Chronic subdural hematomas (SDHs) can recur or progress even after surgical intervention. Middle meningeal artery (MMA) embolization has emerged as a promising option for reducing rates of recurrence of chronic SDHs [5]. Recent efforts have also explored expanding its use to manage acute SDHs, particularly in high-risk patients [3].

Our case describes the appearance of a meningo-meningeal arteriovenous fistula (MMAVF) immediately after routine microparticle embolization of the MMA for treating an SDH. While MMAVFs have traditionally been associated with trauma or cranial surgery [1], our case highlights the potential for MMAVF occurrence during an MMA embolization. As more clinical trials demonstrate the efficacy of MMA embolizations in the treatment of SDHs [2], we anticipate a concurrent increase in the incidence of iatrogenic MMAVFs. The natural history of MMAVFs is unclear, but untreated patients may have an increased risk of intracranial hemorrhage [1]. We report our experience with managing one case.

An elderly patient presented with driveline bleeding associated with anticoagulation for his left ventricular assist device. His history includes coronary artery bypass surgery, percutaneous coronary intervention, and scalp squamous cell carcinoma with exposed underlying calvarium after radiation therapy. He took aspirin 325 mg and warfarin daily. While in the hospital, he had two witnessed episodes with slurred speech and right facial droop with subsequent recovery to his baseline neurological status.

A computed tomography (CT) scan of the head for stroke and seizure workup demonstrated a 5 mm left acute SDH with minimal mass effect on the underlying brain (Fig. 1). An electroencephalogram was not obtained due to his scalp lesion. He was started on anti-epileptic drugs. Surgery was not indicated for the SDH due to its small size. Given the necessity of resuming therapeutic anticoagulation for his underlying comorbidities, we offered to perform a left MMA embolization to prophylactically reduce his risk of developing a chronic SDH.

During the procedure, the patient was heparinized. The MMA trunk was catheterized with a Headway 17 microcatheter (Microvention Terumo, Aliso Viejo, CA, USA). Initial angiogram demonstrated a normal-appearing MMA trunk originating from the internal maxillary artery with an anterior and posterior division (Fig. 2). After confirming no ophthalmic feeders, microparticle embolization (100-300 μM Embosphere^®, Merit Medical, Jordan, UT, USA) of the anterior division was performed. Two 2 mm×8 cm Stryker Target Helical coils (Stryker Neurovascular, Fremont, CA, USA) were then placed within the anterior division of the MMA trunk. Subsequent angiography demonstrated arteriovenous shunting proximal to the coils (Fig. 3).

Further microparticle embolization was aborted and the anterior MMA division was coiled to its origin to shut down the newly discovered MMAVF. Repeat angiography subsequently demonstrated the resolution of the MMAVF within the anterior division; however, another MMAVF was evident within the posterior division after its catheterization (Fig. 4). We, therefore, withdrew our microcatheter and continued our coil construct to the trunk of the MMA, stopping proximal to the foramen spinosum. A final angiogram demonstrated complete occlusion of the MMA (Fig. 5). Angiography of the left internal carotid artery did not reveal abnormalities. A cone beam CT showed no changes to the known SDH, no contrast extravasation, or no new hemorrhage. Protamine was given for heparin reversal, and the patient was extubated without difficulty.

The patient had an intact neurological exam with intact vision immediately after the procedure. An interval head CT confirmed stability of the SDH without evidence of contrast extravasation. A head CT obtained 3 months after the procedure demonstrated complete resolution of the SDH (Fig. 6).

MMAVF is a rare vascular abnormality most often seen after trauma or cranial surgeries [1]. Our case suggests that MMAVFs may also develop immediately after microparticle embolization of the MMA.

A literature review (Pubmed search: ““middle meningeal artery embolization” AND “arteriovenous fistula””) identified four relevant articles detailing seven patients with incidental MMAVFs during MMA embolization [4,6-8]. Three patients had MMAVFs appear immediately after microparticle MMA embolization, similar to our case [4,8], while one patient had an MMAVF appear three months post-embolization [6]. Piergallini et al. treated both MMAVF cases by coiling the MMA trunk, similar to our case [4]. Tavakkoli et al. switched from 250-micron Embozene particles to larger 500-micron particles after MMAVF appearance [8]. All three patients did well on outpatient follow-up.

It remains unclear whether the type of embolization influences MMAVF appearance, but all three patients detailed in the aforementioned publications, as well as our patient, underwent microparticle embolization before intra-procedural development of the MMAVF. Both articles suggest that the lesion may exist as physiological shunts, only to become angiographically apparent after occlusion of the MMA pushes blood and contrast into the fistula [4,8]. Alternatively, Tavakkoli et al. also postulated that vessel tearing and perforation of the MMA into a vein during contrast injection and coil embolization may also cause the MMAVF [8]. We suspect that MMAVF appearance immediately following MMA embolization is a phenomenon seen more often with microparticles. Microcatheter wedging into small distal MMA vessels may inadvertently lead to vessel dissection(s) or MMA rupture during manual injection that allows shunting into veins. This phenomenon may be less often observed with use of liquid embolic agents since these agents also occlude the proximal parent artery. However, more cases are needed before the definitive etiology can be elucidated.

Our treatment was effective. While the appearance of the MMAVF during the procedure raised concern about potential hemorrhage into the subdural or subarachnoid space, subsequent CT studies confirmed the stability of the SDH, supporting the diagnosis of MMAVF. Our current protocol derived from this experience is to avoid microparticle injection when the microcatheter is wedged distally within the MMA with no antegrade flow, in order to prevent vessel injury that may lead to the appearance of an MMAVF. Should an MMAVF appear, we perform coil embolization of the MMA at the level of the MMAVF. If angiographic occlusion is not achieved, we coil the MMA trunk. Review of current literature suggests that switching to larger particles or a liquid embolic agent may also be equally effective in management of an MMAVF appearing during MMA embolization [8].

Meningo-meningeal arteriovenous fistulas may appear immediately following routine middle meningeal artery embolization for treatment of subdural hematomas, particularly when the catheter is wedged distally without antegrade flow or with forceful injection of contrast/embolic material. Appropriate intra-procedural workup should be performed to rule out vessel perforation and associated hemorrhage into the subdural or subarachnoid space. We recommend occlusion of the meningo-meningeal arteriovenous fistula at the level of the fistula until complete angiographic occlusion is achieved.