Arri/icirr/ Organs 20(6):66266.5, Blackwell Science, Inc., Boston 0 1996 International Society for Artificial Organs

Subchronic Use of the St. Jude Centrifugal Pump as a Mechanical Assist Device in Calves

Jack Curtis, Colette Wagner-Mann, *Fred Mann, Todd Demmy, Joseph Walls, and ?James Turk

Division of Cardiothoracic Surgery, Department of Siirgery, School of Medicine, *Department of' Veterinary Medicine and Surgery, College of Veterinary Medicitw, und ?Department of Veterinary Pathology, C o l l ~ g e o j

Veterinary Medicine, University o j Missouri, Columbia, Mirwuri , U . S . A .

Abstract: The purpose of this experiment was to study the effects of the St. Jude Lifestream centrifugal pump on hemodynamic and hematologic parameters and the inci- dence of postmortem findings in a subchronic ex vivo left ventricular assist animal model. Five calves were im- planted with the pump as a left ventricular assist device (left atrial to thoracic aorta bypass) and studied for 96 h of continuous pumping under identical conditions. Heparin (100 IUlkg) was administered only in the initial saline pump prime. Throughout the protocol, mean arterial and central venous pressures averaged 102.1 ? 4.6 and 3.4 t- 2 . 2 mm Hg, respectively. Pump flow was 47.8 * 8.4 mll

kglmin at a mean pump speed of 1,676.3 k 106. I rpm. N o clinical abnormalities or mechanical malfunctions attrib- utable to the pump were detected during the 96 h of con- tinuous pumping for each calf. Mean plasma-free hemo- globin after 96 h was 3.9 ? 3.7 pmol/L (p = 0.337 compared to baseline). At post mortem, renal infarctions were detected in 1 calf. No other pump-associated lesions were detected in any of the other calves. We have con- cluded that the St. Jude Lifestream centrifugal pump functions reliably during 96 h of continuous left heart hy- pass in a calf model. Key Words: Centrifugal p u m p Mechanical assist-%. Jude Medical-Calves.

Centrifugal pumps are used most commonly for postcardiotomy mechanical assist after pharmacol- ogy and intraaortic balloon pumping have failed ( I ) . Although not designed for prolonged mechanical assist, cxtensive clinical use with the BioPump (Medtronic. BioMedicus, Inc., Eden Prairie, Min- nesota, U.S.A.) (2-5) and the Sarns centrifugal pump (Sarns 3M Healthcare, Ann Arbor, Michigan, U.S.A.) (6-9) have shown that some patients who are otherwise unweanable from cardiopulmonary bypass can be salvaged with this technology. The St. Jude Lifestream centrifugal pump (St. Jude Medical, Chelmsford, Massachusetts, U .S.A.) (Fig. 1) compares favorably with other centrifugal pumps in in vitro testing with regard to hemolysis and pump durability (10). However, there are few pub- lished reports of use of the St. Jude Lifestream cen- trifugal pump for cardiopulmonary bypass (1 I ) . We

arc unaware of any extended use of this device clin- ically or in an animal model. Thc purpose of this experiment was to study the effects of the St. Jude Lifestream centrifugal pump on hemodynamic and hematologic parameters in a subchronic ex vivo lcft ventricular assist animal model.

MATERIALS AND METHODS

Holstein bull calves (n = 5 ) weighing between 70 and 90 kg were chosen as the animal model. Each of the 5 calves in this study underwent an initial phys- ical examination and serum chemistry and complete blood cell count screening to determine general fit- ness. For 10 days prior to surgery, each calf was conditioned to the stanchion in which it was to be confined during the 96 h of centrifugal mechanical assist. All animals were managed in accordance with the Institutional Animal Care and Use Com- mittee guidelines.

Received January 1996. Surgical instrumentation Address correspondence to Dr. Jack Curtis, Division of Car- diothoracic Surgery, MA312 Medical Sciences Building, # 1 Hospital Drive, Columbia, MO 65212, U.S.A.

Anesthesia was induced with halothane and oxy- gen administered via a nasotrachel tube and main-

662

SUBCHRONIC ST. JUDE CENTRIFUGAL ASSIST IN CALVES 663

FIG. 1. The St. Jude Lifestream centrifugal pump head is shown. Characteristics of this pump head include curved fin design, lightweight construction, and minimal surface con- tact with the circulating blood. This pump head is manufac- tured and marketed by St. Jude Medical, Inc., Cardiac Arrest Division, Chelmsford, Massachusetts, U.S.A.

tained with isoflurane and oxygen administered via an orotrachel tube. The left neck and thorax were then prepped and draped in a sterile field. By a cervical incision, the jugular vein and carotid artery were cannulated and connected to precalibrated strain gauges (Transpac disposable pressure trans- ducer\; Abbott, North Chicago, Illinois, U.S.A.) for continuous postoperative central venous and systemic arterial pressure monitoring. Pressure tracings were monitored continuously using the Po- Ne-Mah Digital Acquisition Analysis and Archive System (Simsbury, Connecticut, U.S.A.), which was programmed to record 1 minute of every 10 min over the entire instrumentation and postoperative period. The heart and the aorta were then exposed by a left lateral thoracotomy entering the chest through the fifth intercostal space. Venous egress for lefi. ventricular assist was accomplished by can- nulating the left atrial appendage with a Sarns 3M 32Fr x 36Fr wire reinforced cannula with the bas- ket tip cut off. A synthetic vascular graft (12 mm diameter) (Cortex; W.L. Gore and Associates, Flagstaff, Arizona, U.S.A.) was sutured end-to-side to the thoracic aorta just distal to the aortic arch. The inflow cannula (Sarns 3M), a 32Fr wire rein- forced cannula with the basket tip cut-off, was in- serted into the Gortex graft and secured in place with silk sutures. Both the atrial outflow and aortic inflow cannulas exited through the ninth intercostal space in a manner that avoided kinking. The St. Jude Lifestream centrifugal pump was then primed with heparinized (100 IU/kg) saline solution and connected to the inflow and outflow cannulas after care was taken to completely de-air the circuit. The pump head was attached to an adapter (Model 2100

Pump Interface; Aires Medical, Chelmsford, Mas- sachusetts, U.S.A.) designed to interface with the BioConsole (Medtronic, BioMedicus, Inc., Eden Prairie, Minnesota, U.S.A.). Flow was begun grad- ually and then increased to the target pump flow of 50 ml/kg/min before closing the thoracotomy inci- sion to ensure proper function of the pump. After assuring hemostasis and inserting a chest tube, the surgical incisions were closed anatomically. Bupiv- acaine (1.5 mg/kg) was infused through the chest tube for local analgesia. Buprenorphine (0.01 mg/ kg) was administered at the conclusion of the oper- ation and every 6 h thereafter for postoperative an- algesia. No further heparin was administered.

Each calf was provided with intensive care during the postoperative period; data recorded included heart rate, respiratory rate, arterial blood pressure, central venous pressure, pump flow, and pump speed. Temperature was monitored, and blood sam- ples were drawn daily for a variety of laboratory tests including platelet counts, plasma-free hemo- globin (Hb) analysis, and serum lactate dehydroge- nase (LDH) levels. After 96 h of ventricular assist, each calf was euthanized humanely and submitted for a complete gross and histopathologic examina- tion by a veterinary pathologist (JT).

Statistical analysis Data are presented as mean k standard deviation

(SD). The paired t-test was applied for comparison of values from baseline to post-96 h of ventricular assist. Analysis of variance with repeated measures design was used to analyze the LDH, platelet count, and plasma-free Hb data. In all instances, significance was set at p < 0.05.

RESULTS

No clinical abnormalities or mechanical malfunc- tions attributable to the pump were detected during the 96 h of continuous pumping for each calf.

At post mortem, renal infarctions were detected in 1 calf. No other pump-associated lesions were identified in any of the calves.

The data collected in this study are summarized in Tables 1 , 2, and 3, and in Fig. 2. Although plasma-free Hb levels tended to increase slightly over time, the increases were not statistically sig- nificant when compared to baseline values. Only in the case of platelet counts was there a significant change from baseline (p = 0.020) with all postsur- gical values being significantly lower than the base- line (Fig. 2).

Artif Organs, Vol. 20, No. 6 , 1996

664 J . CURTIS ET AL,

TABLE 1. Summary of select serum chemistry and complete blood cell count values from the 5 calves

included in this study, both preoperatively (baseline) und after 96 h of left mechanical assist

TABLE 3. Summary ofplasma-jree hemoglobin and luctute dehydrogenuse values at baseline, 24 h , and

96 h of mechanical assist.for the 5 calves included in this study

Parameter Baseline 96 h Value of p Baseline 24 h 96 h

PC v 26.6 t 1.9 25.4 k 3.5 0.342 WBC 8,920 i- 2,586 7,240 2 1,730 0.219 PP 6.3 rt 0.4 4 .6r t 1 . 1 0.056 UN 5.6 2 2.1 4.6 2 1.1 0.460 Creatinine 0.8 rt 0.1 0.6 5 0.1 0.188

Values are expressed as mean i- SD. PCV, packed cell volume (%); WBC, white blood cell count

(per mm3); PP, plasma protein (rngidl); UN, urea nitrogen (mg/ dl); creatinine (mgidl).

DISCUSSION

Centrifugal pumps have been recognized as being less destructive to blood cellular elements than roller pumps when used for extended periods (10, l l ) . Consequently, in the setting of postcardiot- omy cardiogenic shock, centrifugal pumps are em- ployed commonly. In in vitro testing, the curved vane, St. Jude Lifestream centrifugal pump pro- duces a very low rate of hemolysis (10). However, to data there has been little published information about extended use of the St. Jude Lifestream cen- trifugal pump clinically or in laboratory animals.

In this experiment, we chose the calf model weighing between 70 and 90 kg so that the volume of blood pumped by the centrifugal pump would be similar to that required in humans for postcardiot- omy cardiac assist. Our experience has been that patients who will recover cardiac function after postcardiotomy cardiogenic shock will do so by 96 h of cardiac assist (12). Consequently, a device be- ing considered for this purpose should be able to function without mechanical difficulties for 96 h. In this experimental study using the St . Jude Lifestream centrifugal pump with continuous left heart bypass for 96 h, no mechanical difficulties were encountered.

TABLE 2. Summary of select physiologic and pump parameters for the 5 calves included in this study at 24

h and 96 h of left mechanical assist

Value 24 h 96 h of P

Hean rate bpm 102.4 ? 22.9 108.0 2 24.8 0.404 Respiratory rate (breaths

per min) 42.4 -C 11.9 40.0 f 14.1 0.675 Body temperature 39.1 -C 0.5 39.2 f 0.5 0.830 Central venous pressure

(mm Hs) 4.2 * 3 .3 3 .3 * 4.2 0.933 Mean artenal pressure (mm Hg) 103.0 2 7.2 108.8 f 7.7 0.107

Pump flow (mlikgimin) 52.4 2 8.8 50.5 * 8.3 0.117 Pump speed (rpm) 1,662.0 f 122.6 1,672.0 f 114.3 0.473

Free Hb

LDH (IU) 3,655 ? 2,052 3,472 ? 1,064 3,305 ? 1,349 (pnol/L) 2.2 2 0.9 3.2 -+ 1.8 3.9 -+ 3.7

Values presented are mean i- SD. There was no significant change from baseline for either parameter throughout the 96 h on assist (LDH: p = 0.783; free Hb: p = 0.377).

Hb, hemoglobin; LDH, lactate dehydrogenase.

Conditioned healthy calves tolerate continuous left ventricular assistance well for 96 h. There were no untoward clinical consequences secondary to centrifugal assist with regard to heart rate, blood pressure, and measured serum blood chemistry lev- els. Significant platelet reduction did occur, a phe- nomenon that is observed both in vitro (10) and clinically with other types of centrifugal pumps (2 77).

Although plasma-free Hb did tend to increase over 96 h of continuous centrifugal pumping, it did not reach statistical significance.

One of 5 animals in this study sustained a renal infarction that was not appreciated clinically but was observed at autopsy. We have learned that the clinically recognized incidence of thromboembo- lism secondary to centrifugal pumping underesti- mates the true magnitude of this problem (12,13). It should be noted that heparin was used only in the priming fluid in this experiment and not continu- ously as would be recommended clinically

700 s E 650 E $600

-550

500

450

7 X u) C c)

0

Platelet Counts 7 \ \

\ *

bl 1 4 24 48 72 96 Hours on Assist

FIG. 2. The average (n = 5) platelet counts (expressed at counts x 1 ,000/mm3, using an automated particle counter) over the 96-h mechanical assist period are shown. Data points (*) are significantly lower than the baseline counts (p = 0.020). ~

A r f f O r g a n s , Val 20, N o 6. 1996

SUBCHRONIC S T . JUDE CENTRIFUGAL ASSIST IN CALVES 665

We conclude that the St. Jude Lifestream centrif- ugal pump functioned reliably during 96 h of con- tinuous left heart bypass in a calf model. We would anticipate satisfactory performance in a clinical set- ting. Comparative testing of other centrifugal pumps in this experimental model may yield useful information.

REFERENCES

I . Pac WE Jr, Miller CA, Matthews Y. Pierce WS. Ventricular assist devices for postcardiotomy cardiogenic shock. J Tho- rac Cardioizasc Surg 1992;104(3):541-53.

2. Magovern GJ Jr. Use of the Bio-Medicus pump in postop- erative circulatory support. In: Ott RA, Gutfinger DE, Gaz- zaniga AB. eds. Cardiac surgery: state of the art revieH,s. ed 7 . Philadelphia: Hanley & Belfus, 1993:249-64.

3. Killen DA. Piehler JM, Borkon AM, et al. Bio-Medicus ven- tricular assist device for salvage of cardiac surgical patients. Ann Thoruc Surg 1991;52:23&5.

4. Noon GP. Bio-Medicus ventricular assistance. Ann Thorac Surg 1991;52:18(kl.

5. Golding LAR, Crouch RD. Stewart RW, et al. Postcardiot- omy centrifugal mechanical ventricular support. Ann Thorac Surg 1992;54:1059-64.

6. Curtis JJ . Centrifugal mechanical assist for postcardiotomy

ventricular failure. Semin Thorac Cardiovasc Surg 1994;6(3): 1 4 M .

7. Curtis JJ, Walls JT, Schmaltz R, Boley TM, Nawarawong W, Landreneau RJ. Experience with the Sarns centrifugal pump in postcardiotomy ventricular failure. J Thorac Car- diovasc Surg 1992;104(3):554-60.

8. Curtis J, Walls J , Schmaltz R, Boley T, Landreneau R, Nawarawong W. Prognosis of hospital survivors after sal- vage from cardiopulmonary bypass with centrifugal cardiac assist. Am Soc Arfiflnt Organs Trans 1990:88:811-3.

9. Joyce LD, Kiser JC, Eales F, King RM, Toninato CJ, Hansen J. Experience with the Sarns centrifugal pump as a ventricular assist device. Am Soc Artif Int Organs Trans

10. Curtis JJ, Wagner-Mann CC, Turpin TA, et al. In vitro eval- uation of five commercially available perfusion systems. Int J Angiology 1994;3: 128-33.

11 . Nishinaka T, Nishida H, Endo M, Koyanagi H. Less platelet damage in the curved vane centrifugal pump: a comparative study with the roller pump in open heart surgery. Art i for-

12. Curtis JJ , Walls JT, Schmaltz RA, Demmy TL, Wagner- Mann C, McKenney C, Nawarawong W. Improving clinical outcome with centrifugal mechanical assist for post cardiot- omy ventricular failure. Artif Organs 1995:19(7):761-5.

13. Curtis JJ, Walls JT, Boley TM, Schmaltz RA, Demmy TL. Autopsy findings in patients on postcardiotomy centrifugal ventricular assist. Am Soc Artif Inr Organs Trans 1992;38:

1990;36:M619-M623.

guns 1994: 18(9): 687-90.

M688-M690.

Artif Organs, Vol. 20, No. 6, 1996