Increased cardiac output, not pulmonary artery systolic pressure, increases intrapulmonary shunt in healthy humans breathing room air and 40% O₂

Abstract

Blood flow through intrapulmonary arteriovenous anastomoses (IPAVAs) has been demonstrated to increase in healthy humans during a variety of conditions; however, whether or not this blood flow represents a source of venous admixture (/) that impairs pulmonary gas exchange efficiency (i.e. increases the alveolar-to-arterial difference (A–aDO₂)) remains controversial and unknown. We hypothesized that blood flow through IPAVAs does provide a source of /. To test this, blood flow through IPAVAs was increased in healthy humans at rest breathing room air and 40% O₂: (1) during intravenous adrenaline (epinephrine) infusion at 320 ng kg⁻¹ min⁻¹ (320 ADR), and (2) with vagal blockade (2 mg atropine), before and during intravenous adrenaline infusion at 80 ng kg⁻¹ min⁻¹ (ATR + 80 ADR). When breathing room air the A–aDO₂ increased by 6 ± 2 mmHg during 320 ADR and by 5 ± 2 mmHg during ATR + 80 ADR, and the change in calculated / was +2% in both conditions. When breathing 40% O₂, which minimizes contributions from diffusion limitation and alveolar ventilation-to-perfusion inequality, the A–aDO₂ increased by 12 ± 7 mmHg during 320 ADR, and by 9 ± 6 mmHg during ATR + 80 ADR, and the change in calculated / was +2% in both conditions. During 320 ADR cardiac output () and pulmonary artery systolic pressure (PASP) were significantly increased; however, during ATR + 80 ADR only was significantly increased, yet blood flow through IPAVAs as detected with saline contrast echocardiography was not different between conditions. Accordingly, we suggest that blood flow through IPAVAs provides a source of intrapulmonary shunt, and is mediated primarily by increases in rather than PASP.

Key points

• The contribution of blood flow through intrapulmonary arteriovenous anastomoses (IPAVAs) to pulmonary gas exchange efficiency remains unknown and controversial.

• Intravenous infusion of adrenaline (epinephrine) increases blood flow through IPAVAs detected by the transpulmonary passage of saline contrast and breathing 40% O₂ minimizes potential contributions from ventilation-to-perfusion inequality and diffusion limitation.

• Pulmonary gas exchange efficiency was impaired to the same degree, and the transpulmonary passage of saline contrast was not different, in humans at rest during the intravenous infusion of adrenaline before and after atropine when breathing room air and 40% O₂.

• Cardiac output increased to the same degree during intravenous infusion of adrenaline before and after atropine, but pulmonary artery systolic pressure only increased significantly before atropine.

• These data demonstrate that blood flow through IPAVAs contributes to pulmonary gas exchange efficiency and that blood flow through IPAVAs is predominantly mediated by increases in cardiac output rather than increases in pulmonary artery systolic pressure.

Introduction

Pulmonary gas exchange efficiency is defined and quantified by the alveolar-to-arterial difference (A–aDO₂) where an A–aDO₂ >0 mmHg reflects imperfect pulmonary gas exchange that could be caused by three distinct mechanisms: (1) incomplete end-capillary O₂ diffusion equilibration (diffusion limitation), (2) alveolar ventilation-to-perfusion (/) inequality, and (3) right-to-left shunt. Sources of right-to-left shunt in subjects without an intracardiac shunt (e.g. patent foramen ovale (PFO)) include postpulmonary shunt (e.g. venous blood from the bronchial and Thebesian circulations) and intrapulmonary shunt. In humans without a PFO, blood flow through intrapulmonary arteriovenous anastomoses (IPAVAs) can be detected using saline contrast echocardiography (Meltzer et al. 1981; Roelandt, 1982; Meerbaum, 1993). Although blood flow through IPAVAs in humans is absent or trivial at rest (Elliott et al. 2013), blood flow through IPAVAs increases during exercise (Eldridge et al. 2004; Stickland et al. 2004; Lovering et al. 2008a; Elliott et al. 2011), while breathing hypoxic gas mixtures at rest (Laurie et al. 2010), and during the intravenous (i.v.) infusion of inotropic drugs (Bryan et al. 2012; Laurie et al. 2012). The physiological significance of these findings remains controversial (Hopkins et al. 2009) but it has been hypothesized that blood flow through IPAVAs could provide a source of venous admixture and increase the A–aDO₂ (Stickland et al. 2004; Lovering et al. 2008a).

Bryan et al. (2012) demonstrated that during the i.v. infusion of dobutamine and dopamine in healthy humans at rest breathing room air, the degree of left-sided contrast increased suggesting an increase in blood flow through IPAVAs, yet the A–aDO₂ remained unchanged. However, in accordance with the Fick principle of mass balance, this necessitates an increase in the calculated shunt fraction (Q_S/) due to an increase in mixed venous O₂ content (O₂) secondary to an increased and constant whole body . In that study, increases in left-sided contrast, which never exceeded intermittent boluses of contrast in the left ventricle, occurred concomitant with increases in that were ∼50% above resting values. Similar work by Laurie et al. (2012) in healthy humans at rest without a PFO breathing room air demonstrated that i.v. infusion of adrenaline (ADR) increased blood flow through IPAVAs. In that study the highest dose of ADR (320 ng kg⁻¹ min⁻¹) corresponded to an ∼2-fold increase in and pulmonary artery systolic pressure (PASP) and the qualitative degree of left-sided contrast ranged from intermittent boluses to heterogeneous filling of the left ventricle. Accordingly, the data from Bryan et al. support the potential that blood flow through IPAVAs may provide a source of venous admixture, yet in light of the work by Laurie et al. it remains possible that higher and PASP would have corresponded with greater degrees of left-sided contrast and therefore, potentially more blood flow through IPAVAs resulting in a measurable increase in the A–aDO₂.

In addition to potential contributions from blood flow through IPAVAs, pulmonary gas exchange efficiency may also be impaired by / inequality, diffusion limitation and postpulmonary shunt. Determination of the A–aDO₂ via arterial blood gas analysis and calculation of alveolar () does not yield information concerning the relative contribution of each of these factors. This is particularly relevant during interventions known to increase blood flow through IPAVAs (e.g. during exercise, breathing hypoxic gas mixtures, or the i.v. infusion of catecholamines) where the pulmonary capillary transit time required for complete end-capillary O₂ equilibration, mixed venous , and driving gradient for O₂ across the alveolar–capillary interface may vary considerably. As a result, contributions from / inequality, diffusion limitation and shunt may differ during these different interventions, and thus obscure the ability to draw conclusions regarding the potential contributions from each to the A–aDO₂ unless some of these contributions are held constant.

Accordingly, an experimental protocol that would eliminate / inequality and diffusion limitation while increasing blood flow through IPAVAs in healthy humans at rest would be optimal to determine if blood flow through IPAVAs contributed to pulmonary gas exchange efficiency. Elevating the alveolar partial pressure of O₂ () via breathing 100% O₂ at sea level theoretically eliminates / inequality and diffusion limitation (Riley & Cournand, 1949); however, blood flow through IPAVAs during exercise and the i.v. infusion of ADR has been shown to be significantly reduced or eliminated when breathing 100% O₂ (Lovering et al. 2008b; Elliott et al. 2011; Ljubkovic et al. 2011; Bryan et al. 2012; Laurie et al. 2012). Thus, breathing 100% O₂ isolates the contribution of right-to-left shunt, but this would not include the contribution of blood flow through IPAVAs, and therefore may not represent the same volume of right-to-left shunt when breathing room air. Instead, breathing <100% O₂ but >21% O₂ has the potential to sufficiently elevate such that the contributions of / inequality and diffusion limitation are minimized or prevented, yet still permit blood flow through IPAVAs.

In subjects at sea level, breathing 40% O₂ elevates to the mid-200 mmHg range, so it is increased >2-fold compared to breathing room air. The impact of this increased driving gradient for O₂ on the rate of O₂ uptake across the alveolar–capillary interface can be model-predicted using a computer algorithm implementing a Bohr integration for the lung (Wagner & West, 1972; Wagner, 1982). According to this model, in subjects at rest breathing room air, complete end-capillary O₂ equilibration occurs within ∼0.4 s, which is reduced to ∼0.1 s when breathing 40% O₂. Importantly, mean pulmonary capillary transit time is calculated to be ∼0.7 s whether breathing room air or 40% O₂. Although on average complete end-capillary O₂ equilibration is expected to occur when breathing room air at rest (Wagner et al. 1986), it remains possible that some lung units might have a pulmonary capillary transit time <0.4 s, such that these lung units would be diffusion limited. However, this possible scenario is much less likely when breathing 40% O₂ considering the pulmonary capillary transit time in a diffusion-limited lung unit would need to be <0.1 s. Additionally, although alveolar N₂ is not eliminated when breathing 40% O₂, it is further reduced compared to breathing room air, which serves to minimize the potential contribution of / inequality. In fact, the venous admixture component of / inequality is nearly eliminated when reaches ∼250 mmHg (Forster, 1957). Although regional atelectasis resulting in / inequality may be a concern when breathing a fraction of inspired oxygen () >0.21, previous work in humans has shown breathing 40% O₂ to not be associated with an increase in / inequality (Dantzker et al. 1975). Lastly, preliminary and unpublished observations in our laboratory (Davis JT, Goodman RD, Futral JE, Elliott JE, Duke JW, Lovering AT.) suggest that breathing 40% O₂ does not alter blood flow through IPAVAs during exercise compared to breathing room air. Therefore, breathing 40% O₂ will satisfy the requirements of eliminating/minimizing the potential for diffusion limitation and/or / inequality to contribute to the A–aDO₂, while not altering blood flow through IPAVAs.

Taken together, the experimental protocol used in the current study to investigate the potential for blood flow through IPAVAs to provide a source of venous admixture and impair pulmonary gas exchange was designed to increase blood flow through IPAVAs in healthy humans at rest when breathing room air and 40% O₂. When breathing 40% O₂ the potential contributions from diffusion limitation and / inequality to the A–aDO₂ are prevented, and therefore, the contributions from postpulmonary and intrapulmonary shunt will be the only remaining factors capable of explaining an increase in the A–aDO₂ and calculated volume of venous admixture. To this end we quantified the A–aDO₂ in healthy humans at rest, breathing room air and 40% O₂ during the i.v. infusion of ADR. To facilitate increases in with the i.v. infusion of lower doses of ADR, this protocol was repeated following the administration of atropine. We hypothesized that the A–aDO₂ would increase concomitantly with an increase in blood flow through IPAVAs during the i.v. infusion of ADR when breathing room air, and the resulting increase in the A–aDO₂ and blood flow through IPAVAs would not be ameliorated during the i.v. infusion of ADR when breathing 40% O₂.

Methods

The University of Oregon Office for Protection of Human Subjects approved this project and all subjects provided verbal and written informed consent prior to participation. All studies were performed in accordance with the Declaration of Helsinki.

Subjects

Nine subjects (4 female) volunteered to participate in this study, all of whom were healthy, young, non-smoking and without a history of cardiopulmonary or respiratory disease.

According to the American Thoracic Society and European Respiratory Society standards (ATS/ERS) all subjects demonstrated normal pulmonary function, lung volumes and capacities, and diffusion capacity for carbon monoxide. Pulmonary function was assessed using computerized spirometry (MedGraphics Ultima CardiO₂, St Paul, MN, USA) and involved forced and slow vital capacity manoeuvres (Miller et al. 2005). Lung volumes and capacities were determined using whole body plethysmography (MedGraphics Elite Plethysmograph) (Wanger et al. 2005). Lung diffusion capacity for carbon monoxide (D_LCO) was determined by the single-breath, breath hold method (Knudson et al. 1987; Macintyre et al. 2005) using the Jones and Meade method for timing (Jones & Meade, 1961).

All subjects underwent a comprehensive echocardiographic screening procedure performed by a registered diagnostic cardiac sonographer in both adult and paediatric echocardiography with 25 years of experience (Philips iE33, The Netherlands). The purpose of this screening procedure was to confirm that subjects were without cardiac abnormalities, including a PFO. To this end, subjects were rigorously screened using transthoracic saline contrast echocardiography (TTSCE) as described previously (Lovering & Goodman, 2012; Elliott et al. 2013) and did not have a PFO.

Instrumentation

Following local anaesthesia (1% lidocaine (lignocaine), 2% by volume nitroglycerine (5 mg ml⁻¹) to minimize vasospasm) a 20 gauge × 1.75 inch radial artery catheter (Arrow International, Reading, PA, USA) was placed under aseptic conditions by a board-certified cardiologist. Patency of the arterial catheter was maintained using a pressurized flush system of normal saline. Intravenous catheters (i.v.) (18–22 gauge) were placed in the antecubital fossa of each arm, one for the injection of agitated saline contrast and one for the infusion of adrenaline and atropine. All catheters were connected to an extension set (7.0 inch, 0.3 ml volume) followed by a three-way stopcock. Measures of core body temperature were made using an oesophageal temperature probe (Mon-a-therm, Covidien LP, Mansfield, MA, USA), which was inserted via nasal intubation following the application of an anaesthetic gel (2 ml of 2% lidocaine gel) into the nasal cavity. In subjects who could not tolerate the oesophageal temperature probe placement (4/9 subjects), an ingestible core temperature pill (HQInc, Palmetto, FL, USA) was used. Unpublished observations from our laboratory demonstrate that temperatures obtained using the oesophageal temperature probe and the core temperature pill are in good agreement at rest (r² = 0.85). Lastly, a 12-lead ECG (MedGraphics, Mortara, CardiO₂) was placed with V4 repositioned ∼5 cm inferiorly as described previously (Lovering & Goodman, 2012).

Protocol

Subjects reported to the laboratory at 07.00 h and after familiarization with the study procedure began instrumentation. Thereafter, subjects were repositioned in a reclining chair in the left lateral decubitus position, and remained in this position for the duration of the protocol. The first set of measurements (Fig. 1A) were made breathing room air and 40% O₂ before and during the i.v. infusion of adrenaline (ADR). The second set of measurements (Fig. 1B) were made breathing room air and 40% O₂ following atropine (ATR) before and during the i.v. infusion of adrenaline. In both sets of measurements, before and after ATR, the order of breathing room air or 40% O₂ was done in a random and balanced fashion. However, measurements with ATR always occurred second due to the half-life of ATR being ∼120 min.

Figure 1. Schematic diagram of the protocol.

A, pre-vagal blockade breathing room air and 40% O₂. B, post-vagal blockade breathing room air and 40% O₂. The order of breathing room air or 40% O₂ was randomized.

Adrenaline was diluted in sterile saline to 4000 ng ml⁻¹ and, depending on the subject's body weight, was delivered at a constant rate for 5 min, using a continuous infusion syringe pump (Harvard Apparatus, Pump 22) as described previously (Laurie et al. 2012). In the first set of measurements, before ATR, ADR was infused at 320 ng kg⁻¹ min⁻¹ (320 ADR). In the second set of measurements, after ATR, ADR was infused at 80 ng kg⁻¹ min⁻¹ (ATR + 80 ADR). Atropine (2 mg) was diluted in 10 ml of sterile saline, and intravenously infused at a constant rate of 1 ml min⁻¹ for 10 min. The purpose of using ATR was to remove the parasympathetic involvement in the regulation of heart rate, and ultimately to achieve similar increases in , despite the two different doses of i.v. ADR (320 ng kg⁻¹ min⁻¹ vs. 80 ng kg⁻¹ min⁻¹). Atropine binds to and inhibits muscarinic acetylcholine receptors, and therefore allows the chronotropic effects of i.v. ADR to be unrestricted yet has no influence on the inotropic effects of i.v. ADR. Thus, without ATR (i.e. during 320 ADR) the parasympathetic system can keep heart rate down; however, with ATR (i.e. during ATR + 80 ADR) heart rate is allowed to increase further, despite the lower dose of i.v. ADR. The result is that is increased to a similar degree during 320 ADR (lower heart rate yet greater contractility) compared to ATR + 80 ADR (higher heart rate yet less contractility), despite the two different doses of i.v. ADR.

In both sets of measurements, before and after ATR, subjects were allowed ∼20 min for ADR to be metabolized and a final measurement without i.v. ADR was made. In this measurement, subjects reproduced the respiratory rate and tidal volume they had during the i.v. infusion of ADR. Respiratory rate was coached using a metronome and tidal volume was coached by verbal feedback while observing real time ventilatory data.

The first set of measurements (Fig. 1A) were made breathing room air and 40% O₂ at baseline (Pre ADR), during the i.v. infusion of ADR at 320 ng kg⁻¹ min⁻¹ (320 ADR) and subsequently while reproducing the ventilation () they had at 320 ADR ( @ 320 ADR). Following ATR, the second set of measurements (Fig. 1B) were made breathing room air and 40% O₂ at baseline (Pre ADR (ATR)), during the i.v. infusion of ADR at 80 ng kg⁻¹ min⁻¹ (ATR + 80 ADR) and subsequently while reproducing the ventilation they had at ATR + 80 ADR ( at ATR + 80 ADR). All nine subjects completed the first set of measurements; however, two of these subjects were unable to complete the entire second set of measurements due to clotted arterial catheters and therefore the n for the second set of measurements is 7.

Measurements

Respiratory variables

Breath-by-breath metabolic data was collected (MedGraphics, CardiO₂) and presented as the mid 5 of 7 (i.e. the moving average of five breaths excluding the low and high). In this way continuous measures of pulmonary , , minute ventilation (), and associated parameters, including end-tidal and mixed expired and values, were collected. Gases with known O₂ and CO₂ concentrations were used to calibrate the gas analyser before every test and change from breathing room air or 40% O₂.

Cardiac output and pulmonary artery systolic pressure

Cardiac output was measured using the open circuit acetylene wash-in method developed by Stout et al. (1975), modified by Gan et al. (1993), and subsequently validated in humans using direct Fick (Johnson et al. 2000). Subjects breathed through a non-rebreathing Hans Rudolph mouthpiece connected to an automatic three-way sliding valve allowing rapid and non-interruptive transitions (i.e. during subjects’ expiration) between two different inspiratory ports. When breathing room air, one inspiratory port was open to room air and the other was connected to a normoxic acetylene gas mixture (0.6% acetylene, 9.0% helium, 20.9% O₂, balance N₂). When breathing 40% O₂, one inspiratory port was connected to a hyperoxic gas mixture (40% O₂, balance N₂) and the other was connected to a hyperoxic acetylene gas mixture (0.6% acetylene, 9.0% helium, 40% O₂, balance N₂). During the wash-in phase, breath-by-breath acetylene and helium uptake was measured by a respiratory mass spectrometer (Marquette MGA 1100, MA Tech Services, Saint Louis, MO, USA) and tidal volume was measured by a pneumotachograph (Hans Rudolph, Shawnee, KS, USA) linearized by the method of Yeh et al. (1982) and calibrated using test gas before each study.

Pulmonary artery systolic pressure was determined by measuring the peak velocity (v) of the tricuspid regurgitation jet and estimating right atrial pressure (P_RA) based on the collapsibility of the inferior vena cava, and applying these to the modified Bernoulli equation; 4v² + P_RA = PASP (Yock & Popp, 1984; Currie et al. 1985; Himelman et al. 1989; Rudski et al. 2010).

Blood flow through intrapulmonary arteriovenous anastomoses

Transthoracic saline contrast echocardiography (TTSCE) was used to detect blood flow through IPAVAs as before (Laurie et al. 2012). Briefly, agitated saline contrast was produced by combining 3 ml room air with 1 ml of normal saline and agitating for ∼15 s prior to injecting. Each agitated saline contrast injection was visualized in the apical, four-chamber view and recorded at >30 frames per second for 20 cardiac cycles post-appearance of saline contrast in the right ventricle. The single frame within the 20 cardiac cycle recording with the greatest density and spatial distribution of contrast was qualitatively assessed using a previously published scoring system (Lovering et al. 2008b) similar to others (Barzilai et al. 1991); 0 = no bubbles, 1 = 1–3 bubbles, 2 = 4–12 bubbles, 3 = >12 bubbles appearing as a bolus, 4 = >12 bubbles heterogeneously filling the left ventricle, and 5 = >12 bubbles homogeneously filling the left ventricle.

Arterial blood gas analysis

Arterial blood (3 ml) was drawn anaerobically via the radial artery over 10–15 s, immediately preceding the measurement of cardiac output, into a heparinized syringe and immediately analysed in duplicate (and triplicate if necessary) for arterial (), arterial (), and pH. The blood-gas analyser (Siemens RAPIDLab 248, Munich, Germany) was calibrated daily with tonometered whole human blood by equilibrating three 6 ml samples of human blood, positioned within a heated block (37°C), with gases of known concentrations of O₂ and CO₂ for ∼120 min. The gases were chosen to span the expected range of and anticipated in subjects breathing room air and 40% O₂ at rest ( = 100, 210 and 245 mmHg; = 21, 35, 48 mmHg). Each sample was run in triplicate, and the values were used to create a predicted versus measured slope. A correction factor based on the inverse slope of this relationship was applied to measured values in this study. Arterial blood gases were also corrected for body temperature (Severinghaus, 1966; Kelman & Nunn, 1966; Dempsey & Wagner, 1999). Arterial O₂ saturation () and haemoglobin (Hb) were measured with CO-oximetry (Radiometer OSM3, Copenhagen, Denmark). Haematocrit was analysed using the microcapillary tube centrifugation method (M24 Centrifuge, LW Scientific, Lawrenceville, GA, USA). Blood lactate was analysed in duplicate using an electrochemical analyser (YSI 1500 Sport, Yellow Springs, OH, USA).

Calculations

Alveolar () was calculated using the ideal gas equation, using temperature/tonometry-corrected , and a respiratory quotient (RER) from a 15 s average of breath-by-breath metabolic data corresponding to the time and duration of the arterial blood draw:

where T_B is core body temperature for temperature correction of water vapour pressure, RER is the respiratory exchange ratio (/), and P_B is barometric pressure measured daily using a solid-state transducer barometer (7400 series Perception II, Davis Instruments, Vernon Hills, IL, USA). Pulmonary gas exchange efficiency (A–aDO₂) was then determined at rest and during i.v. infusion of ADR with and without ATR by subtracting the temperature/tonometry-corrected from the corresponding calculated .

Measures of O₂ content were calculated from the standard content equation:

using an O₂-carrying capacity of 1.39 ml O₂ (g Hb)^–1, and subjects directly measured Hb (g Hb dl^–1). For arterial O₂ content () represents CO-oximetry-measured arterial O₂ saturation () and temperature/tonometry-corrected . For end-capillary O₂ content () represents calculated end-capillary O₂ saturation () via the Kelman equation (Kelman, 1966) assuming end-capillary () was equal to . Mixed venous O₂ content was calculated using the Fick equation ( = × ( − )) and rearranging such that, = − (/).

Total venous admixture (/) required to account for the entirety of the A–aDO₂ was calculated from the shunt equation using the previously calculated O₂ contents:

Alveolar ventilation () was calculated using the directly measured and temperature-corrected :

Statistical analyses

Statistical analyses using GraphPad Prism (v. 5.0d) were done using a one-way analysis of variance comparing measured and calculated variables within an (room air or 40% O₂) without and with ATR. A Tukey's multiple comparison post hoc test was used to determine specific pairwise differences within variables. Non-parametric analyses (bubble scores) were done using a Friedman's test with a Dunn's multiple comparison post hoc test. All comparisons were determined a priori and significance was set to P < 0.05.

Results

Subject anthropometric and pulmonary function data are presented in Table 1. Cardiopulmonary and respiratory data when breathing room air are presented in Table 2, and when breathing 40% O₂ are presented in Table 3.

Table 1.

Anthropometric and pulmonary function data

	Absolute	%Predicted
Age (years)	24.6 ± 4.9	—
Height (cm)	172 ± 11	—
Weight (kg)	73.3 ± 11.2	—
FVC (l)	5.12 ± 1.09	(103.6 ± 5)
FEV1 (l)	4.19 ± 0.86	(101.5 ± 6.9)
FEV1/FVC (%)	82 ± 5	(97 ± 5)
FEF25–75	4.18 ± 1.2	(95.4 ± 6.4)
SVC (l)	5.43 ± 1.1	(111.4 ± 6.7)
TLC (l)	6.55 ± 1	(106 ± 7.3)
DLCO (ml min−1 mmHg−1)	36.56 ± 7.46	(122.5 ± 17.9)
DLCO/VA (ml min−1 mmHg−1)	5.69 ± 0.62	(117.1 ± 15.6)

Values are mean ± standard deviation. FVC, forced vital capacity; FEV₁, forced expiratory volume in 1 s; FEF_25–75, forced expiratory flow from 25 to 75% of FVC; SVC, slow vital capacity; TLC, total lung capacity; D_LCO, uncorrected diffusion capacity for carbon monoxide; D_LCO/V_A, D_LCO per alveolar volume.

Table 2.

Cardiopulmonary and respiratory data breathing room air, during the i.v. infusion of adrenaline with and without atropine

	Pre ADR	320 ADR	@ 320 ADR	Pre ADR (ATR)	ATR + 80 ADR	@ ATR + 80 ADR
(l min−1)	10.5 ± 2.2	16.4 ± 2.9*	15.5 ± 2.8*	11.4 ± 2.1	15.7 ± 3.2*	15.5 ± 2.9*
(l min−1)	5.9 ± 1.7	9.5 ± 3.1*	9.4 ± 2.6*	5.5 ± 1.3	8.4 ± 2.1*	8.0 ± 1.6*
RR (breath min−1)	14.3 ± 3.6	19.3 ± 2*	17.5 ± 4	18.2 ± 3.7	21.6 ± 3.4	20.6 ± 3.8
VT (ml)	764 ± 158	851 ± 141	937 ± 384	640 ± 112	730 ± 104	757 ± 117
VD/VT	0.44 ± 0.09	0.43 ± 0.08	0.39 ± 0.08	0.51 ± 0.07	0.47 ± 0.05	0.48 ± 0.05
(l min−1)	0.32 ± 0.08	0.35 ± 0.08	0.34 ± 0.06	0.30 ± 0.06	0.33 ± 0.09	0.31 ± 0.06
(l min−1)	0.28 ± 0.08	0.41 ± 0.10*	0.39 ± 0.08	0.27 ± 0.07	0.38 ± 0.09*	0.35 ± 0.07
RER	0.87 ± 0.04	1.21 ± 0.22*	1.15 ± 0.28*	0.89 ± 0.05	1.17 ± 0.1*	1.13 ± 0.14*
HR (beats min–1)	63 ± 6	80 ± 16*	73 ± 10	82 ± 15	97 ± 20	83 ± 13
(mmHg)	99 ± 4	107 ± 7	110 ± 8*	101 ± 4	107 ± 2*	110 ± 5*
(mmHg)	41 ± 4	38 ± 5	36 ± 4	41 ± 4	39 ± 3	37 ± 3
(mmHg)	99 ± 5	113 ± 7*	112 ± 8*	99 ± 4	111 ± 3*	112 ± 5*
(%)	96.9 ± 0.4	97.2 ± 0.4	97.4 ± 0.3	96.9 ± 0.3	97.2 ± 0.3	97.4 ± 0.2
(ml O2 dl−1)	17.46 ± 1.62	18.47 ± 1.69	17.51 ± 1.8	17.2 ± 1.77	18.07 ± 1.84	17.72 ± 1.93
pH	7.392 ± 0.017	7.402 ± 0.028	7.418 ± 0.023	7.387 ± 0.018	7.391 ± 0.01	7.405 ± 0.015
Hct (%)	39 ± 3	41 ± 3	40 ± 4	39 ± 5	40 ± 4	41 ± 3
Hb (g dl−1)	13.2 ± 1.3	13.9 ± 1.3	13.2 ± 1.4	13.0 ± 1.4	13.6 ± 1.4	13.3 ± 1.5
Core temp. (°C)	36.8 ± 0.3	36.7 ± 0.3	36.8 ± 0.3	36.9 ± 0.3	36.7 ± 0.3	36.7 ± 0.3
Lactate (mmol l−1)	0.94 ± 0.32	1.59 ± 0.38*	1.28 ± 0.34	1.12 ± 0.3	1.35 ± 0.32	1.42 ± 0.3

Values are mean ± standard deviation. _E, minute ventilation; _A, alveolar ventilation; RR, respiratory rate; V_T, tidal volume; V_D/V_T, dead space to tidal volume ratio; , volume of O₂ consumed; , volume of CO₂ produced; RER, respiratory exchange ratio; HR, heart rate; , arterial ; , arterial ; , alveolar ; , arterial oxygen saturation; , arterial oxygen content; Hct, haematocrit; Hb, haemoglobin. *P < 0.05, Pre ADR compared to 320 ADR and _A @ 320 ADR; *P < 0.05, Pre ADR (ATR) compared to ATR + 80 ADR and _A @ ATR + 80 ADR.

Table 3.

Cardiopulmonary and respiratory data breathing 40% O₂, during the i.v. infusion of ADR with and without atropine

	Pre ADR	320 ADR	@ 320 ADR	Pre ADR (ATR)	ATR + 80 ADR	@ ATR + 80 ADR
(l min−1)	10.4 ± 2.5	15.6 ± 4.7*	14.6 ± 4.2	11.3 ± 2.1	15.0 ± 3.0	14.9 ± 4.5
(l min−1)	5.3 ± 1.3	10.9 ± 9.0	9.9 ± 8.0	5.4 ± 1.2	8.8 ± 2.5	7.7 ± 3.1
RR (breath min−1)	15.9 ± 3.8	19.2 ± 5.5	18.2 ± 5.0	18.2 ± 2.8	21.6 ± 4.4	20.3 ± 4.6
VT (ml)	656 ± 59	830 ± 221	840 ± 294	620 ± 62	702 ± 78	740 ± 177
VD/VT	0.48 ± 0.05	0.45 ± 0.05	0.46 ± 0.08	0.53 ± 0.06	0.43 ± 0.06	0.49 ± 0.08
(l min−1)	0.30 ± 0.08	0.33 ± 0.08	0.31 ± 0.09	0.28 ± 0.07	0.31 ± 0.07	0.27 ± 0.09
(l min−1)	0.26 ± 0.06	0.39 ± 0.12*	0.35 ± 0.10	0.26 ± 0.06	0.37 ± 0.09	0.34 ± 0.14
RER	0.86 ± 0.1	1.16 ± 0.21*	1.05 ± 0.14*	0.93 ± 0.12	1.24 ± 0.13*	1.25 ± 0.19*
HR (beats min–1)	58.2 ± 7.6	81.3 ± 22.2*	72.1 ± 11.3	83.2 ± 23.6	92.2 ± 24.2	81 ± 18.2
(mmHg)	223 ± 12	222 ± 8	232 ± 11	220 ± 9	225 ± 4	230 ± 11
(mmHg)	42 ± 5	37 ± 9	36 ± 9	42 ± 4	38 ± 4	39 ± 3
(mmHg)	235 ± 10	245 ± 8	245 ± 12	230 ± 6	243 ± 8*	241 ± 9*
(%)	98.0 ± 0.3	97.9 ± 0.3	98.0 ± 0.2	98.0 ± 0.2	98.0 ± 0.2	98.0 ± 0.2
(ml O2 dl−1)	18.1 ± 1.84	18.92 ± 1.77	18.21 ± 1.7	18.02 ± 1.95	18.59 ± 2.05	18.07 ± 1.94
pH	7.388 ± 0.017	7.400 ± 0.026	7.406 ± 0.04	7.386 ± 0.015	7.384 ± 0.014	7.407 ± 0.018†
Hct (%)	40 ± 4	41 ± 4	40 ± 4	40 ± 3	41 ± 4	39 ± 3
Hb (g dl−1)	13.3 ± 1.4	13.9 ± 1.4	13.3 ± 1.3	13.2 ± 1.5	13.7 ± 1.6	13.2 ± 1.5
Core temp. (°C)	36.7 ± 0.2	36.8 ± 0.3	36.8 ± 0.2	36.8 ± 0.2	36.7 ± 0.3	36.8 ± 0.2
Lactate (mmol l−1)	0.83 ± 0.19	1.54 ± 0.4*	1.42 ± 0.32†	1.03 ± 0.32	1.28 ± 0.21	1.28 ± 0.29

Values are mean ± standard deviation. See Table 2 for definitions of abbreviations and symbols.

Transpulmonary passage of saline contrast and pulmonary gas exchange efficiency

When breathing room air, from Pre ADR to 320 ADR, the transpulmonary passage of saline contrast increased (Fig. 2A) concomitant with an increase in the A–aDO₂ (Fig. 2B) that corresponded with an increase in calculated / (Fig. 2C). Similarly, when breathing 40% O₂, from Pre ADR to 320 ADR, the transpulmonary passage of saline contrast increased (Fig. 2D) concomitant with an increase in the A–aDO₂ (Fig. 2E) that corresponded with an increase in calculated / (Fig. 2F). As expected, when breathing 40% O₂ the magnitude of the increase in A–aDO₂ was greater compared to breathing room air; however, importantly, calculated / increased to the same extent and averaged 2 ± 1%.

Figure 2. Bubble score, pulmonary gas exchange efficiency, and venous admixture.

The transpulmonary passage of saline contrast (bubble scores, A and B), where each subject is identified by a unique and consistent symbol shape/colour, pulmonary gas exchange efficiency (A–aDO₂, C and D), and calculated venous admixture (/, E and F) before (non-hatched bars) and after (hatched bars) vagal blockade breathing room air (open bars) and 40% O₂ (shaded bars). *: Pre ADR compared to 320 ADR and @ 320 ADR; Pre ADR (ATR) compared to ATR + 80 ADR and @ ATR + 80 ADR. †: 320 ADR compared to @ 320 ADR; ATR + 80 ADR compared to @ ATR + 80 ADR.

When breathing room air, from Pre ADR (ATR) to ATR + 80 ADR, the transpulmonary passage of saline contrast increased (Fig. 2A) concomitant with an increase in the A–aDO₂ (Fig. 2B) that corresponded with an increase in calculated / (Fig. 2C). When breathing 40% O₂, from Pre ADR (ATR) to ATR + 80 ADR, the transpulmonary passage of saline contrast increased (Fig. 2D) concomitant with an increase in the A–aDO₂ (Fig. 2E) that corresponded with an increase in calculated / (Fig. 2F). Similarly, as expected when breathing 40% O₂, the magnitude of the increase in A–aDO₂ was greater compared to breathing room air. Importantly, calculated / increased to the same extent and averaged 2 ± 1%. Therefore, a similar impairment in pulmonary gas exchange efficiency and an increase in calculated / occurred with high and low dose ADR with and without ATR.

The respiratory rate and tidal volume, and therefore , that subjects had during 320 ADR was reproduced for the @ 320 ADR measurement. Accordingly, per experimental design there was no difference in or dead space ventilation (V_D/V_T) between 320 ADR and @ 320 ADR when breathing room air and 40% O₂ (Tables 2 and 3). Prior to the @ 320 ADR measurement, ADR from the preceding infusion (i.e. 320 ADR) was allowed time to be metabolized and as such the transpulmonary passage of saline contrast was reduced compared to Pre ADR values at this time point when breathing room air (Fig. 2A) and 40% O₂ (Fig. 2D). Similarly, at @ 320 ADR the A–aDO₂ was reduced to Pre ADR values when breathing room air (Fig. 2B) and 40% O₂ (Fig. 2E), which also corresponded to the calculated / being reduced to Pre ADR values when breathing room air (Fig. 2C) and 40% O₂ (Fig. 2F).

When breathing room air, during @ ATR + 80 ADR the transpulmonary passage of saline contrast was reduced to Pre ADR (ATR) values (Fig. 2A), corresponding to a reduction in calculated / to Pre ADR (ATR) values (Fig. 2C). When breathing 40% O₂, from ATR + 80 ADR to @ ATR + 80 ADR, the transpulmonary passage of saline contrast was reduced to Pre ADR (ATR) values (Fig. 2D) corresponding to a reduction in the A–aDO₂ (Fig. 2E) and calculated / (Fig. 2F) to Pre ADR (ATR) values.

Cardiac output and pulmonary artery systolic pressure

When breathing room air and 40% O₂, increased during 320 ADR and ATR + 80 ADR; however, there were no differences between 320 ADR and ATR + 80 ADR (Fig. 3A and B). Pulmonary artery systolic pressure (PASP) increased during 320 ADR (Fig. 3C and D) when breathing room air and 40% O₂. However, PASP was not significantly increased during ATR + 80 ADR when breathing room air and 40% O₂ (Fig. 3C and D).

Figure 3. Cardiac output and pulmonary artery systolic pressure.

Cardiac output (A and B) and pulmonary artery systolic pressure (PASP, C and D) before (non-hatched) and after (hatched) atropine breathing room air (open symbols) and 40% O₂ (shaded symbols). The whiskers reflect the minimum and maximum value, while the box is constructed of the median of the first and third quartile. The line in the box represents the median of the data set and the plus sign represents the mean of the data set. *: Pre ADR compared to 320 ADR and @ 320 ADR; Pre ADR (ATR) compared to ATR + 80 ADR and @ ATR + 80 ADR. †: 320 ADR compared to @ 320 ADR; ATR + 80 ADR compared to @ ATR + 80 ADR.

Discussion

The main finding of this study was that the i.v. infusion of ADR in healthy humans resulted in an increase in the transpulmonary passage of saline contrast and a concomitant increase in the A–aDO₂ and calculated /, when breathing room air and 40% O₂. Neither the transpulmonary passage of saline contrast, nor the calculated /, differed during the i.v. infusion of ADR when breathing room air and 40% O₂. The increase in A–aDO₂ observed during the i.v. infusion of ADR before or after ATR was not due to the associated increase in ventilation. Furthermore, these findings were present during the i.v. infusion of ADR at 320 ng kg⁻¹ min⁻¹ before ATR, and during the i.v. infusion of ADR at 80 ng kg⁻¹ min⁻¹ after ATR. Lastly, these data support the idea that blood flow through IPAVAs is mediated primarily by increases in in subjects breathing room air and 40% O₂.

Effect of breathing 40% O₂ on diffusion limitation and / inequality

Compared to breathing room air, breathing 40% O₂ increased from ∼100 mmHg to ∼235 mmHg, corresponding to an increase in calculated mixed venous () of ∼40 mmHg when breathing room air, to ∼45 mmHg when breathing 40% O₂. Accordingly, the partial pressure gradient of O₂ across the alveolar–capillary interface was ∼60 mmHg when breathing room air, which increased >3-fold to ∼190 mmHg when breathing 40% O₂. As detailed in the Introduction, the result of this increased driving gradient for the diffusion of O₂ across the alveolar–capillary interface is a markedly increased rate of end-capillary O₂ equilibration, such that when breathing 40% O₂, the potential contribution from diffusion limitation to the A–aDO₂ is minimized compared to breathing room air. Additionally, the potential contribution from / inequality to the A–aDO₂ is also minimized when breathing 40% O₂ compared to breathing room air. This is based on the same principles as those that define the 100% O₂ technique. With the removal of alveolar N₂, as occurs when breathing 100% O₂, the in all lung units must increase to that of the inspired less water vapour pressure and . There is then, very little discrepancy in between lung units, regardless of their respective / ratios, and how poorly ventilated they may be. Following this same logic, although alveolar N₂ is not eliminated when breathing 40% O₂, it is further reduced compared to breathing room air and the venous admixture component of / inequality is nearly eliminated when reaches ∼250 mmHg (Forster, 1957).

Effect of breathing 40% O₂ on blood flow through IPAVAs

The effect of breathing 40% O₂ on blood flow through IPAVAs has not been directly investigated and more extensive work in this area would be of interest. However, it is known that breathing 100% O₂ during exercise (Lovering et al. 2008b; Elliott et al. 2011) and during the i.v. infusion of catecholamines (Bryan et al. 2012; Laurie et al. 2012) significantly reduces, or eliminates, the transpulmonary passage of saline contrast. This has been demonstrated to not be a product of O₂ breathing altering the in vivo dynamics of saline contrast (Elliott et al. 2011) and is also in agreement with work using solid microspheres to detect a reduction in blood flow through IPAVAs in anaesthetized dogs ventilated with 100% O₂ (Niden & Aviado, 1956). Conversely, bubble scores during 320 ADR and ATR + 80 ADR were not different between breathing room air and 40% O₂ (Fig. 2A and B), yet bubble scores have been shown to be reduced or eliminated during 320 ADR when breathing 100% O₂ (Laurie et al. 2012). Thus, our experimental design allowed for blood flow through IPAVAs while preventing contributions from / inequality and diffusion limitation.

Pulmonary gas exchange efficiency during the i.v. infusion of adrenaline

When breathing room air the A–aDO₂ increased by 6 mmHg at 320 ADR compared to Pre ADR, which corresponded to an increase in calculated / of 2%. This 2% increase in calculated / necessary to explain the measured A–aDO₂ could in theory include contributions from diffusion limitation, / inequality, and right-to-left shunt. When breathing 40% O₂ the A–aDO₂ increased by 12 mmHg at 320 ADR compared to Pre ADR, which corresponded to a similar increase in calculated / of 2%. However, the 2% increase in calculated / when breathing 40% O₂ is only explained by an increase in right-to-left shunt. Previous work in dogs has also demonstrated an increase in intrapulmonary shunt during the i.v. infusion of ADR using microspheres (Nomoto et al. 1974) to detect blood flow through IPAVAs and by arterial blood gas analysis to determine the impact on pulmonary gas exchange efficiency. This finding was demonstrated to be eliminated in dogs ventilated with 100% O₂ (Berk et al. 1973) demonstrating that 100% O₂ prevented blood flow through IPAVAs as we have suggested occurs in humans (Lovering et al. 2008b; Elliott et al. 2011).

Cardiac output was similarly increased by ∼2-fold above Pre ADR values in the 320 ADR and ATR + 80 ADR conditions. Assuming no change in pulmonary capillary blood volume, this should correspond to a decrease in mean pulmonary capillary transit time from ∼0.7 s to ∼0.35 s. As previously discussed, when breathing room air and 40% O₂ complete end-capillary O₂ equilibration is expected to occur within ∼0.4 s and ∼0.1 s, respectively. Therefore, if pulmonary capillary blood volume remained constant, some degree of diffusion limitation would be possible when breathing room air, although breathing 40% O₂ should sufficiently minimize this potential. That said, it seems unlikely that pulmonary capillary blood volume would not increase to some degree in response to the ∼2-fold increase in pulmonary blood flow and any increase in pulmonary capillary blood volume would mitigate reductions in mean pulmonary capillary transit time. Furthermore, as expected based on the Fick principle of mass balance, calculated increased secondarily to the ∼2-fold increase in and constant whole body . Specifically, calculated would have increased from ∼40 mmHg at Pre ADR to ∼50 mmHg at 320 ADR when breathing room air, and from 45 mmHg at Pre ADR to ∼55 mmHg at 320 ADR when breathing 40% O₂. In both cases, the ∼10 mmHg increase in would help to accelerate end-capillary O₂ equilibration (Wagner, 1982).

The potential that / inequality was increased during the i.v. infusion of ADR at 320 ADR or at ATR + 80 ADR cannot be excluded, although as discussed it would probably be minimized when breathing 40% O₂ compared to room air. Previous work investigating the effect of inotropic drugs on / inequality in healthy humans is lacking. However, in four patients admitted to the intensive care unit for acute respiratory failure from pulmonary embolism, Manier & Castaing (1992) showed a non-significant decrease in mean / for blood flow distribution (0.67 to 0.57) and a non-significant decrease in log standard deviation of perfusion (logSD) (1.19 to 1.11) during the i.v. infusion of dobutamine at 10 μg kg⁻¹ min⁻¹. In two of these patients, shunt (i.e. / = 0) increased. One of these patients was also assessed during the i.v. infusion of dobutamine when breathing 40% O₂ and in this patient shunt was unchanged compared to breathing room air. In another study, Rennotte et al. (1989) studied 10 patients admitted to the intensive care unit for a variety of reasons and found a significant increase in logSD_Q during the i.v. infusion of dopamine, but not during the i.v. infusion of dobutamine (both drugs at 5 μg kg⁻¹ min⁻¹). Work in dogs with pulmonary oedema secondary to oleic acid infusion has demonstrated that an increase in via the i.v. infusion of isoproterenol (isoprenaline) (Light et al. 1988) and dopamine (Lynch et al. 1979) resulted in a significant increase in shunt. Of note, in the work by Lynch et al. the changes in shunt occurred whether was pharmacologically or mechanically altered. Therefore, as concluded by these authors, the increase in shunt from pharmacologically increased is probably not due to vasoconstrictor or vasodilator activity of these drugs as the increase in shunt is observed when is increased via mechanical means as well. The current study in healthy humans lends support to this conclusion by demonstrating the same increase in shunt in response to the same increase in during 320 ADR and ATR + 80 ADR.

If it is accepted that diffusion limitation and / inequality are probably not significantly contributing to the measured A–aDO₂ in healthy humans breathing room air and 40% O₂ during 320 ADR and ATR + 80 ADR, one possibility other than an increase in intrapulmonary shunt remains; an increase in postpulmonary shunt. The chronotropic and inotropic effects of ADR, and chronotropic effects of ATR may increase Thebesian blood flow, and similarly the increase in ventilation may increase bronchial blood flow. In an attempt to control for the potential confounding effects of increased ventilation, measurements were made in iso-alveolar ventilation (same tidal volume and respiratory rate) conditions after ADR had been allowed time (∼20 min) to be metabolized (i.e. @ 320 ADR and @ ATR + 80 ADR). Accordingly, the increase in calculated / during @ 320 ADR and @ ATR + 80 ADR may serve as a measure of predominantly bronchial shunt. Therefore, subtracting the increase in / during @ 320 ADR and @ ATR + 80 ADR, from 320 ADR and ATR + 80 ADR, respectively, would yield the increase in total right-to-left shunt less the contribution from bronchial blood flow. Doing so reveals an ∼1% increase in / during 320 ADR breathing room air and 40% O₂, and an ∼1% increase in / during ATR + 80 ADR breathing room air and 40% O₂ (Table 4), which could be due to either Thebesian and/or intrapulmonary shunt.

Table 4.

Difference in / with i.v. ADR and during conditions of iso-alveolar ventilation without i.v. ADR

		/		/
Room air	320 ADR	2%	ATR + 80 ADR	2%
	@ 320 ADR	1%	@ ATR + 80 ADR	1%
	Difference	1%	Difference	1%
40% O2	320 ADR	2%	ATR + 80 ADR	2%
	@ 320 ADR	1%	@ ATR + 80 ADR	1%
	Difference	1%	Difference	1%
	Mean	1%	Mean	1%

Considering the unique experimental conditions used in the present study, it is difficult to extrapolate these findings to their potential physiological relevance during other conditions known to increase blood flow through IPAVAs, such as exercise. Nevertheless, it is worthwhile to acknowledge that previous work using the multiple inert gas elimination technique (MIGET) does not report significant intrapulmonary shunt in healthy humans during exercise breathing room air, a condition that does produce significant increases in the transpulmonary passage of saline contrast. Pooling data from several studies using the MIGET (Podolsky et al. 1996; Hopkins et al. 1998; Rice et al. 1999; Olfert et al. 2004; Jonk et al. 2007), intrapulmonary shunt averages 0.2 ± 0.7% of the . Accordingly, these data from studies using the MIGET suggest that intrapulmonary shunt contributes minimally to the increased A–aDO₂ during exercise breathing room air. However, the data from the current study provide two potentially important pieces of information. First, they demonstrate an ∼1% increase in / that is due to right-to-left shunt from blood flow through IPAVAs and/or Thebesian blood flow. Second they demonstrate a combined volume of right-to-left shunt (including blood flow through IPAVAs, Thebesian and bronchial blood flow) of ∼2%. Importantly, a 1–2% increase in right-to-left shunt during exercise could explain between 1/3 and 2/3 of the increase in A–aDO₂ measured during exercise. This would suggest a larger contribution of right-to-left shunt on exercise-induced pulmonary gas exchange impairment than has been previously appreciated, and a physiologically relevant contribution of blood flow through IPAVAs.

Cardiac output and pulmonary artery systolic pressure

Cardiac output and PASP have previously been hypothesized as potential mechanisms responsible for regulating blood flow through IPAVAs (Stickland et al. 2004; La Gerche et al. 2010; Bryan et al. 2012; Laurie et al. 2012; Norris et al. 2014). The primary purpose of the ATR + 80 ADR measurement was to facilitate increasing to a similar extent to that at 320 ADR, with a lower dose of ADR. However, it has also been shown that the chronotropic effects of ATR are associated with a minimal change in mean pulmonary artery pressure measured using right heart catheterization in patients with asthma and emphysema at rest (Williams et al. 1960) and in a different group of patients (Abbott et al. 1955). In support of these observations, albeit in healthy subjects, PASP in the current study was unchanged at Pre ADR (ATR) compared to Pre ADR, despite increasing by ∼2 l min⁻¹ when breathing room air and 40% O₂ (Fig. 3). Pulmonary artery systolic pressure increased significantly by ∼20 mmHg during 320 ADR, but non-significantly by ∼10 mmHg during ATR + 80 ADR when breathing room air and 40% O₂ (Fig. 3). However, was ∼12.5 l min⁻¹ (i.e. not different) during 320 ADR and ATR + 80 ADR when breathing room air and 40% O₂ (Fig. 3). Therefore, 320 ADR and ATR + 80 ADR produced similar increases in , yet PASP was less during ATR + 80 ADR and there was no difference in bubble scores during 320 ADR and ATR + 80 ADR (Fig. 2). Although bubble scores do not represent a means to quantify blood flow through IPAVAs and are a semi-quantitative, non-distributed variable, the calculated / was ∼2% during both 320 ADR and ATR + 80 ADR when breathing room air and 40% O₂. It is important to note that PASP during ATR + 80 ADR was not statistically significantly greater than Pre ADR (ATR) despite relatively low variability in the data. This non-significant finding was not due to insufficient power as we attained a power, calculated post hoc, of 0.92 for room air and 0.91 for 40% O₂. This power is more than sufficient (0.80 is considered ideal) to detect a significant effect on PASP should one have existed. Nevertheless, future work in this area will be helpful in clarifying the roles of pressure and flow under other experimental conditions such as hypoxia and exercise.

Limitations

Saline contrast echocardiography

The detection of blood flow through IPAVAs using TTSCE is based on the rationale that the pulmonary circulation functions as a biological sieve, filtering out bubbles that are larger than the diameter of a pulmonary capillary. Although the size of the injected bubbles remains unknown, it is estimated that, based on the minimum size necessary for bubbles to be stable enough to survive the pulmonary transit time, their initial diameter would need to be >60 μm (Eldridge et al. 2004). To date, there is no direct evidence that either pulmonary capillaries or corner vessels can distend to diameters >20 μm, even under supra-physiological pulmonary pressures for humans. For example, at pulmonary pressures up to 73 mmHg in isolated greyhound lungs the mean capillary diameter was found to be 6.5 μm and the maximum capillary diameter measured was 13 μm (Glazier et al. 1969; Rosenzweig et al. 1970). Thus, the transpulmonary passage of bubbles larger than the diameter of a pulmonary capillary indicates that these bubbles have escaped pulmonary capillary filtration by traversing the pulmonary circulation via large-diameter IPAVAs. However, this technique does not quantify the volume of blood flow through IPAVAs; therefore, the bubble-scoring system, which is not without precedent (Barzilai et al. 1991), offers only a qualitative assessment of the degree and spatial distribution of left-sided contrast.

In subjects without a PFO, the alternative explanation for observing left-sided saline contrast is the possibility that bubbles are, by some means, travelling through pulmonary capillaries and/or corner vessels. In this way, left-sided contrast would not reflect blood flow through IPAVAs. However, this possibility should be considered with several points in mind. First, bubbles small enough to traverse pulmonary capillaries (<10 μm) have an estimated survival time of 180–550 ms (Yang et al. 1971a,b; Meltzer et al. 1980) and in healthy humans at rest, the mean pulmonary transit time is ∼9 s, that decreases to ∼3 s at maximal exercise with in excess of 30 l min⁻¹ (Hopkins et al. 1996). Second, Roelandt (1982) demonstrated that an injection pressure of 300 mmHg through a firmly wedged pulmonary artery catheter was needed to observe left-sided contrast, by forcing bubbles to deform and squeeze through pulmonary capillaries. Third, pulmonary capillary disruption, not distension has been predicted to be the more likely result of pulmonary capillary pressures sufficient to distend pulmonary capillaries to >20 μm in diameter (West, 2000).

Although we cannot definitely rule out the possibility that microbubbles are, by some means, traversing pulmonary capillaries, the transpulmonary passage of 25 and 50 μm microspheres has been demonstrated in isolated human lungs ventilated and perfused with physiological pressures (Lovering et al. 2007). In addition work in animals during exercise (Stickland et al. 2007) and breathing hypoxic gas at rest (Bates et al. 2012) demonstrate the transpulmonary passage of >25 μm and 70 μm microspheres, respectively. Importantly, the transpulmonary passage of non-deformable, large-diameter microspheres can only be explained by blood flow through IPAVAs, and for these aforementioned studies using microspheres, studies using saline contrast microbubbles demonstrate the same findings (Lovering et al. 2008a; Laurie et al. 2010; Elliott et al. 2011).

Subjects in the current study were positioned in a reclining chair, at 45 deg in the left lateral decubitus position. Originally reported by Stickland et al. (2004) and subsequently by Elliott et al. (2013) body positioning appears to have an effect on the detection of left-sided contrast in healthy humans at rest breathing room air. Specifically, the majority of subjects who demonstrate the transpulmonary passage of saline contrast when supine at rest breathing room air, do not do so when upright at rest breathing room air. The reason for this difference in blood flow through IPAVAs based on body positioning is unknown, although subjects who demonstrate left-sided contrast when supine most commonly have a bubble score of 1 (i.e. 1–3 bubbles), which is probably a trivial degree of blood flow through IPAVAs. Lastly, it is unknown if this difference in blood flow through IPAVAs based on body position persists during the i.v. infusion of adrenaline; however, all previous work investigating blood flow through IPAVAs in subjects at rest has been done in subjects reclined in the left lateral decubitus position and therefore we felt it important to use these same physiological conditions in the current study.

Accurately measuring arterial

Accurate measurements of are critically important to the interpretation of these data, and several precautions were taken to ensure very high standards. When breathing room air at Pre ADR was ∼100 mmHg (Eugene, OR, USA; 130 m above sea level), which increased to ∼220 mmHg when breathing 40% O₂. During 320 ADR and ATR + 80 ADR when was doubled, increased to ∼110 mmHg and ∼230 mmHg when breathing room air and 40% O₂, respectively. Accordingly, this >2-fold increase in when breathing 40% O₂ amplifies several potential sources of error compared to when breathing room air, all of which can only decrease the measured , and therefore, increase the measured A–aDO₂. Perhaps the most significant source of error is due to air bubbles within the sampling syringe, which are composed of room air, and thus contain a of ∼70–80 mmHg lower than when breathing 40% O₂. This was minimized by withdrawing all arterial blood samples in such a way as to keep any small air bubble at the top of the sample, which was immediately expelled, and by maintaining a positive pressure on the sample syringe during the analysis to pre-emptively prevent the production of an air bubble.

Additional potential issues are (1) the possible diffusion of O₂ from the sample through the syringe, (2) the consumption of O₂ by white blood cells, and (3) the consumption of O₂ by the electrode during the analysis. The impact of the first two concerns increases over time, and therefore, we minimized this by immediately analysing each sample. All of these concerns can also, at least partially, be corrected for by calibrating the blood gas analyser on each study day using tonometered human blood, as before (Lovering et al. 2013). This practice is similar, but superior, to using standard quality control methods in that it allows for the correction of day-to-day fluctuations and inherent errors in measuring the exact and , which include the previously mentioned concerns. Consistent inherent errors with handling and analysing blood with in the 220–230 mmHg range are also present during tonometry, and therefore this correction factor will help offset these. After correcting and for core body temperature, the tonometry correction resulted in an average increase in of <1 mmHg and <2 mmHg when breathing room air and 40% O₂, respectively. The minimal change in , particularly when breathing 40% O₂ is reflective of our careful handing and immediate analysis of all arterial blood samples.

In the current study there were several negative A–aDO₂ values in the Pre ADR and Pre ADR (ATR) measurement breathing room air. Negative A–aDO₂ values are physiologically impossible, and as detailed by Jonk et al. (2007) are probably the result of the cumulative effects of very small (∼1%) errors in variables used to calculate (e.g. barometric pressure, body temperature, , , and ). One possible explanation for the cause of these errors is that subjects are not in true resting conditions, and anticipating exercise (in other studies) or anticipating the i.v. infusion of adrenaline (in the current study). It is worthwhile noting that the magnitude of these negative values was between 0 and −3.7 mmHg, and therefore was probably a result of measuring a physiological parameter that is approximately 0, using several instruments that each have a certain inherent margin of error.

Ovarian hormones

Female subjects in the current study were not monitored to determine which phase of the menstrual cycle they were in. It is known that resting ventilation is increased in female subjects in the mid-luteal phase of the menstrual cycle when oestrogen and progesterone are highest relative to the early/late follicular phase (MacNutt et al. 2012). However, no difference was observed in ventilation between male and female subjects, as well as in variables relating to pulmonary gas exchange efficiency. Accordingly, the potential confounding effects of ovarian hormones do not appear to be an issue in the current study. Furthermore, previous work has shown no difference in the transpulmonary passage of saline contrast (Kennedy et al. 2012) between male and female subjects. Whether or not pulmonary gas exchange efficiency during exercise differs between male and female subjects remains a controversial area (Dominelli et al. 2013); however, work by Olfert et al. (2004) suggests no difference in exercise-induced pulmonary gas exchange impairment between male and female subjects.

Summary

The current study investigated the potential that blood flow through IPAVAs provides a source of venous admixture that can contribute relevantly to pulmonary gas exchange efficiency. These data demonstrate that during the i.v. infusion of adrenaline (i.e. during conditions of elevated pulmonary blood flow) the transpulmonary passage of saline contrast is increased concomitant with an increase in calculated / and A–aDO₂. Furthermore, this impairment in pulmonary gas exchange efficiency was present to the same extent, with / averaging ∼2% during 320 ADR and ATR + 80 ADR, not only when breathing room air, but also when breathing 40% O₂ when contributions from / inequality and diffusion limitation were minimized. When the potential contribution from predominantly bronchial shunt is subtracted from this total, the average / is again similar between breathing room air and 40% O₂ and averages 1% which may be the result of Thebesian shunt and/or blood flow through IPAVAs. The current paradigm describing mechanisms of impaired pulmonary gas exchange efficiency does not include blood flow through IPAVAs, particularly during exercise. However, the transpulmonary passage of saline contrast is also observed during exercise and thus, blood flow through IPAVAs during exercise may also provide a source of venous admixture.

Glossary

A–aDO₂

alveolar-to-arterial difference in

ADR

adrenaline (epinephrine)

ATR

atropine

arterial oxygen content

mixed venous oxygen content

fraction of inspired oxygen

Hb

haemoglobin

Hct

haematocrit

IPAVAs

intrapulmonary arteriovenous anastomoses

alveolar partial pressure of oxygen

arterial partial pressure of oxygen

arterial partial pressure of carbon dioxide

PASP

pulmonary artery systolic pressure

PFO

patent foramen ovale

mixed venous partial pressure of oxygen

cardiac output

venous admixture

/

percentage of the cardiac output that is venous admixture

arterial oxygen saturation

mixed venous oxygen saturation

TTSCE

transthoracic saline contrast echocardiography

alveolar ventilation

/>

alveolar ventilation-to-perfusion inequality

volume of carbon dioxide produced

volume of oxygen consumed

minute ventilation

Additional information

Competing interests

The authors have no conflicts of interest to declare.

Author contributions

All experiments were conducted in the Cardiopulmonary and Respiratory Physiology Lab at the University of Oregon. J.E.E., J.W.D., J.R.H. and A.T.L. contributed to the conception and design, or analysis and interpretation of data. J.E.E., J.W.D., J.A.H., J.R.H. and A.T.L. contributed to drafting the article or revising it critically for important intellectual content and approved the final version to be published.

Funding

This research was supported by the Department of Defense (W81XWH-10-2-0114). J. E. Elliott was supported by an American Heart Association pre-doctoral fellowship and the Eugene & Clarissa Evonuk Memorial Graduate Fellowship in Environmental, Cardiovascular, or Stress Physiology.