Intrapulmonary arteriovenous anastomoses in humans with chronic obstructive pulmonary disease: implications for cryptogenic stroke?

What is the central question of this study? Do individuals with chronic obstructive pulmonary disease have blood flow through intrapulmonary arteriovenous anastomoses at rest or during exercise? What is the main finding and its importance? Individuals with chronic obstructive pulmonary disease have a greater prevalence of blood flow through intrapulmonary arteriovenous anastomoses at rest than age-matched control subjects. Given that the intrapulmonary arteriovenous anastomoses are large enough to permit venous emboli to pass into the arterial circulation, patients with chronic obstructive pulmonary disease and an elevated risk of thrombus formation may be at risk of intrapulmonary arteriovenous anastomosis-facilitated embolic injury (e.g. stroke or transient ischaemic attack). The pulmonary capillaries prevent stroke by filtering venous emboli from the circulation. Intrapulmonary arteriovenous anastomoses are large-diameter (≥50 μm) vascular connections in the lung that may compromise the integrity of the pulmonary capillary filter and have recently been linked to cryptogenic stroke and transient ischaemic attack. Prothrombotic populations, such as individuals with chronic obstructive pulmonary disease (COPD), may be at increased risk of stroke and transient ischaemic attack facilitated by intrapulmonary arteriovenous anastomoses, but the prevalence and degree of blood flow through intrapulmonary arteriovenous anastomoses in this population has not been fully examined and compared with age-matched healthy control subjects. We used saline contrast echocardiography to assess blood flow through intrapulmonary arteriovenous anastomoses at rest (n = 29 COPD and 19 control subjects) and during exercise (n = 10 COPD and 10 control subjects) in subjects with COPD and age-matched healthy control subjects. Blood flow through intrapulmonary arteriovenous anastomoses was detected in 23% of subjects with COPD at rest and was significantly higher compared with age-matched healthy control subjects. Blood flow through intrapulmonary arteriovenous anastomoses at rest was reduced or eliminated in subjects with COPD after breathing hyperoxic gas. Sixty per cent of subjects with COPD who did not have blood flow through the intrapulmonary arteriovenous anastomoses at rest had blood flow through them during exercise. The combination of blood flow through intrapulmonary arteriovenous anastomoses and potential for thrombus formation in individuals with COPD may permit venous emboli to pass into the arterial circulation and cause stroke and transient ischaemic attack. Breathing supplemental oxygen may reduce this risk in COPD. The link between blood flow through intrapulmonary arteriovenous anastomoses, stroke and transient ischaemic attack is worthy of future investigation in COPD and other populations.