Intracranial aneurysms: individualising the risk of rupture.

The decision of what to do with asymptomatic unruptured intracranial aneurysms is a fairly new problem. Although these aneurysms have been known to exist for centuries through autopsies and for several decades through catheter angiography, their true prevalence did not begin to emerge until the use of non-invasive angiograms became widespread in the recent past. Now we know that nearly 3% of the general population has an unruptured intracranial aneurysm. Comparatively, subarachnoid haemorrhages from aneurysmal rupture are relatively uncommon. This discrepancy indicates that many intracranial aneurysms are not destined to rupture. Yet, the consequences of rupture can be devastating. Treatment of the aneurysm by craniotomy and clipping or by endovascular coiling can eff ectively eliminate the risk of subarachnoid haemorrhages, but treatment of all unruptured intracranial aneurysms is neither prudent, because of the risk of iatrogenic complications, nor parsimonious, because of the high fi nancial cost. Thus, when advising a patient with an asymptomatic unruptured intracranial aneurysm we face a diffi cult problem. An adequate solution can only be based on individualisation of the rupture risk. In The Lancet Neurology, Jacoba Greving and colleagues present their analysis of individual patient data pooled from six prospective cohort studies on the natural history of unruptured intracranial aneurysms (three from Japan, one from the Netherlands, one from Finland, and one from the USA, Canada, and various European countries). The large sample size of the combined population allowed the researchers to refi ne the estimation of rupture risk at 1 and 5 years, defi ne the strongest predictors of rupture among those evaluated in the original studies, and develop a score (named PHASES) to gauge the individual risk of rupture using readily available information. The information provided is solid and has practical value. I plan to use it when advising my patients, although keeping in mind certain caveats. The PHASES score is calculated from the region of origin of the patient (attributing a higher risk to Finnish and Japanese patients than to all others, although specifi c information is not available for most world regions), presence of hypertension, patient’s age (dichotomised at 70 years), maximum diameter of the aneurysm (by far the strongest predictor of rupture), previous history of subarachnoid haemorrhage from another aneurysm, and the site of the aneurysm (with a higher risk assigned to aneurysms arising from the anterior cerebral arteries, posterior communicating arteries, or the posterior circulation vessels). Notable absences include smoking (at entry and on follow-up), hypertension control, family history of subarachnoid haemorrhage, multiplicity of unruptured intracranial aneurysms, other anatomical and haemodynamic factors of the aneurysm (infl ow angle, concentrated infl ow jets, complex fl ow patterns, non-spherical shape, dome-to-neck ratio), and, most importantly, evidence of aneurysm growth over time. Yet, these additional factors also deserve careful consideration. Smoking cessation and control of hypertension might reduce rupture risk. Follow-up imaging to exclude aneurysm growth is necessary because growth is strongly associated with increased risk of subarachnoid haemorrhage, although the optimum frequency and duration of radiological follow up is not well defi ned. The Article also incorporates charts off ering specifi c 5-year risk prediction for diff erent populations. Readers should note that the chart that would apply to North America and European countries other than Finland is based on information from the large ISUIA cohort (1691 patients with a median follow-up of 9 years)