Tmax volume can predict clinical type in patients with acute ischemic stroke

Abstract Objective Endovascular therapy (EVT) is performed for acute ischemic stroke (AIS) with large vessel occlusion (LVO), however, the treatment strategies and clinical outcomes differ between cardiac embolism (CE) and intracranial arteriosclerosis‐derived LVO (ICAS‐LVO). We analyzed whether the time‐to‐max (Tmax) volume derived from perfusion imaging predicted clinical classification in AIS patients before EVT. Methods Consecutive AIS patients with anterior circulation LVO evaluated by automated imaging software were retrospectively identified. Patients were classified into a CE group and an ICAS‐LVO group, and parameters were compared between groups. Results Thirty‐nine patients were included and Tmax volume and Tmax > 6 s volume/Tmax > 4 s volume were significantly greater in the CE group than in the ICAS‐LVO group (Tmax > 4 s volume: 261 mL vs. 149 mL, p = .01, Tmax > 6 s volume: 143 mL vs. 59 mL, p = .001, Tmax > 6 s volume/Tmax > 4 s volume: 0.59 vs. 0.40, p < .001). Multiple logistic regression analysis indicated an association between clinical classification and high Tmax > 6 s volume/Tmax > 4 s volume (p = .04). Conclusion The Tmax volume derived from perfusion imaging predicts the clinical classification of AIS patients before EVT.


Image analysis
Perfusion imaging was conducted using a 3.0-T MR system (Signa HDx ® ; GE Healthcare) or a 64-row detector CT system (Revolution Maxima ® ; GE Healthcare). Either MRP or CTP was taken at the discretion of the doctor. A single-shot spin-echo EPI-DWI sequence was used with the following parameters: b-values, 1000 s/mm 2 ; slice thickness, 5 mm. MRP images were acquired with a gradient-echo echo-planar technique. Gadoteridol was power-injected through a peripheral intravenous catheter at doses standardized by patient body weight (0.2 mL/kg body weight, to a maximum of 20 mL) at 2-4 mL/s, immediately followed by a 20-mL saline flush at the same rate. CTP was performed two times at the parietal and basal sides. Forty milliliters of iopamidol (370 mg iodine/mL) was power-injected through a peripheral intravenous catheter at 4-6 mL/s at one time and immediately followed by a 20-mL saline flush at the same rate. Estimates of the volume of the ischemic core and penumbral regions from CTP or MRI diffusion and perfusion scans were calculated with the use of RAPID software.
Ischemic core volumes were based on a RAPID relative cerebral blood flow (CBF) lesion volume using a ≺30% threshold or a RAPID DWI lesion volume with an apparent diffusion coefficient (ADC) threshold of <620 × 10 −3 mm 2 /s. The volume of penumbral volumes was estimated using the Tmax perfusion parameter with a threshold of >6 s (Tmax > 6 s) on both CTP and MRP . Incorrect estimation of the core regions by RAPID was manually reanalyzed and cases with poor images due to motion artifacts or low cardiac output were excluded (Laughlin et al., 2019 ). Cases were then classified to a CE group or ICAS-LVO group. Clinical type was diagnosed by imaging (cerebral angiography, MRI/MR angiography, and CT/CT angiography

Outcome measures
The following values were measured by analysis of RAPID: DWI- In addition, the collateral score was measured in cases in which cerebral angiography was performed (Higashida et al., 2003), Comparisons between patients in the CE and ICAS-LVO groups were performed for the variables shown in Tables 1 and 2.

Statistical analysis
The descriptive statistics of baseline variables were calculated and are reported as the median and interquartile range (IQR) for non-normal

RESULTS
During the study period, 168 patients underwent perfusion imaging and analysis using RAPID. Of these, 116 patients were excluded for  of AF, a lower median HbA1c, a higher median BNP, a higher median D-dimer, a shorter median last-known-well time to perfusion imaging time, and a lower median DWI-ASPECTS than patients in the ICAS-LVO group (Table 1). Patients in the CE group had a larger median core volume, Tmax volume, and Tmax ratio than patients in the ICAS-LVO group (Table 2). Patients in the CE group had larger median Tmax > 6 s volume/Tmax > 4 s volume than patients in the ICAS-LVO group for every location of occlusion, although there was no significant difference (Table 3). In thirty-eight patients who received cerebral angiography or EVT, patients in the ICAS-LVO group were more likely to have intermediate or excellent collateral flow, whereas those of CE group were more likely to have poor collateral flow (p = .007) ( Table 2). Although no significant difference was observed between every collateral score group due to the small number of cases, the higher the collateral score, the

Case 2
A patient in their 70s presented with hypertension, diabetes, and no AF. NIHSS on arrival was 6. MRA showed occlusion in the M1 segment of the right MCA ( Figure 5A). MRI and MRP showed a core volume of 13 mL and a Tmax > 6 s volume of 50 mL. DWI-PWI mismatch was positive ( Figure 5B). Fluid-attenuated inversion recovery (FLAIR) imaging showed hyperintense vessel sign in the M2 segment of the right MCA ( Figure 5C). Tmax > 6 s volume/Tmax > 4 s volume was 0.31 and Tmax > 8 s/Tmax > 4 s was 0.11 ( Figure 5D). MRA showed severe stenosis of the right MCA before onset ( Figure 5E). This case was diagnosed as ICAS-LVO.

DISCUSSION
The present study showed Tmax volumes and collateral scores were associated with clinical type. In particular, Tmax > 6 s volume/Tmax > 4 s volume was significantly higher in the CE group than Several studies have previously reported on clinical type diagnosis.
Clinical history, including low median baseline NIHSS score, male sex, hypercholesterolemia, and smoking, may possibly support ICAS-LVO, while AF suggests CE (Jin et al., 2020). Although AF is useful for clinical type diagnosis, it may not be recognized at the arrival. One study reported that BNP level was useful in distinguishing CE from non-CE (Igarashi et al., 2019). Imaging features including no hyper-dense vessel sign on noncontrast enhanced CT, no susceptibility vessel sign on susceptibility-weighted MRI, truncal-type occlusion on CTA, occlusion in the M1 segment of the MCA, scattered or borderzone infarct pattern on DWI, arterial calcification and tortuosity on the angiography, and residual arterial stenosis after EVT are considered suggestive of ICAS-LVO (Park et al., 2019). A jet-like appearance on angiography in 20.7% (34 patients) and noted that a jet-like appearance offered an image marker for ICAS-LVO in a retrospective study of 164 patients (Jin et al., 2020). However, the identification of clinical type before EVT by these imaging features remains challenging, particularly in patients with complicated clinical situations. We therefore focused on analysis using RAPID, which can provide data before EVT. Acute diagnostic imaging has become increasingly important with the expansion of eligibility criteria for EVT in AIS to tissue-based selection and vari-ous imaging biomarkers for predicting symptom onset time have been reported in recent years Kishi et al., 2022;Ma et al., 2019;Nogueira et al., 2018). DWI-PWI mismatch or CBF-PWI mismatch by the analysis of RAPID has been used to judge eligibility for EVT in AIS, including in patients with unknown symptom onset time, and has expanded the recommendations of eligibility for EVT . Perfusion imaging has become an essential tool for deciding between EVT and medical care in AIS (Demeestere et al., 2020;Laughlin et al., 2019).
During analysis by RAPID, Tmax volumes seemed to show different characteristics between the ICAS-LVO group and CE group, so we focused on Tmax volumes in the present study. The present results revealed that patients in the CE group had a larger median Tmax volume and Tmax ratio than patients in the ICAS-LVO group. The perfusion parameter Tmax > 6 s is often used to estimate tissue that is likely to progress to infarction if reperfusion is not achieved. Mismatch between core volume and Tmax > 6 s volume estimates salvageable tissue (Gregory W. Albers, 2018;Inoue, 2019). Several studies have reported on associations between Tmax volume and collateral flow. Galinovic et al. (2018) reported that CBF/Tmax > 4 s volume ratio and  (Tu et al., 2015). Suk Jae Kim et al. (2009)  In the future, we plan to verify whether Tmax volume can predict the clinical type of posterior circulatory occlusion.

CONCLUSIONS
Tmax volumes and Tmax > 6 s volume/Tmax > 4 s volume obtained from perfusion imaging can predict clinical type in patients with AIS prior to EVT.

AUTHOR CONTRIBUTIONS
IN and FK conceived of the study, analyzed and interpreted the data, drafted the manuscript, and edited the manuscript. All authors have reviewed and approved the final version of the manuscript.

FUNDING
This work did not receive any grants from funding agencies in the public, commercial, or not-for-profit sectors.

CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT
Data are available upon reasonable request.