The referral is almost always the same sentence: "Incidental 4 mm aneurysm on MRA done for headaches — please advise." By the time the patient arrives they have read three forum threads and one obituary, and the actual medical question — what is this lesion's rupture risk, and does intervening beat watching? — has been buried under a week of 2 a.m. searching.

The service this clinic visit performs is arithmetic plus honesty, in that order.

The natural history: what ISUIA and UCAS actually showed

Two cohorts anchor every number in this conversation. ISUIA followed thousands of unruptured aneurysms and found rupture risk stratified sharply by size and territory: small (<7 mm) anterior-circulation aneurysms in patients without prior SAH ruptured rarely, while risk rose with each size tier and with posterior-circulation or posterior-communicating location.1 The Japanese UCAS cohort — 6,697 aneurysms — put the overall rupture rate near 1% per year, again concentrated in larger lesions, posterior-communicating and anterior-communicating sites, and aneurysms with a daughter sac.2

Both cohorts carry the same honest caveats: managed patients had their scariest lesions repaired (selection bias trims the observed risk), and Japanese and Finnish populations run higher baseline risk — which is exactly why the P in PHASES exists.

PHASES: the estimate you can say out loud

PHASES pools six cohorts into a score built from Population, Hypertension, Age ≥70, Size (the dominant driver: points stack steeply from 7 mm up), Earlier SAH from another aneurysm, and Site (posterior-communicating and posterior circulation highest), returning a 5-year rupture probability that runs from under 1% at low scores to well into double digits at high ones.3 Its value in clinic is less precision than shared language: "your particular aneurysm's five-year risk is around 1–2%" is a sentence a patient can weigh against a repair risk — and against their own sleep. Its limits deserve equal honesty: it does not capture growth, irregular morphology beyond site, family history, or smoking — several of which push real decisions.

The patient heard "brain aneurysm" and priced it at catastrophe. The chart's job is to reprice it at its actual number.

Treat or watch

The AHA/ASA guideline frames the decision as risk-versus-risk, favoring treatment consideration when the lesion side of the scale is loaded — larger size, documented growth, symptomatic aneurysms (a new third-nerve palsy from a posterior-communicating aneurysm is an urgent version of this conversation), posterior location, prior SAH from another aneurysm, and younger patients with more years at risk — and surveillance for the small, stable, incidental majority.4 Repair itself is a neurovascular-team decision between endovascular therapy (coiling and its device descendants — generally lower upfront morbidity, some retreatment) and microsurgical clipping (more durable, more invasive), chosen by anatomy, age, and local expertise more than by ideology.

Surveillance is an active plan, not a shrug: periodic MRA/CTA (commonly at 6–12 months, then spaced out once stability is shown), with growth treated as a rupture-risk signal that reopens the treatment conversation. And every plan carries the same two prescriptions, because they are the modifiable rupture factors the scores undercount: smoking cessation and blood-pressure control.4 If the aneurysm was found during a stroke workup, those overlap entirely with the secondary-prevention work already underway.

Questions the visit must answer before it ends

  • "What if it bursts?" — honest but proportionate: rupture is subarachnoid hemorrhage, a true emergency (its severity graded by scales like Hunt and Hess) — and the entire point of today's math is that this patient's yearly probability is what it is, usually small.
  • "Can I exercise / fly / have coffee?" — normal life continues; no evidence-based prohibition list accompanies a small unruptured aneurysm. The warning that matters: a sudden worst-ever headache is an emergency-department event, full stop — kin to the other seconds-count presentations in what is a stroke.
  • "Should my kids be scanned?" — family screening (MRA) is reserved for ≥2 affected first-degree relatives, or heritable associations like autosomal-dominant polycystic kidney disease; one sporadic aneurysm in one parent does not trigger cascade imaging.4
  • Antithrombotics for other reasons — an unruptured aneurysm is generally not a contraindication to indicated antiplatelets or anticoagulation; the decision runs on its own merits (see antithrombotic selection) with the aneurysm managed in parallel.
  • The psychology — name it. For many patients the aneurysm's chief harm is anxiety; a number, a plan, and a scheduled scan treat that better than reassurance ever does. Screen the ones who aren't sleeping (GAD-7).

The bottom line

An incidental aneurysm is a probability, not a prophecy. Size and site carry most of the information; PHASES turns them into a number the patient can actually use; and the decision weighs that number against repair risk and years at stake. Treat the ones that have earned it, watch the majority on a real schedule, extinguish the cigarettes and the hypertension in everyone — and send the patient home with the number, the plan, and permission to live their life.