| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Neuroradiology |
1 From the Department of Adult Radiology, Centre Hospitalier Universitaire Bretonneau, F 37044 Tours, France (L.B.); the Faculte Necker-Enfants Malades, Institut National de la Santé et de la Recherche Médicale Unité 25, Paris, France (P.L., E.T.L., S.L.); the Department of Radiology, Hôpital Saint-Antoine, Paris, France (C.L.); and the Department of Neurosurgery, Centre Hospitalier Universitaire Côte de Nacre, Paris, France (J.P.H.). Received March 25, 1998; revision requested June 19; final revision received April 26, 1999; accepted June 17. Address reprint requests to L.B. (e-mail: l.brunereau@chu.med.univ-tours.fr).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Between November 1996 and June 1997, 51 non-Hispanic families with familial cavernous angioma were identified. Cerebral MR images in 83 symptomatic subjects and 73 asymptomatic subjects were reviewed. Spin-echo (SE) and gradient-echo (GRE) MR imaging features of cavernous angioma were recorded and, in 91 subjects with both SE and GRE images, lesions were graded as type 1, 2, 3, or 4, according to a published classification scheme. MR imaging features were compared between symptomatic and asymptomatic subjects, and sensitivities of SE and GRE images were determined.
RESULTS: Multiple lesions were more common than single lesions in both symptomatic and asymptomatic subjects, with no difference in mean number of lesions between groups. More lesions were detected on GRE images than on SE images. Type 1 and type 2 lesions were more numerous in symptomatic than in asymptomatic subjects. The numbers of types 2, 3, and 4 lesions increased with age in both groups.
CONCLUSION: The familial form of cavernous angioma is characterized by multiple lesions and by a correlation between lesion number and subject age. The clinical manifestation may be more closely related to the type of lesion than to the number of lesions. GRE MR images are more sensitive than SE images for demonstration of cavernous angioma.
Index terms: Angioma, central nervous system, 10.1494, 10.3141 • Arteriovenous malformations, cerebral, 10.1494, 10.3141 • Brain, MR, 10.121411, 10.121412 • Familial conditions
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Magnetic resonance (MR) imaging is the most sensitive modality for the diagnosis of cavernous angioma (3,6,9,10). With T2-weighted sequences, the lesion is typically characterized by an area of mixed signal intensity, with a central reticulated core and a peripheral rim of decreased signal intensity related to deposition of hemosiderin (17,18). However, other MR imaging descriptions have been reported (10,19). A classification on the basis of MR imaging features has, therefore, been proposed by Zabramski et al (9). This classification includes four types that are differentiated on the basis of spin-echo (SE) and gradient-echo (GRE) MR imaging features.
Two forms of cavernous angioma have been described: a sporadic form, in which patients usually have a single lesion, and a familial form, the hallmarks of which are multiple lesions and autosomal dominant transmission (4,8). The familial form appears to be very uncommon (2) and has mainly been described in the Hispanic American population (8–11).
In the present study, we analyzed the MR imaging features of familial cavernous angioma in a series of 51 non-Hispanic families that included both symptomatic and asymptomatic subjects. Specifically, the number and type of lesions were evaluated according to the classification of Zabramski et al (9).
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Our first goal was to collect the medical records of symptomatic subjects treated or followed up for symptoms related to their lesions and to retrospectively analyze MR imaging studies obtained in these subjects. Our second goal was to prospectively screen asymptomatic relatives whose cavernous angioma status (affected or unaffected) was unknown. Screening of asymptomatic subjects had been approved by our local ethics committee according to the Huriet law concerning biomedical research in France (20). Several asymptomatic "at-risk" relatives were noted in each family (grandparents, parents, children, cousins, aunts, uncles) and were contacted by telephone. Asymptomatic relatives who gave informed consent underwent cerebral MR imaging, the results of which were prospectively analyzed.
Families
Fifty-one families were listed: 49 of French descent, one of Belgian descent, and one of Italian descent. No families were of Hispanic descent. Diagnosis of cavernous angioma was confirmed on the basis of results from surgical samples in 41 families and results of MR imaging in 10 families.
Subjects
Symptomatic subjects.—We collected the medical records in 100 symptomatic subjects who were members of the 51 families. At least one cerebral MR imaging study was available in 83 of the subjects. In the remaining 17 subjects, there was no evidence that cerebral MR imaging had been performed, and these individuals were excluded from our study. The 83 symptomatic subjects were systematically examined by a neurosurgeon or a neurologist. They were 42 male and 41 female subjects aged 6–86 years (mean ± SD, 38.3 years ± 17.8). Clinical symptoms at the time of diagnosis were seizures (n = 37), cerebral hemorrhage (n = 35), focal neurologic deficit (n = 9), and headaches (n = 2).
Asymptomatic subjects.—Two hundred seventy-eight at-risk asymptomatic relatives who were members of the 51 families were contacted, and 164 gave informed consent to undergo cerebral MR imaging. MR imaging findings were normal in 89 subjects, and one or more lesions were demonstrated in 73. In the two remaining subjects, the MR images were of poor quality and were considered to be uninformative. The 73 asymptomatic subjects were systematically examined by a neurosurgeon or a neurologist. They were 40 male and 33 female subjects aged 11–88 years (mean, 43.7 years ± 15.7).
MR Imaging
In all subjects, MR protocols varied because the MR imaging examinations were performed at different radiology departments and with different MR imagers (1.0 or 1.5 T). In the symptomatic subjects, T1-weighted, intermediate-weighted, and T2-weighted SE studies were available in all subjects. An additional T2*-weighted GRE MR study was available in 31 of 83 subjects. In the asymptomatic subjects, T1-weighted and T2-weighted SE sequences were included in the protocol approved by the ethics committee, but no gadolinium chelate was used. T2-weighted sequences were performed with either conventional SE or fast SE techniques. Additional T2-weighted GRE studies were obtained in 60 of the 73 subjects in whom one or more lesions were detected and in 64 of the 89 subjects in whom no lesions were detected.
Data Collection
Two experienced radiologists (L.B., C.L.) who were unaware of the clinical signs in the subjects reviewed the 156 cerebral MR studies (83 in symptomatic subjects, 73 in asymptomatic subjects). The most recent available MR studies in the 83 symptomatic subjects were retrospectively and independently reviewed by the two radiologists. MR studies in the 73 asymptomatic subjects were prospectively and independently analyzed by the radiologists. In cases of disagreement, a consensus was established at a subsequent common reading. All MR studies were analyzed for (a) the number of lesions, (b) the multiple or single character of lesions, and (c) the supra- or infratentorial location of lesions. Findings on SE and GRE (when obtained) MR images were recorded separately. Furthermore, in the 91 subjects (31 symptomatic, 60 asymptomatic) who underwent both SE and GRE imaging, lesions were assigned to one of the four types defined by Zabramski et al (9) (Table 1, Figs 1–4).
|
|
|
|
|
|
|
|
|
Statistical Analysis
Descriptive analysis of MR imaging findings in both the symptomatic group and the asymptomatic group included (a) mean number of lesions, (b) percentage distribution of multiple and single lesions, and (c) percentage distribution of supra- and infratentorial lesions. Results from SE and GRE images were recorded separately.
A comparative analysis of MR imaging findings was performed with the SE imaging findings in all 156 subjects (83 symptomatic, 73 asymptomatic) and with the SE and GRE imaging findings in 91 subjects (31 symptomatic, 60 asymptomatic) in whom GRE images was obtained. We used the nonparametric Mann-Whitney U test to compare the mean number of lesions or the mean number of each type of lesion in both the symptomatic group and the asymptomatic group. We also used the nonparametric Wilcoxon matched pairs test to compare the mean number of lesions diagnosed on the basis of SE images with the mean number diagnosed on the basis of GRE images in both groups. The sensitivity of SE images for the detection of cavernous angioma and the demonstration of their multiple pattern was calculated by using GRE images as a reference. Linear regression was used to analyze the relationship between the number of lesions or the numbers of different types of lesions and subject age. A P value of less than .05 was considered to indicate a significant difference.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
|
|
The number of lesions significantly increased with subject age in symptomatic (r = 0.36, df = 82, P < .01) and asymptomatic (r = 0.316, df = 72, P < .01) groups (Table 5).
|
The mean number of lesions detected on SE images versus the mean number detected on GRE images was significantly different (7.2 vs 20.2 in symptomatic subjects, P < .001; 4.9 vs 16.3 in asymptomatic subjects, P < .001). SE images did not depict lesions in three asymptomatic subjects that were clearly demonstrated on GRE images. GRE images were, therefore, more sensitive than SE images for demonstration of cavernous angioma. In comparison with GRE images, SE images had a sensitivity of 97% (88 true-positive, three false-negative). Moreover, SE images demonstrated multiple lesions in 76 subjects (84%) and a single lesion in 15 (one symptomatic, 14 asymptomatic) subjects (16%). In these 15 subjects with a single lesion seen on SE images, however, GRE images demonstrated multiple lesions in six subjects (one symptomatic, five asymptomatic) and confirmed a single lesion in nine asymptomatic subjects. In comparison with GRE images, SE images had a sensitivity of 93% (76 true-positive, six false-negative) for demonstration of multiple lesions.
There were significant differences between symptomatic subjects and asymptomatic subjects for the mean numbers of type 1 lesions (0.4 vs 0.1; P < .001) and type 2 (1.9 vs 1.3; P < .05) lesions. There were no such differences for type 3 lesions (5.5 vs 3.8; P = .19) and type 4 lesions (12.5 vs 11.2; P = 0.75).
Overall, the number of lesions increased significantly with subject age in both the symptomatic (R = 0.43, df = 30, P < .05) and the asymptomatic (R = 0.44, df = 59, P < .01) groups (Table 6). The number of type 1 lesions did not significantly increase with subject age in either the symptomatic (R = 0.182, df = 30, P = .32) or the asymptomatic (R = 0.129, df = 59, P = .32) groups. The number of type 2 lesions increased significantly with subject age in the symptomatic group (R = 0.355, df = 30, P < .05) and approached significance in the asymptomatic group (R = 0.249, df = 59, P = .054). The number of type 3 lesions significantly increased with subject age in both the symptomatic (R = 0.511, df = 30, P < .01) and the asymptomatic (R = 0.313, df = 59, P < .05) groups, as did the number of type 4 lesions in the symptomatic (R = 0.36, df = 30, P < .05) and the asymptomatic (R = 0.433, df = 59, P < .01) groups.
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The prevalence of cavernous angioma has been estimated on the basis of autopsy or MR imaging findings to be 0.5%–0.7% (3,5,6). Cavernous angioma may occur as a sporadic form or as a familial form with an autosomal dominant pattern of inheritance. To date, studies with approximately 50 families have been published in the clinical, pathology, and radiology literatures (8–13). However, MR imaging results have been reported only in the most recent studies (8–11,13). The frequency of the familial form is not precisely known. In previous series, the prevalence of the familial form was reported to range from 6% to 50% of all lesions (5,10,14) and appears to be particularly high in individuals of Hispanic descent (8–11). The largest series (8,9) of familial cavernous angioma published to date involved six families, most of whom were of Hispanic descent. To our knowledge, our series of 51 families (156 subjects) is the largest series that does not include individuals of Hispanic descent.
The clinical manifestation appears to be similar in both the sporadic and the familial forms. Seizures are reported (3,4,6,8,14,21–23) as the most common symptom, accounting for 38%–55% of patient complaints. Other symptoms include focal neurologic deficits in 12%–45% of patients (4,6,8,22), recurrent hemorrhage in 4%–32% (4,6,8,22), and chronic headaches in 5%–52% (3,4,6,8,9). In the symptomatic subjects in our series, symptoms included seizures in 45%, focal neurologic deficits in 11%, hemorrhage in 41%, and chronic headaches in 3%; this distribution was not different from those in the previous reports. The majority of patients with cavernous angioma reportedly (3,7,9,24) become symptomatic between the 3rd and 5th decades, and there is no definite association between symptoms and sex. Our subject demographics were similar to those reported in the literature.
Asymptomatic patients with cavernous angioma have been previously described (3,5,6,8,9,12–16). However, the frequency of asymptomatic cavernous angioma is not precisely known and has been estimated at 11.0%–95.5% in various series (3,5,6,9,10,14,16) that included patients with the familial and those with the sporadic form. In the series of Zabramski et al (9), which included only patients with the familial form, 39% of patients were asymptomatic. In our study, despite a selection process involving relatives which might have introduced potential bias, the 46% frequency of asymptomatic cavernous angioma was similar to that reported by Zabramski et al. This high frequency of asymptomatic affected relatives could lead to underestimation of the exact frequency of the familial form of cavernous angioma.
The presence of multiple lesions is characteristic of the familial form. In fact, lesions were multiple in the familial form in 50%–84% of cases reported in the literature (8,9,25). A mean of 5.8–6.5 lesions per patient was reported in series (9,11) that included a majority of Hispanic families. Our results confirmed this predominance of multiple lesions in the familial form (mean of 7.3 lesions in symptomatic subjects and 4.8 in asymptomatic subjects on SE images, and mean of 20.2 lesions in symptomatic subjects and 16.3 in asymptomatic subjects on GRE images). The values we obtained with SE images were similar to those reported in the literature (9,11), although those obtained with GRE images were much higher. Both of these results emphasize the fact that the mean number of lesions is similar in the familial form in non-Hispanic, as well as Hispanic, families. We did not find any significant difference in the mean number of lesions between the symptomatic and asymptomatic groups, as determined with SE or GRE images. This would indicate that the clinical manifestation of familial cavernous angioma is probably not related to the number of lesions.
A supratentorial location of lesions has been reported (3–7) to be more frequent than an infratentorial location. Our results are in agreement with the results in these reports, with supratentorial lesions found in 80%–85% of all subjects in both groups on SE and GRE images.
MR imaging is currently the best imaging modality for evaluation of cavernous angioma (3,6,9,10,17,18,26); it is more sensitive than computed tomography for demonstration of this lesion (6,10,11,17,27). The higher sensitivity of GRE images as compared with that of SE images has been reported (9,11,28). Owing to the blood stagnation phenomenon (29) or to true chronic microhemorrhages, cavernous angiomas contain deoxyhemoglobin or hemosiderin, which generate susceptibility effects and cause a decrease in signal intensity. This loss of signal intensity is better demonstrated with a GRE sequence, particularly a T2-weighted GRE sequence.
Our results confirmed the higher sensitivity of GRE images for the demonstration of cavernous angiomas. The usefulness of GRE images, relative to that of SE images, is clearly demonstrated for detection of multiple lesions (Fig 5). SE imaging did not demonstrate lesions in three subjects (97% sensitivity) and multiplicity of lesions in six subjects (93% sensitivity). On the basis of our results, we recommend that in the setting of the familial form of cavernous angioma, relatives undergo MR imaging with both SE and GRE sequences. If no lesions are demonstrated, relatives can be considered to be unaffected. Any patient who demonstrates a single lesion on SE images but type 4 lesions on GRE images, should be suspected of having the familial form of cavernous angioma, and close relatives should undergo evaluation.
|
|
|
The pathologic definition of type 4 cavernous angioma is unclear. According to Rigamonti et al (30), telangiectasia and type 4 cavernous angioma may represent a single entity. Despite this equivocal definition, our results demonstrate that there is a close relationship between type 4 lesions and the familial form of cavernous angioma. In fact, other types of cavernous angioma were present in 78 of 81 subjects with type 4 lesions in our series, and only three asymptomatic subjects had type 4 lesions alone. Two of these subjects had one parent and at least one child with MR imaging evidence of familial cavernous angioma (all four types) and, thus, were obligate transmitters of the disease. We, therefore, assume that type 4 lesions must be included among the MR imaging features of the familial form of cavernous angioma.
Zabramski et al (9) reported that the MR imaging–based classification of familial cavernous angioma has clinical implications. They found that hyperintense lesions with evidence of subacute or mixed-age hemorrhage (type 1 and 2 cavernous angiomas) were more likely to be present in symptomatic patients than were hypointense lesions (type 3 and 4 cavernous angiomas). Willinski et al (26), however, found that this classification was not useful for predicting future symptoms such as bleeding. Our results confirm the conclusions of Zabramski et al (9) by demonstrating that type 1 and type 2 lesions were more numerous in the symptomatic group than in the asymptomatic group, whereas type 3 and type 4 lesions were equally distributed between the groups. Moreover, we introduce the possibility that a type 1 lesion may be detected in asymptomatic individuals (as in two subjects in our series). Sigal et al (18) believe that these asymptomatic hemorrhages may be related to pure intralesional bleeding.
Various authors (11,13) have reported the possibility that the number of lesions may increase with patient age. In a three-generation family, Horowitz and Kondziolka (31) found an average of five to 12 lesions per patient in the generation of children or adolescents, 20 in the generation of parents, and more than 100 in the generation of grandparents. In our study, we confirmed the relationship between number of lesions and age in both the symptomatic group and the asymptomatic group for all types except type 1. Kattapong et al (11) reported that the mean number of lesions per patient increased at a rate of one lesion per decade. We did not confirm such a distribution in our study, because the number of lesions seemed to increase more rapidly after the age of 50 years. Different mechanisms might account for the increased number of lesions and the descriptions of de novo lesions in the literature: "New" lesions might correspond to growth of very small and initially undetectable congenital lesions (owing to repeated hemorrhage or proliferation of capillaries), or they might be true new lesions that were not present at birth but appeared later in life (3,9,11,18,32,33,34).
The results of our MR imaging investigation in a series of 51 non-Hispanic families confirmed that the familial form of cavernous angioma is characterized by a predominance of multiple lesions, an increase in number of lesions with increasing subject age, and a relationship between symptoms and the presence of high-signal-intensity lesions on MR images. Furthermore, our results emphasize that GRE sequences are more sensitive than SE sequences for help in evaluating the number of lesions and determining patient status in the setting of familial cavernous angioma.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, L.B., P.L., E.T.L., C.L., J.P.H.; study concepts, L.B., P.L., E.T.L.; study design, L.B., P.L.; definition of intellectual content, L.B., P.L., E.T.L.; literature research, L.B., P.L.; clinical studies, P.L., S.L.; data acquisition, L.B., P.L., C.L.; data analysis, L.B., P.L.; statistical analysis, L.B., P.L.; manuscript preparation and editing, L.B., P.L.; manuscript review, L.B., P.L., E.T.L., C.L., J.P.H. L.B. and P.L. contributed equally to this work.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J.M. de Souza, R.C. Domingues, L.C.H. Cruz Jr., F.S. Domingues, T. Iasbeck, and E.L. Gasparetto Susceptibility-Weighted Imaging for the Evaluation of Patients with Familial Cerebral Cavernous Malformations: A Comparison with T2-Weighted Fast Spin-Echo and Gradient-Echo Sequences AJNR Am. J. Neuroradiol., January 1, 2008; 29(1): 154 - 158. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Koike, N. Aida, M. Hata, K. Fujita, Y. Ozawa, and T. Inoue Asymptomatic Radiation-induced Telangiectasia in Children after Cranial Irradiation: Frequency, Latency, and Dose Relation Radiology, January 1, 2004; 230(1): 93 - 99. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Labauge, L. Brunereau, S. Laberge, and J.-P. Houtteville Prospective follow-up of 33 asymptomatic patients with familial cerebral cavernous malformations Neurology, November 27, 2001; 57(10): 1825 - 1828. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Rosenow and H. Luders Presurgical evaluation of epilepsy Brain, September 1, 2001; 124(9): 1683 - 1700. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
