In this study, we presented a strategy for persistent feeder obliteration preceding the surgical removal of large hypervascular tumors in the lateral ventricle. Devascularization following feeder obliteration was achieved in all six patients, resulting in decreased blood loss during the subsequent tumor removal in the majority of the patients. Postoperative hemorrhage had not occured, and the outcomes were relatively favorable.
Risks of obliterating choroidal arteries
The normal mean diameter of the AChA is 1.24 mm (range: 0.4–3.4 mm) [12]. The classic clinical symptoms caused by occlusion of the AChA are contralateral hemiparesis, hemisensory loss, and homonymous hemianopia [13]. The AChA is divided into the cisternal and plexal segments, and only the cisternal segment contains the perforating branches, which may be associated with the clinical symptomology [14]. The plexal point observed by DSA has been proposed to identify the entry point into the temporal horn; that is, the point separating the cisternal and plexal segments of the AChA.
Embolization of feeders arising from the AChA can carry the risk of ischemic complications [1, 2], and superselective embolization of choroidal arteries has been reported only in a few case reports of intraventricular meningiomas [15,16,17] and in a few studies of arteriovenous malformations [10, 18, 19]. When considering TAE, the catheter tip must be advanced beyond the plexal point to avoid ischemic complications; however, the risk of procedure-related ischemic or hemorrhagic complications is reportedly as high as 16.7–30.7 % [10, 18, 19]. In those studies, AChAs were dilated enough for deep catheterization, but the safety diameter of catheters was not identified. In our study, TAE was preformed successfully only in one patient in which the AChA’s diameter was 1.8 mm; in five other patients in whom the mean diameter of the AChA was 1.0 mm (range: 0.9–1.2 mm, Table 2), we instead selected microsurgical occlusion after assessing the diagnostic images. Thus, the safety diameter for deep cannulation of the AChA might be 1.5 mm.
The risk of ischemic complications might be avoided by performing microsurgical occlusion of the AChA in the temporal horn. Reported approaches for the temporal horn include the transsylvian, the temporal, and the subtemporal approaches [11, 20]. We mainly selected the subtemporal approach to reduce the risk of postoperative language dysfunction [11]. In the present series, no neurological permanent deficits, including language dysfunction, were observed due to microsurgical occlusion. In a cadaver study [21], it was shown that 38 % of the capsulothalamic artery arises from the first portion of the plexal segment. This anatomical variation could be associated with ischemic complications, even in cases involving the direct obliteration of the AChA. Actually, in our study, postoperative MRI in one patient (Patient 6) detected infarction in the internal capsule, suggesting that this anatomical variation could be associated with ischemic complications.
The LPChA branches to the choroid plexus and the trigone of the lateral ventricle. In addition, it also supplies the crus, commissure, body, and part of the anterior columns of the fornix, as well as the dorsomedial and pulvinar portion of the thalamic nucleus, and a part of the lateral geniculate body [22, 23]. Reports on patients with discrete LPChA infarction are limited, with a visual field defect, typically quadrantanopia or hemianopia, being the primary symptom [22]. Controlling ischemic complications post embolization of the LPChA may be challenging since no angiographic safety point has been identified thus far, unlike the AChA [10]. In our study, small thalamic infarction was detected in all four patients with endovascular feeder occlusion, although the catheter tip was advanced just proximal to the tumor in all patients. This was probably because either the blood flow was stolen towards the tumor and the small branches projecting to the thalamus were invisible, or the LPChA was retrogradely thrombosed. Further studies are warranted to identify the appropriate catheter position for avoiding these types of ischemic events.
Role of feeder obliteration preceding tumor removal
Preoperative endovascular feeder embolization is commonly used for vascular-rich meningiomas, except for those located in the lateral ventricle, to reduce intraoperative blood loss in more devascularized tumors, and also to soften tumors [8, 24]. However, past studies have reported no significant differences in surgical duration, extent of resection, blood transfusion requirements, or measures of morbidity [8, 25]. In the present study, the mean devascularization rate was 69.9 %, and the mean blood loss volume (253 mL) during the tumor removal was much less than the volumes reported in a previous study of intraventricular meningiomas (mean: 530 mL; range: 100–1900 mL) in which feeder obliteration was not performed before the tumor removal [4].
The optimal time interval between feeder embolization and tumor removal remains controversial; it has been suggested that the removal should be performed approximately one week post embolization due to greater softening of the tumor, allowing for easier removal [26], whereas others have recommended within seven days post embolization to avoid the revascularization of tumors [27]. In this study, the mean time interval from the final feeder obliteration procedure to the tumor removal was three days (range: 1–6 days). In two patients (Patients 2 and 3) with extensively devascularized tumors (> 95 %), the tumors softened; in three patients (Patients 4, 5, and 6) with two or more feeder arteries, multiple small pial feeders were observed intraoperatively and the tumors were still very firm, even though all the main feeders had already been obstructed. The presence of small pial feeders may explain the low devascularization rates in tumors supplied by two or more feeders. In a previous case report, tumor volume decreased and the tumor softened 13 months post microsurgical occlusion of feeders arising from the AChA or PChA [9]. Therefore, if a patient’s neurological condition is not critical, it may be better to abstain from treatment until substantial tumor shrinkage is obtained following feeder obliteration in cases where tumors are supplied by multiple feeders.
Trisacryl gelatin microspheres have been demonstrated to achieve more distal penetration and higher subsequent devascularization than similarly-sized polyvinyl alcohol [7]. In our series, in one patient (Patient 2), for whom only TAE of the AChA was performed using microspheres, 97.9 % devascularization was achieved. Likewise, in another patient (Patient 3), for whom microsurgical occlusion of the AChA was performed, 98.6 % devascularization was achieved and the tumor was softened. Microsurgical occlusion of feeders alone also helped to devascularize the tumor, to reduce intraoperative blood loss, and to soften the tumor. Therefore, microsurgical occlusion alone could be an effective and alternative method for feeder obliteration when catheterization of the AChA is difficult.
Approach used for tumor removal and surgical outcomes
The optimal surgical approach for treating tumors in the lateral ventricle remains controversial. The main reported approaches have been described [1,2,3,4,5, 9, 28]. (1) The trans-parietal approach through the superior parietal lobe or intraparietal sulcus is preferably used, since this approach can reduce the risk of injury to the optic radiation, but can be disadvantageous, depending on the tumor location, due to the long distance to the trigone of the lateral ventricle. In this study, in two patients (Patients 5 and 6) in whom the tumor was firm, we abandoned this approach in favor of the wide parieto-occipital corticectomy in a subsequent session. (2) The trans-temporal approach (including the temporo-occipital approach) involves a route through the middle or inferior temporal gyrus. This route provides for a shorter distance to reach the tumor, but may increase the risk of injury to the inferior fibers of the optic radiation and to the language cortices in the dominant hemisphere. We used this approach in one patient (Patient 4) after receiving consent since the patient already had homonymous hemianopia prior to the surgery; the symptom had not resolved by the time of the last-follow up due to injury to the optic radiation. (3) The interhemispheric parieto-occipital precuneus approach involves a route associated with a low incidence of hemianopia, aphasia, and epilepsy. We did not select this approach in any of the six patients since a wide brain retraction was necessary, and due to risks of injury or thrombosis of bridging veins or the sagittal sinus.
In a recent review, the postoperative mortality rate following treatment of intraventricular meningiomas was 4.0 %; most deaths were related to hematomas in the surgical bed, probably due to rich tumor vascularization [6]. In this study, the mortality was 0 % and mRS scores assessed at the last-follow up were ≤ 2 in all patients. Notably, no postoperative hemorrhage had occured in this study, probably due to feeder obliteration prior to the tumor removal; therefore, although the staged treatment was required, preceding feeder obliteration could limit intraoperative blood loss and prevent postoperative hemorrhages, and the treatment outcomes of these patients were relatively favorable.
Limitations
This study has several limitations. First, as large intraventricular tumors are relatively rare, the number of cases we assessed was too small to confirm the genuine efficacy of feeder obliteration preceding the surgical tumor removal. Second, the present study was not randomized, which makes outcome comparisons difficult. However, compared with the outcomes of current treatment options for patients with these types of tumors described in the literature [1, 6], the outcomes observed in this study were relatively favorable. Notably, in this study, no patients had permanent neurological deficits due to feeder obliteration. Further multicenter, prospective studies are needed to verify the efficacy of feeder obliteration preceding direct surgery and/or endovascular treatment of these tumors.