Long-term outcomes of untreated cerebral cavernous malformations: a prospective, population-based cohort study.

Background: Treatment decisions for cerebral cavernous malformations (CCM) currently rely on risk extrapolation from short-term studies, despite potential variations over time. This study aims to quantify the risks and functional outcomes of untreated CCMs during long-term follow-up. Methods: This population-based study included people aged ≥16 years in Scotland who were newly diagnosed with CCM between 1 January 1999 and 31 December 2003 or 1 January 2006 and 31 December 2010, using brain MRI or pathology. We analysed clinical events and functional outcomes using the Oxford Handicap Scale (OHS) during prospective follow-up without CCM intervention until 31 December 2023. The primary outcome was a composite of symptomatic intracranial haemorrhage (ICH) or new, non-haemorrhagic, persistent/progressive focal neurological deficit (FND) definitely or possibly related to CCM (ICH/FND). Findings: Among 300 patients (median age 44 years [IQR 32-57], 159 [53%] female, 48 [16%] brainstem CCM) included during 1999-2003 or 2006-2010, 81 (27%) presented with ICH/FND, 88 (29%) with epileptic seizure(s), and 131 (44%) incidentally. Over 4779 person-years of follow-up (completeness 95%), 44 patients were censored after microsurgical resection (n = 41) and stereotactic radiosurgery (n = 3). During a median untreated follow-up of 15 years (IQR 8-20), 40 (13%) patients experienced ICH/FND, 72 (24%) experienced dependence (OHS score 3-5), and 53 (18%) died (n = 7 [2%] related to CCM). The hazard of recurrent ICH/FND was higher than the first-event hazard (hazard ratio 8.66 [95% CI 4.44-16.90], p < 0.0001). The hazard rate of recurrent ICH/FND declined approximately 50-fold from 0.109 (95% CI 0.074-0.160) over the first five years to 0.002 (0.000-0.015) in the subsequent 20 years (p < 0.0001). Interpretation: The risk of a first ICH/FND from CCM is low, while the risk of recurrence is significantly higher, although the recurrence risk declines dramatically five years after a first ICH/FND. These long-term findings can guide clinical decision-making, and suggest focussing on a 5-year risk horizon rather than extrapolating annual risks to patients' lifetimes. Funding: Medical Research Council, Chief Scientist Office of the Scottish Government, and Stroke Association.