TE is well-known as a life-threatening complication of idiopathic NS. Our patient presented headache and impaired consciousness, and TE was suspected based on elevated thrombotic markers such as D-dimer. It has been reported that thrombosis may occur even in asymptomatic patients in NS. For example, PE was reported to be asymptomatically detected by ventilation perfusion scintigraphy in 27% of children in NS remission [8]. Zhang et al. revealed that subclinical PE occurred in 28% of pediatric NS patients without respiratory symptoms by dual-energy CT pulmonary angiography [9]. Therefore, NS may have venous thrombosis such as PE, even if they are asymptomatic; thus careful evaluation of several TE factors, including thrombotic markers such as D-dimer, is required. To our knowledge, there are no reports having investigated whether patients with NS are asymptomatic for DVT or CVST. The International Pediatric Nephrology Association guidelines mention that there is insufficient evidence to recommend routine prophylactic antithrombotic therapy during the nephrotic state in children and do not clarify the criteria for antithrombotic therapy for NS if preventive antithrombotic therapy is needed [10]. In our case, although dipyridamole was administered during the nephrotic state, TE occurred. The TE event in our patient was considered to result from hypercoagulation due to intravascular dehydration and hypoalbuminemia caused by NS relapse and long-term PSL treatment with high doses due to the timing of the first and second relapses being close. Thus, because NS patients have a high risk for TE events, it is expected that consensus standards will be established in the future for appropriate prevention and treatment of TE in NS patients.

We considered that important factors in the treatment of this patient were reducing the risk of thrombus formation by NS relapse and PSL use. It is necessary to select appropriate immunosuppressants to prevent NS relapse. We did not perform a renal biopsy for a histological diagnosis, because TE occurred during the second relapse, and antithrombotic therapy was initiated. However, our patient had a good response to PSL at the first onset and all relapses, with her SI of proteinuria showing highly selective, suggesting a minimal change nephrotic syndrome as a histological diagnosis. As standard treatments for frequent relapsing NS of minimal change disease, cyclosporine, MMF, and RTX are broadly used. Some studies reported that cyclosporine caused vascular endothelial cell damage and promoted thrombi generation [11]; thus, we considered that cyclosporine is not appropriate for a patient who experienced thrombosis. MMF and RTX treatment is highly effective for frequent relapsing NS that can reduce the frequency of relapse [12]. Therefore, we selected MMF and RTX in this case, and successfully achieved to prevent NS relapse.

Our patient was complicated by multiple CIs, thus other important managements were to consider the underlying pathophysiology of multiple CIs and to prevent recurrence of CI in the future. Generally, the main causes of multiple cerebral infarctions include obvious cardiogenic embolisms such as atrial fibrillation, atherothrombotic brain infarction, and lacunar infarction due to thrombotic occlusion of cerebral arterioles. However, none of them were related to our case. Hypercoagulable state due to NS and PSL use cannot be sufficient to cause multiple CIs. Thus, our case was classified as embolic stroke of undetermined source (ESUS) and we searched other underlying risks and found PFO. The ESUS is a subgroup of nonlacunar cryptogenic ischemic strokes in whom embolism was the likely stroke mechanism. Potential causes of ESUS generally include such as venous thrombosis with right-to-left shunt considering paradoxical cerebral embolism (PCE), arteriogenic emboli, minor-risk potential cardioembolic sources like valvular heart diseases, covert paroxysmal atrial fibrillation, and occult malignancy [13]. Therefore, we assumed the possibility of PFO being related to multiple CIs and decided to perform transcatheter PFO closure. We searched venous embolisms besides CVST. The patient was already under antithrombotic therapy at the time of investigation, which may have affected the result. Regarding PE investigations, only contrast-enhanced CT was carried out instead of pulmonary scintigraphy and it may lack the accuracy of examinations. Although the investigation had been performed in these conditions, we detected no embolisms other than CVST and concluded that the patient had no other potential risks of venous embolisms. There are rare reports of brain-to-brain embolism, where cerebral embolism is caused by intracerebral venous thrombosis through the right-to-left shunt [14, 15]. In this case, brain-to-brain embolism caused by CVST through PFO was also suggested as one of the causes of multiple CIs.

There are indications for the treatment of PFO-causing cryptogenic CI. PCE with PFO has previously been treated with antiplatelet drugs; however, in 2017, catheter therapy for PFO was reported to be more effective [16,17,18]. However, these studies were conducted on adults. Cerebral embolism through a PFO is very rare in pediatric patients; therefore, there are only a few cases in which catheter treatment has been indicated for PFO. Wawrzyńczyk et al. reported that seven children aged 12–16 years underwent successful transcatheter PFO closure, and no procedure-related complications were observed [19]. Although it is necessary to consider appropriate age, body size, risks, and period of postoperative antithrombotic therapy prior to transcatheter PFO closure in children, PFO should be actively managed by transcatheter closure in specific adolescents and young adults. In this case, the patient had a high risk of TE, which is one of the major reasons for choosing transcatheter PFO closure.

CI is a serious complication in patients with NS. The conceivable pathogenesis of multiple CIs is various, including right-to-left shunt through PFO, in addition to the hypercoagulability due to NS and PSL. In conclusion, we suggest that it is important to investigate and manage underlying risks of CI, such as PFO, besides preventing the relapses of NS by aggressive treatments using MMF and RTX, in patients with NS.