Patients with unruptured brain arteriovenous malformations (AVMs) who are allocated to conservative management, or ‘observation’, experience a low rate of neurological events, according to an analysis of the TOBAS registry presented recently by Jean-Christophe Gentric (Cavale Blanche Hospital, Brest, France) at the 2024 European Society of Minimally Invasive Neurological Therapy (ESMINT) congress (4–6 September, Marseille, France).
Gentric delivered an interim report on a total of 434 patients evaluated within the TOBAS registry, noting that the analysis included some patients with ruptured AVMs but primarily consisted of unruptured AVM cases (87%), with a roughly even split of low- and high-grade AVMs—as per the Spetzler-Martin grading system—across the entire cohort. He also disclosed a mean follow-up period of 3.3 years, resulting in a total of more than 1,300 patient years within the analysis.
The primary endpoint for this interim analysis was an AVM-related modified Rankin scale (mRS) score >2 at any time—although, for patients with an initial mRS score >2 at enrolment, the primary endpoint shifted to an increase of one point or more on mRS.
He went on to report that 23 out of 434 patients (5.3%) reached the analysis’ primary endpoint, and that seven of these patients had a low-grade AVM compared to 16 with high-grade AVMs. These numbers correspond to an incidence of death or new/worsened dependency of 1.7 per 100 patient years (95% confidence interval [CI], 1.1–2.5) for all patients in the study, and to an even lower incidence of 0.6 per 100 patient years (95% CI, 0.2–1.7) for low-grade, unruptured AVMs specifically.
Additionally, poor outcomes were found to be more frequent in ruptured AVMs and infratentorial AVMs, and in patients aged 55 years or older.
Regarding instances of intracranial haemorrhage (ICH), Gentric detailed that 38 patients (9%) experienced a new ICH, with 35 of these being major. The overall incidence of any new ICH was therefore found to be 2.8 per 100 patient years. Among patients with low-grade AVMs, the incidence of any new, major ICH was 1.3 per 100 patient years, and major AVM rupture occurred in 4% of these patients with one such occurrence (0.6%) being fatal.
Some of the finer points of Gentric et al’s analysis revealed that new, major haemorrhages were more frequent in AVMs in patients with a history of previous rupture; in high-grade AVMs or those that were large in size; and in AVMs with deep venous drainage.
Arriving at the analysis’ findings on serious adverse events (SAEs), the speaker stated that the rate of all SAEs—including haemorrhages—was 3.6 per 100 patient years across the entire cohort. Among low-grade, unruptured AVMs, this incidence decreased to just 1.8 per 100 patient years. Overall, SAEs were found to have been more common in ruptured, high-grade and infratentorial AVM types.
Outlining the analysis’ results in terms of major, non-haemorrhagic neurological events, Gentric noted an incidence of 0.6 per 100 patient years across the full cohort. He also highlighted the fact that 26 patients (6%) were treated during the follow-up period, including 11 with low-grade, unruptured AVMs. In roughly half of these cases—14 out of 26 and 6 out of 11, respectively—treatment was in response to a major neurological event, he added.
Providing a closer examination of these findings, Gentric reiterated that—across a follow-up period of three years—the analysis showed low risks of death or new/worsened disability (1.7% per year), but that new, major ICHs were the main driver among patients who did experience these negative outcomes (2.6% per year).
“This analysis reports clinical events during the follow-up of patients that clinicians did not want to treat,” Gentric added, also drawing attention to the fact that these results “are in line with previous studies that have almost universally identified rupture status, deep venous drainage, and infratentorial location, as risk factors for future haemorrhages and poor clinical outcomes”.
Gentric further highlighted absolute risks of major ICH in the analysis as being “very similar” to those in the existing literature as well, at 2–3% per year for unruptured and 4–8% per year for ruptured AVMs.
“In conclusion, observation was proposed to nearly half of the patients recruited in TOBAS, and patients observed over a mean of 3.2 years suffered from a relatively low rate of neurological events,” he added.
