Retrospective application of risk scores to ruptured intracranial aneurysms: would they have predicted the risk of bleeding?

As the incidental diagnosis of unruptured intracranial aneurysms has been increasing, several scores were developed to predict risk of rupture and growth to guide the management choice. We retrospectively applied these scores to a multicenter series of patients with subarachnoid hemorrhage to test whether they would have predicted the risk of bleeding in the event of aneurysm discovery previous to its rupture. Demographical, clinical, and radiological information of 245 adults were retrieved from two neurovascular centers' database. Data were pooled and PHASES, UCAS, and ELAPSS scores were retrospectively calculated for the whole population and their performances in identifying aneurysms at risk of rupture were compared. Mean PHASES, UCAS, and ELAPSS scores were 5.12 +/- 3.08, 5.09 +/- 2.62, and 15.88 +/- 8.07, respectively. Around half (46%) of patients would have been assigned to the low- or very low-risk class (5-year rupture risk < 1%) in PHASES. Around 28% of patients would have been in a low-risk class, with a probability of 3-year rupture risk < 1% according to UCAS. Finally, ELAPSS score application showed a wider distribution among the risk classes, but a significant proportion of patients (45.5%) lie in the low- or intermediate-risk class for aneurysm growth. A high percentage of patients with ruptured aneurysms in this multicenter cohort would have been assigned to the lower risk categories for aneurysm growth and rupture with all the tested scores if they had been discovered before the rupture. Based on these observations, physicians should be careful about drawing therapeutic conclusions solely based on application of these scores.