points through Aggressive Lipid Lowering (IDEAL) trial compared atorvastatin 80 mg with simvastatin 20-40 mg in patients with a prior MI and who qualified for statin therapy according to NCEP guidelines. The intensive statin treatment lowered LDL cholesterol by 23 mg/dL, and this resulted in an 11% reduction in the primary endpoint of CAD death, MI, or resuscitated cardiac arrest that just missed being statistically significant (P = .07).[14] However, when the results were reassessed post hoc, using the TNT endpoint (which also included stroke), there was a significant 13% reduction in the primary endpoint (P = .02), and when using the primary endpoint from the PROVE IT-TIMI 22 trial (any cardiovascular event including revascularization), there was a highly significant 16% reduction (P < .001).[14] REVERSAL Additional supportive data for the benefit of intensive statin therapy came from the Reversal of Atherosclerosis with Aggressive Lipid Lowering (REVERSAL) trial.[15] In this study, 654 patients who were referred for coronary angiography were randomized to 18 months with intensive (atorvastatin 80 mg) vs moderate (pravastatin 40 mg) statin therapy. The primary endpoint was change in the actual volume of atheroma in designated segments of the coronary arteries, as measured using an intravascular ultrasound (IVUS) technique. In this study, the LDL levels achieved by the 2 treatments were quite different, 110 mg/dL and 79 mg/dL, for the standard and intensive groups, respectively. A significantly lower rate of progression of coronary artery disease was seen in atheroma volume in the intensive therapy arm (-0.4% vs +2.7%, P = .02). These anatomic data support the clinical events data from the 4 clinical outcomes trials in showing that lowering LDL into the range of 70 mg/dL appears better than just lowering LDL below 100 mg/dL. Side Effects With regard to safety, 2 issues have been raised. The first is the issue of (transient) elevations of liver and muscle enzymes. Increases in liver function tests and creatine kinase levels of 1% to 3% are seen in patients on standard- or high-dose statin therapy, compared with elevations of 1% to 2% in the patients treated with placebo.[10,13,16] Once patients began being treated with the more recent high-dose statin regimens, these liver enzyme function elevations were slightly more common. In the PROVE IT-TIMI 22 trial, there were more patients in the intensive group who had transient increases in their liver function tests (alanine aminotransferase > 3 times the upper limit of normal), appearing in approximately 3% of patients receiving atorvastatin 80 mg vs 1% of patients on pravastatin 40 mg (P < .001). However, CK elevations, myalgias, and cessation of the study medication due to report of muscle aches or CK elevations were similar in the 2 groups (approximately 3%). There were no cases of rhabdomyolysis in either group of the trial. Overall, rhabdomyolysis is very rare; in the recent CTT meta-analysis, there were just 9 cases (0.023%) in 39,884 patients treated with statins, vs 6 cases (0.015%) found in 39,817 treated with placebo, which equates to a nonsignificant excess of just 3 patients (0.01%).[9] Looked at another way, this safety profile of statins is quite impressive, in that 90%-95% of patients can tolerate the statins, even at maximal doses, as tested in the recent trials. Nevertheless, although this class of drugs is quite safe, it does require both physicians and patients to monitor for these side effects and adjust the doses accordingly. How Low Is Still Safe? Is there a bottom? Is there an LDL level below which the decrease in risk of events starts to be outweighed by an increase in risk of deleterious effects? This can be separated into 2 questions: How low is still safe? How low is the most efficacious? Safety What about the safety of achieving ultra-low LDL levels? An analysis by Wiviott and colleagues[17] looked carefully at the results of liver function enzyme tests, creatinine kinase elevations, and other side effects among patients who achieved very low LDL levels (ie, < 40 mg/dL or in the range 40 to 60 mg/dL) and found no differences in any of these adverse safety events.[17] Intriguingly, as the achieved LDL reached this new ultra-low range, the rate of death or cardiovascular events continued to drop lower and lower, suggesting that a target LDL of < 70 mg/dL might not be the lowest target to achieve in order to get maximum benefit. However, the important take-home message from this safety analysis is that if a patient is on a high-dose statin, and his or her LDL comes back at 42 mg/dL, there is no need to reduce the dose. How Low Is Low Enough? So -- how low should we go? The current evidence supports, and the NCEP Guideline Committee suggests, that the "therapeutic option" should be to get LDL levels to < 70 mg/dL for high-risk patients (ie, those with a recent ACS or with established CAD or CAD equivalent). However, I aim for this level in practice with all of my patients, as I believe many cardiologists do also. But -- is this the "optimal" level? We don't know if we have reached the bottom yet. We are currently conducting the IMPROVE IT trial, which will compare 2 of the optimal current regimens, simvastatin vs Vytorin (simvastatin plus ezetimibe). The target will be to reach median achieved LDL levels of approximately 65 mg/dL vs 50 mg/dL, respectively. When we see the results of IMPROVE IT, we should see whether we need to set our LDL cholesterol targets even lower. For the moment, however, I try to get all of my patients with CAD (and/or peripheral arterial disease or prior stroke) to < 70 mg/dL. This seems to be the best evidence-based medicine at the moment. References