Figure 1. Relationship between presence or absence of metabolic syndrome (MetS) and type 2 diabetes and of vascular events. CV indicates cardiovascular; CHD, coronary heart disease; and CI, confidence interval. *Wald P value. †Three-group comparison. ‡Compared with the reference group (neither type 2 diabetes nor MetS). Hazard ratio results reflect adjustment for treatment, region (24% North America, 21% Central Europe, 11% Eastern Europe, 21% Mediterranean, 12% Scandinavia, 2% South Africa, 8% Australia or New Zealand, 1% Latin America), entry event, time since the entry event, sex, and age.
Figure 2. Effect of atorvastatin on the risk of stroke according to the presence or absence of metabolic syndrome (MetS) or type 2 diabetes. CI indicates confidence interval. *Within-group P value. † P value for treatment × subgroup interaction test.
Figure 3. Effect of atorvastatin on cardiac end points and death according to the presence or absence of metabolic syndrome (MetS) or type 2 diabetes. CV indicates cardiovascular; CHD, coronary heart disease; and CI, confidence interval. Results reflect adjustment for region, entry event, time since the entry event, and age. *Within-group P value. † P value for treatment × subgroup interaction test.
Callahan A, Amarenco P, Goldstein LB, Sillesen H, Messig M, Samsa GP, Altafullah I, Ledbetter LY, MacLeod MJ, Scott R, Hennerici M, Zivin JA, Welch KMA, SPARCL Investigators FT. Risk of Stroke and Cardiovascular Events After Ischemic Stroke or Transient Ischemic Attack in Patients With Type 2 Diabetes or Metabolic SyndromeSecondary Analysis of the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) Trial. Arch Neurol. 2011;68(10):1245-1251. doi:10.1001/archneurol.2011.146
Author Affiliations: Vanderbilt University, Nashville (Dr Callahan) and Neurology Center of Middle Tennessee, Columbia (Dr Ledbetter); Institut National de la Santé et de la Recherche Médicale Unit 698 and Denis Diderot University, Paris, France (Dr Amarenco); Duke University Medical Center, Durham, North Carolina (Drs Goldstein and Samsa); University of Copenhagen, Copenhagen, Denmark (Dr Sillesen); Pfizer Inc, New York, New York (Dr Messig); Minneapolis Clinic of Neurology, University of Minnesota, Golden Valley (Dr Altafullah); University of Aberdeen, Aberdeen, Scotland (Dr MacLeod); Lipid and Diabetes Research Group, Christchurch Hospital, Christchurch, New Zealand (Dr Scott); Universitat Heidelberg, Mannheim, Germany (Dr Hennerici); University of California, San Diego (Dr Zivin); and Rosalind Franklin University of Medicine and Science, North Chicago, Illinois (Dr Welch).
Group Information: A list of the SPARCL Investigators was published in N Engl J Med. 2006;355(6):549-559.
Objective To perform a secondary analysis of the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) trial, which tested the effect of treatment with atorvastatin in reducing stroke in subjects with a recent stroke or transient ischemic attack, to explore the effects of treatment in subjects with type 2 diabetes mellitus or metabolic syndrome (MetS).
Methods The 4731 subjects enrolled in the SPARCL trial were classified as having type 2 diabetes mellitus at enrollment (n = 794), MetS retrospectively (n = 642), or neither diabetes nor MetS (n = 3295, the reference group) based on data collected at baseline. Cox regression models were used to determine whether the effect of treatment on the primary end point (combined risk of nonfatal and fatal stroke) and secondary end points (major coronary events, major cardiovascular events, any coronary heart disease event, and any revascularization procedure) varied based on the presence of type 2 diabetes mellitus or MetS.
Results Subjects with type 2 diabetes mellitus had increased risks of stroke (hazard ratio [HR] = 1.62; 95% confidence interval [CI], 1.33-1.98; P < .001), major cardiovascular events (HR = 1.66; 95% CI, 1.39-1.97; P < .001), and revascularization procedures (HR = 2.39; 95% CI, 1.78-3.19; P < .001) compared with the reference group. Subjects with MetS were not at increased risk for stroke (P = .78) or major cardiovascular events (P = .38) but more frequently had revascularization procedures (HR = 1.78; 95% CI, 1.26-2.5; P = .001). There were no treatment × subgroup interactions for the SPARCL primary end point (P = .47).
Conclusions The SPARCL subjects with type 2 diabetes were at higher risk for recurrent stroke and cardiovascular events. This exploratory analysis found no difference in the effect of statin treatment in reducing these events in subjects with or without type 2 diabetes or MetS.
Trial Registration clinicaltrials.gov Identifier: NCT00147602
Type 2 diabetes mellitus is associated with increased risks of both stroke and coronary heart disease (CHD).1- 4 Metabolic syndrome (MetS) or insulin resistance identifies individuals at higher risk for developing type 2 diabetes and is also associated with an increase in stroke and cardiovascular morbidity.5- 7 Each component of the cluster of risk factors defining MetS increases the likelihood of stroke (ie, hypertension,8 high triglycerides level,9 low high-density lipoprotein cholesterol level,10,11 high serum glucose level,4 and abdominal obesity12). Metabolic syndrome has also been associated with symptomatic intracranial stenosis13 and with carotid atherosclerosis, conditions associated with increased stroke risk.14,15
The Cholesterol Treatment Trialists' collaboration's meta-analysis of 18 686 people with diabetes included 14 randomized trials of statins.16 There was a 21% (95% confidence interval, 7-33; P < .001) reduction of first stroke, which was similar to the effect observed in the 71 370 patients without diabetes. No information is available, however, on the effect of statins on secondary stroke prevention in diabetic patients or in those with MetS compared with those without these conditions.
The Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) trial found that statin treatment reduced stroke risk in patients with recent stroke or transient ischemic attack (TIA) and no known CHD.17 The SPARCL cohort provides an opportunity to explore the impact of having type 2 diabetes or MetS on the effect of statin treatment for secondary stroke prevention.
The rationale, design, and methods of the SPARCL trial have been described in detail previously.17,18 They are summarized here.
Eligible patients were women and men older than 18 years who had had an ischemic or hemorrhagic stroke or TIA (diagnosed by a neurologist within 30 days of the event) 1 to 6 months before randomization. Stroke was defined by focal clinical signs of central nervous system dysfunction of vascular origin that lasted for at least 24 hours. Transient ischemic attack was defined by the loss of cerebral or ocular function for less than 24 hours of presumed ischemic cause. Patients had to be ambulatory with a modified Rankin score of 3 or less (scores can range from 0-5, with higher scores indicating greater disability) and a low-density lipoprotein cholesterol level of 100 to 190 mg/dL (to convert to millimoles per liter, multiply by 0.0259). Patients who were taking lipid-lowering medications had to stop taking them 30 days before the screening phase of the study, and these drugs were prohibited during the course of the trial. Excluded patients included those with atrial fibrillation, mechanical prosthetic valves, or subarachnoid hemorrhage. The local research and ethics committee or institutional review board of each participating center approved the study protocol (15 of 205 centers excluded otherwise suitable patients with a low-density lipoprotein cholesterol level >160 mg/dL as required by their institutional review boards), and all patients gave written informed consent. Patients were enrolled between September 1998 and March 2001.
The SPARCL subjects were identified as having type 2 diabetes mellitus or no history of diabetes at baseline without other prespecified criteria. The remaining subjects without type 2 diabetes were classified as having MetS or as having neither diabetes mellitus nor MetS based on data collected at baseline. Metabolic syndrome was defined using World Health Organization and National Cholesterol Education Program Adult Treatment Panel III criteria,19- 21 and these subjects had to have at least 3 of the following: body mass index (calculated as weight in kilograms divided by height in meters squared) higher than 30, fasting blood glucose level of 100 to 125 mg/dL (to convert to millimoles per liter, multiply by 0.0555), systolic blood pressure higher than 130 mm Hg, triglycerides level higher than 150 mg/dL (to convert to millimoles per liter, multiply by 0.0113), and high-density lipoprotein cholesterol level lower than 40 mg/dL (to convert to millimoles per liter, multiply by 0.0259) for men and lower than 50 mg/dL for women. The data were also analyzed using the International Diabetes Federation criteria for MetS.22 Waist circumference and microalbuminuria were not recorded at baseline.
The SPARCL primary outcome was the time from randomization after an ischemic stroke or TIA to the occurrence of a nonfatal or fatal stroke. The SPARCL trial included several secondary composite outcomes, including major coronary events (cardiac death, nonfatal myocardial infarction, or resuscitated cardiac arrest), major cardiovascular events (stroke plus any major coronary event), any CHD event (acute coronary event plus coronary revascularization procedure, unstable angina, or angina or ischemia requiring emergent hospitalization), and any revascularization procedure (coronary, carotid, or peripheral).17 Subjects were followed up until the end of the study or death; only 0.5% of enrolled subjects were lost to follow-up. An independent committee adjudicated all potential end points without knowledge of the patients' treatment status or cholesterol levels.
Laboratory assessments were performed and electrocardiograms were obtained and subsequently interpreted at screening, at 6-month intervals, and on completion of the study. Drug safety was assessed by an evaluation of the type, frequency, severity, and duration of any reported adverse event and on the basis of vital signs, physical examination findings, and laboratory test results.
The relative risk of outcome events was calculated for subjects with type 2 diabetes mellitus or MetS compared with those without these conditions (reference group). Cox proportional hazards models were used to calculate hazard ratios, 95% confidence intervals, and P values based on the SPARCL prespecified adjustments for geographic region, entry event (stroke or TIA), time since the entry event (number of days from the qualifying entry event to randomization), sex, and age at baseline. The group (type 2 diabetes, MetS, or neither diabetes nor MetS) main effect was used in the Cox models to test for overall risk differences among the 3 groups using a χ2 statistic with 2 df. Group × treatment interaction effects were used to test for heterogeneity among the 3 groups. Testing for a treatment × subgroup interaction effect is the most appropriate way of evaluating the potential differential impact of an intervention.23,24 All analyses were based on intention to treat (ie, all randomized subjects were analyzed according to their assigned treatment with atorvastatin or placebo). Two-sided P < .05 was considered statistically significant. There was no adjustment for multiple testing.
At baseline, 794 subjects (16.8%) had type 2 diabetes, 642 (13.6%) had MetS, and 3295 (69.6%) had neither MetS nor type 2 diabetes (reference group). Diabetic patients were on average 4 years older than those with MetS. There was no difference between the 3 groups in the distribution of ischemic stroke subtypes at randomization (Table 1). Among the subjects with MetS, 60% were obese, 39% had glycemic abnormalities, 70% had hypertension, 83% had hypertriglyceridemia, and 77% had reduced high-density lipoprotein cholesterol values.
Table 2 shows lipid values at baseline and while receiving treatment in the 3 groups. Treatment with atorvastatin (80 mg) was associated with similar reductions in low-density lipoprotein cholesterol level regardless of group. Triglycerides levels were reduced by 11.3% in the type 2 diabetes mellitus group, 20.2% in those with MetS, and 9.0% in the reference group. High-density lipoprotein cholesterol levels were similar at baseline and while receiving treatment across the 3 subgroups in the atorvastatin and placebo arms.
The risk of stroke was 11.0% in the reference group, 18.1% in subjects with type 2 diabetes (P < .001), and 10.7% in those with MetS (P = .78) (Figure 1). Subjects with type 2 diabetes mellitus were more likely to have any of the primary and secondary end points, including death (all P < .001) (Figure 1). Subjects with MetS had no increase in the risk of major cardiovascular events (P = .38) or major coronary events (P = .19) but were more likely to have any CHD event (P = .01) or revascularization procedure (P = .001).
Figure 2 shows that there was no treatment × subgroup interaction for the SPARCL primary end point (P = .47). Among those experiencing a stroke during the trial, there was no difference in subtypes (ischemic or hemorrhagic) between groups.
Figure 3 shows that treated subjects with type 2 diabetes had a reduction in major cardiovascular events, major coronary events, any CHD event, and any revascularization procedure but not death. There were no treatment × subgroup interactions for any of these end points.
Using International Diabetes Federation criteria, the number of subjects with MetS was reduced from 642 to 574 (from 13.6% to 12.1% of the overall population). The results were not substantially different when the analyses were repeated for all groups.
As previously reported, transaminase levels were slightly but significantly elevated in the atorvastatin group17 but did not differ among the 3 study groups. Muscle abnormalities were rare and did not differ between groups.
In this post hoc analysis, we found that SPARCL subjects with type 2 diabetes mellitus were at higher risk for recurrent stroke and cardiovascular events but that there was no difference in the effect of statin treatment in reducing these events in subjects with or without type 2 diabetes or MetS (Figure 2 and Figure 3). Although not based on specific criteria, investigators were required to record whether a subject had a history of type 2 diabetes mellitus on the baseline enrollment form. Although this was a preplanned secondary analysis, these results should be viewed as exploratory. It should be recognized that the SPARCL study was not powered to test for subgroup effects and that this analysis is likely underpowered. The results are, however, similar to the findings of the Cholesterol Treatment Trialists' collaboration that statin's effect on stroke risk was similar in diabetic and nondiabetic patients.16
Epidemiological studies indicate that persons with stroke or TIA and type 2 diabetes or MetS have a higher risk of stroke and other cardiovascular events.1- 12 A secondary analysis of subjects with stable CHD randomized in the Treating to New Targets trial found that those with MetS including type 2 diabetes had a higher risk of stroke than those without MetS.25 Similar to our results, Treating to New Targets subjects with MetS and no diabetes had no increase in stroke risk. In contrast to our results, the Treating to New Targets trial found that those with MetS without or with type 2 diabetes had a higher risk of major cardiovascular events; however, those classified as having MetS in the SPARCL trial did not have type 2 diabetes. The difference in the results of epidemiological studies compared with our cohort and that of the Treating to New Targets trial might be due to baseline differences in subject characteristics, lack of power, and the effect of intensive risk factor management for patients enrolled in clinical trials. During the SPARCL trial, investigators carefully controlled blood pressure and other risk factors and all patients received appropriate antithrombotic treatment, which is usually not the case for those included in epidemiological studies. Prior studies show that tightly controlling blood pressure reduces the risk of stroke by 40% in diabetic subjects.26
In conclusion, the SPARCL trial was unique in that it was specifically designed to evaluate the effects of statin treatment in subjects with a recent noncardioembolic stroke or TIA and no known CHD. Although the possibility of variation in the benefit of statin treatment in subjects with or without type 2 diabetes or MetS cannot be excluded by this analysis, there was no evidence of a difference in treatment effect.
Correspondence: Alfred Callahan, MD, The Stroke and Heart Attack Prevention Center, 2000 Glen Echo Rd, Ste 122, Nashville, TN 37215 (firstname.lastname@example.org).
Accepted for Publication: March 22, 2011.
Published Online: June 13, 2011. doi:10.1001/archneurol.2011.146
Author Contributions: All authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis; the SPARCL steering committee developed the study protocol with the sponsor and takes responsibility for the data and data analyses. Study concept and design: Callahan, Amarenco, Sillesen, MacLeod, and Welch. Acquisition of data: Callahan, Amarenco, Altafullah, Ledbetter, MacLeod, Zivin, and Welch. Analysis and interpretation of data: Callahan, Amarenco, Goldstein, Messig, Samsa, Altafullah, MacLeod, Scott, Hennerici, and Welch. Drafting of the manuscript: Callahan, Scott, and Zivin. Critical revision of the manuscript for important intellectual content: Callahan, Amarenco, Goldstein, Sillesen, Messig, Samsa, Altafullah, Ledbetter, MacLeod, Scott, Hennerici, Zivin, and Welch. Statistical analysis: Messig and Samsa. Obtained funding: Zivin. Administrative, technical, and material support: Ledbetter. Study supervision: Callahan, Sillesen, Hennerici, and Welch.
Financial Disclosure: Dr Callahan has received research grant support from Pfizer and honoraria from Pfizer, sanofi-aventis, and Bristol-Myers Squibb. Dr Amarenco has received consulting fees from AstraZeneca, Bristol-Myers Squibb, Daiichi, Eli Lilly, GlaxoSmithKline, Guerbet, Negma, Novartis, Pfizer, Sankyo, sanofi-aventis, and Servier; lecture fees from AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Merck, Otsuka Pharmaceutical, Pfizer, sanofi-aventis, and Servier; and grant support from Boehringer Ingelheim, Bristol-Myers Squibb, Eisai, Pfizer, and sanofi-aventis. Dr Goldstein has received consultancy fees and honoraria from Pfizer. Dr Sillesen has received consulting fees from Pfizer and sanofi-aventis; lecture fees from AstraZeneca, Bristol-Myers Squibb, Merck, Pfizer, and sanofi-aventis; and grant support from Pfizer. Dr Messig is an employee of Pfizer and owns stock in Pfizer. Dr Samsa has received consultancy fees from Boehringer Ingelheim. Dr Altafullah has received honoraria and research support from Pfizer; has been part of the speaker's bureau of Pfizer; and owns stock in Pfizer. Dr Hennerici has received research grant support from Pfizer. Dr Zivin has received consulting fees from Ambit, AstraZeneca, CytRx, Merck, Johnson & Johnson, PhotoThera, PhRMA, Pfizer, and Remedy Pharmaceutical. Dr Welch has received consulting fees from Eisai, GlaxoSmithKline, Medpointe/AstraZeneca NMT Med, Ortho-McNeil, and Pfizer; lecture fees from GlaxoSmithKline and Pfizer; and grant support from Pfizer.
Funding/Support: This study was sponsored by Pfizer Inc.
Role of the Sponsor: Pfizer Inc was involved in the design and conduct of the study; collection, management, and analysis of the data; and review of the manuscript.
Additional Contributions: No content development support was provided, but assistance in preparing the figures and manuscript for submission was provided by Shuang Li, PhD, at UBC Scientific Solutions and was funded by Pfizer Inc. Medpace managed all data. Medpace, Charles River Laboratories Clinical Services, and Pfizer Inc provided site monitoring throughout the study.