On this page

The central question in brain AVM research is whether treating an AVM that has never bled does more good than harm. One randomized trial and several large observational studies have tried to answer it, with different results. Evidence on treating AVMs after a hemorrhage, and on individual procedures, comes mainly from case series and expert guidance.

StudyDesignPatientsMain finding
ARUBA (2014, 2020)Randomized trial226 adults with unruptured AVMsDeath or stroke: 15 of 110 with medical management vs 41 of 116 with intervention over about four years
MATCH (2025)Registry analyzed to emulate a trial1,770 patients with unruptured AVMs in ChinaFive-year hemorrhage-free survival 96.2% with intervention vs 89.0% without; no benefit for high-grade or diffuse AVMs
Scottish cohort (2014)Population-based cohortAdults with unruptured AVMs in ScotlandConservative management associated with better outcomes

ARUBA

ARUBA randomly assigned adults with unruptured AVMs to medical management alone or medical management with intervention (surgery, embolization, radiosurgery or a combination). Randomization was stopped early, at an average follow-up of 33 months, when 10.1% of the medical group and 30.7% of the intervention group had died or had a stroke1. After an average of 50 months, the difference persisted: 15 of 110 against 41 of 1162. Critics note the trial's short follow-up, the mix of treatments, and that some interventions were still incomplete when randomization stopped.

MATCH

The MATCH registry followed 1,770 patients with unruptured AVMs at Chinese centers from 2011 to 2021, and the 2025 analysis was designed to mimic a randomized trial using statistical methods. Intervention was associated with better hemorrhage-free survival at five years (96.2% against 89.0%) and ten years, but not for high-grade AVMs or diffuse nidus, and only microsurgery was consistently associated with lower hemorrhage risk. Patients weren't randomly assigned, and the main outcome was hemorrhage rather than all strokes, disability or seizures3.

Population cohort

A Scottish population-based study found that conservative management of unruptured AVMs was associated with better outcomes than intervention over its follow-up4.

Guidelines and consensus

The American Heart Association's 2017 scientific statement covers the anatomy, grading and treatment options for AVMs5. The Society of NeuroInterventional Surgery's 2022 standards cover imaging and multimodality treatment, including embolization6. A 2024 expert roundtable, ARISE, calls for multidisciplinary care and better prospective studies7. The International Stereotactic Radiosurgery Society's guidance covers radiosurgery for low-grade AVMs8.

Recurrence after removal

A 2023 study found that AVMs can recur after surgery confirmed complete removal on angiography, particularly in younger patients, which is why follow-up imaging matters9.

References

  1. Mohr JP et al. Lancet. 2014;383:614–621. doi:10.1016/S0140-6736(13)62302-8. Source ↩
  2. Mohr JP et al. Lancet Neurol. 2020;19:573–581. doi:10.1016/S1474-4422(20)30181-2. PubMed ↩
  3. Han H, Chen Y, Ma L et al. JAMA Netw Open. 2025;8:e2543408. doi:10.1001/jamanetworkopen.2025.43408. Source ↩
  4. Al-Shahi Salman R et al. JAMA. 2014;311:1661–1669. doi:10.1001/jama.2014.3200. PubMed ↩
  5. Derdeyn CP et al. Stroke. 2017;48:e200–e224. doi:10.1161/STR.0000000000000134. Official Top Things to Know summary. Source ↩
  6. De Leacy R et al. J NeuroIntervent Surg. 2022;14:1118–1124. doi:10.1136/neurintsurg-2021-018632. Source ↩
  7. Samaniego EA et al. Stroke. 2024;55:1449–1463. doi:10.1161/STROKEAHA.124.046725. PubMed ↩
  8. Graffeo CS et al. Neurosurgery. 2020;87:442–452. doi:10.1093/neuros/nyaa004. Source ↩
  9. Järvelin P et al. Neurosurg Rev. 2023;46:99. doi:10.1007/s10143-023-02001-8. Source ↩