In this study, the average ratio of male to female was 1:1.22 (male 134: female 164), with 57 children and 241 adults, and the incidence of preoperative seizure was approximately 8.72%. This is basically consistent with the results of previous studies. For the 941 cases of MMD in Japan, the female: male ratio was 1.98:1. The two peaks of age of onset were 5–9 years and about 40 years. The incidence of seizure was 4%[10]. For children with MMD, Ma reported in 2017 that the proportion of men and women with seizure was approximately 1:0.87, and there was no difference in the sex ratio, with an incidence of 18.1%[11].
Patients with MMD report transient ischemic attack (TIA) before operation, but it is difficult to distinguish from some epileptic attacks in the form of expression. It was found that the occurrence of cerebral infarction before operation and new infarction after operation were the risk factors of early seizure in patients with MMD after revascularization. There are two mechanisms that may lead to early and late epileptic seizures after neurosurgery. One mechanism is the extravasation of blood vessels. Iron released from hemoglobin reacts with hydrogen peroxide in the surrounding tissues to produce free radicals, which may lead to early epileptic seizures and the formation of epileptic foci. Hydrogen peroxide reduces the inhibition of γ-aminobutyric acid (GABA)-mediated cortical and thalamic neurons. Free radicals can promote the excitability of thalamic cortical circuits by changing the neurotransmitter mediated by β-aminobutyric acid. Repeated exposure to free radicals will lead to the formation of epileptic foci. Another mechanism may be through disorder of ion balance in the cell membrane caused by ischemia or hypoxia. The decrease of high energy reserve, such as adenosine triphosphate under the condition of ischemia or hypoxia, further leads to the disorder of ion balance on the cell membrane; this is because the ratio of intracellular and extracellular K + concentration decreases, whereas the ratio of sodium ion concentration increases. This imbalance reduces the hyperpolarization of neurons that may cause transient or early seizures. The imbalance of sodium ions is mainly mediated by the outflow of Ca2+-independent glutamate during ischemia[12].
Choi Ji et al found that 4/7 of the patients with MMD had seizure with hypoperfusion, [13] and stroke and hypoperfusion would lead to cerebrovascular reactive disorder, followed by hypoxia of brain tissue and seizure. Blood flow reconstruction can improve the perfusion of brain tissue and prevent seizure[13]. According to Mikami et al, moyamoya seizure is a typical type of seizure after stroke. However, in a previous study, the severity of vascular stroke shown by MRA was not related to seizure[14]. The prognosis of MMD epileptics with abnormal CT findings was poor[15]. Ma et al suggested that preoperative cerebral infarction could not predict the recurrence of seizure[11].
Ma et al. found that the duration of epileptic attack before operation was an independent risk factor for seizure recurrence after revascularization of children with MMD[11], and revascularization should be performed immediately after the diagnosis of seizure. In our study, there was no correlation between preoperative seizure history and postoperative seizures. In this group of patients, we found that the preoperative status of seizure in patients with MMD may be related to the early postoperative seizures as shown by one-way ANOVA. A few other studies have reported on seizure caused by MMD, but in other studies, it was related to cerebrovascular diseases. According to Englot et al, the seizure time was less than 1 year, which is a predictor of seizure in patient with cavernous malformation[16]. According to Falero and León, the short seizure time of arteriovenous malformation (AVM) was statistically significant with class I after operation[17]. Liu et al believes that the long duration of AVM indicates poor prognosis[18] .
In this group of cases, the seizure forms of patients with seizure before operation include simple partial seizures (SPS), complex partial seizures (CPS), and generalized tonic–clonic seizures (GTCS). Among the 23 epileptic patients before operation, 12 had SPS attack and 6 had CPS attack. Some of them had high incidence, which may be related to focal infarction or ischemic focus before operation. However, we found that patients with persistent state of seizure before operation are more likely to have early seizure after undergoing revascularization. Few researchers examined the relationship between seizure form and seizure after operation for MMD, while in other cerebrovascular disease studies, seizure form is related to the prognosis of patients; moreover, there are also contrary findings. GTCS attack in cavernous malformation (CM) patients indicates poor prognosis[19–22]. According to Falero and León, GTCS attack in AVM indicates a good prognosis[17]. In patients with MMD, there was no correlation between attack pattern and prognosis.
In the present study, we did not find that the choice of operation mode was related to the early postoperative seizure. For patients with ischemic MMD and high risk of preoperative infarction, we chose EDAS operation with shorter operation time to reduce the risk of anesthesia and operation. If there are receptor vessels with appropriate thickness and flow, the bypass operation is more feasible. Therefore, there is a certain bias in the selection of operation method, which is often compared with decision-making. According to Choi et al, EDAS is safe and effective to prevent postoperative seizure[13]. Ma 2017 found no difference between the two surgical methods (direct / indirect) in the prevention of postoperative seizure[11]. There are still many concerns associated with the specific operation. In this group, the blood supply reconstruction operation mode involves cutting the skin craniotomy and milling the bone flap along the superficial temporal artery. Some units use the frontotemporal craniotomy and mill the bone flap. However, it is unclear (1) whether there is any influence of the size of bone window exposing the brain tissue area on the postoperative seizure, (2) whether there is stimulation of the connective tissue around the superficial temporal artery, (3) whether there is traction and contusion of brain tissue during the operation, and (4) whether there is subarachnoid hemorrhage during the operation. This information may have an impact on the postoperative seizure, but no detailed data can be used for in-depth statistics.
Hyperperfusion of brain tissue after STA-MCA bypass can cause transient neurological dysfunction, including postoperative seizure[23]. However, there are similar biological mechanisms between postoperative seizure and postoperative hyperperfusion syndrome[12, 24]. Postoperative high perfusion syndrome can cause headache, eye and facial pain, epileptic seizure, and focal neurological dysfunction secondary to brain edema. The pathological mechanism may be the impairment of cerebrovascular self-regulation function, which makes it difficult to regulate the blood flow from the superficial temporal artery to the middle cerebral artery after bypass surgery. In addition, oxygen free radicals produced during reperfusion after revascularization may damage cerebral vessels, leading to the occurrence or aggravation of high perfusion. The intrinsic response of some revascularization, such as the increase of vascular permeability, may be related to high perfusion after revascularization in MMD patients. Long-term cortical ischemia can induce the overexpression of angiogenic factors and extracellular matrix proteins, thus promoting the formation of new blood vessels and increasing the permeability of blood vessels[25].
According to the new National Institute for Health and Care Excellence guidelines, carbamazepine and lamotrigine are the first-choice drugs for focal seizure[26]. Most of the cases of moyamoya occur in young women. For the long-term control of moyamoya seizure, lamotrigine and levetiracetam can be used as the first-choice drugs in consideration of tolerance and family burden[14].
The present study has several limitations: first, although the total number of patients with MMD included in this study is quite large, the number of seizure cases after operation is insufficient. To confirm our conclusion, a larger sample size and randomized design of future research are needed. Second, it is impossible to collect more detailed information about the operation details and techniques. Perhaps more detailed observation indicators should be taken in the future research design, in order to obtain more detailed conclusions.