Comprehensive Assessment of Arterial, Tissue, and Venous Collaterals for Evaluating the Infarct Growth Rate: The Multimodal Collateral Score

BACKGROUND: Collaterals are crucial factors that influence the infarct growth rate (IGR). We aimed to determine whether a comprehensive multimodal collateral score (MCS), incorporating collateral assessment at the arterial, tissue, and venous levels, is associated with functional independence and provides incremental prognostic value over individual IGR metrics. METHODS: Arterial collaterals were assessed using multiphase CT-angiography (mCTA), ranging from 0 to 5; tissue-level collaterals were evaluated on CT-perfusion by the hypoperfusion intensity ratio (HIR), ranging from 0 to 1; and venous outflow was measured on mCTA with the comprehensive venous outflow (CVO) score, ranging from 0 to 8. These metrics were used to develop the multimodal collateral score (MCS), which encompasses good mCTA collaterals (score 4-5), HIR (< 0.4), and CVO (range 4-8). The MCS ranges from 0 to 3, where 0 indicates no favorable collaterals, 1 means only one good collateral metric among the used ones, 2 signifies just two favorable collateral metrics differently combined among the used ones, and 3 represents the sum of all good collateral metrics. Multivariable logistic and ordinal regression analyses identified clinical and radiological factors independently associated with clinical outcomes (modified Rankin scale 0-2 at 90 days, ordinal mRS at 90 days). RESULTS: Overall, 615 acute ischemic stroke patients treated with thrombectomy within 24 h from symptom onset between January 2017 and December 2023 were included: the median age was 77 [IQR = 67-84], 320 (52%) were females, and 307 (50%) achieved functional independence. The MCS grades of 2 (aOR 2.92 [95% CI = 1.70-5.03]) and 3 (aOR 5.35 [95% CI = 3.03-9.44]) were independently associated with functional independence. The MCS was also independently associated with 90-day ordinal mRS (adjusted common odds ratio [acOR] 1.73 [95% CI = 1.51-2.01] per 1-point increase). The AUC of the MCS for predicting functional independence was 0.71 (95% CI = 0.66-0.75), which was superior to the other IGR metrics (p < 0.001). CONCLUSION: MCS was a strong, independent predictor of favorable functional outcomes and demonstrated significantly superior prognostic performance compared with individual IGR metrics.