Outcome of severe arteriovenous malformation-related intracranial hemorrhage: the importance of cisternal subarachnoid hemorrhage and early seizures

Early Treatment of Ruptured Cerebral Arteriovenous Malformations: Analysis of Neurological Outcomes and Health Care Costs

BACKGROUND

The timing of surgical resection is controversial when managing ruptured arteriovenous malformations (AVMs) and varies considerably among centers.

OBJECTIVE

To retrospectively analyze clinical outcomes and hospital costs associated with delayed treatment in a ruptured cerebral AVM patient cohort.

METHODS

Patients undergoing surgical treatment for a ruptured cerebral AVM (January 1, 2015-December 31, 2020) were retrospectively analyzed. Patients who underwent emergent treatment of a ruptured AVM because of acute herniation were excluded, as were those treated >180 days after rupture. Patients were stratified by the timing of surgical intervention relative to AVM rupture into early (postbleed days 1-20) and delayed (postbleed days 21-180) treatment cohorts.

RESULTS

Eighty-seven patients were identified. The early treatment cohort comprised 75 (86%) patients. The mean (SD) length of time between AVM rupture and surgical resection was 5 (5) days in the early cohort and 73 (60) days in the delayed cohort ( P < .001). The cohorts did not differ with respect to patient demographics, AVM size, Spetzler-Martin grade, frequency of preoperative embolization, or severity of clinical presentation ( P ≥ .15). Follow-up neurological status was equivalent between the cohorts ( P = .65). The associated mean health care costs were higher in the delayed treatment cohort ($241 597 [$99 363]) than in the early treatment cohort ($133 989 [$110 947]) ( P = .02). After adjustment for length of stay, each day of delayed treatment increased cost by a mean of $2465 (95% CI = $967-$3964, P = .002).

CONCLUSION

Early treatment of ruptured AVMs was associated with significantly lower health care costs than delayed treatment, but surgical and neurological outcomes were equivalent.

Seizures and Epileptiform Patterns in SAH and Their Relation to Outcomes

In subarachnoid hemorrhage (SAH), seizures are frequent and occur at different time points, likely reflecting heterogeneous pathophysiology. Young patients, those with more severe SAH (by clot burden or presence of severe mental status changes at onset or focal neurologic deficits at any time), those with associated increased cortical irritation (by infarction or presence of underlying hematoma), and patients undergoing craniotomy are at higher risk. Advanced neurophysiologic monitoring allows for seizure burden quantification, identification of subclinical seizures, and interictal patterns as well as neurovascular complications that may have an independent impact on the outcome in this population. Practice regarding seizure prophylaxis varies widely; its institution is often guided by the risk-benefit ratio of seizures and medication side effects. Newer anticonvulsants seem to be equally effective and may have a more favorable profile. However, questions regarding the association of seizures and vasospasm, the therapeutic dosing, timing, and duration of antiepileptic treatment and the impact of seizures and antiepileptics on the outcome remain unanswered. In this review, we provide a broad overview of the work in this area and offer a diagnostic and therapeutic approach based on our own expert opinion.

Patient Age, Hemorrhage Patterns, and Outcomes of Arteriovenous Malformation

BACKGROUND

The angioarchitecture and presentation of arteriovenous malformations (AVMs) associated with AVM hemorrhage may vary with patient age. Our aim was to determine the influence of patient age at diagnosis on hemorrhage patterns and outcomes.

METHODS

A consecutive case series of 267 cases of ruptured AVMs was retrospectively analyzed. Hemorrhage patterns (intracerebral hemorrhage, subarachnoid hemorrhage, and intraventricular hemorrhage) were confirmed based on computed tomography imaging at initial diagnosis. Clinical outcomes were classified with a modified Rankin Score. These cases were analyzed with respect to age at diagnosis, hemorrhage patterns, Spetzler-Martin grades, and their clinical outcomes.

RESULTS

During a mean 22.2 months follow-up time (range, 7 hours to 10 years), 212 cases (79.4%) were favorable (modified Rankin Score ≤2) and 55 cases (21.6%) were unfavorable (modified Rankin Score ≥3). The mean age was higher in patients with an unfavorable outcome at follow-up. In univariate analyses, different age groups were significantly associated with bleeding patterns (P = 0.022). Unfavorable outcome was associated with intracerebral hemorrhage (odds ratio, 0.330; 95% confidence interval, 0.142-0.768; P = 0.008) and evacuation of hematoma (odds ratio, 0.195; 95% confidence interval, 0.044-0.867; P = 0.025), whereas intraventricular hemorrhage, subarachnoid hemorrhage, and intraventricular drainage were significantly associated with a favorable outcome. Different age groups were not significantly associated with bleeding patterns, sex, and the location of the AVM, and Spetzler-Martin grades did not show a significant association with the severity of outcomes.

CONCLUSIONS

Different age groups were significantly associated with bleeding patterns. Higher patient age, intracerebral hemorrhage, and evacuation of hematoma seem to be associated with an unfavorable outcome after AVM rupture.

Decompressive craniectomy for arteriovenous malformation-related intracerebral hemorrhage

Arteriovenous malformation (AVM)-related intracerebral hemorrhage (ICH) is the cause of approximately 2-3% of ICH and is an important factor in the significant morbidity and mortality in patients with AVM. Decompressive craniectomy (DC) is a surgical procedure to relieve malignant elevation of intracranial pressure. The use of DC to treat patients with AVM-ICH has been much less common. The present study describes our experience with DC for AVM-ICH and discusses the safety of this procedure. The present retrospective analysis compared 12 consecutive patients treated with DC for AVM-ICH with 23 patients treated with DC for hypertensive ICH. Nine patients were male and three were female, aged from 11 to 53 years (mean, 31.7 years). Hematoma volumes ranged from 50 to 106 ml (mean, 75.8 ml). The outcomes were good recovery in one patient, moderate disability in three, severe disability in seven, and vegetative state in one. Complications after DC included subdural hygroma in four patients, hydrocephalus in one, intracranial infection in two, and intracranial hemorrhage in one. No significant difference was found in the incidence of complications between DC for large AVM-ICH and DC for hypertensive ICH. In conclusion, the present study found no significant difference in the incidence of complications between DC for large AVM-ICH and DC for hypertensive ICH. Further investigations including a prospective randomized trial are needed to confirm the safety and efficacy of DC for the treatment of large AVM-ICH.

Relevance of bleeding pattern on clinical appearance and outcome in patients with hemorrhagic brain arteriovenous malformations

Although several descriptions of the angioarchitectural features of brain arteriovenous malformations (AVMs) associated with higher hemorrhagic risk have been reported, the prognostic value of the different bleeding patterns still needs to be elucidated. This study evaluated the influence on clinical appearance and outcome of the parenchymal and non-parenchymal (subarachnoid hemorrhage-SAH-and intraventricular hemorrhage-IVH) bleedings associated with ruptured AVMs. Clinical records and neuroradiological examinations of 30 patients with hemorrhagic AVMs were reviewed in order to identify their angioarchitectural features and the associated bleeding pattern. These data along with demographic characteristics and treatment modality were dichotomized and their relationship with clinical status at admission and follow-up was tested. IVH as well as parenchymal hematomas larger than 20 cm(3) appeared associated with a severe clinical status at admission, whereas SAH involving basal cisterns was significantly associated with unfavorable outcome. Age, sex and angioarchitectural features did not show significant association with the severity of the prognosis. However, none of these bleeding patterns appeared as an independent risk factor of poor outcome at multivariate analysis. In conclusion, our data emphasized the possibility that non-parenchymal bleeding may worsen the outcome of patients with hemorrhagic AVMs.