Reprinted from HEART SUBSTITUTES
by Albert N. Brest, M.D.

CHARLES G THOMAS e PUBLISHER
Springfield, Illinois
pp 225-237, 1966

Chapter 18

HEART FAILURE AND IMPLANTABLE
INTRATHORACIC CIRCULATORY PUMPS:
PHYSIOLOGY AND CLINICAL APPLICATION*

D. Liotta, M.D., G. W. Hatt, M.D., anp M. E. DeBaxzy, M.D.

Dneversiore HEART failure is focusing attention to the need for mechanical
assistance, not only for a short period of time, but for long-term support. We
have previously reported on the implantable left ventricular bypass pump which
requires no anticoagulants (1, 2, 3, 4). The pump was designed to maintain
support for periods of weeks or months.

The prerequisites of an implantable circulatory pump to support left ven-
tricular function have likewise been outlined and previously reported (3).
They are as follows:

1. Long-term support of ventricular function (weeks, months?).

2. Avoidance of heparinizing the patient.

3. Minimal amount of blood trauma.

4. Feasibility of discontinued function of the implantable pump for certain
time intervals (hours, days) with normal resumption of its function after-
wards. This should be accomplished without general heparinization of the
patient and without danger of clotting the pump.

5. Possibility of pumping being synchronized with any preselected time of
the cardiac cycle.

6. Surgical technique of implantation which is simple and safe for the patient.

In this report, preliminary physiological observations and clinical comment
are presented from the initial phase of an active study in progress.

The physiology of the left ventricular extracorporeal bypass was studied, and
compared with similar observations obtained from implanted intrathoracic cir-
culatory pumps.

METHODS

Fifty mongrel dogs were used for the left ventricular extracorporeal bypass
and twelve for the implanted intrathoracic circulatory pump. Included are
data from the initial phase of a physiological study of experimental chronic
heart failure (E.C.H.F.) (Fig. 1). This study involved (1) arterial-venous
fistulas; (2) aortic insufficiency, and (3) mitral insufficiency. From the mor-

*This study was supported in part by Grant HE-09252 from the United States Public Health
Service.

Heart Substitutes, Mechanical and

 

[1] Transplant, 1966, pages 225-237,
2 Heart Substitutes

 

Figure 1. Experimental chronic heart failure in a dog with arterial-venous fistula. Note
the bulging left atrium.

tality standpoint, the best results were obtained by performing two A-V fistulas
on one date, followed by a third fistula three to four weeks later.

Extracorporeal Bypass of the Left Ventricle

A left thoracotomy through the fifth intercostal space was performed.
Respiration was maintained by a Bird* intermittent positive pressure res-
pirator. Both femoral arteries and one femoral vein were exposed. After
heparinization (2 mg/Kg), a partial left ventricular bypass from left atrium
to femoral artery was performed, using a calibrated occlusive DeBakey
roller pump. Pumping was continued for two hours, and pump speed was
varied to keep the mean left atrial pressure slightly positive during expira-
tion. The connection to the left atrium was either by the direct cannulation
of the atrial appendage with the bypass tubing (internal connection) or
by an external molded Silastic prosthesis (external connection) sutured to
the atrium and then sealed with Eastman Adhesive 910.** When the
external connection was used, the atrium was clamped and a small disc
was excised for the blood to pass into the prosthesis. The external con-
nection was then sealed to an adapter attached to the tubing.

*Manufactured by Bird Corporation, Palm Springs, California.
**Manufactured by Ethicon, Inc., Somerville, New Jersey.

Heart Failure and Implantable Pumps 3

A specially designed tubing stand attached to the operating table was used
to hold the inlet and intrathoracic segments of the tubing in a constant pre-
selected position during the procedure. Blood was returned to a femoral artery
through a tapered plastic catheter.

Intracorporeal Bypass of the Left Ventricle

Early models have been previously described (1, 2). The pump consists of
(1) an implantable intrathoracic pump (Fig. 2) and (2) an external gas
energizing and controlling system. The intrathoracic pump is made of 0.040
inch thick Dacron-reinforced Silastic and consists of a blood chamber surrounded
by a gas chamber. The implanted pump is coupled to the external gas system
by a small tube brought out through the chest wall. Pressurized CO, entering
the gas chamber compresses the blood chamber thereby expelling its contents.
Ball valves guarding the inlet and outlet cause unidirectional blood flow. The

Ce

Blood

       
      
 
 

chamber

 
  

Outlet —T Toma <x
valve chamber

Dacron

Graft

poscer tito
Yh or

Figure 2. Schematic representation of the implantable pump. A. The inlet valve chamber

is sealed to the body of the pump with Eastman Adhesive 910. B. The same adhesive

makes the suture line of the inlet valve chamber and the left atrium air-tight. C. Inflatable

chamber for occluding the outlet connection, allowing the isolation of the pump from the
systemic circulation.
4 Heart Substitutes

inlet of the blood chamber is connected to the left atrium (external atrial
connection) (Fig. 3) and the outlet to the descending thoracic aorta. Tech-
nical details of implantation are reported elsewhere (3).

The external gas system consists of a Teflon bellows powered by an electric
motor. A control system has been developed to initate a pump cycle in re-
sponse to an EKG signal (Fig. 4). Motor speed control is provided by phase
variable firing of silicon controlled rectifiers and has a dynamic braking cir-
cuit to bring the motor to a complete stop at the end of each stroke. By ad-
justing the maximum and minimum speed override circuits, the system main-
tains synchronization by initiating a stroke from either each EKG signal, every
second signal, or every third signal. The speed override circuit also insures
continuous operation (not necessarily synchronized) in the event that the EKG
rate becomes very high, or is interrupted. Electrocardiographic synchronization
can be substituted by a system dependent upon the mean left atrial pressure.

    
 

Circumflex
vessels.

eft atrial
Ra appendage

     
   

: ait pulmonary §
ea. Vessels.
=

 

Figure 3. Technique of implantation of the external left atrial connector: The external
connector is sutured medially to the appendage base (A), and laterally to the atrial wall
(B). Care should be taken to allow an atrial clamp to be placed proximal to this anasto-
mosis without encroaching upon the circumflex vessels (C). D. The atrial appendage and

part of the atrial wall was excised.

-26-
-24-

 

  
    

 

 

 

 

 

 

 

DIASTOLIC TRIGGERING IN IMPLANTABLE __
INTRATHORACIC CIRCULATORY PUMP

 

 

 

 

 

 

 

 

ARTIFICIAL VENTRICLE -

~ PRESSURE IN
‘AIR CHAMBER OF

 

€

&

Figure 4, The tracing shows a diastolic synchronization of an implantable pump with the
electrocardiogram,
6 Fleart Substitutes

The control system is transistorized and employs a minimum of switch and
relay contacts. An independent power supply provides for operating the sys-
tem without synchronization, and a battery powered converter is used to
make the entire system portable for periods up to 1 hour.

The following physiological measurements were made (Fig. 5):

uo Ascending
Sy aortic pressure
wy ORNS Pp

   
   

Subclavian a.

LS"
See KE

 
   
  

é

//yertriculary

pres: Ure 5, \
mY

KE,

 
    
  
  
 

... Left
“intraventricular
pressure

Femoral a.
pressure

yp

  

ight atrial Sy

PAR
: ~P essure

Figure 5. Schematic diagram showing the physiological parameters explored during the
present investigation,

Heart Failure and Implantable Pumps 7

Pressure Controt. An eight-channel recorder* with four pressure trans-
ducers was employed to determine the following:

1. Femoral artery pressure.

2. Right atrial pressure.

3. Ascending aortic pressure.

4. Left intraventricular pressure.

5. Left atrial pressure.

All pressure transducers were calibrated to a baseline by placing the tip of
a catheter 0.5 cm below the point of entrance of the left superior pulmonary
vein into the left atrium with the dog in right lateral position.

Mean left atrial pressure values were determined during expiration.

Putmonary Frow Controt. A Microflo Medicon Division Model FM-6
flowmeter was employed to measure pulmonary flow using a 14 or 16 mm
probe. Prior to each measurement, the electrical zero was corrected and the
biological zero was determined by clamping the artery distal to the probe while
the probe was surrounded with physiological saline.

TENSION-TIME INDEX: ‘The Tension-time index was determined by plan-
imetrically integrating in mm Hg secs. the area beneath the systolic portion of
the left intraventricular pressure curve which starts with the beginning of the
isometric contraction and ends at the point which corresponds with the aortic
incisura. All measurements were made on tracings taken at a paper speed
of 50 mm/sec.

Ejection-trme INpEx: The Ejection-time Index is a measurement of the
amount of tension developed in the myocardium of the left ventricle during
the ejection period of systole. As such, it is an indication of the stroke volume
of the ventricle. This value was determined by planimetrically integrating the
area under the systolic portion of the aortic pressure curve, beginning with
the opening of the aortic valve and ending at the aortic incisure. Values are
stated in mm Hg secs., and are from tracings taken at 50 mm/sec. paper speed.

RESULTS

Left Atrial Pressure

Left ventricular extracorporeal bypass in each case decreased the mean left
atrial pressure. Pump speed was varied to keep mean left atrial pressure
(MLAP) slightly positive during expiration. Without constant attention to
the pump speed and MLAP, negative pressure developed and resulted in dan-
gerous left atrial suction. Average mean pressure decrease in normal dogs
was 2.3 mm Hg (50 per cent). In experimental chronic heart failure dogs,
the decrease was 5.7 mm Hg (46 per cent).

A correlation was made between mean left atrial pressure decrease (end-

*Manufactured by Electronics for Medicine.
8 Heart Substitutes

diastolic left intraventricular pressure) and the systolic peak of left intraventri-
cular pressure. The changes in the systolic peak were classified as: (1) alternans,
(2) irregular decrease, and (3) uniform decrease. In seventeen normal dogs
studied, the average MLAP and the decrease during pumping was 4 and 2.5
mm Hg, respectively. The irregularities noted in the systolic peak were probably
due to the decreased intraventricular blood volume and decreased stimulus
to the end diastolic stretch reflex.

These changes in the ventricular contraction were unpredictable when the

 

 

 

 

 

Figure 6. Pressure tracings taken during extracorporeal bypass. A. Before bypass. B. Dur-

ing bypass—note marked alternans coincident with the decrease of MLAP. C. Note the

peak (P) of systolic contraction when MLAP reached zero, reflecting impairment of myo-

cardial contractility. D. The first part shows a mild alternans during pumping; in the

second part the pump is off. E. Marked irregular decrease in the systolic peak of the left
ventricle when the pump is on.

 

Heart Failure and Implantable Pumps 9

MLAP was maintained above zero. However, when the MLAP reached zero
or below, a succession of small beats in conjunction with abnormally high
systolic peaks was observed (Fig. 6).

In fifteen experimental chronic heart failure animals the average MLAP was
16 mm Hg. In this group, the systolic peak of the left ventricle remained
unchanged except in one instance when the flow through the bypass was ex-
tremely high. .

The implantable intrathoracic circulatory pump in each case decreased the
mean left atrial pressure. Average mean pressure decrease in normal dogs was
0.55 mm Hg (13 per cent). In experimental chronic heart failure dogs, the
decrease was 8 mm Hg (50 per cent).

The correlation between the mean left atrial pressure decrease and_ the
peak systolic intraventricular pressure depended upon the time of synchroniza-
tion. During independent rate pumping and during systolic pumping, the
systolic ventricular peak changed to a mild alternans or to a mild irregular
decrease. When only every other R wave was synchronized, a mild alternans was
a constant finding (Fig. 7).

Systemic Pressure

Left ventricular extracorporeal bypass usually increased mean systemic pres-
sure, though this was not a constant finding. Average mean pressure increase
in normal dogs was 1.2 mm Hg (0.6 per cent). In experimental chronic heart
failure dogs, the increase was 4.3 mm Hg (7 per cent).

The implantable intrathoracic circulatory pump generally increased the mean
systemic pressure. Again, this finding was inconstant. Average mean aortic
pressure increase in normal dogs was 3.9 mm Hg (4 per cent) ; in experimental
chronic heart failure dogs, 1.1 mm Hg (1 per cent).

Tension-time Index

The left ventricular extracorporeal bypass decreased TTI in all but one
animal. The average decrease in normal dogs was 16 per cent; in those with
experimental chronic heart failure, 8.2 per cent.

In all animals with the implantable intrathoracic circulatory pump, TTI
was decreased. Average decrease in normal dogs was 15 per cent; in experi-
mental chronic heart failure dogs, 3.8 per cent.

Ejection-time Index

A decrease in ETI was a constant finding with the left ventricular extra-
corporeal bypass. ETI in normal dogs decreased an average of 15 per cent;
experimental chronic heart failure dogs, 43 per cent.

The implantable intrathoracic circulatory pump also caused an ETI de-
crease in each case. Normal animals demonstrated an average decrease of 25
per cent; experimental chronic heart failure animals, 36 per cent.
LEFT VENTRICULAR BYPASS WITH IMP
INTRATHORACIC CIRCULATORY PUMP inmate oS
INDEPENDENT RATE PUMPING So"

/300-PULSE precee eo SEC. HEART RATE=|59/min, "3°"

; So. -26-

 

     
  
   
 
  
 
  
  
   
     

 

SYSTEM | oe
PRESS Lo. Lo a io | -22-
Le : 20.

~—-TTh. i ET. 1. | : ( . -

‘MEAN A
LEFT ATRIAL ple
PRESSURE Le 4-
| \ of a |

igh = l2=
= 1O-
Bo
= 6-
aS q-
- lee
ey = O-
ae mm.He

 

 

 

 

 

 

: “HEART RATE=I36/in. Oe

 

   
 
    
 

-28-
A ont re6-
“ ‘ rea
. -22-
. ob : ~20-
| pee rmer, Lym > —-4'e
'eo- AQRTIG. feet s ia ls-
je0-. PRE f By Cc

ne

Er / : \ 7 b
0- Wwenmcuca i SURE

 

 

Figure 7. Pressure tracings taken during bypass with implantable pump. Note the mild
decrease in MLAP. The pressure pulse from the pump is reflected in the aortic pressure
at the same time that the gas system pressure reaches its systolic peak. A. During inde-
pendent rate pumping. The pump rate is 76/min., and the heart rate is 153/min. Note the
pressure pulse from the pump is unrelated to the cardiac cycle. B. During systolic pumping.
Note that the pressure pulse from the pump is superimposed on the systolic aortic pressure
curve. C. During diastolic pumping. Note that the pressure pulse from the pump occurs
during diastole. ETI = crosshatched area; TTI = shaded area.

Heart Failure and Implantable Pumps 11

Pulmonary Flow

Pulmonary flow was measured in two animals with the implantable intra-
thoracic circulatory pump, and 4 with the extracorporeal bypass. All were
normal dogs and no change was observed.

DISCUSSION

The constant observation in both extra- and intracorporeal bypasses has
been a decrease in mean left atrial pressure (MLAP). Because they are
functionally related, the left ventricular end-diastolic pressure (LVEDP), pul-
monary capillary pressure, and pulmonary venous pressure were consequently
decreased. However, the correlation between MLAP and the systolic peak
of the left ventricular pressure demonstrated that there was a drastic change
in myocardial contractility when MLAP was allowed to approach or reach
negativity.

The intrathoracic pump avoided an important decrease in MLAP in normal
dogs. Indeed, MLAP was constantly autoregulated (2, 3). On the other hand,
a striking contrast occurred when the bypass was performed in dogs with
elevated MLAP (experimental chronic heart failure group). In this group,
MLAP was decreased, but not to a pressure below normal, and only slight
changes were occasionally detected in myocardial contractility.

There is a discrepancy in the reported results of complete left ventricular
bypass when directed toward decreasing the myocardial oxygen consumption
rather than allowing an effective myocardial contraction to occur. Unfortun-
ately, the reported work on this bypass used normal dogs and an extracorporeal
pump, overlooking the provision of an effective left ventricular filling pressure.
It is also evident that the law of Starling (5), which has recently been con-
firmed in man (6-9), warrants consideration.

It appears that both factors (a decrease or an elevation of MLAP) may result
in an impairment of left ventricular contractility.

It is apparent that a mechanism which could continue to maintain left
ventricular end-diastolic pressure at a level adequate for an effective left ven-
tricular contraction, and allow decompression of the left atrium and pulmonary
venous system, would be of clinical interest (Fig. 8).

TTI showed a decrease during extra- and intracorporeal bypass. The myo-
cardial wall tention (TTI) is the primary determinant of cardiac oxygen con-
sumption (10). However, in both groups the ETI demonstrated an important
decrease because the bypass affects predominantly the ejection phase of the
left ventricular contraction.

Pulmonary flow remained unchanged except for a brief period at the be-
ginning and end of the bypass. This observation indicated a blood redistribution
rather than overload of the right ventricle.
12 Heart Substitutes

Pulmonary venous
> Hig, Pressure

NU am Hg.

 
  
  

Mean left
atrial pressure......\...

 

 
 
 
 
  
  
 
 

End diastolic
left ventriclar
pressure...........
LEFT VENTRICULAR FAILURE
Pulmonary venous.
re pressure 35S
Mean left

atrial pressure... 3-5

End diastolic
left ventricular
pressure....2...

Descending
aorta

Decompression of left ventricular diastolic pressure,
mean left atrial pressure and pulmonary venous pressure
during bypass functioning, leaving a filling pressure of
3-5 mm. Hg.

Figure 8. Diagram showing the concept of the decompression of MLAP, LVEDP, and

pulmonary venous pressure during left ventricular bypass in left heart failure.

Negligible hemolysis was noted when a tubing support was used. This
minimized turbulence and suction at the site of the left atrial connection.
Hall (11) reported significant hemolysis during left ventricular bypass with
the closed-chest technique, and one wonders whether placement of a cannula
within the heart could ever overcome significant hemolysis.

Clinical application of the implantable pump has been temporarily suspended
until we have a better understanding of its physiologic implications.

SUMMARY

A physiological evaluation was made of the left ventricular bypass using
both extracorporeal and implantable intrathoracic pumps.
The features of the implantable pump were defined as:
1, Long-term circulatory support;

Heart Failure and Implantable Pumps 13

Avoidance of heparinizing the patient;

Minimal amount of blood trauma;

Synchronization with a preselected time of the cardiac cycle;

Possibility of discontinuing the pump in order to correlate its effects;
6. Autoregulation of MLAP.

MLAP was shown to be critical in the provision of an effective ventricular

Te ON

contraction.

ACKNOWLEDGMENT

We are indebted to B. F. Polk and J. L. Crook, medical students, and F. A.
White, pre-medical student, at Baylor University College of Medicine, for their
valuable assistance durint the experiments of this work.

REFERENCES

1. Liotta, D., Crawford, E.S., Cooley, D.A., DeBakey, M.E., de Urquia, M., and Feldman, L.:
Prolonged partial left ventricular bypass by means of an intrathoracic pump implanted
in the left chest. Trans. Amer. Soc. Artif. Int. Organs, 9:90-99, 1962.

2. Liotta, D., Hall, C.W., Henly, W.S., Cooley, D.A., Crawford, E.S. and DeBakey, M.E.:
Prolonged assisted circulation during and after cardiac or aortic surgery. Am. J. Cardiol.,
12:399-405, 1963.

3. Liotta, D., Hall, C.W., Maness, J.H., and DeBakey, M.E.: The implantable intrathoracic
circulatory pump: surgical technique. Cardiov. Res. Cent. Bull., 3:54-61, 1964.

4. Hall, C.W., Liotta, D., Henly, W.S., Crawford, E.S., and DeBakey, M.E.: The develop-
ment of artificial intrathoracic circulatory pumps. Am. J. Surg., 108:685-692, 1964.

5. Starling, E.H.: The Linacre Lecture on the Law of the Heart, Cambridge, 1915. London,
Longmans, Green and Co., 1918.

6. Frye, R.L., Braunwald, E., and Cohen, E.R.: Studies on Starling’s Law of the Heart.
J. The circulatory response to acute hypervolemia and its modification by ganglionic
blockade. Jf. Clin Invest., 39: 1043-1050, 1960.

7. Braunwald, E., Frye, R.L., Aygen, M.M., and Gilbert, J.W., Jr.: Studies on Starling’s Law
of the Heart. III. Observations in patients with mitral stenosis and atrial fibrillation
on the relationships between left ventricular end-diastolic segment length, filling pres-
sure, and the characteristics of ventricular contraction. J. Clin. Invest., 39:1874-1884,
1960.

8. Braunwald, E., and Frahm, C.J.: Studies on Starling’s Law of the Heart. IV. Observations
on the hemodynamic functions of the left atrium in man. Circulation, 24:633-642, 1961.

9, Gleason, W.L., and Braunwald, E.: Studies on Starling’s Law of the Heart. VI. Relation-
ships between left ventricular end-diastolic volume and stroke volume in man with ob-
servations on the mechanism of pulsus alternans. Circulation, 25:841-848, 1962.

10. Sarnoff, S.J., Braunwald, E., Welch, G.H., Case, R.B., Stainsby, W. N., and Macruz, R.:
Hemodynamic determinants of oxygen consumption of the heart with special reference
to the Tension-Time Index. Am. J. Physiol., 192:148-156, 1958.

11. Hall, D.P., Moreno, J.R., Dennis, C., and Senning, A.: An experimental study of prolonged
left heart bypass without thoracotomy. Amer. Surg., 156:190-196, 1962.