Treatment of complex, non-saccular intracranial aneurysms has increasingly involved endovascular therapies, in particular flow diversion, due to an ability to remodel diseased vessels from the inside in a minimally invasive manner (1). Intrasaccular devices similarly have the potential to change treatment paradigms for selected wide-necked saccular aneurysms (2). Limitations to these technologies nonetheless exist, particularly in the setting of ruptured aneurysms where flow diversion (or stent/coil constructs) requires the use of dual anti-platelet agents (increasing the risk of hemorrhage with ventricular catheters), and both flow diversion and intrasaccular devices can often have continued aneurysm filling (maintaining a small but significant risk of re-rupture) (3, 4). Although newer flow diversion technologies are being investigated that require only a single anti-platelet agent, concerns regarding significant morbidity and mortality from ischemic complications or post-treatment aneurysmal re-bleeding (occurring in up to 21% of patients) remain based on data from preliminary studies (5).

Open management of complex aneurysms, in particular cerebral bypass, remains an important tool for vascular neurosurgeons (6). Combined approaches that leverage the respective advantages of open and endovascular treatments, while avoiding their risks, can also be tailored to optimize the treatment of select aneurysms. For example, for fusiform or complex saccular ruptured aneurysms where a definitive treatment is desired, a combined strategy of open revascularization followed by endovascular deconstruction can be used to limit the risk of additional dissection with open-only techniques (i.e., for clip trapping), and allow for endovascular vessel sacrifice (or aggressive aneurysm treatment that may risk parent vessel occlusion) without downstream ischemic risk. We herein report our experience with combined open bypass and endovascular management of complex cerebral aneurysms.

This study was performed in compliance with Health Insurance Portability and Accountability Act and local institutional review board (IRB) regulations. Patient consent was obtained based on institutional guidelines.

Data from all patients undergoing combined open revascularization and endovascular treatment for the management of intracranial aneurysms from 1/2016 to 6/2022 was retrospectively collected from an IRB-approved, prospectively maintained database. Recorded data included: patient demographics (age, gender), presenting symptoms, clinical status, Glasgow Coma Scale (GCS) score, aneurysm characteristics, surgical and endovascular details, treatment/overall complications, and neurologic outcomes (Glasgow outcomes scale [GOS] and modified Rankin Scale [mRS] scores at discharge and last follow up). Descriptive statistics were used to report continuous data (mean ± standard deviation).

Management and planning of combined open revascularization and endovascular treatment of all patients in this series was based on multidisciplinary (open and endovascular) neurosurgical review of the patient, and neurovascular and pathologic anatomy. Surgical details for open revascularization strategies are described previously (7–11). All patients were maintained on aspirin in the peri-operative and post-operative period to promote bypass patency. Secondary endovascular procedures were performed as early as possible after revascularization to secure the aneurysm (in cases of rupture) and to promote bypass patency. Systemic heparinization was not used during endovascular procedures for ruptured aneurysms or for procedures immediately following open surgery. Neuromonitoring was used for both open and endovascular procedures. Balloon test occlusion (BTO) was performed if feasible based on vascular anatomy to aid in treatment planning prior to endovascular vessel sacrifice.

Ten patients (4 male [40%]; mean age 51.9 ± 8.7 years) underwent combined endovascular and open treatment of intracranial aneurysms (Figure 1; Table 1). Most aneurysms presented ruptured, 9/10 (90%), and 3/10 (30%) patients had undergone prior aneurysm treatments or treatment attempts. Aneurysm locations included: 5 (50%) involving the vertebral artery (VA)/posterior inferior cerebellar artery (PICA) origin, 2 (20%) of the proximal PICA or anterior inferior cerebellar artery (AICA)/PICA complex, 2 (20%) of the anterior communicating artery (Acomm) or A1/2 junction, and 1 (10%) of the proximal posterior cerebral artery (PCA). All (8/8) of the aneurysms involving the posterior circulation were fusiform, while aneurysms involving the anterior circulation were traumatic/pseudoaneurysm (1/2; 50%) or saccular (1/2; 50%) in morphology. Bypass strategies included both extracranial-to-intracranial (EC-IC) and intracranial-to-intracranial (IC-IC) constructs and included PICA-PICA side-to-side (5/10; 50%), V3-PICA with a descending branch of the lateral circumflex artery (DLCFA) interposition graft (2/10; 20%), A3-A3 side-to-side (2/10; 20%), and occipital artery (OA)-P4 with a DLCFA interposition graft (1/10; 10%). Mean temporary clip time was 38.5 ± 8.9 min, and immediate bypass patency was 100%. Initial endovascular procedures were performed early after surgery (mean 0.7 ± 1.5 days) to reduce rupture risk and promote bypass patency, and included vessel sacrifice in 9/10 patients (90%) (Table 2). Two patients underwent multiple endovascular embolizations. In one patient, a non-occlusive post-bypass endovascular strategy was initially employed for a fusiform VA aneurysm involving the proximal left PICA to preserve flow to the dominant anterior spinal artery (ASA) located immediately distal to the aneurysm, although further aneurysmal growth necessitated a delayed vessel sacrifice. The second patient had an initial post-bypass endovascular strategy of aggressive re-coiling of a recurrent/re-ruptured Acomm aneurysm arising from an unpaired A1 (including Acomm sacrifice), followed by delayed flow diversion/coiling for aneurysm re-growth.

Strokes from any etiology occurred in 5/10 patients (50%). Treatment related strokes were diagnosed in 3/10 patients (30%) and included a lateral medullary stroke after PICA-PICA bypass and VA sacrifice, thalamic/occipital strokes after PCA bypass/sacrifice, and an ASA stroke after V3-PICA bypass and delayed VA sacrifice. In 2/10 patients (20%) strokes were from vasospasm. Mean GOS and mRS at discharge were 3.8 ± 0.9 and 2.7 ± 2.0, respectively. Follow up data was available in 6/10 patients (60%). Over a median follow up of 14.0 months (range 4–72), all bypasses were patent and mean GOS and mRS were 4.5 ± 0.5 and 1.2 ± 1.3, respectively. Good outcomes (defined as a GOS ≥ 4 and mRS ≤ 2) occurred in 6/10 patients (60%) overall.

Combining open and endovascular techniques has established utility for a variety of cerebrovascular pathologies, including complex arteriovenous malformations, arteriovenous fistulas, and aneurysms (12–22). Such approaches are especially useful for aneurysms that can be challenging to treat with purely open or endovascular options, including ruptured giant, blister, fusiform, dissecting, heavily calcified, or pseudoaneurysms in difficult to access locations and with associated perforating vessels or scar tissues from prior treatments (13). Described strategies include both endovascular therapy after surgical treatment (i.e., coil embolization after partial clipping of incompletely exposed or recurrent aneurysms, or endovascular occlusion/vessel sacrifice after bypass, as used herein), and open surgical treatment after endovascular therapy (i.e., definitive clipping after dome-protection of ruptured aneurysms unable to be completely coiled, or surgical debulking to reduce mass effect after endovascular occlusion) (12, 13, 15).

Our data add to prior works supporting the use of cerebral bypass combined with endovascular embolization for complex intracranial aneurysms (13, 14, 16, 17, 19, 21, 23–26). Although previously described for lesions throughout the anterior and posterior circulation, we have found these constructs most useful for the treatment of aneurysms less surgically accessible, particularly in the posterior circulation such as those located on the distal VA, proximal PICA, and proximal PCA (80% of cases in this series). In our experience, complex aneurysms of the anterior circulation (internal carotid artery [ICA], middle cerebral artery [MCA], and even ACA) can often be directly addressed with open surgery algorithms due to their more easily accessible anatomy, which facilitates direct visualization of perforators for precise clip application during vessel ligation, trapping or aneurysm/vessel reconstruction (8, 27–31). This is not always feasible or advisable though, as evidenced by the two ACA cases in this series where a combined open and endovascular approach was used to avoid manipulation of an unstable A1/A2 pseudoaneurysm and a multiply ruptured partially coiled wide-necked Acomm aneurysm.

As demonstrated in this series, the combined approach is easily tailored to a variety of aneurysms and anatomical variations. Emerging data supports the technical and safety profile of IC-IC versus EC-IC constructs (8, 9, 32–35), and we prefer to use PICA-PICA bypass for revascularization during treatment of complex proximal PICA or VA/PICA aneurysms when the contralateral PICA anatomy is favorable. This also avoids the need for dissection of the often torturous OA or harvest of an interposition graft (as needed with other EC-IC options), or performance of a deep anastomosis (as needed with excision and re-anastomosis/re-implantation strategies) (9, 33). In cases where the contralateral PICA is not favorable for side-to-side IC-IC bypass, the DLCFA is well-sized for PICA revascularization and can be used as an interposition graft with a V3, OA, or even the contralateral PICA as a donor vessel (10, 11). Regardless of the bypass strategy used, when combined with endovascular sacrifice of the proximal PICA (and VA as needed), extended skull base approaches to reach the distal VA/PICA origin anterolateral to the brainstem can be avoided. With this approach, after vessel sacrifice the distal PICA fills anterograde from its anastomosis at the tonsillar segment, while the proximal PICA (including brainstem perforators) fills in a retrograde fashion to the point of occlusion. A similar concept can be employed for complex proximal ACA aneurysms by combining an A3-A3 bypass with endovascular vessel sacrifice or aggressive aneurysm treatments that risk vessel occlusion. This strategy avoids the need for long interposition grafts to reach into the interhemispheric fissure from the external carotid artery (ECA) circulation for ACA revascularization, while also avoiding surgical manipulation of the perforator rich A1/A2 region for aneurysms with expected scarring or wall friability. For PCA revascularizations as part of combined treatments of complex proximal PCA aneurysms, IC-IC constructs are less favorable, and we prefer to use a DLCFA interposition graft to connect the OA to a P3 or P4 vessel (again avoiding a time-consuming full OA dissection) (7, 10, 11). This can be followed by proximal PCA endovascular sacrifice, avoiding temporal lobe retraction and potential venous compromise from an additional open subtemporal approach, as well as the manipulation of often sub-optimally visualized sensitive proximal PCA perforators.

Although the overall stroke rate was 50% in this series (5 patients), treatment-related strokes occurred in three patients (30%), with the remaining two patients (20%) having delayed strokes from vasospasm. With the use of multiple treatments in a combined approach, the additional risk of each procedure should be offset by its advantages. The 100% bypass patency and low open surgery complication rate herein is similar to the 97% patency reported in a recent large series of 430 cerebral bypasses (36). These data support the relative efficacy of bypass in experienced hands, and justify the addition of an open procedure to minimize the ischemic risk associated with stand-alone end-vessel sacrifice (37, 38). Vessel sacrifice by any means (open or endovascular) is nonetheless associated with the risk of both downstream and perforator strokes, with bypass mitigating the downstream risk but not the risk of perforator stroke if they arise directly from the diseased vessel and a trapping/occlusive (rather than proximal Hunterian ligation) strategy is used. In all the patients in this series, endovascular embolization was selected to avoid unfavorable open dissections, with any significant angiographically visualized perforators preserved and a minimum length of vessel sacrificed to secure the aneurysm.

Understanding perforator anatomy is crucial to designing a treatment strategy that minimizes stroke risk. Along the V4 VA, critical perforators (including the ASA) are variable but thought to be more common distal to the PICA origin (39–41). Accordingly, although brainstem strokes from perforator infarcts can occur from sacrifice anywhere along the V4 vessel, sacrifice of the segment distal to the PICA is thought to be higher risk than more proximally (39, 42). This is true even with selective sacrifice only at the level of disease, where although it is presumed that perforators within the pathologic segment have likely already been occluded, a stump effect from inadequate outflow can extend the length of vessel occlusion based on local arterial anatomy independent of sacrifice technique (42). Even with visualization and early angiographic preservation of perforators, their patency is thus not guaranteed with deconstructive options. This is demonstrated in the patient that experienced an ASA stroke in this series, which occurred after endovascular vessel sacrifice despite angiographic visualization of the ASA just distal to the aneurysm and allowing it to backfill from the contralateral VA. Learning from this case, we have since allowed ASA anatomy to dictate the treatment strategy of VA/PICA aneurysms, with Hunterian ligation or bypass/clip trapping used for ASAs arising from the aneurysm itself or just distal, respectively (allowing for direct, real-time assessments of ASA viability with clip placement and the potential for strategy modifications if changes are seen with vessel flow or neuromonitoring), and combined open bypass and endovascular occlusion used for cases where the ASA is removed from the aneurysm and the potential for a stump effect is low (43). Similarly, with PICA-PICA bypass, avoidance of tension on the PICAs is critical to maintain lateral medullary perforator patency, a potential contributing factor in the patient in this series with a lateral medullary stroke after PICA-PICA bypass and occlusion of a VA/PICA aneurysm. In this patient, mobilization of the ipsilateral PICA was needed to reach the contralateral vessel for bypass, potentially placing the proximal PICA perforators at risk and decreasing retrograde demand within this vessel. Learning from this, EC-IC PICA bypass has since been used if there is any concern for PICA tension with a PICA-PICA construct.

Perforators play a similarly important role in the treatment of proximal PCA aneurysms. Although these vessels can be difficult to visualize, the proximal PCA is perforator-rich and sacrifice should be avoided whenever possible (44). In the patient that experienced a small thalamic stroke from PCA perforators, this occurred despite keeping a portion of the diseased segment of the PCA proximal to the aneurysm open to allow for perforator filling (with the stroke likely from occlusion of a small vessel arising more distally along the sacrificed and grossly dysplastic P2). This situation is difficult to predict, and although it would likely have also occurred with clip trapping, may have been avoided with a bypass/proximal ligation strategy. Emerging technologies such as high-resolution cone beam computed tomography scans, which can visualize small vessels with high resolution, may help identify critical perforators pre-operatively and inform treatment planning focused on their preservation (45). BTOs may be helpful to assess bypass adequacy and predict tolerance of vessel occlusion prior to sacrifice (especially for vessels other than the VA with no expected redundancy, and/or to assess for potential stump affect with nearby critical perforators). It can nonetheless be challenging to position the balloon properly when sacrifice of smaller or more torturous vessels is planned (i.e., an ACA as in patient 8). BTOs are also not entirely predictive, as the BTO and final occlusion sites often differ slightly. Highlighting this point, patient 3 and patient 10 experienced an ASA and PCA perforator stroke, respectively, after unremarkable BTOs of the VA and P1 prior to aneurysm embolization/sacrifice.

With use of combined open and endovascular approaches in general, if there is concern for significant perforator occlusion with aneurysm embolization, proximal vessel occlusion or open clipping at a perforator-free area can be considered, allowing retrograde filling of the aneurysm and perforators through the bypass (i.e., Hunterian ligation). This must nonetheless be balanced by the reduced but ongoing rupture risk with continued aneurysm filling with this approach (46). Open clip trapping can also be used if a significant perforator is near but not emerging from the aneurysm and the proximal and distal vessel are surgically accessible.

An analysis of outcomes for the other main treatment options for complex intracranial aneurysms highlights the significant challenges of managing this patient population. Flow diversion is the main endovascular option for the treatment of nonsaccular ruptured aneurysms, but is associated with a 21 to 26% rate of ischemic/hemorrhagic complications with use of either standard devices and dual anti-platelet regimens, or newer devices and single agent anti-platelets across all anatomic sites (4, 5). While good neurologic outcomes have been reported with flow diversion for the treatment of fusiform VA aneurysms involving the PICA origin (representing 50% of the cases in the series), complete aneurysm occlusion occurs in <60% of cases, while PICA/VA occlusion can occur in up to 10% of cases (47). A variety of other endovascular options exist for complex non-saccular dissecting lesions, such as placement of multiple stents, stent-assisted coiling, and internal trapping with stenting (described as an alternative to flow diversion if a daughter vessel is involved), but are associated with 12.5 to 25% rates of infarction or aneurysm recurrence (48). Similarly, although selected complex saccular aneurysms not amenable to primary coiling and treatable with stenting or intrasaccular devices can have low rates of procedure-related hemorrhagic/thrombotic or external ventricular drainage–related events (<10% overall), persistent aneurysm filling with unclear long-term implications can occur in a significant proportion (17–46%) of cases. (2, 49). Nonetheless, the algorithm for aneurysms at our institution is endovascular management first, with more complex open or combined approaches reserved for lesions deemed to be poor candidates for stand-alone endovascular therapies (including flow diversion) on multidisciplinary review. While a direct comparison of open-only treatment strategies for the heterogenous complex aneurysm mix in this series is challenging, data from a recent series of 42 patients with largely ruptured dissecting aneurysms of the vertebrobasilar system (the majority of patients in our series) reflects the often poor baseline neurologic conditions of these patients, with good outcomes reported in less than 50% of patients with a variety of open treatments (50). The 60% overall rate of good outcomes in our series (despite 90% of patients having ruptured aneurysms), is consistent with prior data demonstrating the utility of a combined endovascular and open approach for otherwise difficult-to-treat aneurysms (16, 17).

Limitations of this study include its single-institution, retrospective design, and small patient size reflective of the relative rarity of complex aneurysms requiring combined treatments. Four patients (40%) were also lost to follow up, a reflection of the quaternary referral nature of our institution where transfer back to the referring center is often required after completion of the acute treatment stage. Finally, as expertise in both complex open revascularization and endovascular techniques is critical for the successful implementation of this strategy, the generalizability of this data is most pertinent to other high-volume bypass centers.

Combined open revascularization and endovascular embolization can be used for select complex aneurysms not amenable to stand-alone open or endovascular techniques.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by University of Southern California Institutional Review Board. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

RR and JR contributed to the design and conception of the study. RR, VN, JR, JC, MT, AAm, and WM developed plans for and participated in the clinical care of included patients. AAb and NA organized and maintained the database. RR, VN, AAb, and NA collected data. RR drafted the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.