Acta Neurochir (Wien) (2000) 142: 843±853
Acta Neurochirurgica
Clinical Articles
Functional Results After Microsurgical Resection of Brain Stem Cavernous
Malformations (Retrospective Study of a 12 Patient Series and Review of the
Recent Literature)
M. Sindou1, J. Yada1, and F. Salord1
1 Department of Neurosurgery, HoÃpital Neurologique, P. Wertheimer, University of Lyon, France
2 Department of Anesthesiology, HoÃpital Neurologique, P. Wertheimer, University of Lyon, France
Summary
Background. Since advent of MRI, brain stem Cavernous Mal-
formations (CM) can be easily diagnosed, and their curative surgical
resection considered under precise conditions. The authors report a
consecutive series of twelve patients with CMs surgically treated and
histopathologically con®rmed. Eleven of the cases had bled (six more
than once). In this study special emphasis has been put on the pre and
post-operative functional status of the patients, by using the 100
Karnofsky scale (KS).
Method. Surgical approaches were: 1 ) supra-occipital trans-
tentorial for 1 thalamomesencephalic and 1 quadrigeminal plate
CM, 2 ) suboccipital infratentorial supracerebellar for 1 dorsolateral
mesencephalic CM, 3 ) retrosigmoid through the cerebello-pontine
angle for 3 pontine and/or medullary CM, 4 ) suboccipital inter-
tonsillar for 6 CM located under the ¯oor of the IVth ventricle.
Completeness of removal was checked by postoperatoire MRI. It
was complete in 11 cases and only partial in 1 (i.e., in the case with
the progressing mass-e¨ect presentation). There was no post-op
death. Follow-up ranged from 1 to 7 years.
Findings. Preoperatively: 2 patients were operated on in a coma-
tose state (KS U 20), 5 were in state of functional dependance
(K U 60) and 5 had severe neurological de®cits but were still of
independant functional status (KS V 60). At one year after surgery:
3 patients had a KS V 80 (i.e., they could resume their prior normal
life), 6 had a KS between 60 and 80 (i.e., they were independant)
and 3 had a KS below 60 (i.e., they were dependant especially for
walking).
Interpretation. Our results, as well as the data harvested from the
literature, plead for advocating radical surgical resection at least in
patients with exophytic CMs having bled. As a matter of fact, study
of the natural history shows that in brain stem CMs, the bleeding risk
amounts to 21% per year per patient. Review of literature shows
evidence that radiosurgery did not prove e¨ective and/or even
innoccuous.
Keywords: Brain stem; cavernoma; Cavernous Malformation;
microsurgical removal.
> Springer-Verlag 2000
Printed in Austria
Introduction
The diagnosis of brain stem Cavernous Malforma-
tions (CMs) has greatly bene®ted from the advent of
MRI. This tool facilitated the identi®cation and survey
of this lesion even in pauci-symptomatic lesions, which
helped to study the natural history of these malforma-
tions. According to data in the literature [8, 9, 10, 14,
17, 21, 27, 31, 35], brain stem CMs account for 17% on
average of all symptomatic intracranial CMs, with an
extreme of 35% in the study on natural history of CMs
by Kondziolka et al. [14].
If considering overall intracranial CMs, the risk of
bleeding is 0.5% per year for cavernomas which have
not bled and 4.5% per year for those that have already
bled [7, 22]. For brain stem CMs, the annual rate of
hemorrhagic risk is much higher [1, 5, 9, 10, 14, 18, 24,
35]; in Fritschi et al's article [9] it has been estimated
at 2.7% for a patient without a prior bleed and at 21%
for a patient with previous hemorrage. In the surgical
series of Porter et al. [21] 97 of the 100 patients referred
had bled. 54 of these 97 patients (56%) had multiple
hemorrhages and 21 of them (22%) had more than two
hemorrhages.
Growth evolution with mass-e¨ect and/or repeated
hemorrhages almost constantly result in irreversible
severe neurological de®cits. In a retrospective study of
a group of 30 patients treated conservatively [9], mor-
tality rate amounted to 20%.
844
Table 1. Karnofsky Rating Scale
Score Findings
100 normal, no complaint
90 normal activity, minor symptoms
80 normal activity with e¨ort. Some symptoms
70 care for self. Unable to carry on normal activity
60 requires occasional assistance
50 requires considerable assistance and care
40 disabled, requires special care and assistance
30 severely disabled, hospitalized
20 very sick, requires supportive measures
10 Moribund
Because it appeared that microsurgical techniques
could render surgical resection of CMs feasible at the
brain stem, an increasing number of neurosurgeons
in the nineties started to treat symptomatic brain stem
CMs surgically, especially those with a super®cal
location under the pial surface [4, 5, 6, 8, 9, 10, 16, 18,
19, 20, 21, 23, 24, 30, 31, 32, 33, 34, 35].
This article is the analysis of a consecutive series of
12 patients a¨ected by a brain stem CM and surgically
treated at the University Neurological Hospital of
Lyon, between 1992 and 1998, by the senior neuro-
surgeon (M.S.) Special emphasis has been put on the
study of the pre-and post-operative functional status of
the patients measured by the Karnofsky scale [12]
(Table 1).
Clinical Material and Methods
Patients' Presentation
The 12 patients were referred to our institution because of a sud-
den or rapidly progressing deterioration of their neurological status.
Nine patients were male and three female. Age ranged from 24 to
63 years, 43 on average. Eleven patients were referred for a hemor-
rhage. For ®ve of them, it was the ®rst bleeding; in the six others, it
was a rebleeding (the second one in 5 patients and the third in
one). Only one patient was referred for a progressing mass-e¨ect
syndrome.
All patients had MRI at the ®rst imaging investigation or after a
CT. All of them had in addition four vessels angiography by femoral
catheterization. Angiography was performed to provide the maxi-
mum vascular anatomy information including a detailed study of the
venous system, before surgical decision was made and the choice of
the surgical approach decided upon.
CM location was: lateral thalamo-mesencephalic in 1 case, lateral
mesencephalic in 2 (one involving the quadrigeminal plate and one
the tegmentum), in the pons in 5 (on the midline in 4 and laterally
in 1), at the ponto-medullary junction in 3 (2 on the midline and 1
laterally), in the medulla, laterally, in 1.
One patient (with a solitary brain stem CM) had a family history
of CMs, and two had multiple localizations. Two patients had an
associated clear developmental venous abnormality in the close
vicinity of the CM, and two others a simple but particularly marked
vein of drainage of the cavernoma. One patient with a midline pon-
M. Sindou et al.
tine CM had a Chiari II malformation (that was treated by cervico-
occipital posterior decompression at the same time as the resection of
the cavernoma).
All patients, with the exception of one, had headaches, more or
less intense. All had multiple neurological symptoms and signs (cra-
nial nerve de®cits, motor weakness, sensory disturbances, cerebellar
discoordination. . . .) depending upon the rostro-caudal level of the
CM and its location within the brain stem. Two patients soon after
admission to the hospital were in coma with severe vegetative dis-
turbances and respiratory distress, necessitating an emergency tra-
cheal intubation with assisted ventilation. The detailed clinical pic-
ture for each patient will not be described in the article; only the
predominant clinical manifestation(s) will be mentioned. Table 2
summarizes the preoperative data of the patients in this series.
Only two patients were operated on within the ®rst three days after
the (last) bleed. The others had surgery in the subacute phase, which
ranged from ten days to 10 months (seven weeks on average), after
the clinical status had improved and was stabilized.
The severity of the patients' functional status at the time of oper-
ation was quanti®ed by evaluating the degree of dependance/activity
with the Karnofsky scale (KS) (see corresponding table). Two pa-
tients were operated on in comatose state (KS U 20), ®ve in a state of
functional dependance (KS U 60) and ®ve with severe neurological
de®cits but still with independant functional status (KS V 60) (Table
2).
Surgical Treatment
All patients of this series were referred by their neurologist for
surgical resection of the CM after consent had been obtained from
both the patient and relatives.
The Choice of the Surgical Approach was essentially based on the
location of the CM, i.e., its topography according to rostro-caudal
level, its ventro-dorsal situation within the brain stem, its more or
less super®ciality under the brain stem surface. The pre-existing
neurological disorders, as well as the foreseeable new de®cits inher-
ent to the intented incision of the overlying parenchyma, were of
course taken into account for designing the surgical approach (Table
3).
The surgical approach was:
± Supraoccipital transtentorial, on the side of the lesion, in two pa-
tients: once for a thalamo-mesencephalic CM located in the pul-
vinar, the subthalamic area and the upper part of the tegmentum;
once for a CM involving the quadrigeminal plate, (Fig. 1).
± Suboccipital infratentorial-supracerebellar, on the side of the
cavernoma, once; it was for a CM located within the mesen-
cephalon and bulging under the dorso-lateral surface of the cere-
bral peduncle.
± Retrosigmoid and through the cerebello-pontine angle, in three
cases of CM located ventro-laterally in the pons and/or the
medulla, (Fig. 2).
± Through a median low suboccipital (and intertonsillar) approach.
This approach passed through the Foramen of Magendie, that
was enlarged by displacing laterally each tonsil after having freed
and opened both cerebello-medullary ®ssures up to the Luschka
lateral foramen of the IVth ventricle. This approach was used
to remove ®ve CMs located at, or in the vicinity of, the midline
under the ¯oor of the IVth ventricle at the pontine or the ponto-
medullary levels. Same approach was used, but had to be com-
pleted by a splitting of the lower part of the vermis in the case in
whom a Chiari malformation was associated with the pontine CM
(Fig. 3).
Surgical Steps. All patients were operated on under general anes-
thesia in the sitting position. Microsurgical techniques were used in
Functional Results in Brain Stem Cavernomas 845

Table 2. Patient Series (pre-op. Data)

Case Sex Location Bleeding (s) Dominant revealing Recent Delay Pre-op.
n age (y) symptom/sign bleeding between functional
on MRI last warning status
symptoms (Karnofsky
and surgery scale)

1 M34 mesenceph. quad. plate (lateral) three ataxia ‡ 2 w 90

2 M63 thal.-mesenceph (lateral) two ataxia ‡ 2 m 70
3 F57 pons (¯oor IVth V) (median) two diplopia ‡ 10 m 90
4 M59 pons (lateral) one hemiparesia ‡ 10 d. 60
5 F24 pons (¯oor IVth V) / tetraparesia ÿ / 60
multilobulated mass
6 M43 pons (¯oor IVth V) (median) one comatose state ‡ 3d 20
7 M63 pons (¯oor IVth V) (median) one headaches, ataxia ‡ 15d 100
8 F46 ponto-medullary (lateral) two dysphagia, dysphonia, ‡ 2d 40
hemiparesia
9 M38 ponto-medullary (¯oor IVth V) one comatose state ‡ 15d 20
(median)
10 M35 medullary (lateral) one headaches, dysphagia, ‡ 26d 60
dysphonia
11 M56 mesenceph. (lateral) two ataxia, diplopia ‡ 15d 50
12 M33 ponto-medullary (¯oot IV th V) two diplopia ataxia ‡ 25d 60
(paramedian)

all patients as soon as the craniotomy was completed. In all patients
the exposed brain stem surface had a xanthochromic coloration at
the site of the CM. In the majority of cases there was an additional
dark-blue coloured small area, which corresponded to the bulging
hematoma surrounding the cavernoma. When present, incision of
the brain stem surface was performed at this very point.
The incision was as small as possible (less than 10 mm in most
cases) to avoid increasing the neurological de®cits. After having
sucked out the clots of the hematoma, and also the blood coming
from inside the malformation as well as the serous liquid from even-
tual microcystic cavities, the cavernoma was collapsed and detached
from the gliotic surrounding tissue. Then the mass was disconnected
by using bipolar coagulation and microscissors from the tiny feeding
and draining vessels. When big, the draining veins were respected
at all costs, the cavernous mass being disconnected from the
vein(s), passing as close as possible to the cavernoma. In the two pa-
tients with an associated developmental venous abnormality, the
venous malformation was not touched after the CM was removed
(Fig. 4). After the CM has been resected and removed, the cavity
was washed with saline, and its walls covered with small pieces of
Surgicel.
After surgery, the patients were admitted to the Intensive Care
Unit for the early post-operative period.
Results (Table 3)
Cure of the Lesion
In the 11 patients in whom CM was revealed by
bleeding, the hematoma was totally evacuated and
the cavernous mass could be entirely removed. Com-
pleteness of the removal was checked again by post-
operative MRI performed before discharge from the
hospital, and checked one year later. This new check-
up at one year was important to be certain of the
total eradication of the lesion especially if at the early
control there was some contamination of the cavity by
residual blood or ``hematic'' surgicel deposits on its
walls.
In the case with the giant multilobulated pontine
CM, which did not bleed, but was operated on because
of threatening pseudo-tumorous evolution, the resec-
tion was only partial. For, it was very di½cult in this
case to ®nd clearcut boundaries between the cavernous
masses and the surrounding parenchyma.
Post-Operative Events
All patients were hospitalized in the Intensive Care
Unit immediately after surgery, whatever their preop-
erative status might be. The two patients operated on
in comatose state recovered consciousness within a few
days after surgery. Assisted ventilation was withdrawn
after three weeks in both patients, one of them after
completion of a tracheostomy that could be removed
some weeks later.
The 10 patients operated on in conditions of a good
level of consciousness awoke normally (i.e., with a
Glasgow Score superior or equal to 14/15) and were
extubated at the 24th hour. But six of them because of
secondary respiratory disturbances necessited tracheal
846 M. Sindou et al.

Table 3. Patient Series (Post-op. Data)

Case Location Pre-op Surgical Removal Main post-op Duration Post-op Functional Overall
n functional approach T (total) event(s) ICU functional status result
status P (partial) (days) status at at one year
(Karnofsky discharge (Karnofsky
scale) (Karnofsky scale)
scale)

1 mesenceph. quad. 90 supra-occipital T respiratory 40 60 60 ‡‡ &

plate (lateral) transtentorial disturbances
(lateral)
2 thal.-mesenceph 70 supra-occipital T frontal 15 70 90 ‡%
(lateral) transtentorial hydroma
(lateral) (transient)
3 pons (¯oor IVth V) 90 subtonsillar, ¯oor T respiratory 16 70 90 ÿ!
(median) of IVth V disturbances
4 pons (lateral) 60 retro-sigmoid, T respiratory 15 60 90 ‡‡ %
cerebello- disturbances
pontine
5 pons (¯oor IVth V) 60 subtonsillar, ¯oor P CSF ®stula 9 40 50 ‡&
multilobulated of IVth V (transient
mass after lumbar
drainage)
6 pons (¯oor IVth V) 20 subtonsillar, ¯oor T hydrocephalus 13 40 60 ‡‡ %
(median) of IVth V (permanent,
CSF shunt)
7 pons (¯oor IVth V) 100 subtonsillar, ¯oor T none 3 90 100 ÿ!
(median) of IVth V
8 ponto-medullary 40 retrosigmoid, T respiratory 13 70 80 ‡‡ %
(lateral) cerebello- disturbances
pontine
9 ponto-medullary 20 subtonsillar, ¯oor T respiratory 9 50 90 ‡‡ %
(¯oor IVth V) of IVth V disturbances
(median)
10 medullary (lateral) 60 retrosigmoid, T respiratory 11 60 70 ‡%
cerebello- disturbances
pontine
11 mesenceph. (lateral) 50 sub-occipital T respiratory 7 60 80 ‡‡ %
infratentorial, disturbances
supra-
cerebellar
(lateral)
12 ponto-medullary 60 subtonsillar, ¯oor T none 5 90 90 ‡‡ %
(¯oor IVth V) of IVth V
(paramedian)

reintubation ± some with assisted ventilation ± be-
tween the third and the ®fth day after surgery. For
these six patients duration of their stay in ICU ranged
from 10 to 15 days according to cases (Table 4).
Of the 8 overall patients who needed respiratory
assistance (i.e., the two operated on in comatose state
and the six who needed secondary arti®cial ventilation
because of respiratory aggravation) four had to be
tracheostomized (transiently in three, for two weeks on
average), so as to facilitate suppression of ventilatory
assistance.
No other major complication(s) occured in our
series with the exception in one patient of kidney fail-
ure which necessitated transient dialysis.
Long-Term Outcome
Follow-up in the series ranged from 1 to 7 years
(5.5 years on average). There was no mortality in this
series, neither in the early post-operative period nor in
the follow-up period.
Functional Results in Brain Stem Cavernomas 847

Fig. 1. Mesencephalic CM located in the right part of the quadrigeminal plate (left Fig., MRI-T2, preoper.), totally removed through a right
supra-occipital transtentorial approach (right Fig., MRI-T1, post-oper.)

Fig. 2. Pontine CM located in the right part of the upper pons with a huge hematoma (left Fig., MRI-T1, pre-oper.), totally removed
through a right cerebello-pontine angle approach with incision of the pons below pars major of the trigeminal nerve (right Fig., MRI-T2, post-
oper.)

At one year after surgery, the functional neurologi-
cal status was evaluated using the Karnofsky scale. 9
patients had a score superior to 60 ± which means that
they were independant for daily life ± three of them
being above 80 (i.e., that they could resume a normal
activity in spite of some symptoms and slight de®cits).
3 had a score inferior to 60: one (at 50) was requiring
occasional help (for walking) and two (at 40) required
considerable assistance and care.
Comparison of the functional status, at one year
after surgery, with the one before surgery, shows
that: 8 patients had signi®cantly improved, 2 were
stabilized, and 2 were worse: one slighlty and one sev-
erely, the latter needing a permanent tracheostomy
and a wheelchair.
For the patients with a follow-up longer than one
year, there was no secondary deterioration except in
one who required a CSF shunt for a normal pressure
hydrocephalus syndrome and did well after the shunt
procedure.
848
Fig. 3. Pontine CM located in the middle part of the lower pons in a patient with a Chiari malformation (left Fig., MRI-T1, pre-oper.), totally
removed through a (transvermian) approach of the ¯oor of the IVth ventricle
Discussion
Results of Surgery
The present series was characterized by the existence
of preoperatively severe clinical disturbances in all the
patients. The mean Karnofsky score of the series was
60% at the time of surgery despite most of the patients
being operated on after an average delay of seven
weeks during which the clinical status had improved
until stabilization. With the exception of one patient
who presented with a progressive mass-e¨ect syn-
drome, all patients had one (or several) bleed(s) from
their malformation. Because, in all these patients,
evolution was threatening, we decided together with
the refering neurologist, to convince the patient and his
family ± that surgical resection was the best therapeu-
tic option. This policy seemed to us logical especially
when considering that CMs treated conservatively
generally lead to severe disability and entail a mor-
tality rate of the order of 20% [35]. The radiosurgical
option, in spite of some articles claiming that radio-
surgery might be e¨ective in CMs [13, 28], did not
prove to be e½cacious or even innocuous [2, 11, 26,
29].
There was no mortality in this series. A majority of
the patients improved after surgery; only one was sig-
ni®cantly worse. The mean Karnofsky activity score
increased from 60 (before) to 70 (after) surgery.
These good surgical results (i.e., 91% of the patients
cured without signi®cant deterioration) are concor-
dant with the ®gures reported in the literature (Table
M. Sindou et al.
5): from 83% to 89% according to the series (Berta-
lan¨y et al. 13 pts surgically treated [4], Fahlbusch and
Strauss. 10 pts [8] Zimmerman et al. 16 pts [34], Frit-
schi et al. 93 pts [9], Sathi et al. 23 pts [25], Amin-
Hanjani et al., 14 pts [3], Porter et al. 86 pts [21], Ziyal
et al. 9 pts [35].
Surgical Indications and Anatomical Considerations
Indications constitute a di½cult problem to deal
with. Most authors agree that among brain stem CMs
those which are asymptomatic, or pauci-asympto-
matic if quiescent, have not to be touched surgically,
especially if they are small and enclosed with a large
rim of parenchyma around them. On the contrary for
symptomatic brain stem CMs, as soon as they manifest
a tendency to rebleed, literature date as well as the
results from our own series, plead for a decision for
surgical resection before mass-e¨ect and/or further
bleed(s) cause severe de®cits. As a matter of fact in
these «symptomatic» forms of CMs, spontaneous
evolution generally leads to irreducible disability.
As optimal timing for surgery, most surgeons prefer
the subacute phase, after the patient's clinical status
has stabilized and the hematoma has become partially
organized.
If the CM corresponds to the so-called «exophytic»
type -i.e., has a super®cial subpial extension ± there is
no problem of localizing the lesion and for removing it
without signi®cant risk. On the contrary if the CM is
deeply situated inside the brain stem parenchyma, the
risk to produce neurological worsening has to be seri-
Functional Results in Brain Stem Cavernomas 849

Fig. 4. Cavernous angioma associated with a developmental venous abnormality. Upper left: paramedian ponto-medullary cavernous mal-
formation under the ¯oor of the IVth ventricle. Upper right: axial views showing the cavernoma and the developmental venous anomaly. Lower
left: sagittal view, and lower right: axial view, after operation, showing complete removal of the cavernoma and preservation of the venous
abnormality

ously evaluated before a decision is taken. A careful
neurological examination and a precise MRI study
with a rigorous elucidation of the brain stem surgical
anatomy are important prerequiritives. As a matter of
fact, in patients in whom the cavernoma is covered by
a signi®cant rim of neural tissue, the new de®cits that
the incision will logically create to have access to the
lesion have to be taken into account and objectively
discussed with the patient and his family.
Although brain stem structures may be distorted
because of the CM pathology, it may be of interest
to have in mind some important elements of surgi-
cal anatomy when considering surgical approaches.
Motor tracts are located in the ventral portion of the
brain stem. Traversing through the ventral aspect and/
or damaging the perforating vessels would result in
pyramidal de®cits. The tegmentum with the reticular
formation and the medial longitudinal fasciculus
850
Table 4. Pre- and Post-Operative Respiratory Status in the Patients' Series
Scale of respiratory status
Grade De®nition
R0 normal respiratory state
R1 slight respiratory disturbances (without
the necessity for tracheal intubation)
R2 respiratory disturbances necessitating
tracheal intubation, but without the
need of assisted ventilation
R3 respiratory disturbances necessitating
tracheal intubation, with the need of
partially controlled ventilation
R4 respiratory disturbances necessitating
tracheal intubation, with the need
for full-time controlled ventilation
* RT Tracheostomy (permanent)
occupy the mid and dorsal portions of the brain stem.
The cranial nerve and vestibular nuclei and the extra-
lemniscal sensory tracts are located in the dorsal and
lateral parts. For lesion which are situated under the
¯oor of the IVth ventricle, the nuclei of the abducens
and facial nerves (facial colliculus) in the paramedian
region and the medial longitudinal fasciculus on the
midline, are structures at risk. Detection of the facial
nuclei by electrical stimulation of the ¯oor of the IVth
ventricle can be a useful tool for identifying their loca-
tion. Two «safe entry zones» have been described: one
above and one below the facial colliculus over 10 mil-
limeters in length heigth each [15]. For lesion situated
ventrolaterally within the brain stem, approach via an
antero-lateral brain stem surface incision is generally
well-tolerated, especially at the level of the pons [21]
but also of the medulla [6].
Conclusions
± Since modern MR imaging, brain stem Cavernous
Malformations (CMs) can be easily detected and
their anatomical situation and relationships pre-
cisely de®ned.
± Thanks to microsurgical techniques, most CMs can
be cured with satisfactory results, i.e., with a com-
plete resection of the lesion and without signi®cant
neurological deterioration, provided they are «sur-
gically accessible» and reached through an «appro-
priate minimally invasive approach».
± «Symptomatic» CMs manifesting with bleeding(s),
M. Sindou et al.
Pre-oper D1 D4 On discharge
Post-oper Post-oper from the hospital
to home
9 7 2 11
1 3 2 1 (RT)
3
1
2 (comatous 2 4
patients)
or a progressing mass-e¨ect syndrome, are better
operated on. As a matter of fact, according to
literature data or experience from our institution,
spontaneous evolution generally leads to severe
and irreversible disability. Radiosurgery has not
been proven to be e¨ective and/or innocuous in the
long-term.
± Optimal timing is after the patient's status has
stabilized.
± Exophytic CMs are generally easy to remove with-
out creating additional de®cits. Enclosed CMs would
have to be operated on only if passing through the
overlying parenchymal rim would not create new
severe de®cits, unless the cavernoma demonstrates a
dangerous tendency to bleed.
± The surgical incision has to be at the very point
where the CM and/or the satellite hematoma is (are)
the most super®cial under the brain stem surface.
The surgical approach has to be designed to reach
this entry point.
± Hospitalization in the Intensive Care Unit for the
®rst post-operative week is a prudent measure as
even patients who awake normally from anesthesia
can secondarily worsen (hopefully, transiently) their
respiratory conditions.
± Asymptomatic CMs or those with minor manifes-
tations should be managed conservatively, espe-
cially if they are small and deeply situated and en-
closed with in a wide rim of neural tissue. They
should be monitored over a limited period and fre-
quent re-examinations.
Functional Results in Brain Stem Cavernomas 851

Table 5. Literature Series

Series of b.s CMs Total nb (nb Nb of totally Outcome Surgical removal recommended
treated surgically presenting with resected
hemorrhage)

Symon et al. 7 7 ± mortality: none with an appropriate incision being made into that part
1991 (7) ± improved: 7 of the brain stem most directly overlying the lesion.
Bertalan¨y et al. 13 13 ± mortality: none In symptomatic CMs having bled, in the subacute
1991 ± improved: 8 phase when the hematoma is partially organized
± unchanged: 1 and the patient's condition has stabilized.
± aggravated: 4
Fahlbusch and 10 10 ± mortality: none in patients with recurrent hemorrhage, persistent or
Strauss 1991 (10) ± improved: 6 progressive neurological de®cit, during the subacute
± unchanged: 3 stage. Late surgery increases the risk of additional
± aggravated 1 neurological de®cit.
Zimmerman et al. 16 13 ± mortality: 1 when the lesion is located super®cially and an
(1991) (10) ± improved or operative approach can spare eloquent tissue.
unchanged: 14
± aggravated: 1
Fritschi et al. 93 86 ± mortality: none for symptomatic lesions with either a sudden onset
1994 (75) ± improved: 78 attributable to a hemorrhage or a slowly progressive
± aggravated neurological de®cit. Deep lesions that lack a
(moderately 14, super®cal extension are unlikely to have a good
severely): 1 surgical outcome because eloquent structures will be
injured.
Pechstein et al. 7 ± mortality: 1 in symptomatic patients in whom CM seems to reach
1997 ± improved: 2 the surface of the brain stem.
± unchanged: 3
± aggravated: 1
Bouillot et al. 17 14 ± mortality: 2 when CMs are presenting with hemorrhage(s) and
1996 (16) ± improved: 10 considered anatomically accessible.
1 unchanged: 2
± aggravated: 3
Lewis et al. 24 22 ± mortality: none in symptomatic lesions, the risk of permanent
1995 ± improved or morbidity being less than the natural history.
± unchanged: 21
± aggravated: 3
Amin-Hanjani 14 11 ± mortality: none when the patient is su½ciently neurologically
et al. 1998 (12) ± improved: 5 compromised to reap the potential bene®ts of
± unchanged: 7 surgical treatment.
± aggravated: 2
Sakai et al. 9 8 ± mortality: none for CMs which are located near the brain stem surface.
1997 (8) ± improved or
± unchanged: 8
Porter et al. 86 94% ± mortality: 3 in patients with symptomatic hemorrhage who harbor
1999 (83) 67 out of 71 MRI ± improved: 3 lesions that approach the pial surface.
checked ± unchanged: 7
± aggravated: 8
Cantore et al. 11 10 ± mortality: 1 In addition to the parenchymal windows produced by
1999 ± improved: 4 the CMs, safe entry zones are ˆ ¯oor of the IVth
unchanged: 6 ventricle and the anterior surface of the medulla and
± aggravated: 1 pons
Ziyal et al. 7 7 ± mortality: none in CMs which are symptomatic, have bled or are
1999 (7) ± improved: 7 growing, and are approachable through one of the
pial surfaces of the brain stem. Radiosurgery is not
recommended in such cases.

References 2. Amin-Hanjani S, Ogilvy CS, Candia GJ, Lyons S, Chapman PH

(1998) Stereotactic radiosurgery for cavernous malformations:
1. Aiba T, Tanaka R, Koike T, Kameyama S, Takeda N, Komata Kjellberg's experience with proton beam therapy in 98 cases at
T (1995) Natural history of intracranial cavernous malforma- the Harvard Cyclotron. Neurosurgery 42: 1229±1238
tions. J Neurosurg 83: 56±59 3. Amin-Hanjani S, Ogilvy CS, Ojemann RG, Crowell R (1998)
852
Risk of surgical management of Cavernous Malformations of
the nervous system. Neurosurgery 42: 1220±1228
4. Bertalan¨y H, Gilbach JM, Eggert HR, Seeger W (1991) Mi-
crosurgery of deep-seated cavernous angiomas: report of 26
cases. Acta Neurochir (Wien) 108: 91±99
5. Bouillot P, Dufour H, Roche Ph, Lena G, Graciani N, Grisoli F
(1996) Malformations vasculaires angiographiquement occultes
du tronc ceÂreÂbral (25 cas). Neurochirurgie 42: 189±201
6. Cantore G, Missori P, Santoro A, Bricolo A (1999) Cavernous
angiomas of the brain stem. Surg Neurol 52: 84±94
7. Curling OD, Kelly DL, Elster AD, Craven TE (1991) An anal-
ysis of the natural history of the cavernous angiomas. J Neuro-
surg 75: 702±708
8. Fahlbusch R, Strauss C (1991) Zur chirurgischen Bedeutung
von cavernoÈsen HaÈmangiomen des Hirnstammes. Zent Bl Neu-
rochir 52: 25±32
9. Fritschi JA, Reulen HJ, Spetzler RF, Zabramski JM (1994)
Cavernous malformations of the brain stem. A review of 139
cases. Acta Neurochir (Wien) 130: 35±46
10. Houteville JP (1995) The surgery of cavernous both supra-
tentorial and infratentorial. Adv Tech Stand Neurosurg 22:
185±259
11. Karlsson B, Kilhstrom L, Lindquist C, Ericson K, Steiner L
(1998) Radiosurgery for cavernous malformations. J Neurosurg
88: 293±297
12. Karnofsky DA, Abelmann WH, Craver LF (1948) The use of
the nitrogen mustards in the palliative treatment of carcinoma.
Cancer I: 534±656
13. Kondziolka D, Lunsford LD, Flickinger JC, Kestle JRW (1995)
Reduction of hemorrhage risk after stereotactic radiosurgery for
cavernous malformations. J Neurosurg 83: 825±831
14. Kondziolka D, Lunsford LD, Kestle JRW (1995) The natural
history of cerebral cavernous malformations. J Neurosurg 83:
820±824
15. Kyoshima K, Kobayashi S, Gibo H et al (1993) A study of safe
entry zones via the ¯oor of the fourth ventricle for brain stem
lesions. Report of three cases. J Neurosurg 78: 987±993
16. Lewis AI, Van Loveren HR, Tew JM Jr (1995) Management of
brain stem vascular malformations: advances in surgical tech-
nique and long-term results. Neurosurg Quarterly 5: 217±228
17. McCormick WF, Nofzinger JD (1966) Cryptic vascular malfor-
mations of the central nervous system. J Neurosurg 24: 865±875
18. Mizoi K, Yoshimoto T, Suzuki J (1992) Clinical analysis of ten
cases with surgically treated brain stem cavernous angiomas.
Tohoku J Exp Med 166: 259±267
19. Ondro SL, Doty JR, Mahla ME, George ED (1988) Surgical
excision of a cavernous hemangioma of the rostral brain stem:
case report. Neurosurgery 23: 490±493
20. Pechstein U, Zestner J, Van Roost D, Schramm J (1997) Surgi-
cal management of brain stem cavernomas. Neurosurgical Re-
view 20: 87±93
21. Porter RW, Detwiler PW, Spetzler RF, Lawton MT, Baskin JJ,
Dersken PT, Zabramski JM (1999) Cavernous malformations of
the brain stem experience with patients. J Neurosurg 90: 50±58
22. Robinson JR, Awad IA, Little JR (1991) Natural history of the
cavernous angioma. J Neurosurg 75: 709±714
23. Sakai N, Andoh T, Nishimura Y, Takenada K, Yamakawa H,
Nakatani K, Sakai H, Yoshimura S (1997) Surgical treatment of
cavernous angioma involving the brain stem. Proceedings od the
11th int. Congress of Neurological Surgery, Amsterdam, July 6±
17 vol. 1, Monduzzi publ, Bologna, pp 1159±1163
24. Sakai N, Yamada H, Tanigawara T, Asano Y, Andoh T,
Tanabe Y, Takada M (1991) Surgical treatment of cavernous
angioma involving the brainstem and review of the literature.
Acta Neurochir (Wien) 113: 138±143
M. Sindou et al.
25. Sathi et al (1996) unpublished data, cited by Porter et al
26. Seo Y, Fukuoka S, Takanashi M et al (1995) Gamma-knife
surgery for angiographically occult vascular malformations.
Stereotactic Funct Neurosurg [Suppl] 1: 98±109
27. Simard JM, Garcia-Bengochea F, Jr Ballinger WE et al (1986)
Cavernous angioma: a review of 126 collected and 12 new clini-
cal cases. Neurosurgery 18: 162±172
28. Stea RA, Schicker L, King GA, Win®eld JA (1994) Stereotactic
linear radiosurgery for cavernous angiomas. Stereotactic Funct
Neurosurg 63: 255±265
29. Steiner L, Lindquist C, Steiner M (1992) Radiosurgery. Adv
Tech Stand Neurosurg 19: 19±102
30. Symon L, Jackowski A, Bills D (1991) Surgical treatment of
ponto-medullary cavernomas. Brit J Neurosurg 5: 339±347
31. Voigt K, Yasargil MG (1976) Cerebral cavernous haemangio-
mas or cavernomas. Neurochirurgica 19: 59±68
32. Weil SM, Tew JM (1990) Surgical management of brain stem
vascular malformations. Acta Neurochir (Wien) 105: 14±23
33. Yoshimoto T, Suzuki J (1986) Radical surgery on Cavernous
Angioma of the brain stem. Surg Neurol 26: 72±78
34. Zimmerman RS, Spetzler RF, Lee KS, Zabramski JM, Har-
graves RW (1991) Cavernous malformations of the brain stem.
J Neurosurg 75: 32±39
35. Ziyal M, Sekhar LN, Salas E, Sen C (1999) Surgical manage-
ment of cavernous malformations of the brain stem. Brit J
Neurosurg 13: 366±375
Comments
In this excellent article, Prof. Sindou and colleagues present a de-
tailed retrospective analysis of a 12-patient series operated on for
brain stem cavernoma, and a concise yet well-represented review of
the literature. Their reported ®ndings are similar to previous litera-
ture published on this lesion, functional results are reasonably good
and conclusions are appropriate. I would like to address the authors'
suggested selection criteria for surgery. I agree in full that ``symp-
tomatic'' cavernous malformations manifesting with bleeding or a
progressive mass-e¨ect syndrome should undergo surgical treat-
ment; however, based on the present skill level of brain stem surgery
as well as my experience with this particular lesion, I am inclined to
believe that ``asymptomatic'' patients should also be considered for
surgery. Young adult patients leading an active life who undergo
MRI for some reason and are discovered to be harboring a cavern-
ous malformation in a critical area such as the brain stem do not
merit a `watch and wait' passive policy as such a ®nding will condi-
tion their life negatively. Now that we have the understanding and
technical capability to remove a brain stem cavernoma with very
low morbidity, I strongly suggest that surgery be o¨ered or at least
considered also in a selected subgroup of so-called ``asymptomatic''
patients.
A. Bricolo
The authors report on a retrospective study of their personal ex-
perience with surgical resection of brain stem cavernous malforma-
tions (CM) in 12 consecutive patients. The surgical results are excel-
lent. In 11 patients the CM was totally resected. The majority of the
patients had improved at one year after surgery. The authors con-
clude that brain stem Cms have a high bleeding and re-bleeding risk.
Considering the grave natural history, the authors argue that these
lesions should be surgically resected, particularly once they have bled.
The authors present no new aspects concerning the natural history
and management of these lesions. Their results and conclusions are
in accordance with the current literature and cited larger series.
Nevertheless, publication of this plead for radical resection is rec-
Functional Results in Brain Stem Cavernomas 853

ommended in view of the still existing reluctance to operate in this Correspondence: Professor Marc Sindou M.D., D.Sc., Neuro-
area and ignorance of the often devastating sequelae when these chirurgie A, Department of Neurosurgery, Hopital Neurologique P.
lesions are left untreated. Wertheimer and University of Lyon, 59 Bd Pinel, 69003 Lyon,
H. Reulen France.