Surgical clipping vs coiling
Through the development of the endovascular technique over time, the success rate of ACoAA occlusion with endovascular treatment has been increased, followed with the increase in the number of the endovascular treatment. Previous studies reported that clipping had more procedure related injury than coiling [OR: 2.17– 24.42], including postoperative infarction [24, 45, 46]. Although coiling has become an efficient treatment for intracranial aneurysms, a higher recurrence rate and retreatment rate are reported than surgical clipping in previous reports [43, 58]. Surgical clipping remains a necessary treatment for an aneurysm with specific characteristics. In our hospital, ACoAA expecting incomplete treatment with coiling, such as aneurysm with 1) broad neck (> 4 mm); 2) large or giant size (≥ 10 mm); 3) unfavorable dome to neck ratio 4) fusiform shape; 5) unstable intraluminal thrombus; 6) perforators which are incorporated into the aneurysm neck; 7) aneurysms in multiple locations; or 8) difficulty in proximal access via endovascular approach, had been treated with surgical clipping. These characteristics of ACoAAs may result in a difficult procedure with clipping; however, it is possible to obtain a better outcome using surgery by the development of microsurgical instruments and microscope, application of ICG-VA, and intraoperative EP monitoring.
Risk factors for postoperative infarction
With demographic data, the patients with hypertension (OR 2.05; p < 0.05) and previous CVA history (OR 2.79; p < 0.05) were independently associated with postoperative infarction. A meta-analysis of association between silent brain infarction and stroke was reviewed [21]. It was reported that even very small cerebrovascular lesions had an association with stroke and mortality. In addition, hypertension is well known as a strong risk factor for stroke [19, 50, 67]. These may also have a high relationship with postoperative infarction. Another hypothesis is that postoperative infarction may be due to the atherosclerotic change of intracranial vessels. Atherosclerosis is known to be one of the major causes of stroke [2]. Additionally, hypertension is a well-known risk factor for intracranial atherosclerosis [1, 66]. Cosar et al. performed an experiment with 12 rabbits with or without atherosclerotic common carotid artery, and evaluated the effect of the temporary clip [14]. The result showed that the microscopic change due to the temporary clip occurred much earlier in the atherosclerotic CCA (in 1 min) than non-atherosclerotic CCA (in 10 min). In our study, the patients with hypertension or previous CVA may have had predisposing intracranial atherosclerosis. Additionally, we usually use temporary clips for most clip surgeries. It may result in a higher tendency of having vessel injury or emboli during manipulation or temporary clipping of parent artery with atherosclerosis, which may need further study. Also, if there are atherosclerotic changes in the aneurysm or its neck, applying a permanent clip on the aneurysm may induce postoperative infarction due to complications, such as shower emboli, slippage of permanent clip, or additional narrowing of inner diameter than outer diameter of parent artery, which are reported in previous studies [3, 13, 51]. These may be the reasons of the postoperative infarction without visible intraoperative complications in the operation field.
The size of the ACoAA is an important factor for the prediction of aneurysm rupture and choice of treatment. An aneurysm with a size less than 7 mm was reported to have a low rupture risk with in ISUIA [68]. However, a high rate of rupture of small cerebral aneurysms has been reported in UCAS [65]. Additionally, it has been reported that the ACoAA has a higher rupture risk than other anterior circulations. In addition, recent studies reported that ACoAAs with a smaller size have a risk of spontaneous rupture [5, 39]. These make it difficult for neurosurgeons to decide whether aggressive treatment is needed for unruptured ACoAAs. Generally, the size of the aneurysm is known to have a relationship with incomplete clipping and intraoperative rupture [15, 28, 56, 61]. Size is also reported to have an association with postoperative infarction. Swiatnicki et al. reported that an intracranial aneurysm with a size more than 9 mm (OR 1.15; p = 0.003) was independently associated with brain ischemia [61]. Additionally, another study evaluated the perforator infarction during surgical clipping of posterior communicating artery aneurysm, and they also showed aneurysm size (OR 1.18; p = 0.02) as an independent predictor [62]. Only one study reported that surgical clipping of larger ACoAAs may come with worse clinical outcomes, which have an mRS score more than 2 [27]. In our study, the mean ACoAA size was 5.21 mm (± 2.38 mm), which is reasonable for treatment. Additionally, size (OR 1.16; p < 0.01) was a significant risk factor for postoperative infarction, especially with aneurysms larger than 5 mm. A large size of aneurysm may obstruct the surgical view and hard to visualize perforators. Moreover, it is also associated with more adhesion with the surrounding structures, including parent and perforating arteries, and parenchyma. These may increase the risk of manipulating the surrounding structures during dissection and clipping, which results in vessel injury. Careful consideration with a large size aneurysm is needed during surgical clipping.
The surgical clipping of a posterior projecting ACoAA is considered to be the most difficult because it is hidden between interhemispheric fissure and can hide ACoA perforators, which originated from the posterior portion of ACoA complex, such as SCA and HThA [10, 25, 31, 38]. However, there are few clinical data for complications with surgical clipping of posterior projecting aneurysm. Proust et al. reported that ACoAAs projecting posteriorly to the axis of the pericallosal arteries had a higher risk of vessel occlusion than anteriorly projecting ACoAAs, which leads to unfavorable outcomes [53]. Similarly, Ivan et al. reported that superiorly and posteriorly projecting ACoAAs have significantly worse outcomes because it requires opening the junctional triangle, which is the intersection between the distal A1 ACA segment and the proximal A2 ACA segment [27]. Our data showed that posterior projection was a strong risk factor for postoperative infarction than any other projection, especially with the pterional approach. More attention should be given for clipping posteriorly projecting aneurysm.
Several studies had evaluated high positioned ACoAAs; however, unified definition has not been determined [20, 32, 33, 52]. In the pterional approach, a high positioned ACoAA is reported to have an association with postoperative anosmia or residual neck [52, 60]. The distance from the skull base to the aneurysm was measured with various modifications. In our study, Measurement was done from the sphenoid planum to the aneurysm dome because we usually secure the whole plane of aneurysm dome before permanent clipping. Additionally, our results showed that a high positioned ACoAA, especially more than 10 mm, is significantly associated with postoperative infarction. To visualize the aneurysm, additional manipulation of the frontal lobe and resection of the rectus gyrus would be needed. It may result in additional cortical injury and bleeding control, leading to vessel injury. Suzuki et al. reported an additional requirement of A2 manipulation and contusion with high positioned ACoAAs, which supports our opinion [60]. In our study, patients who had surgical clipping via an interhemispheric approach had a higher position of the ACoAA than the pterional approached group; however, there had been no significant relationship with postoperative infarction. It has been reported that the treatment of a high positioned ACoAA via an interhemispheric approach was also associated with complete clipping when compared to the pterional approach [33]. Therefore, we suggest the definition of a high positioned ACoAA as an aneurysm located more than 10 mm from the skull base, which should be treated via an interhemispheric approach rather than a pterional approach to avoid postoperative infarction.
Several suggestions were proposed for side selection with the pterional approach. It includes several factors, such as non-dominant hemisphere, projection of aneurysm, dominant A1, and accompanying aneurysms. In our study, an approach from the closed A2 plane showed as an independent risk factor for postoperative infarction (OR 1.98, CI [1.09–3.59], p = 0.024). Considering the anatomical structure of the ACoA complex, an approach from the closed A2 plane requires an additional manipulation of the frontal lobe and A2 due to an obstructed surgical view. Suzuki et al. reported that an approach from the closed A2 plane with a superior directing ACoAA had a higher tendency to have rectus gyrus resection, residual neck, damage of RAH and other perforators, and contusion, which supports our opinion [60]. Hyun et al. also studied 19 superior directing ACoAA, and insisted that a closed A2 plane had a higher incidence of rectus gyrus resection [26]. However, they did not show clinical significance with postoperative infarction, which may be due to the small sample size. Our study showed that an approach from a closed A2 plane has a relationship with postoperative infarction regardless of any projection of the ACoAA, which differs from previous studies and emphasizes the importance of the A2 fork.
Some studies proposed a strategy for side selection by correlating the projection of aneurysm with the A2 fork. They suggested the opening of the A2 plane for superior projection, and closed A2 plane for posteriorly projecting aneurysm [11, 16]. The result of their strategy showed a good clinical outcome with convincing mRS or GOS for more than 88–90.2% of the enrolled patients. Another study showed that the surgical clipping of ACoAAs via supraorbital eyebrow keyhole approach showed no relationship between the A2 plane and postoperative complication [4]. These studies may also be useful to select an approach strategy with preventing postoperative infarction and further evaluation should be needed.
Some studies suggested a right-sided approach, which is believed to avoid retraction of the dominant lobe and provide convenience for most of the right-handed surgeons [16, 57]. However, in our study, there was no significant difference in the postoperative infarction with either approach side. It may be due to our strategy for left-sided approach. We prefer to perform an extended craniotomy to manipulate with convenience for right-sided surgeons and to reduce frontal and temporal lobe retraction. Additionally, it may also reduce the risk of postoperative infarctions. Similar results were also found in previous studies [26, 60].
Clinical outcomes after postoperative infarction
Prognosis of the postoperative infarction was observed to vary with each branch of the ACoA complex. Complications of the postoperative infarction presented as frontal lobe syndrome, amnesia, motor weakness, or pituitary insufficiency. mRS at discharge was the highest in the infarction at A2 cortical branches (mRS = 2.00 ± 1.63), which include orbitofrontal and frontopolar branches, and followed by the infarction at SCA and HthA (mRS = 1.42 ± 0.99).
RAH arises near the A1-ACoA-A2 junction and MLSA originated from the A1 of the anterior cerebral artery [10, 18, 72]. RAH supplies the thalamus, anterior lenticular nucleus, lateral globus pallidus, anterior caudate nucelus, and anterior crus of the internal capsule [10]. Additionally, MLSA supplies the globus pallidus and medial portion of the putamen [54]. Infarction in RAH and MLSA can usually present with weakness in the contralateral face and upper limb [10, 64]. In our study, the infarction in RAH was usually restricted in the head of the caudate nucleus, and the MLSA infarction was mostly presented as a lacunar infarction in the medial portion of the globus pallidus. Infarction in these territories showed good clinical outcomes with an mRS less than 2. They had not been associated with poor outcomes after aneurysm surgery in previous study [55]. Stein et al. reported that the injury localized to the head of the caudate nucleus was presented with vomiting, headache, neck stiffness, and decreased conscious level without neurological focal sign [59]. Additionally, injury in the caudate nucleus showed good clinical outcomes in most studies, which support our results [35, 59].
The SCA and the HThA are known to originate from the posterior region of ACoA. The SCA supplies the genu of corpus callosum, lamina terminalis, anterior commissure, fornix, and septum pellucidum. Additionally, the HThA supplies the hypothalamus and the lamina terminalis [10]. In our study, the infarction of SCA was usually involved the genu of the corpus callosum, and the symptomatic infarction had involved more at the fornix and the anterior commissure, which is similar to previous studies [17, 23, 47]. The fornix is known to be the component of the Papez circuit and plays an important role in the memory formation and consolidation [17, 42]. Additionally, the injury in the fornix may have presented with the memory disturbance, which also affected the clinical outcome. The HThA infarction was rare and presented with pituitary insufficiency. However, for these branches, the mRS after 6 months from discharge has been decreased when compared to the mRS at discharge. The preservation of SCA and HThA during ACoAA clipping is important; however, the infarction of the involved territory may be more critical, and may need further study.
The orbitofrontal artery is the first branch of the A2 segment and supplies the rectus gyrus and medial orbital gyri [41]. Injury of the orbitofrontal area, especially the ventromedial orbitofrontal cortex is known to cause behavioral changes, which is known as the “frontal lobe syndrome” [7, 12]. Our results showed a similar neurologic outcome with previous studies and mRS at discharge and 6 months after discharge showed worse outcomes than other arteries, which require additional care.
Limitations
This study has some limitations. First, a selection bias may have occurred due to the retrospective design of the study. Second, due to the poor documentation of the records, we did not investigate the duration of temporary clipping, range of rectus gyrus resection, which might affect the outcome of postoperative infarction. Third, the precise evaluation of postoperative anosmia, which is a common complication of ACoAA clipping and influence the clinical outcome, had not been performed. Finally, we measured the mRS for the overall clinical outcome; however, it is inaccurate for measuring mild cognitive impairment or amnesia, which needs suitable neuropsychologic tests. Further research is needed in the future.