Objective Although endovascular therapy has been widely adopted for the treatment of cerebral vasospasm after aneurysmal subarachnoid hemorrhage (aSAH), its effect on clinical outcomes remains incompletely understood. The aims of this retrospective cohort study are to evaluate the outcomes of endovascular intervention for post-aSAH vasospasm and identify predictors of functional independence at discharge and repeat endovascular vasospasm treatment.
Methods We assessed the baseline and outcomes data for patients with aSAH who underwent endovascular vasospasm treatment at our institution, including intra-arterial (IA) vasodilator infusion and angioplasty. Statistical analyses were performed to determine factors associated with good outcome at discharge (modified Rankin Scale 0–2) and repeat endovascular vasospasm treatment.
Results The study cohort comprised 159 patients with a mean age of 52 years. Good outcome was achieved in 17% of patients at discharge (26/150 patients), with an in-hospital mortality rate of 22% (33/150 patients). In the multivariate analysis, age (OR 0.895; p=0.009) and positive smoking status (OR 0.206; p=0.040) were negative independent predictors of good outcome. Endovascular retreatment was performed in 34% (53/156 patients). In the multivariate analysis, older age (OR 0.950; p=0.004), symptomatic vasospasm (OR 0.441; p=0.046), initial treatment with angioplasty alone (OR 0.096; p=0.039), and initial treatment with combined IA vasodilator infusion and angioplasty (OR 0.342; p=0.026) were negative independent predictors of retreatment.
Conclusion We found a modest rate of functional independence at discharge in patients with aSAH who underwent endovascular vasospasm treatment. Older patients and smokers had worse functional outcomes at discharge. Initial use of angioplasty appears to decrease the need for subsequent retreatment.
Statistics from Altmetric.com
Aneurysmal subarachnoid hemorrhage (aSAH) accounts for less than 5% of all strokes and affects approximately 21 000–33 000 people in the USA each year.1 2 Despite major advances in ruptured aneurysm treatment and post-aSAH neurological intensive care, cerebral vasospasm and delayed cerebral ischemia (DCI) remain significant causes of delayed morbidity and mortality in this patient population.3 Angiographic vasospasm can be found in up to 70% of patients with SAH, and approximately one-third to one-half of those who develop vasospasm experience neurological symptoms.4 Although the pathophysiology of vasospasm and DCI has been the focus of extensive research, preventative therapy and treatment for these post-aSAH processes are currently limited.5 6
Medical therapy with a combination of induced hypertension, hypervolemia, and hemodilution (ie, ‘triple-H’ therapy) remains the first-line treatment for symptomatic vasospasm, although its efficacy is variable. Endovascular intervention is commonly used to chemically (intra-arterial (IA) infusion of vasodilatory agents) or mechanically (balloon angioplasty) ameliorate vasospasm which has been refractory to medical therapy.7 However, since no randomized controlled trial has evaluated the safety and efficacy of endovascular treatment for vasospasm, the current guidelines are based on clinical case series and expert consensus.8 As such, the clinical effect of endovascular vasospasm treatment remains incompletely understood. Therefore, the aims of this retrospective cohort study are to (1) assess the outcomes of endovascular intervention for vasospasm after aSAH and (2) identify predictors of functional independence at discharge and the need for repeat endovascular vasospasm treatment.
This study was approved by the Institutional Review Board (IRB) at our institution. Patient consent was not required by the committee due to the retrospective nature of the study and de-identification of data presented.
We retrospectively evaluated a database of consecutive patients with aSAH who underwent endovascular treatment for vasospasm at our institution between August 1999 and March 2015. The inclusion criteria were as follows: (1) age ≥18 years; (2) diagnosis of SAH via CT, MRI, and/or lumbar puncture (LP); (3) SAH was aneurysmal in origin, as diagnosed by CT angiography (CTA) or catheter angiography; (4) radiographic diagnosis of cerebral vasospasm by CTA or catheter angiography; and (5) endovascular treatment of vasospasm using IA drug infusion therapy and/or balloon angioplasty. The exclusion criteria were as follows: (1) no evidence of SAH on CT, MRI, or LP; (2) non-aneurysmal SAH; (3) no radiographic evidence of vasospasm on CTA or catheter angiography; and (4) no endovascular intervention performed.
In general, rupture aneurysms were secured within 24 hours of admission by surgical clipping or endovascular coiling. Patients were monitored in the neurological intensive care unit and a euvolemic fluid balance was maintained. Patients who developed a new neurological deficit or change in mental status underwent a head CT and CTA to evaluate for rehemorrhage, hydrocephalus, or vasospasm. Those with radiographically confirmed vasospasm on CTA were then medically treated with induced hypertension, with the goal of elevating the mean arterial pressure by 20%. Patients with symptomatic vasospasm who failed to improve with medical therapy were transferred to the neurointerventional suite for endovascular therapy.
Patients also underwent routine diagnostic catheter angiography approximately 7 days after aneurysm rupture. If vasospasm was detected on angiography, intervention was performed. Endovascular treatment included either IA infusion of a vasodilatory agent (papaverine or verapamil) into the affected vessels or angioplasty. Although the specifics of intervention were operator-dependent, angioplasty was typically reserved for severe vasospasm of the large proximal vessels. In order to reduce the risk of vessel injury, angioplasty was performed to expand the arterial lumen to approximately 50–90% of its original diameter. When angioplasty was undertaken, intravenous heparin was administered using a weight-based protocol (100 units/kg) to achieve an activated clotting time of 250–300 s for the duration of the procedure.
Baseline and outcomes data
Baseline data regarding patient, aneurysms, and vasospasm characteristics were recorded. Patient data included age, sex, history of hypertension, and smoking. Initial neurological assessments at presentation were categorized using the Hunt–Hess (HH) and World Federation of Neurosurgical Societies (WFNS) grades.9 10 Radiographic assessment of SAH was categorized using the Fisher grade.11 Aneurysm location was dichotomized into anterior and posterior circulations. Anterior circulation locations were further classified into internal carotid artery (ICA) terminus, anterior cerebral artery (ACA), middle cerebral artery (MCA), and anterior communicating artery (ACOM). Posterior circulation locations were further classified into posterior communicating artery (PCOM), posterior inferior cerebellar artery (PICA), basilar artery, posterior cerebral artery (PCA), vertebral artery (VA), and superior cerebellar artery (SCA). Aneurysm treatment modalities included microsurgery, endovascular therapy, and combined therapy.
Vasospasm data included time from the ictus of SAH to diagnosis of vasospasm, vasospasm distribution (dichotomized into bilateral vs unilateral), vasospasm location (classified as anterior circulation, posterior circulation, or both), severity of vasospasm, and presence of neurological symptoms. The severity of vasospasm was defined as mild, moderate, or severe, based on reductions in arterial diameter of ≤33%, 34–66%, or ≥67%, respectively.12 Endovascular vasospasm treatment was categorized as IA vasodilator infusion only, angioplasty only, and combined therapy.
Follow-up included radiographic and clinical outcomes data. Reperfusion was defined as successful if 50% reversal of the luminal stenosis was achieved, or failure if <50% reversal of the luminal stenosis was achieved with endovascular treatment. The need for repeat endovascular vasospasm treatment and length of hospital stay were also recorded. Each patient’s functional outcome at discharge was assessed using the modified Rankin Scale (mRS). Good outcome was defined as a mRS of 0–2 (functional independence) whereas poor functional outcome was defined as a mRS of 3–6 (functionally dependent or dead).
All statistical analyses were performed using Stata V. 14.2 (College Station, Texas, USA). Descriptive statistics were computed for all collected data. Comparisons between the good and poor outcome groups were performed using Pearson’s χ2test, Student t-test, or Fisher’s exact test for each independent variable, as appropriate. Covariates associated with functional outcome (p<0.10) were entered into a multivariate logistic regression model to identify independent predictors of functional outcome. In a separate analysis, covariates associated with repeat endovascular vasospasm treatment (p<0.10) were entered into a multivariate logistic regression model to identify independent predictors of repeat vasospasm treatment. The predictors were tested for multicollinearity using tolerance and variance inflation factor, and one variable from closely correlated variables was selected as a predictor in the final model. Goodness-of-fit of the model was assessed using the Hosmer–Lemeshow test. Odds ratios (OR) and 95% CIs were computed for each predictor. Statistical significance was defined as p<0.05 and all tests were two-tailed. Missing data were not imputed.
Patient and aneurysm characteristics
After application of the inclusion and exclusion criteria, the study cohort comprised 159 patients with aSAH who underwent endovascular treatment of vasospasm. Table 1 summarizes the baseline patient and aneurysm characteristics of the study cohort and compares these factors between those with good (mRS 0–2) versus poor (mRS 3–6) outcomes. The mean age was 51.8 years and 76.1% of patients were female. A history of hypertension was noted in 76.7% (79/103 patients) and 71% (71/100 patients) were smokers. The location of aneurysms was anterior circulation in 90 (56.6%) and posterior circulation in 69 (43.2%). Microsurgical, endovascular, and combined treatments for aneurysms were performed in 70 (44.0%), 69 (43.4%), and 20 (12.6%) patients, respectively.
Patients of older age (53.3 vs 44.7 years; p=0.0006) with a history of hypertension (82.7% vs 57.9%; p=0.031) and positive smoking status (76.7% vs 54.6%; p=0.043) were significantly less likely to have a good outcome at discharge. The HH (p<0.001) and WFNS (p=0.016) grades of patients with a good outcome were significantly lower.
Vasospasm characteristics and treatments
Table 1 summarizes the vasospasm characteristics, endovascular therapies, and radiographic outcomes of the study cohort and compares these factors between those with good (mRS 0–2) and those with poor (mRS 3–6) outcomes. The mean time from the ictus of SAH to diagnosis of vasospasm was 5.1 days. Vasospasm was clinically symptomatic in 56 patients (35.2%). The severity of vasospasm was mild, moderate, and severe in 20.5% (31/151 patients), 33.8% (51/151 patients), and 45.7% (69/151 patients), respectively. IA vasodilator infusion alone, angioplasty alone, and combined treatment were performed in 106 (66.7%), 11 (6.9%), and 26.4 (26.4%) patients, respectively. Reperfusion after endovascular therapy was successful (≥50% reversal of luminal stenosis) in 92.8% (103/111 patients). No cases of intraprocedural hypotension or vessel rupture occurred. The mean length of hospital stay was 25.4 days. The length of hospital stay was significantly shorter for patients with a good outcome (15.6 days vs 27.5 days; p=0.0042).
Predictors of good outcome at discharge
Of the 150 patients with available data regarding functional outcome at discharge, 26 had a good outcome (mRS 0–2; 17.3%) and 33 died (in-hospital mortality rate 22%). Table 2 details the multivariate analysis for predictors of good outcome at discharge. Older age (OR 0.895; p=0.009) and positive smoking status (OR 0.206; p=0.040) were found to be negative independent predictors of a good outcome. WFNS grade was removed as covariate due to multicollinearity with HH grade.
Repeat vasospasm treatment
Repeat endovascular vasospasm treatment was performed in 34.0% (53/156 patients). Table 3 compares the characteristics between patients who underwent repeat endovascular vasospasm treatment and those who did not undergo repeat treatment. Patients who underwent repeat endovascular vasospasm treatment were younger (mean age 49.4 vs 53.6 years; p=0.0369), had shorter time intervals from presentation with SAH to the initial diagnosis of vasospasm (mean 4.4 vs 5.4 days; p=0.0341), and were less likely to have symptomatic vasospasm (24.5% vs 41.8%; p=0.034). The initial vasospasm therapies in those who underwent retreatment more frequently used verapamil as the vasodilatory agent (96.2% vs 76.7%; p=0.006) and less frequently employed angioplasty (20.8% vs 39.8%; p=0.017). The association between the use of combined microsurgical and endovascular aneurysm treatment and retreatment trended toward significance (p=0.098).
Table 4 details the multivariate analysis for predictors of repeat endovascular vasospasm treatment. Older age (OR 0.950; p=0.004), symptomatic vasospasm (OR 0.441; p=0.046), the use of angioplasty alone (OR 0.09; p=0.039), and combined therapy with IA vasodilator infusion and angioplasty (OR 0.342; p=0.026) were found to be negative independent predictors of repeat treatment. The type of IA vasodilatory agent infused and use of angioplasty (alone or combined with IA vasodilator infusion) were removed as covariates due to multicollinearity with type of endovascular vasospasm treatment.
Despite a 50% reduction in the case fatality rate of aSAH over the last two decades, many survivors suffer permanent neurological, cognitive, or functional deficits.13 14 Moreover, due to the younger mean age of onset and more substantial extent of morbidity, aSAH carries a greater overall cost to society compared with ischemic stroke or intracerebral hemorrhage.15 Cerebral vasospasm and DCI remain the primary causes of delayed morbidity and mortality after aSAH in patients who survive the initial aneurysm rupture.16 Clinical predictors of DCI include volume, location, persistence and density of subarachnoid blood clot, poor clinical condition on admission, loss of consciousness at ictus, smoking, diabetes, hyperglycemia, and systemic inflammatory response syndrome.6 17 18
Endovascular treatments for vasospasm include IA vasodilator infusion and angioplasty. Numerous vasodilatory agents have been used to treat vasospasm including verapamil, papaverine, nicardipine, nimodipine, fasudil, and milrinone. Angioplasty is a controlled injury to the vasospastic vessel wall via endoluminal balloon inflation to the near-normal vessel diameter, which results in temporary vascular smooth muscle functional impairment.3 Human and animal studies have demonstrated transient flattening of the intima and internal elastic lamina associated with patchy endothelial denudation following angioplasty.19–21 Prophylactic angioplasty for vasospasm has been investigated in a phase II randomized controlled trial of 170 patients with Fisher grade III aSAH; however, no difference in the incidence of delayed neurological deficit or 3-month functional outcome was found between the intervention and control groups.22 Despite the widespread use of IA vasodilator therapy and angioplasty for the treatment of vasospasm in clinical practice, no randomized trials have compared these interventions with medical management alone. Studies of endovascular therapy for vasospasm have, thus far, been limited to retrospective studies. Current American Heart Association/American Stroke Association guidelines for the management of post-aSAH vasospasm state that angioplasty and/or selective IA vasodilator therapy is reasonable in patients with symptomatic cerebral vasospasm, particularly those who are not rapidly responding to hypertensive therapy (Class IIa, Level of Evidence B).7
In our retrospective cohort study of 159 patients who underwent endovascular treatment for vasospasm, functional independence was achieved in 17% of patients at discharge, with an in-hospital mortality rate of 22%. The rate of functional independence is lower and the mortality rate is higher compared with the literature.23 In a recent retrospective study of 116 patients who underwent endovascular treatment for vasospasm, Chalouhi et al reported good functional outcome (Glasgow Outcome Scale (GOS) 4–5) in 73% of patients, with comparable mean age to our study, at discharge.24 Of note, 9.5% of patients in this study had non-aneurysmal SAH, and microsurgical clipping was performed in a lower proportion of patients (27.6%) compared with our cohort. Although our cohort comprised a higher proportion of patients with HH grade 5 aSAH, Chalouhi et al reported a higher proportion of patients with poor admission HH grades (3–5). Hence, the lower rate of functional independence found in our study cannot readily be accounted for by differences in admission HH grades. In an earlier study of 189 patients of similar mean age who underwent endovascular treatment of vasospasm, Jun et al reported good functional outcome (mRS score 0–2) in 61% of patients at a median follow-up of 180 days.25 Microsurgical clipping was more commonly performed for aneurysm treatment (75%) than in our study. This series reported similar rates of poor admission HH grades compared with our study. In a meta-analysis of 55 studies comprising 1571 patients who underwent IA vasodilator infusion alone or combined with angioplasty, Venkatramen et al reported good functional outcome (mRS 0–2 or GOS 4–5) in 66% and mortality in 9%.23 No significant differences in outcome between IA vasodilator infusion alone and medical management alone (OR 0.8, p=0.12) or between IA vasodilator infusion alone and balloon angioplasty alone (OR 1.11, p=0.60) were found in their meta-analysis. However, the meta-analysis was unable to account for differences in patient, aneurysm, and vasospasm characteristics among the included studies. Therefore, the benefits of endovascular therapy for vasospasm cannot be fully evaluated without a randomized controlled trial.
Older age (p=0.009) and smoking (p=0.040) were found to be independent predictors of poor outcome (functional dependence or death) in our study. In contrast, Chalouhi et al reported higher HH grades, pre-procedural hypodensities, posterior circulation aneurysm location, and lack of neurological improvement (p<0.001) following treatment to be significant predictors of poor outcome at discharge.24 In another study of 81 patients with aSAH with symptomatic vasospasm who underwent endovascular treatment, Rabinstein et al found older age (p=0.02) and higher WFNS grade (p=0.02) to be predictors of poor outcome (mRS 3–6) at a median follow-up of 3 months.26 The lack of association between the success of reperfusion and clinical outcome underscores the multifactorial nature of DCI.
Repeat endovascular treatment for recurrent vasospasm was performed in 34% of patients in our study. This retreatment rate is comparable to other studies; Chalouhi et al reported a lower rate of 19%, whereas Jun et al reported a higher rate of 54%.24 25 Older age (p=0.004), symptomatic vasospasm (p=0.046), initial vasospasm treatment with angioplasty alone (p=0.039), and combined initial treatment with IA vasodilator infusion and angioplasty (p=0.026) were negative independent predictors of retreatment. In comparison, Chalouhi et al observed that higher HH grades and treatment with nicardipine alone or angioplasty with nicardipine were independent positive predictors of retreatment.24 Jun et al reported higher rates of retreatment in patients treated with verapamil alone (51%) compared with those treated with angioplasty and verapamil (17%) or angioplasty alone (0%).25 Vasospasm has been postulated to be less common in older patients as a result of an age-related decrease in vessel compliance.27 However, in addition to undergoing fewer retreatments, older patients also had worse functional outcomes. This suggests that a more aggressive approach with regard to the utilization of endovascular therapy for vasospasm should be considered in elderly aSAH patients. Our finding that the initial use of angioplasty decreased the subsequent need for retreatment indicates that angioplasty is more durable than IA vasodilator infusion alone.3 This potential benefit of angioplasty must be weighed against the greater risk of procedural complications associated with angioplasty.28 However, the use of angioplasty was not associated with worse functional outcomes in our study, which suggests that the appropriately selected application of this technique can safely prevent recurrent vasospasm.
It is important to recognize the limitations of the current study. The study’s design subjects it to the inherent selection, treatment, and referral biases of a single-center retrospective cohort analysis. All of the patients underwent endovascular vasospasm treatment, which precludes a comparison of outcomes data to vasospasm patients who were managed with medical therapy alone. Since the institutional practices, individual physicians’ treatment algorithms, and endovascular technology have changed over the course of this study, patients with aSAH who were selected to undergo endovascular intervention may only represent a subset of those with vasospasm. Although rates of major complications such as hypotension and vessel rupture were recorded, we were unable to ascertain whether these post-aSAH period complications developed secondary to aneurysm rupture, aneurysm treatment, or IA vasospasm treatment. The distinction between symptomatic and asymptomatic vasospasm may have been limited by the poor neurological conditions of patients who were intubated. Specifically, the relatively greater proportion of patients with asymptomatic vasospasm compared with those with symptomatic vasospasm may reflect, in part, some patients who were intubated and did not exhibit neurological symptoms attributable to vasospasm, and were therefore categorized as having asymptomatic vasospasm. This may have contributed to the composition of the study cohort by a majority of patients with asymptomatic vasospasm, as well as to the increased likelihood of retreatment in those with asymptomatic vasospasm. We also acknowledge that there was a trend toward a longer duration of hospitalization in patients who underwent repeat treatment (mean 29 vs 23 days; p=0.057). Additionally, the discretion in endovascular vasospasm treatment was operator-dependent, and thus we were unable to adjust for the heterogeneity of therapies that patients received.
Functional independence at discharge was achieved in a modest proportion of patients with aSAH who underwent endovascular treatment of vasospasm. Older patients had worse functional outcomes at discharge but were also less likely to undergo repeat endovascular vasospasm treatment. Therefore, the more aggressive application of endovascular therapy to older aSAH patients may be warranted, although its effect on clinical outcomes needs to be validated in future studies The initial use of balloon angioplasty, with or without IA vasodilator infusion, appears to decrease the rate of retreatment, although the present study is unable to confirm the indications for angioplasty. A randomized controlled trial comparing endovascular intervention to medical management is necessary to ascertain the risk to benefit profile of endovascular vasospasm treatment for patients with aSAH.
JDS and C-JC contributed equally.
Contributors JDS: conception/design of study, acquisition of data, drafting of manuscript, critically revising manuscript, final approval of the manuscript, agreement to be accountable for all aspects of the work; C-JC, DD: conception/design of study, analysis/interpretation of data, drafting of manuscript, critically revising manuscript, final approval of the manuscript, agreement to be accountable for all aspects of the work; TJB, DMR, NI, DGT, RMS, KL: critically revising manuscript, final approval of the manuscript, agreement to be accountable for all aspects of the work.
Competing interests None declared.
Ethics approval IRB of University of Virginia.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.