7/29/2019 Prevalence of left heart contrast in healthy, young, asymptomatic humans at rest breathing room air

http://slidepdf.com/reader/full/prevalence-of-left-heart-contrast-in-healthy-young-asymptomatic-humans-at 1/8

Respiratory Physiology & Neurobiology 188 (2013) 71–78

Contents lists available at SciVerse ScienceDirect

Respiratory Physiology & Neurobiology

 journa l homepage: www.elsevier .com/ locate / resphysiol

Prevalence of left heart contrast in healthy, young, asymptomatic humans at restbreathing room air

 Jonathan E. Elliott a, S. Milind Nigam a, Steven S. Laurie a, Kara M. Beasley a, Randall D. Goodmanb, Jerold A. Hawn a,b, Igor M. Gladstonea,c, Mark S. Chesnuttd, Andrew T. Loveringa,∗

a University of Oreg on,Department of Human Physiology, Eugene, OR,USAb OregonHeart andVascular Institute, RiverBend, Springfield, OR, USAc Department o f Pediatr ics, OregonHealth& Science University,Portland, OR,USAd Dotter Interventional Institute’sHereditary Hemorrhagic Telangiectasia Centerof Excellence,Oregon Health & Science University, Portland, OR, USA

a r t i c l e i n f o

 Article history:

Accepted 23 April 2013

Keywords:

Transthoracic saline contrast

echocardiography

Patent foramen ovale

Right-to-left shunt

Intrapulmonary arteriovenous anastomoses

a b s t r a c t

Our purpose was to report the prevalence of healthy, young, asymptomatic humans who demonstrate left

heart contrast at rest, breathing room air. We evaluated 176 subjects (18–41 years old) using transtho-

racic saline contrast echocardiography. Left heart contrast appearing ≤3 cardiac cycles, consistent with

a patent foramen ovale (PFO), was detected in 67 (38%) subjects. Left heart contrast appearing >3 car-

diac cycles, consistent with the transpulmonary passage of contrast, was detected in 49 (28%) subjects.

Of  these 49 subjects, 31 were re-evaluated after breathing 100% O2 for 10–15min and 6 (19%) contin-

ued to demonstrate the transpulmonary passage of contrast. Additionally, 18 of these 49 subjects were

re-evaluated in the upright position and 1 (5%) continued to demonstrate the transpulmonary passage

of  contrast. These data suggest that ∼30% of  healthy, young, asymptomatic subjects demonstrate the

transpulmonary passage of  contrast at rest which is reduced by breathing 100% O2 and assuming an

upright body position.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

In the last decade, a number of studies using intravenously

injected contrast in combination with echocardiography have

investigated the relationships between the transpulmonary pas-

sage of contrast during exercise (and other conditions) and several

physiologically important functions, including pulmonary gas

exchange efficiency (Lovering et al., 2008a; Stickland and Lovering,

2006; Stickland et al., 2004, 2006) and pulmonary pressure reg-

ulation (La Gerche et al., 2010; Lalande et al., 2012; Stickland

et al., 2004). The majority, but not all, of these studies (Eldridge

et al., 2004; Elliott et al., 2011; Kennedy et al., 2012; Laurie et al.,

2012, 2010; Lovering et al., 2008a,b; Stickland et al., 2004) using

intravenously injected contrast sought to exclude subjects for the

presence of an intracardiac pathway (i.e. PFO) because it is known

that the prevalence of PFO has been reported in various popu-

lations to be ∼25–40% and therefore should be a fairly common

∗ Corresponding author at: Department of Human Physiology, University of Ore-

gon, 1240Universityof Oregon, Eugene,OR 97403-1240,USA.Tel.:+1 5413460831;

fax: +1 541 346 2841.

E-mail addresses: lovering@uoregon.edu, jelliot2@uoregon.edu (A.T. Lovering).

finding (Hagenet al., 1984;Marriott etal.,2013;Woods etal., 2010).

Neglecting to exclude subjects for a PFO could result in ambigu-

ous results as one would be unable to determine if the presence

of left heart contrast was the result of contrast passage via a PFO

or via an intrapulmonary pathway. Thus, excluding subjects with

a PFO would ensure that any potential relationships between the

transpulmonary passage of contrast and the physiological variable

in question would be specific to contrasttraveling through an intra-

pulmonary pathway, rather than a PFO.

Less well accepted is a rationale for screeningand eitherinclud-

ing or excluding subjects without PFO but who demonstrate the

transpulmonary passage of contrastat rest breathing room air prior

to any intervention. Clearly, if subjects demonstrate the transpul-

monary passage of contrast prior to the intervention (i.e. while at

rest breathing room air) there is no baseline without the transpul-

monary passage of contrast on which to form conclusions. Hence,

this is onerationale forscreening subjects andexcludingthosewho

demonstrate thetranspulmonary passage of contrastat restbreath-

ing room air. Recent work by Woods et al. is to our knowledge

the only report of the prevalence of the transpulmonary passage

of saline contrast in humans at rest (Woods et al., 2010). In this

study it was determined that 29/104 (28%) subjects demonstrated

the transpulmonary passage of contrast at rest, however this was

1569-9048/$ – seefrontmatter © 2013 Elsevier B.V. All rights reserved.

http://dx.doi.org/10.1016/j.resp.2013.04.019

7/29/2019 Prevalence of left heart contrast in healthy, young, asymptomatic humans at rest breathing room air

http://slidepdf.com/reader/full/prevalence-of-left-heart-contrast-in-healthy-young-asymptomatic-humans-at 2/8

72 J.E. Elliott et al./ Respiratory Physiology & Neurobiology188 (2013) 71–78

 Table 1

Anthropometric and demographic data of all 176 subjects.

Clear, n= 60 Transpulmonary, n= 49 PFO, n= 67

Height (cm) 175 ± 8 174 ± 10 172 ± 11

Weight (kg) 71 ± 12 68 ± 12 69 ± 12

BMI (kg/m2 ) 23.1 ± 2.8 22.6 ± 2.9 23.2 ± 2.8

Age (yrs) 23 ± 4 24 ± 4 23 ± 3

Female (%) 43 47 51

Values areexpressed as mean± standard deviation.

not an asymptomatic population as the authors were investigating

the potential relationship between left heart contrast and migraine

headache. Also noted in the work by Woods et al. the transpul-

monary passage of contrast in healthy humans at rest who do

not have a PFO is conventionally assumed to be an uncommon

finding. This is presumably because the prevalence of otherwise

healthy, asymptomatic humans who demonstrate the transpul-

monary passage of contrast is not well established. Previous work

done in our laboratory using transthoracic saline contrast echocar-

diography (TTSCE) in healthy humans has meticulously screened

for and excluded subjects with PFO, and for subjects without PFO

who demonstrate the transpulmonary passage of contrast at rest

breathing room air (Elliott et al., 2011; Laurie et al., 2012, 2010).Accordingly, our laboratory has accumulated a large data set of 

healthy, young, asymptomatic human subjects and is in a unique

position to aidin establishingthe prevalence of thetranspulmonary

passage of contrast at rest breathing room air in this population.

Thus, the first aim of this study is to report a retrospective analysis

of these data.

In addition, it has recently been demonstrated that breathing

100% O2 prevents or significantly reduces the transpulmonary pas-

sage of contrast during exercise in healthy human subjects (Elliott

et al., 2011; Lovering et al., 2008b). Furthermore, Stickland and

colleagues reported that subjects who demonstrated the transpul-

monarypassage of contrast at rest breathing room airin the supine

position (2/8), no longer did so after adopting an upright posi-

tion (Stickland et al., 2004). Thus, if breathing 100% O2 or changesin body positioning in healthy human subjects at rest prevented

the detection of the transpulmonary passage of contrast, then this

could potentially alter the findings from TTSCE studies in subjects

breathing supplemental O2 or in various imagingpositions. Accord-

ingly, the second aim of this study was to determine if breathing

100% O2 or standing in the upright position would prevent the

detection of the transpulmonary passage of contrast in a subset

of healthy, young, asymptomatic human subjects at rest.

2. Methods

This study includes the retrospective analysis of echocardi-

ographic screening data collected at the Cardiopulmonary and

Respiratory Physiology Laboratory at the University of Oregon,between 2008 and 2012. This includes data from 179 healthy,

asymptomatic, non-smoking subjects between the ages of 18 and

41 without a self-reported history of cardiopulmonary disease

(Table 1). All179 subjects have enrolled inan ongoing or completed

study within our laboratory, and each subject provided verbal and

writteninformedconsent priorto participation. All studies received

approval from the University of Oregon Committee for the Pro-

tection of Human Subjects Institutional Review Board and were

conducted according to the Declaration of Helsinki.

The comprehensive echocardiographic assessment each subject

underwent was performed to identify and exclude subjects from

participating in studies within our lab on the basis of previously

undetected cardiopulmonary disease, PFO, or if they demonstrated

the transpulmonary passage of contrast at rest breathing room

air. Accordingly, prior work published from our laboratory only

includes subjects without cardiopulmonary disease, PFO or the

transpulmonary passage of contrast at rest breathing room air

(Elliott et al., 2011; Laurieet al., 2012,2010). However,regardless of 

subjects inclusion or exclusion from prior work, the current report

is inclusive of all healthy, asymptomatic subjects (i.e. without

known lung disease) who have undergone this echocardiographic

assessmentin ourlab. Allsonography wasperformed usinga Philips

Sonos 5500, by a registered diagnostic cardiac sonographer in both

adult and pediatric echocardiography with 25 years of experience

using TTSCE. Of the 179 healthy subjects screened within our lab-

oratory between 2008 and 2012, we identified and excluded one

female subject with a small pericardial effusion and two male

subjects with bicuspid aortic valves. The remaining 176 subjects

with normal ventricular function, valvular function, great vessels,

pericardium, and without evidence of myocardial ischemia or con-

genital heart disease are included in this analysis.

 2.1. Transthoracic saline contrast echocardiography

Initial agitated saline contrast studies were performed with

subjects breathing room air and reclined at 45◦ in the left lat-

eral decubitus position where a clear apical, four-chamber view

was obtained. Care was taken to optimally visualize all four cham-bers, interatrial septum and delineate myocardial and valvular

structures by individually adjusting the receiver gain settings.

Each saline contrast injection was created by manually agitating

3 ml of sterile saline with 1 ml of room air for 15s between two

10ml syringes connected in parallel to two 3-way stopcocks. The

saline–air microbubble suspension was then immediately injected

in a constant, forceful manner into a peripheral antecubital vein

via an IV catheter (20–22 G). This mixture of salineand air provides

excellent right sidedcontrast(Fig.1). Followingopacificationof the

right atrium and ventricle, the subsequent 20 cardiac cycles were

recorded at >30frames/s for further analysis.

The appearance of ≥1 microbubble in the left atrium or ventri-

cle in any frame during the subsequent 20 cardiac cycles served as

the criterion that subjects were either positive for an intracardiacright-to-left shunt (i.e. PFO) or demonstrated the transpulmonary

passage of contrast (Freeman and Woods, 2008; Hlastala and Van

Liew, 1975; Kjeldsen et al., 1999; Meerbaum, 1993; Meltzer et al.,

1981; Nanthakumar et al., 2001; Roelandt, 1982; Tsujino and

Shima, 1980; Woods et al., 2010; Y ang et al., 1971a,b). Saline con-

trast injections were performed during normal breathing, as well

as immediately following the release of a Valsalva maneuver in

order to transiently elevate right atrial pressure and create condi-

tions optimal for thedetection of an intracardiacright-to-left shunt.

Effective Valsalva maneuvers were confirmed by a transient left-

ward shift of the interatrial septum. Valsalva maneuvers do not

increase left heart contrast in the absence of a PFO. An intracar-

diac right-to-left shunt was suspected if microbubbles appeared in

the left heart ≤3 cardiac cycles following right heart opacification(Cabanes et al., 2002; Di Tullio et al., 1992; Lamy, 2002). More-

over, in all subjects careful color flow Doppler interrogation of the

interatrial septum in multiple imaging planes was performed and

additional evidence for an atrial septal defect (ASD) was ruled out,

such as a dilated coronary sinus or right-heart chamber dilation

and a normal and separate tricuspid annular plane from the mitral

annulus was confirmed. The transpulmonary passage of contrast

was positive if contrast appeared in the left heart >3 cardiac cycles

following right heart opacification (Hlastala and Van Liew, 1975;

Meerbaum, 1993; Meltzer et al., 1981; Roelandt, 1982; Tsujino and

Shima, 1980; Yang et al., 1971a,b).

Following every saline contrast injection, the recording was

meticulously reviewed frame-by-frame to ensure microbub-

bles were accurately diff erentiated from myocardial or valvular

7/29/2019 Prevalence of left heart contrast in healthy, young, asymptomatic humans at rest breathing room air

http://slidepdf.com/reader/full/prevalence-of-left-heart-contrast-in-healthy-young-asymptomatic-humans-at 3/8

 J.E. Elliott et al. / Respiratory Physiology & Neurobiology188 (2013) 71–78 73

Fig.1. Representative echocardiogramsat rest ina subject without PFOor anyleftheart contrast(A), in a subject with a PFOand bubblescoreof 3 (B)and in a subject without

a PFO with a bubble score of 2 from thetranspulmonarypassage of contrast (C). Note theadequate degreeof right heart contrast in all images.

structuresas well as from potentialpseudocontrast. Multiple saline

contrast injections were made in each subject as needed, during

each condition, to clarify if there was uncertainty and confirm

findings from previous injections. Bubble scores between 0 and

5 were assigned as previously described (Lovering et al., 2008b).

This scoring system is based on both the density and spatial dis-

tribution of microbubbles in the left ventricle and is used solely as

a qualitative assessment of the degree of left heart contrast. One

bubble score was assigned for each contrast injection in the fol-

lowingmanner: 0 = 0 microbubbles; 1 = 1–3microbubbles;2 = 4–12microbubbles; 3 = more than 12 microbubbles in a bolus; 4 = more

than 12 microbubbles heterogeneously distributed throughout the

left ventricle; 5 = more than 12 microbubbles homogeneously dis-

tributed throughout the left ventricle.

In subjects who demonstrated the transpulmonary passage of 

contrast at rest breathing room air, additional saline contrast injec-

tions were performed following breathing 100% O2 via a clinical

rebreathe mask at 6 l/min for 10–15 min. Accordingly, the actual

fraction of inspired oxygen (FIO2) received was between 0.50 and

0.70 confirmed by directly sampling the interior of the mask.

This is in agreement with the American Association of Respira-

tory Care for the expected FIO2 using this apparatus and flow rate

(Kallstrom,2002). Oxygen breathing was maintained for 10–15 min

to ensure near complete partial pressure equilibration and N2washout.Additionally, prior workdone in our laboratory hasshown

that breathing 100% O2 for up to 27 min does not increase the

transpulmonary passage of contrast (Elliott et al., 2011). These sub-

 jects remained reclined at 45◦ in the left lateral decubitus position

during the period of O2 breathing. While still breathing O2 and in

the same position, agitated saline contrast injections were again

performedwithout and immediately followingthe release of a Val-

salva maneuver. The persistence of ≥1 microbubble being observed

within the left atrium/ventricle >3 cardiac cycles following right

heart opacification indicated the period of O2 breathing had no

effect on the transpulmonary passage of contrast. However, if no

microbubbles were observed within the left atrium/ventricle in

any frame during the subsequent 20 cardiac cycles, it was deter-

mined that O2 breathing eliminated the transpulmonary passageof the saline contrast. Additionally, these subjects were invited to

be re-evaluated on a subsequent visit to determine whether or

not positional changes altered the outcome of the saline contrast

injections. Again, bubble scores were assigned as above (Lovering

et al., 2008b). Importantly each subject studied during O2 breath-

ing or with positional changes served as their own control as both

of these interventions have only ever been demonstrated to reduce

the transpulmonary passage of contrast. Thus,O2 breathingor posi-

tional changes do not alter the findings in subjects who do not

demonstrate thetranspulmonary passage of contrastat restbreath-

ing room air in the left lateral decubitus position.

The ultrasound technician was the primary bubble scorer. This

technician has served as the primary bubble scorer for three previ-

ous publications from our group and we foundexcellent agreement

with the secondary scorer (Elliott et al., 2011; Laurie et al., 2012,

2010). To further validate the reproducibility of this scoring system

in our lab for this study, a cardiologist who was blinded to the con-

ditions reviewed 57 saline contrastrecordings, each corresponding

to a single injection. Of theserecordings, there was 100% agreement

on whether there was or was not left heart contrast and 53 were

assigned the same bubble score corresponding to a ∼93% agree-

ment between the registered diagnostic cardiac sonographer and

cardiologist. With respect to the reproducibility of bubble scores

in subjects, we have previously demonstrated that individual sub- jects have the same bubble scores on different days, at rest and

during exercise, breathing either normoxic, hypoxic or hyperoxic

gas mixtures (Elliott et al., 2011; Laurie et al., 2010).

 2.2. Statistics

Bubble scores comparing breathing room air O2 breathing both

withand without a Valsalva maneuverwere analyzedusing a Fried-

man’s test with Dunn’s multiple comparison post-test,  p< .05 was

considered significant.

3. Results

A total of 176 subjects (83 female) passed the initial compre-hensive cardiac assessment, and in all subjects complete TTSCE

was performed. See Fig. 2 for a flow chart illustrating subject pro-

gression through the study and a summary of the overall results.

There were 60 (34%) subjects (26 female) with no indication of an

intracardiac right-to-left shunt and who did not demonstrate the

transpulmonary passage of contrast at rest breathing room air. Of 

the remaining116 (67%) subjects,67 (38%) subjects (34 female) had

evidenceof an intracardiacright-to-left shunt consistent witha PFO

(withor withoutthe release of Valsalva)while 49(28%) subjects (26

female) demonstrated the transpulmonary passage of contrast at

rest breathing room air. No subjects demonstrated any evidence of 

an ASD, includinga sinus venous ASD, coronary sinus ASD, primum

ASD/AVSD or secundum ASD.

The degree of left heart contrast observed in subjects withoutPFO but who demonstrated the transpulmonary passage of con-

trast was qualitatively less (i.e. lower bubble scores) than that

of subjects with PFO (Table 2). Thirty-eight of the 49 (78%) sub-

 jects who demonstrated the transpulmonary passage of contrast

were assigned a bubble score of 1, whereas only 14 out of the 67

(21%) subjects with PFO were assigned a bubble score of 1. Simi-

larly, all subjects withthe transpulmonary passage of contrastwere

assigned a bubble score of ≤3, whereas 18 out of the 67 (27%)

subjects with PFO were assigned bubble scores >3.

In theory TTSCE is not without risk in subjects with either

an intracardiac or intrapulmonary right-to-left pathway, due to

the potential for paradoxical microbubble embolism. Despite this

potential risk, no adverse events characteristic of neurologic

impairment due to paradoxical embolism were observed. These

7/29/2019 Prevalence of left heart contrast in healthy, young, asymptomatic humans at rest breathing room air

http://slidepdf.com/reader/full/prevalence-of-left-heart-contrast-in-healthy-young-asymptomatic-humans-at 4/8

74 J.E. Elliott et al./ Respiratory Physiology & Neurobiology188 (2013) 71–78

179 subjects screened

3 excluded 116 with left heart contrast 60 without left heart contrast

67 consistent with PFO49 with the transpulmonary

passage of contrast

31/49 were re-evaluated

18/31 evaluated for

postural differences

All 31 evaluated after

breathing 100% O2 for 10-15 min

25/31 without

left heart contrast

6/31 with

left heart contrast17/18 without

left heart contrast

1/18 with

left heart contrast

Fig. 2. Flow chart depicting the progression of all 176 subjects screened with transthoracic saline contrast echocardiography, as well as after being re-evaluated following

10–15 min of O2 breathing and postural changes.

results are consistent withpreviousliterature suggesting the safety

profile of SCE is good and warrants minimal risk (Bommer et al.,

1984; van Gent et al., 2009).

 3.1. Effect of O 2 breathing on the transpulmonary passage of 

saline contrast in healthy humans

Thirty-one of the 49 subjects without PFO who demonstrated

the transpulmonary passage of contrast at rest breathing room

air, subsequently breathed 100% O2 via a clinical rebreathe mask

for 10–15 min at 6 l/min which resulted in 25/31 (81%) subjects

no longer demonstrating the transpulmonary passage of contrast

(Fig. 3). Bubble scores comparing all subjects breathing room airand O2 were significantlydifferent. For representative TTSCE recor-

dings of subjects who demonstrated the transpulmonary passage

of contrast at rest breathing room air, that either persisted or was

eliminated after 10–15min of O2 breathing, please refer to video

files1–4 in theonline datasupplement.Note thatall echocardiogra-

phic analysis was done by reviewing each saline contrast injection

frame-by-frame, with any necessary further analysis being done

using the DICOM image viewer OsiriX (v4.1.1). Thus, the video files

in the online database may appear of low resolution, however this

is a reflection of compressing the file sizes for online storage and

did not affect our analysis or interpretation.

 Table 2

Qualitative assessment of left heart contrast (bubble score) in all subjects at rest

breathing room air.

Bubble score Clear Transpulmonary PFO

0 60 0 0

1 0 38 14

2 0 8 20

3 0 3 15

4 0 0 11

5 0 0 7

Total: 60 49 67

Each column represents the number of subjects, in all three groups, who were

assigned each bubble score. Clear , did not demonstrate any left heart contrast;

Transpulmonary, demonstrated left heart contrast consistent with the transpul-

monary passage; PFO, demonstrated left heart contrast consistent with a PFO.

 3.2. Effect of postural changes on the transpulmonary passage of 

saline contrast in healthy humans

We invited the 31 subjects who participated in the period of 

O2 breathing to return to the lab to evaluate the effect of postural

changes on our ability to detect the transpulmonary passage of 

contrast at rest breathing room air. Of these 31 subjects, 18 were

able to return for further evaluation and all 18 subjects demon-

strated identical bubble scores in the left lateral decubitus position

in allconditionsas we originally determined, up to 20 months prior.

However, after adopting the upright position 17 out of the 18 sub-

 jects (94%) no longer demonstrated the transpulmonary passage of 

saline contrast.

4. Discussion

The current work represents a retrospective analysis of 176

healthy, young, asymptomatic subjects studied using TTSCE, and

to our knowledge, is the first report providing data on the preva-

lence of left heart contrast in this population at rest breathingroom

air. We found that 34% of subjects had neither a PFO nor demon-

strated the transpulmonary passage of contrast at rest breathing

room air, 38% of subjects had a PFO, and the remaining 28% of 

subjects demonstrated the transpulmonary passage of contrast

at rest breathing room air. These data provide evidence that the

transpulmonary passage of contrast in healthy, young, asymp-

tomatic subjects at rest should not be an uncommon finding.

Room Air 100% O2

0

1

2

3

4

5

Score 2 or 1 to 1Score 3 to 2 or 1

Score 1 to 0

*

   B  u   b   b   l  e

   S  c  o  r  e

Fig. 3. Bubble scoresfor each subject breathing room airfollowed by O2 breathing.

The shade of each data point signifies the bubble score for each subject during O2

breathing,and these subjects corresponding bubblescore whilebreathingroom air.

7/29/2019 Prevalence of left heart contrast in healthy, young, asymptomatic humans at rest breathing room air

http://slidepdf.com/reader/full/prevalence-of-left-heart-contrast-in-healthy-young-asymptomatic-humans-at 5/8

 J.E. Elliott et al. / Respiratory Physiology & Neurobiology188 (2013) 71–78 75

In addition, in subjects without PFO but who demonstrated the

transpulmonary passage of contrast at rest breathing room air, O2

breathing for 10–15 min andstandingin the upright position elim-

inated the transpulmonary passage of contrast in 81% and 94% of 

subjects, respectively.

4.1. Prevalence of patent foramen ovale in healthy humans

The prevalence of PFO is reported to be ∼30% of the general

population, but varies between ∼25% and 40% depending on the

method of detectionand ageof subjects, i.e.saline contrastechocar-

diography or via a probe in autopsy studies (Hagen et al., 1984;

Marriott et al., 2013; Woods et al., 2010). The current work reports

a 38% prevalence of PFO in our subject population using TTSCE. In

subjects with PFO it is arguably not possible to definitively deter-

mine whether or not observing left heart contrast in >3 cardiac

cycles post right heart opacification is a result of the intracar-

diac pathway or blood flow through an intrapulmonary pathway.

Accordingly, these data demonstrate why screening and excluding

subjects forPFO is importantto do in studies using TTSCE if thegoal

of that study is to detect the transpulmonary passage of contrast

and relate this with some physiologic function.

4.2. Prevalence of the transpulmonary passage of saline contrast 

in healthy humans

Although recent literature has suggested that observing the

transpulmonary passage of contrast is a fairly common finding

(Woods et al., 2010; Woods and Patel, 2006), this is the first study

in a large sample size to report the prevalence of healthy, young,

asymptomatic subjects who demonstrate the transpulmonary pas-

sage of contrast at rest breathing room air (∼28%). Previous

reports which include a description of an otherwise, healthy con-

trol population using SCE exist, however they do not represent

an asymptomatic sample. These other examples in the literature

include reports aimedat characterizing symptomatic patient popu-

lations with hereditary hemorrhagic telangiectasia and pulmonary

arteriovenous malformations, and as such used subjects referredto these speciality clinics who failed to meet criteria for hereditary

hemorrhagic telangiectasia as the control population (Gazzaniga

et al., 2009; Gossage and Kanj, 1998; van Gent et al., 2009).

Considering all 176 subjects in the current study were healthy,

asymptomatic, and free of any cardiopulmonary disease, it is

unlikelythatthe 49 (28%)subjects whodemonstrated the transpul-

monary passage of contrast were due to pulmonary arteriovenous

malformations of a pathologic origin (Gossage and Kanj, 1998;

Shovlin and Letarte, 1999). Thus, other more plausible explana-

tions for the transpulmonary passage of contrast at rest could be

the presence of intrapulmonary arteriovenous anastomoses, which

were first reported to exist in healthy human lungs by Tobin, over

60 years ago (Tobin, 1966; Tobin and Zariquiey, 1953; Tobin and

Zariquiey, 1950) or via distendedpulmonary capillaries and/or cor-ner vessels as recently suggested by others (La Gerche et al., 2010;

Lalande etal., 2012).

Although the current work cannot distinguish between

transpulmonary passage of contrast via large diameter (>50m)

intrapulmonary arteriovenous anastomoses (Lovering et al., 2007)

or distention of pulmonary capillaries, work in animals during

exercise (Stickland et al., 2007) and while breathing hypoxic gas

at rest (Bates et al., 2012) demonstrate the transpulmonary pas-

sage of microspheres >25m and 70m in diameter, respectively.

Interestingly, under these same conditions, during exercise and

breathing hypoxic gas at rest, we and others have demonstrated

the transpulmonary passage of saline contrast in healthy humans

who do not demonstrate the transpulmonary passage of contrast

at rest breathing room air (Elliott et al., 2011; Laurie et al., 2010;

Lovering et al., 2008a; Stickland et al., 2004). Conversely, to date,

there is no direct evidence that either pulmonary capillaries or

corner vessels can distend to diameters greater than 20m, even

under pulmonary pressures that are supra-physiologic for humans.

For example at pulmonary pressures up to 73mm Hg in isolated

greyhound lungs the mean capillary diameter was found to be

6.5m and themaximum capillarydiameter measured was 13m

(Glazier et al., 1969; Rosenzweig et al., 1970), which would not be

a sufficient diameter to allow for the passage of a 25

m, 50

m

or 70m microsphere. Likewise, Manohar and Goetz found that

15m microspheres do not traverse the pulmonary circulation of 

the exercising Thoroughbred horse, prompting these authors to

conclude that pulmonary capillaries do not distend above 15m,

even at cardiac outputs and pulmonary pressures that far surpass

those in humans, e.g. >200l/min and ∼100 mm Hg, respectively

(Manohar and Goetz, 2005). Of note, the pulmonary capillary

distensibility coefficient (˛, fractional diameter change/mm Hg

pressure) of horses has been calculated to be 0.01, which is at the

lower limit of normal as calculated by Reeves and colleagues for

humans (∼0.02) (Reeveset al., 2005). Insummary,it seems unlikely

that capillary distention can account for these findings in healthy

humans at rest.

4.3. Effect of O 2 breathing on the transpulmonary passage of 

saline contrast in healthy humans

In subjects who do not demonstrate the transpulmonary pas-

sage of contrast at rest breathing room air, breathing 100% O2

during exercise prevents or significantly reduces the transpul-

monary passage of contrast (Elliott et al., 2011; Lovering et al.,

2008b). Similarly, breathing 100% O2 for 10–15min in healthy

human subjects eliminated the transpulmonary passage of contrast

in 25 out of the 31 subjects studied, all of which demonstrated

the transpulmonary passage of contrast at rest breathing room

air. In support of these findings is work by Niden and Aviado

which demonstrated that the number of intravenously injected

glass microspheres (up to 420m in diameter) that were recov-

eredfrom thepulmonary venous effluent decreasedin anesthetizeddogs ventilated with 100% O2 (Niden and Aviado, 1956).

Subjects in the current work breathed 100% O2 via a clini-

cal rebreathe mask and as such the actual FIO2 was between

0.50 and 0.70 (Kallstrom, 2002). For this reason, and to allow

adequatetime for nearly complete invivogaspartial pressure equil-

ibration to occur, we chose to evaluate subjects after 10–15 min

of O2 breathing. Indeed, prior work (Lovering et al., 2008b) has

demonstrated O2 breathing to reduce the transpulmonary pas-

sage of contrast in healthy humans during exercise in ∼2min.

This rapid response could potentially be attributed to the actual

FIO2 being 1.0, however, most subjects (5/7) in this prior work

still maintained some degree of left heart contrast (1–3 microbub-

bles), especially at maximal exercise. Therefore, it remains possible

that had subjects continued to exercise while breathing 100%O2, left heart contrast would have eventually been entirely elim-

inated. Additionally, considering the actual FIO2 and duration

of O2 breathing in the current report, it is unlikely that any

potential negative effect of O2 breathing was encountered, such

as an excessive accumulation of reactive oxygen species lead-

ing to CNS toxicity (Dean, 2004). Although it is unknown how

O2 breathing is preventing the transpulmonary passage of con-

trast, previous work has shown that O2 breathing does not affect

the viability/stability of microbubbles created from room air gas.

Accordingly, right heart contrast is unaffected (see videos 1–4

in the online database) and the absence of left heart contrast

does not reflect a disappearance or absence of microbubbles in

the right heart prior to reaching the left heart due to an alter-

ation of  in vivo gas bubble dynamics when the external partial

7/29/2019 Prevalence of left heart contrast in healthy, young, asymptomatic humans at rest breathing room air

http://slidepdf.com/reader/full/prevalence-of-left-heart-contrast-in-healthy-young-asymptomatic-humans-at 6/8

76 J.E. Elliott et al./ Respiratory Physiology & Neurobiology188 (2013) 71–78

pressure environment is changed during O2 breathing (Elliott

et al., 2011). In addition, Melsom and colleagues have shown that

hyperoxia causes a significant redistribution of pulmonary blood

flow in sheep (Melsom et al., 1999) suggesting a hyperoxia-

mediated, active control mechanism for the pulmonary vasculature

which may redirect blood flow away from intrapulmonary arte-

riovenous anastomoses. Taken together, we would speculate, as

others have (McMullan et al., 2004) that blood flow through

intrapulmonary arteriovenous anastomoses may be regulated in

a manner similar to the systemic vasculature or ductus arteriosus

whereby hypoxia causes vasodilation and hyperoxia causes vaso-

constriction, thereby preventing the transpulmonary passage of 

contrast.

4.4. Effect of body positioning on the transpulmonary passage of 

saline contrast in healthy humans

Inadditionto theinspired O2 concentration altering the findings

from TTSCE, body positioning may also be an importantfactor. Pre-

viously, Stickland et al. demonstrated that standing in the upright

position eliminated the transpulmonary passage of contrastin sub-

 jects (2/8) who while supine, did demonstrate the transpulmonary

passage of contrast (Stickland et al., 2004). However, standing in

the upright position did not result in the transpulmonary pas-sage of contrast in subjects (6/8) without left heart contrast while

supine (Stickland et al., 2004). Accordingly, we re-examined 18 of 

the 49 subjects who demonstrated the transpulmonary passage

of contrast at rest breathing room air in the left lateral decubi-

tus position after adopting an upright position. All 18 subjects had

identical baseline TTSCE results while in the left lateral decubitus

position compared to their previous visit (up to 20 months prior),

and in 17 (94%) of these subjects standing in the upright posi-

tion eliminated the transpulmonary passage of contrast. Although

the precise mechanism remains unknown, going from supine to

the upright position would redistribute pulmonary blood flow

away from the apical portion of the lungs toward the dependent

portion of the lungs, suggesting that perhaps in the upright posi-

tion this redistribution of pulmonary blood flow may explain thesudden elimination of left heart contrast. Interestingly, Tobin and

Zariquiey have previously demonstrated thatintrapulmonary arte-

riovenous anastomoses are located within the apex of the lung,

which may explain the transpulmonary passage of contrast while

supine versus theabsenceof leftheart contrastwhileupright(Tobin

and Zariquiey, 1950).

4.5. Implication of the transpulmonary passage of contrast on

 pulmonarymedicine and physiology

The potential physiologic role and/or pathophysiologic impli-

cations of the transpulmonary passage of contrast remain to be

determined. Previous work has suggested that the transpulmonary

passage of contrast is occurring via intrapulmonary arteriove-nous anastomoses which may represent a source of right-to-left

shunt, and thus, may contribute to pulmonary gas exchange effi-

ciency (Lovering et al., 2008a; Stickland et al., 2004; Stickland and

Lovering, 2006). Others have argued that the transpulmonary pas-

sage of contrast may represent pulmonary capillary and/or corner

vessel distention and therefore may be involved with regulating

pulmonary artery pressure (Bryan et al., 2012; La Gerche et al.,

2010; Lalande et al., 2012; Laurie et al., 2012; Stickland et al.,

2004), or influencing pulmonary vascular reserve and minimiz-

ing right ventricular afterload during exercise (La Gerche et al.,

2010; Lalande et al., 2012; Stickland et al., 2004). Others still have

argued that the presence, or absence, of left heart contrast has

limited to no significant physiologic and/or pathophysiologicimpli-

cations (Hopkins et al., 2009; Vogiatzis et al., 2008). Certainly, data

surrounding the potential implications of the transpulmonary

passage of contrast on pulmonary gas exchangeand regulating pul-

monary vascular pressures remains controversial and more work

is neededin this area to eitherconfirm or refute these postulations.

4.6. Limitations

This study exclusively utilized TTSCE and no subjects went on

to further investigation using TEE with agitatedsaline contrast. We

were predominately concerned withdetecting the transpulmonary

passage of contrast,and thus, theoptimal visualization of the inter-

atrial septum that TEE provides was not essential. Because of the

possibility that contrast could be traversing a PFO due to the pres-

sure changes associatedwith normal respiration throughout the 20

cardiac cycle recording, we performedmultipleagitatedsaline con-

trast injections, with and without the use of a Valsalva maneuver

and our 38% prevalence of PFO is evidence of our rigorous inspec-

tion. Indeed this is higher than previous autopsy studies which,

depending on the patients age, detected a 25–34% prevalence of 

PFOusingaprobe(Hagenet al., 1984).Of note,similarto the current

work and also using TTSCE, Woods et al. report a 38% prevalence

of PFO in a population of 104 subjects with a history of migraine

(Woods et al., 2010),whileMarriottet al.report a 35% prevalence of 

PFO in a population of 1162 subjects referred for TTSCE for a multi-tude of reasons, mostcommonly due to transientischemic attack or

to assess right ventricular function (Marriott et al., 2013). Thus, the

higherthan previously reported prevalence of PFOin ourstudyand

that of Woods et al. and Marriott et al. to previous autopsy reports,

is best explainedby thediff erences in detectiontechniques. Also,in

the current study weused3 ml of salineand 1 ml of air tocreate the

salinecontrast suspension whereas Woods et al.used 9 ml of saline

and 1ml of air, while Marriott et al. used 8.5 ml of saline, 0.5ml of 

air, and 1ml of blood. Despite these differences in contrast mixes,

a similar prevalence of PFO was detected in all studies.

When considering the majority of subjects studied demon-

strated low bubble scores, it is important to emphasize our

thorough TTSCE protocol and meticulous analysis. Differentiating

microbubbles from subvalvular structures that are moving in andout of the observational plane can be very difficult, and for this

reason we performed multiple saline contrast injections as nec-

essary, utilized a sonographer with extensive experience as the

primary analyzer and had a cardiologist who was blinded to the

conditions confirm these findings in a subset of 57 saline contrast

injections. Of these recordings there was 100%agreement between

the sonographer and cardiologist for whether there was, or was

not any left heart contrast and 94% agreement on assigning indi-

vidual bubble scores. Furthermore, previous work from our group

demonstrates an excellent degree of intra-subject repeatability on

different days, with respect to bubble scores for a given condition

(Elliott et al., 2011;Laurie et al., 2010). Lastly, we always performed

a baseline, normoxic, saline contrast injection in subjects reclined

in the left lateral decubitus position prior to any intervention toconfirm previous findings in all subjects. This further demonstrates

our ability to repeatedly identify very few microbubbles, often ≤3

microbubbles within each subject.

With respect to the application of our 0–5 bubble scoring sys-

tem to data obtained using TTSCE it deserves to be reiterated that

this scoring system is a 2-dimensional qualitative assessment of 

a 3-dimensional volume. Accordingly, this technique only offers a

qualitative assessment of either more or less contrast in the left

ventricle and has not been directly shown to correlate with more

or less blood flow through intrapulmonary arteriovenous anasto-

moses. The application of a scoring system in studies using TTSCE

does however provide researchers with the ability to categorize

subjects based on the qualitative degree of left sided contrast. Per-

haps the greatest strength of TTSCE is the fact that this technique

7/29/2019 Prevalence of left heart contrast in healthy, young, asymptomatic humans at rest breathing room air

http://slidepdf.com/reader/full/prevalence-of-left-heart-contrast-in-healthy-young-asymptomatic-humans-at 7/8

7/29/2019 Prevalence of left heart contrast in healthy, young, asymptomatic humans at rest breathing room air

http://slidepdf.com/reader/full/prevalence-of-left-heart-contrast-in-healthy-young-asymptomatic-humans-at 8/8

78 J.E. Elliott et al./ Respiratory Physiology & Neurobiology188 (2013) 71–78

Reeves, J., Linehan, J., Stenmark, K., 2005. Distensibility of the normal human lungcirculation during exercise.AmericanJournal of Physiology – Lung Cellular andMolecular Physiology 288, L419-L425.

Roelandt, J., 1982. Contrast echocardiography. Ultrasound in Medicine & Biology 8,471–492.

Rosenzweig, D.Y.,Hughes, J.M.,Glazier, J.B., 1970. Effectsof transpulmonaryand vas-cular pressureson pulmonarybloodvolume in isolated lung. Journal of AppliedPhysiology 28, 553–560.

Shovlin, C.L., Letarte, M., 1999. Hereditary haemorrhagic telangiectasia and pul-monary arteriovenous malformations: issues in clinical management andreview of pathogenic mechanisms. Thorax 54, 714–729.

Stickland, M.K., Lovering, A.T., 2006. Exercise-induced intrapulmonary arteri-ovenous shunting and pulmonary gas exchange. Exercise and Sport SciencesReviews 34, 99–106.

Stickland,M.K., Lovering,A.T.,Eldridge,M.W., 2007. Exercise-induced arteriovenousintrapulmonary shunting in dogs. American Journal of Respiratory and CriticalCare Medicine 176, 300–305.

Stickland, M.K., Welsh, R.C., Haykowsky, M.J ., Pet ers en, S.R., Anderson, W.D.,Taylor, D.A., Bouffard, M.,  Jones, R.L., 2004. Intra-pulmonary shunt and pul-monary gas exchange during exercise in humans. Journal of Physiology 561,321–329.

Stickland,M.K., Welsh, R.C., Haykowsky,M.J., Petersen, S.R.,Anderson, W.D.,Taylor,D.A.,Bouffard, M., Jones, R.L., 2006. Effect of acuteincreasesin pulmonaryvascu-lar pressureson exercisepulmonary gasexchange.Journal ofApplied Physiology100, 1910–1917.

Tobin, C.E., 1966. Arteriovenous shunts in the peropheral pulmonary circulation inthe human lung. Thorax 21, 197–204.

Tobin, C.E., Zariquiey, M.O., 1950. Arteriovenous shunts in the human lung. Pro-ceedings of theSociety forExperimental Biology and Medicine 75, 827–829.

Tobin, C.E., Zariquiey, M.O., 1953. Some observations on the blood supply of thehuman lung. Medical Radiography and Photography 29, 9–19.

Tsujino, T., Shima, A., 1980. The behaviour of gas bubbles in blood subjected to anoscillating pressure. Journal of Biomechanics 13, 407–416.

vanGent,M.W.F.,Post,M.C.,Snijder,R.J.,Swaans,M.J.,Plokker, H.W.M.,Westermann,C.J.J., Overtoom,T.T., Mager, J.J.,2009. Grading of pulmonary right-to-left shuntwith transthoracic contrastechocardiography: does it predict theindication forembolotherapy? Chest 135, 1288–1292.

Vogiatzis, I., Zakynthinos,S., Boushel,R., Athanasopoulos,D., Guenette, J.A.,Wagner,

H.,Roussos, C.,Wagner,P.D.,2008. Thecontributionof intrapulmonaryshuntstothe alveolar-to-arterial oxygen diff erenceduring exercise is very small. Journalof Physiology 586, 2381–2391.

Woods, T.D., Harmann, L., Purath, T., Ramamurthy, S., Subramanian, S., Jackson,S., Tarima, S., 2010. Small- and moderate-size right-to-left shunts identifiedby saline contrast echocardiography are n or mal and unrelated to migraineheadache. Chest 138, 264–269.

Woods,T.D., Patel, A.,2006. A critical review ofpatent foramenovale detectionusingsaline contrast echocardiography: when bubbles lie. Journal of the AmericanSociety of Echocardiography 19, 215–222.

Yang, W.J., Echigo, R., Wotton, D.R., Hwang, J.B., 1971a. Experimental studies of the dissolutionof gasbubbles in whole blood andplasma. I. Stationary bubbles.

 Journal of Biomechanics 4, 275–281.Yang, W.J., Echigo, R.,Wotton, D.R., Hwang,J.B.,1971b. Experimental studies of the

dissolution of gas bubbles in whole blood and plasma. II. Moving bubbles orliquids. Journal of Biomechanics 4, 283–288.