| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 1999;30:1632-1636.)
© 1999 American Heart Association, Inc.
Original Contributions |
From the Departments of Neurosurgery, Division of Vascular Neurosurgery (J.E.C., R.J.T.), and Pathology (G.M.H.), The Johns Hopkins Medical Institutions, Baltimore, Md.
Correspondence to Rafael J. Tamargo, MD, Department of Neurosurgery, Johns Hopkins Hospital, Meyer 7-113, 600 N Wolfe St, Baltimore, MD 21287-7713.
| Abstract |
|---|
|
|
|---|
MethodsThe results of an autopsy series at the Johns Hopkins Hospital of 25 confirmed Marfan syndrome patients from 1939 to the present were reviewed retrospectively. The prevalence of intracranial aneurysms in this Marfan syndrome autopsy series was compared with that in the autopsy population at this institution and with that in the general autopsy population as reported in the literature. In addition, the prevalence of Marfan syndrome in a recent neurosurgical series of 710 consecutive aneurysm cases (19901998) was determined.
ResultsOf the 25 autopsy cases, only 1 had evidence on autopsy of an unruptured, 2-mm aneurysmal dilatation at the anterior communicating artery complex. Three autopsy patients suffered intracranial hemorrhages but had negative angiography and postmortem examinations for intracranial aneurysms. The remaining 21 patients had negative autopsies for intracranial hemorrhages or intracranial aneurysms. The neurosurgical series of 710 patients treated for intracranial aneurysms did not include any patient with Marfan syndrome.
ConclusionsThe prevalence of 1 patient of 25 with an intracranial aneurysm is not statistically different from the 1.3% prevalence of intracranial aneurysms in the autopsy population at this institution (P=0.24) or from the 2.0% prevalence of intracranial aneurysms in the general autopsy population (P=0.31). We therefore conclude that there exists no evidence that Marfan syndrome is associated with an increased prevalence of intracranial aneurysms.
Key Words: cerebral aneurysm Marfan syndrome prevalence
| Introduction |
|---|
|
|
|---|
The diagnosis of Marfan syndrome is based on specific clinical and, most recently, genetic criteria. The diagnostic criteria have been refined since the condition was first described in a 5-year-old girl in 1896 by Antoine Bernard-Jean Marfan, a French professor of pediatrics.7 8 After Marfan's original description, other manifestations of the syndrome were recognized and subsequently codifed in 1955 by McKusick at Johns Hopkins.9 10 In 1986, during the 7th International Congress of Human Genetics in Berlin, Germany, a panel of experts agreed on the specific clinical criteria for the diagnosis of Marfan syndrome.2 These criteria have been widely adopted and are known as the "Berlin nosology." Recently, De Paepe and colleagues8 proposed more stringent clinical criteria and recommended that genetic information be considered for the diagnosis.
Over the past 30 years, an association between Marfan syndrome and intracranial aneurysms has been repeatedly proposed but also vigorously challenged. This association has been based on 10 clinical reports of Marfan patients with intracranial aneurysms,11 12 13 14 15 16 17 18 19 20 1 pathology report describing the early development of an intracranial aneurysm in a Marfan patient,21 and 1 autopsy series of 7 Marfan patients, 2 of whom had intracranial aneurysms.22 By contrast, no association between Marfan syndrome and intracranial aneurysms has been found in 5 clinical series of Marfan patients.23 24 25 26 27 Given the prognostic concerns associated with the presence of an intracranial aneurysm, the proposed increased prevalence of these lesions in Marfan syndrome has raised the question of whether Marfan patients should be evaluated to rule out an intracranial aneurysm.
Our institution has had a dedicated program for patients with Marfan syndrome since the 1950s.9 24 Under the direction of McKusick and colleagues, 300 new or established Marfan patients are seen annually at the Johns Hopkins Hospital,24 and more than 1400 Marfan patients have been evaluated and treated in this program since its inception. As such, there is a large institutional experience with the treatment of complications of Marfan syndrome. We have drawn on our pathological and clinical experience with Marfan syndrome to analyze the purported association between Marfan syndrome and intracranial aneurysms. In this study we report the prevalence of intracranial aneurysms in our autopsy series of 25 Marfan patients (19391996), and in addition we report the prevalence of Marfan syndrome in our recent neurosurgical series of 710 consecutive intracranial aneurysm cases (19901998).
| Subjects and Methods |
|---|
|
|
|---|
Autopsy Series
The pathology autopsy database of the Johns Hopkins Medical
Institutions consists of more than 50 000 reports of all autopsies
performed at the institution during 18891998. This database was
searched to identify potential Marfan cases with the key words
Marfan, ectopia lentis, and
arachnodactyly. The medical records and autopsy results
of all potential Marfan patients were reviewed. Epidemiological
information recorded included the patients' sex, age of death,
race, and the presence of a family history of Marfan syndrome.
Characteristic manifestations of Marfan syndrome, specifically
skeletal, ocular, cardiovascular, pulmonary,
integumentary, and central nervous system manifestations, were
recorded. The Berlin nosology and the De Paepe criteria were then
applied to the cases to select those that fulfilled the
diagnostic criteria for Marfan syndrome by either system.
Intracranial autopsy results from Marfan patients and
histological specimens of reported intracranial
aneurysms were reviewed.
Neurosurgical Series
The Johns Hopkins Hospital neurosurgical series of patients
treated for intracranial aneurysms consists of all patients
admitted to the Johns Hopkins Hospital, Baltimore, Md, and the Johns
Hopkins Bayview Medical Center, Baltimore, Md, with an angiographically
or autopsy-proven diagnosis of an intracranial aneurysm. The
study period was from January 1990 to December 1998. Medical
records of the patients were reviewed to identify either
characteristic manifestations of Marfan syndrome as described above or
a previous diagnosis of Marfan syndrome, according to the Berlin
nosology or the De Paepe criteria.2 8
Statistical Analysis
The epidemiological prevalence of intracranial aneurysms
in the population in general28 and in the Hopkins series
in particular29 was compared with the prevalence of
intracranial aneurysms in our Marfan autopsy series. The
binomial probability test was applied with the use of Stata statistical
software for the Macintosh (version 4.0, Stata Corporation). The power
of the study was determined with the same statistical software.
| Results |
|---|
|
|
|---|
|
|
Of the 25 patients in the autopsy series, only 1 patient had evidence
on autopsy of an intracranial aneurysm. The aneurysm
was an asymptomatic, unruptured, 2-mm dilatation at the
anterior communicating artery complex
(Figure
). This patient was a black man
who had a positive family history of Marfan syndrome and died in 1955
of an aortic dissection at the age of 30 years. Other Marfan syndrome
manifestations exhibited by the patient included arachnodactyly,
hyperextendable joints, and dilatation of the ascending aorta. Cystic
medial necrosis was observed on histological
examination of the aorta. Microscopic histological
examination of the intracranial aneurysm as it originated from
the parent vessel revealed fragmentation and thinning of the medial
smooth muscle layer at the aneurysm neck and absence of the
media at the aneurysm body and dome with replacement of this
region with thin, fibrous connective tissue.
|
Although an additional 3 patients in the autopsy series experienced intracranial hemorrhages, by autopsy none were aneurysmal subarachnoid hemorrhages. One male patient died at the age of 50 years as a result of a cerebellar hematoma that was a complication from surgical resection of an acoustic schwannoma. Another male patient, who died at the age of 21 years from complications of aortic valve and aortic graft surgery, had previously experienced a right frontal hemorrhage while taking warfarin. Cerebral angiography had ruled out an intracranial aneurysm. A female patient died at the age of 37 years from a posterior fossa hemorrhage while taking warfarin. She also had developed a T11-L5 intradural hematoma 4 days before her death. The source of the intracranial hemorrhage could not be identified, but an aneurysm was ruled out at autopsy.
Marfan Syndrome Autopsy Series: Secondary Analysis Using
the De Paepe Criteria
As described above, De Paepe and colleagues8 recently
proposed more stringent clinical criteria for the diagnosis of Marfan
syndrome. With the use of these modified criteria, 4 of the 25 patients
in the autopsy series would have been excluded. One of these patients
would have been the patient with the anterior communicating artery
aneurysm. Therefore, using the modified criteria for the
diagnosis of Marfan syndrome would have resulted in no cases of
intracranial aneurysms in our autopsy series and strengthened
the conclusions of this report.
Statistical Analysis
In the literature, the best estimate for the prevalence of
intracranial aneurysms in the general population is
approximately 2%, which is based on the combined results of
Bannerman's review of 8 autopsy series of 51 360
patients.28 At our institution, we have previously
reported the prevalence of intracranial aneurysms in our
autopsy population as 1.3%.29 In our Marfan autopsy
series, the prevalence of intracranial aneurysms was 1 of 25
patients. There is no statistically significant difference between the
prevalence of intracranial aneurysms in our Marfan autopsy
sample and that reported either in the general autopsy population
(P=0.31, binomial probability test) or in our institution
(P=0.24, binomial probability test).
The prevalence of intracranial aneurysms was 0 of 21 Marfan patients when the more specific criteria of De Paepe were used to determine a Marfan diagnosis. Again, a statistically significant difference does not exist between the prevalence of intracranial aneurysms in this Marfan autopsy series and that reported in the general autopsy population (P=0.65, binomial probability test) or in our institution (P=0.76, binomial probability test).
The size of this autopsy series was probably not large enough to detect small differences, either increases or decreases, in the prevalence of intracranial aneurysms in the Marfan autopsy series compared with that in the general population. This study did, however, possess sufficient power (0.99) to determine whether the alternative prevalence of intracranial aneurysms in Marfan patients (29%) that was reported by Schievink et al22 is the true prevalence of intracranial aneurysms in Marfan patients.
Neurosurgical Series
None of the 710 patients treated for intracranial
aneurysms at the Johns Hopkins Hospital in the past 9 years had
Marfan syndrome.
| Discussion |
|---|
|
|
|---|
By contrast, 5 clinical studies have failed to identify an association between Marfan syndrome and intracranial aneurysms.23 24 25 26 27 Murdoch and colleagues23 from Johns Hopkins reported the cause of death in 56 of 74 deceased Marfan patients and did not identify an intracranial aneurysm or subarachnoid hemorrhage in any of these patients. In addition, a symptomatic intracranial aneurysm was not reported in any patient. Pyeritz and McKusick,24 also from Johns Hopkins, reported the clinical manifestations of Marfan syndrome in 50 consecutive patients and did not identify any patients with a symptomatic intracranial aneurysm. Van den Berg and colleagues27 specifically addressed the association between Marfan syndrome and intracranial aneurysms in a series of 129 Marfan patients with 581 observation years and did not find evidence for such an association. In addition, a diagnosis of Marfan syndrome was not established in any of the 826 patients admitted to their institution's neurology and neurosurgery services with the diagnosis of a subarachnoid hemorrhage or intracranial aneurysm.27
Our institution has had a dedicated program for patients with Marfan syndrome since the 1950s in which more than 1400 Marfan patients have been evaluated and treated since its inception.9 24 We therefore have a large institutional experience with the treatment of complications of Marfan syndrome and have noticed a paucity of Marfan patients with intracranial aneurysms. Two clinical studies of these Marfan patients by Murdoch et al23 and Pyreitz and McKusick24 failed to identify any Marfan patient with a symptomatic intracranial aneurysm. Our autopsy series of 25 Marfan cases has only 1 case with an intracranial aneurysm, which is consistent with the expected prevalence of intracranial aneurysms in the general population and in our autopsy population. Similarly, our neurosurgical series of 710 aneurysm cases does not include any patient with Marfan syndrome. We therefore conclude that there exists no evidence that Marfan syndrome patients are at an increased risk for the development of intracranial aneurysms.
It is of interest that the 1 Marfan syndrome patient in our study who had an intracranial aneurysm had a brother who was also affected by Marfan syndrome. The brother underwent an autopsy as well and was not found to have an intracranial aneurysm. This patient died from a dissection of the ascending aorta and was aged 35 years at his time of death, 5 years older than the age at which his sibling who possessed the aneurysm died. Since this patient did not have an intracranial aneurysm, one could argue that his brother's aneurysm was probably sporadic and not a phenotypic manifestation of the specific FBN1 mutation carried by the brothers.
In regard to the 2 patients who had intracranial hemorrhages while taking warfarin, the evidence strongly suggests that neither of these hemorrhages was caused by a ruptured intracranial aneurysm. The first patient presented with an intraparenchymal hemorrhage in the frontal lobe. There was no subarachnoid hemorrhage. This would be a rare presentation for a ruptured saccular aneurysm but is a well-described complication of warfarin anticoagulation therapy. In addition, a 4-vessel angiogram did not reveal an aneurysm, and an intracranial aneurysm was not discovered during a detailed examination of the intracranial vasculature at autopsy. The second patient initially presented with a T11-L5 subdural hematoma. Neither an aneurysm nor an arteriovenous malformation was identified by angiography. This patient then underwent surgical decompression and evacuation of the spinal subdural hematoma. The source of the hemorrhage was not identified during surgery. The patient did poorly postoperatively, and a CT scan revealed further bleeding into the posterior fossa. The patient died and had a thorough intracranial examination at autopsy. Although the vertebral and basilar arteries were described as "ectatic and tortuous," a saccular aneurysm was not observed in any part of the vasculature. Even though a small or compressed aneurysm may not be detected by angiography, the clinical and postmortem findings in these cases strongly suggest that aneurysmal ruptures were not the cause of the hemorrhage in either case. Furthermore, it would be unlikely that even a small aneurysm would not be identified at autopsy, especially since an aneurysmal source of hemorrhage was specifically sought in each case.
The absence of aneurysm patients with Marfan syndrome in our neurosurgical series of 710 patients and the series of 826 patients reported by van den Berg and colleagues27 is probably explained by the early age of death of most Marfan patients. In our autopsy series the mean and median ages at time of death were 39 and 38 years, respectively. By contrast, aneurysmal subarachnoid hemorrhage peaks in the middle of the sixth decade, with typical mean and median ages at the time of presentation of 56 and 59 years, respectively.32 Therefore, the few Marfan patients with preaneurysmal defects probably die from other causes before they develop a mature aneurysm that ruptures.
We conclude that there exists no evidence that Marfan syndrome is associated with an increased prevalence of intracranial aneurysms. This conclusion implies that the already extensive and costly evaluations pursued in these patients do not need to include serial intracranial radiographic studies intended to rule out these lesions.
| Acknowledgments |
|---|
Received March 26, 1999; revision received April 28, 1999; accepted April 28, 1999.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S. Wills, A. Ronkainen, M. van der Voet, H. Kuivaniemi, K. Helin, E. Leinonen, J. Frosen, M. Niemela, J. Jaaskelainen, J. Hernesniemi, et al. Familial Intracranial Aneurysms: An Analysis of 346 Multiplex Finnish Families Stroke, June 1, 2003; 34(6): 1370 - 1374. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Hofer, M. Hermans, N. Kubassek, M. Sitzer, H. Funke, F. Stogbauer, V. Ivaskevicius, J. Oldenburg, J. Burtscher, U. Knopp, et al. Elastin Polymorphism Haplotype and Intracranial Aneurysms Are Not Associated in Central Europe Stroke, May 1, 2003; 34(5): 1207 - 1211. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. J. Wityk, C. Zanferrari, and S. Oppenheimer Neurovascular Complications of Marfan Syndrome: A Retrospective, Hospital-Based Study Stroke, March 1, 2002; 33(3): 680 - 684. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. E. Conway, G. M. Hutchins, and R. J. Tamargo Lack of Evidence for an Association Between Neurofibromatosis Type I and Intracranial Aneurysms: Autopsy Study and Review of the Literature Stroke, November 1, 2001; 32(11): 2481 - 2485. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. I. Schievink, R. J. Tamargo, J. E. Conway, and G. M. Hutchins Marfan Syndrome and Intracranial Aneurysms • Response Stroke, December 1, 1999; 30 (12): 2759 - 2768. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1999 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |