Middle meningeal artery embolization as an alternative strategy for symptomatic CSF hypovolemia-related chronic subdural hematoma: A case series

Abstract

Background

Chronic subdural hematoma (CSDH) secondary to cerebrospinal fluid (CSF) hypovolemia—most commonly from ventriculoperitoneal (VP) shunt overdrainage or spontaneous intracranial hypotension (SIH)—presents unique therapeutic challenges. In these cases, standard surgical evacuation may be contraindicated if intracranial pressure is not corrected, increasing the risk of complications. While middle meningeal artery embolization (MMAE) has emerged as a treatment option for refractory CSDH, its role in CSF-hypovolemia associated CSDH remains underexplored.

Methods

This retrospective study included seven patients with radiographically confirmed, symptomatic CSDH due to CSF hypovolemia. Etiologies included VP shunt overdrainage (n = 6) and SIH (n = 1). Prior to MMAE, all patients underwent definitive treatment of the underlying cause—either shunt adjustment or ligation, or targeted epidural blood patch. Embolization was performed using 250 μm microspheres via the middle meningeal artery. Patients were followed clinically and with serial CT imaging.

Results

All patients demonstrated neurological improvement within several days following embolization. Hematoma thickness progressively decreased over a period of 2 to 6 months, with complete or near-complete resolution in all cases. No patients experienced recurrence or required surgical evacuation. There were no procedural complications.

Conclusion

MMAE, when performed after correction of the underlying CSF hypovolemia, appears to be a safe and effective therapeutic option for CSDH. This approach may serve as an intermediate strategy between conservative treatment and surgical evacuation, particularly in patients at elevated risk due to underlying CSF hypovolemia.

Introduction

Chronic subdural hematoma (CSDH) is a common neurosurgical condition, especially in elderly patients.^1–4 Beyond traumatic causes, CSDH may also develop insidiously in the setting of intracranial cerebrospinal fluid (CSF) hypovolemia, such as that resulting from ventriculoperitoneal (VP) shunt overdrainage or spontaneous dural CSF leaks.^5,6 In patients with overdraining shunts, negative intracranial pressure causes brain descent and stretching of bridging veins, precipitating subdural collections. CSDH is a recognized complication of VP shunt overdrainage, occurring in approximately 4–10% of cases.^7–9 Similarly, spontaneous intracranial hypotension (SIH), often presenting with orthostatic headache, is frequently accompanied by subdural hygromas or hematomas due to downward displacement of the brain within the cranial vault. ¹⁰ Management of CSF hypovolemia-related CSDH typically prioritizes correcting the leak. In shunted patients this means raising the shunt opening pressure, ligating or revising the shunt, or implanting programmable valves. In SIH, conservative measures (bed rest, hydration, caffeine) and especially targeted epidural blood patch (EBP) are first-line therapies to seal the dural defect. ¹¹ Surgical drainage of the hematoma is reserved for mass effect or neurological deterioration, but it carries a high recurrence rate if the leak remains untreated.^11,12 Previous reports emphasize that treating the hematoma alone can be futile or lead to repeated recurrences unless the underlying CSF leak is addressed.^11,12 Middle meningeal artery embolization (MMAE) has recently emerged as an adjunct or alternative therapy for CSDH in general. Several recent studies have investigated the role of MMAE in the management of CSDH, including its use in septated CSDH, ¹³ comprehensive reviews of past, ongoing, and recent clinical trials, ¹⁴ and single-center reports on patients with unilateral large CSDH and significant midline shift. ¹⁵ Small series and meta-analyses have reported high rates of hematoma volume reduction and low reoperation rates after MMAE in typical CSDH.^1,16 However, its role in CSF hypovolemia-related CSDH remains poorly defined. To our knowledge, only isolated case reports have described MMAE in this context, no series has systematically applied it following leak correction.^17,18 We therefore reviewed our experience with seven patients who had CSDH secondary to CSF hypovolemia (either VP shunt overdrainage or SIH) and were treated with MMAE following standard leak management. Here, we report their clinical and radiographic outcomes and discuss implications for therapy.

Methods

Patient selection and management

We retrospectively identified seven consecutive patients with CSDH attributable to CSF hypovolemia who underwent MMAE at our center between June 2021 and December 2024 (Table 1). Inclusion criteria were: (1) radiographically confirmed CSDH (chronic-appearing density on CT) with measurable thickness, (2) history or evidence of CSF overdrainage (VP shunt with imaging evidence of ventriculomegaly) or SIH (consistent clinical picture with or without spinal imaging), (3) presence of neurological symptoms (gait disturbance, headache, speech deficit, or altered mental status), and (4) no prior burr-hole evacuation for the current hematoma. Although none of the patients in our series underwent surgical evacuation, all had radiographically significant CSDH with clinical symptoms and mass effect—not merely incidental or asymptomatic effusions. All patients received pre-embolization treatment of the CSF leak: six patients with VP shunts had adjustments to valve settings or surgical ligation of the shunt, and one patient with SIH underwent one or more targeted EBPs. The decision to proceed with MMAE was based on the presence of neurological symptoms and imaging findings indicating a risk of ongoing or worsening compression. In the six patients with VP shunts, hematoma thickness was substantial, and symptoms such as gait instability, aphasia, or drowsiness were evident. Importantly, embolization was performed immediately or within 1–2 days following shunt pressure adjustment, with the goal of accelerate hematoma resolution and prevent further neurological decline. In the case of SIH, embolization was deferred until it became apparent that the hematoma continued to enlarge despite initial EBP therapy. This case highlights how embolization can be strategically timed based on hematoma progression and clinical trajectory.

Table 1.

The clinical characteristics and result of patients with CSDH treated with MMAE.

Case	Age/Sex	Etiology	SDH thickness	SDH symptoms	Treatment	Time to clinical improvement	Hematoma resolution	Time to resolution	Recurrence
1	73/F	VP shunt, trauma	7 mm	Drowsy	Shunt ligation, MMAE	4 days	Minimal residual	3 months	No
2	71/M	VP shut	10 mm	Unsteady gait	Adjust pressure, MMAE	3 days	Minimal residual	3 months	No
3	82/F	VP shunt	6 mm	Unsteady gait	Adjust pressure, MMAE	14 days	Complete resolution	6 months	No
4	65/M	VP shunt, trauma	12 mm	Unsteady gait	Adjust pressure, MMAE	3 days	Complete resolution	3 months	No
5	65/M	VP shunt	7 mm	aphasia	Adjust pressure, MMAE	1 day	Complete resolution	3 months	No
6	70/M	VP shunt, trauma	9 mm	headache	Adjust pressure, MMAE	5 days	Complete resolution	2 months	No
7	51/M	Intracranial hypotension	21mm	headache	Epidural patch, MMAE	2 days	Complete resolution	2 months	No

CSDH: chronic subdural hemorrhage; MMAE: middle meningeal artery embolization; VP shunt: ventriculoperitoneal shunt.

MMAE and clinical follow-up

For MMAE, patients underwent angiography under conscious sedation. After common femoral artery access, a guiding catheter was placed in the external carotid artery. A microcatheter was advanced into the main trunk of the middle meningeal artery (MMA) to perform superselective angiography. In the absence of hazardous anastomotic connections, embolization was carried out either within the main trunk or selectively in the anterior (frontal) and posterior (parietal) branches of the MMA. Embozene Microspheres (250 μm; Varian Medical Systems, Palo Alto, CA, USA) were slowly injected until complete stasis of flow was observed and contrast opacification of the distal branches ceased. In cases of bilateral CSDH, bilateral MMA embolization was performed. We monitored patients clinically and with serial head CT approximately monthly. The primary clinical outcome was time to improvement of presenting neurological symptoms (measured in days post-embolization). Radiographic outcomes included change in hematoma thickness and time to hematoma resolution. “Resolution” was defined as disappearance of the hematoma or a tiny residual collection on follow-up CT. We recorded any need for additional interventions (repeat embolization or surgical drainage) and any recurrence of the hematoma during follow-up.

Results

Seven patients (5 men, 2 women; mean age 68 years, range 51–82) met inclusion criteria. Six patients had VP shunts for hydrocephalus or related etiologies; one patient had SIH without shunt. Symptoms at presentation included unsteady gait (n = 3), headache (n = 2), aphasia (n = 1), and drowsiness due to decreased consciousness (n = 1). The mean hematoma thickness on initial CT was approximately 10 mm (range, 6–21 mm). All VP shunt patients had unilateral CSDH, whereas the SIH patient had a 21-mm left-sided hematoma (Figure 1). Importantly, none of the hematomas were evacuated via burr holes prior to embolization. Instead, six patients had their shunts corrected (five by valve adjustment and one by ligation) and one SIH patient underwent a targeted EBP before proceeding to embolization. All patients reported clinical improvement after embolization. Symptom relief was rapid: four patients improved within 3 days, one within 5 days, one by 7 days, and one (presenting with unsteady gait) by 14 days post-embolization. Headaches and focal neurological deficits (including gait disturbance and aphasia) all improved or resolved. Follow-up CT demonstrated gradual hematoma resorption. By 2–3 months, two patients had only minimal residual fluid, and all hematomas were completely or nearly completely resolved by 6 months (mean resolution time, ∼3 months). During follow-up, no patient required repeat embolization or surgical drainage, and no hematoma recurred. In summary, MMAE was associated with prompt symptom improvement and eventual radiographic clearance of the subdural collections in all cases, without recurrence. A representative case is shown in Figure 2: a 71-year-old male with a VP shunt presented with unsteady gait and a 10-mm left-sided hematoma. After adjustment of the shunt valve pressure, the patient underwent MMAE. Clinical symptoms improved within 3 days, and follow-up imaging at 3 months revealed minimal residual hematoma without recurrence.

Figure 1.

51-year-old male with spontaneous intracranial hypotension-associated chronic subdural hematoma. Initial CT scan shows a 21 mm left-sided hematoma with midline shift and symptoms of headache. The patient was treated with a targeted epidural blood patch followed by MMAE. Embolization was performed with 250 μm particles delivered slowly until near-stasis of flow in the distal branches of the middle meningeal artery. Symptomatic relief occurred within 2 days. Follow-up CT at 2 months shows complete resolution of the hematoma with no recurrence.

Figure 2.

71-year-old male with VP shunt overdrainage-associated chronic subdural hematoma. Initial CT scan shows a 10 mm left-sided hematoma with clinical presentation of unsteady gait. After shunt valve pressure adjustment, MMAE was performed using 250 µm particles injected slowly until near stasis of flow in the distal branches of the middle meningeal artery. Symptomatic improvement occurred within 3 days. Follow-up CT at 3 months shows minimal residual hematoma without recurrence.

Discussion

We report a case series demonstrating that MMAE, when combined with prior correction of underlying CSF hypovolemia, can safely and effectively treat CSF hypovolemia-related CSDH. Our findings align with and extend observations from traumatic CSDH. For instance, recent series involving primarily traumatic or idiopathic CSDH have shown that even modest or mildly symptomatic hematomas often improve after MMAE. Tanoue et al. described patients with mild CSDH treated with MMAE alone: 67% had partial or complete neurological recovery within 1 week, with median symptom resolution in less than 1 month. ¹⁹ Similarly, we observed that all seven of our patients (mean hematoma ∼10 mm) experienced clear improvement within 14 days. This suggests that, as in non-hypovolemic CSDH, embolization of the MMA can facilitate recovery even when the hematoma is relatively small or thin. In traditional practice, such small CSDHs might have been managed conservatively ³ and our series supports the idea that embolization can hasten resolution in symptomatic cases. While surgical decompression is generally reserved for patients with significant mass effect or neurological deterioration, ²⁰ many patients with CSF hypovolemia-related CSDH remain symptomatic despite modest hematoma size and corrected CSF pressure. In our cohort, although the primary etiology—either VP shunt overdrainage or spontaneous CSF leak—was treated, patients continued to exhibit clinical symptoms such as unsteady gait, headache, or aphasia. These residual symptoms correlate with the persistent presence of subdural collections, even in the absence of surgical indications. The principal risk of MMAE is inadvertent embolization of critical collateral vessels, such as the lacrimal and petrous branches. Embolic material may pass through existing vascular anastomoses to the facial nerve or orbital circulation, potentially causing post-embolization cranial nerve palsy or blindness. ²¹ Therefore, a comprehensive review of vascular anatomy on DSA and identification of any possible dangerous anastomoses before embolization were crucial. Currently, three primary categories of embolic agents are utilized for MMAE: liquid embolics (e.g. Onyx, N-butyl cyanoacrylate [nBCA], SQUID), particle embolics (e.g. Embozene Microspheres), and coils. Each class possesses distinct technical properties and safety considerations. At our institution, most MMAE procedures are performed under local anesthesia. Using particles avoids the pain and agitation that can be induced by DMSO-based liquid embolic agents, thereby improving patient comfort and procedural quality, and reducing the risk of embolic material migration into dangerous anastomoses. Pathophysiologically, CSDH is associated with neovascularization, ²² with the pathological vessels located at the most distal branches. Compared with the proximal-to-distal occlusion pattern of liquid embolic agents, particles offer superior distal penetration, which we consider more suitable for targeting these fragile neovessels. The rationale for employing MMAE in this context lies in its ability to accelerate hematoma resorption and reduce membrane vascularity, thereby promoting symptom resolution and preventing recurrence. Our data show that patients improved neurologically within days of embolization and experienced complete or near-complete hematoma resolution within months, without requiring surgical evacuation. Therefore, MMAE proved to be a valuable adjunct in symptomatic patients whose condition did not warrant surgery yet failed to improve with conservative management alone. This highlights an important therapeutic niche between observation and surgical intervention, particularly in vulnerable patients at risk of complications from either approach. These considerations suggest that MMAE is not redundant but rather occupies a distinct role in managing symptomatic CSDH in patients who do not meet criteria for burr-hole drainage but remain at risk for persistent mass effect. In line with the existing literature on VP shunt-related CSDH, some cases may improve with shunt revision alone, particularly when asymptomatic. ²³ However, in the presence of clear neurological symptoms or evidence of hematoma enlargement, additional intervention is warranted. Our findings support the use of MMAE in this subset as a minimally invasive, effective adjunct that bridges the gap between conservative care and surgical decompression. Previous reports on MMAE in CSF hypovolemia CSDH are limited. A recent World Neurosurgery series by Meggyesy et al. compared MMAE with standard care in shunted patients. They found hematoma resolution in six of seven MMAE patients, but time to clearance was not accelerated compared to standard care, with median resolution times ranging from 3 to 10 months depending on treatment group. ⁵ Their study lacked uniform inclusion criteria and focused on radiographic endpoints. Notably, clinical improvement was not systematically quantified. In contrast, our cohort included only symptomatic hematomas and we carefully documented clinical recovery. This may explain why we observed rapid clinical benefit even though radiographic resolution remained gradual. This discrepancy highlights that radiographic outcomes alone may underestimate patient improvement, underscoring the importance of clinical, symptom-based assessment. We also note that hematoma absorption is inherently slow. In our series the mean time to complete radiographic resolution was approximately 3 months (range, 2–6 months). This slow course is typical for CSDH. Importantly, however, neurological symptoms often resolved much sooner. Consistent with the findings of Tanoue et al., most of our patients returned to their baseline neurological status well before complete radiographic resolution was observed. ¹⁹ This dissociation implies that clinicians should not be alarmed by residual subdural fluid on CT if the patient is improving clinically. Rather, gradual radiographic resolution over several months is expected after MMAE. A principal takeaway from our series is the necessity of addressing the underlying CSF hypovolemia to ensure durable treatment outcomes. All patients in our cohort had shunt settings adjusted or an EBP performed before embolization. The favorable outcomes suggest that correcting the leak allowed MMAE to be fully effective. This contrasts with previous reports in which MMAE or even surgical evacuation proved ineffective when the underlying CSF leak was left unaddressed. Bhenderu et al. described cases of MMAE (and even surgical evacuation) only succeeded after multiple EBPs were performed for SIH. ¹¹ Similarly, Evran et al. noted that evacuation of SIH-related subdural hematoma without sealing the leak is not favored, as ongoing CSF drainage perpetuates the subdural space. ¹² Together, these observations support our conclusion that MMAE should be considered an adjunctive measure following definitive management of CSF hypovolemia, rather than a standalone intervention. Without correcting the underlying etiology, embolization alone is unlikely to prevent recurrence. Our series suggests that MMAE is a viable option for CSDH in the setting of CSF hypovolemia. It may be particularly useful in patients who are poor surgical candidates or who have recurrent or bilateral hematomas. By embolizing the meningeal arterial supply to the subdural membrane, MMAE appears to accelerate hematoma healing and symptom relief. However, our conclusions are limited by the small, retrospective nature of this series. Larger prospective studies, ideally incorporating standardized neurological outcome measures, are needed to better define patient selection and long-term efficacy of MMAE in this subgroup.

Conclusion

In our experience, MMAE is a safe and effective treatment for CSF hypovolemia-related CSDH when performed following correction of the underlying CSF leak or VP shunt overdrainage. All patients experienced rapid symptomatic improvement and eventual hematoma resolution over several months, with no recurrences. These findings derive from a small, single-center case series and should therefore be interpreted with caution. Our results are consistent with reports in patients with traumatic CSDH, suggesting that the benefits of MMAE may extend to this etiology. Crucially, addressing the underlying CSF hypovolemia must precede embolization, as MMAE alone is not sufficient. On this basis, we recommend that MMAE be considered for CSF hypovolemia-related CSDH as part of a multimodal treatment strategy. Larger, multi-center studies are warranted to further define optimal patient selection and procedural protocols.