# Arteriovenous Malformations: Grading and Multimodal Treatment

## Overview

Brain arteriovenous malformations (AVMs) are congenital vascular lesions consisting of a nidus of abnormal arteries and veins with direct arteriovenous shunting without an intervening capillary bed. They affect approximately 0.1% of the population and are the most common cause of spontaneous intracerebral hemorrhage in young adults. Management options include microsurgical excision, stereotactic radiosurgery (SRS), endovascular embolization, or combinations thereof. The ARUBA trial has created significant controversy regarding treatment of unruptured AVMs.

## Pathophysiology

AVMs consist of abnormal direct arteriovenous connections without an intervening capillary network. High-flow, low-resistance shunting of blood passes from arterial feeders to draining veins. The nidus is a compact tangle of dysplastic vessels lacking normal vessel wall structure. The "steal" phenomenon, in which high-flow shunting diverts blood from surrounding normal brain, is debated. Venous hypertension from arterialized venous pressure may cause perilesional edema, gliosis, and neurological symptoms. The risk of hemorrhage comes from rupture of the nidus or associated aneurysms (either intranidal or flow-related).

## Epidemiology

The prevalence of brain AVMs is approximately 0.1% (1 in 1,000). The annual hemorrhage rate is 2-4% per year for unruptured AVMs and 4.5-7% for previously ruptured AVMs. The lifetime risk of hemorrhage can be roughly estimated as 105 minus the patient's age in years. Peak presentation occurs between 20 and 40 years of age, with no sex predilection. AVMs are rarely associated with hereditary syndromes such as hereditary hemorrhagic telangiectasia (HHT) or Wyburn-Mason syndrome.

## Clinical Presentation

Hemorrhage accounts for approximately 50% of presentations, manifesting as intracerebral, intraventricular, or subarachnoid hemorrhage with a mortality of 10-15% per hemorrhagic event. Seizures account for about 25% of presentations and result from perilesional gliosis or hemosiderin deposition. Headache occurs in approximately 15% and may be migrainous. Progressive neurological deficit from steal, venous hypertension, or mass effect accounts for about 5%. Some AVMs are discovered incidentally on imaging performed for other reasons.

## Risk Factors for Hemorrhage

Prior hemorrhage is the strongest risk factor, increasing the annual risk to 4.5-7%. Deep location (basal ganglia, thalamus, brainstem), deep venous drainage (particularly a single deep draining vein), and associated aneurysms (intranidal or flow-related) all increase risk. Paradoxically, smaller AVMs have higher rupture risk per year because of higher intranidal pressure due to less capacitance. Venous stenosis or ectasia that impairs venous outflow also increases intranidal pressure.

## Spetzler-Martin Grading System

### Components (Total: 1-5 points)

The Spetzler-Martin grade is determined by three factors. Size contributes 1 point for small (under 3 cm), 2 points for medium (3-6 cm), or 3 points for large (over 6 cm). Location contributes 0 for non-eloquent or 1 for eloquent cortex (sensorimotor, language, visual, hypothalamus, thalamus, brainstem, cerebellar peduncles, deep cerebellar nuclei). Venous drainage contributes 0 for superficial only or 1 for deep drainage.

| Feature | Points |
|---------|--------|
| **Size** | |
| Small (<3 cm) | 1 |
| Medium (3-6 cm) | 2 |
| Large (>6 cm) | 3 |
| **Eloquence** | |
| Non-eloquent | 0 |
| Eloquent | 1 |
| **Venous Drainage** | |
| Superficial only | 0 |
| Deep | 1 |

| SM Grade | Surgical Risk | Recommendation |
|----------|---------------|----------------|
| I | Very low | Surgery recommended |
| II | Low | Surgery recommended |
| III | Intermediate | Individualized decision |
| IV | High | Conservative or multimodal |
| V | Very high | Conservative management |

Grades I-II carry low surgical risk and microsurgical excision is recommended if symptomatic or ruptured. Grade III represents intermediate risk with treatment decisions individualized based on specific features. Grades IV-V carry high surgical risk and are often managed conservatively or with multimodal approaches.

### Supplementary Grading (Lawton-Young)

The supplementary Lawton-Young grading adds patient age, hemorrhagic presentation, and nidus compactness to refine surgical risk assessment. The combined Spetzler-Martin and Lawton-Young score improves outcome prediction.

<image>
Illustration of the Spetzler-Martin grading system for brain AVMs. Three panels show the scoring components: (1) nidus size measurement on angiography with small (<3 cm), medium (3-6 cm), and large (>6 cm) examples; (2) eloquent vs. non-eloquent cortex locations shown on a brain map with eloquent regions highlighted; (3) superficial vs. deep venous drainage patterns on lateral angiographic views. A summary table shows the point assignments and total grade calculation. Clean diagnostic teaching illustration.
</image>

## Diagnostic Imaging

### MRI/MRA

MRI shows flow voids within the nidus on T1 and T2 sequences, along with surrounding gliosis and hemosiderin if prior hemorrhage has occurred. MRA provides non-invasive evaluation of feeding arteries and draining veins. Functional MRI assesses the relationship to eloquent cortex.

### CT/CTA

CT identifies acute hemorrhage and may reveal calcification. CTA provides rapid vascular assessment in the acute setting.

### Digital Subtraction Angiography (DSA)

DSA is the gold standard for AVM evaluation. It defines feeding arteries, nidus morphology, draining veins, and flow dynamics. It identifies associated aneurysms (intranidal, flow-related on feeding arteries, or unrelated). It assesses for fistulous versus plexiform nidus morphology and evaluates transit time. Superselective catheterization of feeders provides detailed nidus anatomy and guides embolization.

## Treatment Modalities

### Microsurgical Excision

Microsurgery is the gold standard for immediate, complete AVM obliteration. It is indicated for SM grades I-III (especially I-II), ruptured AVMs with accessible hematoma, and progressive neurological deficits. Cure rates exceed 95% for SM grade I-II but decrease with higher grades.

The technique involves circumferential dissection around the nidus at the gliotic margin, identifying and coagulating or clipping arterial feeders while approaching from the arterial side. The draining veins must be preserved until the very end of resection, as premature venous occlusion causes nidus congestion and hemorrhage. The nidus is removed en bloc, the resection cavity is inspected for hemostasis, and intraoperative or postoperative angiography confirms complete excision.

Normal perfusion pressure breakthrough (NPPB) is theorized to cause hemorrhagic edema in surrounding brain after AVM removal due to loss of autoregulation in chronically hypoperfused tissue. It is managed by staged blood pressure reduction and controlled reperfusion. Occlusive hyperemia (swelling and hemorrhage from blood flow redistribution) may be mitigated by staged embolization and gradual blood pressure management.

### Stereotactic Radiosurgery (SRS)

SRS is indicated for small AVMs (under 3 cm or under 10 mL volume), deep or eloquent locations, poor surgical candidates, or residual AVM after surgery or embolization. The mechanism involves radiation-induced endothelial injury and progressive vessel obliteration over 2-3 years. Obliteration rates reach 70-80% at 3 years for AVMs under 3 cm but decrease with larger volumes. Marginal doses are typically 20-25 Gy for small AVMs. During the 2-3 year latency period, the AVM remains patent and carries ongoing hemorrhage risk. Complications include radiation necrosis (3-5%), radiation-induced edema, cyst formation, and rarely radiation-induced neoplasm. Volume-staged SRS treats portions of the nidus in sequential sessions for larger AVMs.

### Endovascular Embolization

Embolization is rarely curative as monotherapy (complete obliteration in only 10-20%). Its primary role is adjunctive, reducing nidus size and flow before surgery or SRS. Preoperative embolization reduces intraoperative blood loss and can target deep feeders difficult to access surgically. Agents include Onyx (ethylene vinyl alcohol copolymer), NBCA (n-butyl cyanoacrylate), and coils. Targeted embolization of flow-related or intranidal aneurysms reduces hemorrhage risk even if the nidus is not completely treated. Complications include hemorrhage during embolization (2-5%), stroke from non-target embolization, and cranial nerve injury. Curative embolization may be attempted for small, single-feeder AVMs with higher obliteration rates using modern liquid embolic agents.

### Multimodal Treatment

Combined approaches are used for complex AVMs (SM grade III-IV). Common combinations include embolization followed by microsurgery, embolization followed by SRS, or staged embolization with SRS. Treatment plans should be individualized based on nidus morphology, location, and patient factors.

## The ARUBA Trial Controversy

### Trial Design and Results

The ARUBA trial (A Randomized trial of Unruptured Brain Arteriovenous malformations, 2014) randomized 226 patients with unruptured AVMs to interventional therapy (any modality) versus medical management alone. The trial was stopped early due to superiority of medical management: the event rate was 30.7% in the intervention group versus 10.1% in the medical group at a mean follow-up of 33 months. The primary outcome was death or symptomatic stroke.

### Criticisms

The follow-up was short (mean 33 months), which does not capture the lifetime hemorrhage risk of an untreated AVM in a young patient. Treatment arms were heterogeneous (surgery, SRS, embolization, or combinations were not standardized). Predominantly high-grade AVMs populated the treatment arm, introducing selection bias. Only 18 patients had surgery alone, and embolization alone was used in many patients despite its low cure rate. Critically, ARUBA does not apply to SM grade I-II AVMs where surgical cure exceeds 95% with low morbidity. Many neurovascular experts continue to recommend treatment of low-grade unruptured AVMs, especially in young patients.

<image>
Lateral digital subtraction angiography of a Spetzler-Martin grade III brain AVM in the left temporal-parietal region showing enlarged feeding arteries from the MCA and PCA, a compact nidus approximately 4 cm in diameter, and early draining veins (superficial and deep) with rapid arteriovenous shunting. An intranidal aneurysm is visible within the nidus. An adjacent post-embolization angiogram shows reduced nidus filling after Onyx embolization of the dominant MCA feeder. Angiographic teaching illustration with labeled vascular components.
</image>

## Clinical Pearls

The draining vein must never be occluded before the nidus is devascularized. Premature venous outflow occlusion causes acute nidus congestion and massive hemorrhage and is the most common cause of intraoperative catastrophe during AVM surgery. Smaller AVMs have higher hemorrhage rates per year due to the paradox that small AVMs have higher intranidal pressure from less capacitance. ARUBA does not apply to ruptured AVMs, which should be treated to prevent re-hemorrhage (annual re-hemorrhage rate 4.5-7%). SM grade I-II AVMs in young patients should generally be treated surgically despite ARUBA, as the lifetime hemorrhage risk far outweighs the approximately 1-3% surgical morbidity for low-grade lesions. Intranidal aneurysms are the most common source of hemorrhage in AVMs and can be targeted with embolization even if the nidus is not treated. Postoperative angiography is mandatory because even a small residual nidus can recanalize and hemorrhage; DSA should confirm complete excision before leaving the operating room or within 24-48 hours. Normal perfusion pressure breakthrough is managed by strict blood pressure control in the immediate postoperative period, targeting systolic blood pressure of 120-140 mmHg to prevent hemorrhagic edema.

## References
- Spetzler RF, Martin NA. "A Proposed Grading System for Arteriovenous Malformations." *J Neurosurg*. 1986;65(4):476-483.
- Mohr JP, et al. "Medical Management with or without Interventional Therapy for Unruptured Brain Arteriovenous Malformations (ARUBA)." *Lancet*. 2014;383(9917):614-621.
- Lawton MT, et al. "A Supplementary Grading Scale for Selecting Patients with Brain Arteriovenous Malformations for Surgery." *Neurosurgery*. 2010;66(4):702-713.
- Starke RM, et al. "A Practical Grading Scale for Predicting Outcome after Radiosurgery for Arteriovenous Malformations." *J Neurosurg*. 2013;119(4):981-987.
- Derdeyn CP, et al. "Management of Brain Arteriovenous Malformations: AHA/ASA Scientific Statement." *Stroke*. 2017;48(8):e200-e224.