Current decision support tools fail to agree or predict therapeutic decisions in a single cohort of unruptured intracranial aneurysms

Evaluation of PHASES Score for Predicting Rupture of Intracranial Aneurysms: Significance of Aneurysm Size

BACKGROUND

Recent data have identified that certain risk factors for rupture differ between small and larger intracranial aneurysms (IAs). Such differing risk factors make up 5 out of the 6 predictor variables used in the PHASES score, which raises the question on whether IA size has a significant effect on the score's performance.

METHODS

Patients who were diagnosed with an IA incidentally or due to a subarachnoid hemorrhage between 2015 and 2023 were selected for potential inclusion. The median IA size of the cohort was chosen as the cutoff point to categorize small and large (6 mm). The PHASES score was calculated for all patients, and a receiver operating characteristic curve analysis was performed to evaluate the classification accuracy of PHASES in predicting rupture for small and large IAs.

RESULTS

A total of 677 IAs were included. Among the IAs, 400 (58.9%) presented as UIAs and 279 (41.0%) as subarachnoid hemorrhage. The average PHASES score was 2.9 and 6.5 for small (n = 322) and large (n = 355) IAs, respectively. The PHASES score performed significantly lower for predicting rupture in smaller IAs (area under the curve: 0.634) compared with the larger (area under the curve: 0.741) (P = 0.00083).

CONCLUSIONS

PHASES was shown to underperform on small IAs. The decision to treat small unruptured IAs remains highly controversial, and the development of a new score to estimate the annual rupture rate while accounting for IA morphology is of great need. Our findings can help encourage future researchers to develop such a score.

Predictors for Rupture of Small (<7mm) Intracranial Aneurysms: A Systematic Review and Meta-Analysis

INTRODUCTION

Identifying predictors for rupture of small intracranial aneurysms (sIAs) have become a growing topic in the literature given the relative paucity of data on their natural history. The authors performed a meta-analysis to identify reliable predictors.

METHODS

PubMed, Scopus, and Web of Science were used to systematically extract references which involved at least 10 IAs <7mm which including a control group experiencing no rupture. All potential predictors reported in the literature were evaluated in the meta-analysis.

RESULTS

Fifteen studies yielding 4,739 sIAs were included in the meta-analysis. Four studies were prospective and 11 were retrospective. Univariate analysis identified 7 predictors which contradicted or are absent in the current scoring systems, while allowing to perform subgroup analysis for further reliability: patient age (MD -1.97, 95%CI -3.47-0.48; P = 0.01), the size ratio (MD 0.40, 95%CI 0.26-0.53; P < 0.00001), the aspect ratio (MD 0.16, 95%CI 0.11-0.22; P < 0.00001), bifurcation point (OR 3.76, 95%CI 2.41-5.85; P < 0.00001), irregularity (OR 2.95, 95%CI 1.91-4.55; P < 0.00001), the pressure loss coefficient (MD -0.32, 95%CI -0.52-0.11; P = 0.002), wall sheer stress (Pa) (MD -0.16, 95%CI -0.28-0.03; P = 0.01). All morphology related predictors listed above have been confirmed as independent predictors via multivariable analysis among the individual studies.

CONCLUSIONS

Morphology related predictors are superior to the classic patient demographic predictors present in most scoring systems. Given that morphology predictors take time to measure, our findings may be of great interest to developers seeking to incorporate artificial intelligence into the treatment decision-making process.

Trends in the Age of Patients Treated for Unruptured Intracranial Aneurysms from 1990 to 2020

BACKGROUND

The decision for treatment for unruptured intracranial aneurysms (UIAs) is often difficult. Innovation in endovascular devices have improved the benefit-to-risk profile especially for elderly patients; however, the treatment guidelines from the past decade often recommend conservative management. It is unknown how these changes have affected the overall age of the patients selected for treatment. Herein, we aimed to study potential changes in the average age of the patients that are being treated over time.

METHODS

A systematic search of the literature was performed to identify all studies describing the age of the UIAs that were treated by any modality. Scatter diagrams with trend lines were used to plot the age of the patients treated over time and assess the presence of a potential significant trend via statistical correlation tests.

RESULTS

A total of 280 studies including 83,437 UIAs treated between 1987 and 2021 met all eligibility criteria and were entered in the analysis. Mean age of the patients was 55.5 years, and 70.7% were female. There was a significant increasing trend in the age of the treated patients over time (Spearman r: 0.250; P < 0.001), with a 1-year increase in the average age of the treated patients every 5 years since 1987.

CONCLUSIONS

The present study indicates that based on the treated UIA patient data published in the literature, older UIAs are being treated over time. This trend is likely driven by safer treatments while suggesting that re-evaluation of certain UIA treatment decision scores may be of great interest.