Arteriovenous malformations (AVM)
Neurosurgery notes/Vascular/AV shunt/Arteriovenous malformations (AVM)

Arteriovenous malformations (AVM)

View Details
Status
Done

General

  • Abnormal collection of blood vessels where arterial blood flows directly into draining veins without normal interposed capillary beds.
  • No brain parenchyma within nidus
  • AVM are not congenital (not present at birth)

Definition

  • Aggregates of abnormal arteries and veins of variable diameters with direct connections through a nidus or fistula instead of a normal capillary bed.

Classification

Layers

  • Pial
  • Subcortical
  • Paraventricular
  • Combined

Lawton location (7 types of AVMs, which are collectively organised into 32 subtypes)

Frontal AVMs

  • Lateral frontal
  • Medial frontal
  • Paramedian frontal
  • Sylvian frontal

Temporal AVMs

  • Lateral temporal
  • Basal temporal
  • Medial temporal
  • Sylvian temporal

Parieto-Occipital AVMs

  • Lateral parieto-occipital
  • Medial parieto-occipital
  • Paramedian parieto-occipital
  • Basal occipital

Ventricular and Periventricular AVMs

  • Callosal
  • Ventricular body
  • Atrial
  • Temporal horn

Deep AVMs

  • Pure sylvian
  • Insular
  • Basal ganglial
  • Thalamic

Brainstem AVMs

  • Anterior midbrain
  • Posterior midbrain
  • Anterior pontine
  • Lateral pontine
  • Anterior medullary
  • Lateral medullary

Cerebellar AVMs

Mixed AVM

  • 6.5% of the reviewed cases were classified as Mixed AVMs.

Localisation

  • AVM have no predilection for specific central nervous system locations and are distributed proportionately within the brain mass.

Numbers

  • Prevalence
    • 0.14%
    • Adults 10 -100 per 100 000 (Morris et al., 2009).
  • Incidence:
    • 0.9– 1.5 per 100 000 head of population per year (Al- Shahi et al., 2003; Stapf et al., 2003)

Grading

Cerebral arteriovenous malformation grading systems
Cerebral arteriovenous malformation grading systems

Spetzler martin grading

Size
Small <3cm
1
Medium 3-6cm
2
Large >6cm
3
Eloquence
Non eloquent
0
Eloquent
1
Venous drainage
Superficial only
0
Deep
1
  • Surgical outcome as per 100 cases by Spetzler
    • SM Grade
      No. of pt
      No deficit
      Minor deficit
      Major deficit
      1
      23
      100%
      0%
      2
      21
      95%
      5%
      0%
      3
      25
      84%
      12%
      4%
      4
      15
      73%
      20%
      7%
      5
      16
      69%
      19%
      12%
  • Eloquent=Sensorimotor, language, visual cortex, hypothalamus, thalamus, internal capsule, brainstem, cerebellar peduncle, deep cerebellar nuclei
  • Superficial drainage = all drainage are superficial
    • Deep drainage = internal cerebral vein basil vein of Rosenthal or pre-central cerebellar vein
  • AVM mature at age 18 and tend to be more compact
Interpretation
  • Spetzler 2011
    • notion image
      Class
      SM score
      Treatment
      A
      I-2
      • Microsurgical resection is preferred treatment.
      • 8% chance on postoperative deficit
      B
      3
      • Multimodality treatment
      •18% chance on postoperative deficit
      C
      4-5
      • No treatment, with exception of recurrent hemorrhages, progressive neurological deficits, steal-related symptoms, and AVM-related aneurysms.
      • 32% chance on postoperative deficit

Lawton and Young supplementary score

  • = Spetzler martin grade + Lawton and Young supplementary score
  • Formed because the grade 3 or group B is a mixed bag, some are easier to tx and some are harder
  • Limitations
    • Results of high volume surgeons- generalizability?
    • Lack of externally validated outcomes
    • Subjectivity: Diffuse v compact
    • Useful framework for risk-assessment but doesn't replace judgment.
    • Characteristic
      Score
      A
      Age
      0-20
      1
      20-40
      2
      40+
      3
      B
      Bleeding
      No
      1
      Yes
      0
      C
      Compactness
      Compact
      0
      Diffuse
      1
    • Compact AVM has no brain matter in between nidus and diffuse if there is.
  • Interpretation
    • Graph showing receiver-operating characteristic analyses for the Spetzler-Martin grading system (blue curve) and the supplemented Spetzler-Martin (SM-Supp) grading system (red curve) in the combined cohort of 1009 patients. The predictive accuracy of the SM-Supp grading system was greater than that of the Spetzler-Martin grading system (area under the receiver-operating characteristics curve = 0.75 vs. 0.69, respectively; P<0.001)
      Graph showing receiver-operating characteristic analyses for the Spetzler-Martin grading system (blue curve) and the supplemented Spetzler-Martin (SM-Supp) grading system (red curve) in the combined cohort of 1009 patients. The predictive accuracy of the SM-Supp grading system was greater than that of the Spetzler-Martin grading system (area under the receiver-operating characteristics curve = 0.75 vs. 0.69, respectively; P<0.001)
      If below 7 the risk of surgery is significantly higher
      If below 7 the risk of surgery is significantly higher
  • Eg how the LY is a better and a more conservative risk calculator vs martin ponce
    • notion image

Pollock- Flickinger score

  • Calculates the likelihood of obliteration without deficit from focused irradiation
  • AVM score = (0.1)(AVM volume in cm3) + (0.02)(patient age in years) + (0.3)(location of lesion: frontal or temporal) = 0;
    • Location of lesion
      • Parietal, occipital, intraventricular, corpus callosum, cerebellar = 1
      • Basal ganglia, thalamic, or brainstem = 2
  • Outcome
      • Chance (in %) of excellent outcome (with 95% CI)
        • AVM score ≤1.00
          89 (79-94)
          AVM score 1.01 - 1.50
          70 (59-79)
          AVM score 1.51 - 2.00
          64 (51-75)
          AVM score >2.00
          46 (33-60)
      • Chance (in %) of modified Rankin Scale decline (with 95% CI)
        • AVM score ≤1.00:
          0 (0-8)
          AVM score 1.01 - 1.50:
          13 (7-22)
          AVM score 1.51 - 2.00:
          20 (12-32)
          AVM score >2.00:
          36 (24-50)

Comparison vs aneurysm

Description
AVM
Aneurysm
Ratio
1
6
Age of Dx
33
43
Haemorrhage presentation
50%
92%

Aetiology

  • Congenital origin
    • Bonnet- Dechaume Blanc syndrome
    • Wyburn- Mason syndrome
    • Hereditary haemorrhagic telangiectasia (HHT)
    • Autosomal dominant capillary malformation
      • Presents with skin capillary malformations
      • AVM of the brain, limb, or face (no intra- abdominal or intrathoracic organ involvement) has been described with a RASA1 mutation
  • Secondary

Pathophysiology

graph TD subgraph A ["Sporadic causes"] subgraph A1 ["Multiple gene involvement"] A11["Upregulation/downregulation<br>of multiple homeobox genes<br>(HoxD4 HoxB3)"] A12["Involve<br>with angiogenesis"] A11 --> A12 end A2["Somatic Kras mutations<br>(85% of AVM)"] end A --> C subgraph B ["Syndromic causes"] B1["Cerebrofacial arteriovenous<br>metameric syndrome (CAMS)"] B2["Hereditary haemorrhagic<br>telangiectasia (HHT)"] end B --> C subgraph C ["Models of AV shunt formation"] C1["Notch4 induced arteriovenous<br>shunt development from capillaries<br>(NOTCH3 gene In CADASIL)"] C2["Dilation and a primary<br>disorder of venules"] C3["Failure of regression of<br>primitive arteriovenous connections<br>during development"] end C --> D1 style C text-align:left D1["Formation of a nidus (conglomeration<br>of numerous AV shunts) that shunts bloods"] D1 --> E1 D1 --> D2 D2["High output Cardiac failure<br>(<1%)with pulmonary hypertension<br>can occur in neonates and infants"] E1["Due to lack of capillary bed"] E1 --> F1 F1["High pressure arterial bloods<br>enters venous system with low<br>resistance causes high flow of blood"] F1 --> G subgraph G ["Remodeling process"] subgraph G1 ["One"] G11["The high venous pressure and<br>high flow remodels the vein"] G12["dilate and walls to thin"] G11 --> G12 end subgraph G2 ["Two"] G21["The lower arterial pressure<br>remodels the artery"] --> G22["dilate"] end G3["This is why it is not congenitally<br>abnormal it requires time to form)"] G4["All these changes are mediated<br>through upregulation of eNOS and<br>down-regulation of endothelin along<br>with the remodelling vasculogenesis<br>factors (e.g. vascular endothelial<br>growth factor, VEGF)."] G5["There is usually no sharply<br>identifiable point where AVM can<br>be said to begin or end, rather<br>there is a transition zone reflecting<br>the pathological response to the<br>physiological perturbations."] end style G text-align:left G --> H1 G --> H2 G --> H3 G --> H4 G --> H5 G --> H6 H1["High flow vascular in arteries"] H1 --> H1_1 H1_1["Degeneration of remodelled<br>vessel wall"] H1_1 --> H1_2 subgraph H1_2 ["Aneurysm: (7% of AVM)"] H1_2a["Extranidal<br>Arterial aneurysms: located<br>on the wall of feeding arteries"] H1_2b["Unrelated aneurysm: arise<br>from vessels that are not<br>AVM feeders"] H1_2c["Flow related aneurysm: arise<br>from vessels that play a role in the<br>perfusion of the nidus and<br>(hemodynamically related to the AVM)"] H1_2d["Both can be either<br>Prenidal: proximal to the AVM nidus<br>Postnidal: distal to the AVM nidus"] H1_2e["Venous varices: located on<br>the wall of draining veins"] H1_2f["Intranidal:<br>Located within the boundaries of the nidus<br>Angiographically opacified before substantial<br>venous filling has occurred 75% in major<br>feeding arteries (inc. flow)"] end style H1_2 text-align:left H2["Haemorrhage (50%)"] H3["High flow in a low resistant<br>system will cause blood to be<br>stolen through the AVM"] H4["pressure is elevated within<br>the venous sinuses"] H5["If CPA AVM"] H6["Hb extravasation"] style A1 text-align:left style A2 text-align:left style B1 text-align:left style B2 text-align:left style C1 text-align:left style C2 text-align:left style C3 text-align:left style G1 text-align:left style G2 text-align:left style G3 text-align:left style G4 text-align:left style G5 text-align:left

Histopathology

Macroscopic

  • Arteriovenous malformations show dilated surface draining veins and feeding arteries with a deep nidus.

Microscopic

  • Arteriovenous malformations consist of variably sized abnormal arteries and veins with direct fistulous connections and intervening CNS tissue showing gliosis.

Natural history

General

  • Probably a registry data is better than a RCT in finding the best way to treat AVM due to its rarity
  • Patel et al., 2001: rarely spontaneously resolve
  • 25% of patients that do not experience a haemorrhage will have a decline in function within a 10- year period
    • Likely to be confined to larger AVM and be due to seizure or progressive neurological deficits

Bleeding risk

  • Un-ruptured bAVM ICH risk is 1% per year (n=2, 525 IPDMA)
    • Cohort
      Overall events
      Overall rate
      Overall 95% CI
      Hemorrhagic events
      Hemorrhagic rate
      Hemorrhagic 95% CI
      Non‑hemorrhagic events
      Non‑hemorrhagic rate
      Non‑hemorrhagic 95% CI
      All
      141
      2.32
      1.97–2.74
      85
      4.80
      3.88–5.94
      54
      1.30
      1.00–1.69
      UCSF
      28
      2.33
      1.61–3.37
      14
      4.88
      2.89–8.24
      14
      1.53
      0.91–2.58
      COL
      46
      3.50
      2.62–4.67
      35
      8.12
      5.83–11.31
      11
      1.24
      0.69–2.25
      SIVMS
      14
      2.37
      1.40–4.00
      9
      5.54
      2.88–10.85
      5
      1.17
      0.43–2.80
      KPNC
      53
      1.79
      1.37–2.34
      27
      3.04
      2.06–4.43
      26
      1.25
      0.85–1.84
    • CI = confidence interval; COL = Columbia; KPNC = Kaiser Permanente of Northern California; SIVMS = Scottish Intracranial Vascular Malformation Study; UCSF = University of California, San Francisco.
Ave risk of haemorrhage is 3%/year (aneurysm is 3% if untreated and 1% of treated)
  • Annual average haemorrhage rates for various AVM subgroup Stapf et al 2006
    • Venous drainage
      No prior haemorrhage
      Prior haemorrhage
      Nidus location
      No deep venous drainage
      0.9%
      4.5%
      Not deep
      No deep venous drainage
      3.1%
      14.8%
      Deep
      Deep venous drainage
      8.0%
      34.4%
      Deep
      Deep venous drainage
      2.4%
      11.4%
      Not deep
    • Memory
      • If prior haemorrhage: x5
      • If Deep venous drainage: x3
      • If deep nidus location: x3
The risk of bleed is not a constant number and it varies over the number years
  • Probably need to tx young patients more
Bleeding risk in SM Grade 4/5 Sattari et al 2024
Category
Annual Risk of Hemorrhage (Natural History)
Risk of haemorrhage Post-Surgery
Risk of haemorrhage Post-SRS
Risk of haemorrhage Post-Embolization
Cortical AVM
2.68%
0.74%
5.35%
16.96%
Deep-Seated High-Grade AVM
8.37%
5.25%
3.11%
22.33%
Annual and life time risk of haemorrhage
  • Risk of bleeding (at least once) = 1 - (annual risk of not bleeding)^expected years of remaining life
    • Assumption: constant risk of rebleeding after initial bleed
    • No change in risk during lifetime (which is actually false)
    • No difference in various location of AVM
Risk of haemorrhage increased with
(Koester et al 2023)
  • Presence of aneurysm (OR = 1.45 [1.19, 1.77], p < 0.001
  • Deep location (OR = 3.08 [2.56, 3.70], p < 0.001),
  • Infratentorial location (OR = 2.79 [2.08, 3.75], p < 0.001)
  • Exclusive deep venous drainage (OR = 2.50 [1.73, 3.61], p < 0.001) - into the Galenic system
  • Single venous drainage (OR = 2.97 [1.93, 4.56], p < 0.001),
  • Nidus size less than 3 cm (OR = 2.54 [1.41, 4.57], p = 0.002).
  • Previous haemorrhage
Morgan 2017
  • Proximal intracranial aneurysm (APIA)
  • Restriction of venous outflow
Risk of (Morgan 2017)
  • Neurological deficit or death after haemorrhage = 42%
  • Death alone 9%
Clinical presentation with haemorrhage occurs in approximately 50%
  • Untreated brain arteriovenous malformation | Neurology
    • Multicenter AVM research study (MARS) - age and previous haemorrhage predict bleeding
      • Survival curves of time-to-haemorrhage in patients with untreated brain AVM, by MARS cohort
        Survival curves of time-to-haemorrhage in patients with untreated brain AVM, by MARS cohort
        Forest plots of multivariable-adjusted predictors by cohort and combined IPDMA
        Forest plots of multivariable-adjusted predictors by cohort and combined IPDMA
        • 30% increase in risk of haemorrhage for each 10-year increase in age

Epilepsy

  • Josephson 2011
    • 5-year risk of first time seizure in AVM patients
      • If patient had ICH/FND: 23%
      • If patient had incidental AVMs: 8%
  • SIVMS: No difference in seizure outcome between Conservative vs Invasive treatment
  • Josephson 2015: Risk ratio intervention vs conservative (AED) = 0.99 (not much difference)
  • Not enough evidence to prove that surgery can reduce seizure risk
  • Rajeev 2022
    • 19% of seizure naive patient develop epilepsy after surgery
    • 1 year cumulative risk of 9%
    • Higher risk of seizure when
      • Temporal lobe AVM
      • History of haemorrhage

Evaluation

CT

  • Non-ruptured AVM
    • Slightly hyperdense mass with a sharp border with the surrounding normal brain
      • Calcification
  • Ruptured AVM
    • AVM may be obscured by the haematoma

MRI

  • Unruptured
    • Flow void on T1/T2 within AVM
  • Feeding arteries
  • Draining veins
  • Significant oedema around lesion may indicate a tumour that has bled rather than AVM
  • Gradient echo sequences help demonstrate surrounding hemosiderin which suggest a previous significant haemorrhage
  • A complete ring of low density (due to hemosiderin) surrounding lesion suggest AVM over neoplasm

Angiography

  • Allows for identification
    • Draining vein
    • Arterial supply
    • Nidus
      • Draining veins are in the same phase as arteries
    • Angioarchitectural features such as diffuseness
    • Presence of aneurysms
    • Venous stenosis

Differential

  • Proliferative angiopathy
    • A response to infarct or ICH
    • Particularly in the young
    • Should be considered in diffuse vascular lesions that may resemble AVM but lack the very early venous drainage pattern of AVM.
    • Do not share the same propensity to haemorrhage

Treatments

Difficult decision

Risk vs benefit
  • Natural history
    • Death 9%
    • Disability 20%
  • Intervention
    • Death 0-5%
    • Disability 1-24%

Conservative

  • Conservative treatment may be appropriate in large lesions where therapeutic risk exceeds projected natural
    history
  • Observation
    • Clinical signs (seizure/neurological deficits)
    • Radiologically
  • Evidence
    • ARUBA study: Poor study
      • Conservative medical management is better than intervention
      • Conservative management resulted in a 10% risk of stroke or death and a 15% risk of disability over 33 months.
    • Scottish intracranial vascular malformation study
      • Conservative management had better clinical (death/handicap/haemorrhage) outcome
        • Progression to the primary outcome during 12 years of prospective follow-up
          Progression to the primary outcome during 12 years of prospective follow-up

Intervention

General

Partial treatment is worse than natural history. IF YOU ARE NOT CONFIDENT TO COMPLETELY OBLITERATE IT DO NOT TOUCH IT WITH ANY MODALITY OF TREATMENT
  • Haemorrhage risk before treatment
    • n=61
    • 42/61 had haemorrhages before treatment
    • 22 haemorrhages under fu before treatment (3.49 years follow-up)
    • 14/42 with previous haemorrhage
    • 10.4% p.a. (95% CI, 2.2-15.4%)
  • Haemorrhage risk after treatment
    • 14 haemorrhages after treatment
    • 6.1% p.a. (95% CI, 2.5-13.2%)
  • 18/61 complete obliteration (and no haemorrhages)
notion image
  • Hence do not debulk an AVM
  • Hence you cannot palliatively treatment and AVM
Bleeding risk in SM Grade 4/5 Sattari et al 2024
Category
Annual Risk of Hemorrhage (Natural History)
Risk of haemorrhage Post-Surgery
Risk of haemorrhage Post-SRS
Risk of haemorrhage Post-Embolization
Cortical AVM
2.68%
0.74%
5.35%
16.96%
Deep-Seated High-Grade AVM
8.37%
5.25%
3.11%
22.33%
Goal
  • Obliterate the AV shunt completely
Indications for any treatment
  • Bleeding
  • Seizures
  • Fear
By grade
  • Spetzler-Martin Grade I and Il (Class A) AVMs
    • Can be managed with microsurgical resection alone, achieving good outcomes in 96% and 90% of patients, respectively
  • Class B or SpetzIer-Martin Grade Ill AVMs
    • Case-by-case evaluation with special consideration for multimodal therapy.
    • A conservative approach may be warranted in patients with these AVMs when they present without a history of rupture.
    • Surgery if favourable anatomical location and no deep feeders
  • Class C or Spetzler- Martin Grade IV and V AVMs,
    • Conservative
    • Intervention being considered only with progressive neurological decline and/or repetitive haemorrhage.
Evidence
  • ARUBA study RCT
    • Surgery had the highest cure rate
    • 100% when combined with endovascular treatment
    • Comparable morbidity
    • Cons
      • Small sample size
  • What did we learn from ARUBA, SIVMS and MARS?
    • Treatment of AVMs is associated with significant upfront risks.
      • How is risk stratified?
      • Institutional volume vs. outcome?
    • We fail to obliterate AVMs frequently- why?
      • Wrong choice of therapy?
      • Therapy not well executed?
      • Therapy not efficacious?
    • More/better clinical trials are required to test established first-line therapies in comparable lesions? e.g:
      • SMI/2 treated surgically vs. conservative treatment?
      • SM3 AVM multimodality treatment vs. conservative treatment?
      • Are RCTs really how to best address these questions?
Uncontroversial Statements to make in An Exam by Mr Walsh
  • Surgical extirpation is generally preferred in cases where there has been recent haemorrhage- if feasible as
    well as reasonably safe (prompt reduction in rebleeding - LYSScore - 0 for bleed
  • Surgery in most cases can be a planned procedure
    • 1st: decompress clot
    • 2nd come back for AVM
  • SRS is a reasonable alterative for small volume lesions that have haemorrhaged but without other
    reasonably safe treatment alternatives (Balance the natural history of the ruptured AVM against projected
    morbidity, efficacy and lead-time to occlusion)
  • INR may be curative for favourable architectures
    • Role of "palliative" embolisation unclear- one retrospective study suggesting slight improvement in
      rebleeding with partial treatment, many others which do not support that conclusion
    • Targeted endovascular treatment of arterial bleeding points alone may reduce short-term risk of rebleeding
      in high-grade lesions

Surgery

Options
  • Haematoma evacuation + elective AVM treatment
    • Is it safe YES: Beecher 2018 Delay treatment up to 4 weeks post <1% risk to patients
  • Elective AVM treatment
  • Definition of a favourable outcomes
    • Obliteration of AVM + absence of new permanent neurological deficit in first 6 wks of surgery+ MRS>1 at 12 months post OP
  • Timing
    • Ruptured AVM
      • Acute
        • Considered for a patient when a rapidly declining neurological status is attributed to a ruptured AVM.
        • Options
          • Debulking of haematoma
          • Treating AVM + debulking of haematoma
      • Delayed
        • General, a variable “rest period” (1–6 weeks) between the haemorrhage and the conclusive treatment.
        • Pros
          • Allow treatment planning for both radiosurgery and adjunctive embolization.
          • Allow hematoma resorption → better radiological demonstration of the AVM.
  • The treatment of choice for AVMs.
    • When surgical risk is unacceptably high, alternative procedures may be an option
  • Before surgery give 20mg PO QDS propranolol for 3 days to minimise post op normal perfusion pressure breakthrough → prevent post op bleed and oedema
  • Post op keep MAP 70-80mmHg
  • Pros
    • Eliminates risk of bleeding almost immediately.
    • Seizure control improves
    • Best treatment for managing seizures
      • Scottish intracranial vascular malformation study
      • Medication first then surgery for seizure control
  • Cons
    • Invasive
    • Risk of surgery
      • Not suitable for deep/eloquent AVM
    • Cost (high initial cost of treatment may be offset by effectiveness or may be increased by complications)
  • Delayed post op deterioration
    • Normal perfusion pressure break through and Occlusive hyperaemia:
      • Arteriolar
        • Since the AVM has low resistance to flow it causes reduce flow to normal high resistant arterioles in the periphery of the AVM → to compensate for this, the arterioles dilate to dec. resistance → chronic dilation causes arterioles to loose autoregulation
        • When AVM removed, the low resistant, high flow steal is gone → but autoregulation is not present so the remaining good arterioles cannot change in shape to modulate the increased flow of blood to the dilated arteries → increase local blood pressure → cerebral oedema and haemorrhage
      • Venous: post op the flow of blood in venous sinuses has dramatically reduce → predisposition to thrombosis which then causes back pressure to cause oedema and haemorrhage
      • Rebleed from a retained nidus of AVM
      • Seizures
    • Commonest cause of post op bleeding in AVM is due to uncomplete removal of nidus
  • Evidence
    • Against surgery
      • ARUBA study: RCT
      • Scottish intracranial vascular malformation study
    • For surgery
      • Morgan 2017
        • Risk of future AVM haemorrhage
        • Condition
          Time (Years)
          Risk of Haemorrhage (%)
          Annualized Risk (%)
          Without treatment (no haemorrhage)
          10
          16%
          1.8% for unruptured AVMs
          20
          29%
          Without treatment (with haemorrhage)
          10
          35%
          4.7% for 8 years for AVMs with haemorrhage followed by unruptured AVM rate
          20
          45%
      • 8-Year Risk for Unfavourable Outcomes in AVM Patients
        • Condition
          Size (cm)
          Risk of Unfavourable Outcome (%)
          No Deep vein drainage (DVD) or eloquent location
          1
          1%
          6
          9%
          Either DVD or eloquent location (not both)
          1
          4%
          6
          35%
          Both DVD and eloquent location
          1
          12%
          3
          38%

Endovascular techniques

Options
  • Embolization
Not to be used as Curative but supplement SRS or surgery
  • As embolization itself has increased complication rates
Technique
  • Occlude the arterial compartment of the AVM first to avoid bleeding complications associated with early occlusion of venous drainage.
    • In contrast to Spinal AVF which embolizes the vein first
Pros
  • Facilitates surgery
    • By reducing blood loss in large AVM
      • However, the overall management morbidity, mortality and success may not be improved with Embo prior to surgery approach as compared with surgery alone (Morgan et al., 2013; Bervini et al., 2014; Korja et al., 2014)
    • Securing associated aneurysm or deep arterial feeders
  • ? Fascilitates SRS
    • The subsequent radiosurgery is less likely to obliterate residual volume AVM than equivalent volumes of AVM treated by radiosurgery without prior embolization (Andrade- Souza et al., 2007).
Cons
  • Sometimes inadequate by itself to permanently obliterate AVMs,
  • Induces acute hemodynamic changes,
  • May require multiple procedures,
  • Embolization prior to SRS reduces the obliteration rate from 70% (without embolization) to 47% (with embolization)
Agents
  • Liquid agents: Onyx
    • Ethylene-vinyl alcohol (EVOH) copolymer (ethylene and vinyl alcohol) dissolved in dimethyl sulfoxide (DMSO) with micronised tantalum (for radio-opacity).
    • Not an adhesive → greater control with release → the best.
    • Onyx-18 corresponds to the viscosity
      • Onyx-18, Onyx-34 (both for AVMs)
      • Onyx-500 (for aneurysms).
    • When in-contact with aqueous solution (blood, water) → Solidifies through precipitation
    • Pathologic changes (similar to the acrylates) include: endothelial necrosis, acute inflammatory reaction, foreign body giant cells.
    • Onyx is bright on T1WI MRI
  • Particulates: polyvinyl alcohol (PVA) particles
    • Nidus obliteration is slower than with liquid agents → before complete obliteration → nidus is exposed to inc. Pressure → theoratical inc. The risk of haemorrhage
  • Acrylates: Adhesive
    • Can accidentally glue catheter to artery
    • Eg: NBCA (N-butyl cyanoacrylate)
Before using definitive treatment (surg or SRS)
  • Surgery: wait 3 -30 days (??)
  • SRS
    • Wait 30 days → immediate post embo angio looks amazing and you can leave out parts of AVM during SRS planning
    • Do not use radio-opaque material in embolization because will cause CT not useable for SRS planning
Delayed post embolization deterioration
  • Haemorrhage
  • Steal
  • Retrograde venous thrombus
Outcome
  • EVOH
    • Disabling morbidity, mortality, and urgent surgery occurred in 6.6% of cases and complete ablation of arteriovenous shunting in 27% (Morgan et al., 2013)
    • Morbidity was reported in 5.1%, mortality in 4.3% and AVM obliteration in 23.5% (Pierot et al., 2013).
  • Risk of embolisation
  • Anticipated obliteration rates for SPC A and SPC B bAVM treated by radiosurgery extrapolated from a report by Kano and colleagues. They also identified that bAVM of small diameters were more likely to obliterate within each of these two groups.
Obliteration rate for SPC A and B bAVM treated by radiosurgery
Obliteration rate for SPC A and B bAVM treated by radiosurgery
  • The cumulative freedom from haemorrhage following radiosurgery for SPC A bAVM is extrapolated from combining the natural history of haemorrhage after diagnosis derived from a meta-analysis, with the proportion that is not obliterated as reported from Kano and colleagues.
Radiosurgery for SPC A bAVM
Radiosurgery for SPC A bAVM

Radiation treatment

Conventional radiation
  • Effective in ≈ 20% or less of cases.→ not an effective therapy
Stereotactic radiosurgery (SRS) : accepted for some small (≤ 2.5-3 cm nidus), deep AVMs
  • 3 years, 3cm
  • Indicated
    • Deep inaccessible lesion
    • Lesion close to eloquent cortex
  • Mechanism
    • Radiation damages DNA in rapidly dividing endothelial cells → endothelial cells attempt to regenerate → they are depleted over time → eventually Intimal disintegration occurs → exposes smooth muscle cells and triggers a proliferative response in the medial layer → progressive growth of the media thickens the arterial wall and constricts the lumen → eventual vessel occlusion.
  • Technique
    • Leksell gamma knife
    • Treatment planning based on
      • MRI
      • Angiography
      • Both
    • Dose
      • 18 Gy or more
    • The radiation dose prescription was based on the risk of developing radiation-related complications, as predicted by the integrated logistical equation
      • AVM margin dosesAVM volume cm325 Gy220 Gy2-418 Gy4-816 Gy8-12<16 Gy rare12
      • The smaller the volume of AVM irradiated the steeper the fall off in dose of radiation delivered to the surrounding brain
  • Pros
    • Done as an outpatient,
    • Non-invasive,
    • Gradual reduction of AVM flow,
    • No recovery period
  • Cons
    • Takes 1-3 years to work (during that time there is a risk of bleeding, controversial whether it is increased or decreased)
    • Limited to lesions with nidus ≤ 3 cm
Outcome: Wegner 2010 determined at the patient’s last follow-up review
Excellent outcome
Complete nidus obliteration and development of no new neurological deficits.
Good outcome
AVM obliteration was also achieved but was associated with the development of a minor deficit (e.g., quadrantanopsia, ataxia, or cranial nerve injury) that did not interfere with the patient’s normal level of activities
Fair outcome
Obliteration was achieved but the patient developeda major deficit (e.g., hemiparesis, aphasia, or homonymous hemianopsia) that resulted in a decline in his or her level of functioning.
Unchanged
If follow-up imaging confirmed persistent arteriovenous shunting but had no new neurological deficits.
Poor outcome
Any patient who developed a new neurological deficit but had incomplete nidus obliteration
  • Nidus obliteration
    • Total nidus obliteration rate (on MRI/DSA) was documented in 198 patients (68%)
    • Median time to obliteration was 35 months
  • Complication
    • Developed permanent radiation related neurologic deficits
      13%
      Haemorrhage during the latency interval after radiosurgery
      6.5%
    • 10 (53%) died
    • 9 survived
      • 1 had a new permanent deficit from the bleeding event.
Maruyama 2005 et al
  • 11.5%/year bleed rate before SRS
  • 4.2%/year bleed rate after SRS but prior to complete obliteration
  • 0.6%/year bleed rate after complete obliteration post SRS
  • Calculation

Combination techniques

  • e.g. Embolization to shrink nidus then stereotactic radiosurgery