Pretreatment brain CT perfusion thresholds for predicting final infarct volume in distal medium vessel occlusions
Vivek Yedavalli,Omar Hamam,Alireza Mohseni,Kwan Chen,Richard Wang,Hye-Young Heo,Jeremy Heit,Elisabeth Breese Marsh,Raf Llinas,Victor Urrutia,Risheng Xu,Fernando Gonzalez,Greg Albers,Argye Hillis,Kambiz Nael
DOI: https://doi.org/10.1111/jon.13142
Abstract:Background and purpose: Quantitative CT perfusion (CTP) thresholds for assessing the extent of ischemia in patients with acute ischemic stroke (AIS) have been established; relative cerebral blood flow (rCBF) <30% is typically used for estimating estimated ischemic core volume and Tmax (time to maximum) >6 seconds for critical hypoperfused volume in AIS patients with large vessel occlusion (LVO). In this study, we aimed to identify the optimal threshold values for patients presenting with AIS secondary to distal medium vessel occlusions (DMVOs). Methods: In this retrospective study, consecutive AIS patients with anterior circulation DMVO who underwent pretreatment CTP and follow-up MRI/CT were included. The CTP data were processed by RAPID (iSchemaView, Menlo Park, CA) to generate estimated ischemic core volumes using rCBF <20%, <30%, <34%, and <38% and critical hypoperfused volumes using Tmax (seconds) >4, >6, >8, and >10. Final infarct volumes (FIVs) were obtained from follow-up MRI/CT within 5 days of symptom onset. Diagnostic performance between CTP thresholds and FIV was assessed in the successfully and unsuccessfully recanalized groups. Results: Fifty-five patients met our inclusion criteria (32 female [58.2%], 68.0 ± 12.1 years old [mean ± SD]). Recanalization was attempted with intravenous tissue-type plasminogen activator and mechanical thrombectomy in 27.7% and 38.1% of patients, respectively. Twenty-five patients (45.4%) were successfully recanalized. In the successfully recanalized patients, no CTP threshold significantly outperformed what is used in LVO setting (rCBF < 30%). All rCBF CTP thresholds demonstrated fair diagnostic performances for predicting FIV. In unsuccessfully recanalized patients, all Tmax CTP thresholds strongly predicted FIV with relative superiority of Tmax >10 seconds (area under the receiver operating characteristic curve = .875, p = .001). Conclusion: In AIS patients with DMVOs, longer Tmax delays than Tmax > 6 seconds, most notably, Tmax > 10 seconds, best predict FIV in unsuccessfully recanalized patients. No CTP threshold reliably predicts FIV in the successfully recanalized group nor significantly outperformed rCBF < 30%.