This study provides new insight into the “one-stop” noninvasive evaluation of adult MMD with 18F-FDG PET/MRI. The interrelationships of parameters derived from PET and MRI scans, as well as their association with DSA findings, were highlighted.
Brain CT and MRI are currently the main noninvasive imaging diagnostic methods for MMD assessment and have been considered initial alternative imaging modalities that take the place of traditional DSA for making therapeutic decisions. In our study, the MR portion of PET/MRI demonstrated structural abnormalities, including cerebral infarction lesions and ischaemic white matter lesions, which can reflect the severity of disease progression. Thirty hemispheres with infarction lesions were identified, mainly in the mid and late stages, but most patients had lower Fazekas scores. The brain's hemodynamic status, the degree of arteriopathy seen on angiograms and the severity of clinical symptoms frequently differ due to the complicated interactions of various influencing factors. Therefore, an evaluation of compensatory mechanisms contributing to the sustainability of CBF and tissue viability is needed.
Leptomeningeal collaterals are crucial in preserving cerebral perfusion in MMD patients. The degree of proximal stenosis may not be the only determinant of disease severity if the compensation through collateral vessels is adequate. The “ivy sign” on FLAIR is a presentation of leptomeningeal collaterals and has been reported to be related to clinical severity and hemodynamics[10]. According to the semiquantitative analysis of the manifestation of the “ivy sign” in different Suzuki’s stages and the comparisons with collaterals shown on DSA, we found that there were 10.00% regions in the early stage, 56.88% in the mid-stage and 28.00% in the late stage with the “ivy sign”; there were 14.17% regions in the early stage, 70.31% in the mid-stage and 35.00% in the late stage with leptomeningeal collaterals on DSA. Thus, both the number of abnormal cerebral regions with the “ivy sign” and collaterals on DSA in the mid-stage and the late-stage was greater than in the early stage.
The “ivy sign” can reflect the degree of regional cerebral vascularization changes in the downstream tissue to some extent, and it may be useful for predicting the severity of disease progression. These findings were also referred to in previous studies that demonstrated a correlation between a higher total “ivy sign” score and the likelihood of stroke[11]. However, we could not always relate the “ivy sign” to the presence of leptomeningeal collaterals because the “ivy sign” was not detected in all of the regions with collaterals shown on DSA. The “ivy sign” is likely to reflect slow retrograde or turbulent flow in the engorged pial collateral arteries through leptomeningeal anastomoses. In our study, the number of regions with a positive “ivy sign” (41.30%) was statistically less than those with DSA-positive findings (51.30%), which may be associated with the influence of different blood flow statuses. In addition, when analyzed in different Suzuki’s stage subgroups, this phenomenon was also noticed in the mid-stage. Although there was a difference in the ability of FLAIR on the MR portion of PET/MRI and DSA to identify leptomeningeal collaterals in the mid-stage and total stage patients, the “ivy sign” score can positively reflect the degree of leptomeningeal collaterals when taking the direct findings of DSA as a reference. This is not consistent with the research of Annick Kronenburg et al.; however, the subjects enrolled in that study consisted of both MMD and MMS cases, and a limited number were also noticed[18]. Investigation of the “ivy sign” can help characterize the origins of the collateral supply and identify the vascularization of hypoperfusion areas[12–18].
In addition to the assessment of structural abnormalities, variables investigated to reflect the hemodynamics directly are more crucial for the clarification of the severity of adult MMD. Objective and quantitative evaluation of the territory-specific CBF in ischaemic stroke can help neurosurgeons identify patients who could benefit from surgical revascularization[4]. Nevertheless, it is still a challenge to acquire noninvasive perfusion images, particularly when the disease pathology interacts with the imaging system to introduce inaccuracies in the CBF values. The most common application of ASL for cerebral perfusion imaging is in stroke[19]. In particular, ASL can assist in identifying critical areas of intravascular or cortical slow flow with subsequent cerebral hypoperfusion that, if left untreated, could develop into irreversible infarction lesions[20]. In our study, there were 10.00% of regions in the early stage, 51.56% of regions in the mid-stage and 73.00% of regions in the late stage with hypoperfusion on ASL. There were 5.83% regions in the early stage, 55.00% regions in the mid-stage and 76.00% regions in the late stage with decreased irrigation on DSA. The distribution of abnormal perfusion regions in different stages was similar to that of regions with abnormal leptomeningeal collaterals shown on FLAIR and DSA, and the “ivy sign” score was related to the degree of ASL-CBF decline. The “ivy sign” is an effective noninvasive variable for indicating a hemisphere with poor hemodynamics at risk of ischemia with varying performance in different stages, which is consistent with prior research results[21].
Nevertheless, the correlation between the ASL-CBF score and the “ivy sign” score was weak, with a correlation coefficient r = 0.250 (p<0.05). Although the “ivy sign” can be used as a sign to inform the leptomeningeal collateral circulation status in addition to regional hemodynamics, it is still unreliable for assessing the status of regional CBF based on the structural changes revealed by MRI. The use of ASL for the assessment of MMD patients has been reported in various studies with different applications[12, 22, 23–24], and stroke is by far the best-established application of ASL. In some previous studies, a comparable performance of ASL with 15O-H2O PET for assessing perfusion deficits in MMD was demonstrated[23, 25–26]. Meanwhile, repeatability is another benefit of using ASL to define the CBF reserve through a combination of acetazolamide challenge/rest examinations in both preoperative assessments and when monitoring disease progression[24].
However, there are some limitations of ASL, which may lead to inaccurate CBF values. A delay in the arrival times of the labeled arterial water spins caused by the presence of steno-occlusive disease and the presence of collateral moyamoya vessels can result in an underestimation of CBF values[22, 27]. The labeled blood does not reach the imaging slices before the start of image acquisition, resulting in a loss of perfusion signal due to long transit times. Alternatively, if arterial transit times are increased, there is also an overestimation of the CBF value[28]. In addition, the ASL acquisition method and the implementation of flow-crushing gradients to suppress signals from reaching the intravascular labels in arterial blood before imaging can influence the visibility of artifacts.
In further analysis, we found no significant difference between hypoperfusion identified by ASL MRI and decreased irrigation demonstrated by DSA, not only in all patients but also in patients in the mid and late stages. This was then verified by the positive correlation between the ASL-CBF score and the DSA irrigation score. ASL-CBF on MRI can be used as a potential objective indicator for assessing the perfusion status of regional cerebral tissue non-invasively. Other studies reported that standard ASL overestimated the extent of hypoperfusion in patients[24]. Therefore, it was recommended to use the "ivy sign" for assisting ASL-CBF on MRI to assess the status of vascularization and hemodynamic alterations of local brain tissues by verifying each other.
Functional evaluation before brain surgery is also significant and required because a large proportion of stroke survivors suffer from motor deficiencies and functional cognitive impairments even after effective treatment. However, traditional imaging modalities are unable to determine the functional consequences of the underlying pathologic abnormalities, and some previous studies failed to find any evidence of changes in the cortical functional connectivity after stroke as the relevant mechanism for promoting motor recovery using resting-state fMRI[29–30]. Ischaemic brain injury is one of the most prominent causes of disability in stroke patients, and at present, it has been well established that ischaemic brain injury can involve dysfunction of the brain networks demonstrated by 18F-FDG PET[31–32]. 18F-FDG PET is increasingly used in neurological diseases due to its unique and irreplaceable advantage of assessing changes in brain function as well as functional connectivity at the molecular level, and it can add more diagnostic information, verified not only in animals[32–34] but also in clinical experiments[35]. Meanwhile, cumulative energy consumption and presumed steady resting state can be demonstrated by 18F-FDG PET, the findings of which are a critical complement to provide valuable insights into the pathophysiology of disorders.
In our research, further analysis showed that the numbers of cerebral regions with hypometabolism in the mid and late stages were significantly higher than those in the early stage. The numbers of cerebral regions with hypometabolism on PET were less than those with hypoperfusion on ASL in total and all different stages or those with decreased irrigation on DSA in total stage, mid-stage and late stage with significant differences, illustrating that regional cerebral tissues with the same DSA vascularization or hemodynamic status may have completely different glucose metabolism changes. These results are consistent with the research results of Walberer et al., which suggested the presence of viable tissues through a compensatory increase in glucose uptake and phosphorylation in the hypo-perfused tissue[8]. 18F-FDG PET was utilized as a standard modality to evaluate the disease outcome in some previous studies[36]. It can be used to determine the severity of functional abnormalities and to identify salvageable tissues accurately. Further analysis verified that the degree of regional cerebral metabolic changes was positively related to the extent of both hypoperfusion on ASL and decreased irrigation on DSA. Hypoperfusion on ASL may be an independent indicator for predicting hypometabolism on PET with a significant difference, which may be useful for centers without PET/MRI.
Although the noninvasive evaluation of hemodynamics and metabolism makes 18F-FDG PET/MRI outstanding in the assessment of MMD, it is necessary to highlight several limitations of our work. First, there may be a potential selection bias towards races and regions due to this study only including a small number of patients from one single center. Second, the ASL-CBF score, “ivy sign” score and PET score are all acquired with visual assessment methods rather than through absolute quantitative evaluation. Additionally, prospective studies with long-term PET/MRI follow-up with larger sample sizes are needed to illuminate the relationship between the characteristics of altered glucose metabolism and the prognosis of MMD patients.