Sub-analyses of two randomised controlled trials (RCTs) have found that, while the extent of vascular penetration appears to have a minimal bearing on outcomes, both procedural sequencing and surgical technique selection may significantly influence the therapeutic benefits of middle meningeal artery embolisation (MMAe) in chronic subdural haematoma (cSDH) patients.
These analyses from the MEMBRANE and EMBOLISE RCTs were presented for the first time by Hazem Shoirah (Mount Sinai Health System, New York, USA) and Jason Davies (State University of New York, Buffalo, USA), respectively, at the 2025 Society of Vascular and Interventional Neurology (SVIN) annual meeting (19–22 November, Orlando, USA).
After primary findings from both trials were disclosed last year, indicating positive results with MMAe as an adjunctive therapy in cSDH management, investigators set out to elucidate more granular factors that may impact the endovascular procedure’s efficacy.
The MEMBRANE RCT demonstrated a positive treatment effect with MMAe using Trufill n-Butyl cyanoacrylate (n-BCA; Johnson & Johnson) plus standard of care versus standard of care alone, as well as comparable safety profiles between the two approaches. Shoirah and colleagues subsequently conducted a post-hoc, exploratory analysis to assess how the degree of vascular penetration—either ‘good’ (proximal ligation or embolisation without reaching the midline) or ‘excellent’ (reaching the midline or beyond)—may affect patient outcomes up to six months. Across 186 intention-to-treat patients, degree of penetration was found to be excellent in 107, good in 73, and non-existent in six. At baseline, more patients with excellent penetration had hypertension and greater haematoma thickness/volume.
Similar proportions of patients with excellent versus good penetration achieved complete haematoma resolution at six months (75.3% vs 69.6%, respectively) and, while there was a trend towards patients with excellent versus good penetration having a higher percentage change in volume at one month (-40.2% vs -29.1%, respectively), this was not the case at three or six months. Additionally, there was no difference between excellent and good penetration for reoperation/post-randomisation surgical procedures (4.4% vs 3.3%, respectively), nor residual or re-accumulation of cSDH (>10mm; 6.7% vs 7.1%) at six months. Rates of change in haematoma volume were not significantly different between groups; however, effect sizes were largest for excellent versus good penetration at one month, with no difference at three or six months.
These findings led Shoirah and colleagues to conclude that there were “minimal differences” in outcomes between patients with good versus excellent vascular penetration—although the “small trend” towards larger effect sizes seen at one month after embolisation in excellent-penetration patients suggests that deep penetration may improve resorption rates in the immediate post-treatment period. The researchers also believe that the relatively limited numbers of patients in this analysis likely influence the observed outcomes, with larger studies being needed to more thoroughly investigate vascular penetration’s significance in MMAe patients.
The EMBOLISE RCT—an adaptive study primarily comparing MMAe with the Onyx liquid embolic agent (Medtronic) as an adjunct to surgical drainage versus surgical drainage alone—showed that the former approach is able to significantly reduce cSDH recurrence in symptomatic patients. Despite this, Davies and his co-investigators felt that optimal procedural sequences as well as whether or not efficacy varies by surgical technique “remained undefined”, leading them to conduct a subgroup analysis on how these variables might alter clinical outcomes.
Among the 185 patients enrolled in EMBOLISE who received adjunctive MMAe, 78 (42.2%) underwent surgery before embolisation, while the remaining 107 (57.8%) underwent embolisation first. The investigators report that, “remarkably”, all six of the trial’s prespecified primary endpoint failures in the MMAe group occurred when embolisation preceded surgery, as per a 0% failure rate with ‘surgery-first’ compared to 5.6% with ‘embolisation-first’. Furthermore, patients receiving surgery before embolisation demonstrated a superior haematoma volume reduction at 90 days (-65.4% vs -32.6%, respectively), reduced midline shift normalisation (-89.3% vs -79.9%, respectively), and lower rates of neurologic deterioration (9.1% vs 15%, respectively), compared to those undergoing embolisation prior to surgery. Technical success, meanwhile, was 100% regardless of timing.
Moving on to discuss analyses by surgical technique, Davies shared that—among 208 burr hole patients—adjunctive MMAe significantly reduced recurrence (2.9% vs 10.7%, respectively) versus surgery alone. Conversely, among 190 craniotomy patients, the benefits of adjunctive MMAe were not found to be statistically significant compared to surgery alone (5.5% vs 12.1%, respectively). Baseline characteristics also showed that craniotomy patients had larger haematomas by mean thickness (21.8mm vs 20.6mm, respectively) and more complex morphologies as per the rate of septated haematomas (57.9% vs 38.5%, respectively) versus burr hole patients. According to Davies, interaction analyses revealed that the combination of burr hole drainage alongside surgery performed before embolisation yielded optimal outcomes, with zero treatment failures being observed in this subgroup—while surgery-related serious adverse events were higher with craniotomy compared to burr hole (23.1% vs 10.5%, respectively). Embolisation-related events remained low, at 2.2% overall, regardless of timing or surgical approach.
Based on these new data, the researchers conclude that “critical factors” influencing MMAe’s efficacy in cSDH management have now been identified, as performing surgical drainage before embolisation eliminated treatment failures and enhanced haematoma resolution, and the therapeutic benefit of adjunctive embolisation was more pronounced with burr hole compared to craniotomy procedures. As averred by Davies and colleagues, these results have “immediate implications” for clinical practice and protocol development in the neurointerventional management of cSDH patients.
