ESOC 2025 | The MARTA score: predicting complications of treatment for unruptured intracranial aneurysms

The prevalence of intracranial aneurysm is around 5%. Normally we say to patients that we need to treat the aneurysm because the conservative management is associated with a higher risk compared with treatment, but we have very few instruments to weigh our approach and to weigh the complication rates of our approach. So what we did is to first define the characteristics associated with complex intracranial aneurysm. This was done through a Delphi consensus published a year ago, and this allowed us to identify key variables that were associated with complication rates overall. So, morbidity, mortality, and the complex procedure. We then moved to a validation, a development and a validation of the MARTA score, which is a very simple scoring system to predict the procedural complication rate in unruptured intracranial aneurysms treated endovascularly. So the aim was to assess its clinical utility in predicting the complication. So data was collected retrospectively from 15 international hospitals over a 10-year period spanning 2014 to 2024. We had more than 2,000 unruptured intracranial aneurysms treated. And the data were analyzed according to a very simple primary outcome, which is a composite outcome of a safety endpoint, so including any neurological deficit, a modified Rankin score worsening, or procedural death. We had internal and external validation through a temporal validation cohort, and we compared our score with the SAFET and ACC score, which is the most widely used score to predict the complication rates. So our score had a very good ROC, so a very good accuracy in predicting the complications in the testing cohort, which was around 0.87, and this still was very good in the development cohort, so in 0.76. This means that overall the two cohorts, the one where we developed the score and the one where we tested it, were overall behaving in a similar way. When comparing overall the score with the risk stratification, we saw that the risk of complication increased with increasing score. So this was almost linear with very low complication rates with MARTA score below 4 and very high complication rates, as high as 18% in MARTA EVT score, higher than 9. When compared to the SAFET and ACC score, the MARTA EVT score performed more naturally, I would say, it predicted in a more accurate way the complication rate compared to the SAFET and the SCC, and this was significant with the DeLong ROC comparative test. So, overall, what we can say is, considering all the limitations of the study, which is we only enrolled in a retrospective fashion, we had a very high case mix, and we did not account for variables associated with risk of aneurysm rupture, which are unknown, and also that certain aneurysm location and morphologies may have been underrepresented in our cohort. Overall, we can say that the MARTA score seems an accurate, easy to calculate scoring system for predicting the procedural complication. So whenever a patient comes in and we offer treatment for an unruptured aneurysm, we may use the MARTA score to assess what is the risk of procedural complication and inform both the patients and the clinicians.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.