How to manage the patient with a family history of aneurysmal subarachnoid haemorrhage
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
88 PRACTICAL NEUROLOGY
NEUROLOGICAL DILEMMAS
How to manage
the patient with
a family history
of aneurysmal
subarachnoid
haemorrhage
P. M. White
Consultant Neuroradiologist
Department of Clinical Neurosciences, University
of Edinburgh, Western General Hospital,
Crewe Road, Edinburgh, EH4 2XU, UK; E-mail:
pmw@skull.dcn.ed.ac.uk
Practical Neurology, 2004, 4, 88–103
© 2004 Blackwell Publishing Ltd
APRIL 2004 89
INTRODUCTION WHAT IS THE FREQUENCY OF INTRA-
‘The truth is rarely pure and never simple’. So CRANIAL ANEURYSMS IN THE GENERAL
said Oscar Wilde, and this is certainly the case POPULATION?
for individuals with an unruptured intracranial What is anyone’s risk of harbouring an aneu-
aneurysm or a family history of aneurysmal rysm, after all incidental aneurysms are com-
subarachnoid haemorrhage (SAH). Before you monly found at autopsy? Rinkel et al. identified
can counsel these people, it is first necessary for autopsy studies where the prevalence of unrup-
the clinician to have a grasp of this complex tured aneurysms ranged from 0.4% to 3.6% for
area because once you have told the patient retro and prospective studies, respectively, and
something it cannot be untold. You might sub- in retro and prospective angiography studies
sequently expand on the information imparted, from 3.7% to 6%, respectively (Rinkel et al.
but you cannot remove it – correct or incorrect 1998). More recent studies are more in line with
– from the patient’s mind (Fuller 2001). In this the higher figures from the prospective studies ‘The truth is rarely
article I will summarize the relevant knowledge, (Kojima et al. 1998; Ronkainen et al. 1998). In a
and offer some advice, on management in this higher-risk group of patients, because they had pure and never
rather fraught area of medicine. already ruptured an aneurysm, 20–25% had at
least one unruptured aneurysm on angiography simple’
(Lozano & Leblanc 1987). So approximately one
in 20 of the population aged over 30 harbour an
unruptured aneurysm. However, it is important
to remember that, with an aneurysmal SAH in-
cidence of about 8 per 100 000 per annum
(Wardlaw & White 2000), clearly only a very
small proportion of these aneurysms actually
rupture (about 0.1–0.2% pa).
ARE SPECIFIC GROUPS AT PARTICU-
LARLY HIGH RISK OF INTRACRANIAL
ANEURYSMS?
The risk factors for SAH and for having an
unruptured intracranial aneurysm are very
similar (Table 1). In addition, several genetic
conditions are associated with intracranial
aneurysms, although these represent less than
10% of all cases:
• 10–15% of patients with adult polycystic
kidney disease (ADPKD) (Rinkel et al. 1998),
particularly if there is a family history of
There are several stages of assessment and aneurysms and/or SAH (Hughes et al. 1996;
management to consider: Ruggieri et al. 1994);
• What is an individual’s risk of harbouring an • type IV Ehlers–Danlos syndrome (Schievink
aneurysm? et al. 1990);
• How best to detect an aneurysm without • possibly pseudoxanthoma elasticum (Muny-
exposing the patient to unnecessary stress or er & Margulis 1981), although this has been
risk? refuted (van den Berg et al. 1999);
• If an asymptomatic aneurysm is found, what • hereditary haemorrhagic telangiectsaia
is the risk of rupture? (Roman et al. 1978);
• What treatment, if any, should be offered and • neurofibromatosis type I (Mulvihill et al.
what are the risks involved? 1990; Morooka & Waga 1991);
Most importantly, the risk at each stage must be • alpha1-antitrypsin deficiency (Schievink et
weighed against the risk of not doing anything. al. 1996).
The pros and cons of screening for aneurysms Marfan’s syndrome was thought to be associ-
will also be briefly considered to place some of ated with aneurysms but a detailed study of 135
this information in context. patients (mean age 21 years) found no evidence
© 2004 Blackwell Publishing Ltd
90 PRACTICAL NEUROLOGY
Table 1 Risk factors for intracranial aneurysms, and for subarachnoid haemorrhage
RISK OF PREVALENCE OF RELATIVE
RISK FACTOR ANEURYSM SAH ANEURYSMS RISK REFERENCE
Female gender + + 1.6 Linn et al. 1996
Current smoking + + 1.9 Teunissen et al. 1996
Hypertension – + 2.8 Teunissen et al. 1996
Alcohol (heavy consumption) – + 4.7 Teunissen et al. 1996
Oral contraceptive pill ? + 1.5 (low dose) Johnston et al. 1998
1.9 (high dose)
Heavy physical exertion – + 2.7 (odds ratio) Anderson et al. 1993
Atherosclerosis ? + 2.3 Rinkel et al. 1998
Ischaemic heart disease in woman + + (4.3) Uehara et al. 1998
Cholesterol > 6.3 mmol/L ? + 10.2 (odds ratio) Adamson et al. 1994
Autosomal dominant polycystic + + 10–15% 4.4 Rinkel et al. 1998
kidney disease
Familial (2 or more first- or second- + + 9.8% 4.0 Rinkel et al. 1998
degree relatives)
First-degree relatives in families + + 4.5% 1.8 see Table 2
with one affected member
of a relationship (van den Berg et al. 1996), (Table 2). The relative risk for SAH in relatives
although the patients may not have been old compared with the background population was
enough for the aneurysms to have developed. 2.9–6.6 for first-degree relatives and 1.6–2.1 for
In addition, aneurysmal SAH may affect sev- second degree. Whilst these relative risks appear
eral members of a family without any specific large, it is important to remember the small
genetic ‘disease’. Familial SAH has been defined absolute number of relatives affected: only 156/
Approximately inconsistently, so here I mean a family in which 21054 (0.74%) first-degree relatives were them-
two or more close blood relatives (first or second selves affected by SAH. In our Scottish study (in
one in 20 of the degree) have a history of aneurysmal SAH with- press) the risk of SAH among family members
out any other known heritable disease (first- over the decade after the index case had their
population aged degree relatives = parents, siblings, children; SAH was 1.2% for first-degree and 0.5% for
second degree = grandparents, grandchildren, second-degree relatives (some 12 and 5 times,
over 30 harbour aunts and uncles, nieces and nephews; third respectively, the background population SAH
degree = cousins, great grandparents, great risk) (Wardlaw et al., in press).
an unruptured grandchildren) Familial intracranial aneurysms may have
It has been suggested that many cases of distinguishing biological features, including:
aneurysm ‘familial intracranial aneurysms’ might simply rupture at a younger age (most frequently in the
represent accidental aggregation (ter Berg et al. fifth decade compared to the sixth decade for
1992); on the basis of chance alone, each SAH sporadic SAH) (Bromberg et al. 1995); worse
patient has a 5.6% possibility of having a first- to clinical outcome; and an increased prevalence
third-degree relative also affected by SAH (ter of middle cerebral artery aneurysms (Kojima
Berg et al. 1992), and the proportion of SAH et al. 1998; Bromberg et al. 1995). There may
patients with third-degree relatives who had be a younger age of rupture in subsequent
had an SAH is the same as the proportion with generations, implying ‘anticipation’ (Interna-
SAH in the control population (De Braekeleer tional Subarachnoid Aneurysm Trial 2002) but
et al. 1996). familial aneurysms may also have a predilection
Since 1987, 10 studies have examined the towards rupture in the same decade in individu-
prevalence of unruptured aneurysms and/or als of the same family, particularly in siblings
SAH amongst relatives of patients with SAH (Leblanc 1997).
© 2004 Blackwell Publishing LtdTable 2 Summary of population based studies of familial subarachnoid haemorrhage*
NUMBER NUMBER OF RELATIVES SURVEYED NUMBER OF RELATIVES WITH SAH % 1°
OF INDEX WITH
STUDY SUBJECTS LOCATION 1° 2° 3° 1° 2° 3° SAH COMMENT
Norrgard 1987 485 Umea, Sweden 1352 / / 22 / / 4.7 sibs only surveyed –
(sibs only) average six per index case
Wang 1995 149/171† Washington, USA N/S N/S N/S 18 16 / 11.4 OR for SAH in 1°
relative = 1.8, 2° = 2.4
Schievink 1995 76/81† Rochester, USA 608 N/S N/S 11 5 / 1.8 RR of SAH in 1° relative = 4.1
2° = 1.6
Bromberg 1995 163 Utrecht, Netherlands 1290 3588 N/S 10 + 7‡ 4 + 12‡ / 1% RR of SAH in 1° relative = 6.6
definite, and 2.7 possible SAH
de Braekeleer 1996 533 (+ 1599 Quebec, Canada N/S N/S N/S 48 51 77 / RR of SAH in 1° relative = 4.7,
controls) 2° = 2.1, 3° = 1.1 compared to
general pop
Ronkainen§ 1997 91 Kuopio, Finland 716 relatives in total → 76 37 relatives 10.6¶ Risk of unruptured aneurysm in 1°
(1°, 2° and 3°) (2° and 3°) in total → or 2°4x greater than general pop.
Raaymakers 1999 160 Utrecht, Netherlands 626 / / 4 / / 0.6 3.7% of 1° had an unruptured
aneurysm
Gaist 2000 6175 Denmark (1977–95) 14781 / / 19 / / 0.1 RR 2.9, 4.5 for 77% cases admitted
Davie Cooper Study 453 Scotland 3023 5668 87 65 2.9 RR 2.3–2.4. Ten year prospective
risk 1.2% in 1°, 0.5% in 2°
Okamoto 2003 200 Japan N/S / / 29 / / N/S OR for SAH in 1° relative = 4.0
*Wardlaw in press.
†
number surveyed/total sample available.
‡
definite + possible SAH.
§
relatives with SAH or aneurysm.
¶
% of all relatives not just first degree.
OR, odds ratio; RR, relative risk; N/S = not stated
/ = no data.
APRIL 2004
© 2004 Blackwell Publishing Ltd
9192 PRACTICAL NEUROLOGY
WHICH RELATIVES ARE MOST risk was 0.07% (Cloft et al. 1999). Non-invasive
FREQUENTLY AFFECTED BY SAH? tests, such as magnetic resonance angiography
In families with more than one person with (MRA), dynamic spiral CT Angiography (CTA)
SAH, the most frequent relationship is index and transcranial Doppler sonography (TCDS)
patient to sibling, followed by index patient to (Fig. 2) are attractive for detecting asympto-
parent (Wardlaw & White 2000). However, a matic intracranial aneurysms but they have to
parent was affected in only 29% so in less than be extremely accurate to replace IADSA. So how
one-third of affected families was there a clear accurate are they?
warning of the potential for SAH from a previ- Analysis of accuracy per patient rather than
ous generation. Interestingly, one study found per aneurysm is of more clinical relevance to
a much stronger familial link to maternal SAH ‘screening’ an individual for aneurysms. Most
than to paternal or sibling SAH (Okamoto et al. studies reported apparently excellent results but
2003) and in another study, the slightly more were in patients with a recent SAH so the aneu-
commonly affected first-degree relatives were rysm prevalence was high (Table 4) (White et al.
also parents (38%) rather than siblings (33%) 2000). As a result the sensitivities, specificities
(Wardlaw, in press). Only our Scottish study has and predictive values must be interpreted with
prospectively investigated the interaction be- caution, particularly if one is extrapolating to
tween the type of relationship and involvement using the non-invasive test in a low aneurysm
of more than one family member (with SAH) prevalence population where far more positive
on the risk of SAH to the other family members test results will be false positives. Crucially, the
(Table 3). Here, there was a hierarchy of ascend- results for non-invasive imaging methods are
ing risk from a second-degree relative and no significantly poorer for aneurysms < 5 mm in
other family members affected (0.3% cumula- size, which constitute as many as one-third of
tive lifetime SAH risk) to the highest risk in a aneurysms in asymptomatic patients (Kojima
member of a family with at least two first-degree et al. 1998) (Fig. 3). Sensitivities are in the order
relatives affected (7% cumulative lifetime SAH of 35%, 57% and 35% for aneurysms < 5 mm
risk) (Wardlaw in press). diameter for MRA, CTA and TCDS, respectively
(White et al. 2001; White et al. 2001) ICA an-
HOW BEST TO DETECT AN ANEURYSM? eurysms are even harder to detect (White et al.
The gold standard is an intra-arterial digital 2000, 2001). Ultrasound has the advantage of
subtraction angiogram (IADSA) with selective the lowest capital cost, and portability, and with
cerebral arterial injections and multiple projec- contrast agents and 3-D imaging the accuracy
tions, particularly using the 3D DSA technique may improve towards the level of CTA/MRA
(Fig. 1). However, IADSA is invasive, costly, and (Wardlaw & White 2000). Unfortunately, about
not without risk of a permanent neurological 10% of patients will not have an adequate bone
complication. In patients with SAH, suspected window and the technique is operator depend-
aneurysm or arteriovenous malformation, this ent.
Table 3 Relationship between the family history of an index subarachnoid haemorrhage (SAH) case and the
risk of SAH to other family members
NUMBER OF SAH CUMULATIVE SAH RISK
FAMILY HISTORY OF SAH EVENTS OCCURRING (BIRTH TO 70 YEARS)
One second-degree relative with SAH 6 0.3%
One first-degree relative 7 0.8%
No first but at least two second-degree relatives 1 1%
One first and at least one second-degree relative 2 2%
At least two first-degree relatives 4 7%
Analysis based on 3213 relatives of 148 index cases.
Risk prior to 30 years of age is extremely low.
For comparison, the background population risk of SAH is ~0.06% per decade.
© 2004 Blackwell Publishing LtdAPRIL 2004 93
*
(a) (b)
Figure 1 (a) Imaging of a large anterior communicating artery aneurysm (arrow) with standard 2D intra-
arterial digital subtraction angiography (IADSA). By comparison, the 3D IADSA image (b) demonstrates
the detailed anatomy of the aneurysm, particularly the neck, far more clearly. Note additional right middle
carotid artery aneurysm*.
(a) (b) (c)
Figure 2 (a) Right terminal carotid bifurcation aneurysm (arrow) demonstrated on 2D intra-arterial digital subtraction angiography (b) MRA
– Maximum Intensity Projection (MIP) and (c) CTA MIP images. This was a ruptured aneurysm and the associated haematoma results in marked T1
breakthrough effect on the MRA MIP image – highlighting the location of the aneurysm. This effect (or the blood distribution on CT) is one reason
why diagnostic accuracy studies on SAH patients cannot be readily extrapolated into the screening context.
© 2004 Blackwell Publishing Ltd94 PRACTICAL NEUROLOGY
More recently there have been further studies more reliably, than less experienced observers
of non-invasive imaging tests vs. IADSA (White (Okahara et al. 2002; White et al. 2003).
et al. 2001a; White et al. 2001b; Okahara et al.
2002; Chung et al. 1999; Jager et al. 2000; Me- IF AN ASYMPTOMATIC ANEURYSM IS
tens et al. 2000; Suzuki et al. 2003; Pedersen et al. FOUND WHAT IS ITS RISK OF RUPTURE?
2001) and a CTA meta-analysis was published in The best available data on rupture risk are sum-
2002 (Chappell et al. 2003). These later studies marised in Table 5. A systematic review by Rin-
have in general been of better quality although kel et al. found an overall annual rupture rate of
there was still a marked preponderance of 1.9% (Rinkel et al. 1998). Aneurysms were twice
patients with SAH rather than asymptomatic as likely to rupture in women than in men and
unruptured aneurysms. The sensitivities and the risk of rupture increased with age. Symp-
specificities were similar to those in the earlier tomatic aneurysms were significantly more
systematic review. Two small studies suggest likely to rupture than incidental or additional
that dynamic contrast MRA (rapid MRA dur- (asymptomatic) aneurysms (6.5% vs. 0.8% vs.
ing an intravenous injection of gadolinium) 1.4%, respectively). Posterior circulation and
may be more accurate than time of flight MRA large (> 10 mm) aneurysms were significantly
(Metens et al. 2000; Suzuki et al. 2003), but they more likely to rupture. In a logistic regression
do suffer from the methodological problems of model, the only factor significantly related to
the early time-of-flight MRA studies and might aneurysm rupture was the size of the aneurysm
be over-optimistic. Two studies, which looked at – 7 mm or larger aneurysms had a relative risk
the effect of observer experience on aneurysm of rupture of 2.24 compared with smaller aneu-
detection found, not surprisingly, that experts rysms (Juvela et al. 1993). Virtually all patients
detected substantially more aneurysms, and in this study had had a previous SAH. In the
(a) (b)
(d)
(c)
Figure 3 (a) Left terminal carotid bifurcation aneurysm (arrow) on 2D intra-arterial digital
subtraction angiography. (b) Although the terminal carotid is irregular in appearance, the
very small aneurysm cannot be readily appreciated on the MRA MIP image and even the
base image (c) does not clearly demonstrate the aneurysm – illustrating the problems
of detecting very small aneurysms with non-invasive imaging methods. However, the
aneurysm was detected on transcranial Doppler sonography – arrowed ‘An’ (d).
© 2004 Blackwell Publishing LtdAPRIL 2004 95
Table 4 Sensitivity and specificity of non-invasive tests for intracranial aneurysms (White
et al. 2000)
MRA (PP)* CTA (PP)† TCDS (PP)‡ MRA (PA)§ CTA (PA)¶ TCDS (PA)**
Sensitivity 87 92 – 87 90 82
Specificity 92 94 – 95 86 95
*926 subjects in 20 studies; †677 subjects in 16 studies; ‡No studies reported data
per patient; §1596 aneurysms in 926 subjects; ¶1582 aneurysms in 677 subjects; **97
aneurysms in 162 subjects.
PP, identification of a patient as having or not having an aneurysm; PA, identification of an
individual aneurysm correctly; MRA, magnetic resonance angiography; CTA, computerized
tomography angiography; TCDS, transcranial Doppler sonography.
Table 5 Summary of data on annual risk of rupture of unruptured aneurysms
INTERNATIONAL STUDY OF INTERNATIONAL STUDY OF UNRUPTURED
UNRUPTUREDINTRACRANIAL ANEURYSMS INTRACRANIAL ANEURYSMS
RINKEL ET AL. 1998* & INVESTIGATORS 1998 (RETROSPECTIVE) & INVESTIGATORS 2003 (PROSPECTIVE)
No. of subjects 95 1449 1692 (unoperated cohort)
No. of aneurysms – 193 686
Duration of follow-up 3907
(patient years) (mean of mean follow-ups 5.5, 12023 7587
range 2.1–13.7 years) (mean 8.3 years) (mean 4.5 years)
Number ruptured 75 32 51
Overall rupture rate (% pa) 1.9 0.27 0.67
Rupture rate
< 10 mm 0.7 0.05 < 7 mm all sites no prior SAH ~0.07
> 10 mm 4.0 0.5 < 7 mm all sites prior SAH ~0.41
> 7 mm all sites all patients ~0.1†
Cumulative aneurysm 10% per decade 0.5–5% per decade 0.7–1.5% per decade†
rupture rate
(from Juvela et al. 1993)
Incidental aneurysm 6.5 data not extractable data not extractable
rupture rate
Asymptomatic aneurysm 0.8 but 7/8 ruptures in data not extractable < 7 mm ~0.07
rupture rate aneurysms > 10 mm > 7 mm ~2.1
Asymptomatic aneurysm 1.4 data not extractable ~0.36 all sizes
rupture rate
Posterior circulation aneurysm 4.4 data not extractable 0.5 (< 7 mm) – 10% (> 24 mm)
rupture rate
Age
20–39 0 data not extractable data not extractable
40–59 3.5
60–79 5.7
*Additional data to that published in the systematic review were kindly supplied by Dr Gabriel Rinkel to allow calcuation of duration of follow up.
The paper by Juvela et al. 1993 contributed 28% of the patients to the systematic review but almost half the patient-years of follow up.
†
No difference in rates between groups for larger aneurysms.
The rupture rate is much higher than the combined overall figure for the subgroups of posterior circulation aneurysms > 7 mm and for anterior
circulation aneurysms > 12 mm.
© 2004 Blackwell Publishing Ltd96 PRACTICAL NEUROLOGY
same study, repeat angiography during follow- those with prior SAH, the rupture risk was 0.3%
up showed, in patients with later SAH, that the pa for aneurysms < 7 mm with no significant
size of the aneurysms had increased from the difference from the no SAH group for larger
start of follow-up, whereas in those without aneurysms (International Study of Unruptured
later SAH the size did not change, and new an- Intracranial Aneurysms & Investigators 2003).
eurysms formed during the study in 19%, giving 49/51 ruptures occurred within five years of
an approximate rate of formation of 2.2% per diagnosis and only 2/41 ruptures in the no
year (Juvela et al. 1993). prior SAH group were of aneurysms < 7 mm
Rinkel et al. (pers. comm.) identified a total (both were posterior circulation aneurysms),
of 1145 years of patient follow-up in individu- compared with 7/10 in those with prior SAH.
als with asymptomatic aneurysms and no prior It is worth noting that aneurysms < 2 mm were
SAH (the group we are really looking at in this arbitrarily excluded from ISUIA, though these
article). 93% of the aneurysms were 10 mm or can rupture.
less in size. There were nine ruptures – and all
but one occurred in aneurysms 10 mm or larger. WHAT TREATMENT, IF ANY, SHOULD BE
Thus, the annual rupture rate for asymptomatic OFFERED AND WHAT ARE THE RISKS
aneurysms 10 mm or less was about 0.1%. INVOLVED?
The International Study of Unruptured Whether any treatment for an asymptomatic
Intracranial Aneurysms (ISUIA) is the largest aneurysm should be offered or not depends
ever study to follow up unruptured aneurysms critically on the balance between the long-term
(International Study of Unruptured Intracra- rupture risk (i.e. untreated natural history risk)
nial Aneurysms & Investigators 2003). Some vs. the immediate treatment risk. Both of these
patients had asymptomatic aneurysms with no risks in turn depend on the age, gender and life-
prior SAH, and others had previously sustained style factors of the individual patient.
SAH from another aneurysm. In the retrospec-
tive arm, there was a tiny rupture risk of 0.05% Treatment risks
per annum for small aneurysms (< 10 mm Aneurysms are treated by surgical clipping or
diameter) in patients who had not had an SAH by interventional neuroradiology – coiling.
previously, and of 0.5% per annum for large A systematic review of surgical treatment for
aneurysms and all aneurysms in patients who unruptured aneurysms identified 61 studies
had previously sustained an SAH (International between 1966 and 1996, but only eight were
Study of Unruptured Intracranial Aneurysms prospective and unfortunately, in virtually all,
& Investigators 1998). This 0.05% risk is in the neurosurgeon performing the operation
fact quite similar to the re-calculation of the was also the observer of outcome (Raaymakers
rupture risk of 0.1% for aneurysms < 10 mm et al. 1998). Permanent morbidity occurred in
from Rinkel′s systematic review (Rinkel et al. 10.9% of patients and the case fatality was 2.6%.
1998). In the patients that had not previously The lowest morbidity and mortality was found
had an SAH, only 1/12 aneurysmal ruptures with small anterior circulation aneurysms (case
occurred of an aneurysm < 10 mm in diameter fatality 0.8%, morbidity 1.9%), and the high-
(of concern to us in this article), compared with est with large posterior fossa aneurysms (case
17/20 patients in those who had previously had fatality 9.6%, morbidity 38%). The lack of in-
an SAH (International Study of Unruptured In- dependent outcome assessment and the effect
tracranial Aneurysms & Investigators 1998). of publication bias must have underestimated
The more robust prospective data in ISUIA, the surgical risk. The prospective arm of ISUIA
albeit based on rather small numbers, suggest a (Fig. 4) found the surgery-related case fatality
negligible rupture risk for anterior circulation at 1 years was 2.7% in patients with no prior
aneurysms < 7 mm in those without prior SAH SAH and 0.6% in patients who had previously
(none ruptured), 0.5% pa for anterior circula- suffered SAH. Morbidity was the same at 10%
tion aneurysms 7–12 mm, but substantially (International Study of Unruptured Intracra-
higher risk for larger anterior circulation aneu- nial Aneurysms & Investigators 2003). Age was
rysms (3–8% pa). The relative risk of rupture for the only independent predictor of outcome:
posterior circulation (including posterior com- surgery-related morbidity and mortality at
municating) aneurysms was 3–3.2 compared one year was about five times higher in those
with anterior circulation aneurysms of the > 64 years of age compared with patients <
same size. Of no direct concern to this article, in 45 years of age (International Study of Unrup-
© 2004 Blackwell Publishing LtdAPRIL 2004 97
35 tured Intracranial Aneurysms & Investigators
Clipping
Coiling 30
1998).
The effectiveness and risks of aneurysm
25 coiling in unruptured aneurysms are less cer-
tain because the technique is newer and still
Combined 20
death/disability developing. Guigliemi detachable coils (GDC)
rate 15
were introduced in 1991 and revolutionised the
10
endovascular treatment of intracranial aneu-
rysms (Guglielmi et al. 1991) (Fig. 5). System-
5
atic reviews of aneurysm coiling have identified
0 48 studies (all observational, mostly retrospec-
≤ 12 mm > 12 mm ≤ 12 mm > 12 mm
tive) including 1383 patients – but these were
Anterior circulation Posterior circulation
predominantly of ruptured aneurysms (Brilstra
Figure 4 Treatment risks for unruptured aneurysms (adapted from the et al. 1999). Permanent complications (death/
ISUIA data) disability) occurred in 3.7%. However, only
54% of aneurysms were completely occluded
after one procedure, although 88% were sub-
stantially (> 90%) coiled (Brilstra et al. 1999).
The re-bleeding rate from partially treated an-
eurysms is higher, and partial occlusion is well
recognised to be more common with coiling
than clipping. The technology of coils and coil-
ing assist devices is improving all the time, so the
partial occlusion rate is likely to decline.
There is only one large randomised trial
that has compared coiling with clipping – the
International Subarachnoid Aneurysm Trial
(ISAT) – but only in ruptured aneurysms (Inter-
There is a negligible risk of rupture of national Subarachnoid Aneurysm Trial 2002).
This demonstrated a relative risk reduction in
anterior circulation aneurysms < 7mm death/dependency for coiling over clipping of
23% at one year – an absolute risk reduction of
in those without prior subarachnoid 7%. Most ISAT patients had anterior circula-
tion aneurysms < 10 mm so the trial was not
haemorrhage completely representative of the generality of
aneurysms. However, anterior circulation an-
eurysms < 12 mm accounted for about 65% of
the unoperated subjects and 67% of the surgi-
cally treated individuals in ISUIA (International
Study of Unruptured Intracranial Aneurysms &
Investigators 2003). The prospective ISUIA data
on treatment outcomes are interesting although
directly comparing clipping and coiling is dif-
ficult because patient characteristics differ be-
tween the cohorts. There were proportionately
far more elderly patients, posterior circulation
and large aneurysms (all predisposing to poorer
outcome) in the endovascular cohort. Never-
theless, for those with no prior SAH the com-
bined morbidity and mortality at 1 year was
12.6% for clipping and 9.8% for coiling – a 22%
relative risk reduction. For prior SAH patients
at one year the relative risk reduction was 30%
(7.1% vs. 10.1%).
© 2004 Blackwell Publishing Ltd98 PRACTICAL NEUROLOGY
In unruptured aneurysms the long-term du-
rability after treatment is of particular concern.
Although the durability of surgical clipping is
widely accepted, the literature on re-bleeding
from a previously clipped aneurysm is actu-
ally quite scanty and nearly all the information
on re-bleeding rates after coiling or clipping
* relates to previously ruptured aneurysms
which are probably more likely to re-bleed after
treatment than an unruptured aneurysm. In
a purely surgical series, 17 patients presented
with ruptured or symptomatic recurrences post
clipping – 0.1% per annum (Lin & Fox 1989). A
centre, which treated 466 patients with coiling
between 1992 and 2002, found major aneurysm
recurrences in 21% of treated aneurysms. Three
(a) patients re-bled from a coiled aneurysm dur-
ing a mean follow-up of 31 months – a risk of
approximately 0.25% (Raymond et al. 2003).
ISAT will eventually provide 5 and then 10-year
follow-up, which will help resolve the issue of
endovascular treatment durability.
How to assess if treatment should be
offered?
The ‘effective natural history risk’ is the indi-
vidual’s life expectancy multiplied by the annual
rupture risk of their aneurysm. For treatment
to be worthwhile at all, the natural history risk
should substantially exceed the treatment risk
– after all the treatment risk is at once, whereas
the untreated risk accumulates over time. As an
example, a giant posterior circulation aneurysm
has an annual rupture rate of 10% but a treatment
risk of 40% (International Study of Unruptured
Intracranial Aneurysms & Investigators 2003).
(b) If a patient has a life expectancy of 20 years, the
chance of rupture far exceeds the 40% treatment
risk. On the other hand if the patient is a 71 years
old male with, for example, a 4-year life expect-
ancy, the effective natural history risk minus the
40% treatment risk is about zero. In practice, a
sensible figure in order to at least consider treat-
ing an asymptomatic unruptured aneurysm in
an individual is for the effective natural history
risk to exceed the treatment risk by at least 5–10
years. In patients with prior SAH, treatment of
Figure 5 Basilar tip aneurysm additional unruptured aneurysms will often be
with false sac (*) before (a) and indicated but life expectancy, comorbidity and
after coiling (b) with complete treatment risks must be assessed carefully on an
occlusion achieved. Image individual basis. But other factors such as pa-
(c) demonstrates a three- tient anxiety and implications for employment
dimensional GDC coil of the may all have an important impact on the final
(c) type commonly used as an initial treatment decision.
coil in such procedures.
© 2004 Blackwell Publishing LtdAPRIL 2004 99
Are there other worthwhile interven-
tions?
Stopping smoking, careful control of blood
pressure, avoidance of risk factors for athero-
sclerosis (careful diet, regular exercise, etc.),
while unproven, may help reduce both the risk
of formation of aneurysms and their risk of rup-
ture, as well as improving general health.
SCREENING FOR ANEURYSMS It is difficult to know for certain
Unfortunately, we cannot tell when aneurysms
are going to rupture or form de novo. It is diffi- who to screen, when to screen
cult to know for certain who to screen, when to
screen and how often, which aneurysms to treat and how often to screen
(or sometimes how best to treat) and which to
leave alone.
Several groups have applied detailed models
to the screening decision-analysis process for
aneurysms (MARS study Group 1999; Leblanc
et al. 1994; Obuchowski et al. 1995; ter Berg et
al. 1988; King et al. 1995; Kallmes et al. 1998;
Crawley et al. 1999; Baba et al. 2000; Yoshimoto
& Wakai 1999). The most recent studies using half of those with an asymptomatic murmur
up-to-date information on aneurysm rupture became anxious after detection of the murmur
rates, diagnostic test accuracy and treatment despite the normal echocardiogram (McDon-
morbidity rates have all concluded that routine ald et al. 1996). The presence of an unruptured
screening is not warranted. The MARS study asymptomatic aneurysm is considered to be in-
used MRA to screen 626 asymptomatic first- cidental by the UK Driver and Vehicle Licensing
degree relatives of 193 SAH patients and found Authority for ordinary driving licences but for
31 aneurysms in 23 relatives (4%). IADSA heavy goods vehicle and public service vehicle li-
confirmed the presence of all 31 aneurysms, but cences, the licence depends on a specialist assess-
was not performed in the other 603 relatives, so ment of the risks on an individual patient basis.
the number of aneurysms missed is unknown This might have considerable employment and
(MARS study Group 1999). The MARS in- financial implications for some patients. Fur-
vestigators found surgery-related morbidity thermore, one needs to consider the effect of
occurred in 11/18 operated relatives, and no finding a small aneurysm where knowledge of it
ruptures occurred in the non-operated relatives will generate considerable anxiety but where the
with aneurysms but over what period was not simple balance of natural history vs. treatment
stated. Overall surgery increased life expectancy risk does not favour treatment. Such a scenario
by 2.5 years per operation but at the expense could again have significant employment and
of 19 years of decreased function per person insurance implications. Will the patient be able
treated. The study investigators calculated to cope with this situation? It may be in their best
that 298 relatives would need to be screened interests not to find out.
to prevent one fatal SAH (MARS study Group So the message here is that, in general, screen-
1999). This would not be cost effective at about ing of relatives of SAH patients is inappropriate
12 million euros per life (MARS study Group and not cost effective. So only screen for an
1999; Johnston 2000). Similar conclusions were asymptomatic aneurysm if the pre-test prob-
reached in two other recent studies (Baba et al. ability of an aneurysm in an individual is very
2000; Yoshimoto & Wakai 1999). high and if their age and any comorbidity would
The stress of being screened is difficult to not militate against treatment. It is also vital to
quantify but is not insignificant. McDonald et al. explain to the patient before screening that, if an
assessed reassurance after a normal echocar- aneurysm is detected, treatment would only be
diogram in patients with a cardiac murmur. appropriate for about one half of the aneurysms
All those presenting with symptoms remained found. This proportion will be even lower for
anxious despite the normal test result, and over patients over the age of 50.
© 2004 Blackwell Publishing Ltd100 PRACTICAL NEUROLOGY
SO WHAT SHOULD I SAY TO THE PATIENT it. Again it is common sense and good medical
PRESENTING TO MY CLINIC WITH A practice not to investigate ‘just for reassurance’
FAMILY HISTORY OF ANEURYSMAL SAH without discussing first what an aneurysm
WHO IS WORRIED ABOUT HIS OR HER means and what treatment, if indicated (and
OWN RISK? it might not be), involves. If a patient wants to
The first thing is to document their family his- consider investigation and treatment once in the
tory and the presence or absence of other life- full possession of these facts then it is appropri-
style risk factors. ate to refer for imaging. Explain briefly what the
test involves and that they will be reviewed with
If the family history is weak the result. It is worth emphasising that there is
For example, if one first or second degree, or no urgency about investigation or subsequent
even two second-degree relatives, are affected treatment. If there are any compounding risk
the individual should be reassured strongly that factors, such as smoking, emphasise the impor-
further investigation is not indicated because tance of modifying behaviour.
there is a very low risk of SAH (lifetime risk of
no more than 2–3% vs. background population How to screen?
risk of about 0.6%). If they are not reassured, As most aneurysms < 5 mm in asymptomatic
although most are, further explanation about patients without SAH would not merit treat-
what is involved such as the uncertainty of ment, it is reasonable to screen first with CTA or
non-invasive tests, the risk of treatment and MRA – provided a unit has the appropriate ex-
possible implications on employment, driving, perience. Which test to do will depend on local
insurance, etc. if an aneurysm is detected, will availability and patient factors that make one or
lead many people to conclude that it is not in other the better option for that individual. How-
their interests to pursue further investigation. ever, patients must be warned of the limitations
Investigation without a clear discussion of these of a one-off CTA or MRA scan – a negative result
facts is not in the best interests of patients, and does not mean they don’t have an aneurysm, just
will cause considerable anxiety. that they probably don’t have one large enough
This advice assumes that you have the benefit to require treatment now. Also an aneurysm
of a consultation without the patient having could develop in the future. Whether to repeat
too many adverse preconceptions of their risk. screening examinations and if so how and when
If there are adverse preconceptions, then I’m is completely unknown.
afraid it is just going to be very difficult. For
example, if a doctor has explained the concept WHAT TO SAY ABOUT TREATMENT?
of aneurysms that are discovered incidentally to It is reasonable to stick to generalities at this
a patient in terms of a time-bomb ticking away stage. An explanation of the two treatment
in their head and then refers them on to the re- options available and an outline of their risks
gional neuroscience centre! Once a patient has is enough. Explain that the most appropriate
heard an ill-judged phrase such as ‘time-bomb’ treatment option will depend on the site and
and then had time to worry throughout the size of any aneurysm found and possibly the
referral process, no amount of facts and figures patient’s age. Reiterate that for many asympto-
will reassure them. They will almost always de- matic aneurysms treatment is not needed. Local
mand treatment come what may. expertise and availability of treatment may also
come into play but it is not necessary to discuss
If the family history is strong that yet. Finally, this is a complex and difficult
If, for example, two or more first-degree rela- enough field for doctors, let alone the patients.
tives are affected, or there is autosomal domi- We try and provide some basic information the
nant polycysctic kidney disease with a family patients can take away and read, and frequently
history of aneurysmal SAH, individuals are at bring them back for a second clinic visit to go
a substantially greater risk of both harbouring over the facts again before embarking on any in-
an aneurysm and sustaining an SAH compared vestigation. If an aneurysm is identified at a site
with the background population. They may and of a size to warrant treatment, the patient
merit screening for aneurysms provided they should be referred to a specialist multidiscipli-
would contemplate treatment if that proves to nary neurovascular clinic, but if that is unavail-
be appropriate and their life expectancy justifies able to a vascular neurosurgeon.
© 2004 Blackwell Publishing LtdAPRIL 2004 101
The role of coiling as opposed to clipping
in the treatment of unruptured aneurysms
is promising
PRACTICE POINTS overall risk. A substantial number of asymptomatic an-
• Symptomatic aneurysms are those causing SAH follow- eurysms fall into this category.
ing rupture, or exerting symptoms by a space occupying • Treatment risks rise with age and also depend on the
effect (most commonly oculomotor nerve palsy pro- site and size of the aneurysm. The greatest risks are with
duced by a posterior communicating artery aneurysm). larger aneurysms, posterior circulation aneurysms and
• Asymptomatic aneurysms are additional aneurysms in older people (> 50 years). In all these groups coiling
found in patients with a symptomatic aneurysm, which is safer than clipping – the substantially lower risk of
are not responsible for the clinical presentation, or an- treatment favours coiling despite the uncertainty over
eurysms found in patients investigated because they durability.
are at risk (of harbouring an aneurysm). • Treatment risks of coiling and clipping for asympto-
• Incidental aneurysms are those found unexpectedly in matic small, anterior circulation aneurysms in patients
patients undergoing investigation for some other sus- < 50 years are similar, so the unproven long-term dura-
pected problem. bility of coiling is a particular issue in this group.
• About one in 20 of the general population aged over the • In patients with no prior history of SAH, risk benefit anal-
age of 30 harbours an unruptured asymptomatic aneu- ysis favours treatment in people < 50 years old except for
rysm. those with anterior circulation aneurysms < 7 mm in size
• Individuals most at risk of aneurysmal SAH are those (but these account for nearly 50% of the total).
with a family history of at least two first-degree relatives • For people > 50 years, treatment is only clearly favoured
with an SAH, or patients with an underlying genetic for aneurysms > 12 mm, and other factors such as life
condition – most commonly autosomal dominant poly- expectancy and coexistent conditions need to be taken
cystic kidney disease. into account.
• The risk seems highest if at least one of the affected • Routine screening for unruptured intracranial aneu-
relatives is a sibling or mother. Lifestyle factors such rysms in individuals with a family history of SAH is not
as smoking, heavy alcohol consumption and hyperc- warranted.
holesterolaemia may further increase the risk. A single • Asymptomatic individuals with only one family member
relative (especially if second-degree) affected by SAH affected by SAH do not have a sufficiently increased risk
suggests a very small absolute risk of SAH to an indi- to outweigh the risks of screening (and treatment).
vidual without SAH. • Those patients with the greatest risk of having an unrup-
• CT angiography and MR angiography, performed and tured aneurysm, and of it then rupturing, are those aged
reported by experts, can reliably detect aneurysms over 30 years from families with two or more first degree
larger than 5 mm. These techniques are unreliable at relatives affected by SAH. They are perhaps at further
detecting smaller aneurysms. A negative study may increased risk if they have additional risk factors such
therefore offer false reassurance. This problem should as smoking and hypercholesterolaemia. Such individu-
be discussed with anyone referred for imaging to look als and autosomal dominant polycystic kidney disease
for an asymptomatic aneurysm. patients with a family history of SAH should be assessed
• Rupture risk depends predominantly on the site and size on an individual basis – taking all the relevant risk factors
of an aneurysm and whether or not the patient has suf- into account – for screening for intracranial aneurysms.
fered a previous SAH. If not, small aneurysms (< 7 mm), • The role of coiling as opposed to clipping in the treat-
particularly if anterior circulation, have an extremely ment of unruptured aneurysms is promising, particu-
low rupture risk. Other factors such as age and lifestyle larly in patients > 50 years, and for posterior circulation
factors also then need to be considered to determine aneurysms.
© 2004 Blackwell Publishing Ltd102 PRACTICAL NEUROLOGY
Fuller G (2001) What should I tell a patient after an iso-
ACKNOWLEDGEMENTS lated episode of demyelination? Practical Neurology,
This paper was reviewed by Jan van Gijn, 1, 106–12.
Utrecht, the Netherlands. Gaist D, Vaeth M, Tsiropoulos I et al. (2000) Risk of sub-
arachnoid haemorrhage in first-degree relatives of
REFERENCES patients with subarachnoid haemorrhage: Follow up
Adamson J, Humphries SE, Ostergaard JR, Voldby B, Ri- study based on national registries in Denmark. British
chards P & Powell JT (1994) Are Cerebral Aneurysms Medical Journal, 320, 141–5.
Atherosclerotic? Stroke, 25, 963–6. Guglielmi G, Vinuela F, Dion J et al. (1991) Electro-
Anderson C, Mhurchu CN, Scott D, Bennett D, Jamro- thrombosis of saccular aneurysms via endovascular
zik K & Hankey G (2003) Triggers of subarachnoid approach: Part II. Preliminary clinical experience.
hemorrhage. Role of physical exertion, smoking, and Journal of Neurosurgery, 75, 8–14.
alcohol in the Australasian cooperative research on Hughes R, Chapman A, Rubinstein D, Stears J, Johnson A
subarachnoid hemorrhage study (ACROSS). Stroke, & Gabow P (1996) Recurrent intracranial aneurysms
34, 1771–6. in autosomal dominant polycystic kidney disease
Baba Y, Takahashi M & Korogi Y (2000) Cost-effective- (ADPKD). Stroke, 27, 178.
ness of screening unruptured cerebral aneurysms in International Study of Unruptured Intracranial An-
Japan. European Radiology, 10, S362–S365. eurysms (ISUIA) Investigators. (1998) Unruptured
ter Berg HWM, Dippel D, Habbema J et al. (1988) Treat- intracranial aneurysms- risk of rupture and risks of
ment of intact familial intracranial aneurysms: a deci- surgical intervention. New England Journal of Medi-
sion-analytical approach. Neurosurgery, 23, 329–34. cine, 339, 1725–33.
ter Berg HWM, Dippel DWJ, Limburg M, Schievink WI International Study of Unruptured Intracranial An-
& van Gijn J (1992) Familial intracranial aneurysms: eurysms (ISUIA) Investigators. (2003) Unruptured
a review. Stroke, 23, 1024–30. intracranial aneurysms: natural history, clinical
van den Berg JSP, Limburg M & Hennekam RCM (1996) outcome and risks of surgical and endovascular treat-
Is Marfan syndrome associated with symptomatic ment. Lancet, 362, 103–10.
intracranial aneurysms? Stroke, 27, 10–2. International Subarachnoid Aneurysm Trial (2002)
van den Berg JSP, Limburg M, Hennekam RCM, (ISAT) of neurosurgical clipping versus endovascular
Cruysberg JRM, Steijilen PM & Limburg M (1999) coiling in 2143 patients with ruptured aneurysms: a
Prevalence of symptomatic intracranial aneurysm randomised trial. ISAT Collaborative Group. Lancet,
and ischaemic stroke in pseudoxanthoma elasticum. 360, 1267–74.
Cerebrovascular Disease, 9, 3. Jager HR, Mansmann U, Hausmann O, Partzsch U,
Brilstra EH, Rinkel GJE, van der Graaf Y, van Rooij WJJ & Moseley IF & Taylor WJ (2000) MRA versus dig-
Algra A (1999) Treatment of intracranial aneurysms ital subtraction angiography in acute subarachnoid
by embolization with coils: a systematic review. Stroke, haemorrhage: a blinded multireader study of pro-
30, 470–6. spectively recruited patients. Neuroradiology, 42,
Bromberg JEC, Rinkel GJE, Algra A et al. (1995) Sub- 313–26.
arachnoid haemorrhage in first and second degree Johnston SC, Colford JM & Gress DR (1998) Oral con-
relatives of patients with subarachnoid haemorrhage. traceptives and the risk of subarachnoid hemorrhage-
British Medical Journal, 311, 288–9. A meta-analysis. Neurology, 51, 411–8.
Bromberg JEC, Rinkel GJE, Algra A, Limburg M & van Johnston SC (2000) Effect of endovascular services and
Gijn J (1995) Outcome in familial subarachnoid hospital Volumes on cerebral aneurysm treatment
haemorrhage. Stroke, 26, 961–3. outcomes. Stroke, 31, 111–7.
Chappell ET, Moure FC & Good MC (2003) Comparison Juvela S, Porras M & Heiskanen O (1993) Natural history
of Computed Tomographic Angiography with Digital of unruptured intracranial aneurysms: a long-term
Subtraction Angiography in the diagnosis of cerebral follow-up study. Journal of Neurosurgery, 79, 174–82.
aneurysms: Meta-Analysis. Neurosurgery, 52, 624–31. Kallmes DF, Kallmes MH, Cloft HJ & Dion JE (1998)
Chung TS, Joo JY, Lee SK, Chien D & Laub G (1999) Guglielmi detachable coil embolization for un-
Evaluation of cerebral aneurysms with high-resolu- ruptured aneurysms in nonsurgical candidates: a
tion MR angiography using a section interpolation cost-effectiveness exploration. American Journal of
technique: correlation with digital subtraction an- Neuroradiology, 19, 167–76.
giography. American Journal of Neuroradiology, 20, King JT, Glick HA, Mason TJ & Flamm ES (1995) Elective
229–35. surgery for asymptomatic, unruptured, intracranial
Cloft HJ, Joseph GJ & Dion JE (1999) Meta-analysis of aneurysms: a cost-effectiveness analysis. Journal of
risks of cerebral angiography in patients with sub- Neurosurgery, 83, 403–12.
arachnoid hemorrhage, intracranial aneurysm and Kojima M, Nagasawa S, Lee YE, Takeichi Y, Tsuda E &
arteriovenous malformation. Stroke, 30, 317–20. Mabuchi N (1998) Asymptomatic familial cerebral
Crawley F, Clifton A & Brown M (1999) Should we aneurysms. Neuroimaging Clinics of North America,
screen for Familial Intracranial Aneurysm? Stroke, 43, 776–81.
30, 312–6. Leblanc R (1997) Familial cerebral aneurysms. Canadian
De Braekeleer M, Perusse L, Cantin L, Bouchard J-M & Journal of Neurological Science, 24, 191–9.
Mathieu J (1996) A study of inbreeding and kinship in Leblanc R, Worsley KJ, Melanson D & Tampieri D
intracranial aneurysms in the Saguenay-Lac St-Jean (1994) Angiographic screening and elective surgery
regio (Quebec, Canada). Annals of Human Genetics, of familial cerebral aneurysms: a decision-analysis.
60, 99–104. Neurosurgery, 35, 9–18.
© 2004 Blackwell Publishing LtdAPRIL 2004 103
Lin T & Fox (1989) AJ, drake CG. Regrowth of aneurysm of two cases and review of the literature. Annals of
sacs from residual neck following aneurysm clipping. Neurology, 4, 130–44.
Journal of Neurosurgery, 70, 556–60. Ronkainen A, Hernesniemi JA, Puranen M et al. (1997)
Linn FHH, Rinkel GJE, Algra A & van Gijn J (1996) Inci- Familial intracranial aneurysms. Lancet, 349, 380–4.
dence of subarachnoid haemorrhage. Role of region, Ronkainen A, Miettinen H, Karkola K et al. (1998) Risk
year and rate of computed tomography: a meta-analy- of harboring an unruptured intracranial aneurysm.
sis. Stroke, 27, 625–9. Stroke, 29, 359–62.
Lozano AM & Leblanc R (1987) Familial intracranial Ruggieri PM, Poulos N, Masaryk TJ et al. (1994) Occult
aneurysms. Journal of Neurosurgery, 66, 522–8. intracranial aneurysms in polycystic kidney disease:
MARS study Group. (1999) Risks and benefits of screen- screening with MR angiography. Radiology, 191,
ing for intracranial aneurysms in first-degree relatives 33–9.
of patients with sporadic subarachnoid haemorrhage. Schievink WI, Katzmann JA & Piepgras DG (1996)
New England Journal of Medicine, 341, 1344–50. Alpha-1-antitrypsin phenotype among patients with
McDonald IG, Daly J, Jelinek VM, Panetta F & Gutman intracranial aneurysms. Journal of Neurosurgery, 84,
JM (1996) Opening Pandora’s box: the unpredict- 781–4.
ability of reassurance by a normal test result. British Schievink WI, Limburg M, Oorthuys JWE, Fleury P &
Medical Journal, 313, 329–32. Pope FM (1990) Cerebrovascular disease in Ehlers–
Metens T, Rio M, Baleriaux D, Roger T, David P & Rode- Danlos syndrome type IV. Stroke, 21, 626–32.
sch G (2000) Intracranial aneurysms: detection with Schievink WI, Schaid DJ, Michels VV & Piepgras DG
gadolinium-enhanced dynamic three-dimensional (1995) Familial aneurysmal subarachnoid hemor-
MR Angiography – initial results. Radiology, 216, rhage: a community-based study. Journal of Neuro-
39–46. surgery, 83, 426–9.
Morooka Y & Waga S (1991) Cerebral aneurysms as- Suzuki M, Matsui O, Ueda F et al. (2003) contrast-en-
sociated with neurofibromatosis. Surgery Neurology, hanced MR angiography (enhanced 3-D fast gradient
19, 260–2. echo) for diagnosis of cerebral aneurysms. Neurora-
Mulvihill JJ, Parry DM & Sherman JL (1990) Neurofi- diology, 44, 17–20.
bromatosis 1 (Recklinghausen disease) and neurofi- Teunissen LL, Rinkel GJE, Algra A & van Gijn J (1996)
bromatosis 2 (bilateral acoustic neurofibromatosis). Risk factors for subarachnoid haemorrhage: a system-
Annals of International Medicine, 113, 39–52. atic review. Stroke, 27, 544–9.
Munyer TP & Margulis AR (1981) Pseudoxanthoma Uehara T, Tabuchi M & Mori E (1998) High frequency of
elasticum with internal carotid artery aneurysm. unruptured intracranial aneurysms in female patients
American Journal of Roentgenology, 136, 1023–4. with ischaemic heart disease. Journal of Neurology,
Norrgard O, Angquist K, Fodstad H, Forsell A & Limburg Neurosurgery and Psychiatry, 64, 536–8.
M (1987) Intracranial aneurysms and heredity. Neu- Wang P, Longstreth WT & Koepsell TD (1995) Sub-
rosurgery, 20, 236–9. arachnoid hemorrhage and family history: a popula-
Obuchowski NA, Modic MT & Magdinec M (1995) tion based case-control study. Archives of Neurology,
Current implications for the efficacy of noninvasive 52, 202–4.
screening for occult intracranial aneurysms in pa- Wardlaw JM (2001) on behalf of Davie Cooper Study
tients with a family history of aneurysms. Journal of Investigators. Familial Risk of Subarachnoid Haem-
Neurosurgery, 83, 42–9. orrhage in the Scottish Population. Neuroradiology,
Okahara M, Kiyosue H, Yamashita M et al. (2002) Diag- 43, 415–6. [full results in press].
nostic accuracy of Magnetic Resonance Angiography Wardlaw JM & White PM (2000) The detection and
for cerebral aneurysms in correlation with 3D-Digital management of unruptured intracranial aneurysms.
Subtraction Angiogrpahic Images- a study of 133 Brain, 123, 205–21.
aneurysms. Stroke, 33, 1803–8. White PM, Teasdale EM, Wardlaw JM & Easton V
Okamoto K, Horisawa R, Kawamura T et al. (2003) (2001a) CTA and MRA in the detection of intracra-
Family history and risk of subarachnoid hemorrhage. nial aneurysms: a prospective, blinded comparison in
A case-control study in Nagoya, Japan. Stroke, 34, a large patient cohort. Radiology, 219, 739–46.
422–6. White PM, Teasdale EM, Wardlaw JM, Lindsay KW,
Pedersen HK, Bakke SJ, Hald JK et al. (2001) CTA in Sloss S & Patel D (2003) The non-invasive detection
patients with acute subarachnoid haemorrhage. Acta of intracranial aneurysms: do neuroradiologists do it
Radiologica, 42, 43–9. better? European Radiology, 13, 389–96.
Raaymakers TWM, Rinkel GJE, Limburg M & Algra A White PM, Wardlaw JM & Easton V (2000) Can non-
(1998) Mortality and Morbidity of Surgery for Un- invasive imaging tests accurately depict intracra-
ruptured Intracranial Aneurysms. Stroke, 29, 1531–8. nial aneurysms? A systematic review. Radiology, 217,
Raymond J, Guilbert F, Weill A et al. (2003) Long-term 361–70.
angiographic recurrences after selective endovascular White PM, Wardlaw JM, Teasdale E, Sloss S, Cannon J
treatment of aneurysms with detachable coils. Stroke, & Easton V (2001b) Power Transcranial Doppler Ul-
34, 1398–403. trasound in the detection of intracranial aneurysms.
Rinkel GJE, Djibuti M, Algra A & van Gijn J (1998) Prev- Stroke, 32, 1291–7.
alence and risk of rupture of intracranial aneurysms. Yoshimoto Y & Wakai S (1999) Cost-effectiveness
A systematic review. Stroke, 29, 251–6. analysis of screening for asymptomatic, unruptured
Roman G, Fisher M, Perl DP & Poser CM (1978) Neu- intracranial aneurysms. A mathematical model.
rological manifestations of hereditary hemorrhagic Stroke, 30, 1621–7.
telangiectasias (Rendu-Osler-Weber disease): report
© 2004 Blackwell Publishing LtdYou can also read
