Surgical management of brain-stem cavernous malformations: report of 137 cases

Vascular

Surgical Management of Brain-Stem Cavernous Malformations: Report of 137 Cases Chung-cheng Wang, M.D., Ali Liu, M.D., Jun-ting Zhang, M.D., Bo Sun, M.D., and Yuan-li Zhao, M.D. Beijing Neurosurgical Institute, Beijing, People’s Republic of China

Wang C-c, Liu A, Zhang J-t, Sun B, Zhao Y-l. Surgical management of brain-stem cavernous malformations: report of 137 cases. Surg Neurol 2003;59:444 –54. BACKGROUND

With the improvement in neuroimaging and microsurgical techniques, brain stem cavernous malformations are no longer considered inoperable. Surgical indications for brainstem cavernoma are evolving, with better understanding of its natural history and decreasing surgical complications. METHODS

During 1986 through 1998, a series of 137 patients (4 patients each with two brain stem lesions, total number of lesions, 141) with brain stem cavernous malformations were treated microsurgically at Beijing Neurosurgery Institute. The age distribution, lesion location, and clinical presentations were analyzed. The bleeding rate, surgical indications and microsurgical techniques were also discussed.

acute hemorrhage on MRI either inside or outside cavernous malformations with mass effect; (3) cavernoma/ hematoma reaching brainstem surface (⬍2 mm brain tissue between cavernoma /hematoma and pial surface). Grave clinical presentations like coma, respiratory, or cardiac instability are not surgical contraindications. Emergent surgical evacuation may lead to satisfactory outcome. Repeated hemorrhages will worsen the preexisting neurologic deficits and possibly make the surgical dissections more difficult. Patients with minimum, stable neurologic deficits and lesion/hematoma that has not reached the brain stem surface should be followed conservatively. © 2003 Elsevier Inc. All rights reserved. KEY WORDS

Brain stem, cavernous malformation, microsurgical operation, magnetic resonance imaging.

RESULTS

In our series, 92 of 137 cases (67.2%) suffered more than one hemorrhage. Female patients had a higher risk of recurrent hemorrhage than that of male patients. Unlike cavernomas malformations from other locations, repeated hemorrhages from brain stem malformations are much more common and usually lead to new neurologic deficits. Among all 137 surgically treated patients, there was no operative mortality. Ninety-nine patients (72.3%) either improved or remained clinically stable postoperatively. The size of the cavernoma/hematoma does not necessarily correlate with the surgical result. While the acute hematoma can facilitate the surgical dissection, longer clinical history with multiple hemorrhages often makes total surgical resection difficult, partially because of the firmer capsule that may not shrink or collapse after hematoma is released. Pathologically those capsules were associated with more hyaline degeneration, fibrous proliferation and even calcifications. During the follow-up period between 0.5 to 11 years in 129 cases, 115 patients (89.2%) have been working, studying, or doing house work. Three patients (2.3%) suffered recurrent hemorrhages.

rain stem vascular malformation accounted for 31% of all brain stem space-occupying lesions treated in our institute. Most brain stem vascular malformations are cavernous malformations. In the past, this disease was not well recognized, and often diagnosed as “spontaneous brain stem hemorrhage.” Advances in microsurgery and neuroradiology have expanded operative therapy to include these lesions that were once considered inoperable. Between 1986 and 1998, 137 patients with brain stem cavernous malformations were surgically treated at Beijing Neurosurgery Institute. The natural history, clinical presentations, surgical indications, and clinical outcome were reviewed retrospectively; the effects of radiosurgery and repeated hemorrhages after surgery are also discussed.

B

CONCLUSION

Surgical indications of brain stem cavernoma include (1) progressive neurologic deficits; (2) overt acute or sub-

Materials and Methods

Address reprint requests to: Dr. Chung-cheng Wang, Director, Beijing Neurosurgical Institute, Tiantan Xili, Beijing, People’s Republic of China. Received August 17, 2001; accepted February 19, 2003.

GENERAL MATERIALS There were 80 males and 57 females (sex ratio 1.4: 1), ranging in age from 3.5 to 70 years old (mean

0090-3019/03/$–see front matter doi:10.1016/S0090-3019(03)00187-3

© 2003 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010 –1710

Cavernous Malformation

Surg Neurol 445 2003;59:444 –54

The age distribution of the patients with cavernoma of brain stem.

1

33.5 years). The details of age distribution are noted in Figure 1. The mean duration of preoperative clinical history was 21.8 months (ranged from 0.03-288 months). All patients were treated at Beijing Neurosurgical Institute and its affiliated Beijing Tiantan Hospital between 1986 and 1998. Among all 137 patients, four patients had two isolated lesions in the brain stem, making the total lesion number 141. The anatomic locations of these lesions are listed in Table 1. Besides the brain stem lesions, multiple cavernomas were also found in cerebrum and/or cerebellum in 9 patients. One patient had as many as 11 intracranial lesions. CLINICAL PRESENTATIONS Most of the patients presented with sudden-onset symptoms. Unlike brain stem hemorrhage from hypertension, AVM or tumors, brain stem hemorrhage from cavernomas was rarely fatal even with large lesions and hematomas. With conservative treatment, patients’ neurologic deficits usually improve, but brain stem cavernomas are very likely to re-

1

The Anatomical Locations of the Lesions

LOCATION

CASES

Midbrain Brachium conjunctivum Midbrain-pons Pons Brachium pontis Pons-medulla Medulla oblongata Cervicomedullary junction

20 1 9 83* 7 2 18* 1

*Among patients with multiple brain stem cavernomas, 1 patient had a medulla and a pontine lesion; 2 patients each had 2 pontine lesions; 1 patient had a pontine and a midbrain lesion.

bleed, resulting in new or more pronounced neurologic deficits. With this “relapsing and remitting” history, some patients were initially misdiagnosed with multiple sclerosis before a brain MRI was obtained. With repeated hemorrhages, patients’ neurologic deficits became more pronounced and permanent. The clinical presentations differed according to the cavernoma locations (Table 2). Hemiparesis and ataxia are the most common presentations, occurring in all three groups. All medulla cases suffered dysphagia. Increased intracranial pressure (ICP) presented more frequently in

2

The Clinical Presentations According to the Cavernoma Locations

SYMPTOMS AND SIGNS Hemiparesis Ataxia Double vision (diplopia) V⬃VIII cranial nerve palsy Dysphasia Vertigo Hemianesthesia Increased ICP Red nucleus tremor Involuntary laugh Paroxysmal coma Unilateral gaze paralysis Parinaud’s syndrome High fever Intractable hiccup Respiratory difficulties GI hemorrhage Bradycardia

MIDBRAIN (29)

PONS (90)

14 (48.3%) 51 (56.7%) 11 (37.9%) 30 (33.3%) 20 (69%) 29 (32.2%) 1

68 (75.6%)

MEDULLA (18) 8 (44.4%) 8 (44.4%) 1

21 (23.3%) 18 (100%) 40 (44.4%) 3 (16.7%) 1 44 (48.9%) 8 (44.4%) 11 (37.9%) 1 1 3 1 1 15 (16.7%) 1 4 1

5 (27.8%) 3 (16.7%) 1 1

446 Surg Neurol 2003;59:444 –54

Wang et al

Recurrent hemorrhages from a pontine cavernoma over 4 years. (a) A small pontine lesion (arrow) could be identified at the first onset. Patient only had lateral conjugate gaze defect. (b) The lesion is enlarged 2 years later. Patient developed left facial nerve palsy, decreased hearing on left side. (c) The cavernoma continued to grow and bleed over the next 2 years. Besides her previous symptoms, patient also suffered dysphagia, dysphonia, right hemianesthesia, and bilateral lower extremity pyramidal signs.

2

midbrain cases. Red nucleus tremor, involuntary laughing, paroxysmal coma, and vertical gaze paralysis were only seen in patients with midbrain lesions. Cranial nerves V⬃VIII palsy and high fever were unique for pontine lesions. Cardiovascular and respiratory instability, dyspnea, intractable hiccup, and gastrointestinal bleed were often presented in patients with medullary cavernomas. RADIOLOGIC FINDINGS Magnetic resonance imaging (MRI) was routinely obtained in all patients. Half of our patients had either angiogram or magnetic resonance angiography (MRA) examinations. MRI is the most valuable study for the diagnosis of cavernous malformatons. The radiologic characteristics had been well described. The typical findings were “popcorn” or “mulberry” lesion resulting from multiple previous hemorrhages. High-density lesions on T1 and T2weighted images were often surrounded by circular or irregular low-density areas. T2-weighted image is most useful to delineate the surrounding lowdensity hemosiderin deposit around the high density of the methemoglobin, which is characteristic for cavernomas. Gradient echo scanning could increase the sensitivity to diagnose small and multiple cavernomas. On MRA images, methemoglobin signal could be shown clearly. Angiogram generally fails to reveal cavernous malformations because they lack arterial feeders, except for those associated with other vascular malformations. MRIs are especially helpful to demonstrate the relationship between cavernomas and neighboring vital neuro-

structures that are essential to design the surgical approaches. Repeated hemorrhages could significantly change the MRI signals of both cavernoma and hematoma. The recurrent hemorrhage could happen at the exact same location or expand to surrounding area (Figure 2). Four patients in our series had two independent lesions in the brain stem (Figure 3). Another 9 patients had more than two lesions located in the brain stem and cerebrum and/or cerebellum.

3

The pattern of multiple cavernomas of brain stem.

Cavernous Malformation

3

The Effect Frequency

of

Patients’

Surg Neurol 447 2003;59:444 –54

Gender

on

Bleeding

SEX

HEMORRHAGE EPISODES (X ⴞ SD)

Male (N ⫽ 80) Female (N ⫽ 57)

1.95 ⫾ 1.34 2.33 ⫾ 1.34

T test (P ⬍ 0.05).

HISTORY OF HEMORRHAGE Intracerebral hemorrhage was the most common clinical presentation. Because of the low pressure inside the cavernoma and special brain stem structure with compact horizontal and longitudinal fibers, brain stem cavernomas rarely hemorrhage into either subarachnoid space or 4th ventricle. Repeated hemorrhages however, like a “rolling snowball” effect, often resulted in new or more permanent neurologic deficits. Cavernoma hemorrhage was defined as acute or subacute blood either outside or inside the border of cavernoma, in conjunction with sudden onset or aggravation of clinical symptoms. Occasionally patients may present with gradual and progressive symptoms, this also was considered as two hemorrhages. The presence of hemosiderin rings around the cavernous malformations, which was essentially seen in almost all patients, was not considered as a hemorrhage episode. Based on this criteria, 45 patients in our series experienced one hemorrhage, 58 had two, 23 had three, and 11 had more than 3 hemorrhages. Female patients tend to rebleed more frequently than male patients (Table 3). When patients were grouped according to the duration of their clinical symptoms (⬍1 years, 1⬃ 10 yrs, ⬎10 yrs), it was found that rebleed rate was directly related to the length of patients’ clinical history (Table 4). Among patients who harbored brain stem cavernomas that bled more than once, 46.4% rebled within 6 months, another 46.4% rebled between 7 months and 4 years, only 7.2% rehemorrhaged more than 4 years after the initial bleed. In our

4

The Relationship Between Rebleeding Rate and Duration of Clinical History

CLINICAL HISTORY

HEMORRHAGE EPISODES (AVERAGE)

⬍1 yr (N ⫽ 90) 1⬃10 yrs (N ⫽ 41) ⬎10 yrs (N ⫽ 6)

1.6 2.7 4.8

␹2 test (P ⬍ 0.05).

series, the most common rebleeding occurred within 1 month, accounting for 21.8% of all patients who at least bled twice. Overall 137 patients had 288 hemorrhages in 4553.5 patient-years of life. Assuming that the lesions were present since birth, the risk of hemorrhage is 6% per person per year. Once the cavernoma bled, the rebleeding rate increased to 60% per person per year (151 rehemorrhages among 92 patients during a total of 251 lesion observation years after the initial bleed). SURGERY Brain MRI was routinely repeated within one week before surgery. Operative approaches were selected based upon the location of cavernoma and where cavernoma/hematoma reached the brain stem surface. Cavernomas at the tectum of midbrain were resected through Poppen’s suboccipital transtentorial approach; those reaching the lateral surface of the midbrain were managed by subtemporal-occipital approach. Trans-sylvian approach was used for anterior midbrain lesions, posterior fossa lateral approach for ventrolateral lesions of the pons, and posterior fossa midline approach for dorsal pontine and medullary lesions. When the cavernomas were approached through the floor of fourth ventricle, to minimize the dissection and retraction of brain stem tissue, we accessed the lesions through either upper-facial triangle or lower-facial triangle [1,8]. During the operation, the hematoma capsule was opened first, then the hematoma was evacuated. Usually enough room could be obtained to dissect the hematoma together with the malformed vessels. A yellowish hemosiderin-stained layer around the hematoma is a very helpful guide for surgical dissection. All the cavernous malformation should be completely extirpated; partial removal will not prevent recurrent hemorrhage. If a venous angioma is noticed during surgery, it should be protected and preserved. No attempts were made to remove any hemosiderinstained tissue.

Results All 137 patients received a microsurgical operation for evacuation of the hematoma and extirpation of the cavernomas. Judged by the senior surgeon using the microscope, 131 out of 137 patients had total cavernoma resection; 2 patients had a partial resection. Both patients had more than 7 years clinical history; their cavernomas were hard, tenacious and quite adherent to brain stem tissue. Patients’ vital signs became unstable with slight retraction on surrounding brain stem tissue.

448 Surg Neurol 2003;59:444 –54

Wang et al

A dorsal pontine cavernoma. (a, b) Preoperative MRI with evidence of hemorrhage and mass effect. Patient presented with left hemiparesis, ptosis, hypoesthesia and diplopia. (c, d) Postoperative MRI shows the cavernoma was completely removed. All above-mentioned symptoms disappeared.

4

Therefore, only a partial resection could be performed. In 4 patients, because of the large size of the cavernoma and the scar tissue obscuring the dissection plane, the extent of cavernoma resection was uncertain. In cases of multiple brain stem cavernomas, only the hemorrhaged, symptomatic lesions were removed. On 3 months postoperative MRIs that were obtained in 135 patients; total cavernoma resections were achieved in 130 patients, whereas residuals of cavernoma were seen in 5 patients. There was no operative mortality in this series. Postoperatively, clinical symptoms either improved or remained stable in 99 patients (72.3%, see Figures 4 – 6). Thirty-eight patients (27.7%) deteriorated after operation or developed new neurologic deficits. This included 4 patients with medulla

lesions who needed tracheotomy and prolonged ventilation support. Other complications included intracranial infection (1 patient) and recurrent hemorrhage (1 patient); unconsciousness (1 patient with midbrain lesion); stress ulceration (2 patients); Parinaud’s syndrome (1 patient). With proper treatment, these 10 patients with postoperative complications improved before being discharged from the hospital. Follow-up data were obtained by telephone conversation or office visit. With a follow-up between 0.5⬃11 years (mean 52 months) in 129 cases, there were 3 patients who experienced rehemorrhage and underwent a secondary operation. One patient had three operations during hospitalization and finally died of pneumonia 6 months later; 115 patients (89.2%) either re-

Cavernous Malformation

Surg Neurol 449 2003;59:444 –54

A 21-year-old male with a large pontine cavernoma and 18 years clinical history. (a, b) Preoperative MRIs demonstrated a large pontine cavernoma with a “popcorn” like appearance. The lesion occupied ⬎80% of pons on axial MRI. This patient presented with headache, right hemianesthesia, right upper extremity weakness and atrophy, left facial nerve palsy and dysphagia. (c, d) Postop. MRIs: the cavernoma was completely removed, but the cavity did not collapse. Postoperatively, patient’s right arm strength improved.

5

turned to work/school or doing housework, and 10 patients (7.8%) lived dependently.

Discussion NATURAL HISTORY The natural history of intracranial cavernous malformations remains unknown. Cavernomas of the brain stem account for between 8.5 and 22% of all intracranial cavernomas [6,7,9,19]. With the widespread use of MRI, cavernomas are being diagnosed with increased frequency. Most cavernomas occur sporadically and are solitary. Multiple cavernomas can be found in as many as 24% of patients [11]. Overall about 14% of the patients had a familial history of cavernous malformations [11]. The in-

heritance pattern is autosomal dominant with incomplete penetration. Those familial cases tend to have multiple cavernomas and de novo appearance of cavernomas had also been reported [12]. The majority of small asymptomatic lesions without overt hemorrhage are still difficult to diagnose [5,13,19,22]. In our series, the bleeding rate in female patients is significantly higher than that of male patients, indicating a hormonal effect on the cavernoma [1,12,15,16]. The length of preoperative clinical history is directly proportional to the frequency of bleeding. Unlike cavernomas from other locations, hemorrhages from brain stem cavernomas were almost always associated with new neurologic deficits or worsening of preexisting symp-

450 Surg Neurol 2003;59:444 –54

Wang et al

A 7-year-old boy harbors a cavernoma in medulla oblongata. (a, b) Preoperative MRIs show a medullary cavernoma with evidence of hemorrhage and mass effect. Patient had dysphagia, left leg weakness with Babinski sign. (c, d) Postoperative MRIs demonstrated the cavernoma was completely removed. All patient’s symptoms disappeared in 6 months.

6

toms [11,17,20]. Because of the compact brain stem structure, clinically occult hemorrhages were rare [19]. Because of the low pressure inside the lesion, cavernoma hemorrhages rarely extend into either ventricle or subarachnoid space. It has been documented that cavernomas in various intracranial locations have different rate of bleeding. Porter and associates found the hemorrhage rate of infratentorial cavernomas was 30 times greater than that of supratentorial lesions [11]. In our series, the annual bleeding rate of brain stem cavernoma was 6%, higher than that of cavernomas in other intracranial locations. Because of

the strict definition of hemorrhage used in our series, we believe the 6% annual bleeding rate is still a conservative estimate. Once the cavernomas bleed, their annual rebleeding rate increases significantly regardless of the cavernoma location [1,4,6,7,18]. This phenomenon is also documented in our series. The overall rebleeding rate jumped to 60% per year. This figure is much higher than that of cavernomas from other anatomic locations. The high bleeding and rebleeding rate of brain stem cavernomas may result from its biologic uniqueness. It also could reflect the fact that because of the compact, eloquent brain stem structure, caver-

Cavernous Malformation

Surg Neurol 451 2003;59:444 –54

noma hemorrhages are unlikely to be clinically silent, as occurred in many superficial cavernous malformations.

distinguish cavernoma bleed from hemorrhage of other sources, like AVM, hypertension or coagulopathy.

PATHOLOGY The pathologic findings of cavernous malformations include many sinusoidal dilated vessels of various sizes that are compact and discrete from the surrounding brain. Collagenous hyperplasia and fibrosis are present. Almost all-cavernous malformations show a large amount of hemosiderin around them, which provides evidence of multiple previous microscopic hemorrhages. The dilated vascular walls of a cavernous malformation could not be differentiated from that of telangiectasia, because both of these vessels contain no elastic fibers or smooth muscle. This leads to the hypothesis that cavernous malformation and telangiectasia belong to the same kind of vascular malformation because of their common genesis from capillary vessels [7,11]. The dilated capillary vessels have a thin wall and low blood pressure. Therefore, the vessels tend to rupture and bleed in small amount repeatedly. Many small asynchronous hemorrhages can merge to a sizable hematoma. With a long clinical history, the hematomas would become fibrotic; hyaline degeneration or even calcifications could also develop. All these changes could make the surgical dissection difficult.

SURGICAL INDICATIONS Indications of microsurgical resection of brain stem cavernoma include (1) progressive neurologic deficits; (2) grave clinical presentations like coma, cardiac and respiratory instability; (3) overt acute or subacute hemorrhage on MRI, either inside or outside cavernous malformations, with mass effect; (4) either cavernoma or hematoma becomes ⬍2 mm from the pial surface. Although conservative treatment can result in some clinical improvement in patients with first time bleeding who are clinically stable with marked neurologic deficits, we believe that the overall risks of rebleeding outweighs the risks of surgery, considering brain stem cavernomas have high rebleeding rate and each rehemorrhage will almost certainly lead to new or more permanent neurologic deficits. In the past, some surgeons advocated deferring surgery until the lesion bled at least twice. Small AVMs can cause brain stem hemorrhage. It is possible those AVMs could be totally obliterated by the hematoma and never bleed again; such circumstances are extremely rare. The pressure inside the cavernomas is usually low; therefore extirpation of cavernomas by hematoma is only wishful thinking. We recommend surgery once cavernomas bleed and meet the above mentioned criteria. Waiting for the second hemorrhage before considering surgery cannot be justified. However, we do recommend conservative treatment if the diagnosis of cavernoma hemorrhage is uncertain. Nonoperative treatment is indicated when patients have minimal neurologic deficits and small deep lesions deny safe surgical accessibility.

DIFFERENTIAL DIAGNOSIS Brain stem cavernomas should be highly suspected if a young patient without hypertension history presenting with vertigo, hypoesthesia, or other brain stem signs acutely or subacutely, and with MRI findings of brainstem hemorrhage (high-density signal both in T1 and T2-WI). A typical “popcorn” like appearance on T2 weighted image is characteristic, but not always present. Brain stem cavernomas should be distinguished from bleeding of brain stem glioma. The tumor hemorrhage is usually smaller than the tumor itself, and repeated hemorrhage is extremely rare. Brain stem tumors are often associated with long lasting edema, while cavernomas only have edema with acute hemorrhage that usually disappears within a few days. AVM caused hematomas tend to be more homogeneous and lack the typical “popcorn” feature. If the first MRI cannot differentiate cavernoma from AVM bleed, a repeat MRI in a few weeks should help confirm the diagnosis. Occasionally, a patient can have a large brain stem cavernoma with hematoma, but remain neurologically intact or have only minimal neurologic deficits. This feature can help

TIME OF SURGERY Unless patients were in coma or had cardiorespiratory instability which required emergent surgical evacuation, we preferred to treat patients with steroid for a week or two, letting the edema subside, then proceeded with microsurgery. This approach allows the surgeon to take advantage of the dissection from hematoma, and avoid dealing with aged and well-organized hematoma. Usually after evacuating the fresh blood clot, enough room can be obtained for safe excision of cavernoma without any retraction on the brain stem tissue. Once the hematoma is absorbed, gliosis, hyaline degeneration, and calcifications may develop, making the dissection plane obscure, and therefore complicating the surgical resection (Figure 7).

452 Surg Neurol 2003;59:444 –54

Wang et al

A large pontine cavernoma in a 21-year-old male with 18 years clinical history (same patient as in Figure 5). Preoperative MRI shows a large pontine cavernoma with heterogeneous signal and mass effect. Pathologic examination revealed evidence of previous hemorrhage from different times. (a) Samples from upper part of cavernoma demonstrates hyaline degeneration, collagenous and fibrotic scar. (b) Calcification and subacute blood were found in the middle of cavernoma. (c) Sample from the bottom of the cavernoma shows capillary like channels filled with fresh blood.

7

CLINICAL OUTCOME In our series, 17 patients had large lesions: 13 in the pons, (the largest diameter ⬎3 cm); 2 in midbrain-pons (⬎4 cm); and 2 in the medulla oblongata (⬎2.2 cm). Most of these patients had grave clinical symptoms (coma in 1 case, GI bleeding in 1 case, bradycardia in 1 case, respiratory distress in 3 cases, extensive neurologic deficits in 7 cases, and high fever in 4 cases). Postoperatively, all patients had some neurologic improvements. Our results show that poor clini-

cal status and larger lesions do not necessarily correlate with poor clinical outcome. Fritschi et al analyzed 101 cases of brain stem cavernous malformations. The mortality rate was 0% in operative group (93 cases) and 20% in nonoperative group (8 cases) [6]. Porter et al also reported similar results, higher mortality with conservative treatment [10,11]. Our result supports the claim that surgical resection is the most “conservative” treatment once the brain stem cavernomas bleed and reach brain stem surface.

Cavernous Malformation

STEREOTACTIC RADIOSURGERY The role of radiosurgery in the treatment of cavernomas remains unproven [2,4]. In most patients with brain stem cavernoma, the hematoma on MR image is much bigger than the cavernoma itself, so the target of radiation treatment will be unprecise. Moreover, marginal radiation dose must be limited (⬍15 Gy) to avoid damage to brain stem, so complete obliteration of the malformed vessels is difficult to obtain. In our series, 2 patients were treated with radiosurgery before admission, 1 had radiosurgery 4.5 years earlier and the other 3 years. Both patients underwent surgical resection for recurrent hemorrhage. Such limited number of patients did not allow us to draw any definitive conclusion. As cavernomas are angio-occult lesions and no other clinical test or imaging can confirm whether the cavernomas are obliterated or not, long term follow up is the only criteria to evaluate the effectiveness of radiosurgery. Up to now such long term follow up data are not available. However, we did notice during the operation that the cavernomas were more adherent and dissecting plane was less clear if patients had undergone radiosurgery, therefore making the total surgical resection more difficult.

Conclusion Brain stem cavernomas have a high bleeding rate (at least 6% per year). Once a cavernoma bleeds, its annual rebleeding rate will be even higher (60% per year). Most brain stem cavernoma hemorrhages will result in marked neurologic deficits. Therefore, aggressive microsurgical treatment is warranted. Surgical indications include (1) progressive neurologic deficits; (2) grave neurologic deficits like coma, cardiopulmonary instability; (3) overt acute or subacute hemorrhage on MRI either inside or outside cavernous malformations with mass effect; (4) cavernoma or hematoma ⬍2 mm from brain stem surface. Patients with minimum neurologic deficits and small deep lesions that did not reach the brain stem surface should be observed. Once the surgical decision is made, surgery should be performed between 1 to 2 weeks after the brain stem edema subsides and before the hematoma becomes organized. Difficult surgical dissections can be expected in those patients with very long clinical history and multiple previous hemorrhages.

Surg Neurol 453 2003;59:444 –54

3. 4. 5.

6. 7. 8.

9.

10.

11. 12.

13. 14.

15. 16.

17.

18.

19.

REFERENCES 1. Aiba T, Tanaka R, Koike T, Kameyama S, Takeda N, Komata T. Natural history of intracranial cavernous malformations. J Neurosurg 1995;83:56 –9. 2. Bicknell JM, Carlow TJ, Korfeld M, Stovring J, Turner

20.

P. Familial cavernous angiomas. Arch Neurol 1978;35: 746 –9. Contore G, Missori P, Santoro A. Cavernous angiomas of the brain stem. Surg Neurol 1999;52:84 –94. Diomond C, Torvik A, Amundsen P. Angiographic diagnosis of telangiectases with cavernous angiomas of the posterior fossa. Acta Radio 1976;17:281–8. Fahlbusch R, Strauss C, Huk W, Rockelein G, Kompf D, Ruprecht KW. Surgical removal of pontomesencephalic cavernous hemangiomas. Neurosurgery 1990; 26:449 –57. Fritschi JA, Reulen HJ, Spetzler RF, Zabramski JM. Cavernous malformations of the brainstem. Acta Neurochir 1994;130:35–46. Kondziolka D, Lunsford LD, Kestle JRW. The natural history of cerebral cavernous malformations. J Neurosurg 1995;83:820 –4. Kyoshima K, Kobayashi S, Gibo H, Kuroyanagi T. A study of safe entry zones via the floor of the fourth ventricle for brain-stem lesions. J Neurosurg 1993;78: 987–93. Lobato RD, Perez C, Rivas JJ, Cordobes F. Clinical, radiological, and pathological spectrum of angiographically occult intracranial vascular malformations. J Neurosurg 1988;68:518 –31. Porter PJ, Willinsky RA, Harper, Wallace MC. Cerebral cavernous malformations: natural history and prognosis after clinical deterioration with or without hemorrhage. J Neurosurg 1997;87:190 –7. Porter RW, Detwiler PW, Spetzler RF, et al. Cavernous malformations of the brainstem: experience with 100 patients. J Neurosurg 1999;90:50 –8. Pozzati E, Acciarri N, Tognetti F, Marliani F, Giangaspero F. Growth, subsequent bleeding and de novo appearance of cerebral cavernous malformations. Neurosurg 1996;38:662–70. Pozzati E, Giuliani G, Nuzzo G, Poppi M. The growth of cerebral cavernous angiomas. Neurosurgery 1989;25: 92–7. Rigamonti D, Johnson PC, Spetzler RF, Hadley MN, Drayer BP. Cavernous malformations and capillary telangiectasia: a spectrum within a single pathological entity. Neurosurg 1991;28:60 –64. Robinson JR, Awad IA, Little JR. Natural history of the cavernous angioma. J Neurosurg 1991;75:709 –14. Robinson JR, Awad IA, Magdinec M, Paranandi L. Factors predisposing to clinical disability in patients with cavernous malformations of the brain. Neurosurg 1993;32:730 –6. Poberson GH, Kase CS, Wolpow ER. Telangiectases and cavernous angiomas of the brainstem “Cryptic” vascular malformations. Report of a case. Neuroradiology 1974;8:83–9. Sakai N, Yamada H, Tanigawara T, et al. Surgical treatment of cavernous angioma involving the brainstem and review of the literature. Acta Neurochir 1991;113: 138 –43. Simard JM, Garcia-Bengochea F, Ballinger WE, Mickle JP, Quiling RG. Cavernous angioma: a review of 126 collected and 12 new clinical cases. Neurosurgery 1986;18:162–72. Stahl SM, Johnson KP, Malamud N. The clinical and pathological spectrum of brainstem vascular malformations: long-term course simulates multiple sclerosis. Arch Neurol 1980;37:25–9.

454 Surg Neurol 2003;59:444 –54

21. Wowra B, Layer G, Schad R, et al. Three-dimensional Time-of-flight MR-angiography and the surgical indication in brainstem cavernomas. Acta Neurochir 1991;112:77–82. 22. Zimmerman RS, Spetzler RF, Stuart Lee KS, Zabranski JM, Hargraves RW. Cavernous malformations of the brain stem. J Neurosurg 1991;75:32–9. COMMENTARY

The authors present a large series of brain stem cavernomas with impressive surgical results. (No mortality was noted in 137 cases of surgically removed brain stem cavernomas.) Data about frequency of hemorrhage from ruptured and unruptured cavernous malformations is practically important. Having read the paper, I got the main impression that brain stem cavernoma removal is quite an easy task. This feeling is arisen from the lack of description of some important details concerning cavernoma removal. My practice of surgical removal of brain stem cavernomas and hematomas due to cryptogenic vascular malformations shows that this surgery is not simple at all, and sometimes indications for surgery are not as clear and evident as presented by the authors. In some cases, especially in multiple bleeding cavernoma, it is not easy to find the borderline between the surrounding tissue and small compartments of the malformation that not always has a compact structure. Beside that, in the majority of cases a spreading vascular abnormality of the brain stem may be observed, mostly in the shape of capillary telengiectasia in the vicinity as well as in distance to cavernoma. Nevertheless, I would like to congratulate the Chinese colleagues on the remarkable results that reflect a high diagnostic level and perfect surgical technique. Prof. Alexander N. Konovalov, M.D., Ph.D. Director of the Burdenko Neurosurgery Institute Moscow, Russia The paper by Wang et al is an important contribution for a better knowledge of cavernomas arising in the brainstem. The authors analyze homogeneous personal series of 137 cases, which is the largest one found in the international neurosurgical literature. Many data described previously by other authors are confirmed in the Wang series: 1) the frequent pontic localization of the lesion (83%), 2) the high hemorrhagic risk of brainstem cavernomas (6% per person per year increasing to 60% after a previous bleeding), 3) the easier surgical dissection after a recent bleeding than after multiple recurrences, 4) the high rate of good postoperative re-

Wang et al

sults allowing 89% of operated patients to work again. Wang et al open an interesting discussion about surgical indications and make some criticisms about radiosurgical treatment of brainstem cavernomas, which should remain “surgical” lesions. Prof. Jean-Pierre Houtteville, M.D. Service de Neurochirurqie C.H.U. de Caen France

The authors report their operative experience with 137 patients. The authors’ indications for surgery appear to be sound and are consistent with our philosophy. Multiple patients with midbrain lesions presented with increased ICP. If this increase was unrelated to hydrocephalus, we are curious to know the underlying mechanism. The authors reported a higher rate of hemorrhage in females than in males. This finding heightens the controversy about whether females are at higher risk for hemorrhage from these lesions than males. Interestingly, there were no deaths related to surgery. Given the extensive skull base approaches used and the locations of the lesions (intrinsic to the brain stem), this result is surprising. We have treated 2 patients who suffered delayed onset cardiopulmonary arrest after being transferred out of the intensive care unit. Both patients were ambulatory with minimal neurological deficits but appeared to have experienced delayed inhibition of their respiratory drive arising from the medulla. We agree with most of the surgical approaches used. However, for lesions involving the tectum, we use the supracerebellar-infratentorial approach rather than the suboccipital transtentorial approach. We also agree that stereotactic radiosurgery should not be used to treat these lesions. In our opinion, its efficacy is unproven. A hematoma is often the target and will resolve if left untreated. Finally, it would be interesting to know what percentage of the authors’ patients were managed conservatively and their outcome. These data are important to understand the natural history of these lesions. The authors are to be congratulated for their large surgical series. We thank them for their contribution to the management of this disease. Randall W. Porter, M.D. Robert F. Spetzler, M.D. Barrow Neurological Institute Phoenix, Arizona