Comparison of Clazosentan and Nimodipine on Vasospasm and Vasospasm-Related Outcomes after Aneurysmal Subarachnoid Hemorrhage : A Post-hoc Propensity Score-Matched Analysis of Six Randomized Clinical Trials

Article information

J Korean Neurosurg Soc. 2025;.jkns.2024.0195

Publication date (electronic) : 2025 January 17

doi : https://doi.org/10.3340/jkns.2024.0195

Sung Ho Lee^1,*, Kyu-Sun Choi^2,*, Osamu Togo³, Ik Seong Park^,4

¹Department of Neurosurgery, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Korea

²Department of Neurosurgery, Hanyang University Medical Center, Hanyang University College of Medicine, Seoul, Korea

³Data Management & Biometry∙Statistics Division, Nxera Pharma Japan Co., Ltd., Tokyo, Japan

⁴Department of Neurosurgery, Bucheon St. Mary’s Hospital, The Catholic University of Korea, Seoul, Korea

Address for reprints : Ik Seong Park Department of Neurosurgery, Bucheon St. Mary’s Hospital, The Catholic University of Korea, 327 Sosa-ro, Bucheon 14647, Korea Tel : +82-32-340-7212, Fax : +82-32-340-7391, E-mail : ispahk@gmail.com

*Sung Ho Lee and Kyu-Sun Choi contributed equally as first authors.

Received 2024 October 26; Accepted 2024 December 17.

Abstract

INTRODUCTION

Aneurysmal subarachnoid hemorrhage (aSAH) is a critical neurological emergency caused by the rupture of an intracranial aneurysm. Despite advances in early surgical interventions such as surgical clipping and endovascular coiling [20], cerebral vasospasm occurs in up to 70% of aSAH patients, significantly increasing the risk of delayed cerebral ischemia (DCI), which exacerbates neurological injury [31].

For the pharmacological prevention of DCI following aSAH, various international organizations such as the American Heart Association/American Stroke Association (AHA/ASA), European Stroke Organization, and Korean Society of Cerebrovascular Surgeons recommend the oral administration of nimodipine [6,15,33]. However, in Japan, nimodipine is neither approved nor recommended as standard care, according to the Japan Stroke Society guidelines [17].

Although nimodipine is widely used, there is a lack of data comparing the incidence of vasospasm and related complications before and after its introduction. Additionally, the prognosis of aSAH remains poor, with cerebral vasospasm and DCI posing persistent challenges. Even during the widespread use of nimodipine, studies report that approximately 20–50% of patients still develop DCI after aSAH [7], and cerebral vasospasm accounts for 33.7% of deaths within 90 days following aSAH [18]. In patients unresponsive to pharmacological treatments, endovascular therapy is currently the only available option. These limited treatment options underscore the urgent need for new therapeutic strategies to more effectively address vasospasm and DCI.

Clazosentan (PIVLAZ^®), a highly selective endothelin receptor antagonist, was approved in Japan (January 2022) [26] and in South Korea (December 2023) [27] for the prevention of cerebral vasospasm and vasospasm-related cerebral infarction or cerebral ischemic episodes after aSAH. Unlike nimodipine, clazosentan specifically targets endothelin A (ETA) receptors, counteracting the potent vasoconstrictive effects of endothelin-1 (ET-1), which plays a key role in the pathophysiology of cerebral vasospasm following aSAH [19].

Currently, no direct head-to-head comparison between clazosentan and nimodipine following aSAH has been reported. A direct, head-to-head comparison between nimodipine and clazosentan may not be feasible given current clinical practices. Therefore, we performed an exploratory post-hoc analysis using data from six randomized controlled trials (RCTs) of clazosentan to indirectly compare its efficacy with that of nimodipine in preventing vasospasm and vasospasm-related morbidity and all-cause mortality (MM).

MATERIALS AND METHODS

Each randomized controlled trial included in our analyses was approved by the appropriate Institutional Review Board or Ethics Committee and conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. Our secondary analyses were conducted using anonymized data and did not require additional ethical approval.

Data sources

A post-hoc analysis was conducted using data from six RCTs of clazosentan with similar designs : two phase 2 studies (AC-054-201 [CONSCIOUS-1 [22]] and AC-054-202 [11] [hereafter 201 and 202]) and four phase 3 studies (AC-054-301 [CONSCIOUS-2 [19]], AC-054-302 [CONSCIOUS-3 [20]], AC-054-305, and AC-054-306 [10] [hereafter 301, 302, 305, and 306]). Comprehensive details on study designs and eligibility criteria have been published elsewhere [10,11,19-22] and are summarized in Table 1.

Table 1.

Summary of analyzed studies

In these six studies, participants were administered clazosentan intravenously at doses of 1, 5, 10, or 15 mg/h, or a placebo, within 48 to 56 hours of aSAH onset. Treatment continued for up to 14 days following the onset of aSAH. Orally administered nimodipine was permitted as background therapy, except for participants enrolled in Japan. Intra-arterial vasodilators, including nimodipine, were classified as rescue medications, and the prophylactic administration of intravenous nimodipine was prohibited by study protocols.

Study population and reclassification of subgroups

For this post-hoc analysis, participants from six clinical studies were reclassified into three subgroups based on their concomitant use of oral nimodipine : 1) a clazosentan subgroup consisting of those who received clazosentan without nimodipine, 2) a nimodipine subgroup, consisting of those who received placebo along with oral nimodipine, and 3) a placebo subgroup, consisting of those who received placebo without nimodipine (Fig. 1). In the clazosentan subgroup, data from participants who received the 10 mg/h dose of clazosentan, the approved dose in Korean and Japan, were used for efficacy analysis. For overall safety analysis, data from participant who received clazosentan were used, regardless of dose. In the nimodipine subgroup, participants were included if they received nimodipine from 24 hours before the initiation of study treatment and up to 24 hours after its discontinuation. Participants who received nimodipine as a rescue medication were excluded from the nimodipine subgroup.

Fig. 1.

Reclassification of six randomized clinical trials of clazosentan. *Among participants who were originally allocated in the clazosentan arm, those who received nimodipine concomitantly were not included in this post-hoc analysis. ^†Participants who received the approved dose of 10 mg/h in Korean and Japan were included in this post-hoc analysis. ^‡In the nimodipine subgroup, participants were included if they received nimodipine 24 hours before the initiation of study treatment and up to 24 hours after its discontinuation. Participants who received nimodipine as a rescue medication were excluded from the nimodipine subgroup. RCT : randomized controlled trial.

For the efficacy analysis evaluating vasospasm, participants from Studies 201, 202, 305, and 306 were used, as angiographic vasospasm was assessed in these four studies. However, for the analysis of MM and the safety analysis, participants from all of six studies were included.

RESULTS

Within-study comparison and subgroup pooled data

Analysis for vasospasm included data from Studies 202, 305, and 306 for the clazosentan 10 mg/h subgroup, and Studies 201 and 202 for the nimodipine subgroup. In the clazosentan subgroup, vasospasm occurred in 29.59% (79/267) of participants, compared to 56.55% (151/267) in the placebo subgroup (Table 2). This demonstrated a statistically significant RRR of 0.48 (95% CI, 0.35 to 0.58). In the nimodipine subgroup, vasospasm occurred in 68.04% (66/97) of participants compared to 79.31% (46/58) in the placebo subgroup, yielding an RRR of 0.14 (95% CI, -0.04 to 0.29), which was not statistically significant.

Table 2.

Within-study comparison and subgroup pooled results of vasospasm and vasospasm-related morbidity and mortality

Analysis for MM included data from Studies 202, 305, and 306 for the clazosentan 10 mg/h subgroup, and Studies 201, 202, 301, and 302 for the nimodipine subgroup. In the clazosentan subgroup, MM occurred in 20.60% (55/267) of participants, compared to 38.95% (104/267) in the placebo subgroup (Table 2). This demonstrated a statistically significant RRR of 0.47 (95% CI, 0.30 to 0.60). In the nimodipine subgroup, MM occurred in 28.19% (170/603) of participants compared to 35.48% (44/124) in the placebo group, yielding an RRR of 0.21 (95% CI, -0.04 to 0.39), which was not statistically significant.

Propensity score-matched analysis

The propensity score distributions for matched groups were acceptable between the clazosentan and nimodipine or placebo subgroups (Fig. 2). The matched analysis for vasospasm and MM included 89 and 231 participants in each group, respectively. The baseline characteristics of the matched subgroups are shown in Table 3.

Fig. 2.

Propensity score distributions. Propensity score density histogram of each subgroup used to analyze (A) moderate to severe angiographic vasospasm within 14 days after aSAH and (B) vasospasm-related morbidity and all-cause mortality within 6 weeks after aSAH. aSAH : aneurysmal subarachnoid hemorrhage.

Table 3.

Baseline characteristics of participants in the propensity score-matched analysis

Vasospasm occurred in 25.84% (23/89) of participants in the matched clazosentan subgroup, compared to 69.66% (62/89) in the matched nimodipine subgroup and 62.92% (56/89) in the matched placebo subgroup (Fig. 3A). The RRR was statistically significant between clazosentan and placebo (0.59; 95% CI, 0.40 to 0.72) and clazosentan and nimodipine (0.63; 95% CI, 0.46 to 0.75).

Fig. 3.

Propensity score-matched analysis for angiographic vasospasm and vasospasm-related morbidity and all-cause mortality. Event rate (%) and relative risk reductions (95% CIs) are shown for (A) moderate to severe angiographic vasospasm within 14 days after aSAH and (B) vasospasm-related morbidity and all-cause mortality within 6 weeks after aSAH. The numbers on each bar represent the event rates (%) for each subgroup. Only statistically significant p-values based on Fisher’s exact test between two groups are displayed : otherwise, not significant. CI : confidence interval, aSAH : aneurysmal subarachnoid hemorrhage.

MM occurred in 22.08% (51/231) of participants in the matched clazosentan subgroup, compared to 31.17% (72/231) in the matched nimodipine subgroup and 37.66% (87/231) in the matched placebo subgroup (Fig. 3B). The RRR was statistically significant between clazosentan and placebo (0.41; 95% CI, 0.21 to 0.56) and clazosentan and nimodipine (0.29; 95% CI, 0.04 to 0.48).

Overall safety

In the pooled safety analysis, the incidences of TEAEs were comparable among the clazosentan (93.1%), nimodipine (91.3%), and placebo (93.9%) subgroups, while the incidences of serious TEAEs were higher in nimodipine subgroup (34.7%) than clazosentan (23.8%) and placebo (19.6%) subgroups. Deaths were reported in 5.0% (24/479), 4.6% (28/606) and 3.7% (13/347) of clazosentan, nimodipine and placebo subgroups, respectively. Main causes of death included cerebral infarction, cerebral ischemia, cerebral vasoconstriction and brain edema (Table 4).

Table 4.

Safety overview

Among TEAEs, anemia, hypokalemia, pleural effusion, pulmonary edema and cerebral ischemia were more common in clazosentan subgroup compared to nimodipine and placebo subgroups. In contrast, pneumonia, urinary tract infection, hepatic enzyme increased, and hydrocephalus were reported more frequently in nimodipine subgroup (Table 5).

Table 5.

TEAEs reported by ≥3% of clazosentan-treated patients

DISCUSSION

Since global guidelines, including those from the USA, EU, and Korea, recommend nimodipine as the first-line treatment choice after aSAH [6,15,33], supported by a high level of evidence, it is already widely and routinely used in clinical practice. Given this context, conducting a direct head-to-head comparison with the newly approved clazosentan, which has a different mechanism of action, is challenging. At present, a post-hoc analysis using the existing six RCTs offers the most feasible approach.

Therefore, we indirectly assessed the efficacy and safety of these two drugs in preventing vasospasm and its associated outcomes after aSAH using the post-hoc analysis. Within-study comparisons and subgroup pooled data showed that clazosentan 10 mg/h significantly reduced the risk of vasospasm and MM compared to placebo, whereas nimodipine did not achieve statistical significance. Furthermore, in the propensity score-matched analysis, clazosentan 10 mg/h demonstrated a significant risk reduction in cerebral vasospasm and related outcomes when compared to both nimodipine and placebo. However, nimodipine did not yield a statistically significant benefit over placebo. These results suggest that clazosentan could provide a new strategy for preventing vasospasm and related complications. Further comparisons with nimodipine are needed to fully understand their relative efficacy. Throughout its clinical development program, clazosentan has demonstrated angiographic evidence of reducing vasospasm in several key trials (Studies 201, 202, 305, and 306). Furthermore, a recent meta-analysis by Pontes et al. [31] reported that clazosentan significantly reduced vasospasm-related DCI (relative risk [RR], 0.56; 95% CI, 0.38 to 0.81) and angiographic vasospasm (RR, 0.54; 95% CI, 0.47 to 0.61) compared to placebo across seven RCTs, six of which were included in our analysis.

The mechanism of action and evidence base for clazosentan and nimodipine differs in several important aspects. First, nimodipine, an L-type calcium channel antagonist, inhibits the influx of calcium ions in vascular smooth muscles, thereby promoting vasodilation. However, the mechanism of action of nimodipine on the prevention of DCI in aSAH is uncertain [5]. In contrast, clazosentan has a distinct mechanism of action by antagonizing ET-1 and preventing vasoconstriction that contributes to cerebral vasospasm [7]. Second, oral nimodipine lacks sufficient and up-to-date evidence [9,13]. The beneficial effect of oral nimodipine was first reported in 1983 [1] and subsequently investigated in several studies during the 1980s [24,25,28-30], which provided the foundation for its recommendation in global guidelines [6,15,33]. Conversely, clazosentan’s efficacy has been confirmed more recently in patients who underwent endovascular coiling (Study 305) and surgical clipping (Study 306) for aSAH, with data collected between 2016 and 2020. Third, while the clinical data for oral nimodipine primarily originate from Western populations (USA, Cananda, France, and the UK), clazosentan’s development program included a substantial number of patients from Japan (440 patients enrolled in Studies 305 and 306, 105 patients in Study 202) and South Korea (74 patients in Study 202). This broader and more recent demographic representation provides a more comprehensive assessment of clazosentan’s efficacy across diverse populations.

In this analysis, the RRR for vasospasm and MM with nimodipine was not statistically significant compared to placebo, which may be partially attributed to uncontrolled drug administration. Retrospective studies [8,14,32] have suggested that dose reduction or interruption of oral nimodipine is associated with unfavorable outcomes, such as an increased incidence of DCI. However, in the six RCTs analyzed, nimodipine was administered according to best clinical practice, guided by physician discretion and local regulations. Unfortunately, detailed data on dosage and treatment duration were limited. The relatively small number of subjects, particularly in the pooled analysis for vasospasm (58 participants in the placebo group), may have contributed to the marginal effect observed with nimodipine. Similarly, the beneficial effect of oral nimodipine is largely based on a single major study [30] with approximately 270 participants in each group. When this largest study is excluded, meta-analyses fail to demonstrate statistical significance, with the remaining studies including only 30–80 subjects per group [24,25,28-30].

Although the RCTs in our analysis primarily enrolled patients with Fisher grade 3 hemorrhage, the baseline characteristics, particularly clot size and outcomes, varied across the six studies. For example, over 90% of patients in Study 202 had diffuse thick clots, compared to approximately 50% in Studies 305 and 306, which may have contributed to the better outcome observed in the latter studies. Additionally, in the pooled analysis for vasospasm, approximately 80% of patients in the placebo groups of Studies 201 and 202 experienced vasospasms, compared to 50% in Studies 305 and 306. Given the heterogeneity in baseline characteristics and varying event rates, comparisons were made within subgroups of individual studies rather than pooling the results from all studies.

In the pooled safety analysis, the overall incidences of TEAEs were comparable among the subgroups. Clazosentan was associated with a higher incidence of pleural effusion (12.7%) and pulmonary edema (8.1%) compared to placebo (3.5% and 4.3%, respectively) and nimodipine (4.3% and 4.5%, respectively). However, most patients who developed pleural effusion or pulmonary edema were adequately managed and recovered during hospitalization, with no fatal cases reported. These adverse events are consistent with the known class effects of ETA receptor antagonists, with similar or lower placebo-corrected incidence rates of edema or fluid overload observed with other agents such as zibotentan (38%), atrasentan (28%), ambrisentan (18%), avosentan (15%), durasentan (12%), and sitaxentan (8%) [2-4,12,16,23,35].

Our study has several strengths, including the indirect comparison of clazosentan, a newly approved ETA receptor antagonist, with nimodipine. To address the heterogeneity of enrolled subjects and clinical outcomes, we conducted a propensity score-matched analysis to ensure more balanced treatment groups and minimize bias. In the propensity score-matched analysis, 89 and 231 participants were selected from each subgroup for vasospasm and MM, respectively, which are relatively large sample sizes compared to previous studies investigating nimodipine.

However, our study also has limitations including the uncontrolled administration of oral nimodipine and the lack of detailed data on its usage. Furthermore, heterogeneity in the eligibility criteria, baseline characteristics and changes in the standard of care, such as the removal of hypervolemia as a recommended treatment since 2012 [15], may have influenced our results. These limitations highlight the need for further well-designed clinical studies to compare the efficacy and safety of clazosentan and nimodipine, given the evolving standards of care in aSAH management.

CONCLUSION

In this exploratory post-hoc analysis of six RCTs, clazosentan 10 mg/h demonstrated a significant reduction in the incidence of cerebral vasospasm and MM following aSAH, compared to both placebo and nimodipine. The overall safety results were comparable across the three subgroups and consistent with the expected range for ETA receptor antagonists. The TEAEs of pleural effusion and pulmonary edema were generally well managed, despite their higher incidences with clazosentan. While nimodipine has long been used, our findings suggest that clazosentan may offer superior efficacy. Further clinical studies comparing the efficacy of clazosentan, either alone or in combination with nimodipine, are needed to determine the potential treatment position of clazosentan and to develop optimal treatment strategies for aSAH.

Article information Continued