Elsevier

World Neurosurgery

Volume 85, January 2016, Pages 32-41
World Neurosurgery

Original Article
SLAM: Stereotactic Radiosurgery of Large Arteriovenous Malformations: Meta-analysis of Hemorrhage in High-Grade Pollock-Flickinger Arteriovenous Malformations

https://doi.org/10.1016/j.wneu.2015.08.022Get rights and content

Background

The prognosis of arteriovenous malformations (AVMs) after treatment has been predicted largely by 2 grading scales: the Spetzler-Martin and Pollock-Flickinger. Although there are studies that examine the rate of hemorrhage with the Spetzler-Martin scale, there have not been studies examining hemorrhage in which the Pollock-Flickinger score was used. The annual hemorrhage rate after radiosurgery of Pollock-Flickinger AVM scores >2 is analyzed.

Methods

Literature search for radiosurgery of large AVMs from January 1, 2000 to June 1, 2014 was conducted. Articles were examined for individual patient data and aggregate patient data that reported hemorrhage rates and mortality. Patients were only included if they had an AVM score ≥2.

Results

Annual AVM hemorrhage rate after radiosurgery for all patients (n = 673) was 3.22% (99.3 hemorrhages, 3080.5 follow-up years, 95% confidence interval [95% CI] 2.64%-3.89%). Mortality rate from hemorrhage was 40.08% (95% CI 31.21%–49.90%). A total of 203 patients presented with hemorrhage and 395 did not. In patients with first-time hemorrhage, the annual hemorrhage rate was 3.53% (95% CI 2.66%–4.77%). The annual hemorrhage rate of those with hemorrhagic presentation was 6.10% (95% CI 4.65%–8.07%). The odds ratio comparing re-hemorrhage rate versus first-time hemorrhage is 1.768 (95% CI 1.1571–2.7014, P = 0.0084). Complete obliteration of all AVMs was equal to 33.27% (95% CI, 29.25%–37.54%).

Conclusions

The annual hemorrhage rate in AVMs with scores >2 treated with radiosurgery was comparable with baseline rupture rates reported for untreated AVMs. With further stratification by hemorrhagic versus nonhemorrhagic presentation, the subsequent annual hemorrhage rates are similar to their respective natural histories. Considering the mortality rate from hemorrhage at 40.08% (95% CI, 35.54%–44.62%), the consequences of radiosurgical treatment of large AVMs is significantly worse than the reported 10%–30% fatality rate from hemorrhage of an untreated AVM. Additionally, the overall mortality rate was 6.24% however the percentage of mortalities from hemorrhage was 97.62%.

Introduction

Cerebral arteriovenous malformations (AVMs) constitute a rare, congenital cerebrovascular phenomenon that carries a risk for stroke. AVMs present most commonly with hemorrhage in approximately 50% of cases.1, 2, 3, 4, 5, 6, 7, 8, 9, 10 The overall annual risk of bleeding is estimated to be 2%–4% but has shown to vary depending on initial presenting symptoms and characteristics of the AVM.1, 2, 8

AVMs traditionally have been assessed by the use of grading scales. The Spetzler–Martin grade is a prognostic score for outcome after microsurgical resection of AVMs. The score is derived from 3 elements: size (diameter), location in an eloquent area, and presence of deep venous drainage.11 The components are summed to grade AVMs from I to V, with Grades IV and V considered as high grade (Table 1). A Spetzler–Martin grade of IV or V has approximately a 70% chance of an excellent outcome.11

Pollock and Flickinger proposed another grading system to prognosticate outcomes after radiosurgical treatment of AVMs, which was later modified.12 This grading system accounts for volume and location of the AVM as well as the age of the patient (Table 1).12 The algorithm governing the modified grading system is given as:AVMscore=(0.1)(Volume)+(0.02)(Age)+(0.5)(Location)

For AVM scores greater than or equal to 2, considered high grade there is a 46% chance of an excellent outcome, which is defined as complete nidal obliteration with no subsequent development of neurological deficit.12

Annual bleeding rates after radiosurgery for large AVMs range between 3.3% and 12.41%.5, 13, 14 In this paper, we evaluate the rate of hemorrhage after radiosurgical treatment for AVMs with an AVM score greater than 2 and separate the rates by hemorrhagic versus nonhemorrhagic presentation. Previous studies have evaluated the rate of hemorrhage after radiosurgery by use of the Spetzler-Martin grading scale; however, to the authors' knowledge, this is the first study evaluating hemorrhage rate of large cerebral AVMs after radiosurgery with the Pollock-Flickinger grading scale, which is designed to prognose radiosurgical outcomes.

Section snippets

Search Strategy

The PubMed database was searched for publications related to the treatment of large cerebral AVMs with radiotherapy from January 1, 2000, to June 1, 2014. The 2 search terms used were “Large AVM Radiosurgery” and “Large AVM Hemorrhage.” The articles were reviewed for pertinence to the topic of radiosurgical management of large AVMs. Additionally, the references of included articles were searched manually to gather any studies and pertinent background information that may not have been found

Results

Nine individual studies15, 16, 17, 18, 19, 20, 21, 22, 23 were used in the analysis for risk of hemorrhage associated with the radiosurgical treatment of cerebral AVMs with an AVM score greater than 2 (Table 2). Among the 673 patients, the total number of hemorrhages after SRS was 102.3 with a cumulative hemorrhage risk of 15.20% among all patients after treatment. Total number of fatalities throughout the follow-up time period was 42 equaling 6.24% among all patients. Of the 42 total deaths,

Previous Studies on Hemorrhage Rate of High-Grade AVMs

The rate of hemorrhage of high-grade AVMs as defined by Spetzler-Martin Grades IV and V has been previously examined in 4 studies by Han et al.,14 Jayaraman et al.,5 Laakso et al.,24 and Ding et al.13 Han et al., Jayaraman et al., and Ding et al. are studies on the rate of hemorrhage of large AVMs after treatment.5, 13, 14 Laakso et al. is a study on the rate of hemorrhage of large AVMs that went untreated.24 These studies were based on the Spetzler–Martin grading scale and exclusively examined

Conclusion

The treatment of large AVMs presents the dilemma of balancing treatment goals with risk of complications. The number of patients with AVM who present with hemorrhage is approximately 50%, and represents the majority of patients at presentation. With a conservative estimate of a natural risk of re-hemorrhage at 6% after initial presentation of hemorrhage and without treatment, our analysis shows that there is no significant reduction in annual rate of hemorrhage after radiosurgery of high-grade

References (31)

  • S. Yamada et al.

    Risk factors for subsequent hemorrhage in patients with cerebral arteriovenous malformations

    J Neurosurg

    (2007)
  • C.P. Yen et al.

    Hemorrhage risk of cerebral arteriovenous malformations before and during the latency period after GAMMA knife radiosurgery

    Stroke

    (2011)
  • R.F. Spetzler et al.

    A proposed grading system for arteriovenous malformations

    J Neurosurg

    (1986)
  • B.E. Pollock et al.

    A proposed radiosurgery-based grading system for arteriovenous malformations

    J Neurosurg

    (2002)
  • D. Ding et al.

    Outcomes following single-session radiosurgery for high-grade intracranial arteriovenous malformations

    Br J Neurosurg

    (2014)
  • Cited by (0)

    Conflict of interest statement: The authors declare that the article content was composed in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

    Christine Y. Mau and Victor M. Sabourin are co–first authors.

    View full text