Neurosurgical Review (2024) 47:145

https://doi.org/10.1007/s10143-024-02383-3

RESEARCH

Middle meningeal artery patency after surgical evacuation for chronic

subdural hematoma
Georgios S. Sioutas1 · Oleg Shekhtman1 · Mert Marcel Dagli1 · Mohamed M. Salem1 · Sonia Ajmera1 ·
Sandeep Kandregula1 · Jan-Karl Burkhardt1 · Visish M. Srinivasan1 · Brian T. Jankowitz1

Received: 11 December 2023 / Revised: 27 March 2024 / Accepted: 30 March 2024

© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024

Abstract
Background Chronic subdural hematoma (CSDH) often requires surgical evacuation, but recurrence rates remain high.
Middle meningeal artery (MMA) embolization (MMAE) has been proposed as an alternative or adjunct treatment. There
is concern that prior surgery might limit patency, access, penetration, and efficacy of MMAE, such that some recent trials
excluded patients with prior craniotomy. However, the impact of prior open surgery on MMA patency has not been studied.
Methods A retrospective analysis was conducted on patients who underwent MMAE for cSDH (2019–2022), after prior
surgical evacuation or not. MMA patency was assessed using a six-point grading scale.
Results Of the 109 MMAEs (84 patients, median age 72 years, 20.2% females), 58.7% were upfront MMAEs, while 41.3%
were after prior surgery (20 craniotomies, 25 burr holes). Median hematoma thickness was 14 mm and midline shift 3 mm.
Hematoma thickness reduction, surgical rescue, and functional outcome did not differ between MMAE subgroups and were
not affected by MMA patency or total area of craniotomy or burr-holes. MMA patency was reduced in the craniotomy group
only, specifically in the distal portion of the anterior division (p = 0.005), and correlated with craniotomy area (p < 0.001).
Conclusion MMA remains relatively patent after burr-hole evacuation of cSDH, while craniotomy typically only affects the
frontal-distal division. However, MMA patency, evacuation method, and total area do not affect outcomes. These findings
support the use of MMAE regardless of prior surgery and may influence future trial inclusion/exclusion criteria. Further
studies are needed to optimize the timing and techniques for MMAE in cSDH management.

Keywords Chronic subdural hematoma · Middle meningeal artery embolization · Surgical evacuation · Middle
meningeal artery patency · Angiography

Introduction along with pro-inflammatory factors, exudate, and blood

degradation products [2]. Middle meningeal artery embo-
Chronic subdural hematoma (CSDH) is a common neuro- lization (MMAE) is hypothesized to resolve this repetitive
surgical pathology. However, treatment with surgical evacu- hemorrhage and a self-perpetuating inflammation [2, 3], and
ation results in high recurrence rates (up to 33%) [1]. CSDH has been described as an alternative treatment or adjunct to
occurs after initial hemorrhage in the subdural space fol- surgical evacuation of cSDH [4, 5].
lowing minor trauma [2]. Recurrent bleeding, proliferation Many studies have reported the use of MMAE as an
of dural border cells layer, and influx of inflammatory cells adjunct to surgical evacuation [4, 6, 7]. When MMAE is
lead to the organization of outer and inner membranes. Neo- performed after a craniotomy or burr hole evacuation,
vascularization in this layer and formation of fragile ves- the dura is already affected by the surgical operation. It is
sels lead to intermittent bleeding which fuels inflammation unclear if one can still perform adequate MMAE following
surgery. Here, we aimed to angiographically describe the
state of the middle meningeal artery (MMA) after surgical
evacuation at the time of MMAE. Specifically, we aimed
Brian T. Jankowitz
[email protected] to quantify the patency of MMA with a novel scale, assess
how surgical evacuation affects it, and how it correlates with
1
Department of Neurosurgery, Perelman School of Medicine, outcomes after MMAE.
University of Pennsylvania, Philadelphia, PA, USA

13
145 Page 2 of 10 Neurosurgical Review

Methods proximal to the main trunk bifurcation, presence of mem-

branes on selective external carotid artery or selective MMA
Patient selection angiography, midline/falx liquid embolic penetration, pres-
ence of dangerous collaterals (extracranial–intracranial
A prospectively-maintained database of consecutive anastomoses rendering MMAE to be unsafe), intraproce-
patients who underwent MMAE for cSDH in our tertiary dural (e.g., vessel dissection or perforation) and postproce-
cerebrovascular center from May 2019 to December 2022 dural complications (e.g., ischemic, hemorrhagic, blindness
was utilized. In our retrospective analysis, we identified 45 and facial palsy, access-related complications), number
consecutive cases of MMAE (index event) after a prior open and dimensions of burr holes and craniotomies, and MMA
surgery (performed at a prior or current admission). Addi- patency. Catheter position at the time of embolization was
tionally, we included 64 MMAE cases without a prior open defined as: (1) selective (catheterization of the frontal and/
surgery as a control group (“upfront MMAE”, May 2019 to or parietal branches for distal delivery), (2) non-selective
July 2022). Patients with atypical nonconvexity cSDH loca- (the microcatheter is placed proximal to the main trunk
tions (e.g., inter-hemispheric or tentorial) or cSDH due to bifurcation), and (3) using a combination of both previous
underlying lesions (e.g., vascular lesions or arachnoid cysts) techniques.
were excluded. The study was approved by the Institutional To assess the patency of the MMA, a six-point grading
Review Board (IRB), and informed consent was waived scale was created. The MMA territory was divided into 6
because of the study’s retrospective nature. areas. Contrast opacification in the pre-defined area dur-
MMAE was generally utilized in cSDH with no or mild ing pre-embolization angiography allowed one point (any
clinical symptoms and without significant motor deficits or amount of contrast reaching the area). To create the 6 areas,
altered mental status, or for patients with high surgical risk. two horizontal lines were drawn on the lateral view of bi-
Open surgery (alone or with adjunctive MMAE) was uti- plane angiogram. One line was an extension of the planum
lized in cases with significant motor deficits or midline shift sphenoiodale and the other line was drawn in parallel at a
more than 5 mm on CT necessitating immediate evacua- point approximately halfway between the inner table of the
tion according to the neurosurgeon’s discretion. Adjunctive parietal bone and the planum sphenoidale, which would
MMAE was performed either before open surgery or after roughly correlate to the superior temporal line. These two
open surgery (if emergent evacuation was required). parallel lines were bisected by a third, vertical line extend-
ing from the anterior clinoid process to the inner table of
Surgical technique and postoperative management the parietal bone. MMA segment 1 lies below the planum
sphenoidale extended line. This would represent the most
One or preferably two burr holes were drilled in the cranium proximal occlusion of the MMA, typically performed dur-
under general anesthesia, followed by durotomy, irriga- ing a craniotomy that extends to the middle fossa floor.
tion, and subdural drain placement according to the oper- Segments 2 and 3 represent the anterior and posterior divi-
ating neurosurgeon. Craniotomy involved the creation of a sions, respectively, spanning from the planum sphenoidale
large free bone flap under general anesthesia, usually 3 cm extended line to the second horizontal (temporal) line. Like-
or larger, followed by durotomy, irrigation, evacuation of wise, segments 4 and 5 extend above the aforementioned
blood products, subdural drain placement, and replacement line. In posterior-anterior view angiography, one point was
of the flap. Craniotomies (including minicraniotomies) were given if contrast reached the falx cerebri/midline (defined as
mostly performed in cases of cSDH with a significant acute at least 5 mm from the midline; Segment 6; Fig. 1). Three
component or recurrent cSDH with extensive organization independent evaluators assessed the angiographies and
or membranes. The size of the craniotomy was at the dis- rated each MMA with the above scale. Disagreements in
cretion of the operating neurosurgeon. Postoperative man- any MMA segment were resolved by a fourth evaluator.
agement included neurological monitoring, pain control, The longest diameter of the burr hole or craniotomy
prophylactic antiepileptic drugs as needed, labs, DVT pro- was recorded on axial CT slices, whereas the longest
phylaxis, wound care, rehabilitation, early mobilization, and height of craniotomy was recorded on a coronal CT slice.
follow-up imaging. If more than one burr hole was performed on a side, the
areas were summed. We calculated the area of burr holes
Data collection using the circle area formula (π/4)*(diameter)2. We calcu-
lated the area of craniotomies using the ellipse area formula
The following variables were collected for all cases: demo- π*((width)/2)*((height)/2).
graphics (age, gender, race), type of utilized embolic mate-
rial, prior surgical evacuation and type, MMA diameter

13
Neurosurgical Review Page 3 of 10 145

Fig. 1 (A) Drawing, (B) lateral,

and (C) anteroposterior projec-
tion of a selective MMA run
(no prior surgical evacuation),
where the segments of the middle
meningeal artery divided by the
6-point scale are shown. Because
all segments of MMA are visible,
this MMA would be rated with
a total of 6. (D) Drawing, (E)
lateral, and (F) anteroposterior
projection of a selective MMA
run with two previous burr holes.
Because all segments of MMA
are visible, this MMA would be
rated with a total of 6. (*) The
distal posterior segment of the
MMA crosses the parallel line.
(G) Drawing, (H) lateral, and
(I) anteroposterior projection of
a selective MMA run with prior
craniotomy. Because the whole
anterior division is not visible,
and the contrast does not reach
at least 5 mm from midline, this
MMA would be rated with a total
of 3

Outcome definitions Statistical analysis

To define treatment success, the primary clinical outcome For descriptive statistics, categorical variables are reported
was clinical failure (unplanned rescue surgical interven- as proportions, while continuous variables are reported as
tion after the index MMAE). The main radiographic out- mean and standard deviations (SD) or median and interquar-
come was radiographic success (at least 50% reduction in tile range (IQR) based on data normality. We used the chi-
maximal hematoma thickness on last follow-up). Technical squared test, Fisher’s exact test, t-test, Mann-Whitney U,
success was defined as self-reported successful, adequate or the Kruskal-Wallis test based on data normality to com-
embolization, when the target MMA trunk or divisions were pare baseline characteristics and angiographic outcomes
embolized with adequate flow stasis postembolization. between groups. We used linear and logistic regression to
Secondary outcomes included functional outcome at last examine associations between MMA patency and outcomes
follow-up assessed with the modified Rankin Scale (mRS), or total craniotomy/burr hole area.
with a favorable outcome defined as an mRS score < 3, Interrater reliability was calculated with intraclass corre-
cSDH thickness on last follow-up, and hospital length of lation coefficient (ICC) statistics for the total score of MMA
stay. The last visit after MMAE for clinical or radiographic patency. ICC estimates and their 95% confidence intervals
follow-up was defined as the last follow-up. The radio- were based on three raters, absolute-agreement, 2-way
graphic follow-up after MMAE was usually at 2 weeks, 6 mixed-effects model. Values below 0.5 suggested poor reli-
weeks, 3 months, and longer as indicated. ability, values ranging from 0.5 to 0.75 indicated moderate
reliability, values between 0.75 and 0.90 suggested good
reliability, while values exceeding 0.90 indicated excellent
reliability [8].

13
 145 Page 4 of 10 Neurosurgical Review

The statistical significance level was defined as P < 0.05, surgical evacuation, and 20 were MMAE after failed prior
and all analyses were performed with the Stata Statistical surgical evacuation (different admission). The median days
Software: Release 18. College Station, TX: StataCorp LLC. between surgery and MMAE were 5 (IQR 3–32). Twenty
MMAEs were performed after a previous craniotomy,
whereas 25 were after a previous burr hole evacuation. The
Results left side was embolized in 54.1% of the cases, radial access
was utilized in 56.9% of the cases, and 53.2% of procedures
Baseline characteristics were performed under general anesthesia. Various embolic
materials were utilized, with a combination of coils and par-
A total of 84 patients (median age 72 years, 20.2% females) ticles being the most utilized (38.5%). Dangerous extracra-
were included in this study and underwent 109 MMAE nial-intracranial anastomoses were encountered in 8.3% of
procedures. Pre-admission antiplatelet and anticoagula- cases. In most cases, when anastomoses to the orbit were
tion medication was recorded in 27.4% and 15.5% of the found, the catheter was moved distally beyond the anasto-
patients, respectively. The median admission platelet count motic point for embolization, whereas in two cases, only the
was 194.5 (IQR 142–237). Pre-procedural median mRS was parent artery or one branch were embolized with coils. Non-
< 3 in 47.6% of the patients, and 28.6% of patients under- selective catheterization of the MMA was utilized in 33.0%
went bilateral MMAE. The median hematoma thickness of the cases. The median MMA diameter was 1.1 mm (IQR
was 14 mm (IQR 10–19), with a median midline shift of 1.0-1.3) proximal to the main trunk bifurcation. The median
3 mm (IQR 0–7; Table 1). fluoroscopy time was 39.1 min (IQR 29.4–45.4; Table 2).
Maximum SDH thickness and midline shift were signifi- General anesthesia was utilized more often in MMAE
cantly smaller in the upfront MMAE group compared to all after surgical (burr hole) evacuation, while the median
surgical evacuation subgroups, while the admission platelet MMA diameter was greater in MMAE after surgical evacu-
count was higher when MMAE was performed after crani- ation (Supplemental Table 1, Supplemental Table 2).
otomy (Supplemental Table 1, Supplemental Table 2).
Technical success
Procedural details
Technical success was 98.2% because the intended embo-
Of all the MMAEs, 64 (58.7%) were upfront, standalone lization of a division was not performed in 2 cases in the
treatments for cSDH with no prior treatments, while 45 surgical evacuation groups. Specifically, in one case, a
(41.3%) were performed after a prior surgical evacuation microcatheter injection showed truncation of the anterior
(performed at a prior or current admission). Of the latter, MMA branch at the level of the craniotomy, with no supply
17 were adjunctive MMAE after surgical evacuation with- to the subdural neovasculature, and thus no particles were
out previously failed surgical evacuation, 8 were adjunctive injected into this branch. In a second case after burr hole
MMAE after surgical evacuation with previously failed evacuation, the posterior division was not feasible to cath-
eterize safely due to size and tortuosity, and thus was not
Table 1 Baseline characteristics embolized.
Patients n = 84 Midline/falx liquid embolic penetration did not differ
Females 17 (20.2%)
between the upfront MMAE and the surgical evacuation
Age (Years: median; IQR) 72 (64.5–78.2)
subgroups. Likewise, the rate of successful embolization in
Baseline antiplatelets* 23 (27.4%)
at least one MMA division (at least one selective emboliza-
   Aspirin 81 17 (20.2%)
   Clopidogrel 7 (8.3%) tion into divisions) did not differ between subgroups (Sup-
   Aspirin 325 2 (2.4%) plemental Table 1, Supplemental Table 2).
Baseline anticoagulation* 13 (15.5%)
   Warfarin 8 (9.5%) Radiographic and clinical outcomes
   Apixaban 3 (3.6%)
   Other 2 (2.4%) The median length of hospital stay was 6 (IQR 3-10.5)
Admission platelet count (median; IQR) 194.5 (142–237) days. No intraprocedural or postprocedural complications
Pre-treatment mRS < 3 40 (47.6%) occurred, apart from a small wrist hematoma. 78 (92.9%)
Bilateral procedure 24 (28.6%) patients had at least 2 weeks of follow-up. Radiological fol-
Lesions n = 109 low-up imaging was available for 89 cSDHs (81.7%), with a
Maximum SDH thickness (mm) (median; IQR) 14 (10–19)
median of 80 (IQR 34–193) days of radiological follow-up.
Midline shift (median; IQR) 3 (0–7)
At least a 50% reduction in maximal hematoma thickness
*4 patients were on both antiplatelets and anticoagulation medication

13
Neurosurgical Review Page 5 of 10 145

Table 2 Procedural details and outcomes

Procedural characteristics n = 109
Prior surgical evacuation (prior or current admission) 45 (41.3%)
   Adjunctive MMAE after surgical evacuation (same hospitalization) without previously failed surgical evacuation 17 (15.6%)
   Adjunctive MMAE after surgical evacuation (same hospitalization) with previously failed surgical evacuation 8 (7.3%)
   Failed prior surgical evacuation 20 (18.3%)
   Days between surgery and MMAE (median; IQR) 5 (3–32)
Left side 59 (54.1%)
Radial access 62 (56.9%)
General anesthesia 58 (53.2%)
Fluoroscopy time (minutes) (median; IQR) 39.1 (29.4–45.4)
Embolic material
   Particles + coils 42 (38.5%)
   Particles only 27 (24.8%)
   Liquid embolic only 31 (28.4%)
    n-BCA 29 (26.6%)
    Onyx 2 (1.8%)
   Coils only 6 (5.5%)
   Liquid embolic (n-BCA) + coils 3 (2.8%)
Catheter position during MMAE
   Non-selective (MMA trunk) 36 (33.0%)
   Selective (parietal and/or frontal divisions) 37 (33.9%)
   Combined approach 36 (33.0%)
At least one selective embolization into divisions 73 (67.0%)
MMA diameter proximal to bifurcation (mm) (median; IQR) 1.1 (1.0-1.3)
Midline/falx liquid embolic penetration 15/25 (60.0%)
Dangerous extracranial-intracranial anastomoses 9 (8.3%)
Technical success 107 (98.2%)
Radiographic outcomes n = 109
Imaging follow-up time (days) (median; IQR) 80 (34–193)
CSDH thickness on last available follow-up (mm) (median; IQR) 4.0 (0.5–7.5)
≥ 50% decrease in hematoma thickness on last follow-up 73/104 (70.2%)
Clinical outcomes n = 84
Hospital length of stay (days) (Median; IQR) 6 (3-10.5)
Surgical rescue 9 (10.7%)
Clinical follow-up time (days) (median; IQR) 146 (67–296)
mRS < 3 on last follow-up 46/81 (56.8%)

on last follow-up was achieved in 70.2% of the cases, with a thickness and midline shift were significantly smaller in the
median of 4.0 mm (IQR 0.5–7.5) of cSDH thickness on last upfront MMAE group. Also, the admission platelet count
imaging follow-up. Clinical follow-up data were available was higher in MMAE after craniotomy, general anesthesia
for 81 patients (96.4%), and 56.8% of them had a favorable was more often in MMAE after surgical (burr hole) evacua-
mRS score of < 3 on last follow-up. The median clinical tion, and the median MMA diameter was greater in MMAE
follow-up was 146 (IQR 67–296) days. Unplanned rescue after surgery, compared to upfront MMAE.
surgery was required in 10.7% of the patients (Table 2).
Comparing upfront MMAE with MMAE after surgical MMA patency and craniotomy area
evacuation (craniotomy or burr hole evacuation; Table 3,
Supplemental Table 1), craniotomy only, and burr hole Table 3 shows the results of MMA patency using the six-
evacuation only (Supplemental Table 2), there were no point scale. In the upfront MMAE and MMAE after surgi-
differences between utilization of radial access, fluoros- cal evacuation, the median total points were 5 (IQR 5–6),
copy time, embolic agents, catheter position in MMAE, and each part of MMA was equally patent. The MMA was
membranes on selective angiography, at least 50% hema- totally visible in about a third of cases. The ICC for the total
toma thickness reduction on last follow-up, surgical rescue, points of MMA patency was 0.609 (95% CI: 0.511–0.699,
and favorable mRS on last follow-up. Maximum cSDH p < 0.001).

13
145 Page 6 of 10 Neurosurgical Review

Table 3 Comparisons between the upfront MMAE and MMAE after surgical evacuation
MMA patency Upfront MMAE (n = 64) MMAE after surgical evacuation (n = 45) P-value
Bifurcation 64 (100%) 45 (100%) -
Anterior division - proximal 62 (96.9%) 43 (95.6%) 1.000
Posterior division - proximal 63 (98.4%) 45 (100%) 1.000
Anterior division - distal 61 (95.3%) 38 (84.4%) 0.089
Posterior division - distal 59 (92.2%) 43 (95.6%) 0.697
Falx 23 (35.9%) 18 (40.0%) 0.666
Total points, (median; IQR) 5 (5–6) 5 (5–6) 0.874
  ≤4 total points 6 (9.4%) 7 (15.6%) 0.327
   5 total points 37 (57.8%) 22 (48.9%) 0.357
   6 total points 21 (32.8%) 16 (35.6%) 0.766
Outcomes
≥ 50% SDH thickness reduction on last follow-up 41/59 (69.5%) 32/45 (71.1%) 0.858
Surgical rescue 7 (10.9%) 5 (11.1%) 1.000
Technical success 64 (100%) 43 (95.6%) 0.168
mRS < 3 on last follow-up 40/62 (64.5%) 24/43 (55.8%) 0.369

Table 4 Association of the MMA patency with outcomes after upfront MMAE
Variables MMA patency ≤ 5 total points MMA patency = 6 total points P-value
≥ 50% SDH thickness reduction on last follow-up 28/41 (68.3%) 13/18 (72.2%) 0.763
Surgical rescue 5/43 (11.6%) 2/21 (9.5%) 1.000
Technical success 43/43 (100.0%) 21/21 (100.0%) -
mRS < 3 on last follow-up 28/42 (66.7%) 12/20 (60.0%) 0.608

Table 5 Association of the MMA patency with outcomes following MMAE after surgical evacuation
Variables MMA patency ≤ 5 total points MMA patency = 6 total points P-value
≥ 50% SDH thickness reduction on last follow-up 19/29 (65.5%) 13/16 (81.3%) 0.322
Surgical rescue 3/29 (10.3%) 2/16 (12.5%) 1.000
Technical success 28/29 (96.6%) 15/16 (93.8%) 1.000
mRS < 3 on last follow-up 16/28 (64.0%) 8/15 (53.3%) 0.811

Comparing upfront MMAE with MMAE after craniot- 18.6), and the mean total craniotomy area was 3144.2 mm2
omy only, the anterior division–distal portion was patent in (SD 2114.8). The total score of MMA patency was associ-
95.3% and 70.0% of cases, respectively (p = 0.005; Supple- ated with the total area of craniotomy or burr holes combined
mental Table 2). This difference was reflected in the total (p < 0.001), with the area of craniotomy alone (p = 0.002),
points in the six-point scale (p = 0.044). Comparing upfront but not with the area of burr holes alone (p = 0.952; Supple-
MMAE with MMAE after burr hole evacuation only, no mental Table 3; Fig. 2). The total area of craniotomy or burr
differences in MMA patency were found (Supplemental holes did not correlate with at least 50% hematoma thick-
Table 2). MMA patency was not associated with at least ness reduction, percent hematoma thickness reduction, and
50% thickness reduction, percent thickness reduction, and reoperation (Supplemental Table 4).
reoperation in any subgroup (Tables 4 and 5, Supplemental In the cohort with MMAE after surgery, comparisons
Table 3). based on MMA catheterization are presented in Supplemen-
Of the 25 cases with burr hole evacuation, 2 burr holes tal Table 5. Specifically, there were no significant differences
(one frontal and one parietal) were performed in 18/25 in baseline characteristics and outcomes between catheter-
(72.0%) cases and one burr hole (frontal) in 7/25 (28.0%). ization at the main trunk only, selective catheterization only
The mean individual burr hole diameter was 11.7 mm (SD of at least one branch, and combined approach. However,
1.8), the mean individual burr hole area was 109.3 mm2 there were significant differences in the embolic agents used
(SD 31.1), and the mean total burr hole area was 192.5 mm (p < 0.001) and females (p = 0.028) for each subgroup.
(SD 70.9). Of the 20 cases with previous craniotomy, 10/20
(50.0%) were frontal, 5/20 (25.0%) frontoparietal, and 5/20
(25.0%) parietal. The mean craniotomy width was 61.2 mm
(SD 24.0), the mean craniotomy height was 54.5 mm (SD

13
Neurosurgical Review Page 7 of 10 145

Fig. 2 Scatter plot showing the correlation between MMA patency (0–6 total points according to the suggested 6-point scale) and the craniotomy
and/or the burr hole area

Discussion NCT04410146). Despite this ambiguity, our results sug-

gest that MMAE seems to be equally feasible and effective
This is the first study to examine the state of MMA and dura regardless of prior surgical evacuation.
mater after surgical evacuation at the time of MMAE. We We found that MMA patency pre-embolization was not
found that the MMA was less patent in the craniotomy group correlated with radiographic or clinical outcomes. In a mul-
at the distal portion of the anterior division, but patency was ticenter study by Salem et al. [12], along with midline shift,
not influenced by burr holes. MMA patency was associ- superselective MMA catheterization alone (frontal and/or
ated with the total area of craniotomy but not of burr holes. parietal branches, without targeting the main trunk) was
MMA patency and the total area of craniotomy or burr holes an independent predictor of radiographic failure. However,
were not associated with hematoma thickness reduction or radiographic failure was not associated with a combination
surgical rescue. These findings suggest that MMA patency of nonselective and selective catheterization, suggesting the
may be associated with surgical approach but not with out- importance of occluding the main trunk. Also, they found
comes after MMAE. that liquid embolic material use was an independent pre-
MMAE can be performed as a stand-alone treatment or dictor of clinical nonfailure. Interestingly, an in vitro com-
in combination with surgical evacuation for cSDH [4, 6, parison by Sadasivan et al. [13] concluded that the Squid-18
7]. Some prefer adjunctive MMAE after surgery because it liquid embolic agent may be preferable to Contour polyvi-
provides flexibility for the procedure; larger symptomatic nyl alcohol particles if thorough distal MMA penetration
cSDH can be evacuated urgently and then embolized semi- is needed. Kan et al. [14] have suggested that a successful
routinely [9, 10]. In most ongoing clinical trials with an procedure requires catheterization and embolization of both
adjunctive MMAE arm, MMAE is usually performed after distal MMA branches. However, Catapano et al. [15] found
the surgical evacuation (e.g. NCT04372147, NCT04500795, no differences in outcomes when examining the depth of
NCT04065113, NCT04750200, NCT05220826, and embolization (proximal only versus distal and proximal)
NCT04272996). However, others prefer preoperative and branches embolized (one versus two). Samarage et al.
embolization, assuming that the MMA is intact, allowing [16] showed that distal penetration of embolic material into
for distal penetration [11], better access, and efficacy. For the falx (n-butyl cyanoacrylate midline penetration, as evi-
that reason, some trials exclude prior craniotomies and per- denced by the “bright falx” sign on postembolization CT)
form adjunctive MMAE before surgical evacuation (e.g. may lead to more rapid improvement of cSDH. However,

13
145 Page 8 of 10 Neurosurgical Review

they also found no differences in outcomes between one Additionally, to assess the patency of MMA, we relied
versus two MMA branches embolized [16]. Similarly, Kho- on angiographic assessment of the visibility of contrast
rasanizadeh et al. [17] showed that embolizing more than in it, which may not fully reflect the actual patency of the
one branch was not associated with improved outcomes. artery. The scale we used in the study is new and has not
They excluded cases with prior craniotomy or adjunctive been validated before. The ICC for the total score of our
MMAE to avoid disruption of normal MMA anatomy by MMA patency scale indicated moderate reliability [8]. Even
the craniotomy [17]. Weinberg et al. found that membrane though the MMA segments were defined based on anatomi-
presence in cSDHs undergoing MMAE (stand-alone in cal landmarks in angiography, there is likely some inherent
73.7% with membranous and 61.0% with nonmembranous subjectivity in precisely delimiting the borders. The scale
cSDH) was not associated with recurrence, retreatment, or is based on a qualitative judgment of contrast opacification
complications [18]. Overall, embolizing only one branch rather than quantitative measurements. Validating the scale
seems sufficient, and proximal trunk embolization is more in an independent cohort would be important.
beneficial than superselective branch embolization. MMAE The study did not investigate the timing of MMAE in
can be performed even in cases with an absent MMA branch conjunction with surgical evacuation, which warrants fur-
after surgical evacuation by embolizing only one branch. ther investigation. Also, the median clinical follow-up was
According to a recent multicenter study by Chen et al. twice the median radiographic follow-up, and some patients
[19], the SDH thickness in the standalone MMAE group had short follow-up times, which could affect the interpre-
was smaller than the combined surgery and MMAE group. tation of the outcomes. Finally, even though a larger bone
Still, surgical recurrence rates were not significantly dif- opening leads to a higher probability of MMA involvement,
ferent between these two groups, similar to our results. the lack of association between outcomes and MMA patency
Recently, a multicenter study by Salem et al. [12] identified may stem from the fact that outcomes are multifactorial.
predictors of clinical and radiographic failure of MMAE. Future studies with larger sample sizes and comprehensive
They found that concurrent MMAE with surgery was not assessment of outcomes are needed to validate our findings.
associated with the need for surgical evacuation. Also, while
concurrent surgery with MMAE was associated with fewer
radiographic failures, this association declined with longer Conclusion
follow-up. SDH size was not a predictor for surgical recur-
rence after MMAE. Of note, 92.9% of patients in our study This study provides evidence that MMA remains relatively
had at least 2 weeks of follow-up; as demonstrated in previ- patent after surgical evacuation. MMA patency was reduced
ous studies, hematoma reduction started to manifest at least in the craniotomy group only, specifically in the distal por-
2 weeks after MMAE [12, 20]. Overall, the literature sug- tion of the anterior division, and correlated with the total
gests that while combining MMAE with surgical evacuation craniotomy area. However, neither MMA patency nor the
may lead to faster resolution of hematoma, SDH continues type and total area of surgical evacuation were associated
to decrease in size over time, rendering concurrent surgery with outcomes. These findings support the use of MMAE
potentially unnecessary for preventing surgical recurrence. regardless of prior surgical evacuation. Further studies are
Notably, our pre-MMAE and last-follow-up mRS scores warranted to validate these findings and shed more light on
appear somewhat higher than those reported in the literature the optimal timing and techniques for MMAE in conjunc-
[12, 21–23] which could be attributed to selection bias and tion with surgical evacuation for cSDH.
the limited size of our study. Nonetheless, despite the groups’
baseline differences, functional outcomes at the last follow- Supplementary Information The online version contains
supplementary material available at https://doi.org/10.1007/s10143-
up were similar between upfront MMAE and MMAE after 024-02383-3.
surgical evacuation. Future clinical trials should prioritize
comparing these two approaches. Acknowledgements None.

Limitations Author contributions All authors made substantial contributions.

Study conception/design: GSS and BTJ. Data acquisition: GSS, OS,
MMD, MMS, SA, SK. Data analysis: GSS. Data interpretation: all
The retrospective nature of this single-center study may authors. Manuscript drafting: GSS. Illustration: SK. Manuscript criti-
limit the generalizability of the findings. The sample size cal revision for important intellectual content: all authors. Approval of
was relatively small, potentially affecting the statistical final version of manuscript: all authors. Guarantor of the study: BTJ.
power of the study. Adjustments for multiple comparisons
Funding None.
were not made. Matching for baseline characteristics was
not performed due to the small sample size.

13
Neurosurgical Review Page 9 of 10 145

Data availability Deidentified data will be available from the corre- 12. Salem MM, Kuybu O, Nguyen Hoang A, Baig AA, Khorasani-
sponding author upon reasonable request. zadeh M, Baker C, Hunsaker JC, Mendez AA, Cortez G, Davies
JM, Narayanan S, Cawley CM, Riina HA, Moore JM, Spiotta
AM, Khalessi AA, Howard BM, Hanel R, Tanweer O, Levy EI,
Declarations Grandhi R, Lang MJ, Siddiqui AH, Kan P, Ogilvy CS, Gross BA,
Thomas AJ, Jankowitz BT, Burkhardt J-K (2023) Middle menin-
Competing interests The authors declare no competing interests. geal artery embolization for chronic subdural hematoma: predic-
tors of clinical and radiographic failure from 636 embolizations.
Ethical approval The study was approved by the University of Penn- Radiology. ;222045
sylvania Institutional Review Board with a waiver of informed con- 13. Sadasivan C, Dashti N, Marfoglio S, Fiorella D (2023) In vitro
sent. comparison of middle meningeal artery embolization with Squid
liquid embolic agent and contour polyvinyl alcohol particles. J
Neurointerv Surg
14. Kan P, Maragkos GA, Srivatsan A, Srinivasan V, Johnson J, Bur-
References khardt J-K, Robinson TM, Salem MM, Chen S, Riina HA, Tan-
weer O, Levy EI, Spiotta AM, Kasab S, Al, Lena J, Gross BA,
1. Liu W, Bakker NA, Groen RJM (2014) Chronic subdural hema- Cherian J, Cawley CM, Howard BM, Khalessi AA, Pandey AS,
toma: a systematic review and meta-analysis of surgical proce- Ringer AJ, Hanel R, Ortiz RA, Langer D, Kelly CM, Jankowitz
dures. J Neurosurg 121:665–673 BT, Ogilvy CS, Moore JM, Levitt MR, Binning M, Grandhi R,
2. Moshayedi P, Liebeskind DS (2020) Middle meningeal artery Siddiq F, Thomas AJ (2021) Middle meningeal artery emboliza-
embolization in chronic subdural hematoma: implications of tion for chronic subdural hematoma: a Multi-center experience of
pathophysiology in Trial Design. Front Neurol 11:923 154 consecutive embolizations. Neurosurgery 88:268–277
3. Shapiro M, Walker M, Carroll KT, Levitt MR, Raz E, Nossek E, 15. Catapano JS, Ducruet AF, Nguyen CL, Baranoski JF, Cole TS,
Delavari N, Mir O, Nelson PK (2021) Neuroanatomy of cranial Majmundar N, Wilkinson DA, Fredrickson VL, Cavalcanti DD,
dural vessels: implications for subdural hematoma embolization. Albuquerque FC (2021) Middle meningeal artery embolization
J Neurointerv Surg 13:471–477 for chronic subdural hematoma: an institutional technical analy-
4. Di Cristofori A, Remida P, Patassini M, Piergallini L, Buonanno sis. J Neurointerv Surg 13:657–660
R, Bruno R, Carrabba G, Pavesi G, Iaccarino C, Giussani CG 16. Samarage HM, Kim WJ, Zarrin D, Goel K, Chin-Hsiu Wang A,
(2022) Middle meningeal artery embolization for chronic subdu- Johnson J, Kaneko N, Nour M, Szeder V, Tateshima S, Jahan
ral hematomas. A systematic review of the literature focused on R, Duckwiler G, Colby GP (2022) The Bright Falx sign-Mid-
indications, technical aspects, and future possible perspectives. line Embolic Penetration is Associated with faster resolution
Surg Neurol Int 13:94 of chronic subdural Hematoma after Middle Meningeal artery
5. Sattari SA, Yang W, Shahbandi A, Feghali J, Lee RP, Xu R, Jack- embolization: a Case Series. Neurosurgery 91:389–398
son C, Gonzalez LF, Tamargo RJ, Huang J, Caplan JM (2023) 17. Khorasanizadeh M, Shutran M, Garcia A, Enriquez-Marulanda
Middle Meningeal Artery Embolization Versus Conventional A, Moore J, Ogilvy CS, Thomas AJ (2022) Middle meningeal
Management for Patients With Chronic Subdural Hematoma: A artery embolization for treatment of chronic subdural hematomas:
Systematic Review and Meta-Analysis. Neurosurgery does selection of embolized branches affect outcomes? J Neuro-
6. Salah WK, Baker C, Scoville JP, Hunsaker JC, Ogilvy CS, Moore surg. ;1–9
JM, Riina HA, Levy EI, Spiotta AM, Jankowitz BT, Michael 18. Weinberg JH, Akhter A, Zakeri A, Tanweer O, Zyck S, Tjou-
Cawley C 3rd, Khalessi AA, Tanweer O, Hanel R, Gross BA, makaris S, Jabbour P, Kan P, Peng J, Youssef P (2023) Middle
Kuybu O, Hoang AN, Baig AA, Khorasanizadeh MH, Mendez meningeal artery embolization for membranous versus non-mem-
AA, Cortez G, Davies JM, Narayanan S, Howard BM, Lang branous subdural hematomas: a retrospective and multicenter
MJ, Siddiqui AH, Thomas A, Kan P, Burkhardt J-K, Salem MM, cohort study. World Neurosurg
Grandhi R (2023) Middle meningeal artery embolization as a 19. Chen H, Salem MM, Colasurdo M, Sioutas GS, Khalife J, Kuybu
perioperative adjunct to surgical evacuation of nonacute subdural O, Carroll KT, Hoang AN, Baig AA, Salih M, Khorasanizadeh M,
hematomas: an multicenter analysis of safety and efficacy. Interv Baker C, Mendez Ruiz A, Cortez GM, Abecassis Z, Ruiz Rodrí-
Neuroradiol J Peritherapeutic Neuroradiol Surg Proced Relat guez JF, Davies JM, Narayanan S, Cawley CM, Riina H, Moore
Neurosci. ;15910199231162664 J, Spiotta AM, Khalessi A, Howard BM, Hanel RA, Tanweer O,
7. Court J, Touchette CJ, Iorio-Morin C, Westwick HJ, Belzile F, Tonetti D, Siddiqui AH, Lang M, Levy EI, Jovin TG, Grandhi R,
Effendi K (2019) Embolization of the Middle meningeal artery Srinivasan VM, Levitt MR, Ogilvy CS, Jankowitz B, Thomas AJ,
in chronic subdural hematoma - A systematic review. Clin Neurol Gross BA, Burkhardt JK, Kan P (2023) Standalone middle men-
Neurosurg 186:105464 ingeal artery embolization versus middle meningeal artery embo-
8. Koo TK, Li MY (2016) A Guideline of selecting and reporting lization with concurrent surgical evacuation for chronic subdural
Intraclass correlation coefficients for Reliability Research. J Chi- hematomas: a multicenter propensity score matched analysis of
ropr Med 15:155–163 clinical and radiographic outcomes. J Neurointerv Surg
9. Fiorella D, Hirsch JA, Arthur AS (2021) Embolization of the 20. Gomez-Paz S, Akamatsu Y, Salem MM, Enriquez-Marulanda A,
middle meningeal artery for the treatment of chronic subdural Robinson TM, Ogilvy CS, Thomas AJ, Moore JM (2021) Upfront
hematoma: considerations for pragmatic trial design. J Neuroin- middle meningeal artery embolization for treatment of chronic
terv Surg 13:295–297 subdural hematomas in patients with or without midline shift.
10. Schirmer CM, Siddiqui AH (2019) Commentary: middle menin- Interv Neuroradiol J Peritherapeutic Neuroradiol Surg Proced
geal artery embolization for chronic subdural hematoma: a Series Relat Neurosci 27:571–576
of 60 cases. Neurosurgery 85:E1004–E1005 21. Khorasanizadeh M, Chang Y-M, Enriquez-Marulanda A, Mizu-
11. Rudy RF, Catapano JS, Jadhav AP, Albuquerque FC, Ducruet AF hashi S, Salem MM, Gomez-Paz S, Siddiq F, Kan P, Moore J,
(2023) Middle Meningeal Artery Embolization to Treat Chronic Ogilvy CS, Thomas AJ (2021) Morphological changes in chronic
Subdural Hematoma. Stroke Vasc Interv Neurol [Internet]. subdural hematomas following upfront middle meningeal artery
;3:e000490. https://doi.org/10.1161/SVIN.122.000490

13
 145 Page 10 of 10 Neurosurgical Review

embolization: sequence, timing, and association with outcomes. J Publisher’s Note Springer Nature remains neutral with regard to juris-
Neurosurg. ;1–14 dictional claims in published maps and institutional affiliations.
22. Saway BF, Roth W, Salvador CD, Essibayi MA, Porto GBF, Dow-
lati E, Felbaum DR, Rock M, Withington C, Desai SK, Hassan Springer Nature or its licensor (e.g. a society or other partner) holds
AE, Tekle WG, Spiotta A (2023) Subdural evacuation port system exclusive rights to this article under a publishing agreement with the
and middle meningeal artery embolization for chronic subdural author(s) or other rightsholder(s); author self-archiving of the accepted
hematoma: a multicenter experience. J Neurosurg 139:131–138 manuscript version of this article is solely governed by the terms of
23. Salih M, Shutran M, Young M, Vega RA, Stippler M, Papavas- such publishing agreement and applicable law.
siliou E, Alterman RL, Thomas A, Taussky P, Moore J, Ogilvy
CS (2023) Reduced recurrence of chronic subdural hematomas
treated with open surgery followed by middle meningeal artery
embolization compared to open surgery alone: a propensity
score-matched analysis. J Neurosurg 139:124–130

13