Single Centre Experience on Decision Making for Mechanical Thrombectomy Based on Single-Phase CT Angiography by Including NCCT and Maximum Intensity Projection Images – A Comparison with Magnetic Resonance Imaging after Non-Contrast CT

Article information

J Korean Neurosurg Soc. 2020;63(2):188-201

Publication date (electronic) : 2019 October 29

doi : https://doi.org/10.3340/jkns.2019.0131

Myeong Soo Kim^,1, Gi Sung Kim²

¹Department of Neurosurgery, Dong Kang Medical Center, Ulsan, Korea

²Department of Radiology, Dong Kang Medical Center, Ulsan, Korea

Address for reprints : Myeong Soo Kim Department of Neurosurgery, Dong Kang Medical Center, 239 Taehwa-ro, Jung-gu, Ulsan 44455, Korea Tel : +82-52-241-1114, Fax : +82-52-241-1180, E-mail : 1004@jbnu.ac.kr

Received 2019 May 30; Revised 2019 July 23; Accepted 2019 September 1.

Abstract

INTRODUCTION

Early recanalization of occluded vessels in acute stroke patients improves clinical outcomes and reduces mortality [2,3,16,36]. In patients with large vessel occlusion (LVO), the use of recombinant tissue plasminogen activator (r-tPA) only does not lead to high recanalization probability [28]. In recent trials, a variety of instruments have been used and the high recanalization results have been confirmed in those trials. The development of instruments for mechanical thrombectomy (MT) has also permitted us to expect better recanalization [18,37,46]. Numerous institutions are working on realizing early recanalization by attempting to reduce the lagging time to initiate the procedure after the patient is transferred to the emergency room (ER) and to improve the effective procedure for recanalization [1,25,30,45].

Preliminary examinations are essential for decision making for endovascular treatment (EVT). Numerous studies confirmed the use of r-tPA after non-contrast Computed Tomography (NCCT) and performing EVT after confirmation of computed tomography angiography (CTA), CT perfusion (CTP), and magnetic resonance image (MRI) with perfusion images [3,6,17,18,39]. In actual clinical practice, the procedure is performed after confirmation of CTA (or CTP) after NCCT, but there exists a tendency to decide on the procedure after MRI examination with perfusion images [15,25,27].

In CTA protocol, imaging, thin-section NCCT, and maximum intensity projection (MIP) images can identify the presence of hemorrhage, the exact thromboembolic occlusion site, and the structure of the extra-intracranial artery including the aortic arch for the endovascular procedure, and can be evaluated the pial collaterals for prediction of prognosis [15,32].

In this study, we investigated the clinical outcomes of patients who had undergone MT after a single-phase CTA imaging including NCCT and MIP images to reduce the lag time between the patient arrival and the initiation of the procedure. Compared with the group which underwent MRI after NCCT, our new attempt not only led to a reduction in the time to start the procedure but also led to good results in terms of prognosis. Therefore, we propose the importance of this method as the only preliminary examination for decision making.

MATERIALS AND METHODS

This retrospective study was approved by the Institutional Review Board (No. 2019-06-03).

Patient selection

A retrospective cohort study was performed by including patients who underwent MT for LVO performed at Dong Kang Medical Center. A total of 157 patients who underwent MT at our institution from April 2015 to March 2019 were divided into two groups based on the examination performed prior to MT : for all stroke patients who were eligible for inclusion criteria, some of them were allocated into a group (CTA) who had taken CTA with NCCT and MIP as a sole examination prior to procedure at the time of arrival on the ER and the others were allocated into a group (NCCT+MRi) who underwent MRI including perfusion images after NCCT. In the CTA group, CTA imaging with NCCT and MIP was performed for patients with all the neurological symptoms who transferred to the ER. Blood sampling for the decision of r-tPA administration was performed during intravenous (IV) line acquisition for imaging to collect rapid blood test results. After securing the IV line, the patient was immediately transferred to the imaging laboratory for imaging. During the CTA imaging, thin-section NCCT and CTA images were taken sequentially and were confirmed in real time. The prepared MIP images were confirmed within minutes for accurate diagnosis of lesion and decision making. The CTP was not performed (Fig. 1).

Fig. 1.

Single-phase CT angiography images including NCCT and MIP images. A 62-year-old woman with a left M1 MCA occlusion. A : Thin-section NCCT (no hypodense lesion, no hemorrage). B : MIP intracranial axial image (arrow, left M1 MCA occlusion). C : MIP intracranial coronal image (arrow, left M1 MCA occlusion). D : MIP intracranial sagittal image (arrow, left M1 MCA occlusion). E : Contrast-enhanced CT image. F : Axial image showing the aortic arch (arrow, left CCA; origin of Brachiocephalic trunk, left CCA, and left subclavian artery). G : MIP extracranial coronal image (arrow, no stenosis of left CCA-ICA and tortuosity of left cervical ICA). H : MIP extracranial sagittal image (arrow, tortuosity of left mid CCA). CT : computed tomography, NCCT : non-contrast computed tomography, MIP : maximum intensity projection, MCA : middle cerebral artery, CCA : common carotid artery, ICA : internal carotid artery.

The inclusion criteria were as follows : patients who arrived within 4.5 hours after symptom onset, lacked hypodense lesion on CTA image who arrived within 8 hours even after 4.5 hours of symptoms onset, with thromboembolic occlusion of large vessel (internal carotid artery [ICA], middle cerebral artery [MCA], anterior cerebral artery [ACA], posterior cerebral artery [PCA], basilar artery [BA], vertebral artery [VA]), and patients with National Institutes of Health Stroke Scale (NIHSS) 4 or higher.

Patients in the NCCT+MRi group underwent NCCT imaging immediately after they arrived at the ER, followed by MRI imaging with perfusion images. The inclusion criteria for MT applies a universal indication : patients with LVO, patients who arrived within 8 hours after symptom onset, NIHSS 4 or higher, patients without evidence of cerebral hemorrhage, patients with a perfusion-diffusion mismatch in MRI with infarct core volume less than 1/3, and patients with perfusion defect more than 2/3 [39].

Patients with cerebral hemorrhage, patients without LVO, and patients with severe hypodensity on NCCT were excluded before MT through CTA examination. In addition, even if there was an LVO and if the physician could not proceed to decision making with CTA alone, MRI examination with perfusion images was performed and the patients who were not eligible were excluded from the study group. Patients who were transferred, who developed stroke during hospitalization, and who underwent other examination than CTA before the procedure were excluded. Of 157 patients, 49 patients performed NCCT+MRi as pre-examination of MT, 53 patients CTA, 30 patients CTA+MRI, four patients NCCT alone, one MRI alone, one took CT+CTA, nine tested at other hospitals, and 10 patients had in-hospital strokes. The 16 patients excluded from the NCCT+MRi group were four patients with onset-to-door time (OTD) of more than eight hours, seven with unclear onset time (wake-up, un-witnessed daytime strokes), and five patients with less than 4 NIHSS score. The 19 patients excluded from the CTA group were 14 patients with OTD from eight hours to less than 12 hours, and five patients with less than 4 NIHSS score.

IV r-tPA was administered to eligible patients in both the groups following guidelines for the early management of acute ischemic stroke (AIS) [13].

Image protocol and analysis

Single-phase CTA

An Aquilion CT scanner (Toshiba Medical Systems, Tokyo, Japan) was used. After the acquisition of nonenhanced CT data, contrast-enhanced CTA was performed. CTA was performed by axial acquisition from the aortic arch to the vertex using a 100 kVP, 250 mA, 0.5 mm detector collimation, 240 mm D-FOV with a rotation speed of 0.5 sec/rotation. A total of 60 mL of contrast material (iopamidol, Pamiray 370; Dongkook Pharmaceuticals, Seoul, Korea) was administered intravenously using a power injector at 4.0 mL/sec followed by 50 mL normal saline chase at a rate of 4.0 mL/sec. The axial contrast-enhanced images were reconstructed at 3-mm thickness with no intervals. Thick-section axial, coronal and sagittal MIP at 20-mm thickness and 3-mm intervals were also reconstructed. Reconstructed images of thin-section NCCT and CTA were taken from aortic arch to vertex, and at the same time, the images were transferred to a picture archiving and communication system (PACS). For identifying the intracranial lesion, MIP images were reconstructed from the skull base to the vertex in the order of axial, coronal, and sagittal. Extracranial MIP images were reconstructed from the aortic arch to the skull base in the order of coronal and sagittal images in order to identify common carotid artery (CCA), ICA, and VA. Subsequently, the images were uploaded to the PACS sequentially as soon as they were reconstructed so that they could be confirmed in real time.

Non-contrast CT and MRI

NCCT and MRI imaging were acquired by following the usual protocols. NCCT was performed with standard protocols using an Aquilion CT scanner (Toshiba Medical Systems). Axial NCCT images were obtained with 120 kV, 300 mA, and 5-mm-section-thickness reconstruction. MRI imaging with MR angiography and perfusion imaging were performed on either 1.5T or 3T scanners (Avanto 1.5T or Skyra 3T; Siemens, Erlangen, Germany) using standard departmental and vendor protocols.

Pial collateral circulation was defined on the CTA as good, moderate, or poor (using Menon pial arterial filling scale, score 5 and 4 : good, 3 and 2 : moderate, below 1 : poor) [32] (Table 1, Figs. 2-4).

Table 1.

Pial arterial filling score within the symptomatic ischemic territory using single-phase CT angiography (menon pial arterial filling scale^*)

Fig. 2.

A 78-year-old woman with a right distal M1 MCA occlusion. OTD 240 minutes, NIHSS 16, DTI 16 minutes, DTP 82 minutes, mTICI 3, 90 days mRS 1. A : MIP intracranial axial image (arrow, right distal M1 MCA occlusion). B and C : Good pial collaterals (score 5). D : Axial image showing the aortic arch (arrow, Brachiocephalic trunk; origin of Brachiocephalic trunk, left CCA, and left subclavian artery). E : MIP extracranial coronal image (arrow, no significant stenosis of right CCA to cervical ICA). F : MIP extracranial sagittal image (arrow, tortuosity of proximal right cervical CCA). MCA : middle cerebral artery, OTD : onset-to-door time, NIHSS : National Institutes of Health Stroke Scale, DTI : door-to-image time, DTP : door-to-puncture time, mTICI : modified thrombolysis in cerebral infarction, mRS : modified Rankin Scale, MIP : maximum intensity projection, CCA : common carotid artery, ICA : internal carotid artery.

Fig. 3.

A 70-year-old man with a left mid M1 MCA occlusion. OTD 22 minutes, NIHSS 12, DTI 6 minutes, DTP 70 minutes, mTICI 2b, 90 days mRS 2. A : MIP intracranial axial image (arrow, left mid M1 MCA occlusion). B and C : Moderate pial collaterals (score 3). D : Axial image showing the aortic arch (arrow, left CCA; origin of Brachiocephalic trunk, left CCA, aberrant left VA, and left subclavian artery). E : MIP extracranial coronal image (arrow, bovine origin of left CCA). F : MIP extracranial sagittal image (no significant stenosis of left CCA to cervical ICA). MCA : middle cerebral artery, OTD : onset-to-door time, NIHSS : National Institutes of Health Stroke Scale, DTI : door-to-image time, DTP : door-to-puncture time, mTICI : modified thrombolysis in cerebral infarction, mRS : modified Rankin Scale, MIP : maximum intensity projection, CCA : common carotid artery, VA : vertebral artery, ICA : internal carotid artery.

Fig. 4.

A 63-year-old man with a right proximal M1 MCA occlusion. OTD 48 minutes, NIHSS 18, DTI 14 mintues, DTP 67 minutes, mTICI 3, 90 days mRS 5. A : MIP intracranial axial image (arrow, right proximal M1 MCA occlusion). B and C : Poor pial collaterals (score 0). D : Axial image showing the aortic arch (arrow, Brachiocephalic trunk; origin of Brachiocephalic trunk, left CCA, and left subclavian artery). E and F : MIP extracranial coronal and sagittal images (no significant stenosis of right CCA to cervical ICA). MCA : middle cerebral artery, OTD : onset-to-door time, NIHSS : National Institutes of Health Stroke Scale, DTI : door-to-image time, DTP : door-to-puncture time, mTICI : modified thrombolysis in cerebral infarction, mRS : modified Rankin Scale, MIP : maximum intensity projection, CCA : common carotid artery, ICA : internal carotid artery.

RESULTS

From April 2015 to March 2019, 157 MT patients met the inclusion criteria : 34 in the CTA group and 33 in the NCCT+MRi group. Univariate analysis of the demographics, risk factor, occlusion site, pial collaterals, and IV r-tPA usage is shown in Table 2. The mean age of the patients was noted as 71.1±10.9 years (mean±standard deviation) in the CTA group and 67.0±11.5 years in the NCCT+MRi group; 62% in CTA group and 67% in NCCT+MRi group were males. Sites of occlusion were as follows : ICA (13/34 in CTA group vs. 12/33 in NCCT+MRi group); MCA (13/34 vs. 16/33); ACA (3/34 vs. 0/33); PCA (1/34 vs. 1/33); BA (2/34 vs. 4/33); and VA (2/34 vs. 0/33). IV r-tPA was given to 73% of patients in the CTA group and 60% of patients in the NCCT+MRi group.

Table 2.

Baseline characteristics, received r-tPA in the CTA and NCCT+MRi group

The time delay comparison between the two groups is shown in Table 3. We found a significantly shorter delay of door-to-puncture time (mean, 86±22.1 vs. 176±47.5 minutes; p<0.01). Also, a significantly shorter total stroke imaging duration was observed in the CTA group (mean, 13±6.8 vs. 93± 30.8 minutes; p<0.01). A significantly shorter mean interval from the onset of symptoms to arterial puncture was observed (mean, 195±128.0 vs. 314±157.6 minutes; p<0.01). There was no significant difference between the groups from the onset of the symptoms to imaging initiation. Furthermore, there was no significant difference in duration from an arterial puncture to reperfusion time between the two groups. Baseline characteristics, reperfusion result, and clinical outcomes are shown in Table 4.

Table 3.

Comparison of duration (minutes) between the CTA and NCCT+MRi group

Table 4.

Reperfusion result and clinical outcomes in the CTA and NCCT+MRi group

Reperfusion result of mTICI ≥2b was 100% (34/34) in the CTA group and 94% (31/33) in the NCCT+MRi group, and the result of mTICI 3 was 74% (25/34) in the CTA group and 73% (24/33) in the NCCT+MRi group. Baseline NIHSS score was 18.6±5.4 vs. 15.0±5.6, which improved to 7.2±7.4 vs. 8.1± 7.5 by the time of discharge. Favorable functional outcomes (mRS score ≤2 at 90 days) was reported as 68% (23/34) in the CTA group and 60% (20/33) in the NCCT+MRi group, with occurrence of three deaths in the CTA group (cardiac problem 1 and poor pial collaterals 2) and two in the NCCT+MRi group (pulmonary problem 1 and ICH 1). There were three cases of procedure-related complications (ICH 2, SAH 1) in only NCCT+MRi group. Regarding clinical outcomes, although there were significant differences in baseline NIHSS, there was no significant difference between the mRS in both the groups at 90 days (Table 4).

DISCUSSION

Recanalization of acute cerebral artery occlusion is a procedure that significantly improves a patient’s quality of life [9]. In many studies, the main prognostic factors are indicated as early recanalization, good collateral, small core infarct, low NIHSS, and young age [12,16]. Early recanalization is the most important factor for a good prognosis [2,3,16,36]. The IV r-tPA within 4.5 hours of stroke is an effective treatment for AIS [17]. However, low recanalization rates are still limited when the occlusion area is in the proximal region and the narrow therapeutic window [28]. To overcome these limitations, EVT including chemical thrombolysis and various MT were introduced within 6–8 hours of the occurrence of symptoms [20].

MT of large intracranial vessels has been mentioned in most of the previous studies [43]. Currently, MT is a widely used method and many studies have shown good results of early recanalization due to low complications [26]. Generally, the main instrument of EVT for AIS is stent-retrievers, which shows the efficacy and favorable clinical outcomes in several comparative randomized trials [38]. Several recent reports have described manual aspiration techniques known as manual aspiration thrombectomy, forced suction thrombectomy, or the a direct aspiration first pass technique techniques, and demonstrated favorable results on recanalization rates, cost efficiency, procedural duration, and functional outcomes [19,22,49].

To achieve early recanalization, great efforts have been made to minimize the time to arterial puncture after patients arrival at the hospital. Reducing the time required for preliminary examination for decision making can also be used as an approach to reduce the time for early recanalization [41,45]. These efforts have been well discussed in several studies on NCCT+CTA, CTP, and MRI [3,6,16,21]. The five positive trials published in 2015 led to guideline changes in many countries in the process of updating their recommendations for EVT for ischemic stroke [7] (Table 5). In the present study, we proposed that single-phase CTA including NCCT and MIP images, as a sole examination for decision making prior to the procedure, can significantly reduce door-to-puncture time, which can contribute greatly to early recanalization.

Table 5.

Characteristics of endovascular thrombectomy trials [7]

Multimodal MRI shows various useful parameters in determining treatment routes of AIS [24]. This technique is especially useful if stroke diagnosis is uncertain (stroke-mimics), stroke onset time is uncertain (un-witnessed daytime strokes or wake-up), or there are clinic-imaging mismatches. However, despite potential benefits of MRI, there are practical problems limiting use of MRI in AIS [48]. One of the major problems is longer scan duration compared to CT imaging. In general, decision-making of thrombolysis based on MRI extended doorto-needle time, but clinical results were more favorable [9,50]. Thus, use of MRI in treating acute stroke is clinically practical and feasible [23,35]. Conversely, considering the “time is brain” in the hyperacute phase of ischemic stroke, minimizing evaluation time required to determine treatment could expand the benefits of thrombolysis [24].

In our study, 30 of 157 patients who performed MT tested additional MRI after CTA. Additional MRI imaging was required in case decision-making for MT was not possible with CTA alone; if it was not possible to confirm an occlusion site that matched symptoms (high NIHSS score but not sure of occlusion, patients with LVO but low NIHSS score), in cases the onset time was unclear (symptoms onset within 12 hours to 24 hours, wake-up, un-witnessed daytime strokes). If the occlusion site could be confirmed (old lesion or new lesion), and it was similar to symptoms caused by internal disease, it was known there was a stroke (stroke-mimics). If sufficient evidence was not created for determination by CTA alone, further MRI imaging was required, and it was found that MRI was an effective tool for decision-making of MT [24] (Fig. 5).

Fig. 5.

A 80-year-old man with a right proximal cervical ICA occlusion. OTD 304 minutes, NIHSS 21, DTI 78 minutes, DTP 164 minutes, mTICI 3, 90 days mRS 0. A and B : MIP intracranial axial image (no occlusion, good pial collaterals). C : MRA image (occlusion of ICA and weak flow of MCA). D-F : MRI images (perfusion-diffusion mismatch). G : Left ICA angiography (pre-MT, irrigation of contralateral ACA and MCA). H : Right ICA angiography (post-MT, successful recanalization, mTICI 3). ICA : internal carotid artery, OTD : onset-to-door time, NIHSS : National Institutes of Health Stroke Scale, DTI : door-toimage time, DTP : door-to-puncture time, mTICI : modified thrombolysis in cerebral infarction, mRS : modified Rankin Scale, MIP : maximum intensity projection, MRA : magnetic resonance angiography, MCA : middle cerebral artery, MRI : magnetic resonance image, MT : mechanical thrombectomy, ACA : anterior cerebral artery.

CTA has many advantages over NCCT, such as high sensitivity to LVO detection, visualization of carotid vessels, and tortuosity of aortic arch as part of stroke workup and thrombectomy planning, especially under conditions of difficult vessel anatomy (including femoral artery or radial artery approach) to select an appropriate approach to vessel access [29]. Structural variation in the aortic arch is rarely reported and should be considered in terms of the approach in the procedure [10].

Data from previous trials reveals that clinical outcome improves if patients have a salvageable brain at presentation and undergo early recanalization [5,33,40]. Risk of poor clinical outcomes can increase by approximately 14% each time treatment is delayed every 30 minutes [31]. So, the ideal imaging tool should detect the salvageable brain quickly and reliably [32]. Current imaging techniques include non-enhanced CT, single-phase or multiphage CTA, CTP, and MRI imaging. Single-phase CTA and Multiphase CTA provide clinicians with information on degree and extent of pial arterial filling throughout the brain. Multiphase CTA provides clinicians with multiple time-resolution images, providing a more subtle assessment of pial arterial filling in the normal brain and ischemic brain [32]. Single-phase CTA has a limitation on lack of temporal resolution; so, this modality leads to risk of mislabeling pial arterial filling when compared with multiphase CTA [14,34].

Serum creatinine levels are generally evaluated before performing CTA due to potential risk of contrast-induced acute renal damage. However, many studies have shown that incidence (2% to 4%) of contrast-induced nephropathy (>25% increase in creatinine) in AIS patients who have undergone emergency CT scanning or emergency EVT is low and is not significantly different from other conditions [4,47]. There is controversy as to if it is justifiable to perform CTA (or DSA) without baseline serum creatinine in hyperacute ischemic stroke [4]. However, considering the rapid neuronal death rate (2 million per minute) and patients’ outcomes, harm caused by delayed treatment in AIS patients is greater than the benefit of preventing low incidence, and mainly reversible, renal damage. So, in cases of possible stroke with no history of clinical evidence of renal impairment or severe dehydration, doctors do not need to wait for laboratory results before CTA [8,25].

CTA after NCCT before thrombectomy to confirm LVO is widely performed in many institutions based on protocols used for LVO stroke thrombectomy in large clinical trials recently announced [44]. All the positive trials required LVO evidence using non-invasive angiography (CTA). Especially when visualized using thin-slice NCCT, the apparent high density artery appears to have high sensitivity and speciality to acute occlusive thrombus [42]. However, the CTA for occlusion is considered a standard procedure and should be extended from the aorta arch to allow evaluation of proximal vascular approaches to plan an endovascular approach [27,39].

Some of the institutions have attempted to selectively conduct CTA in cases with severe stroke involving frequent LVO. However, patients with mild clinical symptoms still have a significant incidence of LVO (approximately 10% in those with NIHSS <6) and these patients are at a risk of further clinical deterioration [11]. The routine imaging method using both NCCT and CTA is simple, and uniform workflow is likely to result in faster and higher quality acquisition due to the familiarity of radiologists [7].

In the present study, it was observed that avoiding extra vascular images (selective CTA after NCCT or MRI after NCCT) in patients with CTA (including NCCT and MIP) images showing occlusion sites significantly reduced the door-topuncture time. The observed improvement in time in our patient cohort was approximately 90 minutes. Reduced reperfusion time is related to improvement in functional outcomes and by optimizing the time delay before arterial puncture, improvement in time delay could be achieved [5,31,33,40]. Although CTA group did not produce better clinical outcomes than NCCT+MRi group in our study, we achieved as favorable functional outcomes as the NCCT+MRi group. As with recent attempts, it was possible to conclude that reperfusion time is an effort to reduce evaluation time and that it is valuable to consider it as the only preliminary examination for MT.

Our study has several limitations. The data was collected retrospectively without randomized allocation, and the sample size was relatively small and associated with selection bias. The decision of EVT or extra imaging was made by the neurointerventionist on-call doctor, although we had an indication of MT. The usage of variable devices, like the size of the sheath, guiding catheter, intermediate catheter, micro catheter, guidewire, aspiration catheter or stent, and size of materials were decided based on the cases. Furthermore, various limitations, such as prognosis change according to r-tPA usage status, selection of patients with treatment in the NCCT+MRi group, more objective NIHSS/mRS assessment, subdivided assessment of clinical outcomes, and determination of initial imaging selection by on-call doctor, should be considered thoroughly. Also, several clinical and radiological factors which could affect the outcomes should be considered. A larger prospective trial is needed to further confirm our findings.

CONCLUSION

A single-phase CTA including NCCT and MIP images was performed as a single preliminary examination to reduce the time from the arrival of the patient to the hospital to the start of the procedure, which not only reduced the time for the initiation of the procedure but also led to good results in terms of prognosis. So, it is appropriate to conclude that the proposed method holds significant value as the only preliminary examination for decision-making of MT in patients with LVO stroke. Although our study did not achieve better clinical outcomes compared to the NCCT+MRi group, we produced good results like the NCCT+MRi group and confirmed that it was a way to significantly reduce reperfusion time.

Article information Continued