An eight-year experience with porcine bioprosthetic cardiac valves

J

THORAC CARDIOVASC SURG

91:910-917,1986

An eight-year experience with porcine bioprosthetic cardiac valves A total of 589 porcine bioprostheses were implanted in 509 patients from January, 1976, througb December, 1983. Of the valves implanted, 390 were Hancock and 199 were Carpentier-Edwards. A total of 1,633 patient-years was accrued, with a mean follow-up of 38 months per patient. Two hundred eight patients had aortic valve replacement, 209 had mitral valve replacement, and 79 had multiple valve replacements, of which 46 were aortic and mitral replacements. The mortality for isolated aortic valve replacement was 5.8%; for isolated mitral replacement, 8.6%, and for aU patients, 10.9%. Late mortality was 3.9 % per patient-year. The actuarial survival rate at 5 years was 79 % for aortic, 68 % for mitral, and 76% for aortic-mitral valve replacement. There were 12 thromboembolic events (0.73% per patient-year), Two episodes occurred in patients with an aortic bioprosthesis, nine in patients with a porcine mitral valve, and one in a patient with mitral and tricuspid bioprosthetic valves. The probability of remaining free of thromboembolism at 5 years was 99 % for the group having aortic valve replacement, 93% for those having mitral replacement, and 100% for the group having aortic-mitral valve replacements. Thirteen episodes of endocarditis occurred (0.8 % per patient-year). Seven of the 13 patients died as a direct result of endocarditis. The probability of remaining free of prosthetic endocarditis at 5 years was 97 % for the aortic valve replacement group, 95 % for the mitral group, and 97 % for the aortic-mitral group. There were 20 instances of xenograft failure (1.2 % per patient-year). The probability of remaining free of valvefailure at 5 years was 96 % for the aortic valvereplacement group, 93 % for the mitral group, and 93 % for the aortic-mitral replacement group. Primary tissue failure of a prosthesis occurred in seven patients, aU with Hancock valves (0.43 % per patient-year), As yet there has been no primary tissue failure of the Carpentier-Edwards prosthesis. There also appears to bea lower incidence of thromboembolism (Edwards, 0.3 % per patient-year; Hancock, 0.8 % per patient-year) and endocarditis (Edwards, 0.6% per patient-year; Hancock, 1.0% per patient-year). The low incidence of complications with the porcine bioprosthetic valve, especially the Carpentier-Edwards, encourages us to recommend its continued use, especially in situations in which anticoagulation is contraindicated.

Renee S. Hartz, M.D., Elysa B. Fisher, M.D., Betsy Finkelmeier, R.N., M.S., Arthur DeBoer, M.D., John H. Sanders, Jr., M.D., John M. Moran, M.D., and Lawrence L. Michaelis, M.D., Chicago, Ill.

Reine bioprosthetic cardiac valves have been implanted with increasing frequency since 1970. Abundant evidence has accumulated to show that most sizes perform well hemodynamically and that the risk of thromboembolism without anticoagulation is low enough that most patients can be spared lifelong antico-

From the Department of Surgery, Division of Cardiothoracic Surgery, Northwestern University Medical School, Chicago, Ill. Received for publication May 16, 1985. Accepted for publication July 23, 1985. Address for reprints: Renee S. Hartz, M.D., 303 East Chicago Ave., Ward Building 9-105, Chicago, Ill. 60611.

910

agulation with sodium warfarin.'? The most significant unanswered question involves the long-term durability of porcine bioprostheses. Recent data document excellent short-term durability, 1,6-13 and none of these reports proves that the valve fails increasingly with time, except in children. With improvements in the manufacture and preservation of procine valves, primary tissue failure is sufficiently unusual that the continued use of the bioprosthesis is justified." Because most of the accumulated data deal only with the Hancock prosthesis, and because more long-term studies are needed to fully evaluate porcine valve longevity, we analyzed patients with both Hancock and Carpentier-Edwards bioprostheses.

Volume 91

Porcine bioprosthetic cardiac valves

Number 6 June, 1986

100

Table I. Summary of operative procedures Procedure

AVR MVR TVR PVR AVR-MVR MVR-TVR AVR-MVR-TVR

344

137 138

Preop NYHA (%pls)

5

60

1 35 24

0

4

7 11 4 1

Legend: AVR, Aortic valve replacement. MVR, Mitral valve replacement. TVR, Tricuspid valve replacement. PVR, Pulmonary valve replacement. 'Includes one or more of the following: coronary bypass grafting, ventricular aneurysm resection, endocardial resection, ascending aortic reconstruction, valvuloplasty, repair of atrial or ventricular septal defect.

Materials and methods Data base. A total of 589 porcine bioprosthetic valves were implaned in 509 patients during an 8 year period: January, 1976, through December, 1983. Before 1980 the Hancock valve was used almost exclusively (N = 390). Subsequently the Carpentier-Edwards prosthesis (N = 199) was used because of its larger primary orifice':" and because of reports of early structural degeneration of the Hancock valve." These 509 patients represent 84% of all patients undergoing cardiac valve replacement at Northwestern Memorial Hospital. Mechanical prostheses, primarily the St. Jude Medical valve, were reserved for use in young patients, those with a small aorticanulus, and those patients who would require long-term anticoagulation regardless of the prosthesis employed. Three patients in this series received a porcine mitral valve and a St. Jude Medical aortic valve. Aortic valve replacement (AVR) alone or in combination with another procedure was performed in 208 patients, 209 patients had mitral valve replacement (MVR) alone or in combination, and 79 had multiple valve replacements, 46 of which were combined aortic and mitral valve replacements (AVR-MVR). Thirtytwo percent of patients had concomitant procedures in addition to valve replacement (Table I). The mean age was 59 years (range 19 to 88 years) and 54% were men. Follow-upwas considered complete in 91.5% of patients; the remaining 8.5% were known to be alive at most recent follow-up, which was 1 year or more before the end point of data collection. A total of 1,633 years of

TIL II

I

20

165

71 71

ill

40 20

Combined valve replacement*

AVR MVR TVR AVR-MVR MVR-TVR AVR-MVR-TVR

80

No. of patients

Isolated valve replacement

9 11

POSIOp NYHA (% pts )

40 60 80 100

Fig. 1. Comparison of preoperative and postoperative New York Heart Association (NYHA) functional class.

follow-up was accrued (mean 3.25 years per patient). The maximum follow-up was 8.3 years, and 138 patients (27%) were followed up more than 5 years. The New York Heart Association functional class was determined in 507 patients preoperatively and was available at the time of this report in 94% of surviving patients. Before the operation 78% of patients were in Functional Class III or IV. In contrast, 73% of the patients were in Class I or II at most recent follow-up (Fig. 1). Perioperative considerations. Extracorporeal circulation with moderate hypothermia was used consistently, but myocardial preservation techniques changed in 1980. Before that time continuous coronary artery perfusion or crystalloid cardioplegia was employed. In 1980 cold blood hyperkalemic cardioplegia was gradually implemented, and this method is currently used for myocardial protection in all patients. The bioprosthesis was routinely prepared by rinsing it by hand for 3 minutes each in three separate basins of sterile saline solution. Care was taken not to touch the leaflets with any sharp instrument, and during suture placement the valve was frequently rinsed with saline. Pledgets were used only in the face of deficient or heavily calcified annular tissue. Seating of the mitral valve was checked by inserting an angled dental mirror, warmed to body temperature, through the orifice of the prosthesis to inspect the approximation of the sewing ring to the anulus and to verify the absence of strut entrapment by sutures. Anticoagulation. Warfarin sodium was administered to 46% of the patients postoperatively. In the absence of atrial fibrillation, left atrial enlargement or thrombus, or

The Journal of

912

Hartz et al.

Thoracic and Cardiovascular Surgery

100

---

80

IJ'-!2 ---.J

60

0:::

> >

1---1

40

0:::: ~

1fI 20

r

0' 0

RVR [N=208J ~ MVR (N=2091 13------{] RVR-MVR (N=461 & - - - &

2

5

3

6

7

YERRS Fig. 2. Actuarial survival rates of AVR, MVR, and AVR-MVR patients (including combined procedures).

Table

n. Operative mortality Operative procedure

Mortality* (%)

Isolated AVR Isolated MVR Isolated AVR-MVR AVR with or without other procedure MVR with or without other procedure Overall

5.8 8.6 11.4 6.2 14.3 10.9

*Death within 30 days of operation or before discharge from hospital.

Table m. Linearized rates of death and complications* AVR-MVR

Late mortality Thromboembolism Endocarditis Valve failure Primary tissue failure

2.50 0.00 0.64 2.50 1.30

All patients

3.90 0.73 0.80 1.22 0.43

'Pereent per patient-year, including combined procedures.

history of thromboembolism, anticoagulation was discontinued 6 weeks to 3 months after operation. Twentysix percent (8% of AVR patients and 41% of MVR patients) were maintained on long-term warfarin sodium. The operating surgeons recommended warfarin sodium only when there were compelling indications. When the surgeon or managing physician decided long-term warfarin anticoagulation was necessary, a prothrombin time 1Y2 times the control value was regarded as evidence of adequate anticoagulation. All of the patients not given warfarin sodium were treated with antiplatelet drugs (aspirin or aspirin and dipyridamole) for at least 6 to 12 weeks, unless there was a contraindication to their use. Statistical methods. All late morbidity and mortality rates are expressed as linearized occurrence rates and by actuarial analysis. Linearized occurrence rates were calculated by dividing the observed number of occurrences of a particular event (death or complication) by the total number of years of patient follow-up. These rates obviously represent an average over the entire length of the study period and are thus most appropriate

when the rate under consideration remains fairly constant over time. Actuarial analysis, which allows evaluation of time-associated morbidity and mortality, was performed according to the method described by Anderson and associates." All operative deaths were included in actuarial survival analysis. In evaluation of patient morbidity, complication-free deaths were treated in the same manner as patients lost to follow-up. Mortality. Operative mortality, defined as death that occurred within 30 days of operation or before discharge from the hospital, was 10.9% overall. The mortality rates were identical before and after 1980, when cold blood cardioplegia was implemented and when valve preference changed from the Hancock to the Carpentier-Edwards. Isolated AVR carried an operative mortality of 5.8% and a late mortality of 1.9% per patientyear. For isolated MVR the operative mortality was 8.6% and late mortality was 4.6% per patient-year. Isolated AVR-MVR resulted in an 11.4% operative mortality with a late mortality of 2.3% per patient-year.

Volume 91 Number 6 June, 1986

Porcine bioprosthetic cardiac valves

9 13

Table IV. Analysis of thromboembolic events Event

Severity

Pulmonary embolus Cerebral embolus T1A/digital embolus Aorto-iliac occlusion Brachial embolus TIA Cerebral embolus Poss. cerebral embolus T1A/CVA Retinal embolus Cerebellar infarct Retinal embolus

Major Major Minor Major Minor Minor Lethal Minor Major Major Major Major

Valve positionftype"

Warfarin

T/H M/H A/C-E M/H M/H M/H M/H M-T/H-H M/H M/H M/H A/H

No

Years postop. 0.1

Yes Yes

0.6 1.3 1.3 1.5 2.0 2.0 2.0 2.5/7 3.0 4.0 5.0

No Yes

No No Yes

No No Yes

No

Legend: TlA, Transient ischemic attack. CYA, Cerebrovascular accident. 'First letter indicates position (mitral = M, aortic = A, tricuspid = T); second letter indicates type (Hancock = H, Carpentier-Edwards = C-E).

Actuarial survival rates for AVR, MVR, and AVRMVR at 5 years were 79%, 68%, and 76%, respectively (Tables II and III and Fig. 2). Thromboembolism and anticoagulant-related hemorrhage. All patients were issued a questionnaire that was designed to elicit any episode of dizziness, clumsiness, numbness, severe headache, visual disturbance, loss of use of an extremity, or episode of bleeding. Any positive response was directly investigated by one of the authors, by contact with the patient, his family, or his physician. A thromboembolic event was considered to have occurred whenever limb, visceral, or cerebral ischemia could not definitely be related to a noncardiac cause. Events were classified as lethal, major, or minor, the distinction between the latter two being whether or not there was complete resolution of signs and symptoms within 24 hours' (Table IV). Patients who awoke from operation with a focal or generalized neurologic deficit were excluded from analysis of thromboembolic events. Thromboembolic events were documented in 12 patients, for a linearized rate of 0.73% per patient-year (Table III). Eight events occurred in patients with isolated MVR (1.35% per patient-year) and two in patients with isolated AVR (0.28% per patient-year). Thromboembolism did not occur in AVR-MVR patients, Five of the 12 events occurred in patients receiving warfarin sodium. The one fatal event occurred in a patient who was not receiving anticoagulants. The events occurred from 1 month to 7 years after operation; only one event occurred within the first 6 postoperative months. Eleven of the 12 thromboembolic events occurred in patients with the Hancock prosthesis (0.8% per patient-year) and only one in a patient with a Carpentier-Edwards valve (0.3% per patient-year). The actuarial incidence of being free of thromboembolism at

Table V. Primary tissue failure Valve position

Patient age (yr)

Years since valve implantation

Mitral

57

2.0

Mitral

31

2.5

Mitral Aortic

52 5\

4.0 5.6

Aortic

52

5.6

Mitral Mitral

19

50

7.6 7.75

Operative finding

Stiffened leaflets with fresh thrombus Leaflet dehisced from stent Calcified leaflets Valve detached. hole in leaflet One missing leaf, two calcified leaflets Two perforated leaflets Calcified and stenotic leaflets

5 years was 99.4% for AVR, 92.6% for MVR, and 97% for all patients (Table III, Fig. 3). Anticoagulant-related bleeding occurred in five patients, only two of whom were in chronic atrial fibrillation. One patient in normal sinus rhythm with an aortic valve had an intracerebral hemorrhage 6 weeks postoperatively. Another patient in sinus rhythm with a mitral prosthesis had a major intracerebral hemorrhage necessitating craniotomy 3 months after the operation. Of the other three patients, two were in chronic atrial fibrillation and one had a history of thromboembolism. None of the bleeding problems in these three patients were major. The linearized rates of occurrence of anticoagulant-related hemorrhage in patients receiving warfarin sodium were 1.5% per patient-year for AVR patients and 1.3% per patient-year for MVR patients. Endocarditis. Two patients initially operated on for acute endocarditis were excluded from the analysis. Of the remaining 507 patients, 13 developed postoperative

The Journal of

914 Hartz et aJ.

Thoracic and Cardiovascular Surgery

-

..... ......

_::....~..:...-

w

-;

95

.:...._---

-"""-'-, ---..:....-

W 0::::

LL I

>--

91l

f-

o aJ 0::::

o

:z:::

85

Il'[Il

PTF )0----->( TE 13····i] END 6---A VF .---K 2

(N=509l 3

4

5

6

YERRS

Fig. 3. Actuarial morbidity-free rates. PFT, Primary tissue failure. TE, Thromboembolism. END, Endocarditis. VF, Overall valve failure.

endocarditis. Two of these patients were known drug addicts. Only one of the 13 episodes occurred within 60 days of operation and no patient had catastrophic valve failure. Six of the patients (46%) with late endocarditis died as a direct result of their infection (two patients in the operating room). Of the seven survivors, five were treated with antibiotics alone and two required reoperation and valve replacement. The rates of occurrence per patient-year were 0.8% for all patients, 0.65% for AVR, 1.18% for MVR, and 0.64% for AVR-MVR patients. The actuarial probability of being free of endocarditis at 5 years was 97% for AVR, 95% for MVR, 98% for AVR-MVR, and 96% for all patients (Table III, Fig. 3). Valve failure. Bioprosthetic failure was defined by criteria developed by Oyer and associates": (1) development of a new regurgitant murmur, unless proven to be periprosthetic in origin, and (2) reoperation or death because of embolic episodes, thrombotic occlusion of the valve, infective endocarditis, or confirmed hemodynamic valvular dysfunction. There were 20 instances of valve failure; five patients developed peribasilar leaks, eight had valve failure secondary to prosthetic endocarditis, and the remaining seven had primary tissue failure. No patient required reoperation for repeated thromboembolic episodes. The overall linearized rate of valve failure was 1.2% per patient-year (AVR, 1.2% per patient-year; MVR, 1.2% per patient-year; and AVR-MVR, 2.5% per patient-year). The chance of being free of valve failure

at 5 years was 96% for AVR, 93% for MVR, 93% for AVR-MVR, and 95% for all patients. The seven instances of primary tissue failure occurred from 2 to 7V2 years after operation. None of these patients had antecedent infection. The valves in four of the seven patients had definite calcific infiltration and stiffening of the leaflets. In one of these a leaflet was actually missing and in another the sinuses of the prosthesis were partially filled with thrombus. The latter patient had low cardiac output and was not receiving warfarin sodium. The thrombus was quite fresh and was presumed to have formed in the immediate preoperative period. There was no instance of a totally thrombosed prosthesis. In the remaining three patients the valves had leaflet perforations without evidence of calcification (Table V). Linearized rates of primary tissue failure were as follows: 0.4% per patient-year, all patients; 0.3% per patient-year, AVR; and 0.5% per patient-year, MVR. Primary tissue failure occurred in Hancock valves at a rate of 0.62% per patient-year. At the completion of this series, no primary tissue failures had occurred in Carpentier-Edwards valves. The probability of freedom from primary tissue failure after 5 years was 100% for AVR and AVR-MVR patients, 96% for MVR patients, and 99% for all patients (Table III, Fig. 3). Discussion One of the most crucial decisions facing cardiac surgeons on a day-to-day basis is the choice of prosthesis for a given patient requiring cardiac valve replacement. Certainly the prosthetic valve does not provide a cure but rather imposes a different form of disease, which should carry as little additional risk to the patient as possible. The porcine bioprosthesis represented a major advance in heart valve replacement because many patients, especially those in normal sinus rhythm, could subsequently be spared the hazards of lifelong anticoagulation with warfarin sodium. We have used porcine valves in the majority of patients undergoing valve replacement since 1976. We have not used the IonescuShiley pericardial valve. Although it offers the same theoretical advantages as the porcine bioprostheses, we have not been convinced that it is superior to the porcine xenograft. Our operative and late mortality rates are comparable to those of the other major porcine valve series." 6, 7,10,12,13 Our statistics for MVR are especially gratifying in view of the large number of our patients (32%) who had combined operative procedures. Some of these patients had left ventricular aneurysmectomy with endocardial resection to treat life-threatening ventricular rhythm

Volume 91 Number 6 June, 1986

disturbances. Resection of papillary muscle with insertion of a porcine mitral valve was sometimes required to ablate irritable foci." The inclusion of this high-risk group of patients can be expected to elevate both operative and late mortality. We believe our results are also adversely influenced by the choice of the St. Jude Medical prosthesis in younger patients. As stated by Miller," mortality is certainly a definitive measure of valve failure but it is far too crude a yardstick by which to judge overall valve-related prognosis and to compare one type of prosthesis to another. This brings into focus the important issue of durability of the porcine bioprosthesis. It is essential that the various types of valve failure be reported in a way that reflects duration of patient follow-up and allows for variance of occurrence rates over time. As emphasized by McGoon,z° criteria for evaluating valve-related complications such as thromboembolism should be specifically defined and standardized among reported series. Accordingly, we have documented a thromboembolic rate of 0.73% per patient-year, which compares favorably to that reported in several major series.v"7.10.12.13 Only one of the 12 thromboembolic episodes occurred within the first 3 months after operation. This contrasts with the experience of Oyer and associates, 12 who observed at least half their thromboembolic events in the early postoperative period. Because 45% of our patients were given warfarin sodium for 6 to 12 weeks, and virtually all of the remaining patients took antiplatelet drugs temporarily, it is possible that they were better protected from thromboembolism. Treatment with either warfarin sodium or antiplatelet medication may allow the sewing ring of the prosthesis to endothelialize without thrombus formation. Nonetheless, two of the episodes of anticoagulant-related hemorrhage occurred during this early period in patients who had no indication for warfarin other than the recent implantation of a cardiac valve. The issue of whether or not to prescribe warfarin sodium for patients with porcine valves, even temporarily, is still a controversial one. There is no proof that the incidence of thromboembolism is decreased by the long-term administration of warfarin sodium in patients with a porcine bioprosthesis. Because 42% of the episodes of thromboembolism were in patients receiving warfarin, and in consideration of the risks of long-term warfarin sodium administration, we believe that long-term use of this drug in patients with a porcine bioprosthesis should be recommended only in the presence of compelling clinical indications. Our current policy is to treat all patients with anti platelet drugs for 3 months, unless they have a mitral prosthesis in the

Porcine bioprosthetic cardiac valves

H'I11I

EN2

9 15

........ ········I········I········I

8111

W W

0::::

u,

6111

I

>-

l-

I--<

4111

0 I--<

CD 0:::: 0

HRN (N=3901 [-E (N=199)

2111

~

A······"

2::

111

111

2

3

5

6

YERRS Fig. 4. Actuarial valve-related morbidity-free rates for patients with Hancock (Han) and Carpentier-Edwards (C-E) prostheses.

presence of atrial fibrillation or have a history of thromboembolism. We cannot yet agree with investigators who believe that no porcine bioprosthesis requires warfarin anticoagulation, but we do concur with those who do not routinely administer anticoagulants to patients in atrial fibrillation who have a porcine aortic valve.II.21.23 This policy has evolved over the past few years, and the number of patients being discharged receiving warfarin sodium currently is less than that reported herein. There was only one thromboembolic event in the group with the Carpentier-Edwards prostheses. Although linearized rates are not amenable to standard inferential statistical analysis, it would appear from these data that there is a higher propensity for thromboembolic events with the Hancock prosthesis. The Hancock group has accrued a longer duration of patient follow-up years (1,276 versus 391), but the linearized rates should still provide useful information because the incidence of thromboembolism does not appear to be time related. Since many cases of endocarditis can be directly traced to dental manipulation," the fact that prosthetic valve endocarditis was a major cause of valve failure in our patients is of concern. All patients undergoing elective valve replacement are seen by our hospital dental service for preoperative evaluation. They do not undergo the valve operation until the oral cavity is free cf infection. Patients are counseled about dental prophylaxis before discharge from the hospital, are educated

9 16

The Journal 01 Thoracic and Cardiovascular Surgery

Hartz et al.

about the American Heart Association's recommendations," and are encouraged to see a dentist on a routine basis. Because this policy has been in effect for many years, we expected an incidence of prosthetic endocarditis lower than that reported elsewhere. In an effort to determine the results of our educational efforts, we surveyed 225 of our patients having porcine valves. Although 70% saw their dentist regularly, only 21% received intramuscular prophylaxis and a small percentage (9%) had major dental work with no prophylaxis whatsoever. Since adequate antibiotic prophylaxis should reduce the incidence of prosthetic endocarditis in patients not abusing intravenous drugs, this finding warrants attention. It would appear that a major educational effort needs to be directed toward dentists and dental students regarding the potential hazard of endocarditis for their patients with cardiac valve prostheses. Theoretically, the only form of valve failure that should be more common with porcine valves than with rigid prostheses is primary tissue dysfunction. There continue to be major differences of opinion regarding the incidence of primary tissue failure, especially in the mitral position. 9, 11, 26, 27 Pelletier and colleagues," reporting on the use of Carpentier-Edwards valves, reported a primary tissue failure rate of only 0.1% per patient-year at 5 years of follow-up. Miller," however, recently predicted that a full 30% of porcine valves would fail within 10 years, 20% because of primary tissue failure. As in most of the other large porcine valves series, he used one type of prosthesis (Hancock) exclusively. Craver and associates" believe that primary tissue failure is statistically more significant after 6 years of follow-up. In our patients the actuarial incidence of being free of any type of valve failure is 89% at 5 years for Hancock valves and 95% for Carpentier-Edwards valves. At the completion of this series, we had not yet seen a Carpentier-Edwards valve fail because of primary tissue dysfunction. These observations plus the actuarial figures just mentioned tend to suggest that the Edwards valve is failing less often than the Hancock (Fig. 4). Because of the shorter follow-up time for patients with Carpentier-Edwards valves, we do not believe statistical significance can be attached to this difference at present.

improvements in valve manufacture and preservation in the 1980s. 2. The incidence of thromboembolism is very low with the porcine bioprosthesis. The occurrence of early thromboembolism «3 months) in our patients is much lower than in other reported series, perhaps because of our policy of administering warfarin or antiplatelet drugs for 6 to 12 weeks after operation. 3. Long-term anticoagulation with warfarin sodium should be reserved for patients with compelling clinical indications and should not be routine prophylaxis in all patients with a porcine tissue valve. 4. Despite our extensive attempts at education, many of our patients and their dentists fail to follow the recommendations of the American Heart Association regarding antibiotic prophylaxis before dental work.

2 3 4

5

6 7

8

Conclusions

1. Carpentier-Edwards valves implanted since 1980 had lower incidences of thromboembolism and primary tissue failure than did Hancock valves. This fact must be interpreted cautiously in view of the differing lengths of patient follow-up. However, it may well represent

9

10

REFERENCES Borkon AM, McIntosh CL, Jones M, Lipson LC, Kent, KM, Morrow, AG: Hemodynamic function of the Hancock standard orifice aortic valve bioprosthesis. J THoRAc CARDIOVASC SURG 82:601-607, 1981 Cohn LH, Mudge GH, Pratter F, Collins JJ: Five- to eight-year follow-up of patients undergoing porcine heartvalve replacement. N Engl J Med 304:258-262, 1981 Gallo 1, Ruiz B, Duran CMG: Five- to eight-year followup of patients with the Hancock cardiac bioprosthesis, J THoRAc CARDIOVASC SURG 86:897-902, 1983 Oyer PE, Miller DC, Stinson EB, Reitz BA, MorenoCabral RJ, Shumway NE: Clinical durability of the Hancock porcine bioprosthetic valve, J THoRAc CARDIOVASC SURG 80:824-833, 1980 Williams JB, Karp RB, Kirklin JW, Kouchoukos NT, Pacifico AD, Zorn GL, Blackstone EH, Brown RN, Piantadosi S, Bradley EL: Considerations in selection and management of patients undergoing valve replacement with glutaraldehyde-fixed porcine bioprostheses. Ann Thorae Surg 30:247-258, 1980 Angell WW, Angell JD, Kosek JC: Twelve-year experience with glutaraldehyde-preserved porcine xenografts, J THoRAc CARDIOVASC SURG 83:493-502, 1982 Borkon Am, McIntosh CL, Von Rueden TJ, Morrow AG: Mitral valve replacement with the Hancock bioprostheses. Five- to ten-year followup, Ann Thorac Surg 32:127-137, 1981 Craver JM, Jones EL, McKeown P, Bone OK, Hatcher CR, Kandrach M: Porcine cardiac xenograft valves, Analysis of survival, valve failure, and explantation. Ann Thorac Surg 34:16-21,1982 Gallo JL, Ruiz B, Carrion MF, Gutierrez JA, Vega JL, Duran CMG: Heart valve replacement with the Hancock bioprosthesis. A 6-year review, Ann Thorac Surg 31:444449, 1981 Hancock Laboratories: Durabilityassessment of the Han-

Volume 91 Number 6 June, 1986

II

12

13

14

15

16

17

18

cock porcine bioprosthesis. A multicenter retrospective analysis of patients operated prior to 1975. April, 1980 Janusz MT, Jamieson WRE, Allen P, Munro AI, Miyagishima RT, Tutassura H, Burr LH, Gerein AN, Tyers GFO: Experience with the Carpentier-Edwards porcine valve prosthesis in 700 patients. Ann Thorac Surg 34:625633, 1982 Oyer PE, Stinson EB, Reitz BA, Miller DC, Rossiter SJ, Shumway NE: Long-term evaluation of the porcine xenograft bioprosthesis. J THORAC CARDIOVASC SURG 78:343350, 1979 Pelletier C, Chaitman BR, Baillot R, Gutieras Val P, Bonan R, Dyrda I: Clinical and hemodynamic results with the Carpentier-Edwards porcine bioprosthesis. Ann Thorae Surg 34:612-624, 1982 Carpentier A, Dubost C, Lane E, Nashef A, Carpentier S, Reiland J, Deloche A, Fabiani IN, Chauvaud S, Perier P, Maxwell S: Continuing improvements in valvular bioprosthesis. J THORAC CARDIOVASC SURG 83:27-42, 1982 Levine FH, Carter JE, Buckley MJ, Daggett WM, Akins CW, Austen WG: Hemodynamic evaluation of Hancock and Carpentier-Edwards bioprosthesis. Circulation 64:192-195,1981 Borkon AM, Mcintosh CL, Jones M, Roberts WC, Morrow AG: Inward stent-post bending of a porcine bioprosthesis in the mitral position. J THORAC CARDIOVASC SURG 83: 105-107, 1982 Anderson RP, Bonchek LI, Grunkemeier GL, Lambert LE, Starr A: The analysis and presentation of surgical results by actuarial methods. J Surg Res 16:224-230, 1974 Moran JM, Kehoe RF, Loeb JM, Frederiksen JW, Zheutlin TA, Sanders JH, Michaelis LL: The role of papillary muscle resection and mitral valve replacement in

Porcine bioprosthetic cardiac valves

9 17

the control of refractory ventricular arrhythmia. Circulation 68: 154-160, 1983 19 Miller DC: Late results with bioprosthetic valves, Syllabus of the American College of Surgeons' Postgraduate Course in Cardiac Surgery, 1983, pp 39-42 20 McGoon DG: The risk of thromboembolism following valvular operations. How does one know? J THORAC CARDIOVASC SURG 88:782-786, 1984 21 Geha AS, Hammond GL, Laks H, Stansel HC, Glenn WWL: Factors affecting performance and thromboembolism after porcine xenograft cardiac valve replacement. J THORAC CARDIOVASC SURG 83:377-384, 1982 22 Jamieson WRE, Janusz MT, Miyagishima RT, Munro AI, Tutassura H, Gerein AN, Burr LH, Allen P: Embolic complications of porcine heterograft cardiac valves. J THORAC CARDIOVASC SURG 81:626-631, 1981 23 Nunez L, Aguado MC, Ce\emin D, Iglesias A, Larrea JL: Aspirin or Coumadin as the drug of choice for valve replacement with porcine bioprosthesis. Ann Thorac Surg 33:354-358, 1982 24 Jaspers MT, Little JW: Infective endocarditis. A review and update. Oral Surg 57:606-615, 1984 25 Kaplan E, Bisno A, Derrick W, Facklam R, Gordis L, Houser H, Jackson W, Millard H, Shulman S, Taranta A, Wannamaker L: AHA Committee Report: Prevention of Rheumatic Fever. Circulation 55: 1-4, 1977 26 Hooshang B, Mallon S, Kaiser GA, Thurer RJ, Kieval J: Failure of Hancock xenograft valve. Importance of valve position (4- to 9-year followup). Ann Thorac Surg 36:246252, 1983 27 Hetzer R, Hill JD, Kerth WJ, Wilson AJ, Adappa G, Gerbode F: Thrombosis and degeneration of Hancock valves. Clinical and pathological findings. Ann Thorac Surg 26:317-322, 1978