Natural history of cerebral dot-like cavernomas

Natural history of cerebral dot-like cavernomas

Clinical Radiology xxx (2013) e453ee459 Contents lists available at SciVerse ScienceDirect Clinical Radiology journal homepage: www.clinicalradiolog...

742KB Sizes 0 Downloads 41 Views

Clinical Radiology xxx (2013) e453ee459

Contents lists available at SciVerse ScienceDirect

Clinical Radiology journal homepage: www.clinicalradiologyonline.net

Natural history of cerebral dot-like cavernomas O. Nikoubashman a, M. Wiesmann a, *, E. Tournier-Lasserve b, K. Mankad c, M. Bourgeois d, F. Brunelle e, C. Sainte-Rose d, M. Wiesmann a, M. Zerah d, F. Di Rocco d a

Department of Diagnostic and Interventional Neuroradiology, University Hospital Aachen, Aachen, Germany INSERM e UMR-S 740, G en etique des Maladies Vasculaires, Universit e Paris VII, Denis Diderot, Paris, France c Radiology Department, Chelsea and Westminster Hospital, London, UK d Service de Neurochirurgie P ediatrique, Groupe Hospitalier Necker Enfants Malades, Universit e Paris V, Ren e Descartes, Paris, France e Service de Radiologie P ediatrique, Groupe Hospitalier Necker Enfants Malades, Universit e Paris V, Ren e Descartes, Paris, France b

art icl e i nformat ion Article history: Received 3 December 2012 Received in revised form 14 February 2013 Accepted 27 February 2013

AIM: To elucidate the natural history of dot-like or “black spot” cavernomas. MATERIALS AND METHODS: Data of 18 children with black spot cavernomas were analysed retrospectively. RESULTS: Eleven boys and seven girls presented 187 black spot cavernomas during a mean observation period of 5.5 years. Mean and median age at diagnosis of the 187 cavernomas was 9.6 years. There were 70 de novo black spot cavernomas. Boys presented significantly more cavernomas than girls. There were three KRIT1 mutation carriers and four PDCD10 mutation carriers. Children with a PDCD10 mutation presented significantly more lesions than those children with a KRIT1 mutation (mean number of lesions per patient: 23.3 versus 3.3, respectively). There were 10 radiological haemorrhagic events caused by 10 black spot lesions. Two of these events were symptomatic. The haemorrhage rate of black spot cavernomas was 0.7% per lesion-year. CONCLUSIONS: A mean bleeding rate of 0.7% per lesion-year is lower than the overall haemorrhage rates provided in the literature. Nonetheless, black spot cavernomas are not purely benign lesions. Furthermore, genetic mutations may play a role in the natural history of black spot cavernomas. Ó 2013 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

Introduction Cerebral cavernous malformations (CCM), also known as cavernomas, are vascular malformations affecting the central nervous system and have a prevalence of 0.3e0.6% and an annual incidence between one and 5.6 new patients per 1,000,000.1e3 The existence of multiple lesions has been shown in up to 27% of cases.4 Although the imaging appearance of cavernomas may vary, magnetic resonance imaging (MRI) has been proven to * Guarantor and correspondent: M. Wiesmann, Department of Diagnostic and Interventional Neuroradiology, University Hospital Aachen, Pauwelsstr. 30, 52074 Aachen, Germany. E-mail address: [email protected] (M. Wiesmann).

be both a highly specific and sensitive tool in the diagnosis of CCM. In summary, the imaging appearances correspond mostly to the pathophysiological hallmark of CCM, which is recurrent thrombosis and haemorrhage.5 Among other manifestations, such as mulberry-like lesions, small T2*-hypointense dots 1 mm corresponding to small cavernomas are a typical form of CCM. These small cavernomas are defined as type IV lesions by Zabramski et al.6 and occur in the context of multiple cavernomas (Table 1).6 They correspond to susceptibility artefacts in T2*-weighted gradient-echo (T2*-GRE) sequences caused by haemosiderin depositions in small cavernomas that are not or barely visible in spin-echo (SE) sequences.6 Whereas there are numerous publications dealing with cavernomas,7e15 little is known about the nature of these

0009-9260/$ e see front matter Ó 2013 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.crad.2013.02.010

e454

O. Nikoubashman et al. / Clinical Radiology xxx (2013) e453ee459

Table 1 Magnetic resonance imaging (MRI) signal characteristics of cerebral cavernomas according to Zabramski et al.6 Lesion type

MRI signal characteristics

Type I

T1: hyperintense core T2: hyper- or hypointense core with surrounding hypointense rim T1: reticulated mixed signal core T2: reticulated mixed signal core with surrounding hypointense rim T1: iso- or hypointense core T2: hypointense with a hypointense rim that magnifies the size of the lesion GE: hypointense with greater magnification than T2 T1: poorly seen or not visualized at all T2: poorly seen or not visualized at all GE: punctate hypointense lesions

Type II

Type III

Type IV

dot-like or so-called “black spot” cavernomas and their clinical impact. Thus, the aim of the present study was to analyse the natural history of these cavernomas.

Materials and methods Data from the medical files of the Department of Paediatric Neurosurgery of Hopital Necker Enfants Malades, Paris were assessed between 1 January 1993 and 31 December 2009 to provide results concerning epidemiological aspects, imaging features and clinical considerations. Charts and MRI studies of these patients were analysed retrospectively. Among 70 patients with clear radiological and/or pathological criteria of a cerebral cavernoma, 18 patients (26%) had additional black spot lesions in the context of a cavernomatosis and were included in the study. Children with a history of radiotherapy were excluded from the study. Genetic analysis had been intended in all 18 patients. However, consent of the patients’ parents could be obtained only in seven of the 18 cases. In these seven patients, the CCM1, CCM2, and CCM3 genes were analysed for mutations. Results of the genetic analyses were extracted from the medical files.

According the Zabramski classification a black spot cavernoma was defined as a dot-sized T2* hypointense lesion that was not or only poorly visible at T1-weighted (W) and T2W SE imaging (Figs. 1 and 2).6 All remaining lesions that were clearly visible at T1W and T2W imaging were defined as either a Zabramski type 1, 2, or 3 lesion (Table 1; Fig. 2).6 T1W SE, T2W SE, and GRE imaging sequences were consistently acquired. Only MRI studies that were performed at 1.5 T were included, in order to take into account the effect of magnetic field strength on susceptibility artefacts. Assessing the true diameter of a black spot cavernoma is a dilemma, as it is by definition only poorly or not at all visible at SE imaging. Measuring lesion diameters based on susceptibility artefacts alone, however, is not reasonable, as the appearance and size of a susceptibility artefact depend strongly on the amount of haemosiderin deposition and technical aspects, such as section thickness and orientation. Therefore, the diameter of a dot-sized lesion in T2* was arbitrarily defined as 1 mm. In order to differentiate between a haemorrhagic event and simple intralesional thrombosis, a radiological haemorrhagic event was postulated if (1) there was extralesional haemorrhage (acute or subacute degradation products of blood) or (2) there was intralesional haemorrhage (acute or subacute degradation products of blood) accompanied by lesion growth, mass effect, or oedema. In accordance with criteria proposed by Al-Shahi et al.17 lesion growth alone without signal changes and the presence of haemosiderin alone without signs of recent haemorrhage were not considered as haemorrhagic events. The appearance of subacute blood degradation products without lesion growth was also not considered as a haemorrhagic event.16 Lesion diameters in T1W and T2W SE images were averaged whenever possible when the size of cavernomas other than black spot cavernomas was assessed. Overall haemorrhage size was assessed if a distinct lesion within the haemorrhage was not measurable. Only lesions with a radiological follow-up period of >365 days were included for haemorrhage rate calculations. Lesion growth was

ˇ

Figure 1 Axial T2*W (a), T2W (b), and T1W (c) MRI images of a 6-year-old male patient (case 5). There is a black spot cavernoma in the left frontal lobe (arrowhead) with a typical hypointense signal in T2*W MRI images and without corresponding signal changes in T1W or T2W MRI images.

O. Nikoubashman et al. / Clinical Radiology xxx (2013) e453ee459

e455

Figure 2 Coronal T2W (a) and sagittal T1W (b) MRI images of a 10-year-old male patient (case 18). The thick arrow (a) points at a Zabramski type I cavernoma, while the thin arrow (a) points at a Zabrasmi type II cavernoma. A Zabramski type III cavernoma is found in the cerebellum (b: arrowhead; Table 1).

defined as a visually distinct increase of lesion diameter of at least 1 mm. Lesions were regarded as de novo lesions only when their new appearance could be shown in two comparable (i.e., section thickness and orientation) consecutive series. Thus, a lesion would not be regarded as de novo if there was no adequate precedent series available. All statistical analyses were performed with SPSS 16 Ò software.

Results Eighteen children presented 187 black spot cavernomas additionally to other cavernomas during a mean observation period of 5.5 years per child (range: 18 days to 13.1 years). The mean number of black spot lesions was 10.4 ranging from 1 to 24. Bigger dominant cavernomas initially lead to symptoms (seizures, signs of raised intracranial pressure, or focal neurological deficits) in 16 children, but none of the 187 black spot cavernomas could be related to any given symptoms during the observation period as long as these lesions were not haemorrhagic. Overall 104 MRI studies were performed (mean: 5.8 per child, range: 2e12 MRI studies). There were 884 distinct observations of 187 black spot cavernomas with a cumulative observation period of 721.3 lesion-years. Mean observation period per lesion was 4.0 years ranging from 10 days to 13.1 years (SD: 3.2 years). Ten children developed 70 de novo black spot cavernomas during the observation period. Mean observation time of de novo black spot cavernomas was 5.8 years (17 days to 12.3 years). The annual rate of new lesions per patient was 0.7. No risk factors for the development of new cavernomas could be identified.

There were 11 boys and seven girls. Boys presented significantly more cavernomas than girls (p < 0.01): 150 cavernomas were found in boys (mean 13.6; range 1e24) whereas only 37 cavernomas were found in girls (mean 5.3, range 1e12). Mean and median age at first diagnosis of a black spot cavernoma was 8.3 and 7.2 years, respectively (range 0.7e16.6 years; Table 2). Mean and median age at diagnosis of all 187 cavernomas was 9.6 years (range 8 months to 16.6 years). No significant difference between boys and girls was found concerning age at diagnosis and age at last follow-up. Lesion distribution was similar to distributions provided in the literature8e15: 59 (32%) lesions were found in the frontal, 35 (19%) in the temporal, 33 (18%) in the parietal, and 31 (17%) in the occipital lobes. Furthermore, 16 (9%) lesions were in the deep central structures (basal ganglia and thalamus), whereas five (3%) lesions were found in the brainstem and eight (4%) were located in the cerebellum. No side of the brain was predominant, with 93 lesions on the left side, 93 on the right side, and one lesion in the midline. A family history of cavernoma was demonstrated in nine cases. Seven of the remaining nine children had at least one first- or second-degree relative with a history of neurological symptoms loosely suggestive of a cavernoma but without radiological or histopathological proof. No neurological diseases were known in the family of one child. The family history of another child was not available. Seven children (six boys and one girl) were genetically analysed for CCM1, CCM2 and CCM3 mutations. Mutations were found in every tested child. A mutation in the KRIT1 gene (CCM1) was found in four children and a PDCD10 (CCM3) mutation was found in three children. Children with a PDCD10 mutation presented significantly more lesions than those children with a KRIT1 mutation (p < 0.01, Student’s t-test). The mean number of lesions in PDCD10 patients was 23.3 (range 2e24 lesions), whereas mean

e456

O. Nikoubashman et al. / Clinical Radiology xxx (2013) e453ee459

Table 2 Patient characteristics Case

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Sex

M F M F M M F M F M M F M F F M M M

Age at diagnosis of first BSC (years)

Family history

2.7 11.8 5.2 6.2 6.1 13.2 5.8 1.9 9.6 0.7 2.9 8.2 14.0 12.7 13.3 12.0 16.6 5.8

Possibly familial Possibly familial Familial Familial Familial Familial Familial Familial Familial Familial Possibly familial Possibly familial Possibly familial Not familial Familial N.D. Possibly familial Possibly familial

Mutation

N.D. N.D. N.D. N.D. KRIT1 N.D. N.D. N.D. KRIT1 PDCD10 KRIT1 N.D. KRIT1 N.D. N.D. N.D. PDCD10 PDCD10

BSC

All CCM

Initial no.

Final no.

Initial no.

Final no.

3 0 3 1 7 10 12 3 3 3 0 2 1 0 2 22 22 23

3 9 12 7 7 15 12 19 4 24 1 2 1 1 2 22 22 24

8 3 4 2 10 13 17 5 5 4 2 4 2 1 10 41 29 48

8 13 15 9 10 18 17 26 7 28 4 4 2 2 10 41 31 53

Observation period (years) 1.4 9.7 10.4 12.3 2.4 2.3 3.7 13.1 6.7 12.4 3.4 2.3 18 daysa 5.2 1.6 4.7 3.0 4.5

BSC, black spot cavernoma; F, female; M, male; familial, proven family history of cavernoma; possibly familial, at least one first- or second-degree relative with a history of neurological symptoms loosely suggestive of a cavernoma but without radiological or histopathological proof; N.D., not determined; no.: number. a Patient was foreign and lost to follow-up after initial presentation.

number of lesions in KRIT1 patients was 3.3 (range 1e7 lesions; Fig. 3). The radiological observation period was greater than 365 days in 154 of 187 black spot lesions. Only these lesions 26 24 22

No. of black spot cavernomas

20 18 16 14 12 10 8 6 4 2 0 KRIT1

PDCD10

DNA mutation Figure 3 Boxplot showing the number of black spot cavernomas depending on genetic mutation.

were included in the haemorrhage rate calculations. No haemorrhage occurred in the 33 excluded cases. During the observation period there were 10 radiological events of haemorrhage caused by 10 black spot lesions in five different children. The mean diameter of these haemorrhagic lesions was 5.1 mm (range 3e13 mm). Two of the 10 haemorrhagic events were symptomatic: one haemorrhage of a cavernoma in the basal ganglia was associated with a sudden headache (case 3); another haemorrhage within the brainstem was accompanied by signs of raised intracranial pressure and focal neurological deficits (case 10; Fig. 4). Three of the eight haemorrhagic events that were asymptomatic occurred in the white mater of the occipital and temporal lobes. Four of eight asymptomatic haemorrhagic events occurred in subcortical areas of the frontal, temporal, and occipital lobes. There was one haemorrhagic event of a black spot cavernoma located in the cortex of the temporal lobe. A second haemorrhage occurred in three lesions, and a third haemorrhage in one lesion. These five re-haemorrhages were not considered when the haemorrhage rate of black spot lesions was calculated. Twelve different lesions in five children grew and were consequently visible at T1W- and T2W-examinations without radiological proof of recent haemorrhage. The mean size of these lesions was 2.8 mm (range 2e5 mm). Increase of lesion diameter was discovered in a mean time of 353 days (ranging from 50 days to 2.1 years) after the previous MRI examination. The radiological observation period was >365 days for 88 lesions that were present at first imaging. The haemorrhage rate of these lesions was 0.7% per lesion-year. The haemorrhage rate of 58 de novo lesions with radiological observation periods >365 days was 0.6% per lesion-year (two events during a mean observation time of 5.9 years; one haemorrhagic event after 2.7 years, and another

O. Nikoubashman et al. / Clinical Radiology xxx (2013) e453ee459

e457

Figure 4 Axial T2*W (a) and T1W (b) MRI images of an initially asymptomatic 1-year-old male patient (case 10). (a) First MRI in the context of hereditary cavernomatosis. There is an asymptomatic black spot cavernomas in the brainstem (a: arrow). (b) Sudden onset of headache, nausea, nystagmus, and strabismus 10 months after the initial diagnosis. There is haemorrhage with mixed blood degradation products at T1W MRI (b: arrow).

haemorrhagic event 5.9 years after first discovery of the de novo lesions). The mean overall haemorrhage rate of all 154 included black spot cavernomas with a radiological observation period >365 days was 0.7% per lesion-year.

Discussion Genetic analyses and MRI examinations have fundamentally changed our understanding of the nature of intracranial cavernomas. The discovery of de novo lesions, the link to genetic mutations, and a considerable haemorrhage rate underline the clinical need to better understand the natural history of cavernomas. A limitation of this work is the diagnostic uncertainty concerning T2*-black spots. Establishing a definite diagnosis of black spot lesions in T2*W- images always poses a dilemma, especially as histopathological examinations of these lesions are rarely available. Differential diagnosis of black spots in adults also comprises microangiopathic or amyloid angiopathic micro-bleeds. These differential diagnoses are extremely rare in children. In addition, no patient presented with a history of irradiation as a possible cause for cavernomas. Furthermore, there was no history of a long-lasting non-steroidal anti-inflammatory drug (NSAID) medication as a possible cause for micro-bleeds. Thus, black spots in children (especially if accompanied by a bigger cavernoma) are highly suggestive of a cavernoma, even in the absence of histopathological examination. Therefore, it is assumed that the lesions included in the present paper do correspond to small cavernomas. A further problem is the fact that the appearance of small cavernomas depends strongly on MRI settings, such as field strength and section thickness and orientation. Furthermore, recent studies investigating the role of susceptibility-weighted imaging (SWI) raise interesting questions concerning the natural history of black spot cavernomas and cavernomas in general, given that the number

of detected lesions with SWI is higher than the number of lesions detected with T2* imaging.16 In addition, differentiating between an actual haemorrhagic event and slow thrombosis of a lesion is not always possible. Subtle changes of lesion diameter or lesion signal may be missed or misinterpreted as haemorrhagic events. This is why the definition of haemorrhage was kept both simple and strict: a radiological haemorrhagic event was defined as extralesional haemorrhage with acute or subacute degradation products of blood or intralesional haemorrhage with acute or subacute degradation products of blood accompanied by lesion growth, mass effect, or oedema. Further limitations of this work are the small sample size and the retrospective approach. Approximately 25% of cavernoma patients present multiple cavernomas, including black spot lesions, and paediatric cases represent approximately 20% of all cavernoma patients.1e4 Consequently, recruiting a considerable number of paediatric patients with black spot lesions in a single centre can be an onerous task. Nonetheless, there is a need for better understanding of black spot lesions, as, for instance, the diagnosis of such a lesion in the brainstem poses the question as to whether a patient might or might not be at risk of a haemorrhagic event in the future. Therefore, even retrospective data may be helpful in daily clinical practice until there are prospective data dealing with the natural history of black spot cavernomas. Despite all its limitations, this first analysis concerning the natural history of black spot cavernomas in children may lead to a better understanding of these lesions. One hundred and eighty-seven asymptomatic lesions were analysed. The dynamic nature of black spot cavernomas and the rather high mean annual bleeding rate of 0.7% per lesion were not expected. This figure may even underestimate the actual haemorrhage risk in consideration of 12 lesions that grew without radiological proof of recent haemorrhage. On the one hand, lesion growth in these cases may be explained by thrombosis or radiologically occult

O. Nikoubashman et al. / Clinical Radiology xxx (2013) e453ee459

micro bleeding; on the other hand, lesion growth due to a past undiscovered haemorrhagic event cannot be excluded given the long intervals between the examination in which lesion growth was discovered and the previous examination. In summary, the results suggest that black spot cavernomas are anything but benign compared to other cavernomas and should not be underestimated. Even though one must bear in mind different definitions of haemorrhage in the literature, overall haemorrhage rates in prospective studies range from 0.6e1.1% per lesion-year.6,17e21 The haemorrhage risk of black spot cavernomas in this series is small but roughly in the same range. Al-Holou et al.7 who analysed paediatric cases while excluding all black spot cavernomas showed a slightly higher annual haemorrhage rate of 0.9% per lesion.7 All black spot cavernomas in the present series were asymptomatic during the observation period unless there was haemorrhage. In the present series, focal neurological deficits and seizures could be related to bigger cavernomas with the aid of electroencephalography and MRI examinations. Consequently, there is no evidence proving that black spot cavernomas per se cause focal neurological deficits or are epileptogene. However, the converse argument that black spot cavernomas are always without clinical consequences is not correct. The present study showed that black spot cavernomas may bleed and grow to a considerable size. Two of 10 haemorrhagic events caused clinical symptoms depending on haemorrhage location and size. Conversely, bigger cavernomas may evolve slowly from black spot cavernomas as a result of micro-bleeds. It is also conceivable that clinical symptoms are caused by chronic growth of a cavernoma. Although there are several similarities between cavernomas in the paediatric and adult age group, there were some striking discrepancies. Lesion number rises with increasing age, as it has been shown in other publications (Fig. 5).7,21,22 Denier et al.22 reported an eightfold higher prevalence of KRIT1 mutation carriers compared to PDCD10 mutation carriers in a predominantly adult population.22 Both groups presented equal numbers of cavernomas.21,22 In the present paediatric series, however, PDCD10 mutations were almost as frequent as KRIT1 mutations (three versus four of seven). Furthermore, children with a PDCD10 mutation presented significantly more lesions than those children with a KRIT1 mutation at similar age (p < 0.01, Student’s t-test; Fig. 3).22 The high number of lesions in PDCD10 patients and the resulting higher cumulative haemorrhage risk may be an explanation for an earlier onset of clinical symptoms in PDCD10 patients that was observed by Denier et al.22 The discrepancy between lesion numbers in adults and children with regard to genetic mutations and mutation frequencies may suggest that these differences might even out during adolescence. Although there are no statistically significant differences between boys and girls with regard to age at diagnosis or observation period, boys account for significantly more lesions than girls (p ¼ 0.025, Student’s t-test). A possible

26 24 22

No. of black spot cavernomas

e458

20 18 16 14 12 10 8 6 4 2 0

0

2

4

6

8

10

12

14

16

18

20

Age [years]

Figure 5 Cumulative number of black spot cavernomas depending on age at imaging. Lines connecting consecutive examinations of one child. Thick line represents a linear fit line for all values.

explanation may be that boys and girls may differ in the type of genetic mutations. Furthermore, hormonal factors may play a role in lesion development. However, data in the literature concerning the influence of hormonal factors are not conclusive.23,24 The fact that all seven tested children presented mutations and six further children had a proven family history of cavernomas underlines the important role of hereditary factors in cases of multiple cavernomas. In fact, Labauge et al. stated that 75% of supposedly sporadic cases with multiple lesions were hereditary.25 If the results of the present study are confirmed in future prospective analyses, it may be justifiable to perform genetic analyses in all paediatric cases of multiple cavernomas. Early genetic tests may then be a valuable tool to predict clinical risks during childhood, adolescence, and adulthood, and to modify follow-up intervals according to expected clinical risks. In conclusion, in the present series black spot cavernomas were associated with an annual haemorrhage risk of 0.7% per lesion. Twenty percent of haemorrhages of black spot cavernomas in the present series were symptomatic. Thus, black spot cavernomas are not purely benign lesions. Hence, the present authors recommend annual follow-up MRI examinations for patients with black spot cavernomas in strategically important areas, such as, for instance, the brainstem, in order to anticipate possible future clinical symptoms. Even though the total number of analysed cases is small, the present data indicate that genetic mutations may play a pivotal role in cases of multiple cavernomas. Although there are limitations of the present study, this first analysis of the natural history of black spot cavernomas may serve as a first

O. Nikoubashman et al. / Clinical Radiology xxx (2013) e453ee459

step towards a better understanding of black spot cavernomas until there are prospective data.

References € m L. Deep and brainstem cavernomas: a 1. Mathiesen T, Edner G, Kihlstro consecutive 8-year series. J Neurosurg 2003;99:31e7. 2. Brunon J, Nuti C. Natural history of cavernomas of the central nervous system. Neurochirurgie 2007;99:122e30. 3. Al-Shahi R, Bhattacharya JJ, Currie DG, et al. Scottish Intracranial Vascular Malformation Study Collaborators. Prospective, population-based detection of intracranial vascular malformations in adults: the Scottish Intracranial Vascular Malformation Study (SIVMS). Stroke 2003;34:1163e9. 4. Clatterbuck RE, Moriarity JL, Elmaci I, et al. Dynamic nature of cavernous malformations: a prospective magnetic resonance imaging study with volumetric analysis. J Neurosurg 2000;93:981e6. 5. Tomlinson FH, Houser OW, Scheithauer BW, et al. Angiographically occult vascular malformations: a correlative study of features on magnetic resonance imaging and histological examination. Neurosurgery 1994;34:792e800. 6. Zabramski JM, Wascher TM, Spetzler RF, et al. The natural history of familial cavernous malformations: results of an ongoing study. J Neurosurg 1994;80:422e32. 7. Al-Holou WN, O’Lynnger TM, Pandey AS, et al. Natural history and imaging prevalence of cavernous malformations in children and young adults. J Neurosurg Pediatr 2012;9:198e205. 8. Acciarri N, Galassi E, Giulioni M, et al. Cavernous malformations of the central nervous system in the paediatric age group. Pediatr Neurosurg 2009;45:81e104. 9. Alexiou GA, Mpairamidis E, Sfakianos G, et al. Surgical management of brain cavernomas in children. Pediatr Neurosurg 2009;45:375e8. 10. Cavalheiro S, Braga FH. Cavernous hemangioma. In: Choux M, Di Rocco C, editors. Paediatric neurosurgery. London: Churchill Livingstone; 1999. p. 691e701. 11. Edwards JE, Baumgartner MS. Cavernous and other cryptic malformations in the paediatric age group. In: Awad IA, Barrow DL, editors. Cavernous malformations. Park Ridge: AANS Publications; 1993. p. 163e83.

e459

12. Fortuna A, Ferrante L, Mastronardi L, et al. Cerebral cavernous angioma in children. Childs Nerv Syst 1989;5:201e7. 13. Lena G, Ternier J, Paz-Paredes A, et al. Central nervous system cavernomas in children. Neurochirurgie 2007;53:223e37. 14. Mottolese C, Hermier M, Stan H, et al. Central nervous system cavernomas in the paediatric age group. Neurosurg Rev 2001;24:55e73. 15. Xia C, Zhang R, Mao Y, et al. Paediatric cavernous malformation in the central nervous system: report of 66 cases. Pediatr Neurosurg 2009;45:105e13. 16. de Champfleur NM, Langlois C, Ankenbrandt WJ, et al. Magnetic resonance imaging evaluation of cerebral cavernous malformations with susceptibility-weighted imaging. Neurosurgery 2011;68:641e8. 17. Al-Shahi-Salman R, Berg MJ, Morrison L, et al. Haemorrhage from cavernous malformations of the brain: definition and reporting standards. Angioma Alliance Scientific Advisory Board. Stroke 2008;39:3222e30. 18. Kondziolka D, Lunsford LD, Kestle JR. The natural history of cerebral cavernous malformations. J Neurosurg 1995;83:820e4. 19. Moriarity JL, Wetzel M, Clatterbuck RE, et al. The natural history of cavernous malformations: a prospective study of 68 patients. Neurosurgery 1999;44:1166e73. 20. Porter PJ, Willinsky RA, Harper W, et al. Cerebral cavernous malformations: natural history and prognosis after clinical deterioration with or without haemorrhage. J Neurosurg 1999;87:190e7. vy C, et al. The natural history of familial 21. Labauge P, Brunereau L, Le cerebral cavernomas: a retrospective MRI study of 40 patients. Neuroradiology 2000;42:327e32. 22. Denier C, Labauge P, Bergametti F, et al. Genotypeephenotype correlations in cerebral cavernous malformations patients. Ann Neurol 2006;60:550e6. 23. Maraire JR, Awad IA. Intracranial cavernous malformations: lesion behaviour and management strategies. Neurosurgery 1995;37:591e605. 24. Kaya AH, Ulus A, Bayri Y, et al. There are no estrogen and progesterone receptors in cerebral cavernomas: a preliminary immunohistochemical study. Surg Neurol 2009;72:263e5. 25. Labauge P, Laberge S, Brunereau L, et al. Hereditary cerebral cavernous angiomas: clinical and genetic features in 57 French families. Lancet 1998;352:1892e7.