Background: The Pittsburgh Response to Endovascular therapy (PRE) score incorporates patient’s age, initial NIHSS and pre-treatment CT ASPECTS to predict good outcome in anterior circulation large vessel occlusion (LVO) patients. Further validation of this tool is needed before clinical applicability.
Objective: To externally validate the PRE score in a randomized trial cohort of anterior circulation LVO patients.
Methods: Patients in IMS3 who had LVO involving the intracranial internal carotid artery (ICA), M1 or M2 middle cerebral artery (MCA) confirmed either by CT or conventional angiography and had a documented pre-randomization CT ASPECTS, baseline NIHSS and 3-month mRS were included. The PRE score was calculated as follows: Age (years)+ 2xNIHSS - 10xASPECTS. The predictive power of the PRE score for good outcome (3-month mRS 0-2) was compared to THRIVE, SPAN and HIAT-2 scores. Rate of achieving mRS 0-2 was compared across previously published PRE score risk categories.
Results: 407 patients in IMS3 (Endovascular arm: 319 Control arm: 88) had ICA, M1 or M2 occlusions. The cohort had mean age of 66±12 years, median NIHSS 18 [IQR 14-21], median ASPECTS 8 [IQR 6-9] and 38.2% achieved good outcome (mRS 0-2). Age (p<0.001), NIHSS (p<0.001) and ASPECTS (p<0.001) were independent predictors of good outcome. PRE score had good predictive power for mRS 0-2 (AUC=0.73) and Barthel Index≥95 (AUC=0.71) regardless of the treatment arm (AUC=0.72 in both) or reperfusion status. The predictive power for good outcome was similar to HIAT2 (p=0.9) and THRIVE (p=0.10) and slightly superior to SPAN (p=0.06) (Fig 1A). In the endovascular arm, successful reperfusion predicted good outcome (OR 3.1 p<0.001) independently of the PRE score. Rates of good outcome in PRE score risk groups were comparable to the original PRE score derivation cohort (Fig 1B).
Conclusions: Our results confirm the validity of the PRE score as a prognostic tool in anterior circulation LVO stroke patients.