RESEARCH—HUMAN—CLINICAL STUDIES RESEARCH—HUMAN—CLINICAL STUDIES

Surgery for Unruptured Spetzler-Martin Grade 3 Brain Arteriovenous Malformations: A Prospective Surgical Cohort Michael Kerin Morgan, MD* Nazih Assaad, FRACS* Miikka Korja, PhD*‡ *Australian School of Advanced Medicine, Macquarie University, Sydney, NSW, Australia; ‡Department of Neurosurgery, Helsinki University Central Hospital, Helsinki, Finland Correspondence: Michael Kerin Morgan, MD, Department of Neurosurgery, Australian School of Advanced Medicine, Suite 201, 2 Technology Pl, Macquarie University, Sydney, NSW 2109, Australia. E-mail: [email protected] Received, January 5, 2015. Accepted, March 16, 2015. Published Online, April 9, 2015. Copyright © 2015 by the Congress of Neurological Surgeons.

BACKGROUND: There is uncertainty regarding the management of unruptured Spetzler-Martin grade 3 brain arteriovenous malformations (SMG3 ubAVM). OBJECTIVE: To analyze our series of patients treated by surgery. METHODS: A single-surgeon database of consecutively enrolled bAVMs (between 1989 and 2014) was analyzed. Adverse outcomes due to surgery were assigned within the first 6 weeks following surgery and outcome was prospectively recorded and assigned at the last follow-up visit by using modified Rankin Scale (mRS) score. RESULTS: Of the 137 reviewed patients, 112 (82%) were treated by surgery, 15 (11%) were treated elsewhere or by radiosurgery, and 10 (7%) were recommended for conservative management. Surgery for SMG3 ubAVM was associated with adverse outcomes with a new permanent neurological deficit of mRS .1 in 23 of 112 (21%) patients. Permanent neurological deficit leading to a mRS .2 from surgery was 3.6% (95% confidence interval, 1.1%-9.1%). Late recurrence of a bAVM occurred in 3 of 103 (2.9%) patients who had complete obliteration of bAVM confirmed immediately after surgery and who were subsequently later followed with radiological studies during the mean follow-up period of 3.0 years (range, 6 days to 18.8 years). CONCLUSION: When discussing surgical options for SMG3 ubAVM, a thorough understanding of the significance and incidence of adverse events and outcomes is required to fully inform patients. For our series, the additional subclassification of SMG ubAVM (based on variables contributing to the SMG or age) would not have been of use. KEY WORDS: Arteriovenous malformations, Brain, Cohort, Neurosurgery, Surgery, Treatment Neurosurgery 77:362–370, 2015

DOI: 10.1227/NEU.0000000000000774

T

he appropriate management of unruptured brain arteriovenous malformations (ubAVMs) is the subject of current intense interest and controversy.1-3 However, the risks of surgery and the application of risk to a particular subtype of brain arteriovenous malformation (bAVM) are difficult to establish from these studies. Grading systems to assist in estimating the risks of surgery have been developed and validated.4-8 The Spetzler-Martin 5-tier grading system (based on 1 point for size [largest diameter of nidus] less than 3 cm, 2 points for ABBREVIATIONS: CI, confidence interval; DSA, digital subtraction angiography; mRS, modified Rankin Scale; SMG, Spetzler-Martin grade; ubAVM, unruptured brain arteriovenous malformation

362 | VOLUME 77 | NUMBER 3 | SEPTEMBER 2015

www.neurosurgery-online.com

size from 3 to 6 cm, 3 points for size greater than 6 cm, and 1 additional point added for either the presence of deep venous drainage or location in eloquent brain [defined as motor cortex, primary sensory cortex, visual cortex, language cortex, internal capsule, diencephalon, brainstem, deep cerebellar nuclei, and cerebellar peduncle]) is well accepted and the constitutive variables independently contribute to the prediction of surgical risk.4,7,8 A review of reported series concluded that, for the most part, SpetzlerMartin grade 1 and 2 (SMG 1 and 2) bAVMs have a low surgical risk, and SMG 4 and 5 bAVMs are often considered inappropriate for surgery because of high surgical risk.8 The Lawton-Young extension of this grading system (adding to the SMG of 1 point for age in years less than 20, 2 points for age between 20 and

www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

SURGERY FOR UNRUPTURED SPETZLER-MARTIN GRADE 3 BAVMS

40, 3 points for age greater than 40, and an additional point for ubAVM with diffuse nidus) has provided nuance for the SMG and has been reported to be of greater discriminative value for surgical risk than the SMG alone.5,9 This creates a range of Lawton-Young grades for SMG3 ubAVMs of 5 to 8 of a possible 10 points. This corresponds to a reported wide range of adverse outcomes from surgery (or combined embolization and surgery) of between 20% and 63% for SMG3 ubAVM.9 Lawton-Young grades greater than 6 may be inappropriate for consideration of surgery.10 Adding to the surgical risk profile complexity, SMG3 bAVMs have been further subclassified by the perceived influence of variables contributing to the SMG.6 In this subclassification, the SMG3 of size less than 3 cm are considered to be of a lesser surgical risk, and larger SMG3 bAVMs located in eloquent cortex are considered to have a higher risk for the grade.6 With such a large spread of risk for SMG3 ubAVM depending on the variables examined, there exists considerable management doubt. This doubt arises from both the heterogeneity of the mixture of constitutive variables contributing to the grade, the lack of discrimination by outcome scales (eg, modified Rankin Scale),11 and the paucity of reports on surgical outcome. The aim of our study was to analyze our prospectively and consecutively entered database for surgical outcomes for SMG3 ubAVM.

METHODS Ethical Statement This study was approved by the Macquarie University Human Ethics Committee and was performed in accordance with institutional ethics committee guidelines.

Data Collection The data collection procedure has been described recently.4,12 In brief, since 1989, the senior author (MKM) has prospectively collected an AVM database including consecutively enrolled AVM patients with demographic, clinical, radiological, and treatment-related information (including preoperative embolization). Follow-up data include all referred patients (including patients not treated by surgery). Unruptured bAVMs were graded into the assigned SMG before surgery.7 The database included prospective decisions regarding the reasons for the recommended treatment for those not treated by surgery by the senior author (MKM). Patient-related variables included in the analyses were age, sex of patient, length of follow-up, and reason for management by treatment other than surgery. ubAVM-related variables that could be examined from the database included size, eloquent location, deep venous drainage, angiographic outcome, and late recurrence. In addition to demographic variables, clinical variables, and variables contributing to the definition of SMG, other anatomic variables were included in the analysis. These include infratentorial location; exclusively deep location (no anatomic presentation on a pial surface); indirect access (neither convexity surface supratentorial or occipital surface cerebellar location); choroidal or lenticulostriate arterial supply; and diffuseness. Diffuseness of the nidus was inconsistently recorded. However, diffuseness was included in the analysis for completeness.

NEUROSURGERY

All SMG3 ubAVMs were included. Excluded from analysis were SMG3 bAVMs presenting with hemorrhage and ubAVMs of other grades.

Outcome Assessments Adverse outcome due to surgery or preoperative embolization (including cases not proceeding to surgery) was assigned within the first 6 weeks of surgery or treatment. Outcome assessment was performed by using the modified Rankin Scale (mRS)11 administered preoperatively and at each follow-up visit. The mRS score at the last follow-up visit (or on the last review as an in-patient if no postoperative reviews occurred) was used for analyses. In addition to permanent neurological deficits, a near miss was considered to have occurred (irrespective of outcome mRS) in the following cases: a perioperative transfusion of 2.5 L or more; delayed intracranial hemorrhage in the resection bed leading to a neurological deterioration; reoperation; and intracranial hemorrhage during angiography or embolization. Surgical treatment was considered successful when a postoperative digital subtraction angiography (DSA) study showed no residual or new arteriovenous shunting. Studies at later dates, after the initial in-hospital postoperative DSA, were either magnetic resonance imaging/magnetic resonance angiography or DSA.

Statistical Analyses Mortality, morbidity, and near miss during the follow-up was calculated for all patients. For nonoperated patients, mortality and morbidity relating to their AVM was also reported. Morbidity between groups was compared by using x2 and Fisher tests. For continuous variables, comparison was made with Student t test. A P value ,.05 was considered significant. Statistical analysis was performed using Prism (version 6, GraphPad Software Inc).

RESULTS Study Cohort There were 137 referred patients between January 1989 and May 2014 with SMG3 ubAVMs. Table 1 provides a summary of the cohort. Follow-up Completeness of follow-up for surgically treated patients was 100%. The median follow-up time from surgery was 3.0 years (range, 6 days to 14.1 years). Fourteen of 25 nonoperated patients were seen on at least 1 follow-up visit. Mean follow-up time from referral was 6.3 years (2291 days) for nonoperated patients reviewed on more than 1 occasion and 3.4 years (1254 days) for those undergoing surgery (2.9 years [1065 days] from time of surgery) (Table 1). Obliteration Rate and Late Rebleeding Of the 112 operated ubAVM patients, 107 (96%) had complete obliteration on postoperative DSA following a single surgical procedure. Of the 5 patients with incomplete resections after initial surgery, 4 underwent early reoperation during the same hospital admission (3 leading to angiographic complete obliteration), and one was discharged with a small ubAVM residual. The in-hospital complete obliteration rate (including multiple surgeries in 4) was 98% (110 of the 112 patients). The patient with the

VOLUME 77 | NUMBER 3 | SEPTEMBER 2015 | 363

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

MORGAN ET AL

TABLE 1. Detailed Data on Patients With SMG3 ubAVMsa Characteristic

Total (137)

Operated (112)

Advised to have surgery but elected to be treated elsewhere or by radiotherapy (% of all SMG3 ubAVMs) Advised not to be treated due to risks of surgery (% of all SMG3 ubAVM) Advised not to be treated due to age (% of all SMG3 ubAVM) Patient factors Mean age, y (SD) [range] 35 (15) [5-84] 33 (13) [5-64] Females (%) 73 of 137 (53) 60 of 112 (54) ubAVM factors Mean largest diameter, cm (SD) 4.3 (1.5) 4.3 (1.5) Largest diameter ,3 cm, n (%) 8 (5.8) 6 (5.4) Largest diameter 3-6 cm, n (%) 120 (88) 97 (87) Largest diameter .6 cm, n (%) 9 (6.6) 9 (8.0) Deep venous drainage (%) 47 of 137 (34) 39 of 112 (35) Eloquent location (%) 92 of 137 (67) 72 of 112 (64) Infratentorial (%) 9 of 137 (7) 7 of 112 (6) Deep, no pial surface 8 of 137 (5.8) 5 of 112 (4.4) Indirect access (neither convexity or 49 of 137 (36) 39 of 112 (35) occipital surface of cerebellum) Choroidal or lenticulostriate supply 36 of 137 (26) 26 of 112 (23) Diffuse [data available for 55] (%) 25 of 55 (45) 21 of 47 (45) Presentation Seizure (%) 75 of 137 (55) 65 of 112 (58) Neurological incident or deficit 23 of 137 (17) 19 of 112 (17) (excluding migraine) Headache (including migraine) 14 of 137 (10) 11 of 112 (10) Incidental 25 of 137 (18) 17 of 112 (15) mRS 0–1 115 of 137 (84) 96 of 112 (86) mRS 2 17 of 137 (12) 14 of 112 (13) mRS .2 5 of 137 (4) 2 of 112 (2) Treatment Preoperative embolization 21 of 112 (19) Curative embolization 0 (0) Radiotherapy 3 of 137 (2) Outcome In hospital confirmed postoperative 110 of 112 (98) [93.3-99.9] angiographic obliteration (%) [95% CI] Days between referral and last follow-up 1369 (1542) [8-7388] 1254 (1333) [8-5147] (if seen on more than 1 occasion), mean (SD) [range] Days between surgery and last followna 1065 (1208) [6-5075] up, mean (SD) [range] Adverse outcomes Intraoperative blood loss receiving 36 of 112 (32) [24-41] transfusion of 2.5 L or more, or return to OR for hemorrhage in bAVM bed, or permanent new neurological deficit with mRS .1 (%) [95% CI] Intraoperative blood loss receiving 13 of 112 (12) [6.8-19.0] transfusion of 2.5 L or more, or return to OR for hemorrhage in bAVM bed, but mRS ,2 (%) [95% CI]

Nonoperated (25)

Operated vs Nonoperated (P Value)

15 of 137 (11)

7 of 137 (5) 3 of 137 (2)

45 (18) [12-84] 13 of 25 (52)

.001 ..99

4.2 (1.2) 2 (8) 23 (92) 0 8 of 25 (32) 20 of 25 (80) 2 of 25 (8) 3 of 25 (12) 10 of 25 (40)

.76

..99 .16 .67 .16 .65

10 of 25 (40) 4 of 8 (50)

.13 ..99

10 of 25 (40) 4 of 25 (16)

.12 ..99

3 of 25 (12) 8 of 25 (32) 19 of 25 (76) 3 of 25 (12) 3 of 25 (12)

.72 .08 .24 ..99 .04

0 (0) 3 of 25 (12)

2291 (2580) [9-7388]

.017

na

(Continues)

364 | VOLUME 77 | NUMBER 3 | SEPTEMBER 2015

www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

SURGERY FOR UNRUPTURED SPETZLER-MARTIN GRADE 3 BAVMS

TABLE 1. Continued Characteristic Adverse outcome due to surgery or embolization with last outcome mRS = 2 (%) [95% CI] Adverse outcome due to surgery or embolization with last outcome mRS 3 (%) [95% CI] Adverse outcome due to surgery or embolization with last outcome of death (%) [95% CI] Days between referral and first hemorrhage for 6 patients with hemorrhage after referral, mean (SD) [range] Adverse outcome from first hemorrhage after referral leading to a mRS .2 at last follow-up (%) [95% CI] Delayed radiological outcome Delayed MRI/MRA or DSA case number of all surviving surgery (mean days from surgery to last investigation) [SD of days to investigation] {range of days to investigation} New bAVM diagnosed after confirmed complete resection (%; 95% CI) [days between surgery and diagnosis of recurrent bAVM]

Total (137)

Operated (112) 19 of 112 (17) [11.1-25.1]

Nonoperated (25)

Operated vs Nonoperated (P Value)

3 of 112 (2.7) [0.6-7.9]

1 of 112 (0.9) [0.01-5.4]

na

na

1975 (1419) [14-3086]

na

na

5 of 25 (20%) [8.4-39.6]

103 of 111 (1199) [1173] {50-5075}

na

3 of 103 (2.9; 0.6-8.6) [457; 549; 1373]

na

a

CI, confidence interval; DSA, digital subtraction angiography; mRS, modified Rankin Scale; na, not applicable; MRA, magnetic resonance angiography; MRI, magnetic resonance imaging; OR, operating room; SD, standard deviation; SMG3 ubAVM, unruptured Spetzler-Martin grade 3 brain arteriovenous malformations.

small residual AVM was readmitted at a later time for further treatment, but residual arteriovenous shunting was not evident on the DSA at this time. Therefore, complete obliteration was eventually achieved through surgery in 99% (111 of 112 patients) of patients. One hundred three of the surviving operated patients were followed with either magnetic resonance imaging/magnetic resonance angiography or DSA for a delayed recurrence of ubAVM. The last follow-up study was performed at a mean of 3.3 years (1199 days) (Table 1) after surgery and found 3 of 103 (2.9%) patients with a recurrence of bAVM. Therefore, the long-term complete obliteration rate for all operated patients was known to be greater than 90% (100 of 111 surviving patients) but no greater than 97%. None of the operated patients experienced an early or late postoperative hemorrhage. Of the 14 nonoperated patients who were reviewed on more than 1 occasion (after the first consultation visit), 6 had a hemorrhage from the ubAVM during the follow-up period. Of these 6 patients, 5 had a permanent new neurological deficit with a mRS greater than 2. Operative Outcome Adverse outcome or near miss occurred in 32% (36 of 112; 95% confidence interval [CI] 24%-41%) (Table 1). Operative mortality

NEUROSURGERY

was 0.9% (1 of 112 patients). New permanent neurological deficits leading to a mRS greater than 1 occurred in 21% (23 of 112 patients; 95% CI, 14%-29%). Major morbidity (mRS greater than 2) including mortality was 3.6% (4 of 112 patients; 95% CI, 1.1%-9.1%) (Table 1). There was no significant difference between cases undergoing preoperative embolization and those who did not undergo preoperative embolization (Table 1). Outcomes did not change significantly over the study period (Figure) despite changes in management with the elimination of preoperative embolization. Influence of Characteristics Upon Outcome There was no significant difference in size of AVM or age of patient (as continuous variables) for patients who experienced an adverse outcome or near miss (Table 2). There was no significant difference in size of AVM or age of patient (as continuous variables) for patients with adverse outcomes with permanent new neurological deficit with mRS greater than 1 (Table 2). There was no significant difference in the risk of either adverse outcome or near miss, or those with adverse outcomes with permanent new neurological deficit with mRS greater than 1 for the following characteristics: size (as defined for SMG categories);

VOLUME 77 | NUMBER 3 | SEPTEMBER 2015 | 365

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

MORGAN ET AL

FIGURE. Cumulative embolization rate, complication leading to mRS .1 or near miss (hemorrhage into bed of bAVM, return to OR or transfusion of at least 2.5 L) (with 95% CI), and adverse outcome leading to mRS .2 (with 95% CI). CI, confidence interval; mRS, modified Rankin Scale; OR, operating room; SMG3 ubAVM, unruptured Spetzler-Martin grade 3 brain arteriovenous malformations.

deep venous drainage; eloquent location; age (in the LawtonYoung grading); infratentorial location; deep location (no pial surface); indirect access (neither convexity supratentorial surface or occipital surface of the cerebellum); choroidal or lenticulostriate arterial supply; diffuseness; or those undergoing preoperative embolization (Table 2). Change in mRS Over Time Irrespective of cause for the final mRS, a comparison was made between the presentation mRS and last follow-up mRS (Table 3). Of the 25 patients not undergoing surgery, there was subsequent follow-up on 14 which found that 8 were unchanged, 6 were worse (including 2 with mRS 1 and 1 death from hemorrhage) and none improved during the mean 6.3 years of following cases that could be followed. Of the 112 patients who underwent surgery, there was follow-up on all 112 to find 54 unchanged, 44 worse (including 23 with mRS 1 and 1 death from surgery), and 14 improved during the mean 3.4 years from referral to last follow-up. The comparison of those with known improvements between those having surgery (12.5%) and those having no surgery (0%) was not statistically different (P = .36).

DISCUSSION The results of this series of patients undergoing surgery for SMG3 ubAVM suggest that complete long-term obliteration can be achieved in more than 90% of patients with 1% mortality and 3% nonfatal major morbidity (mRS greater than 2), and 21% overall combined morbidity and mortality (mRS greater than 1)

366 | VOLUME 77 | NUMBER 3 | SEPTEMBER 2015

rates. The long-term recurrence rate was low (3% of those having delayed surveillance). Our previous results suggest that the most significant predictors for recurrence are deep venous drainage combined with the age less than 20 years.13 Our results are similar to other surgical series reporting combined ruptured and unruptured outcome results of SMG3 bAVM (eg, 18% negative outcomes combining series by Spetzler and Ponce9). Although outcome comparisons between different cohorts can only be made with caution, the magnitude of the difference in outcome of the current series with that of the treatment arm (including treatment by embolization, radiosurgery and/or surgery) of the ARUBA trial3 or the SIVMS long-term population-based followup study1 is noteworthy. In the ARUBA trial, 16 (57%) of 28 treated patients with SMG3 ubAVM died or had strokes.3 In the SIVMS study, 38 (37%) of 103 bAVM patients (of which 29 had SMG3 ubAVM) in the treatment arm died or had strokes related to their ubAVM or treatment.1 The obliteration rate is in accordance with the results of a recent review, which summarized that long-term obliteration can be achieved in 96% of the patients with surgery.14 However, when considering the broader selection of management options for all ubAVMs, as in the SIVMS population-based study,1 our study compares favorably (SIVMS obliteration rate for ubAVMs of 56%1). With respect to variables or characteristics that might influence outcomes (including size; patient age; deep venous drainage; eloquent location; infratentorial location; deep location [no pial surface]; indirect access [neither convexity supratentorial surface or occipital surface of the cerebellum]; choroidal or lenticulostriate arterial supply; diffuseness; or preoperative embolization), we

www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

NEUROSURGERY

TABLE 2. Data on Characteristics and Outcomes From Surgery for SMG3 ubAVMsa

Total With With Adverse Characteristic Outcome

a

Adverse Adverse Outcome Adverse Outcome or Near or Near Miss Outcome (mRS P Miss with Without P .1) With Value Characteristic Characteristic Value Characteristic

36

36 (13) [13-64] 32 (13) [5-64]

.14

36

4.6 (2.1)

.15

23

36 (13) [13-64] 33 (13) [5-64]

.34

23

4.5 (2.2)

.46

112 21 55 36 60 6 97 9 39 72 7 21

4.1 (1.1)

4.2 (1.3)

36 of 112 4 of 21 19 of 55 13 of 36 20 of 60 2 of 6 29 of 97 5 of 9 13 of 39 20 of 72 3 of 7 8 of 21

(32) (19) (35) (36) (33) (33) (30) (74) (33) (28) (40) (38)

32 of 91 17 of 57 23 of 76 16 of 52 34 of 106 7 of 15 31 of 103 23 of 73 16 of 40 33 of 105 28 of 91

(35) (30) (30) (31) (32) (47) (30) (32) (40) (31) (31)

.20 .69 .67 .84 ..99 .24 .14 .84 .21 .68 .61

23 of 112 3 of 21 12 of 55 8 of 36 15 of 60 2 of 6 18 of 97 3 of 9 9 of 39 13 of 72 3 of 7 5 of 21

(21) (14) (22) (22) (25) (33) (19) (33) (23) (18) (40) (24)

Adverse Outcome (mRS .1) Without Characteristic

20 of 91 11 of 57 15 of 76 8 of 52 21 of 106 5 of 15 20 of 103 14 of 73 10 of 40 20 of 105 18 of 91

P Value

(22) (19) (20) (15) (20) (33) (19) (19) (25) (19) (20)

.56 .82 .80 .25 .60 .19 .39 .63 .47 .17 .77

5 39

2 of 5 (40) 15 of 39 (38)

34 of 107 (32) 21 of 73 (29)

.66 .54

1 of 5 (20) 11 of 39 (28)

22 of 107 (21) 12 of 73 (16)

..99 .15

26

8 of 26 (31)

28 of 86 (33)

..99

6 of 26 (23)

17 of 86 (20)

.78

21

6 of 21 (29)

10 of 26 (38)

.55

3 of 21 (14)

8 of 26 (31)

.30

mRS, modified Rankin Scale; SD, standard deviation; SMG3 ubAVM, unruptured Spetzler-Martin grade 3 brain arteriovenous malformations.

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

SURGERY FOR UNRUPTURED SPETZLER-MARTIN GRADE 3 BAVMS

VOLUME 77 | NUMBER 3 | SEPTEMBER 2015 | 367

Continuous characteristics Mean age in years (SD) [range] for adverse outcome or near miss Mean largest diameter in cm (SD) for adverse outcome or near miss Mean age in years (SD) [range] for adverse outcome with mRS .1 Mean largest diameter in cm (SD) for adverse outcome with mRS .1 Categorical characteristics All cases Age ,20 y Age 20-40 Age .40 y Female Size ,3 cm Size 3-6 cm Size .6 cm Deep venous drainage Eloquent location Infratentorial Preoperative embolization Deep, no pial surface Indirect access (neither convexity or occipital surface of cerebellum) Choroidal or lenticulostriate supply Diffuse (data available for 47)

Without Adverse Outcome

MORGAN ET AL

TABLE 3. Change in mRS for SMG3 ubAVMsa Cases (#) Surgery

a

No

25

Yes

112

Referral mRS

FU mRS 0

FU mRS 1

FU mRS 2

FU mRS 3

FU mRS 4

FU mRS 5

FU mRS 6

NA as Only 1 Consult

0 1 2 3 4 0 1 2 3

5 0 0 0 0 41 7 3 1

2 1 0 0 0 23 5 2 0

1 0 0 0 0 9 8 8 1

2 0 0 1 0 2 0 1 0

0 0 0 0 1 0 0 0 0

0 0 0 0 0 0 0 0 0

0 0 0 1 0 1 0 0 0

7 1 3 0 0 0 0 0 0

FU, follow-up; mRS, modified Rankin Scale; NA, not applicable; SD, standard deviation; SMG3 ubAVM, unruptured Spetzler-Martin grade 3 brain arteriovenous malformations.

found none of significant difference. This differs from studies of Lawton and colleagues who were able to discern a difference in outcomes depending on the mixture of variables constituting the SMG.5,9 This difference in the importance of these characteristics may be due to the a priori selection of cases based on the SMG variables and the small number of cases in our series (only 6 operated cases with size less than 3 cm) or the inadequate examination of variables (eg, diffuseness). Although, if found, a significant difference with these variables would be important, the absence of a difference does not preclude their effect on selection and outcomes. This is because these variables, to be examined adequately, should be examined in the total cohort of cases without SMG prior classification. From the variables we examined, unlike other studies, we could not find support that either bAVM size as a continuous variable or patient age within the SMG35,9 impacted outcomes. Although the small size of our cohort may contribute to this result, our finding is consistent with our previous study that found the adjusted odds ratio for eloquent location, deep venous drainage, and size (when categorized by SMG size) were similar.4 This would predict the combination of variables contributing to a SMG3 bAVM would likely sum to broadly similar risk of surgery irrespective of how the varying combinations contributed to the SMG3. Diffuseness was inadequately examined by this study. However, with regard to diffuseness, it should be noted that our larger cohort contributed to the validation study that concluded diffuseness was an important consideration.4 Our results raise the question as to whether surgery, in general, should be considered the preferred management option for SMG3 ubAVMs, particularly for cases with a low likelihood of obliteration within a reasonable time frame by radiosurgery. We do not have adequate volume studies to truly compare the likelihood of obliteration by radiosurgery. Kano and colleagues,15 examining their experience with radiosurgery for SMG3 bAVM, had a 3-year obliteration rate approximating 60% for bAVM nidus less than 3 cm and 40% for larger nidus size with a cumulative hemorrhage rate of 5.5%, a mortality rate of 4.2% (from hemorrhage), and

368 | VOLUME 77 | NUMBER 3 | SEPTEMBER 2015

a 2.7% incidence of permanent symptomatic adverse radiation effect. However, the majority of treated cases had previously ruptured and this would suggest that the mortality rate from hemorrhage would be considerably less than this for ubAVM. The obliteration rate at 5 years was considerably higher than at 3 years, but the number at risk in the Kaplan-Meier analysis was fewer than 10% and the number of repeat treatments was not mentioned and, therefore, the longer time frame data are less reliable. Only 6% of our cases were smaller than 3 cm, suggesting that radiosurgery was probably not ideal in terms of time to obliteration in the majority of cases. Given the significant event rate of adverse outcome or near miss (including blood transfusions of 2.5 L or more, hemorrhage into the bAVM resection bed, hemorrhage from embolization, or return to the operating room) of 32%, and 3.6% for outcomes resulting in mRS greater than 2, consideration should be given to recommending radiosurgery for SMG3 ubAVM with a high likelihood of obliteration in a short period of time. Although an argument can also be made for conservative management, the potential for cases thus managed deteriorating with time (approximately a quarter of our cases over 6 years) needs to be considered. The use of preoperative embolization changed significantly during the 25 years of the cohort. The incidence of 70% during the first 5 years of embolization declined to 0% during the last 5 years. There was no significant corresponding change to the cumulative adverse outcome rates over this period (Figure). The elimination of embolization for SMG3 ubAVM reduces the complexity and overall cost of management.16 We have not used embolization as a primary strategy for obliteration in SMG3 ubAVM or a strategy for reducing nidus size and subsequent treatment by radiosurgery. The impact of these management strategies may have contributed to the poor result from treatment in the ARUBA study.3 However, the role of strategic and limited embolization used as a prelude to surgery differs significantly from an embolization goal of maximal ablation to cure (or reduction of the nidus to a minimum suitable for targeted radiosurgery). There are likely to be fewer adverse outcomes from embolization

www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

SURGERY FOR UNRUPTURED SPETZLER-MARTIN GRADE 3 BAVMS

in the former approach where risk sharing between the surgery and embolization teams should be carefully considered. This is reflected in our series. There was no significant difference between cases embolized and those not embolized, with the most significant variable impacting the decision to embolize being time from first surgery. This result may only be achieved after considerable experience of surgery and specific surgical techniques that eliminate the need for embolization. Limitations The current study has limitations. First, mRS is not an objective outcome measure, and it does not necessarily take into account nonneurological and minor complications. Second, we cannot exclude referring, patient selection, and outcome reporting biases. However, there may also be some advantages over previously published studies. Completeness of follow-up was 100%. Moreover, our cohort included more patients with SMG3 ubAVM than any previous study. Importantly, 82% of the patients were treated with surgery, and, therefore, the results are not confused with other management strategies that may combine treatments.

CONCLUSION The current study shows the high reliability of robustly eliminating SMG3 ubAVM by surgery. However, it is likely, for SMG3 ubAVM suitable for radiosurgery and where there is an expected a high obliteration rate, that radiosurgery may be a preferred option because of the adverse outcome or near miss rate from surgery. The dilemma for the treating clinicians and their patients with SMG3 ubAVM remains whether to decide on surgical treatment or the natural history of untreated ubAVMs when radiosurgery is not ideal. Disclosures Dr Korja was supported in part by grants from the Sigrid Juselius, Biomedicum Helsinki, Orion-Farmos Research, Instrumentarium research and from the Finnish Medical Association. Drs Morgan and Assaad have no financial support from grants. The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.

REFERENCES 1. Al-Shahi Salman R, White PM, Counsell CE, et al. Outcome after conservative management or intervention for unruptured brain arteriovenous malformations. JAMA. 2014;311(16):1661-1669. 2. Korja M, Hernesniemi J, Lawton MT, Spetzler RF, Morgan MK. Is cerebrovascular neurosurgery sacrified on the altar of RCTs? Lancet. 2014; 384(9937):27-28. 3. Mohr JP, Parides MK, Stapf C, et al. Medical management with or without interventional therapy for unruptured brain arteriovenous malformations (ARUBA): a multicentre, non-blinded, randomised trial. Lancet. 2014;383 (9917):614-621. 4. Kim H, Abla AA, Nelson J, et al. Validation of the supplemented Spetzler-Martin grading system for brain arteriovenous malformations in a multicenter cohort of 1009 surgical patients. Neurosurgery. 2015;76(1):25-33. 5. Korja M, Bervini D, Assaad N, Morgan MK. Role of surgery in the management of brain arteriovenous malformations: prospective cohort study. Stroke. 2014;45(12): 3549-3555.

NEUROSURGERY

6. Lawton MT, Kim H, McCulloch CE, Mikhak B, Young WL. A supplementary grading scale for selecting patients with brain arteriovenous malformations for surgery. Neurosurgery. 2010;66(4):702-713. 7. Lawton MT; UCSF Brain Arteriovenous Malformation Study Project. SpetzlerMartin Grade III arteriovenous malformations: surgical results and a modification of the grading scale. Neurosurgery. 2003;52(4):740-748. 8. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg. 1986;65(4):476-483. 9. Spetzler RF, Ponce FA. A 3-tier classification of cerebral arteriovenous malformations. Clinical article. J Neurosurg. 2011;114(3):842-849. 10. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J. Interobserver agreement for the assessment of handicap in stroke patients. Stroke. 1988;19(5): 604-607. 11. Bervini D, Morgan MK, Ritson EA, Heller G. Surgery for unruptured arteriovenous malformations of the brain is better than conservative management for selected cases: a prospective cohort study. J Neurosurg. 2014;121(4):878-890. 12. Morgan MK, Patel NJ, Simons M, Ritson EA, Heller GZ. Influence of the combination of patient age and deep venous drainage on brain arteriovenous malformation recurrence after surgery. J Neurosurg. 2012;117(5):934-941. 13. Bradac O, Charvat F, Benes V. Treatment for brain arteriovenous malformation in the 1998-2011 period and review of the literature. Acta Neurochir (Wien). 2013; 155(2):199-209. 14. Kano H, Flickinger JC, Yang HC, et al. Stereotactic radiosurgery for Spetzler-Martin Grade III arteriovenous malformations. J Neurosurg. 2014;120(4):973-981. 15. Morgan MK, Davidson AS, Koustais S, Simons M, Ritson EA. The failure of preoperative ethylene-vinyl alcohol copolymer embolization to improve outcomes in arteriovenous malformation management: case series. J Neurosurg. 2013;118(5): 969-977. 16. Berman MF, Hartmann A, Mast H, et al. Determinants of resource utilization in the treatment of brain arteriovenous malformations. AJNR Am J Neuroradiol. 1999;20(10):2004-2008.

COMMENT

T

he Spetzler-Martin1 and revised Spetzler-Ponce2 grading systems were devised to predict surgical risk and offer a guide in the decisionmaking process for the best management of cerebral arteriovenous malformations (AVMs). The Spetzler-Martin grade 3 or Spetzler-Ponce grade B AVMs represent a complex and heterogenous group. The management of these cases must be highly individualized and typically involves a multimodality approach including microsurgery, endovascular embolization, and radiosurgery. The publication of the ARUBA trial,3 despite its limitations,4,5 has heightened the controversy regarding the proper management of patients with unruptured AVMs. One such controversy was an ill-defined and ineffective treatment arm. However, given the high risk of stroke or death shown in medically managed patients, these patients need a safe and effective cure. Therefore, this article is a welcome addition to the scarce literature evaluating treatment strategies. The authors report on a relatively large cohort, for a subclassification of a rare lesion, of Spetzler-Martin grade 3 AVMs managed either conservatively, microsurgically, or microsurgically with preoperative embolization. In this series, they report a low mortality and serious morbidity rate for which they are to be congratulated. However, the rate of minor morbidity resulting in a mRS greater than 1 was not low at 21%. This point, along with the relatively high rate of poor outcomes for nonoperated patients, underscores the need for continued advancement to optimally manage these difficult lesions. Of interest, the authors during the study time period adopted a strategy of foregoing preoperative embolization in all cases. In our practice, preoperative embolization is performed routinely with the goal of facilitating microsurgical resection. We do not believe that partial embolization or embolization with the goal of cure are prudent treatment strategies because they incur significant risk without justifiable benefit. The authors should be commended for maintaining such

VOLUME 77 | NUMBER 3 | SEPTEMBER 2015 | 369

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

MORGAN ET AL

a well-kept database, the safe and effective management of such complex cases, and the reporting of their results. Rami O. Almefty Robert F. Spetzler Phoenix, Arizona 1. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg. 1986;65(4):476-483.

370 | VOLUME 77 | NUMBER 3 | SEPTEMBER 2015

2. Spetzler RF, Ponce FA. A 3-tier classification of cerebral arteriovenous malformations. Clinical article. J Neurosurg. 2011;114(3):842-849. 3. Mohr JP, Parides MK, Stapf C, et al. Medical management with or without interventional therapy for unruptured brain arteriovenous malformations (ARUBA): a multicentre, non-blinded, randomised trial. Lancet. 2014;383 (9917):614-621. 4. Korja M, Hernesniemi J, Lawton MT, Spetzler RF, Morgan MK. Is cerebrovascular neurosurgery sacrificed on the altar of RCTs? Lancet. 2014;384(9937):27-28. 5. Russin J, Spetzler R. Commentary: the ARUBA trial. Neurosurgery. 2014;75(1): E96-E97.

www.neurosurgery-online.com

Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited

Surgery for Unruptured Spetzler-Martin Grade 3 Brain Arteriovenous Malformations: A Prospective Surgical Cohort.

There is uncertainty regarding the management of unruptured Spetzler-Martin grade 3 brain arteriovenous malformations (SMG3 ubAVM)...
520KB Sizes 0 Downloads 8 Views