Abstract DP175: Achieving better outcomes in treatment of vascular malformations in brainstem and other ‘critically’ eloquent brain regions - a novel surgical paradigm

Background: Critically eloquent brain regions such as the brainstem, thalamus, internal capsule, basal ganglia and hypothalamus pose a challenge for treatment of vascular malformations such as cavernous malformations. This is due to their deep location, with significant mortality and morbidity associated with surgical treatment. Surgery remains the treatment of choice, although the core dilemma of how—and how much—to traverse functional brain remains an unaddressed. In principle, reducing the size of pial entry should limit disruption to eloquent structures and improve functional outcomes. Aims: We sought to determine whether the size of pial entry relative to lesion size predicts and/or contributes to determining long-term functional outcomes related to microsurgical treatment of vascular malformations in critically eloquent brain regions. We evaluated whether the use of the flexible ominiguide microtip laser achieves smaller entry geometries and better outcomes as compared to conventional approaches. Methods: We retrospectively analyzed 234 patients (140 laser, 94 conventional) lesion and entry size. A subgroup of 188 patients (101 laser, 87 conventional) had preoperative and long-term follow-up modified Rankin Scale (mRS) scores. Entry-to-lesion size ratios were compared across groups. Functional improvement was defined as a reduction in mRS from baseline to last follow-up. Logistic regression, spline modeling, and a causal mediation analysis were used to evaluate the predictive and mechanistic role of entry geometries. Results: Laser-assisted resection resulted in significantly lower entry-to-lesion ratios than conventional surgery (mean 0.47 vs. 0.66; p < 0.001), with lower average entry size as compared to traditional microsurgery (4.9 mm vs 8 mm). In the outcome-assessable subgroup, lower entry-to-lesion ratios independently predicted functional improvement ( p = 0.045; OR = 0.981, 95% CI: 0.964–0.998, p = 0.029). Spline modeling identified thresholds corresponding to ‘ideal’, ‘safe,’ and ‘risk’ zones for recovery probability. Mediation analysis demonstrated approximately 51% of the benefit of laser surgery was mediated through a reduction in entry-to-lesion ratio ( p = 0.034). Conclusion: Entry-to-lesion ratio is a significant, quantifiable predictor of neurological recovery after resection of deeply placed vascular malformations. Laser microsurgery facilitates smaller, more precise pial and ependymal entry, which may mediate its functional advantage.