Hypoxia, not pulmonary vascular pressure, induces blood flow through intrapulmonary arteriovenous anastomoses

Abstract

Key points: Blood flow through intrapulmonary arteriovenous anastomoses (IPAVA) is increased by acute hypoxia during rest by unknown mechanisms. Oral administration of acetazolamide blunts the pulmonary vascular pressure response to acute hypoxia, thus permitting the observation of IPAVA blood flow with minimal pulmonary pressure change. Hypoxic pulmonary vasoconstriction was attenuated in humans following acetazolamide administration and partially restored with bicarbonate infusion, indicating that the effects of acetazolamide on hypoxic pulmonary vasoconstriction may involve an interaction between arterial pH and PCO2. We observed that IPAVA blood flow during hypoxia was similar before and after acetazolamide administration, even after acid-base status correction, indicating that pulmonary pressure, pH and PCO2 are unlikely regulators of IPAVA blood flow.

Abstract: Blood flow through intrapulmonary arteriovenous anastomoses (IPAVA) is increased with exposure to acute hypoxia and has been associated with pulmonary artery systolic pressure (PASP). We aimed to determine the direct relationship between blood flow through IPAVA and PASP in 10 participants with no detectable intracardiac shunt by comparing: (1) isocapnic hypoxia (control); (2) isocapnic hypoxia with oral administration of acetazolamide (AZ; 250 mg, three times a day for 48 h) to prevent increases in PASP; and (3) isocapnic hypoxia with AZ and 8.4% NaHCO3 infusion (AZ + HCO3 (-) ) to control for AZ-induced acidosis. Isocapnic hypoxia (20 min) was maintained by end-tidal forcing, blood flow through IPAVA was determined by agitated saline contrast echocardiography and PASP was estimated by Doppler ultrasound. Arterial blood samples were collected at rest before each isocapnic-hypoxia condition to determine pH, [HCO3(-)] and Pa,CO2. AZ decreased pH (-0.08 ± 0.01), [HCO3(-)] (-7.1 ± 0.7 mmol l(-1)) and Pa,CO2 (-4.5 ± 1.4 mmHg; P < 0.01), while intravenous NaHCO3 restored arterial blood gas parameters to control levels. Although PASP increased from baseline in all three hypoxic conditions (P < 0.05), a main effect of condition expressed an 11 ± 2% reduction in PASP from control (P < 0.001) following AZ administration while intravenous NaHCO3 partially restored the PASP response to isocapnic hypoxia. Blood flow through IPAVA increased during exposure to isocapnic hypoxia (P < 0.01) and was unrelated to PASP, cardiac output and pulmonary vascular resistance for all conditions. In conclusion, isocapnic hypoxia induces blood flow through IPAVA independent of changes in PASP and the influence of AZ on the PASP response to isocapnic hypoxia is dependent upon the H(+) concentration or Pa,CO2.

Figure 1. End-tidal gases ( and ) and minute ventilation (_E) at baseline and throughout 20 min of isocapnic hypoxia for all participants (n = 10)

Data points are 15 s means ± SE. AZ, acetazolamide intervention; AZ + HCO₃^–, bicarbonate correction intervention; control, isocapnic hypoxia intervention; , end-tidal partial pressure of oxygen; , end-tidal partial pressure of carbon dioxide.

Figure 2. Individual and group mean PASP during normoxia and isocapnic hypoxia during A, control, B, AZ (B) and C, AZ and intravenous NaHCO₃ (AZ + HCO₃^–)

A significant main effect of hypoxia (P < 0.001) and condition (P < 0.01) were found. For the condition interaction, post hoc analysis revealed a difference between the control and AZ conditions. Means ± SE of each condition are provided above the x-axis. *P < 0.001, hypoxia compared to normoxia; ^†P < 0.001, AZ compared to control. AZ, acetazolamide intervention; AZ + HCO₃^–, bicarbonate correction intervention; control, isocapnic hypoxia intervention; PASP, pulmonary artery systolic pressure.

Figure 3. Bubble scores during the isocapnic hypoxia trial in each condition (n = 10)

Each data point represents a participant with their corresponding bubble score on the y-axis. There were no significant differences between the conditions (P > 0.05). AZ, acetazolamide intervention; AZ + HCO₃^–, bicarbonate correction intervention; control, isocapnic hypoxia intervention.

Figure 4. Relationship between bubble score and A, PASP (mmHg), B, _c (l min⁻¹) and C, PVR (WU) during hypoxia in each condition

Multiple linear regression revealed no significant linear relationships (R² = 0.008, 0.08 and 0.0006 for (A), (B) and (C) respectively; all P > 0.05). AZ, acetazolamide intervention; AZ + HCO₃^–, bicarbonate correction intervention; control, isocapnic hypoxia intervention; PASP, pulmonary artery systolic pressure; PVR, pulmonary vascular resistance; _c, cardiac output.

Figure 5. Effect of isocapnic versus poikilocapnic hypoxia on the relationship between cardiac output (_c) and bubble score

Bubble scores and _c obtained during the last minute of exposure to 20 min of isocapnic hypoxia from the current study (n = 10) and poikilocapnic hypoxia from participants in the study by Laurie et al. (2010) (n = 12) breathing 12% O₂ for 30 min.