Achieving optimal lipid values in patients with dyslipidemia is associated with reduced risk of cardiovascular events

Journal of Clinical Lipidology (2008) 2, 343–353

Original Articles

Achieving optimal lipid values in patients with dyslipidemia is associated with reduced risk of cardiovascular events Scott L. Charland, PharmD, Mark J. Cziraky, PharmD,* Ralph Quimbo, MA, Richard H. Karas, MD, PhD, William Insull, Jr., MD, Michael Davidson, MD, Eric J. Stanek, PharmD University of Colorado, School of Pharmacy, Denver, CO, USA (Dr. Charland); HealthCore, Inc., 800 Delaware Avenue, Fifth Floor, Wilmington, DE 19801, USA (Dr. Cziraky, Mr. Quimbo); Tufts University School of Medicine, Boston, MA, USA (Dr. Karas); Baylor College of Medicine, Houston, TX, USA (Dr. Insull); Rush University Medical Center, Chicago, IL, USA (Dr. Davidson); and Independent Healthcare Research Consultant, Thorofare, NJ KEYWORDS: Cardiovascular diseases; Cardiovascular events; Cholesterol; Clinical practice; Dyslipidemia; Lipids; Lipoproteins; Managed care; Population study

BACKGROUND: Cardiovascular (CV) event risk is significantly lower in patients with combined low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TG) at desired levels versus those without lower levels. However, this has not been investigated relative to specific patterns of baseline lipid abnormalities. OBJECTIVE: To evaluate the association between desired combined lipid value achievement and risk of CV events in patients with different baseline lipid profiles. METHODS: A retrospective managed care database analysis among treatment-naïve adults with elevated CV event risk, ⱖ12 months follow-up, and full lipid panel from January 1, 2001 to December 31, 2001 plus ⱖ1 panel before a CV event or study end. Patients were stratified into three baseline cohorts: isolated high LDL-C (Cohort 1), high LDL-C ⫹ low HDL-C or high TG (Cohort 2), and high LDL-C, low HDL-C, and high TG (Cohort 3). CV event risk stratified by combined desired lipid value achievement was assessed in each cohort. RESULTS: Achievement of combined desired lipid values/median days to achievement was 29% in 385 days (Cohort 1), 11% in 413 days (Cohort 2), and 7% in 505 days (Cohort 3). Achievement of combined desired lipid values was associated with an adjusted 25%– 46% lower CV event risk in Cohort 1 (hazards ratio, 0.75; 95% confidence interval 0.65– 0.87), Cohort 2 (hazards ratio, 0.54; 95% confidence interval 0.43– 0.67), and Cohort 3 (hazard ratio, 0.54; 95% confidence interval 0.37– 0.78). CONCLUSION: Patients with combined desired lipid values had lower risk of CV events versus those without such values. The risk reduction was greatest among patients with multiple lipid abnormalities, suggesting a potential benefit of interventions targeting low HDL-C and/or high TG in addition to high LDL-C. © 2008 National Lipid Association. All rights reserved.

* Corresponding author. E-mail address: [email protected] Submitted March 4, 2008. Revised May 1, 2008. Accepted for publication June 18, 2008.

Cardiovascular (CV) disease affects more than 80 million adult Americans and accounts for one of every 2.8 deaths.1 Elevated levels of low-density lipoprotein cholesterol (LDL-C), high triglycerides (TG), and low high-density lipoprotein cholesterol (HDL-C) values are major inde-

1933-2874/$ -see front matter © 2008 National Lipid Association. All rights reserved. doi:10.1016/j.jacl.2008.06.009

344 pendent risk factors for CV disease morbidity and mortality that are frequently encountered in the clinical setting.1,2 Evidence-based guidelines for the modification of lipid risk factors provide definitions for therapeutic goal values for LDL-C and suggest targets for HDL-C and TG in different subgroups of patients.3–5 These guidelines primarily focus on lowering of LDL-C using statins. More recent guidelines suggest even more aggressive LDL-C desirable levels in very high-risk patients.3– 6 However, although aggressive LDL-C lowering is an important component of current approaches to CV risk reduction, recent evidence suggests that this strategy yields small and incremental benefit in terms of reduction in CV events. As such, there is increasing focus on the need to additionally increase HDL-C or lower TG as potential targets of therapy.7–12 Although definitive studies have not been completed regarding further risk reduction by using these as secondary targets, current evidence-based guidelines suggest use of niacin or fibrate therapy when HDL-C is low or non–HDL-C is elevated in high-risk individuals.3–5 Previous research in diverse populations with elevated CV risk has demonstrated that multiple lipid abnormalities, characterized by combinations of undesired or nonoptimal values for LDL-C, HDL-C, non–HDL-C, and TG, are frequent. Further, simultaneous or combined achievement of desired or optimal levels for these lipid fractions is uncommon in routine clinical practice13–16 in the United States. This observed high prevalence of multiple lipid abnormalities and low rates of combined lipid value achievement within desirable levels appears in part to be related to underutilization of guideline-recommended pharmacotherapy, in particular therapy targeted to HDL-C and TG abnormalities.13–16 The potential for this is noteworthy, as the population-based risk of CV events over time appears to be significantly lower in patients who have desired levels of LDL-C, HDL-C, and TG compared to patients who do not.17 Although this prior research provides important information about the relationship between desirable lipid values and CV outcomes on a broad population basis, it does not provide detail into the same on the basis of the pattern of dyslipidemia at initial presentation. The primary purpose of the present investigation is to determine the association between achievement of desired, or optimal, combined lipid values and risk of CV events in groups of untreated patients with distinct patterns of abnormalities in LDL-C, HDL-C, and TG values at baseline.

Methods Study design The study was a longitudinal retrospective cohort analysis based on healthcare claims from a large southeastern United States managed care organization. The organization’s administrative medical and pharmacy claims records provided individual patient data on demographics, clinical

Journal of Clinical Lipidology, Vol 2, No 5, October 2008 diagnosis, serial plasma lipid levels, prescription drug treatment, and incident clinical events due to CV disease over five and a half years.

Data source This longitudinal retrospective cohort analysis was conducted within the HealthCore Integrated Research Database (HIRD) using administrative claims and laboratory data from January 1, 1999 to June 30, 2004 from 2.1 million Medicare and non-Medicare eligible members. The fully integrated dataset included date-stamped, linked medical, pharmacy, and laboratory encounters complete with laboratory results and eligibility files. Patient identity was masked throughout in a limited data set format, in accordance with the Health Insurance Portability and Accountability Act (HIPAA) of 1996.

Patient selection The patient selection scheme is shown in Figure 1. Adult patients (ⱖ18 years of age) with results from at least one full lipid panel present in a 2-year study intake period between January 1, 2000 and December 31, 2001 were included. A full lipid panel was defined as the presence of LDL-C, HDL-C, TG, and total cholesterol results for a patient on the same calendar day, with that date serving as the baseline laboratory date. Patients were required to have a minimum continuous data stream of 12 months pre- and post-baseline laboratory data. To ensure that patients were naïve to lipid pharmacotherapy at baseline, only those without a National Drug Code (NDC) for lipid-altering therapy in the 6-month period prior to the baseline laboratory date were included for the study.

Patient risk stratification Patients with definable CV risk assessed at the time period preceding and including the baseline lipid panel date were included in the analysis, and were categorized into two risk groups: elevated risk primary prevention (ERPP) or CHD/CHD Risk Equivalent (CHD/CHD-RE) (Table 1). Categorization was based on age, gender, baseline lipid values, previous and concurrent diagnoses and procedures, and medication usage. Patients whose risk status could not be clearly identified as ERPP or CHD/CHD-RE were not included for analysis. Achievement of optimal lipid values (Table 2) was determined by risk level as per evidencebased guidelines for LDL-C and suggested acceptable thresholds for HDL-C and triglycerides.2–5

Study cohort definition and selection The remaining sample was stratified into three mutually exclusive cohorts by baseline lipid laboratory values compared to definitions of optimal values appropriate for each

Charland et al

Cardiovascular events and optimal lipid value achievement

345

7,901

Figure 1

Study cohort selection and stratification schematic.

patient’s individual CV disease risk status (ERPP or CHD/ CHD-RE). These optimal values were derived from the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III),2,4 American Heart Association (AHA),18 and American Diabetes Association (ADA)19 guidelines for target LDL-C goals and desirable levels for HDL-C and TG (Table 2). Cohort 1 included patients with an elevated LDL-C level at baseline, but whose baseline triglyceride and high-density lipoprotein cholesterol levels were within acceptable range (ie, isolated nonoptimal LDL-C). Cohort 2 consisted of patients who had elevated LDL-C levels coupled with either high TG levels or low HDL-C levels at baseline. Cohort 3 included those patients for whom all three levels of LDL-C, HDL-C, and TG, were not optimal, as defined in Table 2. Patients with at least one full lipid panel available during the follow-up period were included initially to evaluate optimal lipid value achievement, and only those patients with a follow-up lipid panel present before a censoring event were included in the outcomes analysis dataset.

Study duration Data were gathered for a minimum of 12 months prior and after the initial baseline full lipid panel. Each patient was followed until either a CV event occurred or until the end of the individual data stream in cases of patients without a CV event.

Drug treatment of dyslipidemia A variety of dyslipidemia drugs were available for use at the discretion of the prescribing physicians. These included statins, bile acid sequestrants, cholesterol absorption inhibitors, fibrates, niacin, and both fixed dose and multiple pill combination drugs. Pharmacotherapy assessment was based on the first dyslipidemia therapy initiated during follow-up identified from prescription claims. Data were not available for concurrent diet intervention instruction and compliance.

346 Table 1 status

Journal of Clinical Lipidology, Vol 2, No 5, October 2008 Criteria for inclusion into elevated risk primary prevention or coronary heart disease (CHD)/CHD risk equivalent (RE)

Risk status category

Criteria

Elevated risk primary prevention

Patients without a prior vascular event or procedure with any two of the following risk factors: ● Males ⱖ45 y; Females ⱖ55 y ● HDL-C level at baseline*: Males ⬍40 mg/dL; Females ⬍50 mg/dL ● Presence of hypertension: defined as ICD-9 codes 401.xx and currently receiving ⱖ1 antihypertensive medications† Patients with any one of the following by ICD-9 code or medication use: ● Diagnoses of coronary artery disease, cerebrovascular disease (stroke or TIA), or peripheral arterial disease ● Any arterial revascularization; coronary, cerebrovascular, or peripheral ● Presence of diabetes mellitus and/or currently receiving ⱖ1 antidiabetic medications§

CHD or CHD risk equivalent‡ (CHD/CHD-RE)

HDL-C, high-density lipoprotein cholesterol; ICD-9, International Classification of Diseases, 9th Revision, Clinical Modification; LDL-C, low-density lipoprotein cholesterol; TG, triglycerides; TIA, transient ischemic attack. *This is a modified NCEP ATP III risk categorization as optimal HDL-C for women has been set at 50 mg/dL.18 †Calcium channel antagonist, ␤ blocker, angiotensin receptor blocker, angiotensin-converting enzyme inhibitor, diuretic, ␣ blocker. ‡Risk equivalents include diagnosis of CHD and/or cerebrovascular disease (including stroke and TIA), aortic aneurysm (441.xx) and/or peripheral arterial disease (PAD) (ICD-9 codes 410.xx– 414.xx, 433.xx– 436.xx, 438.xx, 440.xx, 441.xx, 443.xx), and/or diabetes mellitus (ICD-9 code 250.xx). §Insulin, biguanide, sulfonylurea, thiazolidinedione, meglitinide, ␣-glucosidase inhibitor, syringes, test strips, glucose monitors.

Outcome measures The primary outcome measure was the association between the simultaneous achievement of combined optimal values for LDL-C, HDL-C, and TG and the unadjusted and adjusted risk of CV events within each study cohort over the course of follow-up. Cardiovascular events were identified as any emergency or inpatient visit for ischemic heart disease (IHD), peripheral arterial disease (PAD), and stroke or transient ischemic attack (TIA) as defined by the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9), Current Procedural Terminology Version 4.2 (CPT4) 2004, and billing codes. Secondary outcome measures included the proportion of patients with optimal lipid value attainment of individual and combined parameters at baseline and annually for three years of follow-up, and evaluation of treatment patterns. Treatment pattern assessments included proportion of pa-

tients receiving lipid-altering therapy, mean and median values for times to first lipid-altering prescription, as well as mean and median values for time to achieve optimal lipid values.

Co-morbidity assessment Baseline burden of co-morbidity was assessed using the Deyo-Charlson co-morbidity score,20 calculated using medical claims in the 12-month pre-baseline period. History of hypertension, CHD, PAD, stroke or TIA, diabetes mellitus, and metabolic syndrome were obtained using ICD-9, CPT-4, or pharmacy records from the 12 months prior to the baseline lipid panel date. Metabolic syndrome was defined as the presence of a medical diagnosis of metabolic syndrome (ICD-9 code 277.7) or any three of the following conditions: diagnosis codes for obesity (ICD-9 code 178.xx), TG value ⱖ150 mg/dL, hypertension, fasting

Table 2 Lipid levels defined as optimal or goal based on National Cholesterol Education Program (NCEP) Adult Treatment Panel (ATP) III,2 American Heart Association (AHA),18 and American Diabetes Association (ADA) criteria3 Nondiabetics

Diabetics

Lipid fraction*

NCEP ATP III optimal value (men)

AHA optimal value (women) (mg/dL)

ADA cptimal value (mg/dL)

Description

LDL-C (mg/dL) LDL-C (mg/dL) HDL-C (mg/dL) HDL-C (mg/dL) TG (mg/dL)

⬍130 ⬍100 ⬎40 N/A ⬍200

⬍130 ⬍100 N/A ⬎50 ⬍150

N/A ⬍100 ⬎40 ⬎50 ⬍150

“Elevated risk” primary prevention CHD and CHD RE (CHD/CHD-RE) Below optimal considered low HDL-C in males Below threshold considered low HDL-C in females Above threshold considered high TG

CHD, coronary heart disease; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; N/A, not applicable; RE, risk equivalent; TG, triglycerides. *Evidence-based guidelines recommend an optimal goal for LDL-C and desirable levels for HDL-C and TG.

Charland et al

Cardiovascular events and optimal lipid value achievement

blood glucose ⱖ110 mg/dL, and/or HDL-C ⬍40 mg/dL (men) and ⬍50 mg/dL (women).

Statistical analysis Sample characteristics and demographic data are separately presented for each cohort of patients. Univariate analyses of frequencies, medians, and means with standard deviations (SDs) were performed to describe the study population. Statistical differences between cohorts were assessed using Pearson’s ␹2 tests for categorical variables and analysis of variance (ANOVA) F-test for continuous variables. Unadjusted incidence of CV events and corresponding Kaplan-Meier curves (rates estimated over 5 years of follow-up) were derived. Multivariate Cox proportional-hazards regression modeling for each cohort was used to assess the adjusted risk of a CV event (expressed as a hazard ratio [HR] with 95% confidence interval [CI]) for patients who achieved combined optimal values for LDL-C, HDL-C, and TG compared to patients who did not. Attainment of optimal lipid values was evaluated as of the last available lipid value prior to either a CV event or end of follow-up, and was assumed to remain static for the remainder of the study.

Table 3

347

All patients were observed until a censoring event, which was defined as either the date of their first CV event or the end of the data stream for patients without an event. Covariates in the multivariate Cox models included combined optimal lipid value attainment (as a binary variable), age, gender, use of follow-up lipid-altering pharmacotherapy, baseline Deyo-Charlson co-morbidity index, and presence of pre-baseline vascular disease (myocardial infarction, angina, PAD, stroke or TIA), hypertension, diabetes mellitus, and metabolic syndrome. Covariates with insignificant coefficients (P ⬎ 0.05) were excluded from the model. For all analyses, an a priori two-tailed level of significance was set at the 0.05 level using SAS software version 8.2 (SAS Institute Inc., Cary, NC).

Results The study sample consisted of 21,856 patients having at least one follow-up lipid panel prior to a censoring event (Fig. 1). Baseline characteristics of this study population revealed several differences between the three study cohorts (Table 3). Cohort 1 was slightly older, and had a greater

Characteristics of study cohorts at baseline (n ⫽ 21,856)

Characteristic

Cohort 1* (n ⫽ 8,161)

Cohort 2† (n ⫽ 7,901)

Cohort 3‡ (n ⫽ 5,794)

P value§

Age (y) Male Total follow-up in days Baseline lipids (mg/dL) Total cholesterol LDL-C HDL-C TG Non–HDL-C Deyo-Charlson score Baseline diagnoses Hypertension Diabetes mellitus Metabolic syndrome CV risk status Elevated risk primary prevention CHD or CHD risk equivalent Pre-baseline CV disease or event, % (n) CHD Coronary revascularization Stroke or transient ischemic attack Carotid or cerebral revascularization Peripheral artery disease Peripheral revascularization

66 ⫾ 12 52% (4,249) 962 ⫾ 376

63 ⫾ 13 51% (4,015) 968 ⫾ 374

64 ⫾ 12 37% (2,125) 957 ⫾ 372

⬍0.0001 ⬍0.0001 0.2122

226 146 59 105 167 1.11

⫾ ⫾ ⫾ ⫾ ⫾ ⫾

31 27 13 33 29 1.43

227 151 45 156 182 0.98

⫾ ⫾ ⫾ ⫾ ⫾ ⫾

37 29 11 59 33 1.42

239 153 39 235 200 1.03

⫾ ⫾ ⫾ ⫾ ⫾ ⫾

34 30 6 60 33 1.48

⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001

62% (5,074) 22% (1,791) 30% (2,435)

49 (3,843) 22 (1,743) 62 (4,915)

45 (2,601) 26 (1,494) 87 (5,059)

⬍0.0001 ⬍0.0001 ⬍0.0001

34% (2,761) 66% (5,400)

48 (3,780) 52 (4,121)

52 (2,994) 48 (3,800)

⬍0.0001 ⬍0.0001

13% 2% ⬍1% 1% 2% 1%

(1,077) (139) (14) (45) (165) (50)

11% 2% ⬍1% ⬍1% 1% ⬍1%

(847) (136) (4) (33) (102) (21)

9% 2% ⬍1% 1% 2% 1%

(507) (107) (8) (39) (105) (35)

⬍0.0001 0.7957 0.0749 0.1250 0.0012 0.0021

CV, cardiovascular; CHD, coronary heart disease; HDL-C, high-density lipoprotein; LDL-C, low-density lipoprotein cholesterol; TG, triglyceride. Data are presented as percentages (numbers) and mean ⫾ SD. *Cohort 1, patients with nonoptimal LDL-C levels only. †Cohort 2, patients with nonoptimal LDL-C and TG or nonoptimal LDL-C and HDL-C. ‡Cohort 3, patients with nonoptimal levels for all three lipid values (LDL-C, HDL-C, and TG). §Comparisons between cohorts using Pearson’s ␹2 test and F-test.

348 proportion with hypertension and higher CV risk. Cohorts 2 and 3 were more similar, including a less-skewed distribution of baseline CV risk compared to Cohort 1, with the exception of a female predominance in Cohort 3. By design, lipid values were markedly different between cohorts, and as might be anticipated by lipid value-driven stratification, metabolic syndrome and diabetes mellitus prevalence was higher in Cohorts 2 and 3. In the overall study population, the prevalence of the multiple lipid abnormality pattern encompassing nonoptimal HDL-C and/or TG in addition to nonoptimal LDL-C (45%, Cohorts 2 and 3) was greater than isolated elevated LDL-C (27%, Cohort 1).

Optimal lipid value achievement Among the 28,970 patients with a baseline full lipid panel (Fig. 1), the proportion of patients achieving optimal levels of all three lipid fractions was 29% for Cohort 1, 11% for Cohort 2, and 7% in cohort 3 over an average of 3 years of follow-up (Fig. 2). Among those individuals with combined achievement of optimal LDL-C, HDL-C, and TG, this attainment occurred after significant delay with a mean time of: 468 ⫾ 342 days (median 385) in Cohort 1, 498 ⫾ 350 days (median 413) in Cohort 2, and 615 ⫾ 412 days (median 505) in Cohort 3. Patients in Cohort 1 who eventually achieve combined optimal lipid values had lower mean baseline LDL-C values (144 ⫾ 27 vs. 147 ⫾ 27 mg/dL; P ⬍ 0.0001) and higher baseline TG values (107 ⫾ 34 vs. 104 ⫾ 32 mg/dL; P ⬍ 0.0001) compared to patients not achieving combined optimal lipid values (mean baseline HDL-C was not significantly different). Patients in Cohort 2, who eventually achieved combined optimal lipid values, had lower mean baseline LDL-C (149.4 ⫾ 27.2 vs. 151.8 ⫾ 30.2 mg/dL; P ⫽ 0.0003), HDL-C (42.7 ⫾ 10.0 vs. 46 ⫾ 11 mg/dL; P ⬍ 0.0001), and TG (146 ⫾ 53 vs. 161 ⫾ 61 mg/dL; P ⬍ 0.0001) values compared to patients not achieving combined optimal lipid values. Patients in Cohort 3 who eventually achieved combined optimal lipid values had higher mean baseline LDL-C (156 ⫾ 29 vs. 152 ⫾ 31 mg/dL; P ⫽ 0.0011), HDL-C (41 ⫾ 6 vs. 39 ⫾ 6 mg/dL; P ⬍ 0.0001), and lower TG (230 ⫾ 56 vs. 235 ⫾ 60 mg/dL; P ⫽ 0.0333) values than patients not achieving combined optimal lipid values. Achievement of optimal values for individual lipid parameters was analyzed from baseline over the entire duration of follow-up. Among patients in Cohort 1, there was a 36% increase in proportion of patients achieving desirable LDL-C levels and an 11% decrease for HDL-C. In Cohort 2 the proportion increased from 36% to 56% for HDL-C and 0% to 32% for LDL-C. In Cohort 3, the proportion of patients with desirable range HDL-C increased from zero at baseline to 31%, whereas LDL-C increased from 0% to 41%. Desired TG value achievement declined by 14% in Cohort 1, remained practically unchanged in Cohort 2 (64% to 66%), and increased from zero to 36% in Cohort 3 from baseline to study endpoint.

Journal of Clinical Lipidology, Vol 2, No 5, October 2008

Treatment patterns All patients followed in this study were eligible for some form of lipid-altering pharmacotherapy according to published guidelines, on the basis of their CV risk and presence of one or more nonoptimal lipid values. However, pharmacotherapy patterns observed over follow-up based on prescription claims reveal generalized under-treatment in each cohort, as well as considerable delays in the initiation of therapy. Among Cohort 1, 35% were treated with some lipid-altering therapy, whereas slightly more were treated in Cohorts 2 (42%) and 3 (53%). The proportion of patients treated with statin monotherapy was 34% (Cohort 1), 40% (Cohort 2), and 47% (Cohort 3), whereas very few were treated with combination drug therapy (Cohort 1: 2%, Cohort 2: 4%, Cohort 3: 7%). The mean days to first prescription for patients taking pharmacotherapy was 291 ⫾ 338 days (median, 167) for Cohort 1, 291 ⫾ 346 days (median, 140) for Cohort 2, and 268 ⫾ 331 days (median, 125) for Cohort 3.

Association of combined optimal lipid value achievement with CV event risk The risk of CV events over follow-up was significantly lower in patients achieving combined desired lipid values compared to those patients who did not within each of the study cohorts. The unadjusted CV event risk was 18% (HR, 0.82; 95% CI, 0.72– 0.95) lower for patients achieving combined optimal lipid values versus patients who did not in Cohort 1. In Cohorts 2 and 3, the event rates were lower by 37% (HR, 0.63; 95% CI, 0.64 – 0.84) and 43% (HR, 0.57; 95% CI, 0.39 – 0.83) respectively. Figure 3 depicts the Kaplan-Meier curves describing the risk of CV events over time for each cohort, stratified by the achievement of combined optimal values for LDL-C, HDL-C, and TG, as well as the Cox multivariate risk model results. Adjusted relative risk reduction for combined optimal lipid value achievement was 46% in Cohorts 2 and 3, whereas it was 25% in Cohort 1. The frequency of subsequent CV events was similar (3%, 7%– 8%, 10%, and 12% by years 1– 4 respectively) over time among all three cohorts. Furthermore, there was no significant difference in CV risk between Cohort 1 and 2 (HR, 1.02; 95% CI 0.93–1.12, data not shown) and 3 (HR, 1.01; 95% CI, 0.92–1.11, data not shown).

Discussion This retrospective cohort analysis of an administrative managed care claims database demonstrates a strong, independent, and consistent relationship between achievement of combined optimal values for LDL-C, HDL-C, and TG and reduced CV event rates over a median of 2.5 years of follow-up across three cohorts of patients with distinctly

Charland et al

Cardiovascular events and optimal lipid value achievement

Figure 2

349

Optimal lipid values achieved at baseline and during follow-up.

diverse patterns of baseline lipid abnormalities. Results suggest that the greatest benefit in terms of reduction of CV event risk was observed in patients with multiple lipid abnormalities at baseline that encompassed elevated LDL-C with one or both of HDL-C and/or TG. Those achieving optimal values for all parameters had a relative risk reduc-

tion of 46% in Cohorts 2 and 3 as opposed to those with baseline isolated high LDL-C (Cohort 1) whose relative reduction was 25%. The observed benefit of lowering LDL-C in Cohort 1 is concordant with effects of LDL-C reduction demonstrated in prior clinical trials of statins. A meta-analysis of these clinical trials demonstrated a pooled

350

Figure 3 Co-morbidity–adjusted Kaplan Meier curves stratified by combined optimal lipid value attainment, by cohort.

Journal of Clinical Lipidology, Vol 2, No 5, October 2008 21% reduced risk of events per 1mmol/L (ie, 39 mg/dL) reduction in LDL-C.7 Although patients in Cohort 1 were more likely to be hypertensive and to have had prior CV disease as of their baseline laboratory visit, there was no statistically significant difference in the rate of subsequent CV events among the three total cohorts over study followup. Furthermore, the results of this study are consistent with a previous analysis of a managed care population at risk for CV events which reported a 45% adjusted increase in CV events in patients with all three lipid parameters nonoptimal compared to all three parameters optimal.17 Small clinical trials of combined lipid-altering pharmacotherapy8,9,11,12,21,22 have demonstrated marked and sustained effects on atherosclerosis progression/regression as well as CV event rate reductions of ⬎50%. The magnitude of this effect is quite similar to the 46% reduction in relative risk observed in this study among patients in Cohorts 2 and 3 who achieved combined optimal lipid values during follow-up, presumably due to incremental increases in HDL-C in addition to further reductions in LDL-C and TG. However, a word of caution is indicated because these analyses have come from trials in which the primary endpoint was often not clinical events and were often retrospective as with our study. A definitive trial in patients with elevated LDL-C, elevated TG, and low HDL-C has not been completed. It is not altogether surprising that although a significant CV event reduction was demonstrated with optimal LDL-C attainment in Cohort 1, CV event reductions of greater magnitude were observed in Cohorts 2 and 3 with rendering HDL-C and/or TG to the desirable range in addition to LDL-C. This has been corroborated in other epidemiologic analyses23 and clinical studies.8,9,11,12,21,22 Logically, it would be expected that therapy targeting a single independent risk factor among a constellation of similarly independent risk factors would be of limited benefit compared to modification of multiple risk factors that could confer additional benefit. This study also suggests that the time to benefit for reduction of risk in all three cohorts may have been evident after approximately 6 months with residual risk reductions continuing over the study duration. Although these inferences are mainly derived from assessment of visual separation of the Kaplan Meier curves for each cohort, similar indications of risk reduction have been seen in controlled lipid-lowering trials examining the effect of optimal attainment on one or more lipid fractions.7,11,12,24 –28 There is a tremendous opportunity to reduce CV events in an elevated risk population if more patients could be transitioned to two or three lipid values at optimal levels, as demonstrated by these findings. For this to occur, it would require a concerted effort to first identify patients with low HDL-C and/or high TG levels in addition to elevated LDL-C levels, and to then initiate both therapeutic lifestyle changes and appropriate pharmacotherapy tailored to the lipid abnormalities present. The pharmacotherapy patterns and rates of optimal achievement observed in this study population provide us with some insight into this process,

Charland et al

Cardiovascular events and optimal lipid value achievement

and highlight an important need. Although all patients in this analysis were at elevated cardiovascular risk and had at least one nonoptimal lipid value (all had nonoptimal LDL-C), dyslipidemia pharmacotherapy of any type was markedly underutilized (only 34%– 47% of patients treated) and delayed (approximately 9 months before initiation), contributing to very low rates of optimal lipid value attainment (7%–29%). When therapy was initiated, it was predominantly in the form of statin monotherapy, as would be appropriate initial therapy in patients with elevated LDL-C.2 However, initiation of combination therapy in patients with low HDL-C and/or elevated TG was extremely low (2%–7% of patients). Focused treatment targeting HDL-C and, to a lesser extent, triglycerides is of increasing interest but not yet welldefined in existing treatment guidelines.3–5 Combination therapy of a statin plus niacin, fibrate, and/or a bile acid sequestrant has been associated with a 23%–38% increase in HDL-C, 22%– 48% reduction in LDL-C, and 36%– 46% reduction in TG levels. These effects are greater than that achieved by a statin alone.8,11,22 Ongoing studies such as Atherothrombosis Intervention in Metabolic Syndrome with Low HDL/High Triglycerides and Impact on Global Health Outcomes (AIM-HIGH),29 Heart Protection Study 2 (HPS2),30 and Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol-6 HDL and LDL Treatment Strategies in Atherosclerosis) (ARBITER-6HALTS)31 will provide much needed additional evidence upon which to develop the most rational therapeutic approach to combination therapy. This study highlights several significant clinical, policy, and economic issues that provide insight and direction to clinical management at the individual practice level and among large provider groups or managed care organizations. An important lesson from this evaluation is that patients with dyslipidemia in the managed care environment can achieve optimal lipid levels and realize CV event risk reduction as observed in randomized clinical trials. Thus, this naturalistic assessment demonstrates a successful translation from the controlled clinical research environment to practical clinical application. The study also provides an objective database that identifies opportunities for improvement in lipid risk-factor intervention that may be fairly easily achieved with minor patient management modification by clinicians and other disease management decisionmakers. These include greater awareness of multiple lipid abnormalities, development of a simple classification of patients based on these lipid profiles, and attention to more rapid attainment of optimal levels through appropriate treatment of LDL-C, HDL-C, and TG in unison. Some limitations should be considered when interpreting these results. Although the patient population was derived from a single southeastern United States managed care plan, which may limit its generalizability, patient management was at the discretion of the physician, which increases the relevance of this study compared to controlled clinical trials. Because the study design was retrospective and nonrandom-

351

ized, the database did not contain data on sociodemographic factors (eg, race, education, or income), other risk factors (eg, smoking, diet, exercise, and family history, among others), or patient compliance, which may have influenced cohort assignment, treatment patterns, and/or the association between CV events and optimal lipid value attainment. Potentially, enrolled patients categorized as high risk or CHD/CHD-RE status had greater co-morbidities and/or higher CV risk than those who were excluded due to too few lipid assessments (⬍2). Patients with worse dyslipidemia (either more fractions suboptimal and/or greater severity) may have had more physician intervention, more frequent monitoring, and/or possibly greater compliance and treatment persistence. It is reasonable to conclude, however, that the results derived from this study may be conservative because patients who did not have lipid panels measured may remain untreated leading to lower optimal value attainment, and mortality data were not included.

Conclusion Less than one-third of elevated risk primary prevention and CHD/CHD risk equivalent patients with three distinct patterns of abnormal plasma lipids, achieved combined optimal values for LDL-C, HDL-C and TG over an average three-year follow-up in this administrative claims analysis, and the time to reach combined optimal attainment exceeded one year. Lipid-altering drug therapy was underutilized and substantially delayed among all patient groups. Patients achieving combined optimal lipid values had a significantly lower risk of CV events over follow-up compared to those not achieving optimal values in each of the three cohorts analyzed, and the greatest magnitude of benefit was observed in patients with baseline elevated LDL-C plus one or both of HDL-C and/or TG also outside of acceptable levels. Based on these observations, lifestyle and pharmacotherapy interventions sufficient to safely render LDL-C, HDL-C and TG within optimal values may substantially improve CV outcomes in patients with both isolated elevated LDL-C and multiple lipid abnormalities that include elevated LDL-C plus low HDL-C and/or elevated TG.

Acknowledgement This study was funded by Kos Pharmaceuticals, Inc, and Abbott Laboratories. The authors wish to thank Patti Peeples, PhD (HealthEconomics.Com, Ponte Vedra Beach, FL), for her assistance with this manuscript.

Financial disclosures This study was funded by an unrestricted research grant from Kos Laboratories, Inc., a Subsidiary of Abbott Labora-

352 tories and Abbott Laboratories. The sponsor was permitted to review the manuscript, but all final decisions regarding content remained the sole responsibility of the investigators. Dr. Charland received an unrestricted research grant from Abbott Laboratories. Mr. Quimbo and Dr. Cziraky are employees of HealthCore, Inc. and have received research grants from Abbott Laboratories, AstraZeneca PLC, BristolMyers Squibb, Merck & Co., Inc., Novartis International AG, and Pfizer Incorporated. Dr. Karas has served as a consultant and speaker for Abbott Laboratories, and additionally has provided research support for Merck & Co., Inc. Dr. Insull has served as a speaker for Abbott Laboratories, Schering-Plough Corporation, and Merck and Co., Inc., as an editor for AstraZeneca Pharmaceuticals LP, and a consultant for Daichi Sankyo, Inc. Dr. Davidson has served as a speaker/consultant and has received research grants from Abbott Laboratories, AstraZeneca Pharmaceuticals, Daiichi-Sankyo, Inc., Merck & Co., Inc. Merck/ScheringPlough, Pfizer Laboratories, Reliant Pharmaceuticals, Inc., and Takeda Pharmaceuticals. Additionally, Dr. Davidson has received research grants from Oscient Pharmaceuticals (also served as a speaker), Roche Pharmaceuticals (also served as consultant), Access Health, Atherogenics and Xinthria Pharmaceuticals. Dr. Davidson has also served as a consultant for Sanofi Aventis, and diaDexus, Inc. (also served as speaker). Finally, Dr. Davidson serves on the Board of Directors for Angiogen and Sonogene, while serving as chief medical officer for Professional Evaluation, Inc. Dr. Davidson has no stock ownership or options, etc., in Radiant Research, a Division of Swiss BioScience or any of the above listed companies. Dr. Stanek has served as a consultant for Abbott Laboratories, Oscient Pharmaceuticals, Inc., and HealthCore, Inc.

Journal of Clinical Lipidology, Vol 2, No 5, October 2008

8.

9.

10. 11.

12.

13.

14.

15.

16.

17.

18.

19.

References 1. American Heart Association Disease and Stroke Statistics—2008 Update. Dallas, TX: American Heart Association, 2008. 2. NCEP Expert Panel. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III) final report. Circulation. 2002;106:3143–3421. 3. American Diabetes Association. Standards of medical care in diabetes2008. Diabetes Care. 2008;31(suppl 1):S12–S54. 4. Grundy SM, Cleeman JI, Merz CN, et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III Guidelines. J Am Coll Cardiol. 2004;44:720 –732. 5. Mosca L, Banka CL, Benjamin EJ, et al. Evidence-based guidelines for cardiovascular disease prevention in women: 2007 update. Circulation. 2007;115:1481–1501. 6. Graham I, Atar D, Borch-Johnsen K, et al. European guidelines on cardiovascular disease prevention in clinical practice: executive summary. Fourth Joint Task Force of the European Society of Cardiology and other societies on cardiovascular disease prevention in clinical practice (constituted by representatives of nine societies and by invited experts). Eur J Cardiovasc Prev Rehabil. 2007;14(suppl. 2):E1– 40. 7. Baigent C, Keech A, Kearney PM, et al. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from

20.

21.

22.

23.

24.

25.

26.

90,056 participants in 14 randomised trials of statins. Lancet. 2005; 366:1267–1278. Brown BG, Zhao XQ, Chait A, et al. Simvastatin and niacin, antioxidant vitamins, or the combination for the prevention of coronary disease. N Engl J Med. 2001;345:1583–1592. Brown BG, Stukovsky KH, Zhao XQ. Simultaneous low-density lipoprotein-C lowering and high-density lipoprotein-C elevation for optimum cardiovascular disease prevention with various drug classes, and their combinations: a meta-analysis of 23 randomized lipid trials. Curr Opin Lipidol. 2006;17:631– 636. Superko HR, King S III. Lipid management to reduce cardiovascular risk: a new strategy is required. Circulation. 2008;117:560 –568. Taylor AJ, Sullenberger LE, Lee HJ, et al. Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol (ARBITER) 2: a double-blind, placebo-controlled study of extendedrelease niacin on atherosclerosis progression in secondary prevention patients treated with statins. Circulation. 2004;110:3512–3517. Taylor AJ, Lee HJ, Sullenberger LE. The effect of 24 months of combination statin and extended-release niacin on carotid intima-media thickness: ARBITER 3. Curr Med Res Opin. 2006;22:2243–2250. Alsheikh-Ali AA, Lin JL, Abourjaily P, et al. Extent to which accepted serum lipid goals are achieved in a contemporary general medical population with coronary heart disease risk equivalents. Am J Cardiol. 2006;98:1231–1233. Mosca L, Merz NB, Blumenthal RS, et al. Opportunity for intervention to achieve American Heart Association guidelines for optimal lipid levels in high-risk women in a managed care setting. Circulation. 2005;111:488 – 493. Sarawate CA, Cziraky MJ, Stanek EJ, et al. Achievement of optimal combined lipid values in a managed care setting: is a new treatment paradigm needed? Clin Ther. 2007;29:196 –209. Stacy TA, Egger A. Results of retrospective chart review to determine improvement in lipid goal attainment in patients treated by highvolume prescribers of lipid-modifying drugs. J Manag Care Pharm. 2006;12:745–751. Stanek EJ, Sarawate C, Willey VJ, et al. Risk of cardiovascular events in patients at optimal values for combined lipid parameters. Curr Med Res Opin. 2007;23:553–563. Mosca L, Appel LJ, Benjamin EJ, et al. Evidence-based guidelines for cardiovascular disease prevention in women. Circulation. 2004;109: 672– 693. American Diabetes Association. Dyslipidemia management in adults with diabetes Diabetes Care. 2004;27(suppl 1):S68 –S71. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613– 619. Boden WE. High-density lipoprotein cholesterol as an independent risk factor in cardiovascular disease: assessing the data from Framingham to the Veterans Affairs High-Density Lipoprotein Intervention Trial. Am J Cardiol. 2000;86(suppl.):19L–22L. Whitney EJ, Krasuski RA, Personius BE, et al. A randomized trial of a strategy for increasing high-density lipoprotein cholesterol levels: effects on progression of coronary heart disease and clinical events. Ann Intern Med. 2005;142:95–104. Koro CE, Bowlin SJ, Stump TE, et al. The independent correlation between high-density lipoprotein cholesterol and subsequent major adverse coronary events. Am Heart J. 2006;151:755. Arca M, Montali A, Valiante S, et al. Usefulness of atherogenic dyslipidemia for predicting cardiovascular risk in patients with angiographically defined coronary artery disease. Am J Cardiol. 2007;100:1511–1516. Brown B. Maximizing coronary disease risk reduction using nicotinic acid combined with LDL-lowering therapy. Eur Heart J Suppl. 2005; 7:F34 –F40. Gotto AM, Jr.Establishing the benefit of statins in low-to-moderate– risk primary prevention: the Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS). Atheroscler Suppl. 2007;8: 3– 8.

Charland et al

Cardiovascular events and optimal lipid value achievement

27. Robins SJ, Collins D, Wittes JT, et al. Relation of gemfibrozil treatment and lipid levels with major coronary events: VA-HIT: a randomized controlled trial. JAMA. 2001;285:1585–1591. 28. Sharrett AR, Ballantyne CM, Coady SA, et al. Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: The Atherosclerosis Risk in Communities (ARIC) Study. Circulation. 2001;104:1108 –1113. 29. Atherothrombosis Intervention in Metabolic Syndrome with Low HDL/High Triglycerides and Impact on Global Health Outcomes.

353

Available at: http://www.aimhigh-heart.com/background.shtml. Accessed November 11, 2007. 30. Treatment of HDL to Reduce the Incidence of Vascular Events HPS2THRIVE. Available at: http://clinicaltrials.gov/ct2/show/NCT00461630. Accessed February 14, 2008. 31. Devine PJ, Turco MA, Taylor AJ. Design and rationale of the ARBITER 6 trial (Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol)-6-HDL and LDL Treatment Strategies in Atherosclerosis (HALTS). Cardiovasc Drugs Ther. 2007; 21:221–225.