Related Experiment Video
Updated: Aug 25, 2026

Optimized Management of Endovascular Treatment for Acute Ischemic Stroke
Published on: January 18, 2018
Association of Early M2 Pulsatility Index Ratio With Outcomes After Successful Endovascular Thrombectomy
Seyed Behnam Jazayeri1, Cattien Phan1, Yasaman Pirahanchi1
1Department of Neurosciences, Comprehensive Stroke Center, University of California, San Diego, California, USA.
Background And Purpose:
We investigated whether early transcranial Doppler (TCD) ultrasound asymmetry indices could serve as bedside predictors of hemorrhagic transformation (HT) and 90-day outcome after successful endovascular thrombectomy (EVT).
Methods:
Patients with successful reperfusion (thrombolysis in cerebral infarction [TICI] ≥2b) after M1-segment EVT who underwent bilateral TCD within 8 h post-EVT were included. Ipsilateral-to-contralateral ratios for mean flow velocity (MFV) and pulsatility index (PI) were calculated for M1 (segment of the middle cerebral artery) and M2 (segment of the middle cerebral artery) segments. Multivariate logistic regression identified independent predictors of HT and poor 90-day outcome.
Results:
Among seventy patients (mean age 69.9 ± 6.7 years; 58.5% male), TCD was performed 5.3 ± 1.6 h after EVT. HT occurred in 45 patients (64.2%). Independent predictors of HT were younger age (p = 0.012), higher baseline NIHSS (p = 0.025), lower M2 PI ratio (p < 0.0001), and higher M2 MFV ratio (p = 0.010). Parenchymatous hemorrhage was predicted only by lower M2 PI ratio (p < 0.001). Poor 90-day outcome occurred in 35.7% and was associated with older age (p = 0.027), atrial fibrillation (p = 0.002), lower M2 PI ratio (p = 0.004), and higher M1 MFV ratio (p = 0.043). An M2 PI ratio ≤0.79 was associated with poor 90-day outcome with 89% specificity and 80% sensitivity (area under the curve [AUC] 0.91).
Conclusions:
Early M2 PI ratio was associated with PH-type HT and poor 90-day outcome; however, given the small sample size, wide confidence intervals, and lack of external validation, these findings remain hypothesis-generating rather than ready-to-use prognostic tools. Larger studies are needed to confirm these preliminary observations.
