You are on page 1of 8

Journal of Investigative Surgery, 14:55 61, 2001

Copyright c 2001 Taylor & Francis

0894-1939/01 $12.00 + .00

New Methodologies

A Novel Technique in a Sheep Model

for Evaluating Prosthetic Heart
Valve Performance
Matthew M. Puc, MD,
Steven W. Marra, MD, ABSTRACT There have been many various animal studies to evalu-
Hoang S. Tran, MD, ate the structural integrity and antithrombogenicity of prosthetic heart
Jonathan H. Cilley, Jr., MD, valves. We were interested in developing a novel sheep model to study
Charles W. Hewitt, PhD, and
the thrombogenicity of mechanical heart valves placed into the systemic
Anthony J. DelRossi, MD
Department of Surgery, Cooper circulation but without the need for cardiac bypass. Also, we wanted to
Hospital/University Medical minimize the risk of paraplegia from complete thoracic aortic clamping.
Center, University of Medicine Six sheep underwent left lateral thoracotomy for placement of a mechan-
and Dentistry of New Jersey, ical heart valve in parallel with the descending thoracic aorta. A valved
Robert Wood Johnson Medical conduit with a dacron tube graft sutured to the back end was fashioned.
School, Camden, New Jersey,
Employing partial aortic occlusion with a side-biting clamp, the proximal
and distal ends were anastomosed in an end-to-side fashion. Once ow
was con rmed through the graft, the native aorta was occulded with um-
bilical tape. The sheep received no postoperative anticoagulation. The
median operative time and estimated blood loss (EBL) was 170 min and
250 cc, respectively. Patency of the valved conduits was con rmed during
the initial procedure, and there was no incidence of paraplegia postop-
eratively. Two animals expired shortly after extubation and at necropsy
the valved conduits were patent with preserved valve function. The four
survivors were sacri ced a median of 37 days postoperatively. Prior to
euthanasia, the valved conduits were evaluated in situ with ultrasound.
In all cases, the valves had clot formation at the hinges, which prevented
Received 8 October 1999; active movement of the lea ets. This novel in vivo technique provides an
accepted 26 June 2000.
alternative in testing the thrombogenicity of prosthetic heart valves with-
Presented at the 15th Annual Meeting
of the Academy of Surgical Research, out cardiac bypass or the risk of paraplegia in an animal that is extremely
2 4 September 1999, Scottsdale, AZ. sensitive to complete aortic cross-clamp.
Address correspondence to Charles
W. Hewitt, PhD, Cooper Hospital/UMC,
Three Cooper Plaza, Suite 411,
KEYWORDS heart valve, sheep model, partial occlusion
Camden, NJ 08103, USA. E-mail:

New Methodologies

rosthetic heart valves have evolved signi - rable to humans in various ways [ 8, 9]. Their blood
cantly since their initial clinical use in cardiac pressure, cardiac output, heart rate, and blood coagu-
surgery. Today, more than 60,000 heart valve lation pro le mimic those of humans [10, 11]. Sheep
replacements are performed in the United States an- also have a docile temperament that allows their care
nually [1]. The two main types of heart valves are to be relatively simple [ 12]. In addition, anesthetic
either mechanical, which are composed primarily of doses in sheep are similar to humans, which makes
metal or carbon alloys, or bioprostheses. These bio- for easy administration of anesthesia [13, 14].
prostheses are further divided into heterografts, such When designing this model, it was necessary to
as porcine or bovine tissue, or homografts, which are make the model simple and ef cient with a minimal
preserved human aortic valves. expenditure of resources. One way to accomplish
Our area of interest has involved mechanical heart this would be to avoid cardiac bypass and its associ-
valves, in particular the bilea et tilting disk (St. Jude ated risks, but there was a desire to place the valve
Medical) valve. Even though this type of mechan- within the systemic circulation to mimic as closely
ical valve has the lowest thrombogenic potential, as possible the pressure gradients of an orthotopic
patients still require lifelong anticoagulation [2]. In- position. This could potentially be achieved by po-
dividuals taking oral anticoagulation are suscepti- sitioning the valve in the descending thoracic aorta,
ble to hemorrhagic complications, particularly in as previously reported, but would necessitate cross-
the gastrointestinal tract, cerebral circulation, and clamping the aorta [ 15, 16] . This last point is criti-
retroperitonium. The incidence of these complica- cal since our prior experience has shown that sheep
tions has been reported to be approximately 1% to do not tolerate complete thoracic cross-clamping for
4% per patient-year [3]. Our efforts in identifying extended periods. This was seen in our laboratory,
a solution that would eliminate the need for life- where we experienced paralysis following as little
long anticoagulation have been centered around sur- as 16 mins of total aortic cross-clamping. With all
face modi cation of mechanical heart valves to ef- of these factors taken into consideration, the nal
fect thrombogenicity [4]. We have employed two model incorporated a valved conduit anastomosed
separate techniques, which are well described in the in parallel with the descending thoracic aorta that
literature and are only mentioned brie y here. One could be accomplished with a partial cross-clamping
process, called diamond-like carbon coating (DLC), technique.
produces a dense amorphous hydrocarbon polymer
coating [5]. The other process, plasma or glow dis- MATERIALS AND METHODS
charge treatment (GDT), is a stepwise reaction utiliz-
ing gas plasma to generate a new molecular property All experiments were performed according to a re-
on the surface of the mechanical valve [6]. These search protocol approved by the University of Medi-
modi cations change the surface layers without af- cine and Dentistry of New JerseyRobert Wood
fecting the bulk structure or durability. Early reports Johnson Medical School (Camden, N J) Institutional
from our research laboratory employing both types Animal Care and Use Committee (IACUC).
of surface modi cations have shown promising re- Preoperative Care
sults, but the nal interpretation is yet to be deter-
mined [7]. Therefore, we embarked on an in vivo The adult sheep were transported from a pasture
study to determine the true potential of these two farm and housed in a sheep pen in the vivarium facil-
processes. ities. The vivarium staff was instrumental in the care
We rst had to establish a reliable animal model of the sheep prior to the operative procedure and
that would satisfy our requirements in studying the postoperatively. All the necessary materials, medica-
surface modi cation of mechanical heart valves. In tions, and equipment were properly prepared the day
choosing the proper animal model, the sheep was the prior to surgery (Tables 1 and 2). Appropriate feed
most appropriate. As the literature supports, sheep and water were dispensed, and the sheep were fasted
are an excellent animal model, since they are compa- for approximately 12 h prior to surgery.
56 M. M. PUC ET AL.
New Methodologies

TABLE 1 Medications tus position, and prepped and draped in sterile fas-
Anesthetic hion. A left lateral thoracotomy was performed
agents Antibiotics Analgesics Euthanasia through the 5th intercostal space. The descending
Ketamine Ampicillin Meperidine HCL Somelethal thoracic aorta was carefully exposed. The poste-
Thiamylalal Gentamycin Buprenorpherine rior spinal arteries were isolated, so not to be in-
sodium HCL corporated later into the partially occluding aortic
An intravenous bolus of unfractionated heparin
Operative Procedure sodium (150 U/kg, Elkins-Sinn, Inc., Cherry Hill,
NJ) was given at this time. After 5 min, the descend-
At the time of surgery, anesthesia was induced ing aorta was partially cross-clamped distal to the
using intramuscular (im) injections of ketamine (30 left subclavian artery using atraumatic, partially oc-
45 mg/kg, Parke-Davis, Morris Plain, NJ) and thi- cluding vascular clamps. A valved conduit housing
amylalal sodium (814 mg/kg). Once adequately se- a bilea et tilted disk (number 23 St. Jude Medical,
dated, the vivarium staff moved the sheep to the op- St. Jude Medical, Inc., St. Paul, MN) valve, with
erating room and properly positioned the animal for an appropriately sized portion of woven, gelatin-
intubation. After orotracheal intubation, the animal impregnated, tube graft (Gelweave, Sulzer Vascutek
was maintained on 24% halothane gas (Halocar- USA, Inc., Austin, TX) sutured to the back end, was
bon Laboratories, River Edge, NJ) volatilized with anastomosed end-to-side to the isolated segment of
oxygen. Tidal volume was set at 10 mL/kg plus ad- the proximal descending thoracic aorta (Figure 1).
ditional volume for the dead space within the tub- The end-to-side anastomosis was performed at a 30
ing circuit. The respiratory rate was maintained be- to 45 degree angle to minimize turbulence [17]. After
tween 7 and 20 breaths/min. Electrocardiographic the anastomosis was completed using 4-0 mono la-
leads were placed to monitor the heart rate and the ment Prolene ( Johnson & Johnson Co., Somerville,
cardiac rhythm. An intravenous catheter was inserted NJ) sutures, the partial aortic cross-clamp was reposi-
percutaneously into the cephalic vein. Normal tioned to the supradiaphragmatic area for the distal
saline (0.9%) was used for maintenance uid at a anastomosis, while the valved conduit was temporar-
rate of 10 mL/kg. A femoral arterial line for contin- ily clamped. Following completion of this anasto-
uous arterial blood-pressure monitoring was placed mosis, the aortic clamps were removed and ow was
via a cut-down method. Intravenous doses of gen- established through the graft. This essentially cre-
tamycin (2 mg/kg, Elkins-Sinn, Inc., Cherry Hill, ated a continuous tube graft in parallel with the de-
NJ) and ampicillin sodium (7 mg/kg, Bristo-Myers scending thoracic aorta with a centrally positioned
Squibb Company, Princeton, NJ) were then given. valve (Figure 2). Once adequate ow was con rmed
The animal was placed in the right lateral decubi- through the valved conduit, the native aorta was

TABLE 2 Materials and equipment

0.9% Sodium chloride Vaseline gauze Oxygen tank
Sterile water Chest tube number 36 Fr Arterial line monitor
Angiocatheters Endotracheal tubes Suction pump
iv Sets and tubing Laryngoscope Ventilator
Surgicels Sutures Capnometer
Absorbable sponges Heating pad Pulse oximeter
Povidone iodine Bovie unit Pleurovac
Nu Gauze sponges Prep razors EKG leads/monitor
Combine dressing Gowns, caps, gloves Temperature probe
Dacron tube graft Umbilical tape Ultrasound machine
Atraumatic partially occluding vascular clamps
Valved conduit (St. Jude Medical)


New Methodologies

(Ethicon, Somerville, NJ) suture. The animals was

then allowed to awake from anesthesia, and once
fully alert they were immediately extubated and the
arterial and venous lines were removed. The chest
tubes remained in place for a minimum of 2 h post-
operatively and were removed when the hourly chest
tube drain output had fallen to an acceptable level
and there was no air leak detected.

Postoperative Care
The animals continued their recovery in the vi-
varium sheep pens. For postoperative pain manage-
ment, meperidine (26 mg/kg, im, every 23 h;
Wyeth Laboratories, Inc., Philadelphia, PA) and
buprenorpherine (0.005 mg/kg, im, every 12 h;
Rickett &Coleman, Inc., Richmond, VA) were given
for 34 days. After this period, the narcotics were
withheld unless the sheep demonstrated signs of dis-
comfort or pain. The sheep received a 10-day course
of ampicillin sodium (7 mg/kg, sq) and gentamycin
(2 mg/kg, sq) that was administered twice a day.
There was no postoperative anticoagulation given.
Once the animals recovered completely from
surgery without ongoing complications and com-
pleted their postoperative antibiotic regimen, they
FIGURE 1 Diagram of the customed tube graft with a cen- were placed in a transport cage and taken back to the
trally located mechanical heart valve anastomosed to the de-
scending thoracic aorta with two partially occluding vascular
pasture farm.
clamps in place.

ligated with surgical umbilical tape ( Johnson & The sheep remained on the farm until they re-
Johnson Co., Somerville, NJ). Absorbable oxidized turned for sacri ce and explantation of the valves.
regenerated cellulose (Johnson & Johnson Medical Sonographic evaluation of the valves and conduits
Inc., Arlington, TX) was applied to control bleeding were performed prior to explantation. Following re-
at the suture lines. After hemostasis was achieved, moval of the valved conduits, the sheep were eutha-
a number 36 French chest tube (Argyle, Sherwood nized with a lethal dose of Somelethal (J. A. Webster,
Medical, St. Louise, MO) was placed through a sepa- Inc., Sterling, MA).
rate stab incision, connected to a pleurovac (Atrium
Medical Corp., Hudson, NH), and placed to low-
wall suction.
The thoracotomy incision was closed in a lay-
ered fashion. First, the ribs were reapproximated with The mean weight of the adult sheep was 33.8 kg
interrupted 0 Vicryl (Ethicon, Somerville, NJ) su- (range 24 to 43 kg). All of the sheep had a palpable
ture. The muscle and fascia were closed in two layers pulse in the conduit and an audible valvular click
with running 2-0 Vicryl (Ethicon, Somerville, NJ) prior to closure of the thoracotomy incision. It is
suture. The skin was closed with interrupted 3-0 silk important to note that none of the sheep showed

58 M. M. PUC ET AL.
New Methodologies

FIGURE 2 In situ photograph of the customed tube graft. The suture line indicates the position of the mechanical heart valve.

evidence of postoperative paralysis following a me- uneventful perioperative course. It apparently died
dian partial aortic cross-clamp time of 39 min. In from a cardiac arrhythmia. Prior to both of these
addition, all the animals returned to the farm on deaths, there were no air leaks appreciated from the
postoperative day 10 and were all noted to have an chest tubes in either sheep. Also, at necropsy, neither
audible valvular click. animal had a signi cant hemothorax, and the valved
Two sheep died shortly postoperatively (Table 3). conduits were intact and patent with preserved valve
One death was attributed to intraoperative compli- function.
cations. In this case, the proximal anastomosis tore The results of the four survivors are shown in
posteriorly and had to be redone. However, the sheep Table 4. These sheep had an uneventful operative
suffered a signi cant blood loss. Despite adequate course and fully recovered postoperatively. They re-
uid resuscitation, the sheep never fully recovered turned to the pasture until the time of sacri ce. The
from this insult and died 2 h postoperatively. The sheep returned for sacri ce and explantation of valves
second sheep died 4 h after extubation following an at a median of 37 days (range 30 to 46 days). Sono-
graphic evaluation of the valves prior to sacri ce
TABLE 3 Postoperative deaths (n D 2) showed no active movement of the lea ets in any of
the four sheep, but there was noted to be good blood
Weight EBL Operative CC time
Animal (kg) (ml) time (min) Complications (min) ow through the valved conduits. At explantation,
clot formation was found at the hinges of the lea ets
1 24 200 170 Arrhythmia 42
2 36 1000 165 Hypotension 55 that had prevented active movement. The conduits
Note. EBL, estimated blood loss; CC time, partial aortic cross-clamp themselves were fully patent with intact proximal
time. and distal anastomotic sites.


New Methodologies

TABLE 4 Living sheep (n D 4 )

Animal Weight (kg) EBL (ml) OR time (min) CC time (h) Complications CC time (min) POD explant. (day)
1. 43 150 270 18 Air leak 37 46
2. 35 300 240 2 None 41 37
3. 31 200 170 2 None 35 34
4. 34 300 165 2 None 36 30
Note. EBL, estimated blood loss; OR time, total operative time; CT time, total time chest tube remained in place postoperatively; CC time, partial
aortic cross-clamp time; POD explant., postoperative day of explantation of valved conduit.

DISCUSSION tained a mechanical heart valve without completely

occluding the aorta for any period of time. This tech-
This study was initiated to establish a reliable in nique allows adequate distal perfusion and does not
vivo model for studying the thrombogenicity of interrupt critical blood ow to the spinal cord. Our
mechanical heart valves, since the currently avail- median partial cross-clamp time of 39 min (range
able models did not meet our particular require- 35 to 55 min) produced no postoperative paralysis.
ments [15, 16, 1821]. We had three main issues Even the sheep with a partial cross-clamp time of
that we wanted to adhere to: systemic circulation, 55 min had no paraplegic sequela postoperatively
avoid cardiac bypass, and avoid complete aortic (Table 3).
occlusion. The decision to use sheep was rather simple. Upon
In an attempt to minimize experimental variables, review of the literature, sheep have a relatively large
we decided against placing the valve in the venous aorta compared to the other animal models, such
circulation, as reported by Yang et al. in which a heart as dogs, pigs, or calves. This makes sheep physi-
valve was positioned within the inferior vena cava of cally more appealing, along with their similar phys-
a sheep [21]. The different hemodynamic properties iologic parameters (e.g., heart rate and blood pres-
that exist between the venous and arterial circula- sure) as reported by Barnhart and associates [10].
tions will affect the rate and pattern of thrombus for- More importantly, as suggested by Tillman and co-
mation. For this reason, we wanted to place the valve workers, the coagulation pro le of sheep is similar to
into the arterial circulation, to reproduce the true en- that of humans [ 11]. This is critical when studying
vironment in which the valve will need to function. the effects of thrombogenicity of mechanical heart
Obviously, the ideal place is an intracardiac position. valves.
However, we wanted to eliminate the need for car- Overall, this study established a reliable and repro-
diac bypass and its associated risks of cerebrovascular ducible animal model to enable continued research
accidents, end-organ injury, and hemorrhagic com- in identifying a solution for eliminating the need
plications, to name a few. Chanda et al. has described for lifelong anticoagulation with mechanical heart
placing a bioprosthetic heart valve in the descending valves.
thoracic aorta of juvenile sheep, which would satisfy
our rst two requirements but necessitates complete CONCLUSION
aortic cross-clamping. Previous sheep studies in our
laboratory have shown that sheep have a poor toler- This novel technique provides an acceptable alter-
ance for complete aortic occlusion. We experienced native to the orthotopic placement of heart valves for
sheep paralysis following total cross-clamp times of the study of innovative mechanical heart valve ma-
as little as 16 min. Personal communications with terials and designs. We developed an ef cient model
other investigators have witnessed sheep sustaining that enabled us to reach our endpoints. The mechani-
paralysis following only 10 min of complete aortic cal heart valve was placed within the systemic circula-
cross-clamp time. Our solution to this problem was tion with avoidance of cardiac bypass and essentially
to only partially occlude the aorta with a side-biting eliminated the risk of paraplegia. We strongly advo-
technique as illustrated in Figure 1. We were then cate this model for studying the thrombogenicity of
able to bypass the aorta with a tube graft that con- mechanical heart valves.

60 M. M. PUC ET AL.
New Methodologies

REFERENCES 11. Tillman P, Carson SN, Talken L. Platelet function and coag-
ulation parameters in sheep during experimental vascular
1. Vongpatanasin W, Hillis LD, Lange RA. Prosthetic heart surgery. Lab Anim Sci. 1981;31:263 267.
valves. N Engl J Med. 1996;335:407416. 12. Irwin E, Lang G, Clack R, et al. Long-term evalua-
2. Edmunds LH. Thrombotic and bleeding complications of tion of prosthetic mitral valves in sheep. J Invest Surg.
prosthetic heart valves. Ann Thorac Surg. 1987;44:430 1993;6:133141.
445. 13. Thurmon JC, Kumar A, Link RD. Evaluation of ketamine
3. Levine MN, Rakob G, Hirsch J. Risk of hemorrhage as- HCL as an anesthetic in sheep. J Am Vet Med Assoc.
sociated with long-term anticoagulant therapy. Drugs. 1973;162:293 297.
1985;30:444 460. 14. Waterman AE, Livington A. Some physiologic effects of
4. Tran HS, Puc MM, Hewitt CW, et al. Diamond-like car- ketamine in sheep. Eur Surg Res. 1983;25:225233.
bon coating and plasma or glow discharge treatment of 15. Chanda J, Kuribayashi R, Abe T. New-generation valved
mechanical heart valves. J Invest Surg. 1999;12:133 140. conduit: An experimental study. J Thorac Cardiovasc Surg.
5. Kumar N, Singh B, Devital S, et al. Development of mer- 1997;114:218 223.
metic scratch resistant diamond like carbon coatings for 16. Chanda J, Kuribayashi R, Abe T, et al. Is the dog a useful
silica optical bers. Univ PA Mater Sci. 1997;1011(4):10 model for accelerated calcication study of cardiovascular
14. bioprostheses. Artif Organs. 1997;21:391 395.
6. Aronsson BO, Lausmaa J, Kasemo B. Glow discharge 17. Rutherford RB. Vascular Surgery, 5th ed. Philadelphia:
plasma treatment for surface cleaning and modication of W. B. Saunders; 2000:484486.
metallic biomaterials. J Biomed Mater Res. 1997;35:49 18. Schoen FJ, Hirsch D, Bianco RW, Levy RJ. Onset and
73. progression of calcication in porcine aortic biopros-
7. Chrzanowski F, Goodman M, Tarnoff M, et al. The ef- thetic valves implanted as orthotopic mitral valve re-
fects of surface modication of mechanical aortic valves placements in juvenile sheep. J Thorac Cardiovasc Surg.
on thrombogenicity. Scanning. 1998;20:189190. 1994;108:880 887.
8. Ali ML, Kumar SP, Bjornstad K, Duran CMG. The sheep 19. Jones M, Eidbo EE, Hilbert SL, Ferrans VJ, Clark RE. The ef-
as an animal model for heart valve research. Cardiovasc fects of anticalcication treatments on bioprosthetic heart
Surg. 1996;4:543549. valves implanted in sheep. ASAIO Trans. 1988;34:1027
9. Salerno CT, Droel J, Bianco RW. Current state of in 1030.
vivo preclinical heart valve evaluation. J Heart Valve Dis. 20. Jones M, Eidbo EE, Hilbert SL, Ferrans VJ, Clark RE. An-
1998;7:158162. ticalcication treatments of bioprosthetic heart valves: In
10. Barnhart GR, Jones M, Isihara T, et al. Bioprosthetic valvu- vivo studies of sheep. J Cardiac Surg. 1989;4:6973.
lar failure: Clinical and pathological observation in an 21. Yang Y, Franzen SF, Olin CL. In vivo comparison of hemo-
experimental animal model. J Thorac Cardiovasc Surg. compatibility of materials used in mechanical heart valves.
1982;83:618 631. J Heart Valve Dis. 1996;5:532537.


Copyright of Journal of Investigative Surgery is the property of Taylor & Francis Ltd and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.