High-grade arteriovenous malformations andtheir management

JOCN-190.QXD

08/12/01 2:09 PM

Page 37

Journal of Clinical Neuroscience (2002) 9(1), 37–40 © 2002 Harcourt Publishers Ltd DOI: 10.1054/jocn.2000.0927, available online at http://www.idealibrary.com on

Clinical study

High-grade arteriovenous malformations and their management Richard D. Ferch MBBS, Michael K. Morgan MDBS FRACS The North and West Cerebrovascular Unit, The University of Sydney, Sydney, Australia

Summary The aim of this study was to compare operatively and non-operatively managed high-grade arteriovenous malformations (AVMs) and to identify risk factors for surgical morbidity. Three hundred and ninety-one consecutively enrolled patients with AVMs were graded using the Spetzler Martin grading scheme. Forty-six of these patients had grade 4 or 5 AVMs. Twenty-nine patients underwent surgery and 17 were conservatively managed. During an average of 33 months follow-up the non-operative group experienced a decline in function in 27% of cases followed. These deteriorations were due to haemorrhage, progressive neurological deficits and seizures. In the surgical group completing treatment there was a mortality and morbidity impacting on self-care of 15%. In those without deep perforating arterial supply the morbidity was 10% and with deep perforating arterial supply or deep meningeal recruitment there was a combined morbidity and mortality of 44%. This difference in outcome was statistically significant (P:0.01). We conclude that high-grade AVMs have a high operative morbidity. However, these lesions often have a poor natural history and with careful selection (based on the presence or absence of deep perforating arterial supply) a group can be selected that benefits from surgery. Grade 4 and 5 AVMs with supply from lenticulostriate, choroidal, thalamic deep perforating arteries or deep meningeal recruitment may be best treated conservatively or possibly by multimodality treatment utilising radiotherapy and embolisation combined with surgery. © 2002 Harcourt Publishers Ltd Keywords: arteriovenous malformation, brain, surgery, complication

INTRODUCTION

Table 1

Patients harbouring high grade arteriovenous malformations (AVMs) provide difficult management problems. The natural history carries a risk of haemorrhage of 2–4% per year with an annual mortality of 1%.1–6 The annual risk of haemorrhage in the presence of intranidal aneurysms increases to 10% and with feeder aneurysms it increases to 7%.1,2 For the first year after a haemorrhage the risk of further haemorrhage is 6–18%.5,11 Curative treatment is associated with a risk from as low as 2.5% morbidity in AVMs less than 3 cm7 to considerably greater than this.8 The difficulties with surgical resection of AVMs have been recognized for a long time. In order to predict those at greater operative risk a number of grading schemes have been devised. The most widely accepted grading scheme is that proposed by Spetzler and Martin.9 Prospective analysis using this grading scheme has shown an increasing morbidity with increasing grades of AVM.10 This analysis suggested three distinct groups of patients with regard to outcome Grades 1–2, Grades 3, Grades 4–5.10 A further discriminator for grade 3 AVMs is the presence of deep perforators that determines whether they behave as lesser or greater grades.8 The aim of this study was to examine a prospectively enrolled group of patients with Spetzler-Martin grade 4–5 AVMs and attempt to identify discriminators that allow surgical and nonsurgical subgroups to be selected.

Modified Rankin score

Neurological condition

0 1 2

Normal Minor deficit not interfering with lifestyle Minor handicap but insufficient to restrict ability to self care Moderate handicap requiring some help to self care Moderately severe handicap preventing independence but not needing constant attention Severe handicap requiring constant attention Dead

METHODS Three hundred and ninety-one consecutively enrolled patients with AVMs were prospectively entered into a database between Received 29 September 2000 Accepted 2 April 2000 Correspondence to: M. K. Morgan, Department of Neurosurgery, Level 7 Royal North Shore Hospital, St Leonards NSW 2065, Australia. Tel.: ;61 2 99268756; Fax: ;61 2 94375172; E-mail: [email protected]

3 4 5 6

Modified Rankin score

1989 and September 2000. They were clinically and radiologically assessed and then graded according to the Spetzler Martin grading scheme. Forty-six of these patients were Spetzler Martin grade 4 or 5 (no grade 6 classifications were made). Of this group, 31 were offered surgical treatment (two patients refused and were placed in the observation group). Fifteen patients were assessed as having too high an operative risk and not offered surgery. They were treated with palliative embolization or observation alone. In total there were 17 in the non-surgical group. Factors taken into consideration in estimating the AVM natural history and surgical risk included age of the patient, the degree of neurological deficits, the history of haemorrhagic events, the presence of intranidal or feeding artery aneurysms, the number of deep perforating feeding arteries and the grade of the AVM. Deep perforating arterial supply was defined as supply from the lenticulostriate, thalamostriate, postero-choroidal arterial territories or from deep meningeal supply in cases where this had been recruited. Outcome was measured with the modified Rankin score at last follow-up and compared with that on referral (Table 1). Statistical analysis was completed with chi squared (␹2) or paired t-test probability tests. Significance was set at P⬍0.05. 37

JOCN-190.QXD

38

08/12/01 2:09 PM

Page 38

Ferch and Morgan Table 2 Case

Non-operative group Deep arterial supply

MRS on presentation

MRS at last follow-up

Change in MRS

1

Yes

3

2

;1

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

0 1 2 0 1 1 0 1 3 0 3 2 4 0 3 0

0 1 2 2 3 1 0 1 3 2 3 3 NA 0 NA 0

0 0 0 92 92 0 0 0 0 92 0 91

Cause of change

Duration of follow-up (months)

Recovery post haemorrhage

Steal Steal

Haemorrhage Seizures

0 0

45 1 100 12 12 26 15 15 1 30 81 7 81 0 54 0 21

Cases 14 and 16 were lost to follow-up.

Table 4

Table 3

Operative group

Case

Deep arterial supply

MRS on presentation

MRS at follow-up

None None None None None None None None None None Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

0 1 4 3 2 0 2 0 0 0 0 0 2 2 0 2 0 0 3 0 0 0 0 0 0 0

1 1 1 1 1 0 0 0 0 0 6 6 6 3 2 2 1 1 3 0 0 0 0 0 0 0

1 8 12 13 14 15 18 22 24 26 2 3 4 5 6 7 9 10 11 16 17 19 20 21 23 25

Presen- FollowMRS tation up change91

Change in MRS

91 0 ;3 ;2 ;1 0 ;2 0 0 0 96 96 96 91 92 0 91 91 0 0 0 0 0 0 0 0

RESULTS The results are summarized in Tables 2–4. Of the 17 patients in the non-operative group there were eight males and nine females with a mean age of 37. The operative group was composed of 13 males and 16 females and the mean age was 38. There was no significant difference between the two groups in terms of age or sex. All the AVMs in the non-operative group were supratentorial in location. Twenty-seven of the AVMs in the operative group were supratentorial and two were located in the posterior fossa. The mean AVM size in the non-operative group was 5.6 cm and in the operative group was 5.0 cm. This difference was not significant although there did appear to be a trend towards larger AVMs Journal of Clinical Neuroscience (2002) 9(1), 37–40

Outcome summary

Non-operative No deep group: n:15 supply: n:0

MRS: 0,1,2 MRS: 3,4,5,6

0 0

0 0

0 0

MRS: 0,1,2 MRS: 3,4,5,6

12 3

11 4

2 1

Operative No deep MRS: 0,1,2 group: n:26 supply: n:10 MRS: 3,4,5,6

8 2

10 0

2 1

Deep MRS: 0,1,2 supply: n:16 MRS: 3,4,5,6

15 1

11 5

4 0

Deep supply: n:15

One patient lost to follow up excluded. Two patients not completed therapy excluded. Three MRS (Modified Rankin score) improved due to recovery from presentation haemorrhage.

in the non-operative group. Similarly there was a tendency to higher Spetzler-Martin grade in the non-operative group than in the operative group, but this was not significant. The modified Rankin Score at presentation in the non-operative group was worse than in the operative group reflecting poorer grade patients but the difference was not significant. Forty-five per cent of patients in the operative group had a haemorrhagic event prior to presentation whereas 37% of patients in the non-operative group had a haemorrhagic event. Although there was a trend to operative treatment in patients with a haemorrhagic history the difference was not statistically significant. Of the 15 patients in the non-operative group who had an angiogram there was a total of five patients with aneurysms. In the operative group there were nine patients with aneurysms but this difference was not significant. Thirty-four per cent of the operative group had no deep perforating arterial or deep dural supply while all the patients in the non-operative group had deep perforating arterial supply and often from multiple territories. The difference between the groups was statistically significant (P ⬍ 0.01). Non-operative group Two patients in the non-operative group were lost to follow-up. The remaining 15 patients were followed from 1 to 100 months with the mean follow up period of 33 months. During this time there was a functional decline in four patients (27%). Two patients © 2002 Harcourt Publishers Ltd

JOCN-190.QXD

08/12/01 2:10 PM

Page 39

High-grade arteriovenous malformations 39

B

A

Fig. 1 (A) Axial MRI slice of a 31-year-old male presenting with headaches, papilloedema and visual cognitive impairment from this large arteriovenous malformation. (B) A-P angiogram demonstrating the marked meningeal feeding vessels supplying this AVM. Bleeding from these vessels proved fatal during surgery.

had gradually deteriorating function falling 2 points on their modified Rankin score related to ‘steal phenomenon’. One patient had a non-fatal haemorrhage resulting in a decline in function of 2 points and the other patient had recurrent seizures limiting independence. Two patients in the non-operative group were treated with embolization. However, complete ablation could not be achieved in either of them. Neither of these patients experienced a downgrade in their function related to either embolization or the natural history (Table 2). Operative group Pre-operative embolization was used, particularly in the early part of the series. Embolization was attempted if 975% reduction in blood flow was thought possible in a relatively surgical inaccessible arterial feeder and if amytal testing of the catheterized feeding vessel did not produce a neurologic deficit. All patients were commenced on a course of dexamethasone, ␤-blockade and anticonvulsant medication preoperatively and this was carried through the postoperative period. Following surgery the patients were managed in the ITU where strict blood pressure monitoring was used to control the systolic pressure between 90 and 120 mmHg. More recently, focused radiotherapy has been used in large AVMs on the component receiving deep feeding vessels in an effort to obliterate the deep supply and reduce operative morbidity on surgical resection. Three patients in the operative group are undergoing preoperative focused irradiation to the deeper portion of their AVM in preparation for surgery and have not had their surgery as yet. Of the remaining 26 patients there was a mortality of 11.5% with two deaths occurring intraoperatively (Figs 1A and B) due to blood loss and one occurring 36 months postoperatively from pancreatitis. There was a fall in neurological grade in five additional cases with four patients falling 1 grade and one patient falling 2 grades. The combined morbidity and mortality was 31%. If morbidity was confined to deficits that restricted lifestyle the incidence of serious morbidity and mortality was 15%. The outcome for patients who had surgical management was related to the presence of deep perforating arterial supply. In the 10 patients who did not have deep supply to the AVM, there was © 2002 Harcourt Publishers Ltd

one patient who dropped 1 point on their modified Rankin score representing a 10% morbidity. However, in the 16 patients with deep supply, there were two deaths related to intraoperative haemorrhage and one patient who had a modified Rankin score of 4 dying 3 years postoperatively from pancreatitis. Two of these deaths were patients with a deep meningeal supply in addition to a deep perforating contribution. An additional patient experienced normal perfusion pressure breakthrough and had delayed postoperative haemorrhage resulting in a mild hemiparesis and fall in modified Rankin score of 2 being restricted in lifestyle and ability to work. Three patients developed minor neurological deficits causing them to fall one point on their modified Rankin score. This represented a combined morbidity and mortality for surgery in the presence of deep perforating or deep meningeal arterial supply of 44%. This difference was significant (P:0.001). DISCUSSION Larger AVMs may have a more symptomatic natural history than smaller AVMs as they can cause neurological deterioration from chronic ischaemia associated with ‘steal’ as well as from haemorrhage and seizures. This deterioration can be slow and is usually relentless. Our non-operative patients experienced deterioration in 27% of cases. This incidence of neurological deterioration is higher than the 7% described by Brown in 19883 for all cases of non-operative AVMs and occurred over a shorter period of time. The relatively slow course of neurological deterioration associated with steal may lead to it being under-reported in some series. Our series looked at higher grade AVMs whereas other published series have included all AVMs and this may explain the difference in part. However, the small sample size in our study makes it difficult to draw too many conclusions from these observations on the natural history. In selecting patients who will benefit from surgery the patient’s natural history needs to be estimated and compared to the surgical risk. It is known that patients who have bled within the previous year have a risk of further haemorrhage in the order of 6–18%5,11 and this is dramatically higher than in the patient who has not bled. There was a tendency for patients who had a haemorrhagic Journal of Clinical Neuroscience (2002) 9(1), 37–40

JOCN-190.QXD

40

08/12/01 2:10 PM

Page 40

Ferch and Morgan

history to be offered surgery but the difference between the surgical and non-surgical groups was not significant. A distinguishing feature of those undergoing surgery was the absence of deep perforator or meningeal supply. The non-operated group all had deep perforators and significantly more than the operated group. Deep perforating arterial supply has been shown to be a significant indicator of higher surgical risk for Spetzler Martin grade 3 AVMs.8 Despite careful selection of patients with fewer deep perforators the combined series morbidity and mortality for surgery was still 31%. Excluding morbidity not restricting lifestyle the combined surgical morbidity and mortality remained at 15%. This is similar to previously published series.10 As deep perforating and meningeal arterial supply was a significant discriminator in surgical case selection it is likely that those with deep perforating and meningeal arterial supply chosen for surgery were considered more amenable to resection than many with deep arterial supply. However, even in these highly selected cases a combined morbidity and mortality of 44% occurred in comparison to a surgical morbidity and mortality of 10% in the absence of this blood supply. The operative morbidity in lesions with deep perforating or deep meningeal supply is significantly higher than in its absence. A theoretically appealing treatment regimen to reduce the operative morbidity has been made utilizing adjuvant therapy in the form of preoperative embolization and focused radiotherapy to the deep portion of the AVM. This is done in order to reduce blood supply to the AVM and to try and eliminate the deep perforating arterial supply reducing management morbidity. Embolization is not useful unless it can achieve at least 50% obliteration of the AVM and preferably greater than 75% obliteration.12 Embolization of superficial feeding vessels carries a risk of placing increased demand on deep feeding vessels with resultant enlargement and potential attenuation of wall thickness which may make surgery more difficult.13 All three deaths in our series were in patients who underwent preoperative embolization confirming that the lesions potential for haemorrhage was realised but embolization was not able to effect the desired result. It is likely that little is achieved in overall risk reduction unless embolisation can successfully target these deep vessels. Focused radiotherapy acts on AVMs to thicken blood vessels and cause thrombosis leading to obliteration of the lesion. The effect is delayed leaving the patient with the natural history of the lesion and carries a 3% risk of morbidity from the treatment in the form of radiation necrosis. This morbidity is dependent on the marginal dose as is the effectiveness of the treatment in obliterating the lesion. Focused radiosurgery is mainly efficacious in lesions under 3 cm where 85% obliteration can be achieved.14 It has been used in larger AVMs with follow-up surgery for the residual AVM with good results.15

Journal of Clinical Neuroscience (2002) 9(1), 37–40

CONCLUSIONS The combined morbidity and mortality rate in operating on Spetzler Martin grade 4 and 5 AVMs was 31% in our study. In view of this high rate careful attention needs to be placed on selecting which high grade AVMs need to be considered for surgery. In particular, the natural history of the particular AVM in question needs to be taken into account and the risk factors for a higher complication rate need to be taken into account. A subgroup of AVMs with a lower operative risk can be identified by the absence of deep perforating or deep meningeal arterial supply. A theoretically attractive management regimen for high grade AVMs with deep perforating or deep meningeal arterial supply may include multimodality treatment with focused radiosurgery used to treat the deep portion of the AVM supplied by the deep perforating arterial system, embolisation for the deep supply including deep meningeal arteries and surgical resection for the residual. REFERENCES 1. Brown RDJr, Wiebers DO, Forbes G. Unruptured intracranial aneurysms and arteriovenous malformations: frequency of intracranial haemorrhage and relationship of lesions. J Neurosurg 1990; 73: 859–863. 2. Redekop G, TerBrugge K, Montanera W et al. Arterial aneurysms associated with cerebral arteriovenous malformations: classification, incidence and risk of hemorrhage. J Neurosurg 1998; 89: 539–546. 3. Brown RDJr, Wiebers DO, Forbes G et al. The natural history of unruptured intracranial arteriovenous malformations. J Neurosurg 1988; 68: 352–357. 4. Crawford PM, West CR, Chadwick DW et al. Arteriovenous malformations of the brain: natural history in unoperated patients. J Neurol Neurosurg Psychiatry 1986; 49: 1–10. 5. Fults D, Kelly DLJr. Natural history of arteriovenous malformations of the brain: a clinical study. Neurosurgery 1984; 15: 658–662. 6. Ondra SL, Troupp H, George ED et al. The natural history of symptomatic arteriovenous malformations of the brain: a 24 year follow-up assessment. J Neurosurg 1990; 73: 387–391. 7. Pik J, Morgan MK. Microsurgery for small arteriovenous malformations of the brain: Results of 110 consecutive cases. Neurosurgery 2000; 47: 571–577. 8. Morgan MK, Drummond KJ, Grinnell V et al. Surgery for cerebral arteriovenous malformation: risks related to lenticulostriate arterial supply. J Neurosurg 1997; 86: 801–805. 9. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg 1986; 65: 476–483. 10. Hamilton MG, Spetzler RF. The prospective application of a grading system for arteriovenous malformations. Neurosurgery 1994; 34: 2–7. 11. Wilkins RH. Natural history of intracranial vascular malformations: a review. Neurosurgery 1985; 16: 421–430. 12. Vinuela F, Dion JE, Duckwiler G et al. Combined endovascular embolization and surgery in the management of cerebral arteriovenous malformations: experience with 101 cases. J Neurosurg 1991; 75: 856–864. 13. Morgan MK, Sundt TM Jr. The case against staged operative resection of cerebral arteriovenous malformations. Neurosurgery 1989; 25: 429–436. 14. Ogilvy CS. Radiation therapy for arteriovenous malformations: a review. Neurosurgery 1990; 26: 725–735. 15. Steinberg GK, Chang SD, Levy RP et al. Surgical resection of large incompletely treated intracranial arteriovenous malformations following stereotactic radiosurgery. J Neurosurg 1996; 84: 920–928.

© 2002 Harcourt Publishers Ltd