Pulmonary Arteriovenous Malformation (AVM)


  • xxx


  • Hereditary Hemorrhagic Telangiectasia (HHT) (aka Osler Weber-Rendu (OWR) syndrome) (>84% of cases): autosomal dominant

  • Cavopulmonary Shunt/Atriopulmonary Shunt: shunt from SVC to pulmonary arteries

    • Role of hepatic metabolism: probably plays significant role in modulating the pulmonary vascular bed (supported by observation of intrapulmonary shunting occurring in hepatopulmonary syndrome and case reports that demonstrate AVM’s occurring when hepatic veins are unintentionally surgically routed past the pulmonary circulation)

    • After Glenn Anastomosis: by angiogram, 31% develop pulmonary AVM within 6.8 year f/u

    • Presence of pulmonary HTN in contralateral lung may increase risk of developing AVM
    • Microscopic pulmonary AVM (detected only by bubble Echo, not by pulmonary angiogram) may demonstrate pulmonary AVM within 2 hrs of the shunt

    • After birectional cavopulmonary anastomosis (BCPA), 99mTc-MAA studies detect shunts (due to AVM) ranging from 11-64% of cardiac output (vs 3-8% of cardiac output in controls)

  • Trauma (<0.5% of cases):

  • Sporadic Pulmonary AVM: usually single


  • Anatomic shunt from pulmonary to systemic circulation
    -Shunt of material in venous system: results in systemic embolization
    -Shunt of unoxgenated blood: results in hypoxemia
    -Effect of exercise: increased pulmonary blood flow (probably due to a greater than normal decrease in PVR, caused by less resistance to flow through the AVM) and decreased SVR (possibly due to underlying HHT or hypoxemic vasodilation)
    -Orthodeoxia (decreased pO2 when going from supine -> erect position): occurs due to increased flow in lower lobes with erect posture -> increased shunt (assuming lower lobe AVM)


-Resting Hypoxemia: due to shunt (arterial pO2 is inversely related to the size of the shunt fraction)
-Due to shape of oxygen dissociation curve, changes in arterial pO2 is more sensitive at detecting shunt than changes in SAO2
-Orthodeoxia (decreased pO2 when going from supine -> erect position, as vascular diltations are basilar-predominant): for shunts of >20%, SAO2 decreases about 6% (as compared to 0.3% for normals)
-Exercise Decrease in pO2: variable (pO2 decreases 6% for shunts >30% and decreases 1-2% for shunts <12%)

100% FIO2 Shunt Study: assess pO2 on 100% FIO2 to assess degree of shunt
-Actually measures “physiologic shunt”, which is equal to alveolar shunt (caused by V/Q mismatch) + anatomic shunt (caused by AVM)
-Measure arterial pO2 and hemoglobin after breathing 100% FIO2 x 15-20 min (deep breaths periodically to wash out nitrogen)
-Normal: 5% of cardiac output
-Calculate shunt as % of cardiac output (assuming O2 content of 5 mL per 100 mL): this calculation may underestimate the % shunt
-At threshold <535 mm Hg: 62% sensitivity/98% specificity
-Supine and Upright A-a Gradient on 100% FIO2: only 68% sensitive in detecting a shunt from an AVM

Exercise Study: high ventilatory drive during exercise (similar to patients with congential heart disease and right-to-left shunts) with preserved work capacity

Swan: PVR is usually normal (with no pulmonary HTN)

-FEV1: normal
-VC: usually normal (but may be decreased in some patients with large shunt)
-DLCO: moderately decreased (71-78% pred) with large shunt (especially in those with widespread, small AVM’s), but usually normal
-DLCO/VA: usually better preserved

CXR/Chest CT Pattern: CXR is abnormal in 60-90% of cases (but AVM may be subtle)
-Shape: round or oval, slightly lobulated, well-defined nodule (range from 1 to several cm)
-Number (most AVM’s are multiple): single AVM’s may be due to sporadic etiology/multiple AVM’s are usually due to HHT
–72% of cases have AVM’s in both lungs
-Location: lower-lobe predilection
-Calcification: occasional (probably due to phleboliths)
-Cavitation: none
-Maneuvers: AVM may change in size with Valsalva or Mueller maneuvers
-Growth: AVM may increase in size during puberty, pregnancy, or in presence of pulmonary venous hypertension (due to mitral stenosis or LV dysfunction)

Helical CT: useful to demonstrate AVM
-Contrast is usually not required

Chest MRI: less sensitive than CT or pulmonary angio, as small AVM’s with rapid flow are not visualized

Q Scan/Shunt Study: uses 99mTc-MAA and lung/right kidney imaging to evaluate % of shunt fraction (actually measures “anatomic shunt” only), as macroaggregated albumin will escape into systemic circulation if shunt exists
-AVM may produce shunt fraction of up to 60% of cardiac output
-Shunt fraction increases with standing (probably because most AVM are in lower lobes and gravity increases lower lobe flow with standing)
-Shunt fraction increases with exercise in patients with smaller AVM’s and increases with exercise in patients with larger AVM’s (due to fact that larger AVM’s do not enlarge further with exercise and smaller AVM’s are more compliant)
-At threshold of >5%: false-positive rate is 28% (with sensitivity of only 68%)

Pulmonary Angiogram: may demonstrate macroscopic, microscopic, or smaller diffuse AVM’s
-Allows identification of feeder vessels and assessment of rest of lungs prior to surgery

Transthoracic Echo with Bubble Study: best screening test for shunt from an AVM
-Immediate appearance of bubbles in left side: indicates intracardiac shunt (such as PFO, etc.)
-Delayed appearance of bubble in left side (after at least 4 cardiac cycles): indicates intrapulmonary shunt
-Sensitivity: 93-100% (negative predictive value: 97%)
-Specificity: 21-67% (lower specificity since 27% of normals have a PFO and intrapulmonary shunting may occasionally be seen in 5% of normals)
-Negative TTE + Negative CXR: excludes AVM in 100% cases
-Positive TTE with Bubble Study + Negative Spiral CT: suggests that the AVM is below the detection limit for the spiral CT -> should screen these cases with spiral CT every 1- years to periodically assess for AVM’s with >3 mm feeder vessels

Transesophageal Echo (TEE) with Bubble Study: more sensitive to exclude intracardiac shunt
-Immediate appearance of bubbles in left side: indicates intracardiac shunt (such as PFO, etc.)
-Delayed appearance of bubble in left side (after at least 4 cardiac cycles): indicates intrapulmonary shunt

Brain MRI: indicated to screen for brain AVM’s (even in the absence of pulmonary AVM’s)
CBC: polycythemia


Clinical Presentation of AVM:
1) Asymptomatic (50% of cases):

2) Respiratory Symptoms:
a) Dyspnea (48% of cases):
b) Platypnea: hypoxemia with upright position (as vascular diltations are basilar-predominant)
c) Pulmonary Bruit (49% of cases):
d) Clubbing (32% of cases):
e) Cyanosis (30% of cases):
f) Chest Pain (14% of cases):
g) Hemoptysis (11% of cases): due to endobronchial telangiectasia
-Increased risk during pregnancy
h) Hemothorax (<1% of cases): due to pleural telangiectasia

3) Embolic Phenomomena: serious neurologic events occur in 30-40% of AVM’s with feeder vessels >3 mm
a) Embolic CVA or TIA (27% of cases):
b) Brain Abscess (10% of cases):

4) Other Symptoms:
a) Anemia:
b) High-Output CHF: due to intrapulmonary shunt
c) Endocarditis:
d) Migraine:
e) Seizures:


Mortality Rate: 4-22% in untreated symptomatic cases
Spontaneous Regression of Pulmonary AVM: reported in some rare cases


1) AVM Embolization: especially indicated for AVM’s with feeder vessels >3 mm
-Safe in second/third trimester of pregnancy (with acceptable levels of radiation exposure)
-pO2 on 100% FIO2: useful to follow up after embolization
-Q Scan/Shunt Study: useful to follow up after embolization
2) AVM Resection: indicated for complex or large AVM’s, where embolization might not be effective