Pharmacologic Lipid-Lowering Therapy in Type 2 Diabetes Mellitus: Background Paper for the American College of PhysiciansFREE

The literature review was limited to randomized, controlled trials of drug therapy that included patients with diabetes. Only studies that measured major clinical end points were included. Major clinical end points were defined as major cardiovascular events (for example, cardiovascular mortality, myocardial infarction, stroke), cardiovascular mortality, and total mortality. Of note, many of the trials reported somewhat different clinical end points in the patients with diabetes. All included cardiovascular mortality and myocardial infarction in their composite end point; some included stroke and revascularization, and one included unstable angina. We used the primary reported data directly from the published study in our review. We also subdivided the literature review into 2 categories. The first category evaluated the effects of lipid management in primary prevention (that is, patients without known cardiovascular disease); the second evaluated the effects in secondary prevention.

We used several sources to identify the relevant literature. We started with a search of the Cochrane Library. We then performed a MEDLINE search in September 2002. We used the keywords exp diabetes mellitus and exp lipids [therapy or prevention and control] and limited the search to randomized, controlled trials and human studies. The final search produced 919 results. Of these, most were discarded because they did not measure major clinical end points, did not report outcomes for patients with diabetes, were observational in nature, or were reviews or editorials. We then updated the search through consultation with experts and through references from the identified articles, meta-analyses, and review articles.

The primary author extracted data from the primary study reports. Accuracy and quality of the abstraction were confirmed through reabstraction and comparison with the original abstraction. The outcomes were broken into categories as described earlier, and data on absolute and relative risk reduction and numbers needed to treat for benefit were derived from the primary reports or were calculated in standard fashion (11).

The results of the studies were then combined by using meta-analytic techniques. We pooled data for both relative and absolute risks. A Mantel–Haenszel test was done to test for heterogeneity. In the analyses of secondary prevention, the data had substantial heterogeneity, so the pooled risk ratios and differences were calculated by using the DerSimonian and Laird method with a random-effects model. Sensitivity analyses were done by excluding studies that appeared to be outliers to ascertain the source of the heterogeneity. All analyses were done by using the statistical package Stata (Stata Corp., College Station, Texas).

There were no sources of direct funding for this manuscript. Dr. Vijan was a Veterans Affairs Career Development Awardee during preparation, and the American College of Physicians provided an honorarium to the authors.

No studies of lipid-lowering therapy that reported cardiovascular outcomes were conducted solely in patients with diabetes, but several studies reported diabetes subgroup analyses. Sample sizes of participants with diabetes were often small, so most of the studies presented data only on combined cardiovascular events (for example, cardiovascular mortality, nonfatal myocardial infarction, stroke, and coronary revascularization). In several cases, diabetes was an exclusion criterion or very few patients with diabetes were included; for example, in the West of Scotland Coronary Prevention Study, about 1% of patients (76 of 6595) had diabetes, and results were not presented for this subgroup (12). We found a total of 12 lipid-lowering studies that presented diabetes-specific data and reported clinical outcomes. Of these studies, 4 were focused on primary prevention, 6 were focused on secondary prevention, and 2 presented data on both. The general study characteristics, including the effect of interventions of lipid levels, are presented in Table 1.

The principal complication of type 2 diabetes is cardiovascular disease; up to 80% of patients with diabetes will develop or die of some type of major vascular event (27-35). Therefore, the foremost goal of therapy in type 2 diabetes should be preventing cardiovascular disease through optimization of risk factor modification. This includes aggressive treatment of hypertension (7, 36); smoking cessation; aspirin therapy; and, as our review shows, aggressive use of lipid-lowering therapy, particularly with 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins).

Current medical evidence suggests that lipid-lowering medications lead to about a 22% to 24% reduction in major cardiovascular events in patients with diabetes. Because of small diabetes subgroups in most trials, not all individual trials found benefit. However, the meta-analyses clearly suggest a benefit from using a moderate dose of a statin. There is also evidence of benefit from gemfibrozil, at least among those with low HDL cholesterol levels (<1.03 mmol/L [<40 mg/dL]). Although the relative risk reductions were similar for both primary and secondary prevention, the average absolute risk reduction was more than twice as high for those with known coronary artery disease (secondary prevention) than for those without it (primary prevention). This is a reflection of the fact that the secondary prevention studies universally reported higher risks for cardiovascular outcomes, on average. Indeed, the 4.3-year risk in the control group ranged from 3.6% to 18.6% in the primary prevention studies, but the 4.9-year risk ranged from 22.8% to 45.4% in the secondary prevention studies. Of note, even among the secondary prevention trials, the absolute risk reduction was largest in the 3 trials with the highest-risk participants: 4S, which included patients with by far the highest baseline LDL cholesterol levels, and LIPS and POST-CABG, which were conducted in the highest-risk patients (those who had procedures for coronary revascularization). In contrast, only one of the primary prevention studies showed statistically significant benefit in the patients with diabetes, and the observed benefits were quite small or absent in studies in which patients with diabetes had low baseline risk (18, 19). Therefore, we recommend caution in extrapolating the average results found in our primary prevention meta-analysis to patients at lower than average risk (such as younger patients with diabetes who have no other major cardiovascular risk factors). Unfortunately, none of the clinical trials conducted subgroup analyses using a multivariable risk prediction tool, a practice that needs to become more common in clinical trial reporting (37, 38).

Given the absolute risk reductions observed in these studies, treatment will probably be cost-effective under most circumstances. However, none of the trials has published a direct analysis of cost-effectiveness from a diabetes subgroup. At least 2 secondary analyses of cost-effectiveness in patients with diabetes have been published, one by the Centers for Disease Control and Prevention using a simulation model based on Framingham data and the United Kingdom Prospective Diabetes Study (8) and another by Grover and colleagues using data from a Canadian cohort (39). The analyses found that lipid-lowering therapy was reasonably cost-effective compared with commonly adopted medical interventions. Overall, the range of cost-effectiveness in the Grover model was from $4000 to $40000 per year of life saved, depending on various patient characteristics, and secondary prevention was more cost-effective than primary prevention (39). The distinction between primary and secondary prevention was not made in the Centers for Disease Control and Prevention model but instead applied to treatment from diabetes onset; the estimate of cost-effectiveness was about $52000 per quality-adjusted life-year gained, with substantial variation by age (treatment was most cost-effective in those between 55 and 74 years of age). In comparison, analysis of the cost-effectiveness of hypertension treatment in diabetes was found to be cost-saving in most instances (8).

Although the average benefits of statin therapy for patients with diabetes are fairly clear, there are important issues that cannot be easily answered with the available data. The appropriate target for LDL cholesterol levels remains, at best, poorly defined. While the National Cholesterol Education Panel (NCEP) guidelines state that patients with diabetes should be treated to a target LDL cholesterol level of 2.59 mmol/L (100 mg/dL) and that drug therapy should be started if LDL cholesterol levels exceed 3.36 mmol/L (>130 mg/dL) (40), currently available clinical trial data do not firmly support this specific approach. Current trials either did not set specific LDL cholesterol target levels or used different targets than those commonly proposed (Table 2). While the achieved LDL cholesterol levels in trials are consistently under 3.1 mmol/L (<120 mg/dL), different studies have found differing benefit for lower target levels. For example, the CARE study found that if the baseline LDL cholesterol level was less than 3.23 mmol/L (<125 mg/dL), then there was no benefit to lipid-lowering therapy, at least in the general population (21). In contrast, no such thresholds were identified in the LIPID study or in the HPS (17, 22). Indeed, the HPS, which is the largest of the studies and included the most patients with diabetes, showed that there was a consistent, approximately 25% relative risk reduction and a 5% to 7% absolute risk reduction in cardiovascular event rates regardless of starting LDL cholesterol levels, even among those whose starting LDL cholesterol levels were at or near the NCEP's stated target of 2.59 mmol/L (100 mg/dL). This suggests that empirical use of statins for diabetic persons with average or above average cardiovascular risk is much more important than the baseline or target LDL cholesterol level. It could be argued that there is no strict definition of hyperlipidemia in patients with type 2 diabetes, since nearly the entire population qualifies for lipid-lowering treatment.

Although treatment goals remain somewhat poorly defined by randomized trials, an argument could be made for an LDL cholesterol level goal as low as 1.81 to 1.94 mmol/L (70 to 75 mg/dL), based on cohort analyses. However, such conclusions would be speculative at this time because most trials lack specific target lipid levels and cohort studies largely examine naturally occurring LDL cholesterol levels rather than treatment-induced levels. Even if this speculation is found to be accurate, the curvilinear relationship between LDL cholesterol levels and cardiac risk (41) would suggest that the incremental absolute benefit of such a strategy would be relatively low and that the number needed to treat for benefit and the cost-effectiveness of treatment are likely to be much less favorable, especially in those with average cardiovascular risk (42). Furthermore, it is not clear that target LDL cholesterol levels of 1.81 to 1.94 mmol/L (70 to 75 mg/dL) are commonly achievable in practice. Indeed, the studies reviewed here did not achieve LDL cholesterol levels below 2.3 mmol/L (<89 mg/dL), even the few that had statin dose titration or used multiple agents (Table 1).

Even setting an LDL cholesterol target of less than 2.59 mmol/L (<100 mg/dL) rather than simply recommending moderate doses of statins for most or all patients with type 2 diabetes is difficult to justify from the literature. Not only is clinical trial evidence lacking to support titration of therapy to reach this goal, but statins have potential non–lipid-related effects. For example, statins may modulate cardiovascular risk by reducing inflammation, by stabilizing existing plaque, and by improving endothelial function (43-46). One could easily speculate that if statins do have a lipid-independent effect on cardiovascular risk, then wide-scale use of at least moderate doses of statins may be more beneficial than dose titration based on LDL cholesterol levels. However, the importance of these lipid-independent benefits has not been established in clinical trials, and whether other markers (such as C-reactive protein or nitrotyrosine levels) may be better indicators for use and titration of statins remains unclear (47, 48).

Perhaps the best conclusion to be drawn from the currently available data (particularly that from the HPS) is that most patients with diabetes should be taking at least moderate dosages of statins (for example, simvastatin, 40 mg/d; pravastatin, 40 mg/d; lovastatin, 40 mg/d; atorvastatin, 20 mg/d; or an equivalent dose of another statin). Given the absence of clinical trial data on the subject, the decision of whether to further increase statin doses or to use combination lipid-lowering treatment titrated on the basis of LDL cholesterol levels should be left to individual clinicians and patients. Our appraisal of the literature suggests that one could equally advocate for empirical maximization of statin dose as tolerated.

For the reasons mentioned previously, we also do not feel that the evidence is sufficient to make strong recommendations for primary prevention therapy for people with diabetes who have relatively low cardiovascular risk. While overall risk is elevated in patients with diabetes, the relatively low event rates in ASCOT-LLA show that some patients with diabetes are at low risk. This trial had an event rate of only 3.6% over 3.3 years in the control group of patients with diabetes, which is below the 2% per year guideline used to recommend treatment in the NCEP (19). Finally, we strongly advocate reanalysis of the lipid-lowering trials reviewed here using multivariable risk-stratification prediction tools. With such analyses, existing data could shed considerable light on the issue of relative and absolute benefits for higher-risk versus lower-risk patients and could allow more accurate individual tailoring of therapy (37, 38).

In addition, the current literature suggests that statins are extremely safe. While discontinuation and nonadherence rates in clinical trials are reasonably high (≥ 15% in many cases), rates of discontinuation typically are not different from those of placebo. Rates of elevated liver or muscle enzyme levels did not differ between statin and placebo groups in recent large-scale studies. For example, in the HPS, rates of elevated alanine aminotransferase levels above twice the upper limit of normal were 1.8% in the simvastatin group and 1.6% in the placebo group, and rates of elevated creatinine kinase levels were 0.3% in the simvastatin group and 0.2% in the placebo group (17). Neither of these differences were statistically significant. Similarly, among the 5804 patients in PROSPER, only 1 in each group had an alanine aminotransferase or aspartate aminotransferase level more than 3 times the upper limit of normal (18). In addition, there were no cases of rhabdomyolysis and 36 cases of myalgia in the pravastatin group compared with 32 in the placebo group (17, 18, 49, 50). Considering the safety of these drugs, routine monitoring of liver or muscle enzymes is probably not warranted unless patients have symptoms, have liver enzyme abnormalities at baseline, or are taking drugs that interact with the statins to increase the risk for adverse events. This simplifies and reduces the cost of treatment and would be similar, for example, to simply prescribing a daily aspirin for a patient with diabetes (51).

Another possible exception to the “statins for all” rule in patients with diabetes is in those with low HDL and LDL cholesterol levels. These patients may benefit more from gemfibrozil than from a statin, although there are no head-to-head comparisons of the drugs (25). However, unlike therapy with statins, the benefits of gemfibrozil therapy may be attenuated in patients with very low baseline LDL cholesterol levels of around 2.59 mmol/L (100 mg/dL). In patients with LDL cholesterol levels above 2.59 mmol/L (>100 mg/dL) and low HDL cholesterol levels, combination therapy with both a statin and gemfibrozil could be considered. However, there are as yet no efficacy data on combination therapy, and the risks of treatment, while modest, are elevated compared with those for single-agent therapy (52). Some of these questions will be better addressed by future studies that are comparing dose–response effects and the effect of combinations of lipid-lowering therapy on clinical outcomes.

Given the markedly elevated risk for cardiovascular events in people with type 2 diabetes, aggressive management of lipids provides substantial benefit, at least to the average patient. The use of statins should be nearly universal in this population. The current literature offers stronger support for empirical use of at least moderate doses of statins than it does for targeting specific LDL cholesterol levels. An argument can be made for using gemfibrozil as first-line therapy for patients with low HDL cholesterol levels and moderately low LDL cholesterol levels. Future studies should evaluate the relative effectiveness of specific strategies, such as different LDL cholesterol targets versus different doses of empirical statin therapy and combination therapy, and should also consider the potential effects of statins beyond lipid lowering.

Note added in proof: As this review went to press, a study of intensive lipid lowering in patients with acute coronary syndromes demonstrated the superiority of 80 mg of atorvastatin over 40 mg of pravastatin (Cannon CP, Braunwald E, McCabe CH, Rader DJ, Rouleau JL, Belder R, et al. Comparison of intensive and moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med. 2004; Mar 8 [Epub ahead of print] [PMID: 15007110]). This study evaluated secondary prevention in a highly selected population and has limited statistical power for analyses of the diabetes subgroup. The implications of the study for most patients with diabetes is unclear; however, more aggressive lipid-lowering therapy should be considered for patients admitted with acute coronary syndromes.