In 1969, Suzuki et al. proposed the traditional classification method of MMD based on DSA, which is still regarded as the gold standard for diagnosis of MMD by many studies. However, both from our questionnaire results and literature review, the limitations of Suzuki grading at the beginning of the design and the irrationality in the process of use have gradually aroused the thinking of the majority of neurosurgeons.
The interobserver reliability of the questionnaire results
Most neurosurgeon want to work through the use of a variety of clinical grading suzuki, such as "condition" (61.88%), "diagnosis MMD" (61.29%), there are more than half of the neurosurgeon will suzuki grading for "academic discussion" (57.60%), which can be good, traditional classification methods of MMD, seems to be no need for the "gold standard" too much to discuss.
However, our survey results show that most neurosurgeons (73.61%) have questioned the Suzuki grading, the focus of which is "unclear boundaries between grades" and "inconsistent grading between the left and right cerebral hemispheres". Many literatures believe that clear clinical grading is helpful for neurosurgeons to make better treatment plans for patients with MMD and is also helpful for the analysis of prognosis(Wang, Qian et al. 2016, Rosi, Riordan et al. 2019, Deng, Gao et al. 2018). If neurosurgeons question whether the traditional grading method is competent for this role, it may have a negative impact on the treatment and prognosis of patients. In addition, with the deepening of researches on unilateral MMD, there is sufficient evidence to show that the two sides of the damaged cerebral hemispheres of some patients with unilateral MMD can present different grading signs, which was not considered at the beginning of the design of Suzuki grading(Church, Bell-Stephens et al. 2020, Hayashi, Horie et al. 2014, Gatti, Torriente et al. 2021).(Table2)
We set the Suzuki grading test questions, no one question answer more than half of the respondents choose. As can be seen from Figure 2 and Figure 3, since there is no clear quantitative standard, most interviewees can hardly distinguish the degree of posterior cerebral circulation stenosis in cases 2, 4 and 7. Therefore, when interviewees judge Suzuki grading in cases 2, 4 and 7, there is almost no difference between grade 4 and grade 5.Similarly, the Suzuki grading did not clearly quantify the extent of the proliferation and dissipation of smoky blood vessels. In cases 3 and 6, it was also difficult for respondents to make a judgment on levels 3 and 4. (Fig2\Fig3)
The purpose of setting up the access question is to include a questionnaire of subjects with basic diagnostic ability of MMD. The imaging findings presented in this topic were highly consistent with the Suzuki grading's Level 3 definition, and subjects only needed to identify poorly developed anterior cerebral arteries, middle cerebral arteries, and sufficiently significant smoky vessels. However, in the original summary of the questionnaire results, only less than 50% of the 466 respondents chose Level 3, and no less than 30% of the respondents believed that it met the definition of Level 4.The fourth and seventh test questions we set showed almost the same signs as the access question, with no more than half of the respondents choosing level 3. Obviously, because there is no clear quantitative standard for Suzuki classification, the degree of hyperplasia of smoky blood vessels and the degree of internal carotid artery branch stenosis cannot be accurately assessed. Perhaps this is why the answers to our quiz questions are so widely distributed. (Table3)
Literature review of grading methods for MMD
Among the numerous literatures, the questions raised by researchers on the traditional grading methods of MMD have a high coincidence with the results of our questionnaire. "unclear boundaries between grades" and "inconsistent grading between the left and right cerebral hemispheres", the former more reflected in our literature review, most of the research puts forward a new grading method were think traditional grading methods for MMD, lack of quantifiable indicators to grading of imaging signs, these studies are also by introducing a new concept or imaging methods, rebuilding or improving grading system, most of their point of view has been clinical test and verify. Studies that review one hemisphere as a unit also raise the same questions that we found. Most of them use double-blind case reviews conducted by senior experts to make it easier to diagnose and grade both hemispheres as a whole(Table5). We found that neurosurgeons of different ages and fields of expertise had little difference in grading cases using the Suzuki grading system(P>0.05), but there were significant differences in questions raised by neurosurgeons of different ages(P<0.05). Therefore, we believe that there is little difference in the level of knowledge of the Suzuki grading method among the broad neurosurgeon community, perhaps because of the defects in the Suzuki grading system itself.
Our literature review also shows that there are more studies suggest that the traditional grading method of MMD clinical practicability, lack of the practical mainly includes: analysis of prognosis, explain the illness, to predict risk factors, our investigation, according to the results of clinical application, is the purpose of the neurosurgeon use grading system, perhaps the traditional grading methods for MMD don't have advantage and authority in this field. With the development of imaging technology, many researchers are not satisfied with only using a single, nonlinear method that cannot reflect hemodynamic changes to diagnose such a complex and progressive disease as MMD(Yamamoto, Okada et al. 2018, Ladner, Donahue et al. 2017, Ha, Choi et al. 2019). Moreover, the traditional grading method is based on invasive imaging methods, which may bring many complications. Numerous studies on the new grading of MMD have introduced a variety of more comprehensive, more linear, non-invasive imaging methods, and have introduced more quantifiable hemodynamic indicators and imaging markers. These new grading methods not only reconstruct or improve the existing grading system, but also get better clinical validation(Table5).
New progress in grading methods of MMD
In view of the limitations of Suzuki grading or DSA in the diagnosis of MMD, such as "the boundary between different grades is not clear", some studies based on a large number of patients with MMD image data, summarized and analyzed the easily found imaging markers or features, and finally reclassified the MMD grade according to the degree. Although the degree of stenosis is difficult to measure in the intracranial vessels with progressive ischemia, a clear cutoff point can be found. M. Czabanka et al. classified a large number of DSA images based on the presence or absence of stenosis/occlusion lesions, smoky vessels, and intracranial and extracranial compensation, and regarded DSA as only a basic objective basis, combining the presence or absence of MRI signs of ischemia, hemorrhage, atrophy, and whether cerebral vascular reserve volume (CVRC) reflected the phenomenon of blood steal(Czabanka, Pena-Tapia et al. 2011). The new method is no longer the highest DSA as the only or the weight of diagnosis, and proposes the concept of partial hemodynamic despite the increasing number of clinical examination project would bring something extra to the whole process of diagnosis and treatment of workload, but perhaps as MMD diagnosis efficiency, neurologic symptoms assessment of patients with MMD risk of a new type of MMD grading method(Jiang, Yang et al. 2018, Cho, Jo et al. 2017). Moreover，the Berlin grading system, Czanbanka, put forward again to verify the reliability of using a variety of imaging techniques in the diagnosis of MMD, this kind of method based on magnetic resonance imaging (MRI) are of the utmost importance, single photon emission computed tomography (SPECT), and other means of hemodynamic relationship between damage and the incidence of cerebral ischemic lesions, the study is in the process of into the case is based on a single hemisphere unit, so that each hemisphere is likely to have different grading(Teo and Steinberg 2020, Yu, Zhang et al. 2020, Czabanka, Boschi et al. 2016). Berlin grading system, is considered the beginning of the design may appear "the brain hemispheres inconsistent grading", only to begin the design idea is better than the two hemispheres as a unified whole, more importantly, the study may be first consider differences in the incidence of target population, when as an object in the study to east Asian patients with MMD validation to make the corresponding changes(Teo, Furtado et al. 2020, Kashiwazaki, Akioka et al. 2017).
Some researches believe that Suzuki grading, as a traditional grading system for MMD, has a high popularity worldwide. Therefore, it is expected to design a new grading system based on the six grades of Suzuki grading. Lian Duan et al. summarized and reviewed the imaging data of a large number of patients with idiopathic MMD, reclassified the characteristic structure of MMD on the anatomical level, and established the disease severity grade with a relatively clear number stratified. In this study, based on the collateral network of posterior cerebral artery, anterior cerebral artery and middle cerebral artery in the delayed venous stage, the presence of retrograde blood flow in the above regions and blood supply in the cortical boundary region were analyzed, and all kinds of conditions were classified into the new type of MMD collateral circulation grading system on a grading of 1-12 based on the corresponding Suzuki grading system. Anatomical abnormalities reflected in DSA by patients with MMD are the most intuitive and representative abnormalities. Explore the changes and differences of compensatory vessels in patients with various types of MMD, which may better explain why patients with the same Suzuki grade have different neurological symptoms, that is, solve the problem of "grading is not consistent with clinical symptoms". Although the main and branch vascular abnormalities and various brain parenchymal abnormalities in patients with MMD should be taken into account by the new grading system, studies based on collateral circulation are relatively basic(Rosi, Riordan et al. 2019, Ladner, Donahue et al. 2017, Hwang, Cho et al. 2020, Liu, Han et al. 2019).
Most neurosurgical clinicians routinely use DSA and MRI as the diagnosis and grading means of MMD. However, studies in recent years have shown that the use of multimodal imaging means can enable clinicians to obtain more risk prediction information, and can also better summarize the pathogenesis factors and symptoms of MMD. Based on noninvasive arterial and venous phase, for example, CTP - Sis of MMD grading system, not only has the advantage of noninvasive, convenient, and in the evaluation of unilateral MMD in or out of the blood group MMD has the superiority, such research, with the help of the collateral vessels in blood both in arterial and venous phase flow more slowly, and is verified with the help of the DSA technique and revision, with lower layers spiral CT scanner to rapid and continuous scanning, finally to difference in the level of regional cortical blood vessels are damaged, the damaged brain hemispheres are divided into more interested in area, The severity of the disease was classified by ischemia range, ischemia duration and radiographic integrity in these areas. Such studies, although not completely out of the suzuki grading or the restrictions of DSA technology, but for the traditional checking method provides a different perspective, MMD and postoperative risk assessment is helpful for clinical doctors, neural function damage assessment, and a kind of noninvasive and efficient way of checking for the general clinical neurosurgery is a very ideal, it may also be the future major trends of diagnosis and grading of MMD and focus(Kathuveetil, Sylaja et al. 2020, Fan, Khalighi et al. 2019, Lin, Kuo et al. 2019, Nishizawa, Fujimura et al. 2020, Xue, Peng et al. 2019).
Limitations and disadvantages
Our questionnaire distribution scope is relatively limited, 341 of the answers passed the admission question we set , which seems a little weak for a large population of MMD like China. However, neurosurgeons especially full-time neurosurgeons in the field of cardiovascular disease, even in China is very few, and, this survey focused by many authoritative hospital neurosurgery almost covers all over the world the highest incidence of MMD area, namely the northern coastal areas and the central plains region, this greatly reduced the investigation and research of sampling error and system error. We did not validate the new grading methods or improved methods in the literature review, and the validation methods in the literature only included prospective validation on a small grading and sample size. Therefore, we cannot completely abandon the traditional grading method of MMD, nor can we completely apply the new grading method independently.