Intended for healthcare professionals


Screening for familial intracranial aneurysms

BMJ 1999; 319 doi: (Published 11 December 1999) Cite this as: BMJ 1999;319:1512

No justification exists for routine screening

  1. Peter J Kirkpatrick, consultant neurosurgeon,
  2. R S McConnell, senior registrar in neurosurgery
  1. University Department of Neurosurgery, Addenbrooke's Hospital, Cambridge CB2 2QQ

    Familial intracranial aneurysms are those which have been identified in two or more first degree relatives. The relative risk for intracranial aneurysms in such families is about four times that of the general population. 1 2 In the general population the prevalence has been estimated from a meta-analysis to average 2.3% (95% confidence interval 1.7 to 3.1), though the meta-analysis also pointed out the wide variation in prevalence according to method of diagnosis and selection processes.3 In view of the high mortality and morbidity associated with ruptured intracranial aneurysms, screening for unruptured intracranial aneurysms has been suggested, is practised by many, and would at first sight seem a worthwhile goal. Once an aneurysm is discovered, options include treatment or further surveillance to detect growth The recent improved sensitivity of non-invasive magnetic resonance angiography has made screening more practicable. So should we now be seriously considering screening people for familial aneurysms?

    People with familial intracranial aneurysms represent about 5-10% of all patients presenting with an aneurysmal subarachnoid haemorrhage.4 Both genetic and environmental factors have been implicated to explain the clustering of intracranial aneurysms in families. These cases usually bleed at an earlier age and form smaller intracranial aneurysms than those of the general population of people with a subarachnoid haemorrhage, indicating enhanced susceptibility of the vascular wall, whether genetically determined or otherwise.5

    Screening for intracranial aneurysms demands a full understanding of the pathophysiological substrate for aneurysm rupture.6 Effective screening presumes that intracranial aneurysms develop slowly in a similar manner to aneurysms elsewhere (notably those affecting the abdominal aorta), that they can be detected before potential rupture, and that treatment will significantly reduce the incidence of subarachnoid haemorrhage. The intervention of low risk surgery on some unruptured aneurysms may then be justified. However, recent thinking questions these presumptions.

    Firstly, the assumption that intracranial aneurysms progress to rupture may not be correct. An alternative theory suggests that intracranial aneurysms form relatively acutely, possibly over the space of just a few hours, and rupture in only a proportion of cases.7 The acute presentation of a painful third cranial nerve palsy secondary to an ipsilateral posterior communicating artery aneurysm is a precise example of this. Those that do not rupture may stabilise and represent the target for any practicable screening programme. In this case periodic screening for intracranial aneurysms will intercept asymptomatic incidental intracranial aneurysms which may not be a common substrate for aneurysmal subarachnoid haemorrhage. Screening under these circumstances would be futile. Secondly, most ruptured aneurysms causing subarachnoid haemorrhage are angiographically and surgically small (<10 mm), whereas the risk of rupture of incidental aneurysms increases with size. Thirdly, the prevalence of incidental intracranial aneurysms does not match regional variations in the incidence of aneurysmal subarachnoid haemorrhage.8 Finally, if aneurysm formation represents a genetically or environmentally determined vascular fragility, then those who have already suffered an aneurysmal subarachnoid haemorrhage must be at the greatest risk of further haemorrhage rather than their unaffected relatives. Yet follow up cerebral angiography is not conducted in affected individuals because the yield is widely held to be very low.

    Two recent publications raise further concern over the validity of screening programmes for familial aneurysms. The first is derived from the large international study of unruptured intracranial aneurysms (ISUIA), which showed a very low incidence of spontaneous rupture (about 0.05% per year) for most incidental anterior aneurysms less than 10 mm. 9 10 Even those patients found to have incidental aneurysms following a subarachnoid haemorrhage have a low rate of rupture (about 0.5% per year).9 Thus most intracranial aneurysms appear to follow a relatively benign course, and progression in size is rare. In contrast, the risk of treatment for aneurysms was higher than previously reported (mean combined mortality and significant morbidity 13.1-15.7%), regardless of the mode of treatment (surgical or radiological). These data should be interpreted with the following caveats. Firstly, the rupture rate was derived from the retrospective arm of the study, which introduces selection bias; secondly, coincidental aneurysms found during angiography for carotid stenosis were overrepresented; and, thirdly, the rupture rate in young patients with familial aneurysm may be higher.11

    Even if we take a more pessimistic view of the rupture rate of incidental aneurysms (0.8% a year) in familial cases; a highly pessimistic view of outcome after subarachnoid haemorrhage (75% unfavourable); and an optimistic view of mortality (2.6%) and severe morbidity (5.4%)12 after elective treatment, a risk: benefit analysis provided by Crawley et al seems to rule out intervention for most incidental familial aneurysms detected using modern screening methods.13 This will hold true even if magnetic resonance angiography is used in isolation without introducing the small risk inherent in confirmatory digital subtraction angiography. In addition, Crawley et al estimated the investigational and treatment costs of an extremely modest 10 yearly screening programme to lie at £3-5m (US$4.8-8m) for every 1000 patients.

    Decisions based on the sort of generalised modelling analysis used by Crawley et al have limitations, and some types of incidental intracranial aneurysms are known to be more dangerous than others. The risk of rupture is greater in aneurysms measuring over 10 mm, in those located in the posterior circulation, and in patients who have already bled from other intracranial aneurysms.9 Large lesions are rare, especially those over 25 mm (giant), which are the most vulnerable to rupture. 3 6 9 For these higher risk lesions the operative risks escalate, thereby demanding a separate risk: benefit assessment. Aneurysms which declare symptoms (and hence are non-incidental) are also at high risk of rupture and demand treatment.7 We need to be able to identify the relative risks according to aneurysm site, size, and morphology—and indeed such analyses are proposed in a prospective extension of the international study of unruptured intracranial aneurysms. Such information may help in assessing the potential benefit of intervention in individual cases, but until this is done the screening of patients for familial intracranial aneurysms cannot be justified on clinical grounds alone. For non-familial cases, in whom the risks of aneurysm formation and subarachnoid haemorrhage are less, the need for screening is clearly inappropriate. However, advice about contributory systemic risk factors, especially smoking, remains essential.

    Defining higher risk patients and their incorporation into prospective studies is essential if such individuals are to gain from future developments. People with genetically associated conditions such as adult polycystic kidney disease may be more vulnerable to rupture of their incidental aneurysms,14 and such cases demand individual specialist attention. Indeed, the issue of whether screening is appropriate should be decided by those involved in the regular management of intracranial aneurysms Specialist referral thus remains essential.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.