This is a retrospective, single-center analysis of consecutive patients who underwent surgical or endovascular treatment of UIAs between January 2013 and December 2020. Data collection was approved by the institutional review board and written informed consent was obtained from each patient prior to the procedure.

All patients undergoing microsurgical clipping or endovascular coiling (with or without stenting) of MSAs were considered for inclusion in this study. Patients with multiple MSAs and incomplete preoperative angiographic imaging were excluded. Other exclusion criteria were pre-treated, partially thrombosed and giant (>25 mm) aneurysms and vertebrobasilar artery aneurysms.

All patients underwent preprocedure digital subtraction angiography (DSA) to confirm aneurysm location, size, and morphology. Thereafter, all cases were discussed on a case-by-case basis at a weekly neurovascular conference consisting of vascular neurosurgeons, vascular surgeons, interventional neuroradiologists, and neurologists. At these meetings, treatment decisions and allocation to either endovascular or surgical treatment were made by consensus, taking into account the specific aneurysm and patient characteristics. In cases of equipoise between surgical and endovascular treatment, the choice of modality was made by informed patient consent. If endovascular therapy was favored, the choice of endovascular modality (e.g., coil, stent) was at the discretion of the interventionalist, taking into account aneurysm and patient characteristics and patient preference.

Microsurgical clipping was the treatment of choice for middle cerebral artery (MCA) aneurysms, bifurcation aneurysms, and wide-necked anterior communicating artery (Acom) aneurysms. Clipping was performed under general anesthesia using an operating microscope under the guidance of intraoperative microdoppler ultrasound and/or indocyanine green videoangiography. The pterional approach was used for the majority of aneurysms. The interhemispheric approach was used in selected cases of anterior communicating aneurysms. Our institutional follow-up protocol after aneurysm clipping consisted of clinical evaluation and DSA 3–6 months after surgery, to allow the patient time for recovery after surgery. DSA was preferred over computed tomography angiography due to a higher sensitivity to detect aneurysm remnants [11].

Endovascular procedures were performed under general anesthesia via the transfemoral access using a Philips biplane system (Philips). All patients received systemic heparinization during the procedure. Stand-alone coiling was the first-line treatment for internal carotid artery (ICA) aneurysms. If the coils could not be securely placed within the aneurysm cavity and tended to dislocate into the vessel lumen due to complex aneurysm geometry (i.e., neck width >4 mm, dome-toneck ratio <2), additional stent implantation was considered. Patients were routinely treated with dual antiplatelet therapy (75 mg/d clopidogrel, 100 mg/d aspirin) started 5 days prior to the procedure, even when stand-alone coiling was planned in case additional stent implantation became necessary. After 4 months, treatment was continued with continuous single antiplatelet therapy, either clopidogrel or aspirin. Platelet inhibition was assessed in all endovascular patients with aspirin and P2Y12 assays (VerifyNow; Accumetrics). A platelet inhibition level between 30–60% for clopidogrel and 350–550 aspirin response units for aspirin was required. An inadequate response to either drug was counteracted by dose escalation (e.g., clopidogrel 150 mg/d) or substitution with prasugrel (40 mg bolus, 5 mg/d). Follow-up after endovascular aneurysm treatment consisted of clinical evaluation and control DSA at 6 and 24 months.

Patient demographics, aneurysm characteristics, and treatment details were obtained from medical records. We reported all adverse events with new neurological symptoms. Major events were defined as lasting more than 7 days, whereas symptoms of minor events were defined as completely resolving within 7 days. Neurological complications included intracranial hemorrhage, ischemic stroke, symptomatic vasospasm, and seizure. Functional outcome was assessed using the modified Rankin scale (mRS) at discharge and at 1-year follow-up (favorable mRS score, 0-1; unfavorable mRS score, 2–6).

Aneurysm morphology was evaluated using two-dimensional (2D) DSA images and 3D volume-rendered images from rotational angiography data. The classification of aneurysm morphology proposed by the International Study of Unruptured Aneurysms (ISUIA) was used to classify aneurysms into single-sac aneurysms with regular or irregular margins, aneurysms with a DS, and Lobs [2]. The last 2 morphologic groups were considered MSAs and included in this study. Illustrative images of all 4 morphologic subtypes are shown in Fig. 1.

A DS was defined as a distinct protuberance arising from the primary aneurysm dome and accounting for a maximum of 25% of the main aneurysm sac. Lobs are composed of at least 1 additional aneurysm sac arising either directly from a combined neck or from the primary aneurysm dome and covering at least 25% of the main aneurysm sac [1]. Aneurysm size (i.e., the largest diameter of the aneurysm dome), aneurysm height, and neck width were also measured. The dome-to-neck ratio (aneurysm width/neck width) and the aspect ratio (aneurysm height/neck width) were calculated for all aneurysms. Aneurysms with a neck width >4 mm or a dome-to-neck ratio <2 were considered wide-necked.

The Raymond–Roy occlusion classification (RROC) was used to assess aneurysm obliteration at follow-up angiographic imaging: 1) complete aneurysm occlusion, 2) neck remnant, and 3) residual aneurysm. The RROC scale was originally developed for evaluation of coiled aneurysms; however, it is also used for clipped aneurysms [12].

Categorical variables were expressed as numbers and percentages, and continuous variables were expressed as means±standard deviation. The 3 treatment groups were compared using chi-squared statistics with post-hoc Bonferroni pairwise comparison, and the Kruskal-Wallis test with post-hoc Dunn test, as appropriate. Age, sex, aneurysm morphology, aneurysm location, aneurysm size, and treatment modality were tested for predicting neurological complications and incomplete occlusion at angiographic follow-up. Factors predictive in univariate analysis (P<0.2) were entered into a binary logistic regression model to identify independent prognostic factors for each outcome measure. All statistical tests were performed with SPSS version 25.0 (IBM Co.). A P-value<0.05 was considered statistically significant.

During the study period, a total of 529 patients were treated for unruptured aneurysms, whereof 125 (23.6%) patients had 133 MSAs. Thirty-five patients were excluded for the following reasons: treatment of multiple MSAs (8), flow diverter treatment (12), implantation of the Woven EndoBridgeTM (MicroVention) (7), giant aneurysms (3), and vertebrobasilar artery aneurysms (5). The final study population consisted of 90 patients with 33 aneurysms with DSs and 57 Lobs. The mean age of the patients was 53.2±11.0 years, and 73 (81.1%) patients were female.

The most common location of MSAs was the MCA in 44 (48.9%) cases, followed by the anterior cerebral artery (ACA) in 24 (26.7%). The mean aneurysm size was 8.0±3.8 mm and the mean neck width was 4.0±1.5 mm. Of 90 aneurysms, 50 (55.6%) were clipped, 19 (21.1%) were treated with standalone coiling, and 21 (23.3%) with SAC. Baseline patient and aneurysm characteristics for the entire cohort and for each treatment modality are shown in Table 1.

Clipping was most frequently used for MCA aneurysms (P<0.001), while older patients (P=0.019) and aneurysms located at the ICA (P=0.001), the ACA (P=0.003), and the posterior communicating artery (Pcom) (P<0.001) were more likely to be treated endovascularly.

All clipping procedures were technically successful. In the endovascular group, treatment with stand-alone coiling failed in 3 cases. Stand-alone coiling was not successful in a wide-necked aneurysm with a DS (6 mm) at the bifurcation of the MCA. In this case, the coils could not be securely placed in the aneurysm cavity and prolapsed into the parent artery. Although SAC was potentially an alternative endovascular option after antiplatelet loading, we assumed an increased risk of periprocedural complications with SAC and decided by interdisciplinary consensus to treat the aneurysm with microsurgical clipping. The case is presented in Fig. 2.

Furthermore, 2 aneurysms were initially planned to be treated with stand-alone coiling but eventually required stent support. In the first case, a coil was extended from the aneurysm into the Pcom. In the second case, an ICA aneurysm could not be adequately occluded by stand-alone coiling, so a stent was added.

Successfu cases of endovascular and surgical obliterations are shown in Figs. 3 and 4.

Neurological events are shown in Table 2. There was a total of 13 (14.4%) neurological events, of which 2 (2.2%) were classified as major and 11 (12.2%) as minor. The major events consisted of 1 stroke and 1 hemorrhage after clipping. The major stroke was due to prolonged temporary occlusion of the parent artery during clipping of a complex lobulated MCA bifurcation aneurysm. The patient was severely disabled with an mRS of 5. The major subarachnoid hemorrhage was due to a procedural rupture during clipping of an Acom aneurysm; the patient was discharged to a rehabilitation center with an mRS 3. Minor neurological events in the clipping group included 2 minor strokes due to perforator infarction, 2 access-related subdural hematomas, and 4 seizures. All patients fully recovered from these minor complications during the hospital stay.

In the endovascular group, there were 3 minor strokes due to procedural thromboembolic events during coiling of an Acom aneurysm and stent-assisted coil embolization of an Acom and a Pcom aneurysm, respectively. All 3 patients had only mild symptoms and fully recovered during hospitalization.

There was no procedure-related mortality. The overall (P=0.228) and major event rates (P=0.667) were not significantly different across the treatment modalities. On univariate analysis, an increased aspect ratio (2.4±1.6 vs. 1.8±0.8, P=0.031) was associated with adverse events.

At discharge, 48/50 (96.0%) patients in the clipping group and all 40 patients in the endovascular group achieved a favorable outcome. At 1-year follow-up, a favorable outcome was achieved by 97.8% (44/45) in the clipping group and in 100% (34/34) in the endovascular group.

Angiographic follow-up was available for 37/50 (74.0%) patients in the clipping group (mean follow-up 10.9±9.1 months), 17/19 (89.5%) in the coiling group (mean follow-up 13.4±9.2 months), and 15/21 (71.4%) in the SAC group (mean follow-up 13.0±8.1 months) (Table 3).

The complete occlusion rate after clipping (37/37 [100%]) was significantly higher (P<0.001) than that of coiling (8/17 [41.7%]) and that of SAC (11/15 [73.3%]). The complete occlusion rates of SAC were higher than that of coiling without reaching statistical significance (P=0.130). For coiled aneurysms with recanalization were successfully retreated with coiling in 3 cases and SAC in 1. In the SAC group, 1 retreatment with a flow diverter implantation failed in an ICA aneurysm. This aneurysm was subsequently followed angiographically.

In univariate analysis, Pcom location (44.4% vs. 15.0%, P=0.037) and coiling (52.9% vs. 7.7%, P<0.001) were associated with incomplete occlusion at angiographic follow-up. In multivariate analysis, coiling was the only significant independent prognostic factor for incomplete occlusion (odds ratio, 11.7; 95% confidence interval, 2.7–52.6; P=0.001).

Despite the tremendous progress in neurointerventional techniques over the last 15 years [5,6], there is still a fraction of aneurysms that cannot be safely treated by interventional means and for which clipping remains the first-line treatment option. For example, MCA aneurysms are still preferentially treated surgically in most neurovascular centers [14-16]. This treatment regimen was also true for MSAs in our cohort. Clipping of MSAs was associated with high technical success, resulting in complete occlusion of all aneurysms followed. Thus, the high occlusion rates reported for aneurysm clipping were also applicable to MSAs [17].

Major complications with persistent neurological deficits occurred in only 4.0%, which seems reasonable for this complex subset of UIAs. Morbidity rates were 4.0% at discharge and 2.2% at 1-year, which is comparable to previous studies on clipping [17]. In fact, seizures were the most common complication in the clipping group. It can be speculated that clip placement in MSAs is more difficult than in single-sac aneurysms and may be associated with longer operative time and more manipulation of the aneurysm dome. These factors are known to increase the likelihood of complications [18].

As in previous studies, the complication rate appears to be higher than with endovascular procedures, but the different aneurysm subsets of the 2 treatment groups do not allow a direct comparison. In fact, endovascular treatment of MCA aneurysms is challenging and associated with higher complication rates compared to typical endovascular sites such as the ICA and basilar artery. For example, Vendrell et al. [19] reported a complication rate of 19.4% and permanent morbidity of 5.7% for endovascular treatment of MCA aneurysms, which is not lower than the complication rates for clipping focused on complex morphology in the present series.

Although microsurgical clipping is the standard of care for complex morphology anterior circulation aneurysms, an increasing number of complex aneurysms are being treated endovascularly. Therefore, we also evaluated the efficacy of coiling and SAC for the treatment of MSAs. It should be noted that aneurysms were selected for coiling and SAC after interdisciplinary case discussion, whereas the majority of MSAs were considered unfavorable for endovascular treatment and were treated by clipping (50 clipping vs. 40 endovascular). Given this preselection, coiling failed in 3 cases. In our experience, severe lobulation and a wide neck are characteristics that make stand-alone coiling difficult, as the coil cannot unfold properly in the aneurysm sacs and therefore protrudes into the vessel lumen. Coil protrusion can be counteracted with an additional stent or switch to flow diversion if possible. Otherwise, clipping may be necessary, such as for bifurcation aneurysms that are difficult to treat with stents.

In our study, the complication rate was low and we did not observe any case of adverse outcome after coiling of MSAs. Thus, our study of MSAs confirms the excellent safety results of coiling reported in previous studies [20].

A major disadvantage of endovascular treatment may be that it is associated with higher rates of aneurysm recanalization than clipping [14,21]. Similarly, the rate of aneurysm recurrence at the 6-month angiographic follow-up in our endovascular cohort was 53.0% after coiling and 26.7% after SAC. Based on these mid-term angiographic results, the rate of aneurysm recurrence is expected to increase significantly in the long-term. Due to the high rate of aneurysm recanalization, coiled MSAs are associated with an increased risk of aneurysmal subarachnoid hemorrhage in the long-term, which may have a serious impact on patient prognosis [22]. These considerations must be taken into account when developing a treatment strategy.

In the multivariate analysis, stand-alone coiling was an independent prognostic factor for incomplete occlusion of MSAs, whereas SAC carried a reasonable risk of aneurysm recurrence. In general, aneurysm recurrence after coiling of UIAs is reported to be approximately 25% [20]. Therefore, the recurrence rate after coiling was significantly higher than in most other studies that do not differentiate aneurysm morphology. It can be speculated that achieving dense coil packing in MSAs is complicated by their lobulated shape, making these aneurysms susceptible to coil compaction and aneurysm recurrence. To compensate for this risk, MSAs may require longer angiographic follow-up after stand-alone coiling and retreatment of aneurysm remnants, increasing the cumulative risk of procedure-related complications.

The results of our study indicate a high feasibility and safety of endovascular treatment of MSA when used in selected patients after prior interdisciplinary case discussion. However, recanalization rates were comparably high after stand-alone coiling, while SAC appeared to be more robust to coil compaction and may therefore be the preferred endovascular treatment option for the majority of patients with MSAs, even if it requires long-term antiplatelet therapy. However, coiling may be considered in older patients with limited life expectancy. Clipping had a 100% success or complete occlusion rate, but a slightly higher complication rate than endovascular treatment. However, the complication rates do not appear to exceed those of previous studies in unselected cohorts. Therefore, clipping may remain a viable option for MSA when endovascular treatment appears inappropriate or when antiplatelet therapy is not desired, such as in younger patients.

Due to the retrospective nature of this study, the treatment groups were not balanced with respect to patient characteristics and aneurysm location, making direct comparisons between groups difficult. Aneurysms were preselected for either endovascular or surgical treatment at a multidisciplinary neurovascular conference, introducing a clear selection bias. Because this was a single-center analysis, the number of patients enrolled is limited and our findings require validation by further studies. The follow-up was incomplete both in the clipping and the endovascular group. In addition, we did not analyze long-term follow-up results, but it can be expected that recanalization rates might increase further after coiling. Aware of the above limitations, this study provides novel insights into the management and risk profile of UIAs that may help guide treatment decisions.