Statin Use and Risk of Cerebral Aneurysm Rupture: A Hospital-based Case–control Study in Japan

Statin Use and Risk of Cerebral Aneurysm Rupture: A Hospital-based Case–control Study in Japan Yayoi Yoshimura, MD,* Yoshitaka Murakami, PhD,† Makoto Saitoh, MD,* Toshihiro Yokoi, MD,* Tomohiro Aoki, MD, PhD,‡ Katsuyuki Miura, MD, PhD,xjj Hirotsugu Ueshima, MD, PhD,xjj and Kazuhiko Nozaki, MD, PhD,* for the SSS Research Group

Background: Recent reports have showed that some statins have protective effects in experimental cerebral aneurysm models. We conducted a case–control study to investigate an association between statin use and the rupture risk of cerebral aneurysm in Japanese population. Methods: This was a multihospital case–control study; cases and controls were collected from 15 hospitals in Japan. Cases consisted of patients with aneurysmal subarachnoid hemorrhage hospitalized from April 2009 to March 2011. Controls were selected from patients who had newly diagnosed unruptured saccular aneurysms from April 2006 to March 2011. The primary exposure of interest was statin use. Multivariable logistic regression was used to assess the relationship between stain use and the rupture risk of cerebral aneurysm. Results: A total of 117 cases and 304 controls were included in the analyses. Statin was used in 9.4% of cases and 26.0% of controls. Controls had a significantly higher rate of use of statin. The use of any statin was associated with cerebral aneurysm rupture after adjustment of potential confounders (adjusted odds ratio: .30, 95% confidence interval: .14-.66). The association was similar in each stratum of total cholesterol level. Conclusions: This observation from a hospital-based case-control study in Japan suggested that there is inverse relationship between use of statins and cerebral aneurysm rupture. Future clinical studies are needed. Key Words: Statin—cerebral aneurysm—subarachnoid hemorrhage—case-control study. Ó 2014 by National Stroke Association

Introduction With the prevalent use of noninvasive neuroimaging techniques and spread of medical checkup of the brain, many incidental cerebral aneurysms are detected in the clinical practice, and the overall prevalence of cerebral

From the *Department of Neurosurgery, Shiga University of Medical Science, Seta-Tsukinowa-cho, Otsu, Shiga; †Department of Medical Statistics, Shiga University of Medical Science, Seta-Tsukinowa-cho, Otsu, Shiga; ‡Department of Pharmacology, Kyoto University Graduate School of Medicine, Knoe-cho Yoshida Sakyo-ku Kyoto; xDepartment of Health Science, Shiga University of Medical Science, Seta-Tsukinowa-cho, Otsu, Shiga; and jjCenter for Epidemiologic Research in Asia, Shiga University of Medical Science, Seta-Tsukinowacho, Otsu, Shiga, Japan. Received December 14, 2012; revision received February 14, 2013; accepted April 13, 2013.

aneurysms is estimated as 3.2% in a population without comorbidity with a mean age of 50 years.1 Once incidental aneurysms are detected, careful observation and intervention should be continued to prevent their progression and rupture. Subarachnoid hemorrhage (SAH) is a serious

Grant support: This paper is partly supported by a grant of Ministry of Health, Labour and Welfare (H21-Clinical researchgeneral-008). Disclosures: None. Address correspondence to Yayoi Yoshimura, MD, Department of Neurosurgery, Shiga University of Medical Science, Seta-Tsukinowa-cho, Otsu, Shiga 520-2192, Japan. E-mail: yayoi26@belle. shiga-med.ac.jp. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2013.04.022

Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 2 (February), 2014: pp 343-348

343

Y. YOSHIMURA ET AL.

344

disease that has high morbidity and mortality rate, and average onset of SAH is younger than that of other stroke subtypes,2,3 which causes a greater productivity loss in the society. Therefore, SAH prevention has an important role from the public health perspective. Particularly in Japan and Finland, the risk of rupture from cerebral aneurysms is estimated to be higher than other countries.4 Surgical clipping or endovascular treatment is performed to prevent SAH, but there are constant morbidity and mortality.5 To date, medical treatments to prevent a rupture of cerebral aneurysms are not available, and nonsurgical procedure, such as pharmacological treatments, is expected in this area. Several reports have showed that some statins (3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors) have protective effects in experimental cerebral aneurysm models.6-8 However, no evidence of human population has shown any protective effect of statins on cerebral aneurysmal rupture because of the low incidence of the rupture. In the setting of such circumstances, case–control study is the most efficient approach to human population in epidemiology. We conducted a multihospital case–control study in Japan to investigate an association between statin use and the rupture risk of cerebral aneurysms.

Participants and Methods Case and Control Selection In our study, the study end point was a rupture of aneurysms. We investigated the risk factors that might cause this end point using case–control design, which compared the distribution of the retrospective risk factors among cases and controls. Cases and corresponding controls of the study participants were recruited from 15 hospitals in Japan, which were located in Shiga (8 hospitals), Kyoto (5 hospitals), and Tokyo (2 hospitals). Cases, who had ruptured aneurysms, were selected by the following criteria: the patients who were aged 20 years and older with aneurysmal SAH hospitalized from April 1, 2009 to March 31, 2011. SAH was diagnosed by aneurysmal bleeding pattern, with the additional condition that the presence of one or more saccular aneurysms was confirmed by diagnostic imaging in the hospital. SAH patients caused by trauma, dissecting aneurysms, infected aneurysms, autoimmune diseases, and familial diseases were excluded from the case selection. Controls, who had unruptured aneurysms, were consecutively selected from the patients who have the same age range with newly diagnosed unruptured saccular aneurysms from April 1, 2006, to March 31, 2011. Patients who had the history of aneurysmal SAH were also excluded from cases and controls (5 cases [4.1%] and 9 controls [2.9%]) because unruptured aneurysms in patients with a history of SAH have higher rupture rate rather than those in patients without the history of SAH.

Data Collection Demographic characteristics (age, sex, height, and weight), information on cerebral aneurysms (aneurysm location, aneurysm diameter, and bleb), medical history (hypertension, diabetes mellitus, hyperlipidemia, heart disease, and previous stroke), family history of SAH, and behavioral factors (smoking and alcohol consumption) were collected from medical records in each hospital. The smoking status was categorized into 3 (never, past, and current) and the drinking status was categorized into 3 (never, past, and current). Prescription information on statin use and other drugs (antihypertensive drugs, aspirin or other antiplatelet drugs, vitamin K antagonists, nonsteroidal anti-inflammatory drugs, steroids, and other lipid-lowering drugs, including fibrate, bile acid–binding resins, niacin, ethyl icosapentate) were collected from doctors in each hospital and/or supplemented through medical record review. In each hospital, information about statin and other drugs was obtained from all patients at the time of admission by SAH or first visits. This study was approved by the ethics committee of Shiga University of Medical Science Approval number: (20-96). All participants gave written informed consent, including proxy respondents for case subjects who were severely ill, unconscious, or dead, in accordance with the criteria established by the local ethics committees at each participating center. From April 2009 to March 2011, 122 cases and 313 controls were collected from 15 hospitals.

Statistical Analysis The relationship between stain use and rupture risk of cerebral aneurysms were examined by logistic regression and chi-square test. Instead of using case–control matching, we applied a multivariable logistic regression to adjust potential confounders in the model. Statistical significance was set in .05, and all statistical analysis was performed by SPSS 19.0 (IBM, Chicago, IL).

Results A total of 117 cases and 304 controls were included in the analyses. Table 1 shows characteristics of the study participants. In the case group, age tended to be younger (P 5 .007) and the aneurysm size was larger (P , .001) than those of the control group. Significant differences between cases and controls were found in serum total cholesterol (P 5 .02), triglyceride (P 5 .01), current smoking (P , .001), use of statin (P , .001), and antihypertensive drugs (P , .001). Statin was used in 9.4% of cases and 26.0% of controls in any quantity by at least one prescription. No significant difference was observed in aspirin use (P 5 .39) and use of other lipid-lowering drugs (P 5 .79).

STATIN USE AND CEREBRAL ANEURYSM RUPTURE

345

Table 1. Characteristics of cases and controls, 2006-2011 Cases (n 5 117)

Age, y Aneurysm diameter, mm Total cholesterol, mg/dL LDL-C, mg/dL HDL-C, mg/dL Triglyceride, mg/dL

Controls (n 5 304)

Mean

SD

Mean

SD

60.46 6.4 186.12 113.24 56.7 109.34

13.77 3.40 49.89 40.34 18.45 65.90

64.34 5.1 198.71 115.03 59.88 132.48

11.33 3.20 36.02 29.25 16.69 80.98

Number

%

Number

%

41 11 13

35.0 9.4 11.1

104 79 40

34.2 26.0 13.2

58 3 33 5 1

49.6 2.6 28.2 4.3 .9

180 24 105 28 3

59.2 7.9 34.5 9.2 1.0

36 10 2

30.8 8.5 1.7

162 36 24

53.3 11.8 7.9

3 2 4

2.6 1.7 3.4

1 10 13

.3 3.3 4.3

67 13 37

57.3 11.1 31.6

177 83 41

58.2 27.3 13.5

58 3 56

49.6 2.6 47.9

138 13 150

45.4 4.3 49.3

Male Statin use Family history of SAH The concomitant diseases Hypertension* Diabetes mellitus Hyperlipidemia Heart disease Previous stroke (except SAH) Medications Antihypertensive drug use Aspirin Vitamin K antagonist, including warfarin, other antiplatelet drug Steroid Nonsteroidal anti-inflammatory drugs Use of other lipid-lowering drugs Cigarette smoking Never Past Current smoker Alcohol consumption Never Past Current drinking

Abbreviations: HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; SAH, subarachnoid hemorrhage. *Hypertension was defined as doctor-diagnosed hypertension and/or treatment with antihypertensive drugs.

Table 2 shows multivariable adjusted odds ratios of aneurysm rupture. The model includes 6 other potential confounders in addition to statin use. Use of any statin was inversely associated with cerebral aneurysm rupture independently and significantly (adjusted odds ratio: .30, 95% confidence interval [CI]: .14-.66). Even if we used low-density lipoprotein cholesterol as a potential confounder instead of serum total cholesterol, the similar result was obtained (adjusted odds ratio: .30, 95% CI: .12-.65). Higher serum total cholesterol was related to a significantly lower risk of rupture, independently from other confounders including statin use. Table 3 shows the results of their relationships stratified by serum total cholesterol level. Levels of serum total cholesterol were stratified into 3 categories: less than 130 mg/dL, 130-219 mg/dL, and 220 mg/dL or greater.

Of these categories, statin use was inversely associated with cerebral aneurysm rupture in participants with 130-219 mg/dL (adjusted odds ratio: .20, 95% CI: .07.56) and those with 220 mg/dL or greater (adjusted odds ratio: .32, 95% CI: .04-2.28).

Discussion To the best of our knowledge, this is the first case– control study that has examined the association between statin use and cerebral aneurysm rupture in clinical situations. Our results showed that an ever use of any statin was inversely associated with rupture risk in unruptured aneurysm patients. In the study, serum total cholesterol was significantly higher in the control group than those of case group, and

Y. YOSHIMURA ET AL.

346

Table 2. Adjusted odds ratios* of cerebral aneurysm rupture in 117 cases and 304 controls

Variables Statin use Yes No Sex Male Female Age (unit: 10 y) Hypertensiony Yes No Serum total cholesterol (unit: 10 mg/dL) Cigarette smoking Never Past Current Alcohol consumption Never Past Current

Adjusted odds ratios

.30 1 (reference)

assess in participants with less than 130 mg/dL because few patients with low cholesterol levels were included in this study. Statins are frequently prescribed in clinical practice to lower serum cholesterol for the prevention of stroke and cardiovascular diseases. In addition to their cholesterollowering effect, statins have an anti-inflammatory effect on vascular walls known as a pleiotropic effect.9 Several reports have showed that some statins are effective in experimental cerebral aneurysm models. Aoki et al6,7 reported that the administration of simvastatin and pitavastatin reduced the incidence of cerebral aneurysm in rats. On the contrary, Tada et al10 showed that pravastatin and simvastatin induced the growth of cerebral aneurysms in estrogen-deficient rats. They also reported the risk of rupture increased in high-dose pravastatin. The differences in the results of experimental aneurysmal models may depend on the differences in the models and doses used. Recently, Marbacher et al11 reported no significant relationship between statin use and incidence of cerebral aneurysms. They conducted a single-center case–control study and found no overall association between statin use and incidence of cerebral aneurysm formation even when dichotomized into hydrophilic and lipophilic user or between short (#12 months) and long ($36 months) duration of intake, although hypertension and smoking significantly increased the risk of cerebral aneurysm development. However, they did not examine the association between statin use and cerebral aneurysm rupture risk. Recently, it has been reported that statin use prevents clinical worsening in patients with abdominal aortic aneurysms, and the administration of statins in patients with small abdominal aortic aneurysms is recommended.12 Statins also have been the focus of controversy in relation to clinical practice regarding cerebral hemorrhage. In the Stroke Prevention by Aggressive Reduction

95% CI

.14-.66

1 (reference) .80 .88

.39-1.61 .69-1.10

.99 1 (reference) .93

.58-1.69 .87-.99

1 (reference) .33 1.72

.14-.78 .84-3.50

1 (reference) .46 .72

.09-2.42 .40-1.32

Abbreviation: CI, confidence interval. *Adjusted odds ratios were calculated using logistic regression analysis including all variables in the same model. yHypertension was defined as doctor-diagnosed hypertension and/or treatment with antihypertensive drugs.

there is a possibility that higher cholesterol levels are associated with decreased rupture risk. We conducted a stratified analysis by serum total cholesterol level to evaluate whether serum total cholesterol itself has any influence on cerebral aneurysm rupture. We observed that statin use was inversely associated with cerebral aneurysm rupture regardless of serum cholesterol level in participants with 130-219 mg/dL and 220 mg/dL or more. We cannot

Table 3. Adjusted odds ratios* of cerebral aneurysm rupture for statin use, stratified by serum total cholesterol levels Case Serum total cholesterol Statin level use Number % ,130 mg/dL

130-219 mg/dL

$220 mg/dL

Yes No Total Yes No Total Yes No Total

2 13 15 5 60 65 2 23 25

13.3 86.7 100.0 7.7 92.3 100.0 8.0 92.0 100.0

Control Number

%

0 4 4 50 126 176 10 49 59

.0 100.0 100.0 28.4 71.6 100.0 16.9 83.1 100.0

Crude odds ratio 95% CI Adjusted odds ratio 95% CI — —

— —

— —

— —

.21 1 (reference)

.08-.55 —

.20 1 (reference)

.07-.56 —

.43 1 (reference)

.09-2.10 —

.32 1 (reference)

.04-2.28 —

Abbreviation: CI, confidence interval. *Logistic regression analysis was used to estimate odds ratios adjusted for sex, age, hypertension, smoking, and alcohol consumption.

STATIN USE AND CEREBRAL ANEURYSM RUPTURE

in Cholesterol Levels trial, there was a 16% reduction in stroke outcomes favoring atorvastatin therapy.13 However, in a post hoc analysis, there was an excess risk of hemorrhagic stroke (hazard ratio, 1.66) associated with atorvastatin administration.14 A recent systematic review of cohort studies suggested no association between statin administration and intracerebral hemorrhage.15 There has been no article that shows statin use may influence cerebral aneurysm rupture. Unlike the arteriosclerosis or the aortic aneurysm, hyperlipidemia is not a risk factor for rupture of cerebral aneurysms. Several reports have shown that hyperlipidemia appears to reduce the risk of SAH.16,17 Our study also showed that higher cholesterol levels are protective against cerebral aneurysm rupture and that statins appears to be protective against cerebral aneurysm rupture in higher and normal cholesterol-level groups, although the effect of statins in lower cholesterol group is unknown. It is not proved whether protective effects shown by this study are based on the action of cholesterol-lowering effect or pleiotropic effect. Moreover, proper doses of statins to prevent cerebral aneurysm rupture in each level of cholesterol are not determined yet. We must be careful when statin treatment is applied in patients with cerebral aneurysms considering the possible risk of cerebral hemorrhage, and further studies are mandatory before determining mechanisms of protective effects of statins on cerebral aneurysm and proper doses of statins to prevent cerebral aneurysm rupture. There are several limitations of this study. First, as usually occurred in case–control studies, our results may be influenced by selection bias and recall bias. We have made intensive efforts to avoid bias in every study aspect. In the design stage, we decided to select controls and cases from the same hospital to prevent selection bias. In the analytic stage, we have performed the multivariable analysis for adjusting confounding. Though a case– control study is known to suffer a various kind of bias, our considerable effort could minimize the magnitude of bias as far as possible. It is a well-known fact that the controls of hospitalbased case–control study have different characteristics compared with those of general population. This might cause the issue of selection bias that amplified the value of odds ratios. Second, we did not analyze serum lipids before and after statin administration to verify the effects of statins. In this case–control study, total serum cholesterol measurement was not available before statin administration, and we could not directly analyze the relationship between cholesterol-lowering effects of statin use and the rupture of cerebral aneurysm. Because pleiotropic effects of statins without lowering serum cholesterol are supposed to modify inflammatory process,9 the values of serum cholesterol may not exert as significant factors to reduce the risk of SAH. In fact, regardless of serum cholesterol levels,

347

statin use was inversely associated with the occurrence of aneurysmal rupture in this study. Third, we planned unmatched case–control study with multivariable logistic regression. Epidemiologists often mentioned that matching will introduce selection bias when a matched variable is related to exposure.18 To avoid this bias, we decided to use multivariate logistic regression to adjust an imbalance between cases and controls in the study. Though we included the several variables to adjust confoundings, there might be the effect of residual confounding in odds ratios. For example, there is a possibility of residual confounding by age. Because controls are older than cases on average, the control group would have risk factors for cardiovascular disease more often than cases. Actually, controls also seem to have more atherosclerotic factors that may be a cause of occurrence of unruptured aneurysms or more use of statins. Finally and as the important limitation, there is a possibility that patients in case group might not consult a medical institution because of no medical check or no chance of statin use compared with patients in the control group who might be more interested in their health. Participants in the control group could take other drugs such as antihypertensive drugs, anti-inflammatory drugs, and other lipid-lowering drugs more frequently than the case group. This may be the case in Japan, and we carefully conducted statistical analysis and found that there were no significant differences between the 2 groups except statins. Interestingly, a recent article showed that frequent aspirin use might have a protective effect for risk of intracranial aneurysm rupture in a case–control study using a prospective untreated cohort of the International Study of Unruptured Intracranial Aneurysms.19 The differences in patients’ background, dosage of aspirin, and study designs may cause the discrepancy between this study and ours. Although we showed a possible association between statin use and the phenomenon of aneurysmal rupture in a hospital-based case–control study, we did not prove direct effects of statin use on patients with cerebral aneurysms, and it remains uncertain whether statins are beneficial for the prevention of SAH in human population and whether there are any differences in the effect of statin use by the kinds and the dosage of statins. Further clinical studies are needed to prove the protective effects of medical treatments for cerebral aneurysms. In conclusion, the findings from this study provide for the first time an epidemiological evidence of an association of statin use with incidence of cerebral aneurysm rupture. After adjustments for confounders, the inverse relationship remained between use of statin and cerebral aneurysmal rupture. Further and larger studies, particularly prospective randomized trial studies, are necessary to confirm our findings and safety of statins in SAH prevention and treatment of cerebral aneurysms.

348 Acknowledgment: List of participating institutes (alphabetically ordered): Kohka Public Hospital (Kazuyoshi Watanabe), Koto Memorial Hospital (Hisao Hirai), Kusatsu General Hospital (Kenichi Matsumura), Kyoto Kizugawa Hospital (Tatshuhito Yamagami), Kyorin University Hospital (Yoshiaki Shiokawa), NTT Medical Center Tokyo (Akio Morita), Omihachiman Community Medical Center (Masayuki Nakajima), Saiseikai Shigaken Hospital (Akihiko Hino), Second Okamoto General Hospital (Minoru Kidooka), Shiga University of Medical Science Hospital (Kazuhiko Nozaki), Shimizu Hospital (Yukio Shimizu), Soseikai General Hospital (Kazumitsu Kyoushima), Takashima Municipal Hospital (Masaharu Ichikawa), Tanabe Central Hospital (Ryouji Kimura), and Yasu Hospital (Hirohiko Kizuki).

References 1. Vlak MH, Algra A, Brandenburg R, et al. Prevalence of unruptured intracranial aneurysms, with emphasis on sex, age, comorbidity, country, and time period: asystematic review and meta-analysis. Lancet Neurol 2011; 10:626-636. 2. Nieuwkamp DJ, Setz LE, Algra A, et al. Changes in case fatality of aneurysmal subarachnoid haemorrhage over time, according to age, sex, and region: a meta-analysis. Lancet Neurol 2009;8:635-642. 3. Kita Y, Okayama A, Ueshima H, et al. Stroke incidence and case fatality in Shiga, Japan 1989-1993. Int J Epidemiol 1999;28:1059-1065. 4. Wermer MJ, van der Schaaf IC, Algra A, et al. Risk of rupture of unruptured intracranial aneurysms in relation to patient and aneurysm characteristics: an updated metaanalysis. Stroke 2007;38:1404-1410. 5. Wiebers DO, Whisnant JP, Huston J III, et al. Unruptured intracranial aneurysms: natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet 2003;362:103-110. 6. Aoki T, Kataoka H, Ishibashi R, et al. Simvastatin suppresses the progression of experimentally induced cerebral aneurysms in rats. Stroke 2008;39:1276-1285.

Y. YOSHIMURA ET AL. 7. Aoki T, Kataoka H, Ishibashi R, et al. Pitavastatin suppresses the formation and progression of cerebral aneurysms through the inhibition of nuclear factor kappa B pathway. Neurosurgery 2009;64:357-366. 8. Kimura N, Shimizu H, Eldawoody H, et al. Effect of olmesartan and pravastatin on experimental cerebral aneurysms in rats. Brain Res 2010;1322:144-152. 9. Zhou Q, Liao JK. Pleiotropic effects of statins. -Basic research and clinical perspectives. Circ J 2010;74:818-826. 10. Tada Y, Kitazato KT, Yagi K, et al. Statins promote the growth of experimentally induced cerebral aneurysms in estrogen-deficient rats. Stroke 2011;42:2286-2293. 11. Marbacher S, Schl€ appi JA, Fung C, et al. Do statins reduce the risk of aneurysm development: a case-control study. J Neurosurg 2012;116:638-642. 12. Baxter BT, Terrin MC, Dalman RL. Medical management of small abdominal aortic aneurysms. Circulation 2008; 117:1883-1889. 13. Amarenco P, Bogousslavsky J, Callahan A III, et al. Highdose atorvastatin after stroke or transient ischemic attack. N Engl J Med 2006;355:549-559. 14. Goldstein LB, Amarenco P, Szarek M, et al. Hemorrhagic stroke in the Stroke Prevention by Aggressive Reduction in Cholesterol Levels study. Neurology 2008;70: 2364-2370. 15. Hackam DG, Woodward M, Newby LK, et al. Statins and intracerebral hemorrhage: collaborative systematic review and meta-analysis. Circulation 2011; 124:2233-2242. 16. Feigin VL, Rinkel GJE, Lawes CMM, et al. Risk factors for subarachnoid hemorrhage: an updated systematic review of epidemiological studies. Stroke 2005;36: 2773-2780. 17. Feigin V, Parag V, Lawes CMM, et al. Smoking and elevated blood pressure are the most important risk factors for subarachnoid hemorrhage in the Asia-Pacific region: an overview of 26 cohorts involving 306 620 participants. Stroke 2005;36:1360-1365. 18. Rothman K. Epidemiology-An Introduction: Oxford University Press, 2012. 129-133. 19. Hasan DM, Mahaney KB, Brown RD Jr, et al. Aspirin as a promising agent for decreasing incidence of cerebral aneurysm rupture. Stroke 2011;42:3156-3162.