Background: Acute ischemic stroke due to basilar artery occlusion (BAO) is associated with the highest mortality in patients with large vessel occlusion. This study aimed to identify modifiable risk factors of early mortality in patients with BAO. Methods and Results: This was a cohort study of consecutive patients with BAO admitted to 47 stroke centers in China between January 2014 and May 2019. The primary outcome was all-cause mortality within 7 days after hospitalization. Of 829 patients, 164 died (0–3 days: 115; 4–7 days: 49) within 7 days after hospitalization. Among pre- and periprocedural variables, higher admission National Institutes of Health Stroke Scale (NIHSS, adjusted OR, 1.06, 95% CI: 1.04–1.09; p < 0.001), lower admission posterior circulation-Alberta Stroke Program Early Computed Tomography Score (pc-ASPECTS, adjusted OR, 0.88, 95% CI: 0.79–0.98; p = 0.02), lower Basilar Artery on Computed Tomography Angiography score (BATMAN, adjusted OR, 0.84, 95% CI: 0.76–0.93; p = 0.001), and recanalization failure (adjusted OR, 2.99, 95% CI: 2.04–4.38; p < 0.001) were independently associated with a higher risk of early mortality. Herniation (adjusted OR, 2.84, 95% CI: 1.52–5.30; p = 0.001) is an independent postprocedural predictor of early mortality. In patients dying ≤3 days, higher NIHSS (p < 0.001), lower pc-ASPECTS (p = 0.01), lower BATMAN (p = 0.004), recanalization failure (p < 0.001), herniation (p = 0.001), gastrointestinal hemorrhage (p = 0.046), and absence of pneumonia (p < 0.001) were independent predictors of early mortality. Higher NIHSS (p = 0.01), recanalization failure (p < 0.001), and pneumonia (p = 0.03) were independently associated with early mortality between 4 and 7 days. Conclusions: Recanalization failure, herniation, gastrointestinal hemorrhage, and pneumonia are potentially modifiable risk factors for early mortality in basilar artery occlusion.

© 2022 The Author(s). Published by S. Karger AG, Basel

Introduction

Basilar artery occlusion (BAO) is a rare but often catastrophic cause of stroke associated with severe disability and a high mortality. Despite advances in acute stroke management, the mortality rate after BAO remains considerably higher than seen in anterior circulation large vessel occlusion (range 31.1–51.7%) [1-3]. Most death occurs within the first 7 days of hospitalization [4]. Factors that have a significant impact on in-hospital mortality have been assessed in acute anterior circulation large vessel occlusion [4-8]. Modifiable (neurological and medical complications) and nonmodifiable factors (demographic and clinical parameters and comorbidities) that have an independent impact on in-hospital mortality of patients with acute BAO have not been determined. In this study, we aimed to evaluate factors associated with in-hospital mortality in patients with acute BAO in a daily clinical routine to better estimate the odds for early mortality after hospitalization to improve medical decision-making.

MethodsStudy Design and Data Sources

This cohort study used data from the Endovascular Treatment for Acute Basilar Artery Occlusion Study Registry (BASILAR, http://www.chictr.org.cn/, Identifier: ChiCTR1800014759), a nationwide prospective registry including 47 Chinese comprehensive stroke centers and a pool of individual data of consecutive adult patients with acute BAO confirmed by radiology from January 2014 to May 2019. Written informed consent was obtained from all the patients or their legally authorized representatives.

Variables and Outcomes

The posterior circulation-Alberta Stroke Program Early Computed Tomography Score (pc-ASPECTS; range 0–10) was used to evaluate the ischemic changes on baseline imaging according to the previous description [9]. The National Institutes of Health Stroke Scale (NIHSS) was used to quantify neurological deficit at the time of treatment [10]. The presumed stroke causative mechanism was assessed according to the Trail of ORG 10172 in Acute Stroke Treatment (TOAST) classification [11]. The collateral circulation status was assessed using Posterior Circulation Collateral (PC-CS) score and Basilar Artery on Computed Tomography Angiography (BATMAN) score. The modified Thrombolysis in Cerebral Infarction (mTICI) score of 0 to 2a was used to define unsuccessful recanalization [12]. The Heidelberg Bleeding Classification was used to assess intracerebral hemorrhage [13]. The complications compromised pneumonia (clinical or diagnostic findings of pulmonary infection), herniation (related to large cerebellar infarcts and characterized with cerebellum shift syndrome, compression of the fourth ventricle confirmed by computed tomography or magnetic resonance imaging and clinical deterioration including transtentorial herniation and cerebellar tonsillar herniation), gastrointestinal hemorrhage (any recorded episode of hematemesis or melena in the records), and other complications such as recurrent stroke, seizure, and urinary tract infection.

The primary outcome was in-hospital mortality within 7 days. Secondary outcomes were in-hospital mortality within 3 and within 4–7 days.

Statistical Analyses

Data were presented as medians (interquartile ranges) or numbers with percentages. Categorical and binary variables were compared using χ2 or Fisher exact tests. Continuous variables were compared using Student’s t test for variables with normal distributions and Mann-Whitney for variables without normal distributions.

Variables were considered for inclusion in the multivariable logistic analysis if they were significant at the p < 0.10 level in the univariate analysis and other factors with interest were also included. Then multiple variable logistic regression analysis was performed using ENTER method to identify factors which were associated with 7-day mortality. Furthermore, multivariable logistic regression was performed to evaluate death-associated factors after hospitalization within and beyond 3 days. Finally, the prognostic values for in-hospital mortality of previously developed score were compared using receiver operator characteristics: the Stroke subtype, Oxfordshire Community Stroke Project Classification, Age, and prestrike Rankin (SOAR) score is a simple 8-point clinical score to predict early mortality after acute stroke, which was derived from a UK multicenter stroke registry [14]; the National Institutes of Health Stroke Scale and modified-SOAR (mSOAR) score is a modified score, which the National Institutes of Health Stroke Scale was added to the SOAR score to improve the prognostic accuracy [15]; the Predicting Early Mortality from Ischemic Stroke (PREMISE) score included demographic and clinical variables to predict 7-day poststroke mortality treated at stroke units based on the nationwide Austrian Stroke Unit Registry [16]; and the Get With The Guidelines (GWTG) – Stroke risk model could predict an individual’s risk of death in the hospital after admission due to stroke based on the characteristics of patients at admission [17]. The area under the curve (AUC) values of each score were compared with other scores using the DeLong’s method [18]. The cutoff values were calculated with the predictive values of highest sensitivity and specificity (Youden’s Index) for early mortality within 7 days using receiver operator characteristics. SPSS version 23 (IBM Corp., Armonk, NY, USA), R version 3.2 (R Foundation for Statistical Computing, Vienna, Austria), and MedCalc (MedCalc Software Ltd., Ostend, Belgium) were used for conducting the statistical analyses. A two-tailed p < 0.05 was considered statistically significant for all analyses.

ResultsBaseline Characteristics

During the 5-year study period, 829 patients with acute BAO were enrolled. Of these, 164 patients (19.8%) died within 7 days after hospitalization. The main characteristics of the patients included in the study are shown in Table 1. As online supplementary Figure S1 shows, 54 (32.9%) died within 24 h after hospitalization, the mortality rate within the first 3 days after admission was 70.1% (n = 115), and 49 died within the 4–7 days (see www.karger.com/doi/10.1159/000526124 for all online suppl. material).

Table 1.

Characteristics of the study population

Predictors Associated with Early Mortality

In univariate analysis (Table 2), a higher risk of 7-day mortality was observed for older age, diabetes mellitus, history of ischemic stroke, prestroke mRS score, stroke severity (defined by NIHSS), pc-ASPECTS, collateral score, unsuccessful recanalization, and severe adverse events. In a multivariate logistic regression analysis (Table 2), higher admission NIHSS (adjusted OR, 1.06, 95% CI: 1.04–1.09; p < 0.001), lower pc-ASPECTS (adjusted OR, 0.88, 95% CI: 0.79–0.98; p = 0.02), lower BATMAN (adjusted OR, 0.84, 95% CI: 0.76–0.93; p = 0.001), and recanalization failure (adjusted OR, 2.99, 95% CI: 2.04–4.38; p < 0.001) were independent pre- and periprocedural variables of early mortality. As for postprocedural variables, herniation (adjusted OR, 2.84, 95% CI: 1.52–5.30; p = 0.001) remained independently associated with a higher risk of early mortality. As admission NIHSS score increased, the predicted probability of early mortality within 7 days increased, and the risk of in-hospital mortality ≤3 days was higher than that of mortality within 4–7 days (online suppl. Fig. S2).

Table 2.

Multivariable analysis: predictors of 7-day mortality

Multivariable analysis was stratified for length of stay (LOS) because of the differences in the impact of variables on early mortality (Table 3). Higher NIHSS (adjusted OR, 1.07, 95% CI: 1.04–1.10; p < 0.001), lower pc-ASPECTS (adjusted OR, 0.86, 95% CI: 0.77–0.97; p = 0.01), lower BATMAN (adjusted OR, 0.85, 95% CI: 0.75–0.95; p = 0.04), recanalization failure (adjusted OR, 2.58, 95% CI: 1.67–3.97; p < 0.001), herniation (adjusted OR, 3.23, 95% CI: 1.64–6.39; p = 0.001), gastrointestinal hemorrhage (GIH, adjusted OR, 1.69, 95% CI: 1.01–2.81; p = 0.046), and absence of pneumonia (adjusted OR, 0.40, 95% CI: 0.26–0.63; p < 0.001) were independent factors of early mortality in those with LOS ≤3 days, whereas higher NIHSS (adjusted OR, 1.05, 95% CI: 1.01–1.09; p = 0.01), recanalization failure (adjusted OR, 3.00, 95% CI: 1.62–5.54; p < 0.001), and pneumonia (adjusted OR, 3.27, 95% CI: 1.15–9.32; p = 0.03) were independently associated with early mortality between 4 and 7 days.

Table 3.

Multivariable analysis: predictors of early in-hospital mortality by LOS

External Validation of Mortality Risk Score

Previously described mortality risk scores such as SOAR, mSOAR, PREMISE, and GWTG were subsequently used to evaluate the prognostic value of early in-hospital mortality in patients with acute BAO. As shown in Figure 1, the AUCs of the mortality risk score presented with 95% CI were as followed: PREMISE: 0.64 (95% CI: 0.61–0.68); GWTG, 0.54 (95% CI: 0.50–0.57); SOAR, 0.51 (95% CI: 0.48–0.55); and mSOAR, 0.51 (95% CI: 0.47–0.54). On comparing both discriminators with the highest AUCs, it was observed that the PREMISE score was significantly better than GWTG score (p = 0.002), SOAR score (p < 0.001), and mSOAR score (p < 0.001). At the maximum Youden Index, the best cutoff point of the PREMISE score was 8, with a sensitivity of 0.46 and a specificity of 0.72, in terms of 7-day mortality after hospitalization.

Fig. 1.

Comparison of mortality risk scores (SOAR, mSOAR, PREMISE, GWTG) based on receiver operator characteristic (ROC) curves for predicting 7-day mortality (a), early mortality with LOS 0–3 days (b), and early mortality with LOS 4–7 days (c).

Discussion

We used data from the BASILAR study registry to identify risk factors for early mortality after acute BAO. To our knowledge, this is the first study to evaluate risk factors associated with early in-hospital mortality in BAO. We also performed an external validation of previously defined mortality risk scores. According to our results, recanalization failure, postprocedural complications such as herniation, GIH, and pneumonia were independent modifiable predictors for early mortality in patients with acute BAO. Among the mortality risk scores, PREMISE score has the highest value of predicting early mortality compared with other scores.

Previous studies have demonstrated that several independent factors were associated with in-hospital mortality in patients with anterior circulation large vessel occlusion [4, 5, 8]. However, few studies have established the factors which independently increase the risk of early mortality in BAO, a disorder associated with the highest mortality rate among causes of stroke [19, 20]. The majority of deaths in the first week are due to herniation, hemorrhage, and cerebral infarction according to a previous study [4]. While another study illustrated that increased intracranial pressure (iICP) was independently associated with early mortality within 7 days after hospitalization, iICP and pneumonia independently predicted death beyond 7 days according to the Berlin Stroke Register [5]. In our study, herniation in our population was a significant predictor of early mortality within 3 days of admission. Early posterior fossa craniectomy within 3 days, in patients with a large cerebellar infarct, could reduce mortality preventing the occurrence of increased iICP. GIH was another potentially modifiable risk factor for mortality within 3 days of admission. Acute stroke attack may cause GIH due to an acute stress reaction, which is associated with poor outcome and death during the acute phase [21]. Our results showed that pneumonia was a risk factor for mortality with LOS 4–7 days but seemed a protective factor for early mortality within 3 days. This could be explained that patients died within 3 days after admission were due to herniation and ischemic stroke, and there was not enough time to develop pneumonia or perform chest CT scan or detect pneumonia. In the first 3 days after stroke symptoms onset, herniation and GIH presented a major threat to patients’ lives; thus, much attention was paid to these fateful complications. In the contrary, pneumonia could be controlled by oxygen therapy and antibiotic treatment, which usually may not have an obvious impact on early mortality within 3 days. Also, patients who survived in first 3 days were easily susceptible to pneumonia in the next few days. Pneumonia is a more frequent complication in BAO than in other type of strokes as patients are more likely to suffer from immobility, dysphagia, decreased level of consciousness, and be intubated. Measures to prevent aspiration pneumonia, avoiding unnecessary intubation, and minimizing duration of intubation could reduce early mortality. Although the use of prophylactic antibiotic treatment in patients with stroke is still controversial, its use in selected BAO patients with a high risk of pneumonia may reduce early mortality [5].

In our study, stroke severity is significantly associated with increasing risk of early mortality. Death within the early period after stroke is mainly caused by cerebral ischemia [22]. We observed a higher 90-day mortality in the BASILAR study compared with the BASICS trial (51.7% vs. 40.7%) as more patients with severe deficit were enrolled in our study. Recanalization failure is a strong and modifiable independent predictor of 7-day mortality. Endovascular treatment (EVT) has been illustrated to be associated with successful recanalization ranging from 71% to 80.7% [1-3]. Although EVT can increase the proportion of successful recanalization, several previous randomized controlled studies in patients with BAO failed to show a significant improvement in clinical outcome after EVT compared to best medical management, while EVT may increase the risk of SICH. However, two recent randomized clinical trial confirmed the benefits of EVT for patients with acute BAO, which caused wide attention [23, 24].

In our study, SICH was not significantly associated with early mortality. SICH was an important factor for mortality in patients with acute BAO through mass effect. Among the patients with SICH in our study, the onset to treatment time of patients died within 7 days after hospitalization was longer than those survived more than 7 days (median OTT, 322 vs. 170). The timely treatment salvaged the ischemic penumbra and decreased the amount of infarction. In addition, among 18 patients with sICH in our study, 9 patients died within 7 days after admission, others died within 90 days after symptom onset of acute BAO, and 12 patients died during 14-day hospitalization after acute BAO. Therefore, the limited size of patients with SICH and early mortality and inevitable biases of the study design may be another reason.

Multiple mortality risk scores have been reported to predict mortality in patients after stroke [14-17, 25]. We performed an external validation of the mortality risk scores systematically. In our study, we find that the PREMISE score has the highest prognostic value for predicting 7-day mortality in patients with acute BAO in stroke unit. However, the AUC of the PREMISE score was 0.64, which is far less than a prior study (0.879) [16]. This could be explained that this score was set for all the stroke types, and only patients with acute BAO were enrolled in our study. Other scores such as SOAR and mSOAR may be more suitable in predicting early mortality in anterior circulation stroke. Although the GWTG score is comprehensive and many factors are considered in this score, the clinical practicality will be limited in clinical routine. What’s more, predicting patients with high risk of death would be helpful in the communications with family members about the possible fate of the patients.

BAO is still associated with the highest mortality rate among all strokes despite advanced EVT devices and treatment measures. Recanalization could maintain the reperfusion of the occluded basilar artery and salvage the ischemic penumbra. Preventive measures and early treatment for complications after EVT also have an important impact on clinical outcomes. Increased dehydration and early posterior fossa craniectomy could reduce intracranial pressure and risk of herniation in large cerebellar infarct. At the same time, proton pump inhibitors utility to protect the gastric mucosa and reduce risk of GIH, and early utility of antibiotics for patients with high risk of pneumonia, such as coma, intubation, or tracheotomy.

Our study has several limitations. First, this study is based on the BASILAR registry. BASILAR was not a randomized controlled trial and had the inherent drawbacks of an observational registry, which may limit the interpretation of results. Second, we only looked at early mortality, and further studies are needed to look for modifiable risk factors beyond 7 days. Finally, the cause of death could not be determined because of lack of autopsy data.

Conclusions

Recanalization failure, herniation, gastrointestinal hemorrhage, and pneumonia are potentially modifiable risk factors for early mortality in BAO. Preventive measures, early detection, and treatment could reduce mortality rate in patients with BAO.

Acknowledgment

We thank all the BASILAR coinvestigators for their contribution to the study.

Statement of Ethics

This study protocol was approved by Ethics Committee of Xinqiao Hospital (approval number 201308701). Written informed consent was obtained from all participants or their parent/legal guardian.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

This work was supported by the National Natural Science Foundation of China (No. 82071323), Chongqing Natural Science Foundation (cstc2020jcyj-msxmX0926), and Chongqing Science and Health Joint Project (No. 2019ZX001).

Author Contributions

Jiacheng Huang was involved in all aspects of study design, acquisition and analysis of data, statistical analysis, and writing of the manuscript. Maolin Wang, Fengli Li, and Weilin Kong were involved in writing of the manuscript. Da Liu, Huagang Li, Peiyang Zhou, Zhizhong Yan, Yan Wang, Jiaxing Song, Weidong Luo, Shuai Liu, Linyu Li, Jinrong Hu, Yan Tian, Zhongming Qiu, and Rui Xu were involved in acquisition of data. Wouter J. Schonewille was involved in revising the manuscript. Wenjie Zi and Qingwu Yang were involved in study supervision, data analysis, and writing of the manuscript.

Data Availability Statement

All data generated or analyzed during this study are included in this article and its online supplementary material. Further enquiries can be directed to the corresponding author upon reasonable request.

This article is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC). Usage and distribution for commercial purposes requires written permission. Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.