(PDF) The Possible Relationship Between Patent Foramen Ovale and Decompression Sickness:

A

DEFENCE

DEFENSE

The

possible

relationship

between

patent

foramen

ovale

and

decompression

sickness:

A

review

of

the

literature

M.J.

Saary

G.W

Gray

19990413

079

DEFENCE

AND

CIVIL

INSTITUTE

OF

ENVIRONMENTAL

MEDICINE

Technical

Report

DCIEM

TR

1999-001

January

1999

National

D6fense

Defence

nationale

Cwiadcl

DISTRIBUTION STATEMENT

A

Approved

for

Public

Release

Distribution

Unlimited

January

1999

DCIEM

TR

1999-001

THE

POSSIBLE

RELATIONSHIP

BETWEEN

PATENT

FORAMEN

OVALE

AND

DECOMPRESSION

SICKNESS:

A

REVIEW

OF

THE

LITERATURE

M.J.

Saary

G.W.

Gray

Defence

and

Civil

Institute

of

Environmental

Medicine

1133

Sheppard

Avenue

West,

P.O.

Box

2000

Toronto,

Ontario

Canada

M3M

3B9

"©

HER

MAJESTY

THE QUEEN

IN

RIGHT

OF

CANADA

(1999)-

as

represented

by

the

Minister

of

National

Defence

"©

SA

MAJESTE

LA

REINE

EN

DROIT

DU

CANADA

(1999)

D6fense

Nationale

Canada

DEPARTMENT

OF

NATIONAL

DEFENCE

-

CANADA

EXECUTIVE

SUMMARY

A

patent

foramen

ovale (PFO)

is

a

small

opening between

the

right

and

left

cardiac atria,

a

persisting

remnant

of

a

physiologic

communication

present

in

the

fetal

heart.

This

normally

closes

after

birth,

but

remains

patent through

to

adulthood

in

up

to

a

third

of

normal adults.

A

patent

PFO

is

a

potential

conduit

for

blood

clot

(resulting

in

a

stroke),

or

venous

gas

bubbles

during

decompression, (resulting

in

type

II

neurologic

decompression

sickness).

There

has been

considerable

controversy

about

the

significance

of

a

PFO

as

a

possible

mechanism

for

type

II

decompression

sickness.

Despite

the

high prevalence

of

PFO

in

the

general

population,

and

the relatively common

occurrence

of

venous

gas

bubbles

in

diving

and

altitude

exposures,

the

incidence

of

type

II

DCS

in

diving

or

with

altitude

exposure

is

low.

This

paper

reviews the

literature

with

respect

to

the

potential

for

right-

to-left

embolization

through

a

PFO,

relation

of

PFO

to

DCS,

screening

techniques

for

PFO,

and

treatment

options.

The

literature

supports

a

relationship between

the presence

and

size

of

PFO

and

cryptogenic

stroke

(stroke,

generally

in

younger

individuals

with

no

other

identifiable

risk

factors).

The

weight

of

evidence

also

favours

an

increased

relative

risk

of

type

II

DCS

with

a

PFO,

although

the

absolute

increase

in

risk accrued

is

small.

The gold

standard

for

PFO

screening

is

a

trans-esophageal

echocardiographic

(TEE)

and

colour

flow

study,

but

trans-cranial

Doppler

(TCD)

with

contrast

is

a

promising

technique

with

good

accuracy

compared

with

TEE.

ABSTRACT

There

continues

to

be

a

controversy

about

the possible

significance

of

patent

foramen

ovale

(PFO)

in

the

pathophysiology

of

type

II

decompression

sickness

(DCS

with

neurologic

symptoms).

PFO's

are

a

common

finding

in

normal

persons, being

present

in

up

to

a

third

of

the

population.

The

potential

for

right-

to-left

shunting

of

venous

gas

emboli

(VGE)

which

are

known

to

occur

in

even

no-decompression

dives

is

a

theoretical

concern,

yet

the

incidence

of

type

II

DCS

is

remarkably

low

given

the

prevalence

of

PFO.

Altitude decompression

is

analagous

to

decompression

from

a

saturation

dive,

and

VGE

are

observed

above

15,000

feet

(4572m).

The

potential

for

PFO

shunting

of

VGE

is

a

particular

concern

for

space

extra-vehicular

activity

(EVA)

where

the

pressure

in

the

US

EVA

suit

is

4.3

PSI,

equivalent

to

30,000

feet.

This

paper

reviews the

literature

with

respect

to

the

potential

for

right-to-left embolization

through

a

PFO,

relation

of

PFO

to

DCS,

screening techniques

for

PFO,

and

treatment

options.

The

literature

supports

a

relationship between

the

presence

and

size

of PFO

and

cryptogenic stroke

(stroke,

generally

in

younger individuals

with

no

other

identifiable

risk

factors). The

weight

of

evidence

also

favours an

increased

relative

risk

of

type

II

DCS

with

a

PFO,

although

the

absolute

increase

in

risk

accrued

is

small.

The

gold

standard

for

PFO

screening

is

a

trans-esophageal

echocardiographic

(TEE)

and

colour

flow

study,

but

trans-cranial

Doppler

(TCD)

with

contrast

is

a

promising

technique

with

good

accuracy

compared

with

TEE.

ii

TABLE

OF

CONTENTS

Executive

Summary

i

Abstract

ii

Table

of

Contents

iii

1.

Embryology

1

2.

Prevalence

of

PFO

2

3.

Detection

of

PFO

a.

Echocardiography

2

b.

Valsalva/cough

manoeuvers

4

c.

Transcranial

Doppler

5

4.

Clinical

Relevance

of

PFO

6

5.

PFO

and

Stroke

7

6.

Venous

Gas

and

DCS 8

7.

Arterial

Gas

and

DCS 9

8.

PFO

and

Diving

10

9.

Absolute

Risk

of

DCS

in

Diving

15

10.

DCS

and

Altitude

Decompression

15

11.

Screening

for

PFO

in

Divers

16

12.

Management

of

PFO

17

13.

Limitations

of Review

19

14.

Conclusions

19

15.

References

20

16.

Additional

information

26

Acknowledgements

29

iii

Embryology

The

foramen

ovale develops

as

part

of

the

process

of

atrial

septation.

Initially

in

this

process,

a

sagittal

fold

in the

roof

of

the

atrium

develops

into

the

septum

primum,

essentially

separating

the

2

atria.

Interconnection

between

the

atria

persists via

the

foramen

primum

until

the

septum

fuses

with

the

atrio-ventricular

(AV)

endocardial cushions.

In

order

to

maintain interatrial

right

to

left

blood flow with

a

fused

septum and

closed foramen primum,

a

second

foramen,

the

foramen

secundum, forms

just

prior

to

septal

fusion.

Subsequently,

a

second

septum,

the

septum

secundum

develops

to

the

right

of

the

septum

primum.

Like

the

septum

primum,

the

septum

secundum

is

also

an

incomplete

barrier

between

the

atria;

the

septum

secundum's

opening

being

called

the

fossa

ovalis.

Figure

1:

The

Patent

Foramen

Ovale

(From

O'Rahilly

&

Muller,

1992).

FO

ne,,.

I 2 I1

A.

I:n)nl

view

ol"

A

Ilte

heart

At

8

wevic.ks"

11w

• Of

the

awtma

ha.s

bren

p111l-1

'r'Alially

A t ,,

h.-w

thme

dhL'tl,

;rlterit)'.u

Svc

(I)A)

tItiore

dearl'. K.

Right

laterAl

R

aur.

view

41rlcr

(lK"

right atriutm

has

S

he.

'chtcl

ulvened.

C.

and

D..0.%,ti,n.%

L

thntuglu

the atria

tt'

,•how

the

val-

aU'.vr actloen

of

%eptlllm prtinnltn.

LVonu

which

ewoi

d

^cus

the

t,,r~Amen

Mt.

2

(vale

it

we1m

pr-ivmd

against

""svCeptum

.undumn.

E.

Righit

lateral

/7

RP

-

T

view

of

the

Intrattlal ,%ptum

at

A

RV

-- '"week., and (F) in

the

adulh

(Avmntls

L

V

a

rl

, nart..

c

n

t

l

'.s (fletl

m

S S u .

c

m

tn

a

r y

cs

i

Ssinto:

LauL..

left

altile:

IYV.

left

vcrvits%

valve.

R~aur..

right

auricle:

RV

RPV.

right

pulnmwoary

vein;

S.. wrp

tum

spurtsm:T.

tricuspid

valve.

A

C___--_.___.___E_..

_

and

II

art"

ha.i

on

W212ata.

CE

FO

F

ovails

Essentially what

has

taken

place

is

somewhat analogous

to

drawing

a

Venn

diagram,

with

the

development

of

two

apposed

incomplete septa with partially overlapping

fossae. The

foramen

ovale

represents

the opening

that

remains

patent

between

the

septae,

with

the

septum

primum

acting

as

a

valve

allowing

only

R

->

L

passage

of

blood.

Post-natally,

circulatory

changes

including

the

transfer

of

gas

exchange

from

the

placenta

to

the

lungs,

with

resultant

decrease

in

pulmonary

vascular

resistance

and

increased

pulmonary

blood

flow

lead

to

increased

L

atrial

pressure that

presses

the

interseptal

valve

against

the

septum

secundum.

Usually

within

the

first

2

years

of

life

the

septae

permanently

fuse

due

to the

development

of

fibrous

adhesions.

(O'Rahilly &

Muller,

1992).

Prevalence

of

PFO

In

some

however,

the

foramen

ovale

fails

to

fuse.

In

1984,

Hagen,

Scholz,

&

Edwards

studied

the

prevalence

of

patent

foramina

in

965

normal

hearts.

Their

autopsy

study

revealed

an

overall

prevalence

of

27.3%

(see

Table

1).

Some have

suggested

that

when

functional

imaging

such

as

trans-esophageal

echocardiography (TEE)

is

used

in

the live

patient,

the

prevalence

of

patent

foramen

ovale (PFO)

is

decreased

(Moon

et

al,

1989).

In

fact,

in

Hagen's

study

the

prevalence

of

PFO was

found

to

decline

with

increasing

age

(34.3%

in

first

3

decades

to

25.4%

in

the

4"'

to

8'h

decades),

so

sample

age

distribution

could

lead

to

varying

estimates

of

prevalence

among

studies. Hagen

et

al

cite several

articles

in

which high

prevalence

rates

of

approximately

31%

were

found among

older

patients.

They

attribute

this

lack

of

age-related

decline

to

the

fact

that

many

were

based

on

results

from

abnormal

hearts;

the

implication

being

that

altered

hemodynamics

of

heart

disease

among

older

patients

may

hinder

late PFO

closure.

Finally, among

hearts

with

patent

foramina, Hagen

et

al

found

the

average

foramen

size

to

be

4.9

mm, with

size

increasing

as

age

increased.

Table

1.

Incidence

of

Patent Foramen

Ovale

(PFO)

in

Various Studies

(From

Hagen

et

al,

1984)

Number

of

Incidence

of

Age

(yr)

of

Year

Authors

Hearts

PFO

(%)

Patients

1897

Parsons

and Keith

24

399

26

All

ages

1900

Fawcett

and Blachford

25

306

31.7

>10

1918

Scammon

and

Norris

26

1,809*

29

>1

1931

Patten

4,083*

24.6

Mostly

adults

1934

Seib

2

500

17

20

1948

Wright

et

a,

28

492

22.9

Mostly

adults

1972

Schroeckenstein

et

al

4

144

35.4

>20

1979

Sweeney

and

Rosenquist

29

64

31

>10

1984

Hagen

et

al

965

27.3

>1

*

Combined

review

of

literature

Clinical

Detection

of

PFO

a.

ECHOCARDIOGRAPHY

The

most definitive

studies

on the

prevalence

of

PFO have been

autopsy

studies.

To

determine

the

clinical

relevance

of

PFO,

one

must

have

a

detection

method that

is

useful

in

living

subjects.

Transesophageal

echocardiography

(TEE)

has

been

considered

the

gold

standard

in

this

regard,

although

many

other

detection methods

have

been assessed.

Indeed,

many

studies

support

the

view

that

the

most

accurate

method

for

detecting

PFO

is

TEE,

but

some

disagree

that

it

is

the

most

practical.

2

In

their

study

of

150

consecutive

patients, Siostrzonek

et

al

(1991)

compared

the

detection

rates

of

PFO

in

bubble contrast trans-thoracic

echocardiography

(TTE) and

TEE

(see

Table

2).

Detection

was

significantly

better with

contrast

TEE

imaging

than with

TTE

(30/150

vs

9/150,

p<.0001).

There

were

no

false

positives with

TTE

imaging,

however

there were

15

(50%)

false

negatives

and

6

(20%)

undetermined

cases.

All

of

the

patients with

a

positive

TTE

study

also

had

a

positive

TEE,

thus

an

unequivocal

contrast

TTE

study

negates

the

need

for further

TEE

imaging.

If

however,

the

TTE

is

negative

these authors recommend

TEE

in

all

such

patients

to

assess

for

presence

of

PFO.

Table

2:

Incidence

of

Positive,

Negative

and

Undetermined

Contrast

Studies

with

Transthoracic

and

Transesophageal

Contrast

Echocardiography

in

150

Patients

(From

Siostrzonek

et

al,

1991)

Transthoracic

Contrast

Transesophageal

Contrast

Echocardiography

Normal

Valsalva

Normal

Valsalva

Respiration

Maneuver

Respiration Maneuver

+

7

(5%)

9

(6%)

18

(12%)

30(20%)

0 125

(83%)

119(79%)

132(88%)

120

(80%)

Undetermined

18

(12%)

22(15%)

0

(0%)

0 (0%)

+

=

positive;

0

=

negative

Fischer

et

al

(1995)

assessed

the

prevalence

of

PFO

in live

patients. After summarizing

the

results

of

16

studies with

over

100

patients,

they concluded

that

with

TTE

the

prevalence

of

PFO

was

9.3%,

and

with

TEE

was

11.2%.

They

went

on

to

retro-spectively

examine

1000

of

their

own patients

with colour

Doppler and

contrast

TEE

for

the

presence

of

PFO. The

prevalence

of

PFO

was

found

to

be 9.2%,

with

all

92/1000

cases detected

by

TEE

and

only

22/1000

detected

by

colour

flow.

In addition,

they

found

that

atrial

septal aneurysms

(ASA)

were

more frequent

among

those

with

PFOs, with

ASA

present

in

15.2%

with

PFO

compared

to

6.1%

of

patients

without

PFO (p=.001).

Like

Hagen

et

al

(1984),

Fischer

et

al

found

an

age-related

decline

in

the

prevalence

of

PFO.

Belkin

et

al (1994)

compared

contrast TEE with

colour

flow

TEE,

contrast TTE

and

colour

flow

TTE in

43

patients.

Results

of

their

study

are

summarized below

in

Table

3,

and

indicate

that

PFO

is

more

frequently

detected

with

TEE

methods,

and

that slightly

more

PFOs

were detected

with

colour Doppler

than

with contrast

TEE.

3

Table

3:

Data

summarized

from

Belkin

et

al,

1994.

Parameter

contrast

TEE

colour

TEE contrast

TTE

colour

TTE

nPFO

(%)

14

(37%)

17

(45%)

9

(24%)

1 (3%)

Sens gold

79% 50%

7%

Spec

gold

75%

92%

100%

Schneider

et

al

(1996)

correlated

TEE

directly

with

subsequent

autopsy

findings

in

35

patients

in

order

to

assess

the

diagnostic

accuracy

of

both

colour

Doppler

and

contrast TEE.

A

PFO

was

found

in

9/35

patients

at

autopsy,

all

of

which

were

correctly

diagnosed

by

colour Doppler

TEE,

with

8/9

diagnosed

correctly

by

contrast

TEE.

Others

have

suggested that

contrast

TEE

is

superior

to

both

colour

flow TEE

and

TTE

in

general

(Moon,

1989;

Hausmann,

1992;

Luotolahti

et

al,

1995).

To

further delineate

TEE, Chezbraun

et

al

(1993)

compared

the

vertical

to

horizontal

plane

of

biplane

TEE

in

19

contrast-positive

PFO

patients.

They found

that

in

the

vertical plane

53%

(10/19)

of

PFOs could

be

seen

and

sized,

but

none

of

these

were

visible

in

the

horizontal

plane.

Although

biplane echocardiography

is

clearly

not

the

method

of

choice

for PFO

detection,

it

can

be

useful

for determining

the size

and

morphology

of

the

PFO,

which

may

be

relevant

for

therapeutic decision

making.

Although

TEE provides

better

resolution

than

TTE,

it

is

not

without risks. These

include

esophageal

injury,

laryngospasm,

aspiration,

hypoxia,

bronchospasm,

and

dysrhythmias.

(Porembka,

1996).

James

(1990)

notes

that

the

Contrast

Committee

of

the

American

Society

of

Echocardiography

has

record

of

28

transient

neurological

side

effects

occurring

in

41,000

contrast

echo

studies.

TEE

is

generally

considered

an

unpleasant

procedure

and

IV

sedation

is

often required.

In

addition,

many

find

performing

a

Valsalva

maneuver

difficult

with

the

probe

in

place.

Because

of

the

low sensitivity

of

TTE

and

the

relative

complexity

of

TEE,

a

simpler,

but

acceptably

sensitive

method

for

PFO

screening

was

introduced

in1991

by

Teague

and

Sharma,

this

being

transcranial Doppler

(TCD).

b.

VALSALVA/COUGH

MANOEVERS

It

is

common

practice

to

assess

for

the

presence

of

PFO using the

Valsalva

maneuver.

The

rationale

is

that

this maneuver

will

momentarily

increase right heart

pressure,

thereby

accentuating

any

right

to

left

shunt. Some

however,

believe that

the

cough

test

is

superior

to

the

Valsalva

in

identifying

the

presence

of

PFO.

(Dubourg

et

al,

1984;

Stoddard

et

al, 1993)

In

1994,

Jauss

et

al

simultaneously

performed

TEE and

TCD

in

50

patients

(galactose

microbubbles)

with

and

without

Valsalva. Compared

to

TEE,

the

sensitivity

of

TCD

was

100%

in

both

conditions.

Specificity

without

Valsalva

was

47%,

and

increased

to

93%

with

Valsalva

(see

Table

4).

4

Table

4:

Cross Table

for

Transesophageal

Echocardiography Compared

WithTranscranial

Doppler

Sonography

With

Valsalva

Maneuver

(Table

2.

From

Jauss et

al,

1994)

TEE

TCD

+ -

Sum

+

14

0

14

-

1

35

36

Sum

15

35

50

TCD-trancranial

Doppler

sonography. TEE-transesophageal

echo

+

=

detection

of

PFO;

-

no

detection

of

PFO

Sensitivity

=

0.93;

Specificity

=

1;

P<.O1,

Fischer's

exact

test

c.

TRANSCRANIAL

DOPPLER

In

their

study

of

111

patients,

Klotzsch

et al

(1994)

compared

contrast TEE, TTE

and

TCD

of

the

left MCA

as

methods

to

identify PFO

(see

Table

5).

With

TEE,

46

PFOs

were

found,

of

which

15

were

missed

by

TTE

(accuracy

of

TTE

44%).

In

comparison,

the

accuracy

of

TCD

was

found

to

be

92.8%. The sensitivity

and

specificity

of

TCD

compared

to

TEE were

91.3%

and

93.8%

respectively.

Table

5:

Comparison

of

the

ability

of

TEE and

contrast-TCD

to detect

a

PFO

in

111

patients

with

cerebral

ischemia

(Table

I

From Klotzsch

et

a[,

1994.)

TCD/TEE>

Permanent

Valsalva

Negative Totals

Permanent

19

4

2

25

Valsalva

3

16

2 21

Negative

1 3

61 65

Totals

23

23 65

111

Other

studies

comparing

TCD

to

TEE

demonstrate

sensitivities

ranging

from

68

to

100%

and

with specificities

repeatedly

in

the

order

of

100%

(Di

Tullio

et

al,

1993,

Kwiecinski

et

al,

1994).

Such

results

led

to

the

conclusion

that

TCD

is

the

method

of

choice

for

screening

for

PFO

because the

high

sensitivity could

spare

patients

a

TEE

exam.

Furthermore, TCD costs

less,

and

one

can

easily

monitor

effectiveness

of

the

Valsalva

by

observing decreased

cerebral blood

flow

(Klotzsch,

1994).

Kerut

et

al

(1997)

compared

the

ability

of

TTE,

TEE

and

TCD

to

detect PFOs

in

both

control

subjects

and divers referred for

neurological

DCS

(see

Table

6).

TEE

was

the most

sensitive

method

for

detecting

PFOs

in

both

controls

and divers.

However,

only

the

TCD

method

of

imaging differentiated

between divers

and

controls.

The

authors

suggest

that

the

TCD

method

only

detects clinically

significant

PFOs

since

only

strongly

positive

TEE

also

had

positive

TCDs.

5

They

calculated

the

positive

and

negative

predictive

values

for

detection

of

shunts

in

DCS

divers

for

all

3

imaging modalities.

The

PPV and

NPV

for each

respectively

was

52%

&

59%

(TEE),

62%

&

58%

(TTE),

and

65%

&

64%

(TCD).

Unfortunately

the

authours

do

not define

"clinically

relevant"

and

in

fact

when

"possible

DCS cases" were

removed

from

the

sample,

TCD

no

longer

differentiated between

DCS

and

control

groups.

Table

6:

Right

to

Left Shunting During

the

Valsalva Maneuver

(Table

1

From

Kerut

et

al,

1997)

Control

Probable

+

Subjects

Definite

DS

All

Divers

(n=

30)

(n=

15)

(n=

26)

Positive

Studies

TTE

5(17%)

3)(20%)

8

(31%)

TEE

14

(47%)

9

(60%)

15

(58%)

TD

7

(23%)

7

(47%)

13

*

(50%)

*

p

=

0.05

versus

control

DS

=

decompression

sickness;

TD

=

transcranial Doppler;

TEE

=

transesophageal

echocardiography;

TTE

=

transthoracic

echocardiography

Other

methods

such

as

carotid

duplex

monitoring,

dye

dilution,

and

oximetry have

been

tested

as

a

possible method

for PFO

detection.

Karttunun

et

al (1998)

assessed

the

value

of

dye

dilution

and

oximetry in

detecting

PFOs. They

found

concordance between

the

two

methods ,

both

of

which

detected

PFOs

in

24/59

(41%)

patients.

Unfortunately,

they did

not compare

these

methods

to

TEE,

the

gold

standard, and

so

sensitivity

and

specificity

cannot

be

determined.

Nygren &

Jogestrand

compared

TCD

of

the

MCA

and

duplex

monitoring

of

the

ICA with

TEE

and

found sensitivities

of

100%

(TCD)

and

58%

(duplex),

and

specificities

of

82%

and

91%

respectively.

The

conclusion

of

the

study

was

that

TCD

but

not

duplex

of

the

ICA

could

be used

for

PFO

screening.

Clinical

Relevance

of PFO

The

rationale

for

such

extensive

investigation

into

the

best

method

for

detecting

PFO

is,

of

course,

that

detection

is

clinically

relevant.

As

previously

mentioned,

the

fact

that

the

pressure

in

the

left

atrium

is

greater than that

in

the

right

atrium

post-natally

usually

leads

to

PFO

closure.

In

some

however,

the foramen

remains

patent.

This

is

generally

of

no

significance

since

the

higher

left

atrial

pressure

keeps

the

valve

functionally

closed.

However,

in

situations

where

the

right

atrial

pressure

becomes

significantly

higher

than

that

on

the

left,

a

gradient

reversal

can

occur,

causing

right-to-left

shunting

through

the

foramen.

Gradient

reversal

can

occur when

pulmonary vessels

are

obstructed

(e.g.

from

overload

of

venous

bubbles),

vasoconstriction

causing

increased vascular

resistance

and

subsequent

decrease

in

cardiac

output

(CO)

and

thus left

atrial

pressure,

release

6

of

Valsalva,

coughing, cessation

of

positive

pressure breathing,

negative

pressure

breathing,

restricted

breathing, or

any

other

situations

leading

to

substantial

increase

in

venous

return

to

the

right heart.

Moon

et

al

(1989)

speculate

that

the

prevalence

of

shunting

in

divers

may

be

underestimated

by

echocardiography

done

in

the lab

because

immersion

in

water might

increase

a

shunt

as

a

result

of

increased right

atrial

pressure

and

cardiac

dilation.

In some

cases,

right-to-left

shunting

has

been

shown

to

occur occasionally

during

quiet

breathing

without

complication

(Fraker

et al,

1979;

Lynch

et

al,

1984;

Smith

et

al,

1990,

Apr)

and

generally

such intermittent shunting

in a

normal

individual

may

cause

transient

decreased

oxygen saturation,

but little

else.

Of

greater

concern

is

when

such

right-to-left

shunting

causes

paradoxical

embolization

to

occur.

This

phenomenon

has

been

well-studied

among stroke

patients, particularly

among

those

who

experience

stroke

despite having

no

risk

factors,

or

cryptogenic

stroke.

PFO and

Stroke

Jones

et

al

(1994)

examined

the

prevalence

of

PFO

in 220

patients

with

cerebral ischemia,

compared

to

202

controls.

Prevalence

was

no

different

in

the

two

groups

(16%

vs

15%

respectively).

When

subdivided

by

age

groups,

prevalence

in

PFO

vs

control

groups

remained

similar

in each

of

three

age

categories,

<50,

50-69,

>70.

Similarly,

Fischer

et

al

(1995)

did

not

find

a

higher

prevalence

of

PFO

among

those

with

a

history

of

cerebro-vascular accident

(CVA).

Jones

et

al

recommended

subsequent

longitudinal

studies

in

which

a

group

of

known

PFO

patients

would

be

followed

to

assess incidence

of

stroke. In

other

words,

the

majority

of

studies

to

date

have

assessed

the

prevalence

of

PFO

in stroke

patients,

but

few

if

any

have

attempted

to

prove

an

increased

incidence

of

stroke in a

known

PFO

group

compared

to

non-PFO

controls.

de

Belder

et

al

(1992)

examined

the rates

of

PFO

in

stroke

patients with and

without

risk

factors,

and

controls.

In

this

study,

patients

with

cryptogenic

strokes

were

10

times

more likely

than

controls

to

have

PFO,

and

those

with

risk-positive

strokes

were

5

times more likely

than

controls

to

have PFOs.

However,

those

with

cryptogenic

and

risk-positive

strokes were

equally likely

to

have

PFOs.

Although

PFOs

are

more

frequent

among

patients

with

cryptogenic

than other

types

of

stoke, they

seem

to

also have

a

high

frequency

among stroke

patients

in

general

compared

to

controls. This

conclusion

was

also

put

forth

by

Chen

et

al

(1991),

and

Petty

et

al

(1997).

Similarly,

Lechat

et

al

(1989,

abstract

only) found

the

prevalence

of

PFO

to

be

higher

among

stroke

patients

(40%)

than controls

(10%).

When subdivided

by

the

cause

of

stroke

i.e.

known

cause,

known

risk

factor,

or

cryptogenic,

the

prevalence

of

PFO

rose

respectively

from

21%,

to

40%,

to 54%.

Based

on these

results,

authours

suggested

that

paradoxical emboli

through

PFOs

causing strokes

may

be

more

frequent

than

is

generally believed.

Di

Tullio

et

al

(1992)

used

multiple logistic regression

to

evaluate

the

strength

of

association

between

PFO and

cryptogenic

stroke

after

correcting

for age

and stroke

risk

factors.

They

found

that

patients

with

cryptogenic

stroke

were

7.2

times

more likely

to

have

a

PFO

than

were

those

with

a

known

cause

for

stroke,

thus

supporting

PFO

as a

risk factor

for cryptogenic

stroke.

Klotzsch

et

al,

1994

also

found

PFO

to

occur significantly

more frequently among

those

with

cryptogenic

stroke

than

with

other known

causes

of

stroke

(see

Table

7).

7

Table

7:

PFO

in

111

Patients with

Known

and

Cryptogenic

etiology

of

Cerebral

Ischemia

(Table

2

From

Klotzsch,

1994)

PFO

PFO(-)

Total

Cryptogenic

31

(77.5)

9

(22.5)

40

Large

vessel

disease

8

(26.5)

22

(73.3)

30

Small

vessel

disease

6

(30%)

14

(70%)

20

Cardioembolism

5(26.3%)

14

(73.7%)

19

Miscellaneous

-

2

Frequency

of

PFO

was

significantly

different

in

known

and

cryptogenic cerebral

ischemia,

p

<

0.001

(chi-square test).

Honmma

et

al

(1994)

studied

characteristics

of

PFOs

that

could

differentiate

between

patients

with cryptogenic

strokes

or

strokes

of

known

cause,

since

PFOs

are

known

to

be

present

in

both

types

of

stroke. They

found that

those

with cryptogenic

strokes

were more

likely

to

have

larger

PFOs

with

more

extensive

shunting,

hence

suggesting

that

the

clinical

significance

of

individual

foramina

may

be

in

part

determined

by

echocardiographically

identifiable

characteristics.

Stone

et

al

(1996)

used

contrast TEE

to

subdivide

a

group

of

34

patients

with known

PFO

into

2

groups: a large

shunt

(>20 bubbles) group and

a

small

shunt

(>3)

bubbles

group.

They

followed

the

groups

prospectively

and

found

5/16 (31%)

of

the

large

shunt

group

had

embolic

events

despite

anticoagulation,

whereas

none

of

the

small

shunt

group did

(p=.03).

These

results

indicate

an

association

between shunt

size

and

risk

of

future embolic events.

Venous

Gas

and

DCS

Knowing

that

PFOs

exist,

can

be

clinically

detected,

and

can

lead

to

strokes

if

clot

passes

paradoxically through

a

functional

right-to-left

shunt,

the

question

is

whether

of

not

one

can

now

extrapolate

to

the

decompression

situation.

As

early

as

1969,

reports existed

suggesting

that

early

neurological symptoms

after

diving

could

be

caused

by

intracardiac shunts,

and

specifically

by

PFOs

(Fryer,

1969).

In

both

diving

and

altitude,

venous

gas

bubbles

may

develop when

dissolved

gas

comes

out

of

solution

as

the

ambient

pressure decreases

during

ascent,

and

the

depressurized

gas

volume

expands

(and

is

dissipated).

The

filtration

of

bubbles

by

the lung

means

they

are

usually

asymptomatic.

However,

if

the

lungs

are

overwhelmed,

or

if

there

is

a

right

to

left

shunt

as

would

exist with

PFO (or

other

atrial-septal

defects)

then

venous bubbles

could

bypass

the

lung filter

and

directly

enter

the

arterial

circulation. Considering

that

the

prevalence

of

PFO

in

the

population

is

about

25-30%,

the

incidence

of

type

II

DCS

is

less

than

might

be

expected

given

the

known

prevalence

of

PFOs

and

the

documented

common

occurrence

of

decompression-

induced

venous

gas

bubbles.

This

may

be

because bubbles

will only

pass

from

the

right

to the

left

atria

if

the

normal pressure gradient

is

reversed.

8

Pilmanis

et al

(1996)

cite

a

study

supporting

the

view that

some

cerebral

gas

emboli

may

be

tolerable,

and may travel

back

to

the

venous

side

without

causing

obstruction.

However,

at

least

50%

of

such embolized

gas

is

thought

to

stay

on

the

arterial

side.

Although

gas

emboli

behave

differently

than

clots

in

that

they

are

not

rigid

and

so

can

conform

to

vessel

shape,

their

presence

remains

a

key

factor

in

the

explanation

of

neurological decompression

sickness.

Spencer

(1976)

found that

venous

gas

emboli

were

detectable

in 4/11

divers

(36%)

after

a

no-

decompression

(USN

tables)

18

m

chamber

dive

for

60

min.

He

also

noted

that

for

the

same

profile,

bubbles

were more likely in

open

water

rather

than

chamber

dives.

Later,

Dunford

et

al

(1988)

found

venous

bubbles

in

17%

of

a

sample

of

sport

divers

undertaking

dives

between

6

and

39

msw.

Gas

bubbles

have

been

found

in

the

venous

circulation

after

ascents from

as

shallow

as

3

m

(Eckenhoffet

al,

1990).

Eckenhoff

et

al

(1990)

studied

the

dose-response relationship

for

decompression

magnitude

and

endogenous venous

gas

bubble

formation

in

humans.

Subjects

were

exposed

to

pressure

of

12,

16,

and

20.5

fsw for

48

hrs

then returned

to

surface

in

less

than

5

minutes. There

were

no DCS

cases

but

a

large

incidence

of

venous

bubbling. Using

a

Hill

dose-response

equation,

highly

significant

fits

were obtained

and

they

concluded

that

50%

of

humans

generate

bubbles

after

decompression

from

steady

state

exposures

to

11

fsw,

implying

that

endogenous

bubbles

form

from pre-existing

gas

collections.

Despite

a

clear relationship

between

decompression

and

development

of

venous

bubbles,

some

believe

the

relationship

of

VGE

to

DCS

is

less

conclusive

(Bayne

et

al,

1985).

This

is

in contrast

to

more recent

and

extensive

work by

Ron

Nishi

at

DCIEM

(1993)

who

states

that

although

large

numbers

of

bubbles

are

not necessarily

accompanied

by

DCS, the

opposite

is

usually

true

i.e.

DCS

is

usually

accompanied

by

bubbles.

Arterial

Gas

and

DCS

In

animal

studies

using

pigs,

Vik

et

al

(1992,

1993)

investigated whether

arterial

gas

was

more

likely

when

a

PFO

was

present. Pigs

are

increasingly

being

used

in

research because

of

their

physiological

similarity

to

humans

particularly

with respect

to

the

cardiovascular

system

(Broome

et

al,

1995).

In

1992,

they

compared

the rate

of

paradoxical

embolization

in PFO

to

non-PFO

pigs

at

various

rates

of

air

infusion,

into

either

the

RA

or

RV

(in

the

PFO group). The

incidence

of

PAE tended

to

be

higher

at

all

infusion

rates

in

the

PFO

groups compared

to

controls. In

addition,

less

air

needed

to

be

infused

in the PFO

pigs

before

arterial

bubbles

were

seen,

than

in

the

non-PFO

pigs.

Finally, the

size

of

the PFO

was

found

to

be

unrelated

to the

occurrence

of

arterial

gas.

Then

in

1993,

the same

investigators

tested

the

hypothesis

that

after

rapid

decompression

pigs

with

a

PFO

would

be more

likely

than

those

without

one

to

have

arterialized

bubbles.

Of

14

pigs,

6

were

found

to

have

a

PFO

and

8

did

not.

TEE

was used

to

detect

arterial

bubbles

which

were

found

in

all

6/6

of

the

PFO

pigs, but

only

2/8

in

the

non-PFO

group

(p<.00

9).

In

addition,

venous

bubble

counts

in

the PFO

pigs

were

lower

than

in

non-PFO

pigs.

This means

that

arterial

gas

bubbles

occurred

at

lower

venous bubble

loads

in

PFO pigs,

and

that

pigs

with

a

PFO were

more

likely

to

have arterialized

gas.

9

Table

8:

Incidence

and

Time

of

Detection

of

Arterial

Gas

Bubbles

in

Pigs

with

PFO

and

in

Pigs

Without

a

PFO

(Table

I

From

Vik

et

at, 1993)

Arterial

Gas

Bubbles

Group n

Incidence Time,

a

min

PFO

6

6/6

b(

100%)

<4

', 7, 8,

10,

13,

15

Non-PFO

8

1/8

(25%)

10,

12

n

Minutes

after decompression,

bp

=

0.009

compared

to

the

incidence

in

the

non-PFO

group;

C

exact

time

for the occurrence

of

arterial

gas

bubbles

is

not

available

(see

text)

In

a

human population,

Glen

et

al

(1995)

used transcranial

Doppler

to

determine

the

incidence

of

bubbles

in the

cerebral circulation

of

divers

with

and

without

PFO

at

various

times

during

safe

decompression

from

air dives. They

found

4/17 divers with

shunts

identifiable

by

TCD,

but

none

of

the

divers

either

with

or

without

PFO

had

detectable bubbles

in

the

cerebral

circulation.

PFO

and

Divine

Interest

in

the idea

that

PFO

might

be

a

risk

for

DCS

developed rapidly

after

1986

when

Wilmshurst

et

al

published

a

case

of

Type

II

DCS

in

a

diver

with

an

ASD, and

then hypothesized

that this

resulted

from

venous

gas

passing through

the

defect.

Subsequently,

Moon

et

al

(1989)

noted

that

in1987,

122

cases

of

Type

II

DCS

occurred

in

US

sports divers,

most

of

whom

had

conformed

to

USN Tables,

and

postulated

that

PFO

could

be

a

risk

for

DCS

(see

Table

9).

They

went

on

to

examine

30

divers with

a

history

of

DCS.

These

were

subdivided

into

those

with

serious symptoms

(18/30)

and

those

with

minor

symptoms

(12/3

0,

3

of

which

were

not

type

II

DCS).

Controls

were

healthy

non-diver volunteers

from

2

other

studies

on

PFO

prevalence.

Table

9:

Relation

Between

Decompression

Sickness

and

Right-to-Left Shunting

During

Bubble

Contrast,

Two-Dimensional

Echocardiography

(From Moon

et

at,

1989)

Decompression

Sickness

(n

=

30)

Controls

*

(n

=176)

Right-to-left

shunt

t

Yes (n

20)

11

91

No

(n=

186)

1

19

167

*

Controls

from

refs

4

& 6.

t

During breathing

at

rest

:

Decompression

sickness

vs.

controls

=

25.62,

p

=

0.0001.

10

The

percentage

of

divers

with

R

to

L

shunting

was

37%,

and

of

these

11/30

who

had

shunting,

all

experienced

serious

DCS

symptoms.

The

percentage

of

cases

of

severe

neurological

DCS

with

PFO

was

therefore

11/18

or

61%.

There

were

no

cases

of

shunting

in

those with

only mild

DCS.

The

authours

concluded that

PFO

represents

a

risk

for the

development

of

DCS.

This

article

by

Moon et

al

led

to

much

discussion.

Also

in

1989,

Wilmshurst

et

al

(1989,

Apr), in a

letter

to

the

editor

of

The

Lancet

reported

their

belief

that

cardiac

shunts

are

associated

with

early

neurological symptoms, and usually

occur

in

the

context

of

"safe"

dives, but

only in

dives

which

have

produced

venous

bubbles

(thus

not

all

decompression

tables prevent

bubble formation).

In

contrast,

these

authors

proposed

that

symptoms

occurring

later after

a

dive

are

caused

by large

tissue

nitrogen

loads and

unsafe

decompression procedures.

Eight

months later,

Wilmshurst

et

al

(1989,

Dec)

followed

up

their

commentary

with

an

article

assessing

the

relation

between

shunts

and the

timing

of

neurological

symptoms

after

diving

(see

Table

10).

They

examined

61

divers with decompression

sickness

with

saline

contrast

echocardiography,

and

divided them

into

4

subgroups: Ia

(n=29)

neurological symptoms within

30

min

of

surfacing,

Ib:

(n=24)

neurological

symptoms

with onset

greater than

30

min

after

surfacing,

Ic:

(n=6)

joint

pain

only,

and

Id

(=2)

cutaneous

symptoms

only.

The control

group

was

63

divers

with

no

history

of

DCS.

The

prevalence

of

shunting

was

significantly

higher

in group Ia

than

in

controls

or in

group

1b.

Of

those

with

rapid

onset

neurological

symptoms,

66%

were

found

to

have PFOs.

The

prevalence

of

PFO

in

the

control

group

was

24%,

similar

to

that

in

the general

population.

Risk

factors

the

dive were

significantly

less

prevalent

in

group Ia

than

group

Ib.

Thus,

Wilmshurst

et

al

concluded that those

with

shunts

represent

a

high

proportion

of

cases

of

early

neurologic

DCS

and

they

also

constitute

a

majority

of

cases

in

which

DCS

is

not

explained

by

the

dive

profile.

Table

10:

Prevalence

of

Interatrial

Shunt

in

the

Groups

of

Divers

(Table

I

From

Wilmshurst

et

al,

1989,

Dec)

Group

Ia

Ib

Ic

Id

Group

II

No.

of

divers

29

24

6

2

63

No.

with shunt

19

4

1 1 15

Shunt

on

Valsalva

only

9

0

1

0

7

%

with

Shunt

66*

17 17

50 24

*

Difference

from

group

II,

p<0.001;

difference from

group

Ib, p<0.001.

One

further

conclusion,

that

led

to

several letters

to the

editor

of

The

Lancet

by

Smith

et

al

(1990,

April,

June),

was

that

while

the

cause

of

hemiparesis

is

generally

accepted

to

be

cerebral

gas

embolism,

paraparesis

may

not

be

caused

by

autologous

bubble

formation

in the spinal cord

as

previously

believed,

but

rather

by

arterial

gas

bubbles

from

a

shunt

or

pulmonary barotrauma.

11

Smith

et

al

(1990

Apr,

June)

refuted

the

conclusion

that spinal

DCS

could

be

due

to

arterialized

gas

on

the

basis

of

support

for

the

autochthonous

hypothesis

by

animal

research

and

by

the

fact

that

Wilmshurst

does not

explain

histologically

how

intravascular

arterial

gas

could

be

found

in

myelin. Smith

concludes

that

a

statistical

association

between

PFO

and

neurological

DCS

is

not

proof

of

the

mechanism

that

causes

DCS.

Furthermore,

Smith

et

al

(1990,

Apr,

Oct)

raised

concerns about

the

methodology

used

in

Wilmshurst's

study. The

questions

of

adequate

blinding

of

echocardiographers,

selection

bias,

and

variation

in

methodology

from

that

of

Moon's

1989

study

were

raised.

Wilmshurst's

responses

refuted these

claims

and

notes

that

Moon's

study

was,

in

fact,

neither

blinded

nor

controlled.

Cross et

al

(1990,

Sept),

report

an

uncontrolled

series

of

19

patients referred

for

DCS

who

were

routinely screened for

PFO

with

contrast

echo.

They found

that

in

their

sample

32%

had

PFOs,

and

50%

of

neurological

DCS

cases

had

shunts.

These

results

were

significantly

different

from

Wilmshurst's

(1989,

Dec)

66%

of

cases

of

early

neurological

DCS

with

shunts

(chi

<.05).

They

concluded

that

a

shunt

does

not

predispose

divers

to

neurological

DCS.

These

results

were

subsequently

severely criticized

by

Wilmshurst

(1990)

on

the

basis

of

comparison

of

Cross'

pooled

group

to

a

selected

subgroup

of

Wilmshurst's

(1989

Dec).

In

1992,

Cross

et

al

(1992,

BMJ)

examined

PFOs

among divers

with

no

history

of

DCS.

They

found

that

31%

of

their

sample

of

78

divers had PFO

and

concluded

that shunts

in

those

without

a

history

of

DCS

may

be

irrelevant.

Again, Wilmshurst

(1992)

had

an

opportunity

to

be

critical

of

Cross

et

al,

when

he

pointed

out

that

it

is

to

be

expected that

a

number

of

divers

without

DCS

would have shunts

because

of

the

frequency

of

shunt

occurrence

in

the

general

population

and

presumed

lack

of

effect

this

would

have

on

diver

recruitment.

They

liken

Cross'

argument

to

one

stating

that

a

finding

of

a

given

number

of

stroke

patients

without

hypertension

indicates

that

hypertension

is

not

a

risk

factor

for

stroke; clearly

an

illogical

statement.

Also

in

1992

(Sept),

in

Switzerland Cross

et

al

present

a

larger

data set

in

which

the

prevalence

of

PFO

as

determined

by

contrast TTE

in

neurological

DCS

cases

(49%)

is

compared

to

no-DCS

cases (32.7%)

and

non-divers

(39.3%).

There

were

no

significant differences

among

these

groups

(see

Table

11).

They

also

found

that

the

number

of

dives

undertaken

by

those

with

multiply treated

neurological

DCS

was

higher

then

in

those

with

only

one

DCS

episode.

This

finding

is

not

surprising

though, because

it

makes

sense

that

as

frequency

of

diving increases,

so

does

the

chance

for

DCS.

Table

11:

Prevalence

of

Right-to-Left

Shunt

in

Divers and

Controls

(Table

I

From

Cross

et

al,

1992)

Number

of

Number

With

Subjects

Shunt

Neuro

DCS

51

25

(49.0%)

No

DCS

98 32

(32.7%)

Non-divers

28

11

(39.3%)

DCS

=

decompression

sickness;

Neuro

=

neurological

12

Rather than

assessing

for

PFO

in

divers

with

known

DCS,

Wilmshurst

et

al

(1994)

assessed

DCS

among

a

group

of

divers with

known

PFO.

The

goal

was

to

determine

the

relation

between

PFO

and

any

significant

arterial

desaturation,

heart

rate

and

blood

pressure

responses

during

physiological

maneuvers

such

as

exercise

and

passive tilt. Their

study

involved

three

groups:

PFO

divers

with

type

II

DCS,

PFO

divers

with

no

DCS,

and

age

and

sex-matched

control

divers.

They found

no

significant

differences between these

groups

on

the

above-mentioned

measures.

Because

2

divers

in group

1

with

the

most

frequent

DCS

episodes

developed

substantial

desaturation

during

exercise

they

concluded

that

a large

PFO

might

be

associated with

clinically

relevant desaturation,

although

this

hypothesis

was

not

actually

tested.

In

an

Offshore

Technology

Report

written

by

Shields

et

al

(1996),

the

prevalence

of

PFO

in DCS

divers

(41.4%)

was compared

with that

of

non-DCS

divers

(18.5%)

and

non-diving

controls

(22.2%). No

significant

differences

were

found

among

these

groups

(see

Table

12).

The

authours

note

however

that

lack

of

significant difference

may

be

due

to

the small

number

of

subjects

in

the

non-diving

controls.

Table

12:

Distribution

of

Occurrence

of

PFO

(Table

25

From Shields

et

al,

1996)

Group

A

Group

B

Group

C

I

Total

No

PFO

17

22

7

46

(58.6%)

(81.5%) (77.8%)

(70.8%)

PFO

12

5 2 19

(41.4%)

*18.5%)

(22.2%)

(29.2%)

Note:

In

addition

to

the

3

subjects

that

did

not

take

part

in

this

test,

in

one case

the

procedure could

not

be

carried

out

and

in

an

additional

3

cases,

the

outcome

could

not

be

visualized.

More

recently,

in

a

well-designed

study

by

Germonpre

et

al

(1998)

contrast

TEE was

used

to

compare

the

prevalence

of

PFO

in

subjects with

neurological

DCS

to

a

matched

population

of

control

divers

without

DCS

(see

Table

13).

They

also

examined

PFO

in

relation

to

spinal

and

cerebral

DCS.

The

prevalence

of

PFO

in

DCS divers

was

59.5%

compared

with

36.1%

in

matched

controls,

but

this

difference did

not

quite reach

significance.

However

when subgroup

analysis

was

performed,

PFO

was

significantly

correlated with

cerebral

but

not

spinal

DCS.

When

they examined

divers

with

an

unexplained

DCS episode,

significantly

more cerebral

(9/12)

than

spinal DCS

cases

(4/14)

had

>20

bubbles

passing through

the

PFO.

13

Table

13:

Prevalence

of

PFO

(Table

I

From

Germonpre

et

al,

1998)

Number

of

Number

of

Divers

Divers With

With PFO

Grade

2

PFO

All

types

ofDCS

(n

= 37) 22

(59.5)

19

(51.3)

All

control

(n = 36)

13

(36.1)

9

(25)

P

0.06 0.03

Cerebral

DCS

(n

=

20)

16

(80)

14

(70)

Matched control

(n =

20)

5

(25)

3

(15)

P

0.012

0.002

Spinal DCS

(n=

17) 6

35.2)

5

(29.4)

Matched

control

(n =

16)

8

(50)

6

(37.5)

P

0.49

0.29

The

authors

conclude that

because

all

known confounding

factors

have

either been matched,

or

have

shown

no

significant

difference

between

groups,

the

correlation between

PFO

and

cerebral

but not spinal

DCS lends support

to

the

hypothesis that

"PFO

is

a

cause

of

DCS with

cerebral

localization".

Bove

(1998)

performed

a

metaanalysis

of

studies previously

published

by

Wilmshurst

et

al

(1989,

Dec),

Moon

et

al

(1991),

and

Cross

et

al (1992,

BMJ).

All

three

of

the

studies

were

used

to

calculate

risk

for

all

DCS,

but

for

type

II

DCS

specifically,

only

data from

2

studies

was

used

(see

Table

14).

In

both

the

"all

DCS" and

"type

II

DCS"

analyses,

the

odds

ratio

was

significantly greater

than

1.

The

presence

of

PFO

increases

the

risk

of

DCS

in

divers

with

PFO

by

1.93

times compared

to

divers

without

PFO.

For

type

II

DCS

the

risk

is

2.52

times

higher

in

those

with

PFO.

Table

14:

Calculated

Probabilities

of

DCS

witth

PFO Using

Bayes'

Theorema

(Table

4

From

Bove,

1998)

All

DCS

Type

II

DCS

P

(DCS

+/PFO+)

0.00053 0.00047

P

(DCS+/PFO-)

0.00028

0.00019

Odds

ratio

1.93

22.52

P

value

<0.001

'Odds

ratio

and

P

values

are

derived from

logistic

regression

calculations.

14

Absolute

Risk

of

DCS

in

Divine

Cross

et

al

(1994)

eventually acknowledge

that

risk

of

DCS

may

be

increased

by

a

shunt,

but

then

argue

that

this increase

is

small.

They note

that

of

approximately

50,000

divers

in

Britain

15,000

(30%)

might

be

expected

to have

a

PFO.

Noting

that

the

number

of

neurological

DCS

cases

per

year

is

about

100,

and

that

not

all

shunts

will

invariably

result

in

DCS, the

risk

of

DCS

from

a

shunt

in

the total

diving

population

is

quite

low.

Bove

(1998)

calculated

the

combined

frequency

of

type

II

DCS

among

military, sport,

and

commercial

divers

to

be

2.28

per

10,000

dives. An

increase

of

2.52

times

this

frequency

would

lead

to

an

absolute

number

of

5.7

per

10,000

cases

of

type

II

DCS

among

divers with

PFOs.

He

concludes

that

despite

a

2.5

times

greater

risk

of

type

II

DCS

in

the

presence

of

PFO,

the

absolute

risk

is

small

enough that

there

is

no

basis

for

recommendations

against diving

in

those

with

PFO,

and

that

screening

is

not

warranted.

DCS

and Altitude

Because

altitude decompression

is

analagous

to

decompression

from

saturation,

it

is

thought

that

more

venous

bubbling

occurs

in

altitude

than with

subsaturation decompression

in

diving,

the

result

being

a

greater

likelihood

of

paradoxical cross-over.

One

might

therefore expect

to

find

more cerebral

symptoms

among

altitude

rather

than

diving

decompressions, and

this

indeed

has

been

the

case

(Garrett,

1990).

Although

the

mechanisms

for

development

of

DCS,

as

well

as

the

proposed

pathophysiology

for

arterialization

are

similar

in

diving

and

flying, much

less

research

on

the

phenomenon

of

paradoxical

gas

embolism

has

been

reported

in

the

altitude

literature.

Overall,

there

has

been

some

suspicion that

DCS

symptoms

in

altitude

situations

tend

to

underreported

to

a

greater

extent

than

they

do

in

diving

due

the

perceived negative

career-related

consequences.

At

the

1991

meeting

of

the

Aerospace

Medical

Society

in

Cincinnati,

Clarke

&

Hayes

presented

their

examination

of

the prevalence

of

PFO

among

24

cases

of

Type

II

altitude

DCS

in

naval

aviation

personnel.

They

identified

4

cases

(16%)

of

PFO

by

contrast

TTE.

They

used

Moon's

1989

control data

to

conclude that

there

was

no

significant

relationship

between

PFO

and

type

II

altitude

DCS.

Powell et

al

(1995)

report

a

single case

of

a

research

subject participating

in

NASA

hypobaric

decompression testing

who

was

found

to

have

a

PFO

and

was

presumed

to

be

at

risk

for

DCS.

At

ground

level,

TTE clearly

demonstrated

left-sided

cardiac

bubbles.

During

3

hours

of

hypobaric

decompression

to

21,000

ft

the

MCA

was

monitored

for

the

appearance

of

arterialized

bubbles; none

were

identified,

nor

were there

any

DCS

symptoms,

despite the

presence

of

grade

IV

precordial bubbles.

In this

case,

the

saline

contrast

bubbles

at

ground level

were

clearly

arterialized,

but

the

decompression bubbles

were

not.

The

authors

propose

that

perhaps

the

decompression-induced

bubbles

load

was

not

substantial

enough

to

cause

flow

reversal.

They

suggest

that

the

presumption

of

an

increased risk

of

DCS

among

those who

screen

positive

for

PFO

is

one

to

be

made

with difficulty.

15

Pilmanis

et

al

(1996)

present

the

first

documented right-to-left

shunting

of

venous bubbles

after

exposure

to

altitude.

Retrospectively

examining

a

database containing

1500

subject-flights to

altitudes

ranging

from

15,000

to

35,0000

feet and

exposure

times

to

8

hours, they

identified

6

subjects who

demonstrated

left

ventricular

gas

emboli.

Five

subjects became

symptomatic

at

the

time

of

embolization

(with

joint

pain

or skin

mottling),

but

no

cerebral

symptoms

were

reported.

Of

the

3

cases

investigated

with

TEE,

PFOs

were

found

in

2

cases,

and

not

in

1,

despite

known

embolization.

This

suggests

that

more

than

one

mechanism

is

involved.

In

light

of

the

fact

that

in

all

cases

the

venous

gas

score

was

high

at

the

time

of

embolization, overload

of

pulmonary

filtration

is

the

second

suspected

mechanism.

Overall

the

conclusion

was

that

situations

which

expose

subjects

to

altitudes

which

result

in

high

venous

bubble

loads

should

be

avoided.

Webb,

Pilmanis,

and

O'Connor

(1998)

went

on

to

determine

the

altitudes

at

which high

venous

gas

loading

occurs.

One

hundred

and

twenty four subjects

were

exposed

to

simulated

altitudes

ranging from

11,500

to

25,000

feet

for

4

to

8

hours,

and

were

monitored

for

DCS

and for venous

bubbling. Venous

bubbles

were

first

seen

at

15,000 ft

and were

present

in

70%

of

cases

above

22,500

ft.

In

terms

of

DCS

symptoms,

the

5% threshold

for symptoms

was

20,500

ft

with

an

abrupt increase

in

symptoms

beyond 21,200

ft.

These results

led the

authors

to

recommend

reconsideration

of

current

altitude

exposure

guidelines.

Screening

for

PFO

James

(1990)

cites

several studies

indicating that

nervous

system damage

can

occur

without

neurological

signs. This,

in

light

of

the

28

cases

of

known transient

neurological

symptoms

that

have

occurred

after

41,0000

contrast

echo

studies,

led

him

to

argue

against

the

use

of

contrast

echocardiography

in

screening

divers.

On

the

contrary, Knauth

et

al

(1997)

noted that

the

prevalence

of

PFO

in

the

general

population

was

roughly

equal

to

the

percentage

of

divers

found

to

have

multiple

brain

lesions

on

magnetic

resonance imaging

(MRI)

in

a

by

Reul

et

al

(1995).

Knauth

et

al

then

postulated

that

divers

with

multiple

brain

lesions

may have

PFO

and

that

arterialization

of

bubbles

may

be

the

cause.

They

examined

87

divers

without

a

history

of

cerebral

disease

or

DCS

using

TCD

and

found

25

to

have

PFO,

which

they

considered

to

be

hemodynamically

highly relevant

(>20

bubbles)

in

13

cases.

The

prevalence

of

multiple

brain

lesions

was

significantly

higher

among

divers

with

PFO

than

those

without.

Among those

with

PFOs,

the

prevalence

of

multiple

brain

lesions

on

MRI

was

highest

among

those

with

hemodynamically

relevant PFOs.

Considering

that

none

of

these

divers

had

a

history

of

DCS

the

authors

suggest

that such

lesions

are

a

consequence

of

subclinical

cerebral

gas

embolism.

When

this

work

was

presented

at

the

American

Academy

of

Neurology

annual

meeting

in 1998,

Ries

(co-author

with

Knauth)

advocated that

the

$325

cost

of

one-time

screening

is

reasonable

when

compared

to

the

cost

of

diving

equipment

(Jeffrey,

1998).

Murrison

et

al

(1995)

compared

the

EEGs

of

divers

with

type

II

DCS

to

non-diver

controls

and

found

them

to

be

indistinguishable,

which raises

the

question

of

the

functional

significance

of

the

earlier

described brain

lesions.

Cross

et

al

(1990,

Dec)

argue

that

in

most

cases

a

contrast

study

is

not

performed

in

isolation

of

a

non-contrast

echo study,

such

that

the

latter would

not

be

performed

if

an

obvious

shunt

was

detected

earlier.

They

also

note

that

at

the

time

of

screening,

tissues

are

not

nitrogen

loaded,

so

16

arterialized bubbles

would

not

be

expected

to

expand.

Finally,

they

compare

the

morbidity

rate

of

contrast

echo

(.07%)

to

that

of

a

well-accepted

screening

test,

exercise

stress

testing,

which

has

a

complication

rate

of

.09%

The

implications

of

Bove's

conclusions

are

that

divers

need

not

be

screened

prior

to

initiation

of

diving, and

that

those

who

already

know

(for

some

other

reason)

that

they

have

a

PFO

can

still

go

ahead

and

dive.

But

what

about

the

situation

in

which

DCS

has

already

occurred.

Should

one

then

consider

evaluation

for

a

shunt?

Wilmshurst

(1998,

pers

comm)

recommends three

ways

to

reduce

the

risk

of

recurrent

DCS

in

a

diver

with known

PFO;

1.

stop diving,

2.

modify

diving

to

either

stay

above

15m

depth or,

for

depths

greater

than

15m

use

nitrox

or decompress

with DCIEM

tables,

or

3.

close the

PFO

preferably

by

transcutaneous

transvenous

methods

which

do not

risk

lung

injury

as

open

methods

would.

In

Britain

the

HSE have

required

applicants for professional

diving

with

PFO

to have

transcatheter

closure

(Wilmshurst,

1998

pers

comm). Likewise

in

the

Allied

Guide

to

Diving

Medical

Disorders

published

by

NATO

(1997)

states that

"significant

right-to-left

shunts

are

incompatible

with

diving

unless

surgically corrected".

Management of

PFO

In

the

stroke

literature,

several

methods

of

prevention

of

recurrent

stroke

have

been

used

for

patients

with

PFOs, these

being

antiplatelet medications,

anticoagulants,

transcatheter

closure,

and

surgery.

Mas

(1996)

argues

that closure

is

the

best

option

in cases

of

known

paradoxical

embolism,

which

are rare

and

require

visualization

of

thrombus

straddling

the

PFO.

In presumed

cases

however,

the

best

treatment

is

arguably controversial

and

requires

further

risk/benefit

analysis

to

prevent

exposure

to

unnecessary

treatment

complications.

Nendaz

et

al

(1998)

considered risk

of

recurrence

of

neurological events,

complications,

quality-

adjusted

life

years,

and

death after

5

years

in

their

decision

analysis

model

assessing

PFO

closure

methods.

They

determined

that

if

the

risk

of

recurrence

was

.8

to

7%

per

year, defect closure

was

the

best

management

strategy.

At

risk

levels

of

.8%

and

1.4%

per year,

anticoagulation

and

antithrombotic

therapies

were

better

than

therapeutic

abstention.

If

however,

the risk

of

recurrence

was

low

(i.e.

less

than .8%

per

year)

then

the

best management

option

was

no

treatment.

They

found that

the key

considerations

influencing

choice

of

therapy

aside

from

estimated recurrence

risk

included

bleeding rates,

age,

and

surgery-related

case

fatality

rates.

Several

studies have

assessed open

surgery

as

a

method

of

closure. Giroud

et

al

(1998)

studied

8

stroke

patients

and

found

no

surgical

complications,

no

recurrence

of

neurological

events,

and

no

residual shunting

after

PFO

closure

without post-op

anticoagulation.

Ruchat

et

al

(1997)

also

found

no

post-op

complication

among

32

patients, although

residual

shunts were

present

in

2/32

cases.

Homma

et

al (1997)

followed

28

patients

with

a

history

of

cryptogenic stroke

and

who

underwent

surgical

PFO closure

and

found recurrence

rate

for

neurological events

of

19.5%

overall.

This

rate

was variable

when

age

was

considered

and

proportional

hazard

regression

17

analysis revealed

an

increase

in

relative risk

of

recurrence

of

2.76

per

10

years

of

age.

They

concluded

that although

surgical

closure

is

easy

to

perform,

it

does

not

guarantee

prevention

of

recurrence.

Non-operative

closure

of

atrial

septal

defects

have been

reported

since

1976

(Formigari

et

al,

1998;

King

et

al,

1976).

Closure

by

transcatheter

methods

remains

impossible

for

some

defects

especially those

greater

than

25

mm

in

size.

In

addition

there

are

relative

contraindications

for

closure,

particularly

morphological constraints.

Wilmshurst

et

al

(1996)

write about

2

cases

of

PFO

in

divers

with

neurological

DCS

who

were

successfully

treated

with

an

inverted

adjustable

button

device,

one

with

no

residual

and

the other

with

a

tiny

residual shunt.

Both

divers

returned

to

diving.

There

is

no

mention

of

whether

either

diver

experienced

repeated

DCS

post-procedure. Johnston

et al

(1996)

believe that

wider

application

of

invasive

shunt closure

methods should

not

occur

before

the

relation between

PFO

and DCS

is

further

delineated,

noting that

one

must

consider

the

shunt

size

and

not

just

patency

in DCS

risk

evaluation.

Ende

et

al

(1996)

report

on

their

experience with

10

adults

who

had

ASDs or PFOs

closed

with

button

devices.

Aspirin

(5-10mg/kg/day)

or

Coumadin

was

administered for

6-12

weeks

post-

procedure,

or

until

the

shunt

was

completely

closed.

Closure

was

complete

in

78%

of

cases

at

6

months

and

100%

of

cases

by

1

year.

There

were

complications

in

3

cases.

In

one

case

the

device

slipped repeatedly

across

the

septum

into

the

left

atrium

necessitating

standard

surgical

repair.

In

a

second

case,

the

patient

experienced

palpitations

and

orthostatic

lightheadedness,

thought

to

be

due

to

mechanical irritation

and

required B-blockade.

A

third

case

developed

what

was

presumed

to

be

a

left atrial

thrombus

after

23

months

of

follow-up

and

had

to

be

recoumadinized.

There

were

no

subsequent neurological

events

at

an

average

of

32

months

follow-up.

The later

intermediate-term,

phase

1

FDA trials for buttoned

devices

concluded

that

after

5.5

years

of

follow-up,

98%

of

cases

had

effective

ASD

closure.

Residual shunting

remained

in

27%

of

cases after

-60

months.

Residual

defects

were

significant enough

to

require

further intervention

in

4%

of

cases.

Justo

et

al

(1996)

reviewed

the

effectiveness

of

ASD closure

in

45

children

using

the

Clamshell

double

umbrella

device.

Device

placement

was

optimal

in

43

(96%)

patients.

Closure

was

complete

in

only

23+/-14

%

of

cases

by

6

months,

and

complete

in

-64%

by

4

years.

Complications

necessitating

the

surgical closure

of

the

ASD

in

two

cases

were

due

to

device

embolization

to

the

right

pulmonary

artery

in

one case,

and

malposition

in

the

septum

with

significant

residual shunting

in

the

second. Other complications

included

pulmonary edema

in

one

case,

and

transient

loss

of

femoral

pulse

(resolved

with heparin)

in

another.

The

most

concerning drawback

however,

was the

prevalence

of

device

arm

fracture

of

71%

(+/- 21%)

at

4

years

which

led

to

withdrawal

of

the

device

from clinical trials.

The device has

subsequently

been

redesigned.

The

redesigned

version

(CardioSEAL)

was

evaluated

by

Kaulitz

et

al

(1998)

in

the

context

of

7

cases

of

morphologically

variant

ASDs.

The only

complication

was

in

one

patient

who

experienced

non-sustained

SVT;

otherwise

there

were

no

cases

of

device

embolization,

device

fractures,

thromboembolism,

or

pericardial

effusion.

Residual

shunting

was

trivial

in

3

cases

and

mild

in

1

case.

18

Formigari

et al

(1998)

report

on

the

techniques

and

results

of

28

ASD

closures

(in

children)

using

three

different

percutaneous

devices,

these

being

the

Sideris

"Buttoned

Device",

the

Das

"Angel

Wings",

and

the

"Amplatzer".

For

all

groups, fluoroscopy

times

were

similar,

but

procedure time

was

shortest

for

the

Amplatzer

and

longest

for the

buttoned

device.

Definitive

closure

occurred

in all

cases

except

1

buttoned

device.

Follow-up

times

were

longest

for

buttoned

devices

at

40+/-

2

months,

compared with

27+/-

2

mo.

for

the

Angel

Wings,

and

5+13

mo

for

the

Amplatzer.

In

terms

of

complications,

there

were

2

cases

of

transient

myocardial

ischemia secondary

to

coronary

air

embolism

in

the

buttoned

devices,

and

1

case

of

pericardial

tamponade

with

the

Angel

Wings.

Others have

reported

failures

with

this

device

also

requiring

emergency

surgical intervention

(Agarwal

et

al,

1996).

There

had

been

no

complications

with

the

Amplatzer

device.

Cost,

according

to

these

authours

was least

expensive for

the

buttoned

devices. Overall

they

concluded

that

the

Amplatzer

device

is

preferable.

Wilmshurst

has

subsequently

stated

(1998,

pers

comm) that

PFOs

in

most individuals

are

approximately

1-2

mm in

size

compared

to

PFOs

of

10

mm

or

greater

among those

who

get

DCS.

He

has

moved

from

a

buttoned

device

to

using

the Amplatz septal

occluder,

which

he

considers

to

be

the

best

device

currently

available.

He

prescribes

low

dose

ASA

for

6

months

post-procedure

until endothelialization

occurs.

The

most recent

results

of

the

World Study

on

closure

with

the

Amplatzer indicate

that

a

total

of

936

ASDs have

been

closed

as

well

as

86

PFOs.

Closure

rates for

PFOs

are

good,

with

100%

being

closed

at

24

hours,

compared

with

100%

at

1

year for

the

ASD cases

(98.9%

at

1

month).

There

were

24

complications

among

-1000

patients,

the

majority

of

which included

device

embolization

(9/24),

TIAlembolization

(4/24),

and arrhythmia

(3/24).

Limitations

There

are

several

factors

that

limit

the

generalizability

and hence

the

conclusions

that

can

be

drawn

from

the

studies

performed

to

date.

These

include

variation

in

study

groups

used

(i.e.

sport,

commercial, or

military

divers),

variation in

control

groups

used

(ie.

matched

vs.

unmatched, diver

vs.

non-diver),

differing

techniques

for

PFO detection

(ie.

TTE

vs.

TEE),

and

variability in

definition

of

DCS

or

severity

of

cases

selected

to

be

members

of

the study

group.

CONCLUSIONS

Nonetheless,

several conclusions

can

be

tentatively

drawn

on

the

basis

of

available

research:

1.

For

detection

of

PFO,

TCD

is

probably

adequate, but

contrast

TEE

is

the

gold standard

and

remains

more

commonly used.

2.

There

seems

to

be

a

relationship between

crypogenic

stroke and

the

presence

of

PFO,

as

well

as

the

size

of

the

PFO.

19

3.

Animal studies

show increased

arterial

bubbles

at

lower

venous

bubble

loads

in

pigs

with

PFO

than

in

those without.

4.

The

weight

of

evidence favours

an

association

between

diving DCS

and PFO.

This

association

remains

less

clear

in

the

case

of

altitude

DCS,

with

fewer

studies

available

on

this topic. PFO

increases

the

relative

risk

for

type

II

DCS

but

the

absolute

risk

remains

low.

5.

With altitude,

high

bubbles loads

may

favour

pulmonary

overload

as

a

mechanism

for

embolization.

6.

The

issues

of

screening remains

controversial,

although

the

absolute

increase

in

risk

of

DCS

as

a

result

of

PFO

seems small.

7.

Should closure

be

chosen

for

management,

the

transvenous

Amplatzer

appears

to

be

the

best

available option

at

this

time.

REFERENCES

1.

Agarwal,

S.K.,

Ghosh,

P.K.,

&

Mittal,

P.K.

Failure

of

devices

used

for closure

of

atrial

septal defects:

mechanisms

and

management.

J

Thorac

Cardiovasc Surg

1996;

112(1):

21-6.

2.

Bayne,

C.G.,

Hunt,

W.S.,

Johanson,

D.C.,

Flynn,

E.T.,

&

Weathersby,

P.K.

Doppler

bubble

detection

and

decompression

sickness:

a

prospective

clinical

trial.

Undersea Biomed

Res

1985;

12:

327-32.

3.

Belkin,

R.N.,

Pollack,

B.D.,

Ruggiero,

M.L.,

Alas,

L.L.,

&

Tatani,

U.

Comparison

of

transesophageal

and transthoracic

echocardiography

with contrast

and

flow

Doppler

in the

detection

of

patent

foramen

ovale.

Am Heart

J

1994; 128(3):

520-5.

4.

Bove,

A.A.

Risk

of

decompression

sickness with patent

foramen

ovale.

Undersea

Hyperb

Med

1998;

25(3):

175-78.

5.

Broome,

J.R.,

Dutka,

A.J.,

&

McNamee,

G.A.

Exercise

conditioning reduces

the

risk

of

neurologic decompression

illness

in

swine.

Undersea

Hyperb

Med

1995;

22(1): 73-85.

6.

Chen,

W.J.,

Lin,

S.L.,

Cheng,

J.J.,

&

Lien

W.P.

The frequency

of

patent

foramen

ovale

in

patients

with

ischemic

stroke:

a

transesophageal

echocardiographic

study.

J

Formos

Med

Assoc

1991;

90(8):

744-8.

7.

Chezbraun,

A.,

Pinto,

F.J.,

&

Schnittger,

I.

Biplane

transesophageal

echocardiography

in the

diagnosis

of

patent

foramen

ovale.

J

AM

Soc

Echocardiogr

1993;

6(4):

417-21.

8.

Clarke,

J.B.,

&

Hayes,

G.B.

Patent foramen

ovale

and

type

II

altitude

decompression

sickness.

(Abstract).

Aviat

Space

Environ

Med

1991;

62:

445.

9.

Cross,

S.J.,

Evans,

S.A.,

Thomson,

L.F.,

Lee, H.S.,

Jennings,

K.P.,

&

Shields,

T.G.

Right

to

left

shunts

in

neurological

decompression sickness.

XVIII

Annual

Meeting

of

EUBS,

Sept

1992,

Basel

Switzerland;

pp.

19-21.

20

10.

Cross,

S.J.,

Evans,

S.A.,

Thomson,

L.F.,

Lee,

H.S.,

Jennings,

K.P.,

&

Shields,

T.G.

Safety

of

subaqua diving with

a

patent

foramen

ovale. BMJ

1992;

304(6825):

48

1-2.

11.

Cross,

S.J.,

Jennings,

K.,

& Thomson,

L.

Decompression

sickness.

Role

of

patent

foramen

ovale

is

limited.

BMJ

1994;

309(6956): 743-4.

12.Cross,

S.J.,

Thomson,

L.F.,

Jennings,

K.P.,

&

Shields, T.G.

Right-to-left

shunt

and

neurological

decompression sickness

in divers.

Lancet

1990

Sept

1;

336(8714):

568.

13.

Cross,

S.J.,

Thomson,

L.F.,

Lee, H.S.,

Shields,

T.G.,

&

Jennings,

K.P.

Safety

of

contrast

echocardiography

in screening

divers. Lancet

1990

Dec

22-29;

336(8730):

1595-6.

14.

de

Belder,

M.A.,

Tourikis,

L.,

Leech,

G.,

&

Camm,

A.J.

Risk

of

patent foramen

ovale

for

thromboembolic

events

in

all

age

groups.

Am

J

Cardiol

1992;

69(16): 1316-20.

15.

Dubourg,

0.,

Bourdarias,

J.P.,

Farcot,

J.C.,

Gueret,

P.,

Terdjman,

M.,

Ferrier,

A.,

Rigaud,

M.,

&

Bardet,

J.C.

Contrast

echocardiographic visualization

of

cough-induced

right

to

left

shunting

through

a

patent

foramen

ovale.

J

AM

Coll

Cardiol

1984;

4(3):

587-94.

16.

Dunford,

R.,

Waccholz,

C.J.,

Irwin,

J.,

Mitchell,

P.R.,

&

Bennett,

P.B.

Ultrasonic

Doppler

bubble

incidence

following sport dives.

Undersea

Biomed

Res

1988;

15(supp):

45-46.

17.

Di

Tullio,

M.,

Sacco,

R.L.,

Gopal,

A.,

Mohr,

J.P.,

&

Homma,

S.

Patent

foramen

ovale

as

a

risk

factor for

cryptogenic

stroke.

Ann Intern

Med

1992;

117(6):

461-5.

18.

Di

Tullio,

M.,

Sacco,

R.L.,

Venketasubramanian,

N.,

Sherman,

D.,

Mohr,

J.P.,

&

Homma,

S.

Comparison

of

diagnostic

techniques for

the

detection

of

a

patent

foramen

ovale

in

stroke

patients.

Stroke

1993;

24:

1020-1024.

19.Eckenhoff,

R.G.,

Olstad,

C.S.,

&

Carrod,

G.

Human

dose-response

relationship

for

decompression

and

endogenous

bubble formation.

J

Appl

Physiol

1990;

69(3): 914-18.

20.

Ende,

D.J.,

Chopra,

P.S.,

&

Rao,

P.S.

Transcatheter

closure

of

atrial

septal

defect

or patent

foramen

ovale

with

the

buttoned

device

for prevention

of

recurrence

of

paradoxic embolism.

Am

J

Cardiol

1996;

78(2):

233-6.

21.

Fairbourne,

E.

Closure

with

the

Amplatzer

septal

occluder:

world

study.

Pers

comm

to

Dr.

K.P.

Walsh,

Oct.1998.

22.

Formigari,

R.,

Santoro,

G.,

Rossetti,

L.,

Rinelli,

G.,

Guccione,

P.,

Ballerini,

L.

Comparison

of

three

different

atrial

septal

defect occlusion

devices.

Am

J

Cardiol

1998;

82(5):

690-2,

A9.

23.Fraker,

T.D.

Jr.,

Harris,

P.J.,

Behar,

V.S.,

& Kisslo,

J.A.

Detection

and

exclusion

of

interatrial

shunts

by

two-dimensional echocardiography

and

peripheral

venous injection.

Circulation

1979;

59(2):

380-384.

24.

Fryer,

D.I. Sub-atmospheric

decompression

sickness

in

man.

Slough:

Technivision,

1969;

209-10.

25.

Garrett,

J.L.

The

role

of

patent

foramen

ovale

in

altitude-induced

decompression

sickness.

In:

Pilmanis,

Ed.

Hypobaric

decompression

sickness: proceedings

of

a

workshop

held

at

Armstrong

Laboratory,

Brooks

AFB,

TX,

16-18

Oct

1990.

Alexandria,

VA:

Aerospace

Medical Association

&

Undersea

Hyperbaric

Medical

Society,

1995:

81-96.

21

26.Germonpre,

P.,

Dendale,

P.,

Unger,

P.,

&

Balestra,

C.

Patent

foramen

ovale

and

decompression

sickness

in

sports

divers.

J

Appl

Physiol

1998;

84(5): 1622-6.

27.

Glen,

S.K.,

Georgiadis,

D.,

Grosset,

D.G.,

Douglas,

J.D.,

&

Lees,

K.R.

Transcranial

Doppler

ultrasound

in

commercial

air

divers:

a

field

study including

cases

with

right-to-left

shunting.

Undersea

Hyperb

Med

1995;

22(2):

129-35.

28.

Giroud,

M.,

Tatou,

E.,

Steinmetz,

E.

Lemesle,

M.,

Cottin,

Y.,

Wolf,

J.E.,

Moreau,

T.,

&

David,

M.

The

interest

of

surgical

closure

of

patent

foramen

ovale

after

stroke:

a

preliminary

open study

of

8

cases.

Neurol

Res

1998;

20(4):

297-301.

29.

Hagen,

P.,

Scholz,

D.,

&

Edwards,

W.

Incidence

and size

of

patent foramen

ovale during

the

first

10

decades

of

life:

an

autopsy

study

of

965

normal

hearts.

Mayo

Clinic

Proc

1984;

59:

17-20.

30.

Hausmann,

D.,

Mugge,

A.,

Becht,

I.,

&

Daniel

W.G.

Diagnosis

of

patent

foramen

ovale

by

transesophageal

echocardiography

and

association

with cerebral

and

peripheral

embolic

events.

Am

J

Cardiol

1992;

70(6):

668-72.

31.Homma,

S.,

Di

Tullio,

M.R.,

Sacco, R.L.,

Mihalatos,

D.,

Li

Mandri,

G,.

& Mohr,

J.P.

Characteristics

of

patent

foramen

ovale

associated

with

cryptogenic stroke.

A

biplane

transesophageal

echocardiographic

study. Stroke

1994;

25(3):

582-6.

32.

Homma,

S.,

Di

Tullio,

M.R.,

Sacco,

R.L.,

Sciacca,

R.R.,

Smith,

C.,

&

Mohr,

J.P.

Surgical

closure

of

patent

foramen

ovale in

cryptogenic

stroke

patients.

Stroke

1997;

28(12):

2376-

81.

33.

James, P.B. Safety

of

contrast

echocardiography

in

screening divers.

Lancet

1990

Dec

1;

336(9727):

1389-90.

34.Jauss,

M.,

Kaps,

M.,

Keberle,

M.,

Haberbosch,

W.,

&

Dorndorf,

W.

A

comparison of

transesophageal

echocardiography

and

transcranial Doppler

sonography

with contrast

medium

for

detection

of

patent foramen

ovale.

Stroke

1994;

25(6):

1265-7.

35.

Jeffrey,

S.

Scuba

divers

should

be

screened

for

heart shunt.

The

Medical

Post,

May

19,

1998;

p.

18.

36.Johnston,

R.P.,

Broome,

J.R.,

Hunt,

P.D.,

&

Benton,

P.J.

Patent foramen

ovale

and

decompression

illness

in

divers. Lancet

1996

Nov

30;

348(9040):

1515.

37.

Jones,

E.F.,

Calafiore,

P.

Donnan,

G.A.,

&

Tonkin,

A.M. Evidence that

patent

foramen

ovale

is

not

a

risk factor

for

cerebral

ischemia

in

the

elderly. Am

J

Cardiol

1994;

74(6): 596-9.

38.Justo,

R.N.,

Nykanen,

D.G.,

Boutin,

C.,

McCrindle,

B.W.,

Freedom,

R.M.,

Benson,

L.N.

Clinical

impact

of

transcatheter

closure

of

secundum

atrial

septal

defects

with

the

double

umbrella

device. Am

J

Cardiol

1996;

77(10):

889-92.

39.

Karttunen,

V.,

Ventila,

M.,

Hillbom,

M.,

Salonen,

0.,

Haapaniemi,

H.,

&

Kaste,

M.

Dye

dilution

and

oximetry

for

detection

of

patent

foramen

ovale.

Acta

Neurol

Scand

1998;

97(4):

231-6.

40.

Kaulitz,

R.,

Paul,

T.,

&

Hausdorf,

G.

Extending

the

limits

of

transcatheter

closure

of

atrial

septal

defects

with

the

double

umbrella device (CardioSEAL).

Heart

1998;

80(1):

54-9.

22

41.

Kerut,

E.K.,

Truax,

W.D.,

Borreson,

T.E.,

Van Meter,

K.W.,

Given,

M.B.,

&

Giles,

T.D.

Detection

of

right

to

left shunts

in

decompression

sickness

in

divers.

Am

J

Cardiol

1997;

79(3): 377-8.

42.

King,T.D.,

Thompson,

S.L.,

Steiner,

C.,

&

Mills,

N.L.

Secundum

atrial

septal

defects:

non-

operative

closure

during

cardiac

catheterization.

JAMA

1976;

235:

2506-9.

43.

Klotzsch,

C.,

Janssen,

G.,

&

Berlit,

P.

Transesophageal

echocardiography

and contrast-TCD

in

the

detection

of

a

patent

foramen

ovale:

experiences

with

111

patients.

Neurology

1994;

44(9):

1603-6.

44.

Knauth,

M.,

Ries,

S.,

Pohimann,

S.,

Kerby,

T.,

Forsting,

M.,

Daffertshofer,

M.,

Hennerici,

M.,

&

Sartor,

K.

Cohort study

of

multiple

brain

lesions

in

sport

divers:

role

of

a

patent

foramen

ovale. BMJ

1997;

314(7082):

701-5.

45.

Kwiecinski,

H.,

Mieszkowski,

J.,

Torbicki,

A.,

Pniewski,

J.,

&

Pruszczyk,

P.

[Detection

of

patent

foramen

ovale

by

transcranial Doppler

ultrasonography].

Abstract

only.

Neurol

Neurochir

Pol

1994;

28(1

Supp1):

29-34.

46.

Laden,

G.D.

Patent

foramen ovale

and

decompression

sickness

in

divers.

Lancet

1997

Jan

25;

349(9047):

288.

47. Lechat,

P.,

Lascault,

G.,

Mas,

J.L.,

Loron,

P.,

Klimczac,

K.,

Guggiari,

M.,

Drobinski,

G.,

Fraysse,

J.B.,

Thomas,

D.,

&

Grosgogeat,

Y.

[Prevalence

of

patent

foramen

ovale

in

young

patients

with

ischemic

cerebral

complications].

Arch

Mal

Coeur Vaiss

1989;

82(6):847-52.

48.Luotolahti,

M.,

Saraste,

M.,

&

Hartiala,

J.

Saline

contrast

and

colour

Doppler

transesophageal

echocardiography

in

detecting

a

patent

foramen ovale

and

right-to-left

shunts

in

stroke

patients.

Clin

Physiol

1995;

15(3):

265-73.

49.Lynch,

J.J.,

Schuchard,

G.H.,

Gross,

C.M.,

&

Wann,

L.S.

Prevalence

of

right-to-left

shunting

in

a

healthy population:

detection

by

Valsalva

maneuver contrast echocardiography.

Am

J

Cardiol

1984;

53:

1478-80.

50.

Mas,

J.L.

Diagnosis

and

management

of

paradoxical

embolism

and

patent

foramen

ovale.

Curr Opin

Cardiol

1996;

11(5):

519-24.

51.

Moon,

R.E.,

Camporesi,

E.

M.,

&

Kisslo,

J.A.

Patent

foramen

ovale

and

decompression

sickness

in

divers.

Lancet

1989

Mar

11;

1(8637):

513-4.

52.

Moon,

R.E.,

Kisslo,

J.A.,

Massey,

E.W.,

Fawcett,

T.A.,

&

Theil,

D.R.

Patent

foramen

ovale

(PFO)

and

decompression

illness.

Undersea

Biomed

Res

1991;

13(suppl):

15.

53.Murrison,

A.W.,

Glasspool,

E.,

Pethybridge,

R.J.,

Francis,

T.J.,

&

Sedgwick,

E.M.

Electroencephalographic

study

of

divers

with

histories

of

neurological

decompression

illness.

Occup

Environ

Med

1995;

52(7):

451-3.

54.

NATO.

Allied

Guide

to

Diving

Medical

Disorders. NATO

1997.

Pp.

1-4

tol-5.

55.

Nendaz,

M.R.,

Sarasin,

F.P.,

Junod,

A.F.,

&

Bogousslavsky,

J.

Preventing stroke

recurrence

in

patients

with

patent foramen

ovale:

antithrombotic

therapy,

foramen

closure,

or

therapeutic

abstention?

A

decision analytic perspective.

Am Heart

J

1998;

135(3):

532-41.

56.

Nishi,

R.Y.

Doppler

and

ultrasonic

bubble detection.

In:

Bennett,

P.

&

Elliot,

D.

(Eds),

The

Physiology

and

Medicine

of

Diving,

4t'

Ed.

WB

Saunders

Co.,

Ltd:

London.

pp

433-

23

57.

453.

58.

Nygren,

A.T.,

&

Jogestrand,

T.

Detection

of

patent

foramen

ovale

by

transcranial

Doppler

and

carotid duplex ultrasonography:

a

comparison

with

transoesophageal

echocardiography.

Clin

Physiol

1998; 18(4):

327-30.

59.

O'Rahilly,

R.,

&

Muller,

F.

Human

Embryology

&

Teratology,

2

nd

Edition.

1992;

Wiley-

Liss,

New

York,

pp.

159-206.

60.

Petty,

G.W.,

Khandheria,

B.K.,

Chu,

C.P.,

Sicks,

J.D.,

&

Whisnant,

J.P.

Patent

foramen

ovale in

patients with

cerebral

infarction.

A

transesophageal echocardiographic

study.

Arch

Neurol

1997;

54(7):

819-22.

61.

Pilmanis,

A.A.,

Meissner,

F.W.,

&

Olson,

R.M.

Left

ventricular

gas

emboli

in

six cases

of

altitude-induced decompression

sickness. Aviat

Space

Environ

Med

1996;

67(11):

1092-6.

62.

Porembka,

D.T. Transesophageal

echocardiography.

Critical

Care

Clinics

1996; 12(4):

875-

918.

63.Powell,

M.R.,

Norfleet,

W.T.,

Kumar,

K.V.,

& Butler,

B.D.

Patent foramen

ovale

and

hypobaric

decompression.

Aviat

Space

Environ

Med

1995;

66(3):

273-5.

64.

Reul,

J.,

Weis,

J.,

Jung,

A.,

Willmes,

K.,

&

Thron,

A.

Central

nervous

system lesions

and

cervical

disc

herniations

in

amateur

divers.

Lancet

1995;

345:

1403-5.

65.

Ruchat,

P.,

Bogousslavsky,

J.,

Hurni,

M.,

Fischer,

A.P.,

Jeanrenaud,

X.,

&

von

Segesser,

L.K.

Systematic

surgical closure

of

patent

foramen

ovale

in

selected

patients

with

cerebrovascular

events

due

to

paradoxical embolism.

Early

results

of

a

preliminary

study.

Eur

J

Cardiothorac

Surg

1997; 11(5):

824-7.

66.

Schneider,

B.,

Zienkiewicz,

T.,

Jansen,

V.,

Hofmann,

T.,

Noltenius,

H.,

& Meinertz,

T.

Diagnosis

of

patent

foramen

ovale

by

transesophageal

echocardiography

and

correlation

with

autopsy

findings.

Am

J

Cardiol

1996;

77(14):

1202-9.

67.

Shields,

T.G.,

Cattanach,

S.,

Duff,

P.M.,

Evans,

S.A.,

&

Wilcock,

S.E.

Investigation

into

possible contributory

factors

to

decompression

sickness

in

commercial

air

diving

and

the

potential

long-term neurological consequences. Offshore Technology

Report

-

OTO

96

953,

Sept

1996;

pp. 36-37.

68.Siostrzonek,

P.,

Zangeneh,

M.,

Gossinger,

H.,

Lang,

W.,

Rosenmayr,

G.,

Heinz,

G.,

Stumpflen,

A.,

Zeiler

k.,

Schwarz,

M.,

&

Mosslacher,

H.

Comparison

of

transesophageal

and

transthoracic

contrast echocardiography

for

detection

of

a

patent foramen

ovale.

Am

J

Cardiol

1991;

68(11):

1247-9.

69.

Smith,

D.J.

Interatrial

shunts

and

decompression

sickness.

Lancet

1990

Oct

6;

336(8719):

879-80.

70.

Smith,

D.J.,

Francis,

T.J.,

Hodgson,

M.,

Murrison,

A.W.,

&

Sykes,

J.J.

Interatrial shunts

and

decompression

sickness

in

divers.

Lancet

1990

Apr

14;

335(8694):

914-15.

71.

Smith,

D.J.,

Francis,

T.J.,

Hodgson,

M.,

Murrison,

A.W.,

&

Sykes,

J.J.

Interatrial

shunts

and

decompression

sickness

in

divers.

Lancet

1990

June

30;

335(8705):

1593.

72.Spencer,

M.P.

Decompression

limits

for

compressed

air

determined

by

ultrasonically

detected

blood

bubbles.

J

Appl

Physiol

1976; 40:

229-35.

24

73.

Stoddard,

M.F.,

Keedy,

D.L.,

&

Dawkins,

P.R.

The cough

test

is

superior

to the

Valsalva

maneuver

in the

delineation

of

right-to-left

shunting

through

a

patent

foramen

ovale during

contrast transesophageal echocardiography.

Am

Heart

J

1993;

125(1):

185-9.

74.

Stone, D.A.,

Godard,

J.,

Corretti,

M.C.,

Kittner,

S.J.,

Sample,

C.,

Price,

T.R.,

&

Plotnick,

G.D.

Patent

foramen

ovale:

association

between

the

degree

of

shunt

by

contrast

transesophageal

echocardiography

and the

risk

of

future

ischemic

neurologic events.

Am

Heart

J

1996;

131(1):

158-61.

75.

Teague,

S.M.,

&

Sharma, M.K.

Detection

of

paradoxical

cerebral echo

contrast

embolization

by

transcranial

Doppler

ultrasound. Stroke1991;

22:

740-745.

76.

Turner,

M.

Patent

foramen ovale

and

decompression

illness

in

divers.

Lancet

1996

Nov

30;

348(9040):

1515.

77.

Webb, J.T.,

Pilmanis,

A.A.,

&

O'Connor,

R.B.

An

abrupt

zero-preoxygenation

altitude

threshold

for decompression

sickness

symptoms.

Aviat

Space

Environ

Med

1998;

69(4):

335-40.

78.

Wilmshurst,

P.T.

Right-to-left

shunt

and

decompression

sickness

in

divers.

Lancet

1990

Oct

27;

336(8722):

1071-2.

79.

Wilmshurst,

P.T. Patent foramen

ovale

and

subaqua

diving.

BMJ

1992;

304(6837):

1312.

80.

Wilmshurst,

P.T.

Pers

comm

to

Dr.

L.

Thompson,

Oct

9,

1998.

81.

Wilmshurst,

P.T.,

Byrne,

J.C.,

&

Webb-Peploe,

M.M.

Neurological

decompression

sickness.

Lancet

1989

Apr

1;

1(8640):

731.

82.

Wilmshurst,

P.T., Byrne,

J.C.,

&

Webb-Peploe,

M.M.

Relation between

interatrial

shunts

and

decompression

sickness in

divers.

Lancet

1989

Dec

2;

2(8675):

1302-6.

83.

Wilmshurst,

P.T.,

Ellis,

B.G.,

&

Jenkins,

B.S.

Paradoxical air

embolism

in

a

scuba

diver

with

an

atrial

septal defect.

Br

Med

J

1986;

293:

1277.

84.

Wilmshurst,

P.T. Treacher,

D.F.,

Crowther,

A.,

&

Smith,

S.E.

Effects

of

a

patent

foramen

ovale on

arterial saturation during exercise

and on

cardiovascular

responses

to

deep

breathing,

Valsalva

maneuvre, and

passive

tilt:

relation

to

history

of

decompression

illness

in

divers.

Br Heart

J

1994;

71(3): 229-3

1.

85.

Wilmshurst,

P.T.,

Walsh,

K.,

&

Morrison,

L.

Transcatheter

occlusion

of

foramen

ovale with

a

button

device after

neurological

decompression

illness

in

professional divers.

Lancet

1996

Sept

14;

348(9029):

752-3.

86.

Wilmshurst,

P.T., Walsh,

K.,

&

Morrison,

L.

Patent

foramen ovale and

decompression

illness

in

divers.

Lancet

1997

Jan

25;

349(9047):288.

87.

Vik,

A.,

Jenssen,

B.M.,

&

Brubakk,

A.O.

Paradoxical

air

embolism

in

pigs with

a

patent

foramen

ovale.

Undersea

Biomed

Res

1992;

19(5):

361-74.

88.

Vik,

A.,

Jenssen,

B.M.,

&

Brubakk,

A.O.

Arterial

gas

bubbles

after

decompression

in

pigs

with

patent

foramen

ovale.

Undersea

Hyperb Med

1993;

20(2):

121-131.

89.

Zamora,

R.,

Rao, P.S.,

Lloyd, T.R.,

Beekman,

R.H.

3r), &

Sideris,

E.B.

Intermediate-term

results

of

Phase

I

Food

and

Drug

Administration

Trials

of

buttoned device

occlusion

of

secundum

atrial

septal

defects.

J

Am

Coll

Cardiol

1998;

3

1(3):

674-6.

25

Additional

info

and

quotes,

not

included

but

possibly

relevant.

*

Risk

of

venous

bubbling

seems

to

be

reduced

in

aerobically trained

runners compared to

sedentary

subjects

in

some

studies,

but

others

have

shown

no

relation

to

fitness

and

DCS

(Broome

et

al,

1995).

*

Knowing

the

incidence

of

PFO

in

the

general

population,

one

can

assume

that

the

incidence

of

PFO

in

divers

would

be

similar

since there

is

no

selection

bias

in

a

person's

choice

to

take

up

diving based

on

the

presence

of

an

"unknown"

PFO.

*

Is

there

any

info

on whether known

PFO cases

choose

NOT

to

dive more

frequently

than

unknown

cases

later

discovered?

*

PFO

rates

in divers

and

controls,

not

in

stroke

patients

Moon

(1989)

37%

diver

Cross

(1990)

31.6% diver

Wilmshurst

(1989b)

25/61

diver,

24%

control

Germonpre

(1998)

36.1

%

control

Lynch

(1984) in

Glen

et

al

(1995)

TTE

5-20%

From

Kerut:

Zhu

et

al

(1991)

-

38%

From

Kerut

:

Job

et

al

(1994)

-

43%

*

Difference

in

contrast

media:

Jauss

(1994):

Galactose particle

suspension

is

stable for

about

60 sec

after

intravenous injection,

which

may

lead

to

higher

sensitivity

in

PFO

detection

than

with

other contrast

media

such

as

air

or gelatine.

This

however

necessitates

determining

a

time limit

to

prevent

false

positives

due

to

lung

passage.

Definition

of

presence

of

PFO

based

on

number

of

bubbles passed

per

cardiac

cycle

etc.

*

From

an

epidemiological

perspective,

the

characteristics

of

a

useful

population

based

screening

measure

are:

(From

Chan

Shah's

book

on

Public

Health)

1)

Conditions for which

screening

is

used

should

be

important

health

problems

i.e.

The incidence

should

be

sufficiently

high

that

the

cost

of

screening

is

not

prohibitive.

2)

Facilities for

diagnosis

and

treatment should

be

available

3)

Effective, non-controversial

treatment

for

patients

with

confirmed

condition

should

be

available

4)

Tests

should

have

high

sensitivity

and

specificity;

screening must

be

safe,

rapidly

applied,

and

acceptable

to

the

population

being screened

5)

The natural

history

of

the

condition should

be

understood,

such

that

if

detection

and

treatment

do

not alter

the

natural history,

screening

should not

be

implemented

6)

Policy

must

stipulate

what

action

will

be

taken

in

borderline

cases

to

avoid

overdiagnosis

7)

Maximum

benefit

for

minimum

cost must

be

achieved

by

comparing

the

costs

and

efficiency

of

various

screening

methods

26

8)

Control and

screened

groups

should

be

compared

at

regular

intervals

to

determine

whether

the

screening

procedure

and

subsequent investigations

have

an

effect

on

the

control

group

that

is

greater than

just

regular observation

(placebo effect)

9)

Compliance

with

screening

recommendations

10)

Screening

programs

should

be

a

continuous

process

27

Table

4.

FLOW-PATENT

FORAMEN

OVALE

AND

ECHOCARDIOGRAPHY

Number

of

In

Vivo

Resting Augmentation

Study

Patients

Modality

Conditions

(%)

Maneuvers

(%)

Chen

et

at"

32

TTE

25

38

TEE

44

63

Konstadt

et

all',-5

50

TEE

10

22

Porembka

el

al"'•

30

TEE

27

Stollberger

el

al""1

264"

TEE

15

Lechat

et

al'l

100 TTE

5

10

60

18

24

Hausmann

el

al"l

198

TTE

8

TEE

22

Jaffe

et

al'"

30

TEE

10

Unchanged

Guggiari

et

al,"

189

TTE

8

10

Black

et

a0"

101

TTE

6

Unchanged

51

TEE

8

-

Siostrzonek

et

al"

7O

150

r-E

5

6

160

TEE

12

20

'Suspected

embolic

events.

From

Porembka,

1996

Table

1:

Frequency

of

DCS

in

Sport,

Military,

and

Commercial

Air

Diving

Populations

Source

Military

Sport

Commercial

Reference

(13)

(11,12)

(14)

All

Total

dives'

648,488

2,577,680

43,063 3.269,231

Total

DCS4

172

878

152

1,202

Type

1l

DCS"

86

649

9

744

Incidents

DCSb

2.65

3.41

35.3

3.68

Incidents

DCS

]i

1.33

2.52

2.09

2.28

"Values

are

number

of

events;

bincidents

per 10,000

dives,

DCS

II

-

DCS

type

1I.

From

Bove,

1998

Accuracy

of

Different

Echocardiographic

Criteria

for

Identifying

an

Autopsy-Proven

Patent

Foramen Ovale

SensitivitySpecificityPos

PredictiveNeg PredictivePrevalence

PFO

Definition

(%)

Value

(%)

Value

(%) (%)

Contrast

TEE

Bubbles

LA/heart

cycles

>1/3

89 100 100

96

23

>2/3

67

100 100

90

17

>5/3

55

100

87

14

>2/immediately

44

100

100 84

11

Multiple/

1-2

33 100

100

81

9

Color

Doppler

TEE

Shunt direction

Right-to-left

and/or

left-to-right

100 100 100

100

26

Right-to-left,

not

left-to-right

89 100

100

96

23

From

Schneider

et

al,

1996

28

ACKNOWLEDGEMENTS

Dr.

Joan Saary

is

a

second

year

resident

in

Occupational

Medicine

at

the

University

of

Toronto.

She

completed

this

thorough

review

during

an

elective

in

Diving

Medicine

at

the

Defence

and

Civil

Institute

of

Environmental

Medicine

in

November

and

December

1998.

Dr.

Saary

gratefully

acknowledges

the

contribution

of

Dr.

David

Sawatzky

and

Mr.

Ron

Nishi

to

this

review.

Dr.

Sawatzky

provided

a

thorough

tutorship

in

diving

medicine

during

the

elective.

Both

Dr.

Sawatzky

and

Ron

Nishi

were

kind

enough

to

provide

a

careful

review

of

this

paper.

II2

SECUýflY

-1CLASS1FfCA

-iCNCPCM

(Nichest

c~aSs-'flCadcr

of

7de.

Abstract.

Ka,,r.VcrsJ

OCCUIMENT

CONTROL.

DATA

(Securit'

cdassiicaticn

ot

title,

tccy

of

sastrac

and

indexing annotation

must

to

=entered

when

the

overall

tccu~ment

is

CiassiffLed)

I.

ORIGINA7OR

(the

name

and

address

of

:h-e

argarizatrcr

preparing

the

documen't.

2.

SECURITYi

CLASSIF.CAT

TCN

(overall

security

Organizatocns

cr,.vhcm

the

documert

was

prepareed

-.

eg.

Establishment

c!assiffioabcn

of

!,,e

docurment.

including

sponscring

a

ccntrac-or's

report,

cr

zasking

agency.

are

entered

in

section

8)

special

woarning

terms

it

applicable)

UNCLAS

OCCEW~edical

Assessment

and

Training

3. 17TITLEE

(the

complete

document

JUe

as

indicated

an

the Itite

page.

Its

Olassiflcadcn

snculd

ýe

indicated'by

the

appropriate

abbreviation

(S. C

or

U) in

parentheses

atear

the

tite)

T

he

Possible

Relationship

Between

Patent

Foramen

Ovalle

and

Cecompression

Sickness:

A

Review

at

the

Ute~rature

4.

AUTHORS

(Last

name, first

name,

middle

initial)

Saary

MJ.

Gray

G-W

5.

DATE

OF PUBUCATION

(month

and

year

ot

publication

at

Sa.

NO.

OF

PAGES

(total

containing

8b.

NO.

OF

REFS

(total

document)

information.

Include Annexes,

cited

in

document)

January

1999

28

as8

7.

DESCRIPTIVE

NOTES

(the

The Possible Relationship Between Patent Foramen Ovale and Decompression Sickness:

Abstract and Figures

Recommended publications

Inner-ear decompression sickness in nine trimix recreational divers

A review of the relationship between patent foramen ovale and type II decompression sickness

Transcranial contrast doppler improves assessment of significance of patient foramen ovale in divers...

Transcranial Doppler Detection of Venous-to-Arterial Circulation Shunts: Criteria for Patent Foramen...

Paradoxical embolization and patent foramen ovale in scuba divers: Screening possibilities