Sie sind auf Seite 1von 14

REVIEWS

Full text, figures, tables, and references are available online at www.neurosurgery-online.com

Bradley A. Gross, M.D.


BRAINSTEM CAVERNOUS MALFORMATIONS
Department of Neurological Surgery,
Feinberg School of Medicine and BRAINSTEM CAVERNOUS MALFORMATIONS (CMs) continue to present a consider-
McGaw Medical Center,
Northwestern University, able source of controversy in the neurosurgical community, with an accumulating vol-
Chicago, Illinois, and ume of literature detailing their natural history and their surgical and radiosurgical man-
Division of Neurosurgery, agement. As part of a systematic review of the literature, 12 natural history studies, 52
Evanston Northwestern Healthcare,
Evanston, Illinois surgical series, and 14 radiosurgical series were tabulated. Annual bleeding rates for
brainstem CMs ranged from 2.3% to 4.1% in natural history studies and from 2.68% to
H. Hunt Batjer, M.D. 6.8% in surgical series before intervention. Rebleed rates as high as 21.5% in natural
Department of Neurological Surgery, history studies and greater than once per year in surgical series were reported. A total
Feinberg School of Medicine and of 684 of 745 CMs (92%) were documented as completely resected in 46 series that
McGaw Medical Center,
Northwestern University,
provided specific information on resection rates. Early postoperative morbidity ranged
Chicago, Illinois, and from 29% to 67% in larger surgical series, although it was often transient. Thirty of 61
Division of Neurosurgery, partially resected lesions rebled; 4 of these rebleeds were fatal. Twelve additional patients
Evanston Northwestern Healthcare,
Evanston, Illinois
died from surgically related causes for a combined postoperative rebleeding and sur-
gically related mortality rate of 1.9%. Across 45 series (683 patients), 85% of patients
Issam A. Awad, M.D. were reported as the same or improved, 14% were worse, and 1.9% died from surgi-
Department of Neurological Surgery, cally related causes at long-term follow-up. Patients with anterolateral pontine lesions
Feinberg School of Medicine and generally appeared to have a better functional recovery, whereas those requiring exci-
McGaw Medical Center,
Northwestern University,
sion via the floor of the fourth ventricle had relatively worse long-term outcomes.
Chicago, Illinois, and Radiosurgical series demonstrated conflicting data; some reported a statistically signif-
Division of Neurosurgery, icant decrease in CM rebleeding rates after 2 years, whereas others did not, partially
Evanston Northwestern Healthcare,
Evanston, Illinois
related to dosimetry. Postradiosurgical morbidity was nonetheless significantly greater
for CMs than for arteriovenous malformations.
Bernard R. Bendok, M.D. Please refer to the online version of this article for a more detailed natural history and
Department of Neurological Surgery, radiosurgical review and a discussion of brainstem CM clinical presentation and diag-
Feinberg School of Medicine and
McGaw Medical Center,
nostic evaluation.
Northwestern University, KEY WORDS: Angioma, Brainstem, Cavernous malformation, Cranial base approach, Natural history,
Chicago, Illinois, and Radiosurgery, Skull base approach, Surgery
Division of Neurosurgery,
Evanston Northwestern Healthcare, Neurosurgery 64:805–818, 2009 DOI: 10.1227/01.NEU.0000343667.14177.72 www.neurosurgery-online.com
Evanston, Illinois

Reprint requests:

A
H. Hunt Batjer, M.D., lthough the first reported surgical expe- their ultimate diagnosis after further debilitat-
Department of Neurological Surgery, rience with brainstem cavernous mal- ing or even lethal hemorrhage (92). However,
Feinberg School of Medicine, formations (CMs) dates to 1928, when with widespread use of magnetic resonance
Northwestern University, Dandy evacuated a brainstem hematoma imaging (MRI) and the accumulation of numer-
676 North St. Clair Street,
Suite 2210,
attributed to CM hemorrhage (18), studies of ous case reports and surgical series, awareness
Chicago, IL 60611. the natural history and management of these and a progressively more aggressive attitude
Email: h-batjer lesions have only begun to accumulate since toward these lesions has arisen in the literature.
the late 1980s. In the interim, many patients In this article, we comprehensively review nat-
Received, May 28, 2008. with angiographically occult vascular malfor- ural history, surgical, and radiosurgical studies
Accepted, November 5, 2008. mations (AOVMs) of the brainstem were diag- of these lesions, synthesizing clear recommen-
Copyright © 2009 by the
nosed with demyelinating disease (36, 63, 89), dations and highlighting the nuances of partic-
Congress of Neurological Surgeons neoplastic disease (65, 74, 75, 92), infection (36), ular surgical approaches to these malforma-
or brainstem infarction (36), only to receive tions.

ABBREVIATIONS: AOVM, angiographically occult vascular malformation; CM, cavernous malformation;


CSF, cerebrospinal fluid; MRI, magnetic resonance imaging

NEUROSURGERY VOLUME 64 | NUMBER 5 | MAY 2009 | 805


TABLE 1. Review of cavernous malformation natural history studies including lesions of all locationsa
No. of Factors Factors with
Series Lesion Study design Bleed Bleed Subdivided
patients increasing no effect
(ref. no.) distribution and follow-up definition rates bleed rates
(lesions) bleed rate on bleed rate
GROSS ET AL.

Aiba et al., 110 Given for hemorrhagic Prospective New Sx ⫹ Given 0.39%/pt-y if seizures Prior hem, age Location
1995 (1) (110) lesions: 53% lobar Mean, 4.7 y intralesional or subdivided 22.9%/pt-y if prior hem ⬍40 y, F sex
24% BS perilesional bleed 0%/pt-y if 21.5%/pt-y if BS 40 y, F sex
11% BG/T incidental 34%/pt-y if ⬍40 y and F
8% cerebellar
3% callosal
Cantu et al., 133 (133) 55% lobar Retrospective “MRI 1.7%/pt-y 2.33%/pt-y if BS Nonlobar Age, sex
2005 (14) 19% BS Mean, 5 y documented” location

806 | VOLUME 64 | NUMBER 5 | MAY 2009


11% other ⫹ new Sx
10% supratent deep
6% cerebellar
Del Curling et al., 32 (76) 86% supratent Retrospective Clinically significant, 0.25%/pt-yb
1991 (19) 14% infratent 1195 pt-y “radiographically 0.1%/les-yb
2890 les-y identifiable”
Ghannane et al., 39 (39) 51% supratent Retrospective 6.27%/pt-y Infratentorial
2007 (29) 26% infratent Mean, 3.4 y location
15% “multiple”
8% spinal
Kim et al., 62 (108) For solitary lesions: Retrospective Clinical (new Sx) 1.4%/les-yb Prospective Prior hem
1997 (41) 63% lobar 2509.6 pt-y; 2.3%/pt-yb rebleeding:
20% BS Prospective 3.8%/pt-y
10% BG/T Mean, 1.9 y for 2.6%/les-y
6% cerebellar rebleed rates
Kondziolka et al., 122 (Referred patients): Retrospective Gross bleeding Prospective: Prospective: Prior hem Location, sex
1995 (42) 48% lobar 4550.6 pt-y; on MRI 2.6%/pt-y; 0.6%/pt-y if incidental
35% BS Prospective Retrospective: 4.5%/pt-y if prior hem
17% BG/T Mean, 2.8 y 1.3%/pt-yb 2.4% if BS 5%
if BS with prior hem
Labauge et al., 40 (232) 76% supratent Familial Radiographic only 16.5%/pt-y Infratentorial Age, size
2000 (44) 24% infratent Retrospective (1 of 3 symptomatic) 2.5%/les-y
11% BS Mean, 3.2 y
127 pt-y
844 les-y
Labauge et al., 33 (234) 86% supratent ASx, familial MRI bleeding 4.3%/pt-y
2001 (43) 14% infratent Prospective (1 of 3 symptomatic) 0.6%/les-y
Mean, 2.1 y
70 pt-y
Moriarity et al., 68 (228) 66% lobar Prospective “Extralesional hemor- 3.1%/pt-y 4.2%/pt-y if F F sex Prior hem,
1999 (55) 14% BS Mean, 5.2 y rhage accompanied 0.9%/pt-y if M location, size
8% BG/T 352.9 pt-y by a change in 5.2%/pt-y if superficial
7% other clinical status” 3.1%/pt-y if deep
5% cerebellar

www.neurosurgery-online.com
BRAINSTEM CAVERNOUS MALFORMATIONS

NATURAL HISTORY
on bleed rate
Factors with
no effect

Four percent to 35% of CMs (mean, 15% [Table 1]) are located

BS, brainstem; BG, basal ganglia; T, thalamus; Sx, symptoms; pt-y, patient-year; hem, hemorrhage; supratent, supratentorial; MRI, magnetic resonance imaging; infratent, infratentorial; les-y, lesion-year; ASx,
in the brainstem. Despite the wide range of CM bleed rates
Sex

across our review of natural history studies (Table 1), the range
narrows to 0.7% to 3.1% per patient-year among nonfamilial,
prospective studies that defined hemorrhage as a symptomatic,
radiographically verified extralesional bleed (55, 70); these
deep location
increasing
bleed rate
Factors

Prior hem,c

studies also demonstrated an increased bleed rate among


Age, size
females (55, 70). Although CM size does not affect the bleeding
rate (43, 55, 70), 4 of 5 studies (1, 41, 42, 55, 64) demonstrated
prior hemorrhage as a significant risk factor for subsequent
bleeding, citing annual rebleed rates ranging from 3.8% (41) to
10.6%/pt-y if BS/BG/T

22.9% (1). Although the latter study’s inclusion of intralesional


0%/pt-y if superficial
4.1%/pt-y if BS/BG/T

0%/pt-y if superficial
Subdivided
bleed rates

Overall, 4.2%/pt-y

or perilesional bleeding likely elevates reported hemorrhage


rates, the potential for hemorrhage clustering and aggressive
Including ASx
Event rates:

lesion behavior is well documented in the literature (6, 65, 86).


13%/pt-y
2%/les-y
bleeds:

Barker et al. (6) reported a decrease in monthly bleed rates


F sex

from 2.1% to 0.8% after 28 months, an important concept to


consider when reviewing radiosurgical results.
Bleed rates for brainstem CMs range from 2.33% to 4.1% per
patient-year in natural history studies (14, 42, 55, 64) and from
Bleed
rates

Patients with a presentation of radiographic hemorrhage and a focal neurological deficit had increased hemorrhage rates (P ⫽ 0.03).

2.68% to 6.8% per patient-year before intervention in surgical


1.1%/les-y
0.7%/les-y
1.6%/pt-y

6.5%/pt-y

series (12, 25, 26, 34, 57, 63, 73, 89). Most studies with more lax
definitions of CM hemorrhage (such as a change in symptoms
alone) reported increased hemorrhage rates among brainstem
Acute/subacute blood out-

lesions (14, 22, 43, 64), underscoring the clinical significance of


objective worsening of Sx

New Sx ⫹ radiographic
outside ring or increase

side ring, prior hem on


20% (all symptomatic)

mere morphological changes such as thrombosis and intrale-


Acute/subacute blood

Event ⫽ subjective ⫹

LP, fresh clot outside

sional or perilesional bleeding in the brainstem. Porter et al. (64)


bleeding (includes
definition

in lesion size by
Bleed

reported a significantly greater “clinical event rate” of 10.6% per


ring at surgery

intralesional)

patient-year for deep lesions. Kondziolka et al. (42) and Aiba


et al. (1) reported brainstem CM rebleed rates of 5% and 21.5%,
respectively. Rebleed rates ranged from 17.7% per patient-year to
greater than 1 rebleed per patient-year before intervention in
Prospective and

surgical series (12, 73). Hemorrhage clustering (6, 89) may con-
and follow-up
Study design

retrospective

tribute to these increased rates, whereas early intervention and


Mean, 3.8 y

Mean, 2.2 y
Mean: 2.2 y

Prospective
Prospective

143 les-y

282 les-y

selection bias likely inflate rebleed rates in surgical series.


427 pt-y

Familial

46 pt-y

SURGICAL APPROACHES AND RESULTS


A total of 52 surgical series and 821 brainstem CMs were
5% deep supratent
distribution

reviewed (Table 2). Across 46 series providing data on brain-


Rates calculated assuming lesion presence since birth.
12% cerebellar

4% cerebellum
Lesion

6% cerebellar

stem CM resection rates, 684 of 745 CMs (92%) were completely


4% callosal
55% lobar

72% lobar

83% lobar

resected. Follow-up for 10 partially resected lesions was not


7% BG/T

5% BG/T
30% BS

11% BS

4% BS

reported; however, 30 of the remaining 51 CMs rebled, result-


ing in 4 deaths. Four of these lesions were considered “recur-
rences” after complete resection (12, 62, 63, 97). To curb the
asymptomatic; LP, lumbar puncture.
(lesions)
patients
No. of

31 (128)

possibility of complicating residual CM with recurrence, intra-


57 (66)

operative ultrasound may be useful to avoid leaving a small,


173
TABLE 1. Continued

less visible component of the malformation in the gliotic plane


(48) or complicating the organized wall of a hematoma cavity
Zabramski et al.,
Robinson et al.,

with the peripheral cleavage plane. The determination of a


(ref. no.)
Series

Porter et al.,
1997 (64)

1991 (70)

1994 (93)

complete surgical removal is based on a combination of the


surgeon’s intraoperative impression and postoperative MRI to
confirm the absence of blood in a remaining segment of CM.
This approach is imperfect, however, as evidenced by recur-
b
a

NEUROSURGERY VOLUME 64 | NUMBER 5 | MAY 2009 | 807


GROSS ET AL.

TABLE 2. Review of brainstem cavernous malformation seriesa


No. of Resection
Series (ref. no.) Early morbidity and mortality Long-term outcome
patients rate
Alves de Sousa, 2007 (2) 13 100% 69% morbidity (23% trach/PEG, all trans) 69% improved/same,
7.7% worse, 7.7% dead, 15% lost
Amin-Hanjani et al., 1998 (4) 14 79% 57% morbidity (14% perm) 36% improved, 50% same, 14% worse;
64% no major def, RTW
Attar et al., 2001 (5) 4 25% morbidity (trans III), 25% dead 75% improved, 25% dead
Batay et al., 2007 (7) 4 75% 75% improved, 25% worse (rebled)
Bertalanffy et al., 1991 (9) 13 92% 46% morbidity (31% perm) 69% improved, 31% worse
Bertalanffy et al., 2002 (8) 24 100% 58% morbidity (8.3% perm) 63% improved, 29% same, 8.3% worse
Mean KPS 85.8 (76.8 preop)
Bouillot et al., 1996 (10) 17 65% 59% improved (29% ASx), 12%
same, 18% worse, 12% dead
Bricolo and Turazzi, 1995 (11) 18 94% 6% dead (early resp failure) 89% improved, 6% same, 6% dead
Bruneau et al., 2006 (12) 22 86% 61% morbidity (8.6% perm) 91% improved (41% ASx),
4.6% worse, 4.6% lost
Mean GOS 4.24 (3.22 preop)
Cantore et al., 1999 (13) 11 91% 67% morbidity, 27% VPS, 8% dead 83% independent, 17% dead
(1 postop rebleed, 1 shunt infection)
Chaskis and Brotchi, 1998 (16) 7 100% 29% sig improved (14% ASx),
71% slight improved/same
Chen et al., 2007 (17) 7 86% 28% morbidity (1 III, 1 VI; 14% perm) 100% improved (14% ASx),
all HS, HP Sx improved
DiRocco et al., 1997 (20) 3 100% 33% morbidity (0% perm) 100% improved (33% ASx)
Eisner et al., 1995 (22) 7 100% 29% morbidity 100% improved/same
Esposito et al., 2003 (23) 13 69% 54% morbidity 38% improved, 23% same, 38% worse
Mean KPS 66.9 (67.7 preop)
Fahlbusch et al., 1991 (24) 9 89% 89% improved, 11% worse
Ferroli et al., 2005 (25) 52 92% 44% morbidity (19% CN, 19% perm) 81% improved/same,
19% worse, 2% mortality
Mean KPS 82.5 (76.2 preop)b
Fritschi et al., 1994 (26)c 9c 89%c 56% ASx, 33% min def, 11% mod defc
Fukui et al., 1998 (27) 3 67% 67% improved, 33% same
Houtteville, 1995 (33) 5 80% 100% improved (60% ASx)
Isamat and Conesa, 1993 (34) 6 100% 50% morbidity (17% perm) 100% improved, RTW; 50% ASx
Kikuta et al., 2004 (39) 10 70% 30% improved, 50% same, 20% worse
Mean mRS 1.9 (2.7 preop)
Lapras et al., 1989 (45) 9 100% 11% trach, 11% dead 78% improved, 11% same, 11% dead
Lena et al., 2002 (46) 7 43% 14% morbidity (14% perm), 14% dead 71% improved (57% ASx),
14% worse (GOS 4), 14% dead (rebled)
Mao et al., 2001 (51) 14 100% 21% morbidity (7% perm) 93% improved/same, 7% worse
Mathiesen et al., 2003 (52)d 17 86%d 69% morbidityd 69% improved, 10% same,
17% worse (3/5 rebled), 3% deadd
Mizoi et al., 1992 (53) 6 100% 100% improved/same
Morcos et al., 1999 (54) 5 80% 60% improved (20% ASx),
20% worse (rebled), 20% lost
Nataf et al., 2007 (57) 19 89% 37% neuro morbidity (all CN) 63% improved, 26% same,
5% worse, 5% dead
Pechstein et al., 1997 (58) 7 100% 86% morbidity 29% improved, 57% same, 14% worse
Pendl et al., 1990, 5 80% improved, 20% same
1991 (59, 60) Mean KPS 86 (50 preop)

Continues

808 | VOLUME 64 | NUMBER 5 | MAY 2009 www.neurosurgery-online.com


BRAINSTEM CAVERNOUS MALFORMATIONS

TABLE 2. Continued
No. of Resection
Series (ref. no.) Early morbidity and mortality Long-term outcome
patients rate
Porter et al., 1999, 117 97% 35% morbidity (12% perm) 87% improved/same,
2000 (62, 63) 10% trach/PEG (2% perm) 9% worse, 3.5% dead
3.5% dead (CP arrest/hem ven infarct) Mean GOS 4.53 (4.4 preop)
Sakai et al., 1991 (71) 5 100% 100% improved; 60% excellent
Samii et al., 2001 (72) 36 100% 67% morbidity (47% CN) Mean KPS 78.2 (70.3 preop), 44% KPS 90
Sandalcioglu et al., 12 100% 83% morbidity (8% perm, VI) –100 42% improved, 50% same, 8.3% worse
2002 (73) Mean mRS 1.6 (2.2 preop)
Scott et al., 1990, 1992 (74, 75) 4 0% 50% morbidity (25% perm) 75% improved (50% ASx), 25% worse
Seifert et al., 2003 (76) 4 100% 75% morbidity (0% perm) 25% ASx, 50% min def, 25% dependent
Shehab et al., 2001 (77) 4 75% 100% morbidity
75% hearing loss, 25% CSF leak
75% mild worse hearing, 25% rebled (residual)
Sindou et al., 2000 (78) 12 92% 83% morbidity (17% perm) Mean KPS 79.2 (60 preop)
58% resp difficulty (8% perm)
67% improved, 17% same, 17% worse
Smith et al., 2003 (79) 7 100% 43% morbidity (0% perm) 100% improved/same (28% ASx)
Mean GOS 4.6
Sola et al., 2007 (80) 17 88% 24% morbidity (12% CN, 12% sensory) 76% improved, 24% same; 82% RTW
Mean KPS 90.6 (70 preop)
Steiger et al., 2006 (82) 3 33% 33% morbidity (hearing) 67% improved, 33% worse (hearing)
Steinberg et al., 2000 (83)e 42 93%e 29% worsee 52% improved, 43% same, 5% worsee
Symon et al., 1991 (85) 7 100% 100% improved
Viñas et al., 2002 (87) 8 63% morbidity (50% CN, 0% perm) 100% improved/same
Wang et al., 2003 (89) 137 96% 72% improved/same,
28% worse, 1% dead; 89% working
Weil and Tew, 1990 (90)f 6 67%f 83% morbidityf 83% improved, RTW; 17% worsef
Yaşargil et al., 1988 (91) 5 100% 100% improved
Yoshimoto and Suzuki, 1986 (92) 3 100% 67% morbidity 33% improved, 67% same
Zausinger et al., 2006 (94) 13 100% 24% morbidity (15% HP, perm)
8% trach (all transient) 85 improved/same
14% slightly worse (mild HP)
Zhao et al., 2007 (95) 9 100% 11% neuro morbidity (0% perm) 100% improved/same
Ziyal et al., 1999 (97) 9 78% 56% morbidity (18% perm), 67% improved,
11% trach/PEG 22% same, 11% worse
Mean GOS 4.6 (4 preop)

a
Trach/PEG, tracheostomy and/or percutaneous gastrostomy; trans, transient; perm, permanent; def, deficit; RTW, return to work; III, third nerve palsy; KPS, Karnofsky Performance
Scale score; preop, preoperative; ASx, asymptomatic; resp, respiratory; GOS, Glasgow Outcome Scale score; VPS, ventriculoperitoneal shunt; postop, postoperative; VI, sixth nerve
palsy; sig, significant; HS, hemisensory; HP, hemiparesis; Sx, symptoms; CN, any cranial nerve palsy; neuro, neurological; min, minimal/minor; mod, moderate; mRS, modified Rankin
Scale score; CP, cardiopulmonary; hem ven, hemorrhagic venous; CSF, cerebrospinal fluid.
b
Eight percent of patients had KPS ⬍70; all of these patients had partially resected lesions that rebled (25).
c
Includes results from Bern only (26); results from Phoenix incorporated in the series of Porter et al. (62, 63).
d
Data for entire surgical series of 29 patients with cavernous malformations of brainstem, basal ganglia, and thalamus (52).
e
Results for entire surgical series of 57 cavernous malformations of the brainstem, basal ganglia, and thalamus (83).
f
Includes results from Kashiwagi et al. (36). A subsequent report by Lewis et al. (48) from the same group documented recurrence (by rebleeding) of 2 lesions. This report incorporated
the results of Weil and Tew (90) and mixed them with results for subsequent brainstem cavernous malformation and arteriovenous malformation extirpation, citing a 93% resection
rate, 56% early neurological morbidity rate, and 85% long-term symptomatic improvement rate.

rence rates as high as 3.5% (2 of 56 patients) and 3.4% (1 of 29 attributable to postoperative edema, intraoperative manipula-
patients) in the deep CM series of Steinberg et al. (83) and tion, and changes in microcirculation (8, 72, 83). In the series of
Mathiesen et al. (52), respectively. Porter et al. (63), 10% of patients required tracheostomy and/or
Early postoperative morbidity is high, ranging from 29% to placement of a feeding tube; similar rates were seen in the
67% in larger surgical series (8, 12, 25, 57, 62, 63, 72, 83). series of Ziyal et al. (97) and Lapras et al. (45), although these
Fortunately, many complications are transient and are likely measures, fortunately, were most often temporary. Long-term

NEUROSURGERY VOLUME 64 | NUMBER 5 | MAY 2009 | 809


GROSS ET AL.

condition was reported as improved in 71%, the same in 19%, Visualization of the quadrigeminal plate is improved, and the
worse in 10%, and dead in 1.9% across 35 series (316 patients). risk of damage to the colliculi is decreased (10). Occipital drain-
Across 45 series (683 patients), 85% of patients were the same ing veins are coagulated, the medial aspect of the occipital lobe
or improved, 14% were worse, and 1.9% were dead. Overall, 16 is then retracted superolaterally, and the tentorium is opened to
of 820 patients (2.0%) died from partially resected CM rebleed- afford clear visualization of the cerebellomesencephalic fissure.
ing (4 cases) or surgically related causes: 5 from infectious com- Challenging dissection of the deep Galenic drainage system
plications, 2 from sudden cardiopulmonary arrest, 2 from ensues. As emphasized by Morcos et al. (54), this approach is
venous infarction, 1 from respiratory failure, and 2 from an limited to unilateral lesions, as the deep drainage system, the
unspecified cause. Importantly, these long-term outcome fig- falx, and the straight sinus obscure the view to lesions extend-
ures are heavily biased by larger surgical series and thus by the ing contralaterally. Despite lumbar drainage, postoperative
work of a few neurosurgeons who are uniquely experienced visual field deficits may be observed, owing to occipital lobe
with these generally rare lesions. Although the overall results retraction, and, more rarely, venous infarction may occur from
from Table 2 are therefore not applicable to the entire neurosur- sacrifice or stretching of bridging veins.
gical community at large, lessons learned from particular sur-
gical approaches (Table 3) to these lesions require emphasis. Anterior and Anterolateral Midbrain CMs
Complete resection of anterior and anterolateral mesencephalic
Posterior Midbrain CMs CMs appeared to meet with more difficulty when subgroups in
Long-term outcomes after resection of posterior midbrain surgical series were reviewed. As seen in all 7 patients (all same
CMs were generally quite good (Table 4): 26 of 30 patients were or improved) in the series of Porter et al. (63), good outcomes
the same or improved at follow-up. In the update of the series of may still be expected for the majority of these lesions, although
Zimmerman et al. (96) by Porter et al. (63), 6 additional patients the increased prevalence of incomplete resection among them (5
were reported as improved or the same. Postoperative oculomo- of 19 patients in Table 4) relatively hampered long-term out-
tor deficits may occur (10, 79, 88, 90), and, less frequently, come, compared with that of more posterior lesions. These CMs
obstructive hydrocephalus from scarring around the cerebral are typically approached via either a subtemporal (8, 24, 54) or
aqueduct has been reported (90). The most commonly used transsylvian approach (80, 97). The transsylvian approach pro-
approach was the supracerebellar infratentorial approach (Fig. 1) vides an ideal angle of attack for anterior or interpeduncular
(62, 83), which afforded excellent results in the majority of cases. CMs, whereas the subtemporal approach is well suited for
After unroofing of the transverse sinus and retraction of the anterolateral CMs. In addition, the subtemporal approach may
dural flap, bridging veins between the vermis and tentorium also be advantageous for more caudal pontomesencephalic CMs,
are divided close to the cerebellar surface. The dissection contin- as, unlike a retrosigmoid approach, it affords a direct viewing
ues with preservation of the precentral cerebellar veins until trajectory toward posterior aspects of the CM without requiring
exposure of the quadrigeminal plate. As illustrated by the 1 fatal- significant brainstem retraction. Lumbar drainage is used to min-
ity from rebleeding in the series of Cantore et al. (13), larger imize retraction injury to the temporal lobe, which, along with
lesions with a significant caudal extent into the pons and fourth potential injury to the third and/or fourth cranial nerves during
ventricle may not be amply exposed by this approach. Although tentorial manipulation (12), comprises the major disadvantage of
a paramedian supracerebellar infratentorial approach may be this approach. Fatal postoperative venous thrombosis was
useful for somewhat posterolateral lesions, more radical resec- reported in 1 case that we reviewed (13).
tion of bone superiorly (extending above the transverse sinus)
and laterally with skeletonization of the sigmoid and transverse Anterior and Anterolateral Pontine CMs
sinus may be used for more laterally situated lesions (88). The Modern, larger surgical series have noted improved long-term
extreme lateral supracerebellar infratentorial approach (modi- recovery after surgery among patients with anterolateral pontine
fied retrosigmoid), described by Vishteh et al. (88), averts the lesions (25, 63). All 15 patients with these lesions in the series of
problem of residual bone between the lateral aspect of the cran- Porter et al. (63) were the same or improved. Only 1 of 20 patients
iotomy and the sigmoid sinus impeding exposure of the lesion. with anterolateral pontine lesions in the series of Ferroli et al. (25)
The far posterior subtemporal approach is another option, had permanent postoperative deficits, compared with 29% of the
described by Smith et al. (79). The authors describe a remaining patients. The mean long-term Karnofsky Performance
horseshoe-shaped incision centered behind the ear. The cran- Scale score in the recent series of Sola et al. (80) was 90.6, and
iotomy is defined caudally by the temporal fossa floor, the ten- notably, the majority of lesions were in the anterolateral pons.
torium is divided, and the cerebellum may then be retracted if With the use of a retrosigmoid approach, 8 of 9 lesions in this
needed. Less cerebellar retraction and a potentially decreased series were completely resected, with all patients improving post-
risk of injury to cranial nerves are advantages of this approach; operatively (except 3 patients with preoperative Karnofsky
however, temporal lobe retraction-related edema and increased Performance Scale scores of 100) to a mean Karnofsky Perform-
risk of injury to the vein of Labbé are potential limitations (79). ance Scale score of 95.6. The retrosigmoid approach is a common
The occipital transtentorial approach (53, 54) is another alter- choice for lesions of the anterolateral pons (8, 46, 78, 80). Despite
native for patients with a steeply sloping tentorium, lateral limited ventral exposure, this approach was aptly chosen in 30
lesions, or lesions somewhat ventral to the tectal plate. cases reviewed, leading to complete resection in 93%. Only 3 of 30

810 | VOLUME 64 | NUMBER 5 | MAY 2009 www.neurosurgery-online.com


BRAINSTEM CAVERNOUS MALFORMATIONS

TABLE 3. Potential surgical approaches to cavernous malformations of the brainstema


Location Approach Comments
Posteromedial midbrain Supracerebellar infratentorial Most common
Occipital transtentorial Unilateral lesions; patients with steep slope to tentorium
Posterolateral midbrain Paramedian supracerebellar
infratentorial
Occipital transtentorial Unilateral lesions; patients with steep slope to tentorium
Far posterior subtemporal (79) Caution with vein of Labbé and temporal lobe lobe retraction
Extreme lateral supra- Modified retrosigmoid. More lateral exposure than paramedian supracerebellar infratentorial
cerebellar infratentorial (88)
Anterior midbrain OZ/pterional transsylvian Interpeduncular lesions
Anterolateral midbrain Subtemporal
Combined transsylvian-
subtemporal (8, 12)
Anterolateral pons Retrosigmoid Ideal, but least anterior exposure
Extended retrosigmoid (66) Limited posterior mastoidectomy, skeletonization and anterior mobilization of the sigmoid sinus with the
dural flap. This proadded ventral exposure; offset by greater risk of CSF leak and and venous sinus thrombosis
Subtemporal Rostral lesions
Combined retrosigmoid- Potential alternative to transpetrosal
subtemporal (8)
Retrolabyrinthine This approach, with or without combination with a far lateral or subtemporal approach, is often
(32, 76, 81) adequate for almost all anterior/anterolateral pontine CMs. Petrous bone is removed anterior to the
sigmoid sinus and posterior to semi-circular canals, skeletonizing them. The sigmoid sinus is
skeletonized, the tentorium is cut, and the sigmoid sinus is then mobilized. If nondominant, it may
be ligated and divided to provide even a wider angle between the cerebellum and petrous bone.
Postoperative hearing loss may occur and is almost ubiquitously transient, likely owing to accumu-
lation of fluid and soft tissue within the mastoid (76). Meticulous closure is crucial in all posterior
trans petrosal approaches to decrease the risk of postoperative CSF fistula
Transcrusal (partial Partial labyrinthectomy (superior and posterior semicircular canals removedfrom ampullae to
labyrinthine) common crus) with improved exposure via resectionof the petrous apex. Potentially hearing-
(32, 77) preserving alternative to the translabyrinthine approach
Translabyrinthine (81) Labyrinthectomy and skeletonization of the posterior half of the internal auditory meatus.
Hearing is sacrificed
Transotic (32) Complete drilling of the otic capsule and closure of the external
Hearing is sacrificed
Trancochlear (32, 81) Same as transotic except seventh nerve translocated posteriorly after cutting the chorda tympani and
GSPN; cochlear removal complete. Hearing is sacrificed. Greatest risk of facial nerve palsy. Widest
exposure/least tangential trajectory
Anterior transpetrosal Subtemporal approach followed by drilling of the medial aspect of the petrous the petrous apex. Extra-
(Kawase) dural drilling allows for better visualization and control of the horizontal petrous carotid artery and less
(37, 54, 82) risk of damage to the temporal lobe and vein of Labbé (54). Transdural drilling allows customization of the
extent of resection of the petrous apex; possibly decreased risk of CSF fistula (82). Hearing preserved. Best
for ventral lesions of the middle third of the pons. Disadvantages: manipulation of fifth nerve, relative
depth of posterior fossa in surgical field; hemostasis and complete extirpation more challenging
Transoral, transclival (69) Exposure up to midclivus Rare, often avoided for risk of CSF fistula/infection
Abutting floor of IV Transcerebellomedullary Ideal for lesions of medulla and caudal pons
fissure (intertonsillar) (56) Postoperative truncal ataxia (31, 90) is possible
Vermis splitting
Anterior/anterolateral Far lateral partial transcondylar Extent of condyle drilling proportional to degree of ventral exposure
medulla Retrosigmoid Rostral, lateral lesions
Transoral, transclival (69) Rare, often avoided for risk of CSF fistula/infection

a
OZ, orbitozygomatic; CSF, cerebrospinal fluid; CM, cavernous malformation; GSPN, greater superficial petrosal nerve; IV, fourth ventricle.

patients worsened postoperatively, whereas 14 of 30 patients were the ear, followed by craniotomy, limited posterior mastoidectomy,
neurologically intact at follow-up. The extended retrosigmoid skeletonization of the sigmoid sinus, and anterior mobilization of
approach is an alternative with added ventral exposure that was the sinus with a dural flap. The added exposure from this
recently described by Quiñones-Hinojosa et al. (66). The authors approach is balanced by the potential for venous sinus thrombo-
described a C-shaped skin incision from the mastoid tip around sis and increased risk of cerebrospinal fluid (CSF) fistula.

NEUROSURGERY VOLUME 64 | NUMBER 5 | MAY 2009 | 811


GROSS ET AL.

TABLE 4. Results for brainstem cavernous malformation resection, subdivided by location and approacha,b
No. of Resection Early
Series (ref. no.) Approach Long-term outcome
patients rate morbidity
Posterior midbrain
Bertalanffy et al., 1991 (9) 3 SI 3/3 100% intact (all obstructive HCP preop)
Bertalanffy et al., 2002 (8) 2 OT 2/2 1 same, 1 worse (cereb)
Bouillot et al., 1996 (10) 2 SI 0/2 1 IV 1 improved, 1 worse
Pendl et al., 1990, 1991 (59, 60) 2 SI Both improved (KPS 90, 100 from 20, 40 preop)
Sandalcioglu et al., 2002 (73) 2 SI 2/2 Both improved (mRS 0, 1 from 1, 2 preop)
Sindou et al., 2000 (78) 2 OT 2/2 1 improved, 1 worse (needing trach)
Smith et al., 2003 (79) 7 Far posterior ST 7/7 2 INO 100% same or improved (mean GOS
1 confusion, aphasia 4.6, 28% intact)
Vishteh et al., 1999 (88) 5 ELSI 4/5 1 IV All improved (all GOS 1)
Weil and Tew, 1990 (90) 3 Infratent 3/3 2 new CN, 1 HCP 2 improved, 1 worse
Vermis splitting 1 worse ataxia
Zimmerman et al., 1991 (96)c 2 SI 2/2 1 improved, 1 same
Anterior midbrain
Bertalanffy et al., 2002 (8) 3 ST 3/3 1 KPS 100, 2 KPS 90
Bouillot et al., 1996 (10) 4 ST 3/4 2 intact, 1 improved, 1 dead (VT)
Fahlbusch et al., 1991 (24) 2 ST Both improved
Houtteville, 1995 (33) 2 ST 1/2 Both improved, 1 intact
Kikuta et al., 2004 (39) 2 ST 1/2 mRS 3, 5 (mRS 2, 5 preop)
Morcos et al., 1999 (54) 2 ST 1/2 1 improved, 1 worse (rebled)
Sola et al., 2007 (80) 2 Pterional TS 1/2 1 transient diplopia Both KPS 100 (30 and 90 preop)
Viñas et al., 2002 (87) 2 ST 1 transient CN Both same/improved
Zimmerman et al., 1991 (96)c 2 ST 2/2 1 transient HP Both improved
Ziyal et al., 1999 (97) 2 OZ TS 2/2 No early neuro Both improved, mean
morbidity GOS 4.5 (4 preop)
Anterior/anterolateral pons
Bertalanffy et al., 2002 (8) 3 Retrosigmoid 3/3 1 CSF leak 2 improved, 1 same
Mean KPS 93.3 (70 preop)
Bertalanffy et al., 2002 (8) 2 ST ⫹ retrosigmoid 2/2 1 CSF leak Both improved to KPS 100
Lena et al., 2002 (46) 2 Retrosigmoid 1/2 Both intact
Sandalcioglu et al., 2002 (73) 4 Retromastoid 4/4 3 improved/same, 1 worse
Mean mRS 1.5 (2 preop)
Seifert et al., 2003 (76) 4 Retrolabyrinthine 4/4 1 CSF leak, 1 1 intact, 2 minor deficits,
hearing, 1 TP 1 dependent
Shehab et al., 2001 (77) 4 Partial translab 3/4 1 CSF leak, 3 3 mildly worse hearing,
hearing, 1 CN 1 residual rebled
Sindou et al., 2000 (78) 2 Retrosigmoid 2/2 2 trans resp disturbances 2 improved (KPS 80, 90)
Sola et al., 2007 (80) 9 Retrosigmoid 8/9 6 improved, 3 same
Mean KPS 95.6 (84.4 preop)
Steiger et al., 2006 (82) 3 Transdural ant petrosal 1/3 1 worse hearing 2 improved, 1 worse
Viñas et al., 2002 (87) 4 Transpetrosal 1 CN All improved/same
Zausinger et al., 2006 (94) 10 Retromastoid 10/10 2 new HP 8 improved (4 intact), 2 new HP
Floor of fourth ventricle
Bertalanffy et al., 1991 (9) 7 Median suboccipital 7/7 4/7 morbidity 6 improved (3 intact), 1
worse (air embo, bedridden)
Bertalanffy et al., 2002 (8) 6 IT 6/6 5 improved (1 intact), 1 same
Mean KPS 78.3 (65 preop)

Continues

812 | VOLUME 64 | NUMBER 5 | MAY 2009 www.neurosurgery-online.com


BRAINSTEM CAVERNOUS MALFORMATIONS

TABLE 4. Continued
No. of Resection Early
Series (ref. no.) Approach Long-term outcome
patients rate morbidity
Bertalanffy et al., 2002 (8) 2 Occip interhem 2/2 1 improved, 1 worse (worse ataxia, diplopia)
transtent transvelum Mean KPS 70 (70 preop)
Bouillot et al., 1996 (10) 9 Median suboccipital 7/9 1 rebled, 1 dead 7 improved (3 intact), 1 same, 1 dead
Bruneau et al., 2006 (12) 8 IT 8/8 2/8 neuro morbidity All improved/same
Di Rocco et al., 1997 (20) 3 Median suboccipital 3/3 1 transient VII All improved (1 intact)
Eisner et al., 1995 (22) 7 Inf vermis splitting 7/7 2 CN, 1 HP, 2 ocular All improved/same
Fahlbusch et al., 1991 (24) 6 Vermis splitting motility 5 improved, 1 worse
Isamat and Conesa, 1993 (34) 6 IT/inf vermis splitting 6/6 3/6 morbidity (3 CN) All improved, RTW; 3 intact
Kikuta et al., 2004 (39) 7 Median suboccipital 5/7 4 improved (2 intact); 2 same, 1 worse
Mean mRS 1.4 (2.7 preop)
Mizoi et al., 1992 (53) 4 Vermis splitting 4/4 All improved/same with “good” outcomes
Sakai et al., 1991 (71) 3 Vermis splitting 3/3 All improved, 1 “excellent”
Sandalcioglu et al., 2002 (73) 3 Subcerebellar 3/3 All same mRS score
Mean mRS 1.3
Sindou et al., 2000 (78) 6 IT 5/6 4/6 morbidity 3 improved, 2 same, 1 worse
2 resp disturbances, Mean KPS 80 (70 preop,
1 CSF fistula, 1 HCP 2 comatose)
Symon et al., 1991 (85) 6 Inf vermis splitting 6/6 1 trans loss of gag All improved
Weil and Tew, 1990 (90) 3 Inf vermis splitting 3/3 2/3 morbidity All improved, RTW
1 INO, ataxia; 1 VI, VII
Yoshimoto and Suzuki, 1986 (92) 2 Vermis splitting 2/2 1 trans oculomotor deficit Both improved
Zausinger et al., 2006 (94) 4 Median suboccipital 4/4 1 trach All improved
Ziyal et al., 1999 (97) 2 IT 1/2 Both worse (CN) Both improved, both GOS 5
1 trach, PEG; 1 VPS

a
SI, supracerebellar infratentorial; HCP, hydrocephalus; preop, preoperative; OT, occipital transtentorial; cereb, any cerebellar deficit; IV, fourth nerve palsy; KPS,
Karnofsky Performance Scale score; mRS, modified Rankin Scale score; trach, tracheostomy; ST, subtemporal; INO, internuclear ophthalmoplegia; GOS, Glasgow
Outcome Scale score; ELSI, extreme lateral supracerebellar infratentorial; infratent, infratentorial; CN, any cranial nerve deficit; VT, venous thrombosis; TS, transsylvian;
OZ, orbitozygomatic; neuro, neurological; CSF, cerebrospinal fluid; TP, tetraparesis; translab, translabyrinthine; trans, transient; resp, respiratory; ant, anterior; HP,
hemiparesis; embo, embolism; IT, intertonsillar/transcerebellomedullary fissure; occip interhem transtent transvelum, occipital interhemispheric transtentorial transvelum
medullare; VII, seventh nerve palsy; inf, inferior; RTW, return to work; VI, sixth nerve palsy; PEG, percutaneous gastrostomy; VPS, ventriculoperitoneal shunt.
b
From a review of series providing sufficient individual patient data for at least 2 patients undergoing the same approach.
c
Earlier detailed report of the first set of patients in the larger series of Porter et al. (62, 63).

Although combined far lateral-subtemporal and retrosigmoid- observed in 3 of 4 patients, although it was mild. Translabyrin-
subtemporal approaches are always options, the wider, more thine and transcochlear exposures, considerations for patients
ventral exposure afforded by the transpetrosal approaches with preoperative complete hearing loss, are exceedingly rarely
makes them an attractive choice for surgeons who are well indicated. The anterior transpetrosal (Kawase) approach has
versed in petrous bone drilling for added ventral exposure (Table been described for 4 CMs in our reviewed series (8, 82). Its added
3). The posterior transpetrosal approaches fall along a spectrum caudal exposure, as compared with the standard subtemporal
of increasing petrous bone resection, affording progressively approach, makes it an option for ventral lesions of the middle
wider, more anterior exposure. As more bone is resected, the third of the pons (54). In the series of Steiger et al. (82), which
risk of complications such as CSF leaks and seventh and eighth emphasized the ability to tailor petrous apex resection via trans-
nerve palsies progressively increases as well. We feel that the dural drilling, 1 of 3 lesions was completely resected. Although
retrolabyrinthine approach, combined with a far lateral or sub- the medial aspect of the petrous bone can theoretically be
temporal exposure if needed, is sufficient for the vast majority of resected “with impunity,” 1 of 3 patients with CMs (1 of 9 over-
anterior/anterolaterally situated brainstem CMs. Among larger all) had postoperative hearing loss.
series, Porter et al. (63) used a retrolabyrinthine approach in 5 of
86 cases; Ferroli et al. (25) used it for 3 of 52 cases. The partial CMs Abutting the Floor of the Fourth Ventricle
labyrinthine approach is a hearing-preserving alternative with Lesions abutting or protruding into the floor of the fourth
added exposure via drilling of the petrous apex (32, 77). When ventricle were traditionally resected after a median suboccipi-
applied to CM resection by Shehab et al. (77), hearing loss was tal, vermis-splitting approach (24, 53, 71, 90, 92). The potential

NEUROSURGERY VOLUME 64 | NUMBER 5 | MAY 2009 | 813


GROSS ET AL.

transcerebellomedullary fissure approach (8, 12, 78, 97); only 1 of


A B 22 patients was reported to be in worse condition at long-term
follow-up (78). Although long-term results in our small review
are generally good, early postoperative deficits are often signif-
icant. Transient internuclear ophthalmoplegia and sixth and/or
seventh nerve deficits are common (20, 34, 83, 90, 97). The need
for an oftentimes transient tracheostomy and/or feeding tube
can be expected for 10% to 20% of patients (62, 78, 85, 97), and
the possibility for Ondine’s curse in otherwise improving
patients should not be forgotten (62, 83). Interestingly, the same
large, recent series that suggested improved surgical outcomes
for anterolateral pontine lesions also found worse long-term out-
comes among patients with lesions excised through the floor of
C D the fourth ventricle (25, 63). In the series of Porter et al. (63), 2 of
14 patients died from cardiopulmonary arrest postoperatively,
and 2 of 14 patients (both with lesions under the facial colliculus)
were worse at follow-up. In the series of Ferroli et al. (25), 36%
of patients with CMs excised through the floor of the fourth ven-
tricle had permanent neurological deficits at follow-up. As
emphasized by Ferroli et al. (25) and Porter et al. (63), respect for
the floor of the fourth ventricle is the crucial lesson learned;
excised CMs should be visible on the surface of the floor.
E
Anterior and Anterolateral Medullary Cavernomas
Anterolateral lesions of the medulla can be removed via the
FIGURE 1. Magnetic resonance far lateral partial transcondylar approach (25, 54, 63). After a
imaging scans (A and C, axial low lateral suboccipital craniotomy and removal of part of the
views; B, sagittal view) and ipsilateral posterior arch of C1, the occipital condyle is par-
intraoperative photographs (D tially drilled to provide ample exposure of the ventral brain-
and E) of a 24-year-old woman stem (primarily ventral lesions require more drilling). Potential
who initially presented with
complications include vertebral artery injury and CSF leaks.
diplopia after her second cav-
ernous malformation (CM) hemorrhage (A). We initially observed her as the Transient postoperative lower cranial nerve deficits are not
lesion did not abut the pial surface of the brainstem. After 3 additional suc- uncommon (87, 97). Rostral pontomedullary lesions may be
cessive symptomatic hemorrhages over the course of 6 months, the enlarging, removed via the retrosigmoid approach (12, 78). In the series of
exophytic lesion (B and C) was excised via a supracerebellar infratentorial Porter et al. (63), 10 of 13 patients with anterolateral medullary
approach (D). Care was taken not to disrupt the associated venous angioma CMs were the same or improved, 2 were worsened from dam-
(arrow, D). After meticulous dissection of the Galenic system (E), the lesion age to the pyramidal tracts, and 1 died from venous outflow
was excised. Despite requiring permanent ventriculoperitoneal shunting obstruction. Two patients each in the series of Bertalanffy et al.
for her preoperative hydrocephalus, the patient improved postoperatively (8) and Ziyal et al. (97) improved at long-term follow-up after
without new neurological deficits. After 2 years of follow-up, she remains
excision of their medullary CMs via a far lateral approach,
improved without radiographic recurrence of her CM.
although transient worsening of deglutition was noted in 2
cases (97). Although the transoral, transclival route is a poten-
for postoperative truncal ataxia (31, 90) after vermis splitting tial alternative for a direct route to particularly ventrally situ-
makes the transcerebellomedullary fissure (intertonsillar) ated lesions (69), the narrow operative field and risk of postop-
approach (8, 10, 56) an attractive alternative. The tonsils are erative CSF fistula and/or infection has generally swayed
elevated and spread to expose the tela choroidea, which is cut surgeons toward transcondylar approaches. Reisch et al. (69)
bilaterally at the taenia. If needed, the inferior medullary velum used this approach for 2 patients with ventral medullary CMs.
can be opened for added superolateral exposure (complete Although both patients were ultimately able to return to work,
telovelar approach) (56). This ultimately affords access to cau- they each experienced transient postoperative worsening of
dal pontine and medullary CMs; however, in our experience, their hemiparesis, and 1 patient developed a CSF fistula.
more rostral lesions need added exposure by splitting a small
portion of the inferior vermis. RADIOSURGERY
In our review of series providing specific details on lesions
resected through the floor of the fourth ventricle (Table 4), we see According to retrospective series of brainstem AOVM radio-
that 82 of 86 lesions (95%) were documented as completely surgery, 9% to 44% of patients will rebleed after radiosurgery
resected. Twenty of 22 lesions were completely resected via the (4%–15.2% hemorrhage rate per patient-year), with total long-

814 | VOLUME 64 | NUMBER 5 | MAY 2009 www.neurosurgery-online.com


BRAINSTEM CAVERNOUS MALFORMATIONS

term permanent morbidity rates from radiation injury and shown to be significantly greater than radiosurgery of arteri-
rebleeding ranging from 7% to 27% and mortality rates ranging ovenous malformations (35, 61). As hemorrhage clustering
from 0% to 13% (3, 21, 28, 40, 49). The general radiosurgical among aggressive lesions may abate after a “natural” 2-year
series of Hasegawa et al. (30) and Pollock et al. (61) have latency period (6), the efficacy of radiosurgery remains
reported permanent radiation-induced morbidity rates of 12% unproven, making it an alternative to observation, not to sur-
(6 of 52 patients) and 33% (4 of 12 patients), respectively, for gery, of aggressive lesions.
patients with brainstem AOVMs. Several general radiosurgical
series have demonstrated dramatic decreases of annual CM
Surgical Approach and Excision
rebleed rates after a 2-year “latency period” to as low as 0.8% The vast majority of authors advocate subacute CM removal
(15, 30, 50, 61). This phenomenon was not seen in all radiosur- (12, 24, 34, 48, 57, 80, 83, 89), allowing for posthemorrhage
gical series, however (35, 38), and may merely reflect the natu- edema to diminish and hematoma liquification. At times, the
ral history of the lesions (6). hematoma can facilitate excision by creating a clean dissection
plane (12, 24, 48, 83, 89). In the chronic stage after hematoma
DISCUSSION organization, gliosis, and malformation adherence to the gliotic
capsule, dissection planes are often obscured (12, 48, 83, 89).
Patient Selection and Timing Several series have shown late surgery to be significantly asso-
ciated with worse outcomes (12, 24, 73); however, early relief of
While increasing familiarity with the surgical treatment of mass effect, at least in part, contributes to improved results of
brainstem CMs may broaden the neurosurgeon’s criteria for early surgery. Frameless stereotaxy, brainstem auditory evoked
surgical management, lessons learned from natural history potentials, somatosensory evoked potentials, and intraopera-
studies and surgical results have led us to the following inclu- tive mapping of cranial nerve nuclei (22, 84) are important intra-
sion criteria: 1) at least 2 prior clinically significant hemorrhagic operative adjuncts (8, 48, 62, 73, 80, 83, 97). Ultrasound may be
events, and 2) pial or ependymal representation on T1-weighted useful as well, assisting with intraoperative brainshift affecting
MRI scans (Fig. 1). With the exception of life-threatening emer- neuronavigation (95) and ensuring complete excision (48).
gent cases and patients with significant, progressive neurologi- Several recent series advocate a respect for the floor of the
cal deterioration, we generally await a second clinical event to fourth ventricle (12, 25, 63). We concur; if a CM is not visible on
essentially define the CM as aggressive. It is clear from natural the floor, we will conclude the operation. This conservative
history studies that some lesions have the potential for a benign approach is based on recent results showing worse outcomes
course (23, 47, 49). Asymptomatic patients who may very well for lesions of the floor of the fourth ventricle (23, 25, 63) and the
never experience hemorrhage from their lesion should not face fact that a small rim of functional parenchymal tissue may be
the 29% to 67% risk of early morbidity and 15% risk of long- compressed between the lesion and ependymal surface.
term worsening (in the best of hands) after surgery (rates based Neuroanatomic landmarks and intraoperative mapping may
on series with primarily symptomatic patients). CMs in patients become unreliable (63). The literature describing “safe entry
who have experienced 2 hemorrhages are far more likely to fol- zones” through the floor of the fourth ventricle is conflicted,
low the aggressive paths described by Pozzati et al. (65) and with some authors advocating a paramedian incision to avoid
Tung et al. (86). The neurological risk of surgery at this point is damage to the medial longitudinal fasciculus (13, 24) and oth-
less than the expected morbidity of a third hemorrhage, which, ers advocating a median incision to avoid the risk of disrupting
as these 2 aforementioned studies have shown, is likely to be the dorsolateral pontomedullary vascular supply (8, 10, 11).
more severe and persistent. On the other hand, with intraoperative neuronavigation, we
The eloquence of the brainstem necessitates being afforded advocate the use of small (⬍1 cm) longitudinal incisions on
with a direct surgical view of the lesion, namely via pial or the anterolateral aspect of the pons in the peritrigeminal area if
ependymal representation. Preoperatively, this is best assessed needed (68). These incisions are generally well tolerated, as the
by T1-weighted MRI scans to avoid being misled by the poten- area contains the transverse fibers of the middle cerebellar
tial “blooming effect” from the ferromagnetic properties of peduncle (25, 63, 78). Recalde et al. (68) recently presented
hemosiderin (83) on T2-weighted imaging. Nonetheless, even detailed anatomic descriptions of the “relatively safe” entry
T1-weighted images may inaccurately depict the lesion’s rela- zones to the mesencephalon (lateral mesencephalic sulcus) and
tion to the brainstem surface (58, 94). Although some series medulla (olivary body) as well. We never allow a simple
have demonstrated impressive results after the extirpation of “bulge” of the brainstem to dictate an incision point for CM
deep-seated CMs (8, 11, 12), our experience is more consistent removal; it may contain vital functional tissue (84), and the CM
with that of Ferroli et al. (25) in their large series (9% worse itself may be unexpectedly deep.
long-term condition if superficial versus 29% worse if deep), Once the lesion and hematoma are reached, the hematoma is
fashioning our surgical inclusion criteria. drained and a peripheral cleavage plane is defined. Small feed-
Patients with deep- seated, surgically inaccessible, and ers to the CM are coagulated at low power at the surface of the
aggressive CMs may be considered for radiosurgery at modest CM with ongoing irrigation. Larger lesions are removed piece-
doses (12–14 Gy marginal dose, to decrease complications [50]). meal to minimize retraction. After removal, the cavity is thor-
Nonetheless, the morbidity of AOVM radiosurgery has been oughly inspected for hemostasis and residual lesion. Complete

NEUROSURGERY VOLUME 64 | NUMBER 5 | MAY 2009 | 815


GROSS ET AL.

extirpation is the rule, leaving the surrounding hemosiderin- 6. Barker FG 2nd, Amin-Hanjani S, Butler WE, Lyons S, Ojemann RG, Chapman
stained parenchyma and developmental venous anomaly intact PH, Ogilvy CS: Temporal clustering of hemorrhages from untreated cavernous
malformations of the central nervous system. Neurosurgery 49:15–24, 2001.
(25, 63, 73, 83, 89). 7. Batay F, Bademci G, Deda H: Critically located cavernous malformations.
Minim Invas Neurosurg 50:71–76, 2007.
Recent Advances 8. Bertalanffy H, Benes L, Miyazawa T, Alberti O, Siegel AM, Sure U: Cerebral
cavernomas in the adult. Review of the literature and analysis of 72 surgically
The burden of inaccurate preoperative MRI may soon abate
treated patients. Neurosurg Rev 25:1–55, 2002.
with preoperative 3-dimensional constructive interference in 9. Bertalanffy H, Gilsbach JM, Eggert HR, Seeger W: Microsurgery of deep-
steady state MRI, as described by Zausinger et al. (94). These seated cavernous angiomas: Report of 26 cases. Acta Neurochir (Wien)
authors reported improved judgment of the thickness of the 108:91–99, 1991.
10. Bouillot P, Dufour H, Roche PH, Lena G, Graziani N, Grisoli F: Angio-
parenchymal layer over the lesion, a sharper peripheral rim of
graphically occult vascular malformations of the brain stem: Apropos of 25
the lesion, and clearer visualization of the spatial relations of cases [in French]. Neurochirurgie 42:189–201, 1996.
the lesion to fairly safe entry zones. In addition to neuronavi- 11. Bricolo A, Turazzi S: Surgery for gliomas and other mass lesions of the brain-
gation, intraoperative endoscopy to ensure complete lesion stem. Adv Tech Stand Neurosurg 22:261–341, 1995.
12. Bruneau M, Bijlenga P, Reverdin A, Rilliet B, Regli L, Villemure JG, Porchet F,
excision was reported by Sandalcioglu et al. (73). Quiñones-
de Tribolet N: Early surgery for brainstem cavernomas. Acta Neurochir
Hinojosa et al. (67) used a modified Kartush stimulator to intra- (Wien) 148:405–414, 2006.
operatively map descending corticospinal tracts in the cerebral 13. Cantore G, Missori P, Santoro A: Cavernous angiomas of the brain stem. Intra-
peduncle before excision of a peduncular CM with hemo- axial anatomical pitfalls and surgical strategies. Surg Neurol 52:84–94, 1999.
14. Cantu C, Murillo-Bonilla L, Arauz A, Higuera J, Padilla J, Barinagarrementeria
siderin staining at the brainstem surface. Chen et al. (17) used F: Predictive factors for intracerebral hemorrhage in patients with cavernous
diffusion tensor imaging with white matter tractography trans- angiomas. Neurol Res 27:314–318, 2005.
ferred to their intraoperative neuronavigation system to clearly 15. Chang SD, Levy RP, Adler JR Jr, Martin DP, Krakovitz PR, Steinberg GK:
visualize the relation of the lesion to the lemnisci and descend- Stereotactic radiosurgery of angiographically occult vascular malformations:
14-year experience. Neurosurgery 43:213–221, 1998.
ing corticospinal tracts. Results from this series were impres- 16. Chaskis C, Brotchi J: The surgical management of cerebral cavernous
sive (Table 2); all pyramidal and sensory deficits were the same angiomas. Neurol Res 20:597–606, 1998.
or improved postoperatively. 17. Chen X, Weigel D, Ganslandt O, Buchfelder M, Nimsky C: Diffusion tensor
imaging and white matter tractography in patients with brainstem lesions.
Acta Neurochir (Wien) 149:1117–1131, 2007.
CONCLUSIONS 18. Dandy WE: Venous abnormalities and angiomas of the brain. Arch Surg
17:715–793, 1928.
Brainstem location, prior hemorrhage, and female sex all 19. Del Curling O Jr, Kelly DL Jr, Elster AD, Craven TE: An analysis of the natu-
increase the risk for subsequent clinically significant bleeding ral history of cavernous angiomas. J Neurosurg 75:702–708, 1991.
20. Di Rocco C, Iannelli A, Tamburrini G: Cavernous angiomas of the brain stem
from CMs, whereas lesion size does not. Patients with at least in children. Pediatr Neurosurg 27:92–99, 1997.
2 prior symptomatic bleeds or progressive neurological deficits 21. Duma CM, Lunsford D, Kondziolka D, Bissonette DJ, Somaza S, Flickinger
with pial or ependymal representation of their CM are good JC: Radiosurgery for vascular malformations of the brain stem. Acta
candidates for microsurgical extirpation. Patients with asymp- Neurochir Suppl (Wien) 58:92–97, 1993.
22. Eisner W, Schmid UD, Reulen HJ, Oeckler R, Olteanu-Nerbe V, Gall C,
tomatic lesions should be observed. Complete microsurgical Kothbauer K: The mapping and continuous monitoring of the intrinsic motor
excision is the rule. Retrosigmoidal or other approaches to the nuclei during brain stem surgery. Neurosurgery 37:255–265, 1995.
anterolateral pons are generally less morbid than excision via 23. Esposito P, Coulbois S, Kehrli P, Boyer P, Dietemann JL, Rousseaux P, Auque
the floor of the fourth ventricle; extirpation of intrinsic lesions J, Maitrot D: Place of surgery in the management of brainstem cavernomas:
Results of a multicentric study [in French]. Neurochirurgie 49:5–12, 2003.
through the floor of the fourth ventricle should be avoided. 24. Fahlbusch R, Strauss C, Huk W: Pontine- mesencephalic cavernomas:
Indications for surgery and operative results. Acta Neurochir Suppl (Wien)
Disclosure 53:37–41, 1991.
The authors have no personal financial or institutional interest in any of the 25. Ferroli P, Sinisi M, Franzini A, Giombini S, Solero CL, Broggi G: Brainstem
drugs, materials, or devices described in this article. cavernomas: Long-term results of microsurgical resection in 52 patients.
Neurosurgery 56:1203–1214, 2005.
26. Fritschi JA, Reulen HJ, Spetzler RF, Zabramski JM: Cavernous malformations of
REFERENCES the brain stem: A review of 139 cases. Acta Neurochir (Wien) 130:35–46, 1994.
27. Fukui M, Matsushima T, Ikezaki K, Natori Y, Inamura T, Ohara S, Kawamura
1. Aiba T, Tanaka R, Koike T, Kameyama S, Takeda N, Komata T: Natural his- T: Surgery of angiomas in the brainstem with a stress on the presence of
tory of intracranial cavernous malformations. J Neurosurg 83:56–59, 1995. telangiectasia. Neurol Med Chir (Tokyo) 38 [Suppl]:250–254, 1998.
2. Alves de Sousa A: Deep-seated (corpus callosum, intraventricular, basal gan- 28. García-Muñoz L, Velasco-Campos F, Lujan-Castilla P, Enriquez-Barrera M,
glia and insula) and brain stem cavernous angiomas. Experience in Brazil [in Cervantes-Martínez A, Carrillo-Ruiz J: Radiosurgery in the treatment of brain
French]. Neurochirurgie 53:182–191, 2007. cavernomas. Experience with 17 lesions treated in 15 patients [in French].
3. Amin-Hanjani S, Ogilvy CS, Candia GJ, Lyons S, Chapman PH: Stereotactic Neurochirurgie 53:243–250, 2007.
radiosurgery for cavernous malformations: Kjellberg’s experience with pro- 29. Ghannane H, Khalil T, Sakka L, Chazal J: Analysis of a series of cavernomas
ton beam therapy in 98 cases at the Harvard cyclotron. Neurosurgery of the central nervous system: 39 non operated cases, 39 operated cases, 1
42:1229–1238, 1998. dead [in French]. Neurochirurgie 53:217–222, 2007.
4. Amin-Hanjani S, Ogilvy CS, Ojemann RG, Crowell RM: Risks of surgical man- 30. Hasegawa T, McInerney J, Kondziolka D, Lee JY, Flickinger JC, Lunsford LD:
agement for cavernous malformations of the nervous system. Neurosurgery Long-term results after stereotactic radiosurgery for patients with cavernous
42:1220–1228, 1998. malformations. Neurosurgery 50:1190–1198, 2002.
5. Attar A, Ugur HC, Savas A, Yüceer N, Egemen N: Surgical treatment of 31. Heffez DS, Zinreich SJ, Long DM: Surgical resection of intrinsic brain stem
intracranial cavernous angiomas. J Clin Neurosci 8:235–239, 2001. lesions: An overview. Neurosurgery 27:789–798, 1990.

816 | VOLUME 64 | NUMBER 5 | MAY 2009 www.neurosurgery-online.com


BRAINSTEM CAVERNOUS MALFORMATIONS

32. Horgan MA, Delashaw JB, Schwartz MS, Kellogg JX, Spektor S, McMenomey 59. Pendl G, Vorkapic P: Microsurgery of intrinsic midbrain lesions. Acta
SO: Transcrusal approach to the petroclival region with hearing preservation. Neurochir Suppl (Wien) 53:137–143, 1991.
Technical note and illustrative cases. J Neurosurg 94:660–666, 2001. 60. Pendl G, Vorkapic P, Koniyama M: Microsurgery of midbrain lesions.
33. Houtteville JP: The surgery of cavernomas both supra-tentorial and infra- Neurosurgery 26:641–648, 1990.
tentorial. Adv Tech Stand Neurosurg 22:185–259, 1995. 61. Pollock BE, Garces YI, Stafford SL, Foote RL, Schomberg PJ, Link MJ:
34. Isamat F, Conesa G: Cavernous angiomas of the brain stem. Neurosurg Clin Stereotactic radiosurgery for cavernous malformations. J Neurosurg
N Am 4:507–518, 1993. 93:987–991, 2000.
35. Karlsson B, Kihlström L, Lindquist C, Ericson K, Steiner L: Radiosurgery for 62. Porter RW, Detwiler PW, Spetzler RF: Surgical technique for resection of cav-
cavernous malformations. J Neurosurg 88:293–297, 1998. ernous malformations of the brain stem. Op Tech Neurosurg 3:124–130, 2000.
36. Kashiwagi S, van Loveren HR, Tew JM Jr, Wiot JG, Weil SM, Lukin RA: 63. Porter RW, Detwiler PW, Spetzler RF, Lawton MT, Baskin JJ, Derksen PT,
Diagnosis and treatment of vascular brain-stem malformations. J Neurosurg Zabramski JM: Cavernous malformations of the brainstem: Experience with
72:27–34, 1990. 100 patients. J Neurosurg 90:50–58, 1999.
37. Kawase T, Shiobara R, Toya S: Anterior transpetrosal-transtentorial approach 64. Porter PJ, Willinsky RA, Harper W, Wallace MC: Cerebral cavernous malfor-
for sphenopetroclival meningiomas: Surgical methods and results in 10 mations: Natural history and prognosis after clinical deterioration with or
patients. Neurosurgery 28:869–876, 1991. without hemorrhage. J Neurosurg 87:190–197, 1997.
38. Kida Y, Kobayashi T, Mori Y: Radiosurgery of angiographically occult vascu- 65. Pozzati E, Acciarri N, Tognetti F, Marliani F, Giangaspero F: Growth, subse-
lar malformations. Neurosurg Clin N Am 10:291–303, 1999. quent bleeding, and de novo appearance of cerebral cavernous angiomas.
39. Kikuta K, Nozaki K, Takahashi JA, Miyamoto S, Kikuchi H, Hashimoto N: Neurosurgery 38:662–670, 1996.
Postoperative evaluation of microsurgical resection for cavernous malforma- 66. Quiñones-Hinojosa A, Chang EF, Lawton MT: The extended retrosigmoid
tions of the brainstem. J Neurosurg 101:607–612, 2004. approach: An alternative to radical cranial base approaches for posterior
40. Kim DG, Choe WJ, Paek SH, Chung HT, Kim IH, Han DH: Radiosurgery of fossa lesions. Neurosurgery 58 [Suppl 2]:ONS-208–ONS-214, 2006.
intracranial cavernous malformations. Acta Neurochir (Wien) 144:869–878, 67. Quiñones-Hinojosa A, Lyon R, Du R, Lawton MT: Intraoperative motor map-
2002. ping of the cerebral peduncle during resection of a midbrain cavernous mal-
41. Kim DS, Park YG, Choi JU, Chung SS, Lee KC: An analysis of the natural his- formation: Technical case report. Neurosurgery 56 [Suppl 2]:ONS-439, 2005.
tory of cavernous malformations. Surg Neurol 48:9–18, 1997. 68. Recalde RJ, Figueiredo EG, de Oliveira E: Microsurgical anatomy of the safe
42. Kondziolka D, Lunsford LD, Kestle JR: The natural history of cerebral cav- entry zones on the anterolateral brainstem related to surgical approaches to
ernous malformations. J Neurosurg 83:820–824, 1995.
cavernous malformations. Neurosurgery 62 [Suppl 1]:9–17, 2008.
43. Labauge P, Brunereau L, Laberge S, Houtteville JP: Prospective follow-up of
69. Reisch R, Bettag M, Perneczky A, Kawase T, Toya S, Shiobara R, Mine T:
33 asymptomatic patients with familial cerebral cavernous malformations.
Transoral transclival removal of anteriorly placed cavernous malformations
Neurology 57:1825–1828, 2001.
of the brainstem. Surg Neurol 56:106–116, 2001.
44. Labauge P, Brunereau L, Lévy C, Laberge S, Houtteville JP: The natural his-
70. Robinson JR, Awad IA, Little JR: Natural history of the cavernous angioma.
tory of familial cerebral cavernomas: A retrospective MRI study of 40 patients.
J Neurosurg 75:709–714, 1991.
Neuroradiology 42:327–332, 2000.
71. Sakai N, Yamada H, Tanigawara T, Asano Y, Andoh T, Tanabe Y, Takada M:
45. Lapras C, Lapras C, Deruty R, Patet JD, Mottolèse C: Surgical treatment of
Surgical treatment of cavernous angioma involving the brainstem and review
intracerebral cavernomas [in French]. Neurochirurgie 35:120–125, 1989.
of the literature. Acta Neurochir (Wien) 113:138–143, 1991.
46. Lena G, Paz-Paredes A, Choux M: Brain stem cavernomas in children. Nine
72. Samii M, Eghbal R, Carvalho GA, Matthies C: Surgical management of brain-
case reports and literature review [in French]. Neurochirurgie 48:319–325,
stem cavernomas. J Neurosurg 95:825–832, 2001.
2002.
73. Sandalcioglu IE, Wiedemayer H, Secer S, Asgari S, Stolke D: Surgical removal
47. Lesniak MJ, Clatterbuck RE, Rigamonti D: Vascular malformations of the
of brain stem cavernous malformations: Surgical indications, technical consid-
posterior fossa: Clinical features, treatment and outcomes. Neurosurg Q
erations, and results. J Neurol Neurosurg Psychiatry 72:351–355, 2002.
13:257–267, 2003.
74. Scott RM: Brain stem cavernous angiomas in children. Pediatr Neurosurg
48. Lewis AI, van Loveren HR, Tew JM Jr: Management of brain stem vascular
malformations: Advances in surgical technique and long- term results. 16:281–286, 1990.
Neurosurg Q 5:217–228, 1995. 75. Scott RM, Barnes P, Kupsky W, Adelman LS: Cavernous angiomas of the
49. Liscák R, Vladyka V, Simonová G, Vymazal J, Novotny J Jr: Gamma knife central nervous system in children. J Neurosurg 76:38–46, 1992.
radiosurgery of the brain stem cavernomas. Minim Invasive Neurosurg 76. Seifert V, Raabe A, Zimmermann M: Conservative (labyrinth-preserving)
43:201–207, 2000. transpetrosal approach to the clivus and petroclival region—Indications, compli-
50. Liu KD, Chung WY, Wu HM, Shiau CY, Wang LW, Guo WY, Pan DH: Gamma cations, results and lessons learned. Acta Neurochir (Wien) 145:631–642, 2003.
knife surgery for cavernous hemangiomas: An analysis of 125 patients. 77. Shehab ZP, Walsh RM, Thorp MA, Wallace MC, Tymianski M, Rutka JA: Partial
J Neurosurg 102 [Suppl]:81–86, 2005. labyrinthectomy approach for brainstem vascular lesions. J Otolaryngol
51. Mao Y, Zhou L, Du G: Brain-stem cavernous hemangioma: Surgical indications 30:224–230, 2001.
and approaches [in Chinese]. Zhonghua Wai Ke Za Zhi 39:672–674, 2001. 78. Sindou M, Yada J, Salord F: Functional results after microsurgical resection of
52. Mathiesen T, Edner G, Kihlström L: Deep and brainstem cavernomas: A con- brain stem cavernous malformations (retrospective study of a 12 patient series
secutive 8-year series. J Neurosurg 99:31–37, 2003. and review of the recent literature). Acta Neurochir (Wien) 142:843–853, 2000.
53. Mizoi K, Yoshimoto T, Suzuki J: Clinical analysis of ten cases with surgically 79. Smith ER, Chapman PH, Ogilvy CS: Far posterior subtemporal approach to
treated brain stem cavernous angiomas. Tohoku J Exp Med 166:259–267, 1992. the dorsolateral brainstem and tentorial ring: Technique and clinical experi-
54. Morcos JJ, Heros RC, Frank DE: Microsurgical treatment of infratentorial ence. Neurosurgery 52:364–368, 2003.
malformations. Neurosurg Clin N Am 10:441–474, 1999. 80. Sola RG, Pulido P, Pastor J, Ochoa M, Castedo J: Surgical treatment of symp-
55. Moriarity JL, Wetzel M, Clatterbuck RE, Javedan S, Sheppard JM, Hoenig- tomatic cavernous malformations of the brainstem. Acta Neurochir (Wien)
Rigamonti K, Crone NE, Breiter SN, Lee RR, Rigamonti D: The natural history 149:463–470, 2007.
of cavernous malformations. A prospective study of 68 patients. Neurosurgery 81. Spetzler RF, Daspit CP, Pappas CT: The combined supra- and infratentorial
44:1166–1173, 1999. approach for lesions of the petrous and clival regions: Experience with 46
56. Mussi AC, Rhoton AL Jr: Telovelar approach to the fourth ventricle: cases. J Neurosurg 76:588–599, 1992.
Microsurgical anatomy. J Neurosurg 92:812–823, 2000. 82. Steiger HJ, Hänggi D, Stummer W, Winkler PA: Custom-tailored transdural
57. Nataf F, Roux FX, Devaux B, Page P, Turak B, Dezamis E, Abi Lahoud G: anterior transpetrosal approach to ventral pons and retroclival regions.
Brainstem cavernomas: Surgical experience at the CH Sainte-Anne general J Neurosurg 104:38–46, 2006.
hospital [in French]. Neurochirurgie 53:192–201, 2007. 83. Steinberg GK, Chang SD, Gewirtz RJ, Lopez JR: Microsurgical resection of
58. Pechstein U, Zentner J, Van Roost D, Schramm J: Surgical management of brainstem, thalamic and basal ganglia angiographically occult vascular mal-
brain-stem cavernomas. Neurosurg Rev 20:87–93, 1997. formations. Neurosurgery 46:260–271, 2000.

NEUROSURGERY VOLUME 64 | NUMBER 5 | MAY 2009 | 817


GROSS ET AL.

84. Strauss C, Romstöck J, Nimsky C, Fahlbusch R: Intraoperative identification location and the clinical outcome associated with these approaches. A
of motor areas of the rhomboid fossa using direct stimulation. J Neurosurg limited but concise review of the radiosurgical data is given, and a
79:393–399, 1993. brief discussion of recent advances, including neuronavigation, intra-
85. Symon L, Jackowski A, Bills D: Surgical treatment of pontomedullary caver- operative endoscopy, and various imaging techniques (constructive
nomas. Br J Neurosurg 5:339–347, 1991.
interference in steady state, diffusion tensor imaging, white matter
86. Tung H, Giannotta SL, Chandrasoma PT, Zee CS: Recurrent intraparenchymal
hemorrhages from angiographically occult vascular malformations. tractography), completes this review.
J Neurosurg 73:174–180, 1990. We agree with many points the authors have made regarding the
87. Viñas FC, Gordon V, Guthikonda M, Diaz FG: Surgical management of cav- management of brainstem CMs, and we pursue a similar management
ernous malformations of the brainstem. Neurol Res 24:61–72, 2002. strategy. However, we do not necessarily insist on a second clinically
88. Vishteh AG, David CA, Marciano FF, Ernesto C, Spetzler RF: Extreme lateral significant hemorrhage before deciding on surgical treatment, if the
supracerebellar infratentorial approach to the posterolateral mesencephalon: first bleed caused a significant neurological deficit. Additionally, there
Technique and clinical experience. Neurosurgery 46:384, 1999. are occasional cases of repeated hemorrhage or hemorrhage causing
89. Wang CC, Liu A, Zhang JT, Sun B, Zhao YL: Surgical management of brain-
serious disability in which we would still recommend surgery even
stem cavernous malformations: Report of 137 cases. Surg Neurol 59:444–454,
2003.
though the lesion does not present to a pial or ependymal surface, pur-
90. Weil SM, Tew JM Jr: Surgical management of brain stem vascular malforma- suing a relatively safe corridor to lateral pontine or mesencephalic
tions. Acta Neurochir (Wien) 105:14–23, 1990. lesions (as described in this review).
91. Yaşargil MG: AVM of the brain, in Yaşargil MG (ed): Microneurosurgery IIIB. We also agree that there is currently a lack of good evidence to sup-
Stuttgart, Thieme, 1988, pp 405–438. port radiosurgical treatment of these lesions, and we concur that tem-
92. Yoshimoto T, Suzuki J: Radical surgery on cavernous angioma of the brain- poral clustering of the hemorrhage rate may confound the interpreta-
stem. Surg Neurol 26:72–78, 1986. tion of the radiosurgical data. The review also helps to highlight the
93. Zabramski JM, Wascher TM, Spetzler RF, Johnson B, Golfinos J, Drayer BP,
difficulty in building consensus among the varied data on hemorrhage
Brown B, Rigamonti D, Brown G: The natural history of familial cavernous
rates (2.3%–6.8% per patient-year) and rehemorrhage rates (3.8%–22.9%
malformations: Results of an ongoing study. J Neurosurg 80:422–432, 1994.
94. Zausinger S, Yousry I, Brueckmann H, Schmid-Elsaesser R, Tonn JC: Caver- annually). Perhaps of more importance is finding consensus on clini-
nous malformations of the brainstem: Three-dimensional-constructive inter- cally significant event rates.
ference in steady-state magnetic resonance imaging for improvement of sur-
Marco Lee
gical approach and clinical results. Neurosurgery 58:322–330, 2006.
95. Zhao J, Wang Y, Kang S, Wang S, Wang J, Wang R, Zhao Y: The benefit of neu- Gary K. Steinberg
ronavigation for the treatment of patients with intracerebral cavernous mal- Stanford, California
formations. Neurosurg Rev 30:313–319, 2007.
96. Zimmerman RS, Spetzler RF, Lee KS, Zabramski JM, Hargraves RW:
Cavernous malformations of the brain stem. J Neurosurg 75:32–39, 1991.
97. Ziyal IM, Sekhar LN, Salas E, Sen C: Surgical management of cavernous mal-
T he authors present an extremely thorough review of brainstem CMs
based on the relevant literature and their own expertise in cere-
brovascular neurosurgery. Cogent points in the natural history section
formations of the brain stem. Br J Neurosurg 13:366–375, 1999. include a discussion on how the estimated annual bleeding rate of CMs
varies widely in previous reports on the basis of the definition of hem-
COMMENTS orrhage, and the observation that many patients have complete neuro-
logical recovery after a single bleeding episode. The recommendation

G ross et al. present a systematic review of the literature of brainstem


cavernous malformations (CMs). Natural history studies, surgical
series, and radiosurgical series are included. The authors nicely summa-
that excision of these difficult lesions should be limited to patients with
more than 1 hemorrhage and a pial or ependymal presentation is pru-
dent. Referral of these patients to a center or surgeon with extensive
rize these studies, including patient outcomes and surgical approaches, experience could also have been included in their criteria.
and provide treatment recommendations based on their analysis.
We agree that management strategies must weigh the risks of surgi- Bruce E. Pollock
cal intervention against those of observation. Surgical resection is indi- Rochester, Minnesota
cated for symptomatic patients with accessible lesions. We agree with
the authors that image guidance, cranial nerve monitoring, and motor
and somatosensory evoked potential monitoring are important T his is an excellent review article that performs a very critical analy-
sis of brainstem CMs. Since the previous publication of my small
series (1), I have operated on an additional 20 patients with brainstem
adjuncts for safe resection of these difficult lesions. We strongly agree
that associated venous anomalies should be preserved to maintain cavernomas. The results have remained approximately the same, with
venous drainage of normal brain. We also agree that surgical excision 70% of the patients making a good recovery at 3 months. This means
of brainstem and other parenchymal CMs is the treatment of choice. that 30% also did worse after the surgery. The natural history of these
lesions is still not well defined, although it is clear that it is much worse
Peter H. Maughan than for cavernomas in other parts of the brain. I am not impressed
Robert F. Spetzler with the results of radiosurgery, although my colleagues from
Phoenix, Arizona Pittsburgh believe otherwise. I believe that, with better imaging of the
brainstem nuclei and tracts and better surgical technique (curvilinear
T he authors have produced a comprehensive, thoughtful, and excel-
lent account of current controversies and clinical management of
brainstem CMs. Of particular note, they have gathered all the studies
methods of approach rather than linear, as at present), the results of
surgical excision can be improved further.
published in the English and French language on brainstem CMs and Laligam N. Sekhar
gathered the natural history, surgical, and radiosurgical data for analy- Seattle, Washington
sis and discussion. Data on the hemorrhage rate, clinically significant
event rate, rehemorrhage rate, and morbidity and mortality rates are
discussed. This is followed by a review of the relevant published series 1. Ziyal IM, Sekhar LN, Salas E, Sen C: Surgical management of cavernous mal-
of the surgical approaches to these challenging lesions according to formations of the brain stem. Br J Neurosurg 13:366–375, 1999.

818 | VOLUME 64 | NUMBER 5 | MAY 2009 www.neurosurgery-online.com

Das könnte Ihnen auch gefallen