A taxonomy for deep cerebral cavernous malformations: subtypes of basal ganglia lesions

OBJECTIVE Anatomical taxonomy is a practical tool that has successfully guided clinical decision-making for patients with brain arteriovenous malformations and brainstem cavernous malformations (CMs). Deep CMs are similarly complex lesions that are difficult to access and highly variable in size, shape, and position. The authors propose a novel taxonomy for deep CMs in the basal ganglia based on clinical presentation (syndromes) and anatomical location. METHODS The taxonomy system was developed and applied to an extensive 2-surgeon experience over 19 years (2001–2019). Lesions involving the basal ganglia were identified and subtyped on the basis of the predominant superficial presentation identified on preoperative MRI. Three subtypes of basal ganglia CMs were defined: caudate (31, 57%), putaminal (16, 30%), and pallidal (7, 13%). Neurological outcomes were assessed using the modified Rankin Scale (mRS). Postoperative mRS scores ≤ 2 were defined as a favorable outcome, and scores > 2 were defined as a poor outcome. Clinical and surgical characteristics and neurological outcomes were compared among subtypes. RESULTS Fifty-four basal ganglia lesions were identified in 54 patients. Each basal ganglia CM subtype was associated with a recognizable constellation of neurological symptoms. The most common symptoms at presentation were severe or worsening headaches (25, 43%), mild hemiparesis (13, 24%), seizures (7, 13%), and dysmetria or ataxia (6, 11%). Patients with caudate CMs were the most likely to present with headaches and constitutional symptoms. Patients with putaminal CMs were the most likely to present with hemibody sensory deficits and dysmetria or ataxia. Patients with pallidal CMs were the most likely to present with mild hemiparesis and visual field deficits. A single surgical approach was preferred (> 80% of cases) for each basal ganglia subtype: caudate (contralateral transcallosal-transventricular, 28/31, 90%), putaminal (transsylvian-anterior transinsular, 13/16, 81%), and pallidal (transsylvian supracarotid-infrafrontal, 7/7, 100%). Most patients with follow-up had stable or improved mRS scores postoperatively (94%, 44/47); mRS scores of > 2 at final follow-up did not differ among the 3 basal ganglia subtypes. CONCLUSIONS The study confirms the authors’ hypothesis that this taxonomy for basal ganglia CMs meaningfully guides the selection of surgical approach and resection strategy. Furthermore, the proposed taxonomy can increase the diagnostic acumen at the patient bedside, help identify optimal surgical approaches, enhance the consistency of clinical communications and publications, and improve patient outcomes.