Spontaneous cerebellar hemorrhage: clinical remarks on 50 cases

Spontaneous Cerebellar Hemorrhage: Clinical Remarks on 50 Cases M. Salvati, M.D., L. Cervoni, M.D.,* A. Raco, M.D.,† and R. Delfini, M.D. Department of Neurological Sciences, Neurotraumatology, “La Sapienza” University of Rome; *Department of Neurosurgery, Mediterranean Neurological “Neuromed” Institute, IRCCS, Pozzilli (Isernia); †Department of Neurological Sciences, 1 Chair of Neurosurgery, “La Sapienza” University of Rome, Rome, Italy

Salvati M, Cervoni L, Raco A, Delfini R. Spontaneous cerebellar hemorrhage: clinical remarks on 50 cases. Surg Neurol 2001;55:156 – 61.

KEY WORDS

Spontaneous cerebellar hematoma, hydrocephalus, computed tomography, magnetic resonance imaging.

BACKGROUND

Only during the past 10 years have spontaneous cerebellar hemorrhages became a well-defined nosological entity. The surgical indication remeins debatable. Our primary objective in this study was to set the criteria for undertaking surgery by determining the critical diameter of the hematoma and considering the patients’ neurological status (Glasgow Coma Scale). METHODS

During the 8-year period 1990 through 1997 a series of 50 consecutive patients with spontaneous cerebellar hemorrhage were admitted to the Emergency Neurosurgery Unit, University of Rome “La Sapienza” (Italy). On admission all patients underwent a standard neurological examination, (Glasgow Coma Scale) and a computed tomographic scan. The diameter and the site of the hematoma, a coexisting tight posterior fossa, and the presence of hypertensive hydrocephalus were the criteria, in association with the patients’ neurological status, used as indications for surgery. RESULTS

Operative mortality was nil; and perioperative mortality eight patients (16%, increasing to 24% including the four patients who were deeply comatose on admission). Most patients who died (seven of eight) had two or more general medical risk factors (arterial hypertension and diabetes mellitus; arterial hypertension and liver disease; or liver disease and hematological disorders). CONCLUSION

In patients presenting with spontaneous cerebellar hemorrhage the essential criteria indicating surgery are a hematoma 40 mm ⫻ 30 mm on CT imaging in the cerebellar hemisphere or 35 mm ⫻ 25 mm on CT imaging in the vermis, the presence of a tight posterior fossa (critical size reduced by 10 mm), and a Glasgow Coma Score less than 13. © 2001 by Elsevier Science Inc. Address reprint requests to: Dr. Luigi Cervoni, Via Conteverde 50, scala C, int 10, 00185 Rome, Italy. Received July 25, 2000; accepted October 23, 2000. 0090-3019/01/$–see front matter PII S0090-3019(01)00347-0

he distinguishing features of spontaneous cerebellar hemorrhage (SCH) have only recently been defined [1– 6,8 –29,31–56]. The use in day-today clinical practice of diagnostic techniques such as computed tomography (CT) and magnetic resonance imaging (MRI) has helped to delineate the features of a syndrome distinguished by the absence of vascular and nonvascular anomalies [1– 6, 8 –29,31–56]. We present our results in 50 patients with SCH seen over an 8-year period in a University Emergency Neurosurgery Unit (Table 1).

T

Material and Methods From January 1, 1990 to December 31, 1997, a series of 50 consecutive patients with SCH were admitted to the Emergency Neurosurgery Unit of the Department of Neurosciences at the University of Rome “La Sapienza” (Italy). On admission all patients underwent a standard neurological examination, a complete clinical examination, and routine laboratory tests. The level of consciousness was graded according to the Glasgow Coma Scale (GCS) [7,30] assigning patients to three groups: those with scores of 3 to 7 (coma and semicoma), scores of 8 to 11 (stupor and somnolence), and scores of 12 to 14 (drowsiness and alertness). The diagnostic imaging protocol included computed tomographic (CT) and magnetic resonance imaging (MRI) scans of the brain and in doubtful cases, angiography. CT and MRI scans were obtained to provide information on variables known © 2001 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010

Cerebellar Hemorrhage

1

Surg Neurol 157 2001;55:156 –61

Summary of Clinical Features of 50 Cases

CLINICAL

M:F ratio Mean age Symptoms at admission

GCS ⫽ 14 GCS ⬎ 12 GCS ⬎ 9 GCS ⬍ 9 Coexisting illnesses Site Vermis Hemisphere Vermis-hemisphere Surgical management Operative mortality Perioperative mortality Outcome: At discharge At 6 months At 1 year

PATIENTS

FINDINGS

1.5:1 63.7 Coma ⫽ 30 Intracranial hypertension ⫽ 11 Headache ⫽ 9 9 26 11 14 30 16 30 4 28 0 12 KPS ⬎ 70 ⫽ 28 KPS ⬍ 10 ⫽ 10 6 lost 6 dead KPS ⬎ 70 ⫽ 20 Unchanged

to influence therapy and prognosis: the location of bleeding (vermis, cerebellar hemispheres, or both sites); the size of the lesion (diameter in mm); evidence of blood leakage into the ventricles (notably the fourth); evidence of blood invasion into the brainstem; the presence of hydrocephalus; signs of brainstem compression; the presence and amount of perilesional edema; and evidence of a tight posterior fossa (TPF). A TPF was defined according to the criteria proposed by Weisberg et al [54] as nonvisualization of the basal cisterns in the posterior cranial fossa, an increase in the size of the third ventricle, and lateral ventricles including the temporal horns, and nonvisualization of the fourth ventricle (inconstant). Information was also obtained on pre-existing or coexisting illnesses including diabetes mellitus, arterial hypertension, blood and coagulation disorders, and liver disease. The initial assessment also included standardized serial monitoring of clinical progression with the criteria used at the time of admission (general clinical and neurological examination, GCS, and routine laboratory tests). The standardized clinical examinations were conducted as far as possible at the same times each day. To assess the possible influence of age on disease outcome we also grouped the 50 patients according to age: under 60 years and over 60 years.

2

Statistical Analysis of Factors Influencing Outcome

Age and outcome Intraventricular extension of hemorrhage and GCS score Extension of hemorrhage, hypertensive hydrocephalus, and GCS score at admission Long-term disability and GCS ⬎6 at admission Survival and GCS ⬎6 at admission Two or more risk factors and mortality Two or more risk factors and outcome Single risk factor and outcome Site of hematomas and outcome

P ⫽ 0.9 P ⫽ 0.3 P ⫽ 0.04 P ⫽ 0.06 P ⫽ 0.07 P ⫽ 0.03 P ⫽ 0.04 P ⫽ 0.07 P ⫽ 0.9

Three patients also underwent single photon emission computed tomography (SPECT) after the onset of cerebellar hemorrhage and one of them had a repeat scan 3 months later. The patients’ social and functional recovery was monitored with Karnofsky Performance Status (KPS) scale. In the statistical analysis, chi-square t-tests were used to assess outcome variables including survival and Student’s t-test to compare continuous variables including age and size of hemorrhage in the two age groups. P values equal to or less than 0.05 were considered to indicate statistical significance (Table 2).

Results The study comprised 50 patients, 30 males (60%) and 20 females (40%), with an average age of 61 years. Eight patients (16%) (2 men and 6 women; M:F ratio 1:3) were older than 80 years. No significant differences were found between the sexes. Thirty patients (60%) were comatose on admission. At neurological examination 26 patients (52%) had GCS scores ⬎9 and 11 (22%) scores ⬎12. On admission four patients (8%) were in a deep coma (GCS score of 3). The mean time elapsed between the onset of signs and symptoms of hemorrhage and admission to the hospital was 13 hours (range 4 –72 hours). In most patients (80%) the first symptom was headache. Most patients had coexisting illnesses: 18 patients diabetes mellitus (36%), 16 (32%) arterial hypertension, 7 (14%) blood coagulation disorders, and 8 (16%) liver disease. Nineteen patients (38%) had two or more of these conditions. Hematomas in 16 patients (32%) involved the vermis, in 30 patients (60%) the cerebellar hemisphere, and in the remaining 4 patients (8%) both sites. Neuroimaging studies revealed hydrocephalus in 24 patients (48%), evidence of blood leakage into

158 Surg Neurol 2001;55:156 –61

the ventricular system in 19 patients (38%), and coexisting TPF in 19 cases (38%). No statistical relation was found between intraventricular extension of hemorrhage and GCS scores (P ⫽ 0.3) except in patients with hypertensive hydrocephalus (P ⫽ 0.04). Surgical treatment was undertaken in patients with hemispheric hematomas measuring more than 40 mm in maximum and 30 mm in minimum diameter on CT imaging if located in the cerebellar hemispheres, and in those with vermian hematomas larger than 35 mm and 25 mm in diameter and a GCS score of ⬍13. In patients with coexisting TPF these indexes were lowered by 5 mm. Surgical management consisted of clot evacuation; combined lobectomy was done only for hematomas involving more than 80% of the total surface of the lobe. In patients with GCS scores of ⬎13 (18 cases, 36%) the mean diameter of the hematoma was usually smaller, median 18 ⫻ 16 mm. None of these 18 patients underwent surgery. Of the remaining 32 patients, 28 underwent surgery, whereas the 4 patients who arrived at the hospital in a state of deep coma did not. Operative mortality was nil, and perioperative mortality 8 patients (16%). Most patients who died (7 of 8) had two or more general medical risk factors. The mean duration of the hospital stay was 16 days. In seven cases (14%) bronchopulmonary infections developed. In one patient wound dehiscence complicated the postoperative course but responded to treatment. Eighteen patients (36%) needed mechanical ventilation. Hypertensive hydrocephalus invariably responded to early surgical intervention, so patients with this complication did not have more adverse outcomes than those without. In patients with GCS scores ⬎6 who survived, clinical status at admission did not significantly affect survival (P ⫽ 0.07), but it had a weakly significant effect on the long-term disability outcome (P ⫽ 0.06). Patients with hematomas in the vermis did not have a worse outcome than patients with hemispheric lesions. Nor did they have a longer stay. These data may reflect differences in the criteria used for defining the indications for surgery. No association was found between age and outcome. Conversely, the presence of two or more risk factors significantly influenced mortality (P ⫽ 0.03) and the patients’ quality of life (P ⫽ 0.04). No single general risk factor significantly worsened the outcome (P ⫽ 0.07), possibly because patients with general risk factors (coexisting illnesses) received emergency treatment to correct the imbalance before all other forms of therapy. The 38 patients who survived (four patients in a vegetative state) all

Salvati et al

started the standardized follow-up protocol, but six survivors (15%) were lost to further follow-up. Clinical follow-up observations at 1 year showed that all apallic patients had died as had two others (of bronchopneumonia and myocardial infarction); 20 patients were able to look after themselves. SPECT scans in the three cases studied showed that the areas that returned earliest to satisfactory metabolic levels were the frontal lobes followed by the parietal, occipital, and cerebellar lobes. Temporal lobe function recovered later as did the deeper structures (brainstem). These data in one case contrasted with the clinical data that seemed congruous with a “normal” person of the same age and sex. Neurophysiological techniques (brainstem evoked potentials) in this case showed slightly prolonged conduction times.

Discussion Even though Hillairet and Brown-Sequard had described the disorder more than a century earlier [8,21], the first meaningful description of an SCH dates back some decades to the report by Mitchell published in 1942 [33]. This was soon joined by other undoubtedly important studies, including those of McKissock in 1960 [30] and Ott in 1974 [38] describing the early clinical signs. Yet for its full delineation the syndrome had to await the coming of the CT era. With the radiological and semeiologic means available before this advance, recognizing a true SCH usually was an unusually difficult undertaking. In addition, the surgical indications remained imprecise because they relied upon indirect signs of the lesions. The first noteworthy report on the surgical indications comes from the study conducted by Little [26] with the help of CT scanning. Equally fundamental to the surgical indications was the paper by Weisberg in 1986 introducing the concept of the TPF [54]. The article underlined the extreme discrepancy between the various mortality rates reported over the years (ranging from the 73.5% reported by McKissock in 1960 to 20% reported by Luparello in 1995) [28,30]. The advanced age of the patients with SCH in our series agreed with the published data indicating that SCH occurs in the sixth to eighth decades of life [1–2,7,11,14 –15,17,22–27,30 –31,33,38,40 – 42,48 – 51,53]. This tendency to occur in elderly persons does not necessarily lead to a poor prognosis. Indeed, we found no significant difference in the clinical course of disease among the four age groups (neither for survival nor for disability outcome).

Cerebellar Hemorrhage

The patients’ clinical conditions on arrival at the emergency department had no more than a marginal effect on survival, only approaching statistical significance for patients with GCS scores above 6. In our experience the location of the hematoma— vermian or cerebellar hemisphere— did not seem to be fundamental in determining prognosis. Neither did we find that the presence of hypertensive hydrocephalus worsened the outcome, provided the condition was recognized in time and promptly treated. Conversely when we analyzed the influence on survival outcome of the other single general risk factors (diabetes, arterial hypertension, blood dyscrasia, and liver disease), the presence of two or more of these statistically worsened the outcome, as did the presence of brainstem blood invasion. The neuroradiological diagnosis plays a fundamental role in the staging of these patients [3,10 – 11,15,19 –20,23,29,36,44,47,54 –56]. For early recognition of the disease, MRI seems not to be sufficiently superior to CT to justify the higher costs of this technique. Features such as blood invasion of the brainstem, the presence of hypertensive hydrocephalus, blood in the ventricular system, the size of the hematoma, and the presence of a TPF are all easily visualized on a normal CT scan obtained without injecting contrast material. The real advantage of MRI lies in its ability to give detailed information, to visualize small vascular malformations, and to provide detailed brainstem images. In addition, noninvasive MRI angiography gives excellent images of the regional vascular anatomy. Our experience suggests that a detailed study of these structures can usually be postponed until after the emergency, unless it is needed for therapy. Edema surrounding the SCH— unlike edema surrounding expansive neoplasms in the cerebellum— seems to play a minor role. It is also of little practical interest, being overshadowed by the compression and structural injuries caused by the hematoma. The surgical indications remain controversial, especially regarding the criteria for evaluating the size of SCH [1–3,7,10 –11,14 –15,17,22–31,33,38,40 – 42,47–51,54 –56]. In our opinion, a vermis hematoma measuring ⬎35 ⫻ 25 mm on CT imaging or a hemispheric hematoma measuring more than 40 ⫻ 30 mm requires surgery. The reason the criteria related to size differ for vermian and hemispheric hematomas is that these two regions are intrinsically different (vermian hematomas lie closer to the brainstem and to the CSF pathways). As Kobayashi et al [23] have emphasized, the question of whether to treat patients belonging to the groups previously considered to be candidates for surgical treatment

Surg Neurol 159 2001;55:156 –61

but with a GSC above 13 remains controversial. We saw none of these cases in our series. Nonetheless, patients presenting with these clinical and radiological features probably warrant a watchful, armed, wait-and-see approach. Evidence supporting this conclusion includes the minimal evolution of disease we saw in patients who had no surgery and who were in reasonably good clinical condition on admission. Another indication to consider is a TPF [54]. Evidence of a TPF justifies altering the aforementioned management criteria (reducing by 10 mm the hematoma diameter considered to indicate surgical treatment). Subdividing patients presenting with evidence of a TPF in relation to other factors seems to us of little practical value. Unfortunately we have no data on assessing a hematoma with CT-guided stereotactic fibrinolysis [10,34]. The duration of these patients’ hospital stay reflects the crucial functional importance of the cerebral region involved by the hematoma. Though undeniably interesting, our SPECT data on the functional recovery of the various encephalic regions lacks scientific value owing to the small study sample leading to statistically nonsignificant results. We can find no reasonable scientific basis for performing surgery in patients who arrive at the hospital in a deep coma of more than 2 hours’ duration. Support for this conclusion comes from accumulated data published over the years and also from these patients’ advanced age. Although the use of simple assessment techniques that are easy to standardize makes assessment of the outcome variables slightly less precise, this limitation is outweighed by the more easily reproducible study data they provide. In our study the GCS and KPS amply fulfilled such requirements. An especially encouraging finding was our patients’ return to daily life. The ability to look after oneself is a far more important achievement that owes much to the efforts of social workers, psychologists, and physiotherapists, mainly working on a voluntary basis.

Conclusions SCH in middle-aged or elderly patients is a treatable disease. Despite its presentation the clinical course of disease is often favorable. In assessing the surgical indications for SCH the essential criteria are the diameter of the hematoma on CT imaging (40 ⫻ 30 mm in the cerebellar hemisphere and 35 mm ⫻ 25 mm in the vermis), the presence of a TPF (reduces the critical size by 10 mm), and the GCS score.

160 Surg Neurol 2001;55:156 –61

Patients with GCS scores ⬎13 should be managed conservatively irrespective of the diameter of the hematoma. Invasion of blood into the brainstem and the presence of two or more general medical risk factors significantly worsen the outcome. The presence of hypertensive hydrocephalus, if recognized and treated in time, does not adversely affect the outcome. Deeply comatose patients should receive conservative management. As long as they receive adequate supportive care even elderly patients with spontaneous cerebellar hemorrhage can make a surprisingly good recovery. REFERENCES 1. Abud-Ortega AF, Rajput A, Rozdilsky B. Observations in five cases of spontaneous cerebellar hemorrhage. Can Med Assoc J 1972;106:40. 2. Acampora S, Guarnieri L, Troisi F. Spontaneous intracerebellar hematoma. Report of ten cases. Acta Neurochir 1982;66:83– 6. 3. Amacher AL. Spontaneous cerebellar hemorrhage: evolution of an entity. Crit Rev Neurosurg 1996;6: 73–7. 4. Aoki N, Mizuguchi K. Expanding intracerebellar hematoma: a possible clinicopathological entity. Neurosurgery 1986;18:94 – 6. 5. Auer LM, Auer T, Sayama L. Indications for surgical treatment of cerebellar hemorrhage and infarction. Acta Neurochir 1986;79:74 –9. 6. Bogousslavsky J, Regli F, Jeanrenaud X. Benign outcome in unoperated large cerebellar hemorrhage. Report of two cases. Acta Neurochir 1984;73:59 – 65. 7. Brennan RW, Bergland RM. Acute cerebellar hemorrhage: analysis and clinical findings and outcome in 12 cases. Neurology 1977;27:527–32. 8. Brown-Sequard CE. Diagnosis of hemorrhage in the cerebellum. Lancet 1861;2:391. 9. Chadduck WM, Duong DH, Kast JM, Donahue DJ. Pediatric cerebellar hemorrhages. Child’s Nerv Syst 1995;11:579 – 83. 10. Chin D, Carney P. Acute cerebellar hemorrhage with brainstem compression in contrast with benign cerebellar hemorrhage. Surg Neurol 1983;19:406 –9. 11. Da Pian R, Bazzan A, Pasqualin A. Surgical versus medical treatment of spontaneous posterior fossa haematomas: a cooperative study on 205 cases. Neurol Res 1984;6:145–51. 12. De Freixo MF, Garcia MJ, Alcelay LG. Cerebellar hemorrhage, nonsurgical forms. Ann Neurol 1979;6:84. 13. Dunne JW, Chakera T, Kermode S. Cerebellar haemorrhage: diagnosis and treatment. A study of 75 cases. Q J Med 1987;64:739 –54. 14. Fisher CM, Picard EH, Polak A, Dalal P, Ojemann RJ. Acute hypertensive cerebellar hemorrhage. J Nerv Ment Dis 1965;140:38 –57. 15. Gilliard C, Mathurin P, Passagia JG, Kallel S, Thauvoy C, Stroobandt G. L’hematome spontane du cervelet. Neurochirurgie 1990;36:347–53. 16. Greenberg J, Shubick D, Shenkin H. Acute hydrocephalus in cerebellar infarct and hemorrhage. Neurology 1979;29:409 –13. 17. Gullermain P, Lena G, Reynier Y, Vincentelli F, Jazzar

Salvati et al

18. 19. 20. 21. 22.

23.

24.

25.

26. 27. 28. 29.

30. 31.

32. 33. 34.

35. 36.

37.

A. Hematomes intracerebelieux spontanes de l’adulte: 44 cas. Rev Neurol 1990;146:478 – 83. Heiman TD, Satya-Murti S. Benign cerebellar hemorrhages. Ann Neurol 1978;3:366 – 8. Heros RC. Cerebellar hemorrhage and infarction. Stroke 1981;12:17–22. Heros RC. Cerebellar hemorrhage and infarction. Stroke 1982;13:106 –9. Hillairet JB. De l’haemorrhagie cerebelleuse. Arch Gen Med 1858;1:149 – 69, 324 – 40, 411–32, 549 – 68. Kase CS. Cerebellar hemorrhage. In: Kase CS, Caplan LR, eds. Intracerebral hemorrhage. Boston: Butterworth-Heinemann, 1994:425– 43. Kobayashi S, Sato A, Kageyama Y, Nakamura H, Watanabe Y, Yamaura A. Treatment of hypertensive cerebellar hemorrhage: surgical or conservative management? Neurosurgery 1994;34:246 –51. Kubo T, Sakata Y, Sakai SI, Koizuka I, Matsunaga T, Nogawa T. Clinical observations in the acute phase of cerebellar hemorrhage and infarction. Acta Otolaryngol 1988;447:81– 8. Labauge R, Boukobza M, Zinszner J, Blard JM, Paces M, Salvaing P. Hematomes spontanes du cervelet. Vingthuit observations personnelles. Rev Neurol 1983;139:193–204. Little JR, Tubman DE, Ethier R. Cerebellar hemorrhage in adults. J Neurosurg 1978;48:574 – 8. Lui T, Fairholm DJ, Shu T, Chang C, Lee S, Chein H. Surgical treatment of spontaneous hemorrhage. Surg Neurol 1985;23:555– 8. Luparello V, Canavero S. Treatment of hypertensive cerebellar hemorrhage. Surgical or conservative management? Neurosurg 1995;37:552–3. Mathew P, Teasdale G, Bannan A, Oluoch-Olunya D. Neurosurgical management of cerebellar hematoma and infarct. J Neurol Neurosurg Psychiatr 1995;59: 287–92. McKissock W, Richardson A, Walsh L. Spontaneous cerebellar haemorrhage. A study of 34 cases treated surgically. Brain 1960;38:1–9. Medrazzi JJM, Otero JM, Ottino CA. Management of 50 spontaneous cerebellar haemorrhages. Importance of obstructive hydrocephalus. Acta Neurochir 1993; 122:39 – 44. Melamed N, Satya-Murti S. Cerebellar hemorrhage: a review and reappraisal of benign cases. Arch Neurol 1984;41:425– 8. Mitchell N, Angrist A. Spontaneous cerebellar hemorrhage. Report of fifteen cases. Am J Pathol 1942;18: 935– 46. Mohadjer M, Eggert R, May J, Mayfrank L. CT-guided stereotactic fibrinolysis of spontaneous and hypertensive cerebellar hemorrhage: long-term results. J Neurosurg 1990;73:217–22. Neubauer T, Schwenk B. Therapy and prognosis in spontaneous cerebellar hematomas. Adv Neurosurg 1993;21:57– 60. Niizuma H, Suzuki J. Computed tomography-guided stereotactic aspiration of posterior fossa hematomas: a supine lateral retromastoid approach. Neurosurgery 1987;21:422–7. Noterman J, Dewitte O, Baleriaux D, Vandesteene A, Raftopoulos C, Brotchi J. Les hemorragies sousarachnoidiennes au dela de 65 ans. Etude retrospec-

Cerebellar Hemorrhage

38.

39.

40.

41. 42. 43.

44. 45. 46. 47.

48. 49. 50. 51.

52. 53.

tive d’une serie de 72 patients dont 65 cas d’origine aneurismale. Neurochir 1995;41:51–7. Ott KH, Kase CS, Ojemann RG, Mohr JP. Cerebellar hemorrhage: diagnosis and treatment. A review of 56 cases. Arch Neurol 1974;31:160 –7. Patterson RH. Comment on treatment of hypertensive cerebellar hemorrhage. Surgical or conservative management. Neurosurgery 1994;34:250. Puzzilli E, Ruggeri A, Salvati M, Rocchi G, Delfini R. Emorragie cerebellari spontanee. Considerazioni clinico-terapeutiche. Rivista Neurologia 1998;7:31– 4. Richardson AE. Spontaneous cerebellar hemorrhage. In: Vinken PJ, Bruyn GJ, eds. Handbook of clinical neurology, Vol 12. New York: Elsevier, 1972:54 – 67. Rosenberg GA, Kaufmenn DM. Cerebellar hemorrhage. Reliability of stroke. Stroke 1976;7:332– 6. Rousseaux M, Lesoin F, Combelle G, Petit H, Jomin M (1984) Interet et limites de la derivation ventriculaire isolee dans les hematomes cerebelleux non traumatiques. Neurochirurgie 30:41– 44. Salazar J, Vaquero J, Martinez P. Clinical and CT scan assessment of benign versus fatal spontaneous cerebellar haematomas. Acta Neurochir 1986;79:80 – 6. Seelig JM, Selhorst JS, Young HF, Lipper M. Ventricolostomy for hydrocephalus in cerebellar hemorrhage. Neurology 1981;31:1537– 40. Sypert GW. Cerebellar hemorrhage and infarction: Compr Ther 1977;3:42–7. Janeda M, Hayakawa T, Mogami H. Primary cerebellar hemorrhage. The quadrigeminal cistern obliteration on CT scans as a predictor of outcome. J Neurosurg 1987;67:545–52. Turgut M, Ozcan OE, Erturk O, Saribas O, Erbengi A. Spontaneous cerebellar strokes. Clinical observations in 60 patients. Angiology 1996;47:841– 8. Van Calenbergh F, Van Havenbergh T, Goffin J, Plets C. Les hematomes spontanes du cervelet au-dela’de 60 ans. Neurochirururgie 1996;42:162– 8. Van der hoop RG, Vermeullen M, Van Gijn J. Cerebellar hemorrhage: diagnosis and treatment. Surg Neurol 1988;29:6 –10. Van Loon J, Van Calenbergh F, Goffin J, Plets C. Controversies in the management of spontaneous cerebellar hemorrhage. A consecutive series of 49 cases and review of the literature. Acta Neurochir 1993;122: 187–93. Yoshida S, Sasaki M, Oka H. Acute hypertensive cerebellar hemorrhage with signs of lower brainstem compression. Surg Neurol 1993;10:79 – 83. Wai Dhauser E, Hamburger C, Marguth F. Neurosur-

Surg Neurol 161 2001;55:156 –61

gical management of cerebellar hemorrhage. Neurosurg Rev 1990;13:211–7. 54. Weisberg LA. Acute cerebellar hemorrhage and CT evidence of tight posterior fossa. Neurology 1986;36: 858 – 60. 55. Wizer B, Wall M, Weisberg L. The clinical and computed tomographic features of cerebellar peduncular hemorrhage. Neurology 1988;38:1485–7. 56. Zieger A, Vonofakos D, Steudel WI. Nontraumatic intracerebellar hematomas: prognostic value of volumetric evaluation by computed tomography. Surg Neurol 1984;22:491– 4.

COMMENTARY

The authors have reviewed 50 consecutive patients with spontaneous cerebellar hemorrhage to review the criteria for undertaking surgical intervention. They particularly looked at the diameter of the hematoma and the patient’s Glasgow Coma Score. The series had a significant mortality rate, but the patients who died had considerable medical risk factors. The authors feel that the criteria for surgery are a large hematoma and a Glasgow Coma Score less than 13. Patients who are deeply comatose should not undergo surgical intervention, as their outcome is invariably poor. There is nothing startling or new in this article. However, it does add to the literature and confirms the long-held view that a significant hematoma in patients of poor neurological grade should have surgery as a life-saving measure. Unfortunately, several of these patients had very significant medical problems such as diabetes, hypertension, and liver disease, which contributed to the high mortality and morbidity. The authors note that there is a large number of patients in the older age group with this disease, but there was no difference in outcome unless the patients had significant medical risk factors. Michael Besser, MBBS, FRACS Neurosurgeon Sydney, Australia