Measuring blood flow through intrapulmonary and intracardiac shunts: a technical labyrinth

Abstract

Intracardiac and intrapulmonary pathways permitting the right-to-left shunting of venous blood can impair pulmonary gas exchange efficiency and provide passageways for the arterialization of venous blood clots. Intrapulmonary arteriovenous anastomoses (IPAVAs) are recruited with increasing exercise intensity, acute hypoxia, and when cardiac output is increased by physiological or pharmacological stress. Whether IPAVA recruitment is sufficient in magnitude to impair pulmonary gas exchange efficiency has been a matter of great debate. As with most fields of research, settling this debate probably requires the development of novel methodology. In this case, methodology is needed to permit the accurate measurement of blood flow and gas exchange through shunt pathways that may be susceptible to pre-capillary gas exchange. This is a void remaining to be filled; yet, several tools have been used separately to approximate blood flow through IPAVAs or gas exchange deficits due to shunt. Measuring blood flow through either pathway is technically challenging and typically relies on the lung’s ability to filter blood. It is reasoned that if particles larger in diameter than the pulmonary capillaries are injected into a peripheral vein and subsequently observed in the systemic arterial circulation, they must have travelled by way of cardiac or pulmonary shunt. In animal models, solid 25-μm-diameter microspheres can be injected into a peripheral vein and physically collected in the arterial circulation. Anatomical shunt fraction can be calculated using the fraction of microspheres collected and measures of cardiac output. In humans, shunt fraction can be assessed by injecting radiolabelled macroaggregates and using nuclear medicine to count the particles that have bypassed the lung. Agitated saline contrast echocardiography, another approach, uses a mixture of small air bubbles injected into a peripheral vein; detection of saline contrast in the left ventricle is a sure sign of intracardiac or intrapulmonary shunt. Saline contrast is more feasible, and, therefore, frequently used in humans, but it provides little detail with respect to the magnitude of shunted blood owing to methodological limitations (Hackett et al. 2016; Boulet et al. 2017). Finally, the multiple inert gas elimination technique (MIGET) provides an assessment of gas exchange through the infusion of six inert gases of varying blood solubilities. Measurements of gas retained in the blood and excreted from the lung are mathematically modelled to approximate the lung’s ventilation– perfusion distribution, including an estimated shunt fraction. MIGET is a highly complex technique successfully used by just a few worldwide. In this issue of The Journal of Physiology, Stickland et al (2019) conducted a highly technical study involving the combination of three techniques for measuring intracardiac/intrapulmonary shunt in anaesthetized canines. Measurements were made at rest, with dopamine and dobutamine to increase IPAVA recruitment, and finally, with a surgically induced intra-atrial shunt. Shunt magnitude was quantified using (1) left ventricular ultrasound contrast scores (0–5 ordinal scale) for agitated saline and microspheres, (2) the fraction of collected 25-μm microspheres and (3) the gas exchange shunt fraction as approximated by MIGET. Across all conditions the magnitude of shunt fraction measured by microspheres was 2.3 ± 7.4% and was reasonably similar to that measured by MIGET. However, the shunt fraction was much smaller when excluding the few intracardiac shunt studies (n = 4, 2 observations each). In this case, the mean shunt fraction measured by microspheres was <1% and the shared variance between MIGET and microsphere shunt measurements was greatly reduced. Sensitivity and specificity were determined for microsphere quantified shunt fractions of <1% and 1%. For ultrasound contrast scores of 0 and 1, saline contrast and microspheres were highly sensitive for shunt (100%) but specificity was poor (22% and 36%, respectively). When ultrasound contrast scores were 0–1 and 2 sensitivity declined (86% and 71%) but specificity improved (48% and 88%). In summary, when shunt fraction is small (<1%), there is modest agreement between MIGET and microsphere methods but saline contrast lacks in specificity. Although the results from Stickland et al. (2019) far from settle the debate on the contribution of cardiac and intrapulmonary shunt on the gas exchange deficits observed during exercise, they do highlight the limits of agitated saline contrast echocardiography in the detection of small shunts (<1%). Ultrasound contrast scoring systems are a non-linear function of shunt fraction (Duke et al. 2017). As a result, contrast scores of 1–2 represent small inconsequential shunt fractions while contrast scores > 3 may reflect a widening of the alveolar-to-arterial difference of oxygen (Elliott et al. 2014). Saline contrast is highly unstable and its rapid decay is sensitive to environmental and blood pressure, blood gases, temperature, and transpulmonary transit time (Hackett et al. 2016; Boulet et al. 2017). Technical improvement in how agitated saline contrast detection is quantified is clearly required to improve its specificity. Additionally, the appearance of contrast in the left ventricle is not a finite event, yet ultrasound contrast scoring systems treat them as such. Utilizing the entire ultrasound video loop and acoustic intensity from the right and left ventricle may permit a calculation of shunt fraction based on indicator dilution theory (Hackett et al. 2016). With these and other improvements in methodology, the use of agitated saline contrast echocardiography may have improved specificity for measuring blood flow through both small and large cardiac and intrapulmonary shunts.