Effectiveness of Ezetimibe Monotherapy in Patients With Hypercholesterolemia

Although the death rates from coronary heart disease (CHD) have declined over the years, CHD is still the leading cause of mortality in the United States.^1,2 In 2004, CHD accounted for 869,724 deaths in the United States alone.¹ In addition, the economic burden of CHD is increasing, with total direct and indirect costs in the United States estimated at $448.5 billion for 2008.¹

Hyperlipidemia is recognized as a major modifi able risk factor for CHD, particularly when it occurs in association with other risk factors such as hypertension, smoking, and diabetes.^3,4 A strong correlation between elevated levels of low-density lipoprotein cholesterol (LDL-C) and increased risk of CHD has been well established.⁴ Clinical trials have shown that 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins) can signifi cantly lower LDL-C levels and reduce CHD-related morbidity and mortality.^5-9 These findings have led to the National Cholesterol Education Program (NCEP) Adult Treatment Panel III (ATP III) guidelines and the American Heart Association/American College of Cardiology (AHA/ACC) guideline update (2006) for aggressive cardiovascular risk management.^4,10 The NCEP ATP III guidelines placed patients with CHD into 5 risk categories, each with specifi c LDL-C treatment goals. The AHA/ACC 2006 guideline update states that the target LDLC goals for all patients with CHD and other clinical forms of atherosclerotic disease should be less than 100 mg/dL and that it is reasonable to treat such patients to a goal of less than 70 mg/dL (

Table 1

).^4,10 Patients with established CHD and other clinical forms of atherosclerotic disease are in the very high—risk category and those without atherosclerotic disease but with CHD risk equivalents are in the high-risk group. Patients without CHD risk equivalents are at moderate risk, and patients with zero to one (0-1) risk factor are in the lowest-risk group. CHD risk equivalents include diabetes and multiple risk factors that confer a 10-year risk for CHD that is greater than 20%. In addition to reducing LDL-C levels to less than 100 mg/dL, an LDL-C level less than 70 mg/dL is a reasonable therapeutic goal for all patients at very high risk of myocardial infarction or death.^4,10

Several studies have reported poor goal attainment in CHD populations treated in clinical practice^11,12; these patients are far less likely to achieve LDL-C targets than those treated in clinical trials.¹³ Frolkis et al compared expected LDL-C lowering based on the package inserts of atorvastatin, simvastatin, and pravastatin with the actual reductions achieved by hyperlipidemic patients in a preventive cardiology practice.¹⁴ The mean observed reduction in LDL-C of 26% (±20%) was significantly lower than the expected reduction of 34% (±7%).¹⁴

Treatment in clinical practice is rarely delivered as effectively or uniformly as in controlled efficacy trials. Factors such as population heterogeneity, patient adherence, and actual costs of treatment to the patient all contribute to the discrepancy seen between efficacy and effectiveness studies.¹⁵

Studies with statins have shown there is a lack of adherence to treatment by patients with CHD, and patient adherence worsens over time.^16-18 A significant portion of these patients are reluctant to take statins because of associated musculoskeletal symptoms, or as a result of reports in the media regarding myopathy.¹⁹ Myopathy and rhabdomyolysis have been rarely reported for all statins,^20,21 and fatal rhabdomyolysis has been reported for all statins except fluvastatin.²² Polypharmacy also can contribute to poor patient adherence, particularly in the elderly population, because of the higher number of drugs prescribed to elderly patients compared with the rest of the population.^23,24 Drug—drug interactions have limited the use of statins in a number of patients, including patients taking select fibrates, protease inhibitors, and anticoagulants such as warfarin.²¹ One specific example is the voluntary removal of cerivastatin from the market due to reports of fatal rhabdomyolysis—particularly in association with gemfibrozil.²⁵

Current alternatives for patients with hypercholesterolemia who are intolerant to statins include bile acid sequestrants, niacin, and esters from plant sterols. Although these medications provide some reduction in LDL-C levels, the major barrier to their use is lack of patient compliance due to a range of associated side effects.³ These data suggest the need for new interventions that are effective, safe, and well tolerated to increase the percentage of patients attaining LDL-C goals and, ultimately, to reduce the risk and burden of CHD.

Ezetimibe has a mechanism of action different from that of other currently available lipid-lowering medications in that it aims to block cholesterol absorption at the intestinal brush border to reduce LDL-C levels without affecting absorption of triglycerides, lipid-soluble vitamins, or most concurrently administered medications.^26,27

The efficacy of ezetimibe monotherapy was demonstrated in a pooled analysis of 2 phase III clinical trials with a total of 1179 patients with primary hypercholesterolemia. Ezetimibe monotherapy significantly reduced mean LDL-C by 18.2% (P <.01) and resulted in a statistically significant small increase in high-density lipoprotein cholesterol (HDL-C) (1%; P <.01) and a statistically significant reduction in triglycerides (−8.0%; P <.01). Ezetimibe was well tolerated, with a safety profile similar to that of placebo.²⁸

As outlined above, several studies have found that patients receiving various lipid-lowering therapies do not achieve levels of LDL-C reduction in clinical practice similar to those obtained in clinical trials. In this study, we examined LDL-C lowering by ezetimibe monotherapy in a managed care population to assess its effectiveness compared with the results obtained in clinical trials.

METHODS

A retrospective cohort study was conducted using administrative claims data from the 2004 to 2007 Ingenix Claims Database (i3 Innovus, Eden Prairie, MN). This database contains pharmacy and medical claims information for more than 20 million employed, commercially insured patients and their dependents, and laboratory results for approximately 15% of the population.

Study Population

Patients with newly filled prescriptions for ezetimibe were identified from July 1, 2004, through December 31, 2007. The period of study was from 6 months prior to the index date (the index date was the date of the first claim for ezetimibe) to 3 months following the index date. The length of time a patient was treated was variable between the index and postindex dates. A subcohort of patients with diabetes also was investigated. The study sample was selected from the population of insured health plan members age 18 years or older who had pharmacy benefits and were continuously enrolled throughout the study period.

All patients included in the study had no previous claim for lipid-lowering therapy for at least 6 months prior to the index date, at least 1 LDL-C measurement within 6 months prior to the index date, and at least 1 LDL-C measurement between 1 and 3 months after the index date, and they had no change in their treatment regimen between the index date and the first posttreatment LDL-C measurement (months 1-3). Patients were categorized by risk level according to the updated AHA/ACC 2006 guidelines, and only those in moderate-risk and high-risk categories were included. Patients identified as low risk for CHD (no CHD or CHD risk equivalents and 0-1 major CHD risk factors such as age, smoking, and family history of premature CHD) were excluded because of the potential misclassification of this risk category. As risk levels were assigned based on data provided by administration claims, a patient without CHD or CHD risk equivalents may be inadvertently classified as low risk because of the underreporting of certain risk factors such as smoking or family history of premature CHD. Patients who took medications that impact lipid levels or interact with ezetimibe at any time during the study period also were excluded. Patients could not have filled a prescription for any lipid-lowering therapy including statins, ezetimibe, bile acid sequestrants, fibrates, and prescription-strength niacin during the 6 months prior to the index date. Patients could not have filled a prescription for any lipid-lowering therapy besides ezetimibe during the time period from the index date to the first laboratory date.

Patients with liver disease identified using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnostic codes, and patients at or below target LDL-C at baseline also were excluded (

Figure 1

). Optional goals were used for very high—risk and moderately high–risk patients (Table 1).

Data Collection

Data on variables related to ezetimibe therapy, adherence, and goal attainment were obtained from the medical claims database; these included demographics, cardiovascular history, and serum lipid levels. Information on demographic variables including sex and age was obtained from enrollment files. Drug use and prescription fill dates were obtained from the pharmacy claims database using National Drug Codes. Information regarding patients’ risk for CHD or CHD equivalents were obtained from the medical and pharmacy claims database prior to the index date using the AHA/ACC guidelines, ICD-9-CM codes, Current Procedural Terminology codes, and prescription records.

Comorbidities were identified using the ICD-9-CM diagnostic codes during the 180-day time frame before the index date. Serum lipid level data for total cholesterol, LDL-C, HDL-C, and triglycerides were collected for 2 time periods: (1) baseline measurement taken within 6 months before the index date, and (2) postindex measurement taken 1 to 3 months after the index date. Attainment of the LDL-C goal was based on the 2006 AHA/ACC guidelines (Table 1).

Adherence was based on prescription refill activity, measured as the proportion of days covered by filled prescriptions for ezetimibe between the index date and the first postindex measurement.²⁹ For patients with multiple LDL-C laboratory values available within 1 to 3 months after the index date, the value from the first test following the index date was used for the adherence calculation.

RESULTS

Descriptive Analyses

A total of 562 patients were identified as new ezetimibe users who were at moderate to high risk for CHD and met all the study inclusion criteria; low-risk patients were excluded. From this population, a subcohort of 227 patients with diabetes also was identified (Figure 1).

In the overall population, the most common comorbidities included hypertension, diabetes, low HDL-C (<40 mg/dL), chronic ischemic heart disease, and benign neoplasm. Similar trends were also seen within the diabetic subcohort (

Table 2

).

Lipid Measurements

The Wilcoxon signed rank test was used to evaluate changes in LDL-C measurements. Results showed a significant decrease in LDL-C values (P <.001) from baseline to postindex measurements. This finding was similar in both the total population and the diabetic subcohort (Table 3 and

Table 4

).

The average percent LDL-C decline within 1 to 3 months of initiating ezetimibe monotherapy treatment was 18.5% ± 18.2% in the total population (Table 3). Measurements for the diabetic subcohort were similar, with an LDL-C decline of 17.4% ± 19.4% (Table 4). Analyses of the baseline and postindex LDL-C distributions show the decrease in LDL-C measurements for both the total population and the diabetic subcohort (

Figure 2

).

Based on the 2006 AHA/ACC guidelines, approximately 35% of the total population treated with ezetimibe monotherapy reached their target LDL-C goal within 1 to 3 months of initiating therapy. Approximately 30% of the diabetic subcohort also reached their target LDL-C goal within 1 to 3 months on ezetimibe monotherapy (Table 4).

Adherence

As seen in

Figure 3

, patients in the total population whose adherence rate (percent of days covered) was 75% or greater had statistically significant LDL-C reductions (P<.0001), with an average of 20.0% ± 16.3% decline from the baseline LDL-C measurement. Patients whose adherence was 75% or greater had higher percentage reductions in LDL-C than patients whose adherence was less than 50% (P <.0001).

DISCUSSION AND CONCLUSIONS

This study demonstrates that ezetimibe is an effective LDL-C—lowering agent in patients with moderate- to very high–risk hypercholesterolemia, including diabetic patients in a managed care setting.

Ezetimibe was associated with a mean LDL-C percent decrease from baseline to endpoint of −18.5% ± 18.2% in the overall population and −17.4% ± 19.4% in the diabetic subcohort. This decline in LDL-C is similar to rates previously reported in clinical trials (−18%).28 In addition, within 1 to 3 months of initiation of ezetimibe therapy, a reduction in triglycerides and a slight increase in HDL-C were observed in both the total population and diabetic subcohort (Tables 3 and 4). Increased levels of HDL-C and decreased levels of triglycerides have been reported to lower the risk of CHD.³⁰

The effectiveness of ezetimibe, as with other CHD medications, depends on patient adherence.³¹ In the present study, overall patient adherence (based on prescription fill activity) was 85% in the total population and 83% in the diabetic subcohort (Tables 3 and 4). Patients with poor adherence rates had more limited LDL-C percent declines (Figure 3). Parris et al observed a similar relationship between adherence and LDL-C reductions in a study examining the relationship of adherence to statins and LDL-C goal attainment.³² Of the 44% of patients in the Parris et al study who achieved their LDL-C goal, the mean adherence rate was 82% compared with a mean of 61% for the patients who did not attain their goal during the 9-month study.³²

Because this study was a retrospective analysis of an administrative claims database, several limitations should be taken into consideration. First, risk categories were assigned based on data provided by administrative claims. As such, patients without CHD or CHD risk equivalents may have been inadvertently assigned to a risk level lower than their true medical condition due to underreporting of certain risk factors such as smoking or family history of premature CHD. That would have had no impact on the observed LDL-C change and percent LDL-C change, but would have resulted in a lower assigned LDL-C goal and a higher percentage of patients achieving that goal. To avoid this potential skewing of the analysis and confounding of the interpretation of the results, the analysis was limited to patients in the very high—risk, high-risk, and moderate-risk categories.

Additional limitations inherent to the current study include the following: (1) data regarding medication samples used by patients were not available in retrospective claims; (2) adherence was measured based on prescription fill activity rather than patient reporting; and (3) the sample size was limited, partially due to availability of LDL-C laboratory data.

Despite these limitations, this study is able to offer insights into the effectiveness of ezetimibe monotherapy in reducing LDL-C levels in a managed care setting, and further emphasizes the importance of patient adherence for improved health-related outcomes. Although lowering LDL-C has a well-established relationship with reducing the risk of cardiovascular events, studies demonstrating the efficacy of ezetimibe with respect to clinical outcomes have not been completed, and the analysis of clinical outcomes was beyond the scope of the current study. The clinical benefit of ezetimibe is currently being evaluated in the IMPROVE-IT trial, where cardiovascular outcomes with simvastatin or simvastatin plus ezetimibe are being evaluated in patients with acute coronary syndrome.³³ IMPROVE-IT is expected to include approximately 18,000 patients and has an estimated completion

date of June 2013.³³

In conclusion, ezetimibe monotherapy may offer an important alternative for patients with hypercholesterolemia, including those with diabetes. It may be particularly useful for those who are intolerant to statins and those in whom statins are contraindicated.