JNS-12880; No of Pages 6 Journal of the Neurological Sciences xxx (2013) xxx–xxx
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis Peng-Peng Niu a, Ge Yang b, Ying-Qi Xing a, Zhen-Ni Guo a, Yi Yang a,⁎ a b
Department of Neurology, The First Hospital of Jilin University, Xinmin Street 71#, Changchun 130021, China Department of Geriatrics Medicine, The Affiliated Hospital to Changchun University of Chinese Medicine, Changchun, China
a r t i c l e
i n f o
Article history: Received 30 August 2013 Received in revised form 11 October 2013 Accepted 18 October 2013 Available online xxxx Keywords: Aneurysm Cilostazol Meta-analysis Subarachnoid hemorrhage Systematic review Vasospasm
a b s t r a c t Previous studies with small sample size have shown that cilostazol can reduce the risk of cerebral vasospasm in patients with aneurysmal subarachnoid hemorrhage (SAH). The purpose of this study was to determine whether cilostazol is effective in patients with aneurysmal SAH. Studies investigating the effect of cilostazol in patients with aneurysmal SAH were identified using Embase.com without language or publication-type restrictions. We used the random-effect model to combine data. Pooled risk ratios (RR) and 95% confidence intervals (CI) were calculated. Two randomized controlled trials and two quasi-randomized controlled trials with a total of 340 patients were included. The incidence of symptomatic vasospasm (RR = 0.47; 95% CI, 0.31–0.72; p b 0.001), severe vasospasm (RR = 0.48; 95% CI, 0.28–0.82; p = 0.007), vasospasm-related new cerebral infarctions (RR = 0.38; 95% CI, 0.22–0.67; p = 0.001), and poor outcome (RR = 0.57; 95% CI, 0.37–0.88; p = 0.011) were significantly lower in the cilostazol group. The numbers needed to treat for these outcomes were 6.4, 6.3, 5.7, and 5.4, respectively. Mortality rate differences between the two groups were insignificant. No statistical heterogeneity was found for all outcomes. These results show that cilostazol can decrease the incidence of symptomatic vasospasm, severe vasospasm, vasospasm-related new cerebral infarctions, and poor outcome in patients with aneurysmal SAH. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Subarachnoid hemorrhage (SAH) is a stroke subtype. Although it accounts for only 5% of all strokes, it generally occurs at a fairly young age with a relative high rate of morbidity and mortality [1]. Cerebral vasospasm and vasospasm-related cerebral infarction are common and serious complications in patients with aneurysmal SAH, and are important causes of death and dependency [2,3]. A Cochrane review showed that oral administration of nimodipine likely reduces the risk of cerebral ischemia and poor outcome after aneurysmal SAH, but the evidence for other calcium antagonists is inconclusive [4]. An updated systematic review that investigated the effects of statins in patients with aneurysmal SAH also showed no significant results [5]. Observational studies suggested that antiplatelet agents may reduce the risk of secondary ischemia and poor outcome after aneurysmal SAH [6], but randomized controlled trials (RCTs) and the Cochrane review failed to show any significant results [7]. However, studies that investigated the effect of cilostazol on aneurysmal SAH were not included in this Cochrane review. Cilostazol, a selective inhibitor of phosphodiesterase 3, is an antiplatelet agent marketed in Japan that is used to treat the ischemic symptoms
⁎ Corresponding author. Tel.: +86 15943020037; fax: +86 43188782378. E-mail addresses: [email protected], [email protected] (Y. Yang).
of peripheral vascular disease. Unlike other antiplatelet agents, cilostazol not only inhibits the platelet aggregation, but also has effects of direct vasodilatation [8] and anti-inflammation [9]. Moreover, cilostazol is more effective than aspirin in the secondary prevention of ischemic stroke without increasing the risk of bleeding complications [10]. Both animal studies and clinical studies have shown that cilostazol can reduce the risk of cerebral vasospasm in patients with aneurysmal SAH [9,11]. However, the sample sizes of previous studies were small, and there has been no relevant systematic review regarding this topic. We performed a systematic literature review of all clinical controlled trials to evaluate the effect of cilostazol in patients with aneurysmal SAH.
2. Methods 2.1. Search strategy Studies that investigated the effect of cilostazol in aneurysmal SAH were identified until August 10, 2013 through electronic searches in Embase.com, which includes records from Medline and Embase, without language or publication-type restrictions. Search results were limited to human. The search strategy (Table 1S, Supplementary materials) combined the terms “cilostazol” and “subarachnoid hemorrhage.” Two authors (P.P.N. and Y.Y.) independently performed the literature search. The references of identified articles were screened to find further studies.
0022-510X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jns.2013.10.027
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
2
P.-P. Niu et al. / Journal of the Neurological Sciences xxx (2013) xxx–xxx
2.2. Study selection and eligibility criteria Two authors (P.P.N. and G.Y.) independently screened the articles that were retrieved using the search strategy. Articles that failed to meet our inclusion criteria were excluded after reading the title, keywords, and abstract. Articles that were not excluded in the initial screen were evaluated by reading the full text. For duplicate articles, we included those with the largest sample size or the most complete information. Any disagreements were resolved by discussion. Inclusion criteria for the studies were the following: (1) randomized controlled clinical trial or controlled clinical trial; (2) study that investigated the effect of cilostazol in patients with aneurysmal SAH; (3) study that reported the incidence of vasospasm, the incidence of cerebral infarction, or functional outcome of patients; and (4) rational diagnosis methods for aneurysmal SAH and cerebral infarction were described. 2.3. Data collection Data from the articles included in this study were extracted by two independent authors (Y.Q.X. and Z.N.G.) using a standardized form. The following information were extracted from each study: name of the first author, publication year, study period, country or geographical origin of investigation, baseline characteristics of patients, number of patient in each group, and outcome data in each group. All disagreements were resolved by consensus. 2.4. Assessment of risk of bias for the included studies Two authors (G.Y. and Y.Q.X.) independently assessed the internal validity of the included studies by using the criteria list described by the Editorial Board of the Cochrane Back Review Group [12]. Twelve criteria that reflect the risk of bias were included in this appraisal system. Each criterion can be scored as yes, unclear, or no, where yes indicates the criterion has been met and therefore suggests a low risk of bias. Studies that meet at least 6 of the 12 criteria and have no serious flaws (e.g., 80% dropout rate in one group) can be defined as having a “low risk of bias.” 2.5. Outcome measures The primary outcome was the incidence of symptomatic vasospasm. Secondary outcomes were the incidence of severe vasospasm, incidence of vasospasm-related new cerebral infarctions, incidence of poor outcome, and mortality. The definitions of the outcome measures were made according to the descriptions provided in the included studies. Symptomatic vasospasm was defined as any new unexplainable focal or global neurological deficit or a decrease of at least 2 points on the Glasgow Coma Scale (not associated with rebleeding, intracerebral hematoma, hydrocephalus, metabolic disturbances, fever, or infection), regardless of cerebral vasospasm. Vasospasm-related new cerebral infarction was defined as new follow-up computerized tomography (CT) or magnetic resonance imaging (MRI) lesions not attributable to other causes, regardless of any association with symptoms. Severe vasospasm was defined as N 50% decrease in vessel diameter detected by digital subtraction angiography (DSA), computerized tomography angiography (CTA), or magnetic resonance angiography (MRA). Poor functional outcome was defined as a modified Rankin Scale (mRS) of 3 to 6 determined during the follow-up in the study. 2.6. Statistical analysis STATA, version 12.0 (Stata Corporation, College Station, Texas) was used for all statistical analyses. The risk ratio (RR) and 95% confidence interval (CI) were used as the measure of effect of interest. We measured statistical heterogeneity using the Q statistic and I2 statistic.
Heterogeneity existed if the Q statistic was pQ b 0.1 [13,14]. I2 statistic is expressed as a value between 0% and 100% that represents heterogeneity as absent to extreme. Because only a few studies with low numbers of patients were included, the DerSimonian–Laird random-effect model was used in all analyses [5]. The number needed to treat (NNT) for each outcome was calculated. Subgroup analyses by different study designs were performed. Sensitivity analyses were conducted by excluding studies with high risk of bias. 3. Results 3.1. Studies included in the meta-analysis According to our search strategies, 222 articles were initially identified. Ultimately, three articles written in English [11,15,16] and one article written in Japanese [17] with a total of 340 patients were included in this analysis. Two studies were RCTs [11,16] and two were quasi-randomized controlled trials (q-RCTs) [15,17]. The selected studies were published between 2009 and 2013. There were no significant differences in any of the characteristics between the two groups in the studies. Postoperative management was also uniform between the two groups for each study. The baseline characteristics of all qualified studies (Table 1), outcome data (Table 2), and main results of this metaanalysis (Table 3, Fig. 1) are presented. The risk of bias is also illustrated (Table 2S, Supplementary materials). 3.2. Assessment of risk of bias for the included studies The results for the risk of bias assessment are presented in Table 2S. The four included studies were judged to have low risk of bias because ≥6 items were met for each of them. However, the two q-RCTs just met 6 of the 12 items. In the study by Senbokuya et al., six patients (11.11%) of the cilostazol group dropped out because of adverse events. However, the authors included these six patients in the cilostazol group analyses. 3.3. Quantitative data synthesis 3.3.1. Symptomatic vasospasm All studies reported the incidence of symptomatic vasospasm. Using a random-effect model to combine data, the number of patients who had symptomatic vasospasm was significantly higher in the control group (55/186, 29.57%) than in the cilostazol group (23/154, 14.94%) (RR = 0.47, 95% CI = 0.31–0.72, p b 0.001), with no statistical heterogeneity showed by the Q statistic (p = 0.586) and I2 statistic (0.0%). The NNT for this outcome is 6.4 (95% CI = 4.9–12.1). The significant difference was essentially unchanged after excluding the two q-RCTs (RR = 0.46, 95% CI = 0.25–0.84, p = 0.012). No statistical heterogeneity was found between the two RCTs (pQ = 0.216, I2 = 34.5%). 3.3.2. Severe vasospasm Yoshimoto et al. reported the incidence of severe vasospasm between 7 and 9 days (or when any neurological deterioration was detected) after the onset of SAH detected by cerebral angiography. Senbokuya N et al. and Murahashi et al. reported the incidence of severe vasospasm between 7 and 10 days after the onset of SAH detected by CTA or DSA. All of them used the definition of N 50% decrease in vessel diameter for severe vasospasm. Using a random-effect model to combine data, the number of patients who had severe vasospasm was significantly higher in the control group (41/135, 30.37%) (RR = 0.48, 95% CI = 0.28–0.82, p = 0.007) than in the cilostazol group (16/105, 15.24%). The NNT for this outcome is 6.3 (95% CI = 4.6–18.3). Data on severe vasospasm did not indicate statistical heterogeneity (pQ = 0.347, I2 = 5.5%). The RR was essentially unchanged after excluding the two q-RCTs (RR = 0.42, 95% CI = 0.22–0.80, p = 0.008).
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
P.-P. Niu et al. / Journal of the Neurological Sciences xxx (2013) xxx–xxx
3
Table 1 Characteristics of included studies. Cilostazol group Study
Study period Study Study No. (F/M) region design
Control group
Age, year
Surgery type No. (F/M) (clip/coil)
Age, year
Surgery type Time of surgery (clip/coil)
60
22/4
24 (14/10) 58
19/5
Yoshimoto T 2004–2006 et al. (2009)
Japan
q-RCT
26 (23/3)
Suzuki S 2006–2008 et al. (2011) Senbokuya N 2009–2010 et al. (2013)
Japan
RCT
49 (36/13) 62 ± 13
49/0
51(40/11) 64 ± 14
51/0
Japan
RCT
54 (35/19) 60.0 ± 12.5
54/0
55 (33/22) 61.3 ± 12.7
55/0
Murahashi T 2008–2009 et al. (2013)
Japan
q-RCT
25 (21/4)
64.2 (44–82) 25/0
56 (37/19) 60.6 (35–86) 56/0
Intervention of cilostazol group
Within a couple of 200 mg/d from day days after SAH onset one after surgery for 2 weeks Within 72 h 100 mg twice per day after SAH onset for 14 days after SAH Within 72 h 100 mg twice per day after SAH onset for 14 days from 48 hours after surgery Within 72 h 200 mg/d from day after SAH onset one after surgery for 2 or 4 weeks
RCT indicates randomized controlled trial; q-RCT, quasi-randomized controlled trial.
3.3.3. New cerebral infarctions related to cerebral vasospasm Two RCTs and one q-RCT reported the incidence of new cerebral infarctions related to cerebral vasospasm. Using a random-effect model to combine data, the number of patients who had vasospasmrelated new cerebral infarctions was significantly higher in the control group (37/130, 28.46%) than in the cilostazol group (14/129, 10.85%) (RR = 0.38, 95% CI = 0.22–0.67, p = 0.001), with no statistical heterogeneity showed by the Q statistic (p = 0.997) and I2 statistic (0.0%). The NNT for this outcome is 5.7 (95% CI = 4.5–10.6). The significant difference was essentially unchanged (RR = 0.38, 95% CI = 0.20–0.71, p = 0.003) after excluding the one q-RCT. No statistical heterogeneity was found between the two RCTs (pQ = 0.967, I2 = 0.0%). 3.3.4. Poor outcome Three studies reported functional outcome. Yoshimoto et al. reported the mRS score at discharge or one month after SAH. Suzuki et al. reported the mRS score at discharge (the mean hospital stay was 40 days). Senbokuya et al. reported the functional outcome using the mRS score at 1, 3, and 6 months after SAH. In order to ensure that the time point is consistent across studies, we used the 1-month data from the Senbokuya et al. study for our analysis. Using a random-effect model to combine the data, the number of patients who had poor outcome (mRS ≥ 3) was significantly higher in the control group (56/130, 43.08%) than in the cilostazol group (31/129, 24.03%) (RR = 0.57, 95% CI = 0.37–0.88, p = 0.011). Data on poor outcome did not indicate statistical heterogeneity (pQ = 0.276, I2 = 22.2%). The NNT for this outcome is 5.4 (95% CI = 3.7–19.3). The significant difference was essentially unchanged (RR = 0.53, 95% CI = 0.30–0.94, p = 0.030, pQ = 0.155) after excluding the one q-RCT. 3.3.5. Mortality Three studies reported the mortality rates. No statistical differences were found between the cilostazol (3/129, 2.33%) and control
groups (5/130, 3.85%) (RR = 0.64, 95% CI = 0.15–2.76, p = 0.552, pQ = 0.472).
4. Discussion This meta-analysis showed that cilostazol significantly decreases the incidence of symptomatic vasospasm, severe vasospasm, vasospasmrelated new cerebral infarctions, and poor outcome in patients with aneurysmal SAH. Subgroup analysis by pooling data from the two RCTs also showed similar results. However, no significant differences in mortality rate were found between the cilostazol and control groups. As reported by Ito et al. in 2008, cilostazol can significantly attenuate delayed cerebral vasospasm partially by suppressing lipid peroxidation in SAH rats [18]. Subsequent animal studies further confirmed the effect of cilostazol on cerebral vasospasm after SAH [9,19]. Moreover, both controlled clinical trial [15] and randomized controlled clinical trial [11] confirmed the effect of cilostazol on attenuated cerebral vasospasm after SAH. The multifaceted function of cilostazol in reducing reactive oxygen species generation, suppressing apoptotic death, inducing nitric oxide production, anti-inflammation, and anti-lipid peroxidation may account for its effect on attenuated cerebral vasospasm [9,18,20]. Because cilostazol could not suppress all vasospasm, patients with poor outcome also exist in the cilostazol group. Moreover, not all infarctions and poor outcomes were caused by vasospasm. Patients with quite modest vasospasm also could develop symptomatic cerebral infarction [21], which could lead to poor outcome. Factors such as distal microcirculatory failure, poor collateral anatomy, and genetic or physiological variations in cellular ischemic tolerance likely also contribute to the development of delayed cerebral infarction and poor outcome [22]. In one of the studies included in this review [16], the authors reported that cilostazol could improve the outcome of patients with aneurysmal SAH without significantly reducing the incidence of symptomatic vasospasm and the incidence of cerebral infarction, suggesting that
Table 2 Number of events for each outcome measure. Number of events for each outcome (cilostazol group/control group) Study
Sample size (cilostazol group/control group)
Symptomatic vasospasm
Severe vasospasm
Vasospasm-related new cerebral infarctions
Poor outcome
Death
Yoshimoto T et al. (2009) Suzuki S et al. (2011) Senbokuya N et al. (2013) Murahashi T et al. (2013)
26/24 49/51 54/55 25/56
5/9 11/19 7/22 0/5
2/7 NA/NA 10/24 4/10
3/7 5/14 6/16 NA/NA
6/7 10/27 15/22 NA/NA
0/0 1/3 2/2 NA/NA
NA indicates not applicable.
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
4
P.-P. Niu et al. / Journal of the Neurological Sciences xxx (2013) xxx–xxx
Table 3 Estimate of RR for each outcome measure. RR (95% CI); p; number of studies; pQ; I2 Group
Symptomatic vasospasm
Severe vasospasm
Vasospasm-related new cerebral infarctions
Poor outcome
Death
All studies
0.47 (0.31, 0.72); b0.001; 4; 0.586; 0.0% 0.46 (0.25, 0.84); 0.012; 2; 0.216; 34.5% 0.47 (0.19, 1.14); 0.096; 2; 0.519; 0.0%
0.48 (0.28, 0.82); 0.007; 3; 0.347; 5.5% 0.42 (0.22, 0.80); 0.008;1; –; – 0.54 (0.17, 1.78); 0.007; 2; 0.183; 43.5%
0.38 (0.22, 0.67); 0.001; 3; 0.997; 0.0% 0.38 (0.20, 0.71); 0.003; 2; 0.967; 0.0% 0.40 (0.12, 1.36); 0.141;1; –; –
0.57 (0.37, 0.88); 0.011; 3; 0.276; 22.2% 0.53 (0.30, 0.94); 0.030; 2; 0.155; 50.5% 0.79 (0.31, 2.02); 0.625; 1;-; -
0.64 (0.15, 2.76); 0.552; 3; 0.472; 0.0% 0.64 (0.15, 2.76); 0.552; 2; 0.472; 0.0% –⁎
RCTs q-RCTs
⁎ Yoshimoto et al. reported that there was no any case of death in both groups.
cilostazol may improve outcomes via other mechanisms in addition to anti-arteriospasm and antiplatelet effects. Other effects of cilostazol, including improving microcirculation [23] and preventing endothelial damage [20] may contribute to the improved outcome of patients with aneurysmal SAH. Regarding the definition of outcome measures, all four studies reported the incidence of symptomatic vasospasm, and three studies reported the incidence of vasospasm-related cerebral infarction. However, the definitions of symptomatic vasospasm and vasospasmrelated new cerebral infarction were not the same among the included studies. The definition of symptomatic vasospasm in the
two RCTs was similar to what we described above. However, for the two q-RCTs, the authors provided the incidence of symptomatic vasospasm and listed all symptoms with no clear definition. For vasospasm-related infarction, only two studies [11,15] mentioned their definition that was similar to our description. Suzuki et al. [16] stated that vasospasm-related infarction was confirmed by head CT scan or MRI. The lack of consistent definitions and diagnostic methods among studies is a limitation of our study. However, even considering this limitation, we believe that the combined data analyses of the outcome measures still have some significance.
Fig. 1. RR estimates with the corresponding 95% CI for each outcome measure. (A, symptomatic vasospasm. B, severe vasospasm. C, vasospasm-related new cerebral infarctions. D, poor outcome. RCT indicates randomized controlled trial; q-RCT, quasi-randomized controlled trial; CI, confidence interval; and RR, risk ratio.)
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
P.-P. Niu et al. / Journal of the Neurological Sciences xxx (2013) xxx–xxx
In this study, we chose the incidence of symptomatic vasospasm as the primary outcome rather than the incidence of cerebral vasospasm because obvious variations existed in the definition of cerebral vasospasm and cerebral vasospasm data were lacking. For example, Senbokuya et al. [11] defined moderate angiographic vasospasm as a 25%–50% reduction in vessel diameter, which differs from the definition of b50% decrease in vessel diameter reported by Murahashi et al. [17]. Because there were three studies that reported the incidence of severe angiographic vasospasm with the same definition (greater than 50% decrease in vessel diameter), a meta-analysis using the random-effect model to combine these data was performed. The results showed that cilostazol can significantly attenuate severe cerebral vasospasm (RR =0.48, 95% CI = 0.28– 0.82, p = 0.007). Data on severe vasospasm did not indicate statistical heterogeneity (pQ = 0.347, I2 = 5.5%). Moreover, the result was essentially unchanged after excluding the two q-RCTs. Regarding the actual incidence of vasospasm and cerebral infarction, the data were reported in only one study for each of them. Yoshimoto et al. [15] reported that there were no significant differences in the incidence of angiographic vasospasm between the cilostazol group (11/26, 42.3%) and the control group (14/24, 58.3%). Senbokuya et al. [11] reported that there were no significant differences in the actual incidence of cerebral infarction between the cilostazol group (11/54, 20.4%) and the control group (21/55, 38.2%) (p = 0.057). Because the sample sizes were small, future studies with larger sample sizes may produce significant results. We used the criteria list described by the Editorial Board of the Cochrane Back Review Group to assess the risk of bias of the included studies. Although the four studies did not show high risk of bias because they met 6, 9, 10, and 6 of the 12 criteria items, respectively, the two qRCTs also might have a certain risk of bias because they did not use proper methods of allocation or any blind methods. However, for the parameters that analyzed in our study, the significant and non-significant results remained essentially unchanged after excluding the two q-RCTs. Indeed, RCT is the most powerful type of experimental study because the proper randomization procedure reduces the risk of serious imbalance in known and unknown factors that could influence the clinical course of patients [24]. However, only two included studies used the proper methods of allocation. In the other two q-RCTs included in this study, subjects in the cilostazol and control groups were recruited from two different centers for one study and during two different periods in the other study. Even the methods of allocation of these two q-RCTs were inadequate, the baseline characteristics of the patients showed no significant differences between the two groups for each study, which just meet the main purpose of random allocation. Although it is insufficient, the similar baseline characteristics between groups may partly compensate for the risk of selection bias caused by improper methods of allocation. The adverse effects of cilostazol were mentioned in one of the studies that were included in this study. The authors [11] reported that there were three hemorrhagic events (gastrointestinal hemorrhage, epidural hematoma, and intracerebral hemorrhage) and three cardiac events (sinus tachycardia in one patient and paroxysmal atrial fibrillation in two patients) in the cilostazol group. Even if these events were related to cilostazol, there were no significant differences in the incidence of adverse events between the two groups (6/54 vs. 4/55), and each of patients improved after discontinuing cilostazol. A previous Cochrane review investigated the effect of antiplatelet therapy on aneurysmal SAH showed that there were no obvious side effects of antiplatelet agents (including aspirin, cataclot, and ticlopidine) [7]. Furthermore, in a more recent meta-analysis [25], data from four RCTs showed that cilostazol is significantly better than aspirin in the secondary prevention of ischemic stroke with significantly fewer adverse events. All of these suggested that cilostazol is safe for patients with aneurysmal SAH. All in all, the results of the present systematic review should be interpreted with caution. Although there is a relative low risk of clinical heterogeneity because the dose, start time, duration of cilostazol, the
5
study population, and time of surgery were the same or similar among groups and studies, the total number of studies and the total number of patients included in this review are only 4 and 340 respectively. And there were two q-RCTs included with relative high risk of bias because they did not use proper methods of allocation or any blind methods. Furthermore, all studies were conducted in the Japanese population, which makes the results may not be generalized beyond the Japanese population [25]. In summary, this systematic review showed that the use of cilostazol can decrease the incidence of symptomatic vasospasm, severe vasospasm, vasospasm-related new cerebral infarctions, and poor outcome in patients with aneurysmal SAH with no obvious side effects, and thus supporting the use of cilostazol in patients with aneurysmal SAH. However, further larger randomized, controlled trials are needed to confirm these conclusions, particularly for non-Japanese populations. Moreover, because all the four studies had the same or similar usage of cilostazol and most of the patients included in this review were treated with clipping, future studies are needed to determine the optimal timing, dosage, duration, and safety of cilostazol treatment, as well as to evaluate the effect of cilostazol on patients treated with coiling. In addition, future studies should determine the effects of cilostazol on other outcome measures and investigate the protective mechanisms of cilostazol. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jns.2013.10.027. Conflicts of interest The authors declare that they have no conflict of interest. Acknowledgments None. References [1] van Gijn J, Kerr RS, Rinkel GJ. Subarachnoid haemorrhage. Lancet 2007;369(9558): 306–18. [2] Roos YB, de Haan RJ, Beenen LF, Groen RJ, Albrecht KW, Vermeulen M. Complications and outcome in patients with aneurysmal subarachnoid haemorrhage: a prospective hospital based cohort study in the Netherlands. J Neurol Neurosurg Psychiatry 2000;68(3):337–41. [3] Rabinstein AA, Friedman JA, Weigand SD, McClelland RL, Fulgham JR, Manno EM, et al. Predictors of cerebral infarction in aneurysmal subarachnoid hemorrhage. Stroke 2004;35(8):1862–6. [4] Dorhout Mees SM, Rinkel GJ, Feigin VL, Algra A, van den Bergh WM, Vermeulen M, et al. Calcium antagonists for aneurysmal subarachnoid haemorrhage. Cochrane Database Syst Rev 2007;3:CD000277. [5] Vergouwen MD, de Haan RJ, Vermeulen M, Roos YB. Effect of statin treatment on vasospasm, delayed cerebral ischemia, and functional outcome in patients with aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis update. Stroke 2010;41(1):e47–52. [6] Toussaint 3rd LG, Friedman JA, Wijdicks EF, Piepgras DG, Pichelmann MA, McIver JI, et al. Influence of aspirin on outcome following aneurysmal subarachnoid hemorrhage. J Neurosurg 2004;101(6):921–5. [7] Dorhout Mees SM, van den Bergh WM, Algra A, Rinkel GJ. Antiplatelet therapy for aneurysmal subarachnoid haemorrhage. Cochrane Database Syst Rev 2007;4:CD006184. [8] Goto S. Cilostazol: potential mechanism of action for antithrombotic effects accompanied by a low rate of bleeding. Atheroscler Suppl 2005;6(4):3–11. [9] Nishino A, Umegaki M, Fujinaka T, Yoshimine T. Cilostazol attenuates cerebral vasospasm after experimental subarachnoid hemorrhage. Neurol Res 2010;32(8): 873–8. [10] Kamal AK, Naqvi I, Husain MR, Khealani BA. Cilostazol versus aspirin for secondary prevention of vascular events after stroke of arterial origin. Cochrane Database Syst Rev 2011;1:CD008076. [11] Senbokuya N, Kinouchi H, Kanemaru K, Ohashi Y, Fukamachi A, Yagi S, et al. Effects of cilostazol on cerebral vasospasm after aneurysmal subarachnoid hemorrhage: a multicenter prospective, randomized, open-label blinded end point trial. J Neurosurg 2013; 118(1):121–30. [12] Furlan AD, Pennick V, Bombardier C, van Tulder M, Editorial Board CBRG. 2009 updated method guidelines for systematic reviews in the Cochrane Back Review Group. Spine (Phila Pa 1976) 2009;34(18):1929–41. [13] Kramer AH, Fletcher JJ. Statins in the management of patients with aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis. Neurocrit Care 2010;12(2):285–96.
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
6
P.-P. Niu et al. / Journal of the Neurological Sciences xxx (2013) xxx–xxx
[14] He J, Li X, Yang J, Huang J, Fu X, Zhang Y, et al. The association between the methionine/valine (M/V) polymorphism (rs1799990) in the PRNP gene and the risk of Alzheimer disease: an update by meta-analysis. J Neurol Sci 2013;326(1–2):89–95. [15] Yoshimoto T, Shirasaka T, Fujimoto S, Yoshidumi T, Yamauchi T, Tokuda K, et al. Cilostazol may prevent cerebral vasospasm following subarachnoid hemorrhage. Neurol Med Chir 2009;49(6):235–41. [16] Suzuki S, Sayama T, Nakamura T, Nishimura H, Ohta M, Inoue T, et al. Cilostazol improves outcome after subarachnoid hemorrhage: a preliminary report. Cerebrovasc Dis 2011;32(1):89–93. [17] Murahashi T, Kamiyama K, Hara K, Ozaki M, Mikamoto M, Nakagaki Y, et al. The efficiency of cilostazol for cerebral vasospasm following subarachnoid hemorrhage. No Shinkei Geka 2013;41(5):393–400. [18] Ito H, Fukunaga M, Suzuki H, Miyakoda G, Ishikawa M, Yabuuchi Y, et al. Effect of cilostazol on delayed cerebral vasospasm after subarachnoid hemorrhage in rats: evaluation using black blood magnetic resonance imaging. Neurobiol Dis 2008;32(1):157–61. [19] Yamaguchi-Okada M, Nishizawa S, Mizutani A, Namba H. Multifaceted effects of selective inhibitor of phosphodiesterase III, cilostazol, for cerebral vasospasm after subarachnoid hemorrhage in a dog model. Cerebrovasc Dis 2009;28(2):135–42.
[20] Kim KY, Shin HK, Choi JM, Hong KW. Inhibition of lipopolysaccharide-induced apoptosis by cilostazol in human umbilical vein endothelial cells. J Pharmacol Exp Ther 2002;300(2):709–15. [21] Vergouwen MD, Vermeulen M, van Gijn J, Rinkel GJ, Wijdicks EF, Muizelaar JP, et al. Definition of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage as an outcome event in clinical trials and observational studies: proposal of a multidisciplinary research group. Stroke 2010;41(10):2391–5. [22] Connolly Jr ES, Rabinstein AA, Carhuapoma JR, Derdeyn CP, Dion J, Higashida RT, et al. Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2012;43(6):1711–37. [23] Tanaka K, Gotoh F, Fukuuchi Y, Amano T, Uematsu D, Kawamura J, et al. Effects of a selective inhibitor of cyclic AMP phosphodiesterase on the pial microcirculation in feline cerebral ischemia. Stroke 1989;20(5):668–73. [24] Stolberg HO, Norman G, Trop I. Randomized controlled trials. AJR Am J Roentgenol 2004;183(6):1539–44. [25] Dinicolantonio JJ, Lavie CJ, Fares H, Menezes AR, O'Keefe JH, Bangalore S, et al. Metaanalysis of cilostazol versus aspirin for the secondary prevention of stroke. Am J Cardiol 2013. http://dx.doi.org/10.1016/j.amjcard.2013.05.067.
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis Peng-Peng Niu a, Ge Yang b, Ying-Qi Xing a, Zhen-Ni Guo a, Yi Yang a,⁎ a b
Department of Neurology, The First Hospital of Jilin University, Xinmin Street 71#, Changchun 130021, China Department of Geriatrics Medicine, The Affiliated Hospital to Changchun University of Chinese Medicine, Changchun, China
a r t i c l e
i n f o
Article history: Received 30 August 2013 Received in revised form 11 October 2013 Accepted 18 October 2013 Available online xxxx Keywords: Aneurysm Cilostazol Meta-analysis Subarachnoid hemorrhage Systematic review Vasospasm
a b s t r a c t Previous studies with small sample size have shown that cilostazol can reduce the risk of cerebral vasospasm in patients with aneurysmal subarachnoid hemorrhage (SAH). The purpose of this study was to determine whether cilostazol is effective in patients with aneurysmal SAH. Studies investigating the effect of cilostazol in patients with aneurysmal SAH were identified using Embase.com without language or publication-type restrictions. We used the random-effect model to combine data. Pooled risk ratios (RR) and 95% confidence intervals (CI) were calculated. Two randomized controlled trials and two quasi-randomized controlled trials with a total of 340 patients were included. The incidence of symptomatic vasospasm (RR = 0.47; 95% CI, 0.31–0.72; p b 0.001), severe vasospasm (RR = 0.48; 95% CI, 0.28–0.82; p = 0.007), vasospasm-related new cerebral infarctions (RR = 0.38; 95% CI, 0.22–0.67; p = 0.001), and poor outcome (RR = 0.57; 95% CI, 0.37–0.88; p = 0.011) were significantly lower in the cilostazol group. The numbers needed to treat for these outcomes were 6.4, 6.3, 5.7, and 5.4, respectively. Mortality rate differences between the two groups were insignificant. No statistical heterogeneity was found for all outcomes. These results show that cilostazol can decrease the incidence of symptomatic vasospasm, severe vasospasm, vasospasm-related new cerebral infarctions, and poor outcome in patients with aneurysmal SAH. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Subarachnoid hemorrhage (SAH) is a stroke subtype. Although it accounts for only 5% of all strokes, it generally occurs at a fairly young age with a relative high rate of morbidity and mortality [1]. Cerebral vasospasm and vasospasm-related cerebral infarction are common and serious complications in patients with aneurysmal SAH, and are important causes of death and dependency [2,3]. A Cochrane review showed that oral administration of nimodipine likely reduces the risk of cerebral ischemia and poor outcome after aneurysmal SAH, but the evidence for other calcium antagonists is inconclusive [4]. An updated systematic review that investigated the effects of statins in patients with aneurysmal SAH also showed no significant results [5]. Observational studies suggested that antiplatelet agents may reduce the risk of secondary ischemia and poor outcome after aneurysmal SAH [6], but randomized controlled trials (RCTs) and the Cochrane review failed to show any significant results [7]. However, studies that investigated the effect of cilostazol on aneurysmal SAH were not included in this Cochrane review. Cilostazol, a selective inhibitor of phosphodiesterase 3, is an antiplatelet agent marketed in Japan that is used to treat the ischemic symptoms
⁎ Corresponding author. Tel.: +86 15943020037; fax: +86 43188782378. E-mail addresses: [email protected], [email protected] (Y. Yang).
of peripheral vascular disease. Unlike other antiplatelet agents, cilostazol not only inhibits the platelet aggregation, but also has effects of direct vasodilatation [8] and anti-inflammation [9]. Moreover, cilostazol is more effective than aspirin in the secondary prevention of ischemic stroke without increasing the risk of bleeding complications [10]. Both animal studies and clinical studies have shown that cilostazol can reduce the risk of cerebral vasospasm in patients with aneurysmal SAH [9,11]. However, the sample sizes of previous studies were small, and there has been no relevant systematic review regarding this topic. We performed a systematic literature review of all clinical controlled trials to evaluate the effect of cilostazol in patients with aneurysmal SAH.
2. Methods 2.1. Search strategy Studies that investigated the effect of cilostazol in aneurysmal SAH were identified until August 10, 2013 through electronic searches in Embase.com, which includes records from Medline and Embase, without language or publication-type restrictions. Search results were limited to human. The search strategy (Table 1S, Supplementary materials) combined the terms “cilostazol” and “subarachnoid hemorrhage.” Two authors (P.P.N. and Y.Y.) independently performed the literature search. The references of identified articles were screened to find further studies.
0022-510X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jns.2013.10.027
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
2
P.-P. Niu et al. / Journal of the Neurological Sciences xxx (2013) xxx–xxx
2.2. Study selection and eligibility criteria Two authors (P.P.N. and G.Y.) independently screened the articles that were retrieved using the search strategy. Articles that failed to meet our inclusion criteria were excluded after reading the title, keywords, and abstract. Articles that were not excluded in the initial screen were evaluated by reading the full text. For duplicate articles, we included those with the largest sample size or the most complete information. Any disagreements were resolved by discussion. Inclusion criteria for the studies were the following: (1) randomized controlled clinical trial or controlled clinical trial; (2) study that investigated the effect of cilostazol in patients with aneurysmal SAH; (3) study that reported the incidence of vasospasm, the incidence of cerebral infarction, or functional outcome of patients; and (4) rational diagnosis methods for aneurysmal SAH and cerebral infarction were described. 2.3. Data collection Data from the articles included in this study were extracted by two independent authors (Y.Q.X. and Z.N.G.) using a standardized form. The following information were extracted from each study: name of the first author, publication year, study period, country or geographical origin of investigation, baseline characteristics of patients, number of patient in each group, and outcome data in each group. All disagreements were resolved by consensus. 2.4. Assessment of risk of bias for the included studies Two authors (G.Y. and Y.Q.X.) independently assessed the internal validity of the included studies by using the criteria list described by the Editorial Board of the Cochrane Back Review Group [12]. Twelve criteria that reflect the risk of bias were included in this appraisal system. Each criterion can be scored as yes, unclear, or no, where yes indicates the criterion has been met and therefore suggests a low risk of bias. Studies that meet at least 6 of the 12 criteria and have no serious flaws (e.g., 80% dropout rate in one group) can be defined as having a “low risk of bias.” 2.5. Outcome measures The primary outcome was the incidence of symptomatic vasospasm. Secondary outcomes were the incidence of severe vasospasm, incidence of vasospasm-related new cerebral infarctions, incidence of poor outcome, and mortality. The definitions of the outcome measures were made according to the descriptions provided in the included studies. Symptomatic vasospasm was defined as any new unexplainable focal or global neurological deficit or a decrease of at least 2 points on the Glasgow Coma Scale (not associated with rebleeding, intracerebral hematoma, hydrocephalus, metabolic disturbances, fever, or infection), regardless of cerebral vasospasm. Vasospasm-related new cerebral infarction was defined as new follow-up computerized tomography (CT) or magnetic resonance imaging (MRI) lesions not attributable to other causes, regardless of any association with symptoms. Severe vasospasm was defined as N 50% decrease in vessel diameter detected by digital subtraction angiography (DSA), computerized tomography angiography (CTA), or magnetic resonance angiography (MRA). Poor functional outcome was defined as a modified Rankin Scale (mRS) of 3 to 6 determined during the follow-up in the study. 2.6. Statistical analysis STATA, version 12.0 (Stata Corporation, College Station, Texas) was used for all statistical analyses. The risk ratio (RR) and 95% confidence interval (CI) were used as the measure of effect of interest. We measured statistical heterogeneity using the Q statistic and I2 statistic.
Heterogeneity existed if the Q statistic was pQ b 0.1 [13,14]. I2 statistic is expressed as a value between 0% and 100% that represents heterogeneity as absent to extreme. Because only a few studies with low numbers of patients were included, the DerSimonian–Laird random-effect model was used in all analyses [5]. The number needed to treat (NNT) for each outcome was calculated. Subgroup analyses by different study designs were performed. Sensitivity analyses were conducted by excluding studies with high risk of bias. 3. Results 3.1. Studies included in the meta-analysis According to our search strategies, 222 articles were initially identified. Ultimately, three articles written in English [11,15,16] and one article written in Japanese [17] with a total of 340 patients were included in this analysis. Two studies were RCTs [11,16] and two were quasi-randomized controlled trials (q-RCTs) [15,17]. The selected studies were published between 2009 and 2013. There were no significant differences in any of the characteristics between the two groups in the studies. Postoperative management was also uniform between the two groups for each study. The baseline characteristics of all qualified studies (Table 1), outcome data (Table 2), and main results of this metaanalysis (Table 3, Fig. 1) are presented. The risk of bias is also illustrated (Table 2S, Supplementary materials). 3.2. Assessment of risk of bias for the included studies The results for the risk of bias assessment are presented in Table 2S. The four included studies were judged to have low risk of bias because ≥6 items were met for each of them. However, the two q-RCTs just met 6 of the 12 items. In the study by Senbokuya et al., six patients (11.11%) of the cilostazol group dropped out because of adverse events. However, the authors included these six patients in the cilostazol group analyses. 3.3. Quantitative data synthesis 3.3.1. Symptomatic vasospasm All studies reported the incidence of symptomatic vasospasm. Using a random-effect model to combine data, the number of patients who had symptomatic vasospasm was significantly higher in the control group (55/186, 29.57%) than in the cilostazol group (23/154, 14.94%) (RR = 0.47, 95% CI = 0.31–0.72, p b 0.001), with no statistical heterogeneity showed by the Q statistic (p = 0.586) and I2 statistic (0.0%). The NNT for this outcome is 6.4 (95% CI = 4.9–12.1). The significant difference was essentially unchanged after excluding the two q-RCTs (RR = 0.46, 95% CI = 0.25–0.84, p = 0.012). No statistical heterogeneity was found between the two RCTs (pQ = 0.216, I2 = 34.5%). 3.3.2. Severe vasospasm Yoshimoto et al. reported the incidence of severe vasospasm between 7 and 9 days (or when any neurological deterioration was detected) after the onset of SAH detected by cerebral angiography. Senbokuya N et al. and Murahashi et al. reported the incidence of severe vasospasm between 7 and 10 days after the onset of SAH detected by CTA or DSA. All of them used the definition of N 50% decrease in vessel diameter for severe vasospasm. Using a random-effect model to combine data, the number of patients who had severe vasospasm was significantly higher in the control group (41/135, 30.37%) (RR = 0.48, 95% CI = 0.28–0.82, p = 0.007) than in the cilostazol group (16/105, 15.24%). The NNT for this outcome is 6.3 (95% CI = 4.6–18.3). Data on severe vasospasm did not indicate statistical heterogeneity (pQ = 0.347, I2 = 5.5%). The RR was essentially unchanged after excluding the two q-RCTs (RR = 0.42, 95% CI = 0.22–0.80, p = 0.008).
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
P.-P. Niu et al. / Journal of the Neurological Sciences xxx (2013) xxx–xxx
3
Table 1 Characteristics of included studies. Cilostazol group Study
Study period Study Study No. (F/M) region design
Control group
Age, year
Surgery type No. (F/M) (clip/coil)
Age, year
Surgery type Time of surgery (clip/coil)
60
22/4
24 (14/10) 58
19/5
Yoshimoto T 2004–2006 et al. (2009)
Japan
q-RCT
26 (23/3)
Suzuki S 2006–2008 et al. (2011) Senbokuya N 2009–2010 et al. (2013)
Japan
RCT
49 (36/13) 62 ± 13
49/0
51(40/11) 64 ± 14
51/0
Japan
RCT
54 (35/19) 60.0 ± 12.5
54/0
55 (33/22) 61.3 ± 12.7
55/0
Murahashi T 2008–2009 et al. (2013)
Japan
q-RCT
25 (21/4)
64.2 (44–82) 25/0
56 (37/19) 60.6 (35–86) 56/0
Intervention of cilostazol group
Within a couple of 200 mg/d from day days after SAH onset one after surgery for 2 weeks Within 72 h 100 mg twice per day after SAH onset for 14 days after SAH Within 72 h 100 mg twice per day after SAH onset for 14 days from 48 hours after surgery Within 72 h 200 mg/d from day after SAH onset one after surgery for 2 or 4 weeks
RCT indicates randomized controlled trial; q-RCT, quasi-randomized controlled trial.
3.3.3. New cerebral infarctions related to cerebral vasospasm Two RCTs and one q-RCT reported the incidence of new cerebral infarctions related to cerebral vasospasm. Using a random-effect model to combine data, the number of patients who had vasospasmrelated new cerebral infarctions was significantly higher in the control group (37/130, 28.46%) than in the cilostazol group (14/129, 10.85%) (RR = 0.38, 95% CI = 0.22–0.67, p = 0.001), with no statistical heterogeneity showed by the Q statistic (p = 0.997) and I2 statistic (0.0%). The NNT for this outcome is 5.7 (95% CI = 4.5–10.6). The significant difference was essentially unchanged (RR = 0.38, 95% CI = 0.20–0.71, p = 0.003) after excluding the one q-RCT. No statistical heterogeneity was found between the two RCTs (pQ = 0.967, I2 = 0.0%). 3.3.4. Poor outcome Three studies reported functional outcome. Yoshimoto et al. reported the mRS score at discharge or one month after SAH. Suzuki et al. reported the mRS score at discharge (the mean hospital stay was 40 days). Senbokuya et al. reported the functional outcome using the mRS score at 1, 3, and 6 months after SAH. In order to ensure that the time point is consistent across studies, we used the 1-month data from the Senbokuya et al. study for our analysis. Using a random-effect model to combine the data, the number of patients who had poor outcome (mRS ≥ 3) was significantly higher in the control group (56/130, 43.08%) than in the cilostazol group (31/129, 24.03%) (RR = 0.57, 95% CI = 0.37–0.88, p = 0.011). Data on poor outcome did not indicate statistical heterogeneity (pQ = 0.276, I2 = 22.2%). The NNT for this outcome is 5.4 (95% CI = 3.7–19.3). The significant difference was essentially unchanged (RR = 0.53, 95% CI = 0.30–0.94, p = 0.030, pQ = 0.155) after excluding the one q-RCT. 3.3.5. Mortality Three studies reported the mortality rates. No statistical differences were found between the cilostazol (3/129, 2.33%) and control
groups (5/130, 3.85%) (RR = 0.64, 95% CI = 0.15–2.76, p = 0.552, pQ = 0.472).
4. Discussion This meta-analysis showed that cilostazol significantly decreases the incidence of symptomatic vasospasm, severe vasospasm, vasospasmrelated new cerebral infarctions, and poor outcome in patients with aneurysmal SAH. Subgroup analysis by pooling data from the two RCTs also showed similar results. However, no significant differences in mortality rate were found between the cilostazol and control groups. As reported by Ito et al. in 2008, cilostazol can significantly attenuate delayed cerebral vasospasm partially by suppressing lipid peroxidation in SAH rats [18]. Subsequent animal studies further confirmed the effect of cilostazol on cerebral vasospasm after SAH [9,19]. Moreover, both controlled clinical trial [15] and randomized controlled clinical trial [11] confirmed the effect of cilostazol on attenuated cerebral vasospasm after SAH. The multifaceted function of cilostazol in reducing reactive oxygen species generation, suppressing apoptotic death, inducing nitric oxide production, anti-inflammation, and anti-lipid peroxidation may account for its effect on attenuated cerebral vasospasm [9,18,20]. Because cilostazol could not suppress all vasospasm, patients with poor outcome also exist in the cilostazol group. Moreover, not all infarctions and poor outcomes were caused by vasospasm. Patients with quite modest vasospasm also could develop symptomatic cerebral infarction [21], which could lead to poor outcome. Factors such as distal microcirculatory failure, poor collateral anatomy, and genetic or physiological variations in cellular ischemic tolerance likely also contribute to the development of delayed cerebral infarction and poor outcome [22]. In one of the studies included in this review [16], the authors reported that cilostazol could improve the outcome of patients with aneurysmal SAH without significantly reducing the incidence of symptomatic vasospasm and the incidence of cerebral infarction, suggesting that
Table 2 Number of events for each outcome measure. Number of events for each outcome (cilostazol group/control group) Study
Sample size (cilostazol group/control group)
Symptomatic vasospasm
Severe vasospasm
Vasospasm-related new cerebral infarctions
Poor outcome
Death
Yoshimoto T et al. (2009) Suzuki S et al. (2011) Senbokuya N et al. (2013) Murahashi T et al. (2013)
26/24 49/51 54/55 25/56
5/9 11/19 7/22 0/5
2/7 NA/NA 10/24 4/10
3/7 5/14 6/16 NA/NA
6/7 10/27 15/22 NA/NA
0/0 1/3 2/2 NA/NA
NA indicates not applicable.
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
4
P.-P. Niu et al. / Journal of the Neurological Sciences xxx (2013) xxx–xxx
Table 3 Estimate of RR for each outcome measure. RR (95% CI); p; number of studies; pQ; I2 Group
Symptomatic vasospasm
Severe vasospasm
Vasospasm-related new cerebral infarctions
Poor outcome
Death
All studies
0.47 (0.31, 0.72); b0.001; 4; 0.586; 0.0% 0.46 (0.25, 0.84); 0.012; 2; 0.216; 34.5% 0.47 (0.19, 1.14); 0.096; 2; 0.519; 0.0%
0.48 (0.28, 0.82); 0.007; 3; 0.347; 5.5% 0.42 (0.22, 0.80); 0.008;1; –; – 0.54 (0.17, 1.78); 0.007; 2; 0.183; 43.5%
0.38 (0.22, 0.67); 0.001; 3; 0.997; 0.0% 0.38 (0.20, 0.71); 0.003; 2; 0.967; 0.0% 0.40 (0.12, 1.36); 0.141;1; –; –
0.57 (0.37, 0.88); 0.011; 3; 0.276; 22.2% 0.53 (0.30, 0.94); 0.030; 2; 0.155; 50.5% 0.79 (0.31, 2.02); 0.625; 1;-; -
0.64 (0.15, 2.76); 0.552; 3; 0.472; 0.0% 0.64 (0.15, 2.76); 0.552; 2; 0.472; 0.0% –⁎
RCTs q-RCTs
⁎ Yoshimoto et al. reported that there was no any case of death in both groups.
cilostazol may improve outcomes via other mechanisms in addition to anti-arteriospasm and antiplatelet effects. Other effects of cilostazol, including improving microcirculation [23] and preventing endothelial damage [20] may contribute to the improved outcome of patients with aneurysmal SAH. Regarding the definition of outcome measures, all four studies reported the incidence of symptomatic vasospasm, and three studies reported the incidence of vasospasm-related cerebral infarction. However, the definitions of symptomatic vasospasm and vasospasmrelated new cerebral infarction were not the same among the included studies. The definition of symptomatic vasospasm in the
two RCTs was similar to what we described above. However, for the two q-RCTs, the authors provided the incidence of symptomatic vasospasm and listed all symptoms with no clear definition. For vasospasm-related infarction, only two studies [11,15] mentioned their definition that was similar to our description. Suzuki et al. [16] stated that vasospasm-related infarction was confirmed by head CT scan or MRI. The lack of consistent definitions and diagnostic methods among studies is a limitation of our study. However, even considering this limitation, we believe that the combined data analyses of the outcome measures still have some significance.
Fig. 1. RR estimates with the corresponding 95% CI for each outcome measure. (A, symptomatic vasospasm. B, severe vasospasm. C, vasospasm-related new cerebral infarctions. D, poor outcome. RCT indicates randomized controlled trial; q-RCT, quasi-randomized controlled trial; CI, confidence interval; and RR, risk ratio.)
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
P.-P. Niu et al. / Journal of the Neurological Sciences xxx (2013) xxx–xxx
In this study, we chose the incidence of symptomatic vasospasm as the primary outcome rather than the incidence of cerebral vasospasm because obvious variations existed in the definition of cerebral vasospasm and cerebral vasospasm data were lacking. For example, Senbokuya et al. [11] defined moderate angiographic vasospasm as a 25%–50% reduction in vessel diameter, which differs from the definition of b50% decrease in vessel diameter reported by Murahashi et al. [17]. Because there were three studies that reported the incidence of severe angiographic vasospasm with the same definition (greater than 50% decrease in vessel diameter), a meta-analysis using the random-effect model to combine these data was performed. The results showed that cilostazol can significantly attenuate severe cerebral vasospasm (RR =0.48, 95% CI = 0.28– 0.82, p = 0.007). Data on severe vasospasm did not indicate statistical heterogeneity (pQ = 0.347, I2 = 5.5%). Moreover, the result was essentially unchanged after excluding the two q-RCTs. Regarding the actual incidence of vasospasm and cerebral infarction, the data were reported in only one study for each of them. Yoshimoto et al. [15] reported that there were no significant differences in the incidence of angiographic vasospasm between the cilostazol group (11/26, 42.3%) and the control group (14/24, 58.3%). Senbokuya et al. [11] reported that there were no significant differences in the actual incidence of cerebral infarction between the cilostazol group (11/54, 20.4%) and the control group (21/55, 38.2%) (p = 0.057). Because the sample sizes were small, future studies with larger sample sizes may produce significant results. We used the criteria list described by the Editorial Board of the Cochrane Back Review Group to assess the risk of bias of the included studies. Although the four studies did not show high risk of bias because they met 6, 9, 10, and 6 of the 12 criteria items, respectively, the two qRCTs also might have a certain risk of bias because they did not use proper methods of allocation or any blind methods. However, for the parameters that analyzed in our study, the significant and non-significant results remained essentially unchanged after excluding the two q-RCTs. Indeed, RCT is the most powerful type of experimental study because the proper randomization procedure reduces the risk of serious imbalance in known and unknown factors that could influence the clinical course of patients [24]. However, only two included studies used the proper methods of allocation. In the other two q-RCTs included in this study, subjects in the cilostazol and control groups were recruited from two different centers for one study and during two different periods in the other study. Even the methods of allocation of these two q-RCTs were inadequate, the baseline characteristics of the patients showed no significant differences between the two groups for each study, which just meet the main purpose of random allocation. Although it is insufficient, the similar baseline characteristics between groups may partly compensate for the risk of selection bias caused by improper methods of allocation. The adverse effects of cilostazol were mentioned in one of the studies that were included in this study. The authors [11] reported that there were three hemorrhagic events (gastrointestinal hemorrhage, epidural hematoma, and intracerebral hemorrhage) and three cardiac events (sinus tachycardia in one patient and paroxysmal atrial fibrillation in two patients) in the cilostazol group. Even if these events were related to cilostazol, there were no significant differences in the incidence of adverse events between the two groups (6/54 vs. 4/55), and each of patients improved after discontinuing cilostazol. A previous Cochrane review investigated the effect of antiplatelet therapy on aneurysmal SAH showed that there were no obvious side effects of antiplatelet agents (including aspirin, cataclot, and ticlopidine) [7]. Furthermore, in a more recent meta-analysis [25], data from four RCTs showed that cilostazol is significantly better than aspirin in the secondary prevention of ischemic stroke with significantly fewer adverse events. All of these suggested that cilostazol is safe for patients with aneurysmal SAH. All in all, the results of the present systematic review should be interpreted with caution. Although there is a relative low risk of clinical heterogeneity because the dose, start time, duration of cilostazol, the
5
study population, and time of surgery were the same or similar among groups and studies, the total number of studies and the total number of patients included in this review are only 4 and 340 respectively. And there were two q-RCTs included with relative high risk of bias because they did not use proper methods of allocation or any blind methods. Furthermore, all studies were conducted in the Japanese population, which makes the results may not be generalized beyond the Japanese population [25]. In summary, this systematic review showed that the use of cilostazol can decrease the incidence of symptomatic vasospasm, severe vasospasm, vasospasm-related new cerebral infarctions, and poor outcome in patients with aneurysmal SAH with no obvious side effects, and thus supporting the use of cilostazol in patients with aneurysmal SAH. However, further larger randomized, controlled trials are needed to confirm these conclusions, particularly for non-Japanese populations. Moreover, because all the four studies had the same or similar usage of cilostazol and most of the patients included in this review were treated with clipping, future studies are needed to determine the optimal timing, dosage, duration, and safety of cilostazol treatment, as well as to evaluate the effect of cilostazol on patients treated with coiling. In addition, future studies should determine the effects of cilostazol on other outcome measures and investigate the protective mechanisms of cilostazol. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jns.2013.10.027. Conflicts of interest The authors declare that they have no conflict of interest. Acknowledgments None. References [1] van Gijn J, Kerr RS, Rinkel GJ. Subarachnoid haemorrhage. Lancet 2007;369(9558): 306–18. [2] Roos YB, de Haan RJ, Beenen LF, Groen RJ, Albrecht KW, Vermeulen M. Complications and outcome in patients with aneurysmal subarachnoid haemorrhage: a prospective hospital based cohort study in the Netherlands. J Neurol Neurosurg Psychiatry 2000;68(3):337–41. [3] Rabinstein AA, Friedman JA, Weigand SD, McClelland RL, Fulgham JR, Manno EM, et al. Predictors of cerebral infarction in aneurysmal subarachnoid hemorrhage. Stroke 2004;35(8):1862–6. [4] Dorhout Mees SM, Rinkel GJ, Feigin VL, Algra A, van den Bergh WM, Vermeulen M, et al. Calcium antagonists for aneurysmal subarachnoid haemorrhage. Cochrane Database Syst Rev 2007;3:CD000277. [5] Vergouwen MD, de Haan RJ, Vermeulen M, Roos YB. Effect of statin treatment on vasospasm, delayed cerebral ischemia, and functional outcome in patients with aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis update. Stroke 2010;41(1):e47–52. [6] Toussaint 3rd LG, Friedman JA, Wijdicks EF, Piepgras DG, Pichelmann MA, McIver JI, et al. Influence of aspirin on outcome following aneurysmal subarachnoid hemorrhage. J Neurosurg 2004;101(6):921–5. [7] Dorhout Mees SM, van den Bergh WM, Algra A, Rinkel GJ. Antiplatelet therapy for aneurysmal subarachnoid haemorrhage. Cochrane Database Syst Rev 2007;4:CD006184. [8] Goto S. Cilostazol: potential mechanism of action for antithrombotic effects accompanied by a low rate of bleeding. Atheroscler Suppl 2005;6(4):3–11. [9] Nishino A, Umegaki M, Fujinaka T, Yoshimine T. Cilostazol attenuates cerebral vasospasm after experimental subarachnoid hemorrhage. Neurol Res 2010;32(8): 873–8. [10] Kamal AK, Naqvi I, Husain MR, Khealani BA. Cilostazol versus aspirin for secondary prevention of vascular events after stroke of arterial origin. Cochrane Database Syst Rev 2011;1:CD008076. [11] Senbokuya N, Kinouchi H, Kanemaru K, Ohashi Y, Fukamachi A, Yagi S, et al. Effects of cilostazol on cerebral vasospasm after aneurysmal subarachnoid hemorrhage: a multicenter prospective, randomized, open-label blinded end point trial. J Neurosurg 2013; 118(1):121–30. [12] Furlan AD, Pennick V, Bombardier C, van Tulder M, Editorial Board CBRG. 2009 updated method guidelines for systematic reviews in the Cochrane Back Review Group. Spine (Phila Pa 1976) 2009;34(18):1929–41. [13] Kramer AH, Fletcher JJ. Statins in the management of patients with aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis. Neurocrit Care 2010;12(2):285–96.
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
6
P.-P. Niu et al. / Journal of the Neurological Sciences xxx (2013) xxx–xxx
[14] He J, Li X, Yang J, Huang J, Fu X, Zhang Y, et al. The association between the methionine/valine (M/V) polymorphism (rs1799990) in the PRNP gene and the risk of Alzheimer disease: an update by meta-analysis. J Neurol Sci 2013;326(1–2):89–95. [15] Yoshimoto T, Shirasaka T, Fujimoto S, Yoshidumi T, Yamauchi T, Tokuda K, et al. Cilostazol may prevent cerebral vasospasm following subarachnoid hemorrhage. Neurol Med Chir 2009;49(6):235–41. [16] Suzuki S, Sayama T, Nakamura T, Nishimura H, Ohta M, Inoue T, et al. Cilostazol improves outcome after subarachnoid hemorrhage: a preliminary report. Cerebrovasc Dis 2011;32(1):89–93. [17] Murahashi T, Kamiyama K, Hara K, Ozaki M, Mikamoto M, Nakagaki Y, et al. The efficiency of cilostazol for cerebral vasospasm following subarachnoid hemorrhage. No Shinkei Geka 2013;41(5):393–400. [18] Ito H, Fukunaga M, Suzuki H, Miyakoda G, Ishikawa M, Yabuuchi Y, et al. Effect of cilostazol on delayed cerebral vasospasm after subarachnoid hemorrhage in rats: evaluation using black blood magnetic resonance imaging. Neurobiol Dis 2008;32(1):157–61. [19] Yamaguchi-Okada M, Nishizawa S, Mizutani A, Namba H. Multifaceted effects of selective inhibitor of phosphodiesterase III, cilostazol, for cerebral vasospasm after subarachnoid hemorrhage in a dog model. Cerebrovasc Dis 2009;28(2):135–42.
[20] Kim KY, Shin HK, Choi JM, Hong KW. Inhibition of lipopolysaccharide-induced apoptosis by cilostazol in human umbilical vein endothelial cells. J Pharmacol Exp Ther 2002;300(2):709–15. [21] Vergouwen MD, Vermeulen M, van Gijn J, Rinkel GJ, Wijdicks EF, Muizelaar JP, et al. Definition of delayed cerebral ischemia after aneurysmal subarachnoid hemorrhage as an outcome event in clinical trials and observational studies: proposal of a multidisciplinary research group. Stroke 2010;41(10):2391–5. [22] Connolly Jr ES, Rabinstein AA, Carhuapoma JR, Derdeyn CP, Dion J, Higashida RT, et al. Guidelines for the management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2012;43(6):1711–37. [23] Tanaka K, Gotoh F, Fukuuchi Y, Amano T, Uematsu D, Kawamura J, et al. Effects of a selective inhibitor of cyclic AMP phosphodiesterase on the pial microcirculation in feline cerebral ischemia. Stroke 1989;20(5):668–73. [24] Stolberg HO, Norman G, Trop I. Randomized controlled trials. AJR Am J Roentgenol 2004;183(6):1539–44. [25] Dinicolantonio JJ, Lavie CJ, Fares H, Menezes AR, O'Keefe JH, Bangalore S, et al. Metaanalysis of cilostazol versus aspirin for the secondary prevention of stroke. Am J Cardiol 2013. http://dx.doi.org/10.1016/j.amjcard.2013.05.067.
Please cite this article as: Niu P-P, et al, Effect of cilostazol in patients with aneurysmal subarachnoid hemorrhage: A systematic review and meta-analysis, J Neurol Sci (2013), http://dx.doi.org/10.1016/j.jns.2013.10.027
