Intra- and inter-rater agreement of the Menon, ASITN/SIR, and Calgary collateral grading systems using multiphase computed tomography angiography, and their association with clinical outcomes in acute stroke
Neuroradiology, 2026 (SCI-Expanded, Scopus)
- Publication Type: Article / Article
- Publication Date: 2026
- Doi Number: 10.1007/s00234-026-04184-5
- Journal Name: Neuroradiology
- Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Aerospace Database, CINAHL, EMBASE, MEDLINE, Academic Search Ultimate (EBSCO), Natural Science Collection (ProQuest), Biological Science Database (ProQuest), Biomedical Reference Collection: Corporate Edition (EBSCO), Health Research Premium Collection (ProQuest), Pharma Collection (ProQuest), Technology Collection (ProQuest)
- Keywords: Acute ischemic stroke, Cerebral collateral circulation, Inter-rater agreement, Large vessel occlusion, Modified Rankin Scale, Multiphase CT angiography
- Hacettepe University Affiliated: Yes
Abstract
Purpose: Cerebral collateral circulation is a key determinant of functional outcome in AIS with LVO. We aimed to compare intra- and interobserver agreement among the Menon, ASITN/SIR, and Calgary collateral grading systems on mCTA and to assess their association with clinical outcomes. Methods: In this retrospective study, mCTA images from 108 patients with anterior circulation LVO were independently assessed for collateral status by two radiologists using the Menon, ASITN/SIR, and Calgary systems; intra-observer agreement was assessed in 72 patients. The PCA-MCA component of the Calgary score was analyzed separately as Posterior-CS. All scores, except Calgary, were analyzed as 3-category versions. Agreement was assessed using weighted kappa, PABAK, and ICC. The association with mRS at 90-day was assessed using ordinal logistic regression and ROC analysis in 104 patients. Results: Intra-observer agreement was substantial to almost perfect for the Menon (κ = 0.636–0.914) and ASITN/SIR scores (κ = 0.727–0.943), and excellent for Calgary (ICC = 0.931). Inter-observer agreement was moderate to almost perfect for the Menon (κ = 0.462–0.815) and ASITN/SIR (κ = 0.469–0.813), and good for Calgary (ICC = 0.865). Posterior-CS demonstrated moderate intra-observer (κ = 0.474) and poor inter-observer agreement (κ = 0.187). All collateral scores were significantly associated with 90-day mRS (p ≤ 0.039), except 3-category Posterior-CS (p = 0.073). Higher scores predicted favorable outcome (mRS 0–2) and decreased mortality (mRS 6). Conclusion: The Menon and ASITN/SIR systems demonstrated substantial to almost perfect intra-observer and moderate to almost perfect inter-observer agreement, while Calgary demonstrated excellent intra-observer and good inter-observer agreement. All three systems demonstrated comparable prognostic performance in AIS with LVO. Higher collateral scores predicted favorable outcomes and decreased mortality.