- Email: [email protected]
Long-term clinical and hemodynamic evaluation of the Ionescu-Shiley pericardial xenograft and the Braunwald-Cutter and Björk-Shiley prostheses in the mitral position
Long-term clinical and hemodynamic evaluation of the Ionescu-Shiley pericardial xenograft and the Braunwald-Cutter and Bjork-Shiley prostheses in the mitral position Single mitral valve replacement was undertaken in 220 patients between March, 1971, and October, 1977. Bjork-Shiley prostheses (BS) were inserted in 42 patients, Braunwald-Cutter prostheses (BC) in 52, and Ionescu-Shiley pericardial xenografts (PX) in 126. No attempt was made at randomization. The actuarial survival rate was 81.9 ± 12.8 percent at 7 years for patients with BS prostheses, 41.7 ± 22.9 percent at 6 years for patients with BC prostheses, and 89.0 ± 9.3 percent for patients with PX valves 7 years following valve replacement. Of the late deaths in patients with BC prostheses, 62.5 percent were valve related. The incidence of thromboembolism was 4.7, 1.8, and 1.5 episodes per 100 patient-years in the BS, BC, and PX groups, respectively. Long-term anticoagulation was used only in patients with BS and BC prostheses. Late postoperative hemodynamic studies were performed in six patients each with BS and BC prostheses and in 29 patients with PX valves. The mean diastolic gradients at rest were 6.2, 8.3, and 6.4 mm. Hg in the respective groups. The corresponding figures for calculated valve area were 1.8, 1.6, and 2.0 sq. cm. There was no statistically significant difference between the data recorded from the three groups of patients except for the survival rate of the BC prosthesis and the PX valve. The BS prosthesis and the PX valve have similar durability and hemodynamic performance for an almost identical duration of follow-up. Because of the improved quality of life and reduced morbidity without anticoagulants, we are using the PX valve exclusively for heart valve replacement^^
Anand P. Tandon, M.D., M.R.C.P. (by invitation), Somen M. Sengupta, M.B., F.R.C.S. (by invitation), Laszlo Lukacs, M.D. (by invitation), and Marian I. Ionescu, M.D., F.A.C.S. (by invitation), Leeds, England Sponsored by Harris B Shumaker, Jr., M.D., Indianapolis, Ind.
O i n c e the first successful clinical implantation of ball-valve prostheses in the mitral position,1 a number of valve substitutes have been developed and made available to the cardiac surgeon. A long-term prospective randomized clinical trial on a large scale is required to establish which is the best choice. To date no such study has been carried out, and the possibility of having such large groups of patients from one institution over a reasonably short period of time is rather remote.
Since March, 1971, Bjork-Shiley (BS) and Braunwald-Cutter (BC) prostheses and glutaraldehydestabilized Ionescu-Shiley pericardial xenograft (PX) valves have been used for isolated mitral valve replacement at our institution. The present study reports the clinical and hemodynamic evaluation of 220 patients who underwent mitral valve replacement between March, 1971, and October, 1977.
From the Departments of Cardiology and Cardiothoracic Surgery, The General Infirmary, Leeds, England. Read at the Fifty-eighth Annual Meeting of The American Association for Thoracic Surgery, New Orleans, La., May 8, 9, and 10, 1978. Address for reprints: Marian I. Ionescu, M.D., Department of Cardiothoracic Surgery, The General Infirmary, Great George Street, Leeds, LSI 3EX, England.
Of the 220 patients, 42 received a BS prosthesis, 52 had a BC prosthesis, and 126 patients had a PX valve. No attempt was made at randomization. Prior to 1976, when the PX valve became available from Shiley Laboratories, it was manufactured in small numbers in our hospital laboratory and its use in patients depended on availability. Concerning the choice of the two prosthe-
Patients and methods
0022-5223/78/120763+08$00.80/0 © 1978 The C. V. Mosby Co.
763
The Journal of Thoracic and Cardiovascular Surgery
7 6 4 Tandon et al.
Table n . Preoperative clinical details of 220 patients who underwent mitral valve replacement
Table I. Number, age, and sex of pat ients Valve substitute Bjork-Shiley Braunwald-Cutter Pericardial xenograft
Sex
No. of patients
Age range (yr.)
Male
Female
42 52 126
32-63 14-64 15-68
14 24 44
28 28 82
ses, the BC valve was inserted in patients considered to have a higher propensity for thromboembolism. The preoperative clinical details of the patients are outlined in Tables I and II. At the time of valve replacement 55 additional surgical procedures were performed in 46 patients, as shown in Table HI. Description of the valve substitutes. The BS prosthesis* consists of a low-profile Stellite valve housing with a sewing rim made of thin Teflon fabric.2 It contains a suspended tilting pyrolitic disc occluder that opens to 60 degrees. In the closed position the disc does not overlap the cage ring but fits within the orifice area, a design that is thought to diminish blood cell trauma. The BC prosthesist is a ball valve with a clothcovered cage. 3 The cage of the valve is open ended and the poppet is made of silicone elastomer improved by the use of a mild cure technique. An ultrathin polypropylene fabric mesh is used to cover the inflow orifice and a single layer of porous Dacron fabric tubing covers the struts. The Ionescu-Shiley pericardial xenograft^ consists of bovine pericardium mounted as a three-cusp valve onto a Dacron-covered titanium frame. The pericardium is stabilized with 0.5 percent purified glutaraldehyde and sterilized and stored in 4 percent buffered formaldehyde. It has a universal sewing rim.4 Operative technique. All valve replacements were performed by the same surgeon through a median sternotomy with the aid of euthermic conventional cardiopulmonary bypass. Intermittent hypoxia with local hypothermia was the routine procedure until the end of 1976. Since January, 1977, moderate total body hypothermia with cold cardioplegia and topical hypothermia has been used. The great majority of valves were inserted with continuous suture. The BS prosthesis was orientated with the larger opening of the disc toward the posterior wall of the left ventricle. Anticoagulant regimen. All patients with BS and BC valves received long-term oral warfarin sodium starting on the first postoperative day; the dose was *Manufactured by Shiley Laboratories, Irvine, Calif. tManufactured by Cutter Laboratories, Inc., Berkeley, Calif. ^Manufactured by Shiley Laboratories, Irvine, Calif.
BjorkShiley
BraunwaldCutter
Pericardial xenograft No.
No.
%
No.
3 33 6
7 79 14
6 34 12
11 66 23
16 86 24
13 68 19
13 5 19 5
31 12 45 11
11 7 32 2
21 13 62 4
44 27 51 4
35 21 41 3
9 33
21 79
9 43
17 83
24 102
19 81
N.Y.H.A. Class* I
n
in
IV Valve lesion Stenosis Incompetence Mixed disease Malfunctioning of previous valve substitutes Rhythm Sinus Atrial fibrillation
*New York Heart Association classification.
adjusted so as to maintain the prothrombin time 2.5 times the normal value. All these patients were regularly followed in a special anticoagulant clinic at this institution. The first 70 patients with PX received no prothrombin depressants during their follow-up. Since May, 1976, the 56 patients operated upon in this group received warfarin sodium for the first 4 to 6 weeks postoperatively. The prothrombin time was maintained at or slightly below twice the normal value. Hemodynamic study. Of the total number of 220 patients, 41 underwent postoperative cardiac catheterization and left ventricular angiography. The only criterion for selection was the informed consent of the patient. Six patients each from the groups with BS and BC valves were catheterized at mean durations of 22.7 ± 2.3 and 43.0 ± 2.8 months, respectively, following valve replacement. The valve sizes were 27 and 29 mm. in patients with BS prostheses and 32 and 34 in patients with BC prostheses. The 29 patients with mitral PX valves underwent hemodynamic investigations at a mean duration of 40.3 ± 2.8 months postoperatively. The valve sizes were 25 (two patients), 27 (15 patients), and 29 (12 patients). All patients were hospitalized 24 hours prior to the hemodynamic investigation. Right and left heart catheterization was performed in the postabsorptive state without any prior sedation. Pulmonary and systemic pressures were transduced by strain-gauge manometers (SEM* 486). Cardiac output was measured *Shields and Epstein Monitoring, Middlesex, England.
Volume 76 Number 6 December, 1978
765
Mitral valve replacement
Table III. Concomitant surgical procedures performed at the time of valve replacement in 220 patients Braunwald-Cutter
Pericardial xenograft
Total
5 1
3 2
-
-
5 11 1 9 2 1 2
13 14 1 18 3 4 2
31 (25)
55(46)
Bjork-Shiley
Surgical procedure Aorta-coronary SVBG Removal of left atrial thrombus Resection of hypertrophied LVOT Tricuspid annuloplasty Implantation of permanent pacemaker Closure of ASD Decalcification of left atrial wall
-
1
5 1 2
-
-
4
Totals
11(10)
13(11)
Legend: SVBG, Saphenous vein bypass graft. LVOT, Left ventricular outflow tract. ASD, Atrial septal defect. Figures in parentheses denote number of patients.
Table IV. Incidence of early mortality and follow-up of 220 patients Valve substitute Bjork-Shiley BraunwaldCutter Pericardial xenograft
Early deaths No.
Hospital
%
Table V. Details of late deaths in 203 hospital survivors
Follow-up (mo.) Range
Cause
Total
9.5 9.6
38 47
7-82 25-69
1,282 1,947
6.3
118
6-83
3,270
by the direct Fick method using 2 minutes of gas collection. Hemodynamic data were obtained during a 4 minute period of rest and between the fourth and sixth minutes of a 6 minute period of supine leg exercise on a bicycle ergometer. Left ventricular angiograms were performed in all patients at the end of the study. Pulmonary vascular resistance was calculated by the standard formula. The mean diastolic gradient across the mitral valve substitutes was measured by planifnetric integration of at least five simultaneously recorded phasic left ventricular and pulmonary wedge tracings. The effective valve surface area was calculated according to the formula of Gorlin and Gorlin.5 Results The definition of terms used is as reported in previous publications.6,7 Actuarial analysis of the data was carried out by the method of Anderson and associates.8 Standard statistical formulas were used for the analysis of results. Early death. There were 17 early deaths; four (9.5 percent) in the BS group, five (9.6 percent) in the BC, and eight (6.3 percent) in the PX group (Table IV). In none of these patients could the death be related to the valve substitute used. Follow-up. All 203 hospital survivors have been followed regularly at 6 month intervals and examined by
BjorkShiley
BraunwaldCutter
Pericardial xenograft
Valve related Bacterial endocarditis Cardiac Noncardiac Totals
5(11.9%)
8(15.4%)
6(4.!
one of us. In addition to the clinical history, estimation of functional capability, and physical examination, a chest radiograph and a 12-lead electrocardiogram were obtained in all patients at each visit. The total and range of follow-up are given in Table IV. Late deaths. The incidence and causes of late deaths are detailed in Table V. In the BS group the one valverelated death was due to thrombosis of the prosthesis in a 63-year-old woman who died soon after hospitalization of acute heart failure. Postmortem examination showed a thrombosed BS prosthesis. The second patient died 8 months postoperatively of biventricular failure while she was being treated with antibiotics for staphylococcal endocarditis. In the BC group 62.5 percent of late deaths were valve related. They were the consequence of bacterial endocarditis in two and gross wear and disruption of the cloth covering of the cage in three patients. Two of the deaths owing to mechanical failure occurred suddenly at home, and the third patient died before any definitive therapy could be instituted. All these patients underwent postmortem examination, and in none was there any evidence of recent or old myocardial infarction. None of the late deaths in patients with mitral PX valves were due to mechanical dysfunction of the valve substitute. Two patients with bacterial endocarditis were successfully reoperated upon but both died, 1 and
The Journal of
7 6 6 Tandon et al.
Thoracic and Cardiovascular Surgery
/o 100
SURVIVAL RATE
100
- • 89.0193%. -A 81.911287.
W
THROMBO EMBOLISM
1
90
958:21/ 932*53%
60 848184% 41.7122.9Z
40
80
lonescu-Shiley
lonescu-Shiley (1.5 episodes/100 patient years) Braunwald-Cutter (1.8 episodes/100 patient years)
Bjork-Shiley
Bjork-Shiley (4.7 episodes/100 patient years)
Braunwald - Cutter 0^
2
3
4
5
YEARS FOLLOW-UP
o-l
1
2
3
4
6
YEARS FOLLOW-UP
Fig. 1. Actuarial representation of survival rate in patients with mitral valve replacement. The data are expressed as percent expected survival rate ± standard error.
Fig. 2. Actuarial representation of thromboembolism free rate in patients with mitral valve replacement. The data are expressed as percent event free rate ± standard error.
3 months following reoperation, of myocardial infarction and heart failure, respectively. Actuarial analysis of the late results has shown an expected survival rate of 81.9 ± 1.28 percent at 7 years in patients with the BS prosthesis, 41.7 ± 22.9 percent at 6 years in patients with the BC prosthesis, and 89.0 ± 9.3 percent in patients with the PX valve 7 years following valve replacement (Fig. 1). At the 95 percent confidence limit, there is no significant difference between the survival rates in patients with BS and BC valves, but there is a significant difference (p < 0.05) in the survival rates of patients with BC and PX valves. Thromboembolism. Five patients (11.9 percent) with BS valves had nonfatal systemic emboli and one patient (2.4 percent) died of valve thrombosis. Four of the embolic episodes occurred in the first year following valve replacement, and the remaining one occurred in the second postoperative year. One patient had mesenteric artery embolization leading to gangrene of the bowel, which was successfully resected 7 months following valve replacement. The other four patients had cerebral emboli which left residual pareses in two patients. The incidence of thromboembolism is 4.7 episodes per 100 patient-years. Three patients (5.8 percent) with BC prostheses had cerebral emboli 3, 12, and 13 months following valve replacement. In two there was almost complete recovery but the third patient had residual weakness and is still subject to recurrent occular emboli. The incidence of thromboembolism in this group is 1.8 episodes per 100 patient-years. In all patients with BS and BC prostheses a correct level of anticoagulation was maintained at all times.
Of the patients with PX valves, four (3.2 percent) had embolic episodes 3 to 42 days postoperatively. All episodes were mild and left no sequelae. In the group of 56 patients operated upon since May, 1976, who received 4 to 6 weeks of anticoagulation, there have been no episodes of thromboembolism. Thus the incidence of embolization for the PX group is 1.5 episodes per 100 patient-years. Valve thrombosis has not been encountered in the entire PX series. Actuarial analysis shows a predicted thromboembolism-free rate of 84.8 ± 8.4 percent in patients with the BS prosthesis, 93.2 ± 5.3 percent in patients with the BC prosthesis, and 95.8 ± 2.1 percent in patients with the PX valve 7 years following valve replacement. There is no statistically significant difference in the incidence of thromboembolism among the three groups of patients. Clinical condition. The clinical condition of the patients was assessed according to the New York Heart Association (N.Y.H.A.) classification. The preoperative and the latest postoperative statuses of the 184 long-term survivors were compared and are shown in Fig. 3. At the latest follow-up examination 80 percent of patients with BS, 56.4 percent with BC, and 91.0 with PX valves were in N.Y.H.A. Class I. The mean values of N.Y.H.A. class for the three groups of patients were 3.06, 3.08, and 3.05 preoperatively and 1.23, 1.51, and 1.09, respectively, at the latest postoperative evaluation. The difference in postoperative N.Y.H.A. status between patients with BC and PX valves was statistically significant (p < 0.05). Hemolysis. Studies for chronic intravascular hemolysis were undertaken in 10 patients each with BS and
Volume 76
Mitral valve replacement
Number 6 December, 1978
767
INTRAVASCULAR HAEMOLYSIS 800-
N.Y.H.A. Status 3
\ \\ \\ \
ws,
B.S. valve
K> patients 6.3±0.7mths
V V
2
VX/''1'!,,
1-
BC
I I
B.S PX.
—— — 15.4 6.7 13.0
HI 61.5 80.0 68.0 IV
23.1 13.3 19.0
Pre-Op
B.C. valve
10 patients «.6±1.7mths 12 patients 10.7±1.2mths
□
i
400-
200-
3 8 4 167 9.0 2.6 3 3 — 2.6 — — Post-Op
A " N A Braunwald-Cutter valve: 52 patients. 3 9 long term survivors. ■■"■"'• Bjork-Shiley valve: 4 2 patients. 3 3 long term survivors. • ■ ^
l-S PX.
BC B.S PX 5 6 4 80.0 91.0
600-
L.D.H. I.U./L.
1.0
Haemopexin g/L (0.7-1.3) r-f-
2.0
t-S Pericardial Xenograft:126 patients. 112 long term survivors.
Fig. 3. Preoperative and latest postoperative New York Heart Association (N.Y.H.A.) status of the long-term survivors. The lines represent the mean of the individual groups. BC prostheses and in 12 patients with PX valves, at mean postoperative durations of 6.3 ± 0 . 7 , 16.6 ± 1.7, and 10.7 ± 1.2 months. The investigations performed included hemoglobin level, packed cell volume, reticulocyte count, total red blood cell count, fragmented red blood cell count, serum bilirubin, serum glutamic oxaloacetic transaminase level, serum lactic acid dehydrogenase level (LDH), hemopexin level, and haptoglobin estimation. Patients with mitral systolic murmur or with clinical suspicion of perivalvular leak were excluded from the study. The LDH hemopexin, and haptoglobin levels are depicted in Fig. 4. Patients with the BS valve had higher LDH and low serum hemopexin levels indicative of mild intravascular hemolysis. Patients with the BC valve had abnormally elevated LDH but normal levels of hemopexin and haptoglobin, also indicative of mild hemolysis. Patients with the PX valve did not show any evidence of hemolysis. Hemodynamic study. The results of the hemodynamic investigations are given in Tables VI and VII and in Fig. 5. All three mitral valve substitutes failed to restore normal hemodynamics in the 41 patients studied. Furthermore, no statistically significant difference was noted in the postoperative hemodynamic findings in the three groups of patients. The cardiac output at rest was near normal in patients with BS prostheses and normal in patients with BC and PX valves. However, the cardiac output response to exercise was abnormal in all patients studied. The mean
0.5-
Haptoglobin g'L (0.5-1.5)
i+i
i
(150-540)
1.0
Fig. 4. Serum lactic acid dehyrogenase (L.D.H.), hemopexin, and haptoglobin levels (mean values ± S.E.M.) in 32 patients with mitral valve replacement. pulmonary wedge pressure at rest was almost normal in patients with the mitral PX valve but was elevated in patients with BS and BC prostheses. The pulmonary artery and wedge pressures rose abnormally during exercise, the minimal rise being in patients with mitral PX valves. The mean diastolic gradients at rest were 6.2 ± 1.0, 8.3 ± 1.1, and 6.4 ± 0.5 mm. Hg in patients with BS, BC, and PX valves, respectively. The corresponding figures during exercise were 17.6 ± 0.6, 18.1 ± 1.3, and 15.3 ± 0.9 mm. Hg. The calculated valve surface areas at rest were 1.8 ± 0.3, 1.6 =t 0.1, and 2.0 ± T 0.1 sq. cm. in patients with the BS, BC, and PX valves, respectively. Although the surface area increased significantly during exercise in patients with BS prostheses and PX valves, only a marginal increase was noted in patients with BC prostheses (Fig. 5). The pulmonary vascular resistance was abnormally elevated both at rest and during exercise in all three groups of patients. Left ventricular angiography demonstrated mild incompetence in two patients with BS valves, but the rest of the prostheses and all 29 PX valves were competent.
The Journal of Thoracic and Cardiovascular Surgery
7 6 8 Tandon et al.
Table VI. Postoperative hemodynamic data (mean values ± S.E.M.) at rest in 41 patients with mitral valve replacement Variable Oxygen uptake (ml./min/sq. M.) Cardiac index (L./min/sq. M.) Mean pulmonary artery pressure (mm. Hg) Mean pulmonary wedge pressure (mm. Hg) Pulmonary vascular resistance (dynes sec. cm."' sq. M.) Mean diastolic gradient (mm. Hg) Calculated valve surface area (sq. cm.) No. of patients Mean duration postop. (mo.)
BjorkShiley
BraunwaldCutter
Pericardial xenograft
110.7 ±8.5 2.2 ±0.3 26.0 ±1.5 16.4 ±1.2 348.8 ±24.6 6.2 ±1.0 1.8 ±0.3
116.8 ±7.1 2.5 ±0.2 26.8 ±3.7 16.7 ±2.4 323.2 ±28.2 8.3 ±1.1 1.6 ±0.1
136.3 ±4.0 2.6 ±0.1 22.8 ±1.0 13.8 ±0.8 285.0 ±20.7 6.4 ±0.5 2.0 ±0.1
Discussion Three different mitral valve substitutes were used in the present series of 220 patients and follow-up data to 7 years are presented. Although this was not a randomized study, comparison of the preoperative clinical variables of the three groups of patients did not reveal any statistically significant differences. All the valve replacement procedures were undertaken by the same surgical team employing identical operative techniques. It is generally agreed that the hospital mortality rate is independent of the valve substitute employed. 9-11 None of the early deaths in the present series could be ascribed to the type of valve substitute used. Although the long-term survival following mitral valve replacement is related to a number of factors, such as age, the extent and duration of preoperative hemodynamic impairment, and coexistent myocardial and coronary artery disease, the complications directly related to various valve substitutes continue to contribute significantly to the over-all mortality rate." - 1 3 In the actuarial analysis of the long-term results of this series, the hospital mortality data have been excluded.14 The expected survival rate 7 years following valve replacement is 81.9 ± 12.8 percent for patients with BS prostheses and 89.0 ± 9.3 percent for patients with the PX valve, whereas with the BC prostheses the survival rate at 6 years was only 41.7 ± 22.9 percent. Valve-related complications were responsible for 62.5 percent of late deaths in patients with BC prostheses, whereas 40 percent and 33.3 percent of late deaths in
Table VII. Postoperative hemodynamic data (mean values ± S.E.M.) during exercise in 41 patients with mitral valve replacement Variable Oxygen uptake (ml./min./sq. M.) Cardiac index (L./min./sq. M.) Mean pulmonary artery pressure (mm. Hg) Mean pulmonary wedge pressure (mm. Hg) Pulmonary vascular resistance (dynes sec. cm. - 5 sq. M.) Mean diastolic gradient (mm. Hg) Calculated valve surface area (sq. cm.) No. of patients Mean duration postop (mo.)
BjorkShiley
BraunwaldCutter
Pericardial xenograft
390.5 ±24.3 4.3 ±0.6 48.4 ±4.5 34.2 ±2.1 264.2 ±32.3 17.6 ±0.6 2.2 ±0.4 6 22.7 ±2.3
365.7 ±20.7 4.3 ±0.3 48.5 ±6.2 32.5 ±3.6 297.7 ±24.6 18.1 ±1.3 1.7 ±0.1 6 43.0 ±2.8
406.5 ±14.4 4.1 ±0.2 42.0 ±1.9 29.4 ±1.5 260.3 ±26.3 15.3 ±0.9 2.3 ±0.1 29 40.3 ±2.8
patients with BS and PX valves, respectively, were valve related. Mechanical dysfunction of the prosthesis was responsible for three of eight late deaths in patients with the BC valve, whereas thrombotic occlusion was the cause of one of five late deaths in patients with BS prostheses. None of the late deaths in patients with PX valves was due to valvular dysfunction. Bacterial endocarditis was responsible for two valve-related deaths occurring in this group of patients. Systemic thromboembolism still remains the most significant complication associated with mitral valve replacement. The embolic rates for the three groups of patients in this series ranged from 3.2 for the PX group to 11.9 percent for patients with the BS prosthesis. The predicted fhromboembolism-free rate was 95.8 ± 2.1 percent in patients with PX valves and 93.2 ± 5.3 percent in patients with BC prostheses. Only 84.8 ± 8.4 percent of patients with the BS prosthesis are expected to be free of thromboembolism at 7 years' follow-up. Fatal valve thrombosis occurred in one patient with a BS prosthesis. Of the five embolic episodes in the BS group, three were associated with considerable morbidity; one patient underwent bowel resection and two continue to have residual neurologic deficit. All these embolic episodes occurred in the presence of a correct level of anticoagulation. All the embolic episodes in the PX group occurred in patients without any anticoagulant treatment and during the first 6 weeks following mitral valve replacement. Among the 56 patients operated upon since May, 1976,
Volume 76 Number 6 December, 1978
769
Mitral valve replacement
3L./min.Vm! X
g
T _
4"
mm. Hg. 15-
/A
-■'/
■■" /
O 2-
f R
T<
S
S
10-
5E
<
Cm.'
UJ
1
<25-
r^
UJ
5? S
UJ
I
R
E
_j
R
E
1-
* — - * Bjork-Shiley valve,(valve size 27,29), 6 patients 22.7*2.3 mths. m-—■■ Braunwald-Cutter valve,(valve size 32,34), 6 patients 4 3 . 0 * 2 . 8 mths. • — • l - S Pericardial Xenograft valve, (valve size 25,27,29), 29 patients 40.3±2.8 mths.
Fig. 5. Postoperative hemodynamic data (mean values valve replacement. R, Rest. E, Exercise. and receiving prothrombin depressants for 4 to 6 weeks, there have been no thromboembolic complications despite the suboptimal levels of anticoagulation. Valve thrombosis has not been encountered in patients with PX valves.15 Thus the PX valve is definitely superior to prosthetic valves in reducing the incidence of thromboembolism even without use of long-term anticoagulation. The benefits of a low embolic rate and the absence of valve thrombosis, without long-term anticoagulation, become even more significant when considering the life-threatening complications resulting from the use of prothrombin depressants per se.16, 17 In the present series fatal hemorrhage occurred in one patient each with BC (1.9 percent) and BS (2.4 percent) prostheses (Table VIII). The latest follow-up assessment of long-term survivors showed that 91 percent of patients with PX, 80 percent with BS, and 56.4 percent with BC prostheses were in N.Y.H.A. Class I. The hemodynamic investigations demonstrated significant circulatory improvement following mitral valve replacement in all patients studied. However, the investigations failed to reveal any significant difference among the results obtained with the three types of valve substitutes. A certain degree of obstruction to forward flow across the mitral valve was found with all three valve substitutes, and this was significant during exercise. The mean diastolic gradients were identical with BS and PX valves at rest, but during exercise the PX had a marginal advantage over the BS prosthesis. The most significant single factor in producing obstruction to forward flow is the presence of rigid rings in all currently used valve substitutes that do not expand during diastole and therefore encroach upon the normal functional mechanism of the mitral valve anulus.10
standard error of the mean) of 41 patients with mitral Table VLU. Long-term complication rate of the valve substitutes used for mitral valve replacement Valve substitute
Thrombosed Embolism (%) valve (%)
BjorkShiley BraunwaldCutter Pericardial xenograft
Fatal hemorrhage (%)
Late mortality (%)
2.4
11.9
2.4
11.9
-
6.8
1.9
15.4
-
3.2
-
4.8
In practical terms the most important hemodynamic determinant of a valve substitute is the effective surface area as measured in vivo. In the present study the largest surface areas were obtained in patients with PX valves, both at rest and during exercise.18 The BS prosthesis was found to have a surface area comparable to that of the PX valve, but that of the BC prosthesis was obviously suboptimal. Conclusions The BC prosthesis is associated with a comparatively low survival rate and with serious mechanical complications owing to wear of the cloth fabric. We have discontinued its use and are now in the process of electively replacing the remaining valves. Despite subjective impressions, there was no statistically significant difference between the clinical and hemodynamic performance of the BS prosthesis and the Ionescu-Shiley PX valve. The BS prosthesis and the PX valve offer similar mechanical durability and hemodynamic performance for an almost identical duration of follow-up. Use of the BS valve is associated with increased
770
Tandon et al.
morbidity despite the imperative need for permanent anticoagulation. Since the PX valve is associated with an improved quality of life and minimal thromboembolic complications in the absence of long-term anticoagulation, we are using this preparation exclusively for all heart valve replacements. We gratefully acknowledge the help given by Miss Ann Tunnicliffe in the preparation of this manuscript. REFERENCES 1 Starr A, Edwards ML: Mitral replacement. Clinical experience with a ball valve prosthesis. Ann Surg 154:726740, 1961 2 Technical information on Bjbrk-Shiley heart valves, Shiley Laboratories, Irvine, Calif., 1972 3 Braunwald NS, Tatooles C, Turina M, Detmer D: New development in the design of fabric-covered prosthetic heart valves. J THORAC CARDIOVASC SURG 62:673-682,
1971 4 Technical information on the Ionescu-Shiley pericardial xenograft heart valve, Shiley Laboratories, Irvine, Calif., 1977 5 Gorlin R, Gorlin SG: Hydraulic formula for calculation of the area of the stenotic mitral valve, other cardiac valves, and central circulatory shunt. Am Heart J 41:1-29, 1951 6 Ionescu MI, Pakrashi BC, Holden MP, Mary DAS, Wooler GH: Results of aortic valve replacement with frame-supported fascia lata and pericardidal grafts. J THORAC CARDIOVASC SURG 64:340-353, 1972
7 Mary DAS, Pakrashi BC, Catchpole RW, Ionescu MI: Tissue valves in the mitral position. Five year's experience. Br Heart J 37:1123-1132, 1975 8 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
The Journal of Thoracic and Cardiovascular Surgery
9 Simonsen S, Forfang K, Anderson A, Efskind L: Hospital mortality after mitral valve replacement. Prognostic significance of pre-operative clinical and haemodynmaic factors. Acta Med Scand 195:243-246, 1974 10 Brawley RK, Donahoo RK, Gott VL: Current status of the Beall, Bjbrk-Shiley, Braunwald-Cutter, LilleheiKasterand Smeloff-Cutter cardiac valve prostheses. Am J Cardiol 35:855-865, 1975 11 Salomon NW, Stinson EB, Griepp RB, Shumway NE: Mitral valve replacement. Long-term evaluation of prosthesis related mortality and morbidity. Circulation 56: (Suppl.):94-101, 1977 12 Angell WW, Angell JD, Sywak A: Selection of tissue or prosthetic valve. A five-year prospective, randomized comparison. J THORAC CARDIOVASC SURG 73:43-53,
1977 13 Leading article: Which prosthetic valve? Lancet 1:619620, 1976 14 Rahimtoola SH: Valve replacement—a perspective. Am J Cardiol 35:711-715, 1975 15 Ionescu MI, Tandon AP, Smith DR: Long-term clinical and hemodynamic evaluation of the Ionescu-Shiley pericardial xenograft heart valve. Presented at the Fourteenth Congress of the Pan Pacific Surgical Association, Honolulu, Hawaii, April 1-7, 1978 16 Hume M, Sevitt S, Thomas DP: Venous Thrombosis and Pulmonary Embolism. Cambridge, Mass., 1970, Harvard University Press, p 455 17 Isom OW, William CD, Falk EA, Spencer FL, Glassman E: Evaluation of anticoagulant therapy in cloth covered prosthetic valves. Circulation 47:Suppl 3:48-50, 1973 18 Tandon AP, Smith DR, Ionescu Ml: Hemodynamic evaluation of the Ionescu-Shiley pericardial xenograft in the mitral position. Am Heart J 95:595-601, 1978 (For discussion see page 785)
Anand P. Tandon, M.D., M.R.C.P. (by invitation), Somen M. Sengupta, M.B., F.R.C.S. (by invitation), Laszlo Lukacs, M.D. (by invitation), and Marian I. Ionescu, M.D., F.A.C.S. (by invitation), Leeds, England Sponsored by Harris B Shumaker, Jr., M.D., Indianapolis, Ind.
O i n c e the first successful clinical implantation of ball-valve prostheses in the mitral position,1 a number of valve substitutes have been developed and made available to the cardiac surgeon. A long-term prospective randomized clinical trial on a large scale is required to establish which is the best choice. To date no such study has been carried out, and the possibility of having such large groups of patients from one institution over a reasonably short period of time is rather remote.
Since March, 1971, Bjork-Shiley (BS) and Braunwald-Cutter (BC) prostheses and glutaraldehydestabilized Ionescu-Shiley pericardial xenograft (PX) valves have been used for isolated mitral valve replacement at our institution. The present study reports the clinical and hemodynamic evaluation of 220 patients who underwent mitral valve replacement between March, 1971, and October, 1977.
From the Departments of Cardiology and Cardiothoracic Surgery, The General Infirmary, Leeds, England. Read at the Fifty-eighth Annual Meeting of The American Association for Thoracic Surgery, New Orleans, La., May 8, 9, and 10, 1978. Address for reprints: Marian I. Ionescu, M.D., Department of Cardiothoracic Surgery, The General Infirmary, Great George Street, Leeds, LSI 3EX, England.
Of the 220 patients, 42 received a BS prosthesis, 52 had a BC prosthesis, and 126 patients had a PX valve. No attempt was made at randomization. Prior to 1976, when the PX valve became available from Shiley Laboratories, it was manufactured in small numbers in our hospital laboratory and its use in patients depended on availability. Concerning the choice of the two prosthe-
Patients and methods
0022-5223/78/120763+08$00.80/0 © 1978 The C. V. Mosby Co.
763
The Journal of Thoracic and Cardiovascular Surgery
7 6 4 Tandon et al.
Table n . Preoperative clinical details of 220 patients who underwent mitral valve replacement
Table I. Number, age, and sex of pat ients Valve substitute Bjork-Shiley Braunwald-Cutter Pericardial xenograft
Sex
No. of patients
Age range (yr.)
Male
Female
42 52 126
32-63 14-64 15-68
14 24 44
28 28 82
ses, the BC valve was inserted in patients considered to have a higher propensity for thromboembolism. The preoperative clinical details of the patients are outlined in Tables I and II. At the time of valve replacement 55 additional surgical procedures were performed in 46 patients, as shown in Table HI. Description of the valve substitutes. The BS prosthesis* consists of a low-profile Stellite valve housing with a sewing rim made of thin Teflon fabric.2 It contains a suspended tilting pyrolitic disc occluder that opens to 60 degrees. In the closed position the disc does not overlap the cage ring but fits within the orifice area, a design that is thought to diminish blood cell trauma. The BC prosthesist is a ball valve with a clothcovered cage. 3 The cage of the valve is open ended and the poppet is made of silicone elastomer improved by the use of a mild cure technique. An ultrathin polypropylene fabric mesh is used to cover the inflow orifice and a single layer of porous Dacron fabric tubing covers the struts. The Ionescu-Shiley pericardial xenograft^ consists of bovine pericardium mounted as a three-cusp valve onto a Dacron-covered titanium frame. The pericardium is stabilized with 0.5 percent purified glutaraldehyde and sterilized and stored in 4 percent buffered formaldehyde. It has a universal sewing rim.4 Operative technique. All valve replacements were performed by the same surgeon through a median sternotomy with the aid of euthermic conventional cardiopulmonary bypass. Intermittent hypoxia with local hypothermia was the routine procedure until the end of 1976. Since January, 1977, moderate total body hypothermia with cold cardioplegia and topical hypothermia has been used. The great majority of valves were inserted with continuous suture. The BS prosthesis was orientated with the larger opening of the disc toward the posterior wall of the left ventricle. Anticoagulant regimen. All patients with BS and BC valves received long-term oral warfarin sodium starting on the first postoperative day; the dose was *Manufactured by Shiley Laboratories, Irvine, Calif. tManufactured by Cutter Laboratories, Inc., Berkeley, Calif. ^Manufactured by Shiley Laboratories, Irvine, Calif.
BjorkShiley
BraunwaldCutter
Pericardial xenograft No.
No.
%
No.
3 33 6
7 79 14
6 34 12
11 66 23
16 86 24
13 68 19
13 5 19 5
31 12 45 11
11 7 32 2
21 13 62 4
44 27 51 4
35 21 41 3
9 33
21 79
9 43
17 83
24 102
19 81
N.Y.H.A. Class* I
n
in
IV Valve lesion Stenosis Incompetence Mixed disease Malfunctioning of previous valve substitutes Rhythm Sinus Atrial fibrillation
*New York Heart Association classification.
adjusted so as to maintain the prothrombin time 2.5 times the normal value. All these patients were regularly followed in a special anticoagulant clinic at this institution. The first 70 patients with PX received no prothrombin depressants during their follow-up. Since May, 1976, the 56 patients operated upon in this group received warfarin sodium for the first 4 to 6 weeks postoperatively. The prothrombin time was maintained at or slightly below twice the normal value. Hemodynamic study. Of the total number of 220 patients, 41 underwent postoperative cardiac catheterization and left ventricular angiography. The only criterion for selection was the informed consent of the patient. Six patients each from the groups with BS and BC valves were catheterized at mean durations of 22.7 ± 2.3 and 43.0 ± 2.8 months, respectively, following valve replacement. The valve sizes were 27 and 29 mm. in patients with BS prostheses and 32 and 34 in patients with BC prostheses. The 29 patients with mitral PX valves underwent hemodynamic investigations at a mean duration of 40.3 ± 2.8 months postoperatively. The valve sizes were 25 (two patients), 27 (15 patients), and 29 (12 patients). All patients were hospitalized 24 hours prior to the hemodynamic investigation. Right and left heart catheterization was performed in the postabsorptive state without any prior sedation. Pulmonary and systemic pressures were transduced by strain-gauge manometers (SEM* 486). Cardiac output was measured *Shields and Epstein Monitoring, Middlesex, England.
Volume 76 Number 6 December, 1978
765
Mitral valve replacement
Table III. Concomitant surgical procedures performed at the time of valve replacement in 220 patients Braunwald-Cutter
Pericardial xenograft
Total
5 1
3 2
-
-
5 11 1 9 2 1 2
13 14 1 18 3 4 2
31 (25)
55(46)
Bjork-Shiley
Surgical procedure Aorta-coronary SVBG Removal of left atrial thrombus Resection of hypertrophied LVOT Tricuspid annuloplasty Implantation of permanent pacemaker Closure of ASD Decalcification of left atrial wall
-
1
5 1 2
-
-
4
Totals
11(10)
13(11)
Legend: SVBG, Saphenous vein bypass graft. LVOT, Left ventricular outflow tract. ASD, Atrial septal defect. Figures in parentheses denote number of patients.
Table IV. Incidence of early mortality and follow-up of 220 patients Valve substitute Bjork-Shiley BraunwaldCutter Pericardial xenograft
Early deaths No.
Hospital
%
Table V. Details of late deaths in 203 hospital survivors
Follow-up (mo.) Range
Cause
Total
9.5 9.6
38 47
7-82 25-69
1,282 1,947
6.3
118
6-83
3,270
by the direct Fick method using 2 minutes of gas collection. Hemodynamic data were obtained during a 4 minute period of rest and between the fourth and sixth minutes of a 6 minute period of supine leg exercise on a bicycle ergometer. Left ventricular angiograms were performed in all patients at the end of the study. Pulmonary vascular resistance was calculated by the standard formula. The mean diastolic gradient across the mitral valve substitutes was measured by planifnetric integration of at least five simultaneously recorded phasic left ventricular and pulmonary wedge tracings. The effective valve surface area was calculated according to the formula of Gorlin and Gorlin.5 Results The definition of terms used is as reported in previous publications.6,7 Actuarial analysis of the data was carried out by the method of Anderson and associates.8 Standard statistical formulas were used for the analysis of results. Early death. There were 17 early deaths; four (9.5 percent) in the BS group, five (9.6 percent) in the BC, and eight (6.3 percent) in the PX group (Table IV). In none of these patients could the death be related to the valve substitute used. Follow-up. All 203 hospital survivors have been followed regularly at 6 month intervals and examined by
BjorkShiley
BraunwaldCutter
Pericardial xenograft
Valve related Bacterial endocarditis Cardiac Noncardiac Totals
5(11.9%)
8(15.4%)
6(4.!
one of us. In addition to the clinical history, estimation of functional capability, and physical examination, a chest radiograph and a 12-lead electrocardiogram were obtained in all patients at each visit. The total and range of follow-up are given in Table IV. Late deaths. The incidence and causes of late deaths are detailed in Table V. In the BS group the one valverelated death was due to thrombosis of the prosthesis in a 63-year-old woman who died soon after hospitalization of acute heart failure. Postmortem examination showed a thrombosed BS prosthesis. The second patient died 8 months postoperatively of biventricular failure while she was being treated with antibiotics for staphylococcal endocarditis. In the BC group 62.5 percent of late deaths were valve related. They were the consequence of bacterial endocarditis in two and gross wear and disruption of the cloth covering of the cage in three patients. Two of the deaths owing to mechanical failure occurred suddenly at home, and the third patient died before any definitive therapy could be instituted. All these patients underwent postmortem examination, and in none was there any evidence of recent or old myocardial infarction. None of the late deaths in patients with mitral PX valves were due to mechanical dysfunction of the valve substitute. Two patients with bacterial endocarditis were successfully reoperated upon but both died, 1 and
The Journal of
7 6 6 Tandon et al.
Thoracic and Cardiovascular Surgery
/o 100
SURVIVAL RATE
100
- • 89.0193%. -A 81.911287.
W
THROMBO EMBOLISM
1
90
958:21/ 932*53%
60 848184% 41.7122.9Z
40
80
lonescu-Shiley
lonescu-Shiley (1.5 episodes/100 patient years) Braunwald-Cutter (1.8 episodes/100 patient years)
Bjork-Shiley
Bjork-Shiley (4.7 episodes/100 patient years)
Braunwald - Cutter 0^
2
3
4
5
YEARS FOLLOW-UP
o-l
1
2
3
4
6
YEARS FOLLOW-UP
Fig. 1. Actuarial representation of survival rate in patients with mitral valve replacement. The data are expressed as percent expected survival rate ± standard error.
Fig. 2. Actuarial representation of thromboembolism free rate in patients with mitral valve replacement. The data are expressed as percent event free rate ± standard error.
3 months following reoperation, of myocardial infarction and heart failure, respectively. Actuarial analysis of the late results has shown an expected survival rate of 81.9 ± 1.28 percent at 7 years in patients with the BS prosthesis, 41.7 ± 22.9 percent at 6 years in patients with the BC prosthesis, and 89.0 ± 9.3 percent in patients with the PX valve 7 years following valve replacement (Fig. 1). At the 95 percent confidence limit, there is no significant difference between the survival rates in patients with BS and BC valves, but there is a significant difference (p < 0.05) in the survival rates of patients with BC and PX valves. Thromboembolism. Five patients (11.9 percent) with BS valves had nonfatal systemic emboli and one patient (2.4 percent) died of valve thrombosis. Four of the embolic episodes occurred in the first year following valve replacement, and the remaining one occurred in the second postoperative year. One patient had mesenteric artery embolization leading to gangrene of the bowel, which was successfully resected 7 months following valve replacement. The other four patients had cerebral emboli which left residual pareses in two patients. The incidence of thromboembolism is 4.7 episodes per 100 patient-years. Three patients (5.8 percent) with BC prostheses had cerebral emboli 3, 12, and 13 months following valve replacement. In two there was almost complete recovery but the third patient had residual weakness and is still subject to recurrent occular emboli. The incidence of thromboembolism in this group is 1.8 episodes per 100 patient-years. In all patients with BS and BC prostheses a correct level of anticoagulation was maintained at all times.
Of the patients with PX valves, four (3.2 percent) had embolic episodes 3 to 42 days postoperatively. All episodes were mild and left no sequelae. In the group of 56 patients operated upon since May, 1976, who received 4 to 6 weeks of anticoagulation, there have been no episodes of thromboembolism. Thus the incidence of embolization for the PX group is 1.5 episodes per 100 patient-years. Valve thrombosis has not been encountered in the entire PX series. Actuarial analysis shows a predicted thromboembolism-free rate of 84.8 ± 8.4 percent in patients with the BS prosthesis, 93.2 ± 5.3 percent in patients with the BC prosthesis, and 95.8 ± 2.1 percent in patients with the PX valve 7 years following valve replacement. There is no statistically significant difference in the incidence of thromboembolism among the three groups of patients. Clinical condition. The clinical condition of the patients was assessed according to the New York Heart Association (N.Y.H.A.) classification. The preoperative and the latest postoperative statuses of the 184 long-term survivors were compared and are shown in Fig. 3. At the latest follow-up examination 80 percent of patients with BS, 56.4 percent with BC, and 91.0 with PX valves were in N.Y.H.A. Class I. The mean values of N.Y.H.A. class for the three groups of patients were 3.06, 3.08, and 3.05 preoperatively and 1.23, 1.51, and 1.09, respectively, at the latest postoperative evaluation. The difference in postoperative N.Y.H.A. status between patients with BC and PX valves was statistically significant (p < 0.05). Hemolysis. Studies for chronic intravascular hemolysis were undertaken in 10 patients each with BS and
Volume 76
Mitral valve replacement
Number 6 December, 1978
767
INTRAVASCULAR HAEMOLYSIS 800-
N.Y.H.A. Status 3
\ \\ \\ \
ws,
B.S. valve
K> patients 6.3±0.7mths
V V
2
VX/''1'!,,
1-
BC
I I
B.S PX.
—— — 15.4 6.7 13.0
HI 61.5 80.0 68.0 IV
23.1 13.3 19.0
Pre-Op
B.C. valve
10 patients «.6±1.7mths 12 patients 10.7±1.2mths
□
i
400-
200-
3 8 4 167 9.0 2.6 3 3 — 2.6 — — Post-Op
A " N A Braunwald-Cutter valve: 52 patients. 3 9 long term survivors. ■■"■"'• Bjork-Shiley valve: 4 2 patients. 3 3 long term survivors. • ■ ^
l-S PX.
BC B.S PX 5 6 4 80.0 91.0
600-
L.D.H. I.U./L.
1.0
Haemopexin g/L (0.7-1.3) r-f-
2.0
t-S Pericardial Xenograft:126 patients. 112 long term survivors.
Fig. 3. Preoperative and latest postoperative New York Heart Association (N.Y.H.A.) status of the long-term survivors. The lines represent the mean of the individual groups. BC prostheses and in 12 patients with PX valves, at mean postoperative durations of 6.3 ± 0 . 7 , 16.6 ± 1.7, and 10.7 ± 1.2 months. The investigations performed included hemoglobin level, packed cell volume, reticulocyte count, total red blood cell count, fragmented red blood cell count, serum bilirubin, serum glutamic oxaloacetic transaminase level, serum lactic acid dehydrogenase level (LDH), hemopexin level, and haptoglobin estimation. Patients with mitral systolic murmur or with clinical suspicion of perivalvular leak were excluded from the study. The LDH hemopexin, and haptoglobin levels are depicted in Fig. 4. Patients with the BS valve had higher LDH and low serum hemopexin levels indicative of mild intravascular hemolysis. Patients with the BC valve had abnormally elevated LDH but normal levels of hemopexin and haptoglobin, also indicative of mild hemolysis. Patients with the PX valve did not show any evidence of hemolysis. Hemodynamic study. The results of the hemodynamic investigations are given in Tables VI and VII and in Fig. 5. All three mitral valve substitutes failed to restore normal hemodynamics in the 41 patients studied. Furthermore, no statistically significant difference was noted in the postoperative hemodynamic findings in the three groups of patients. The cardiac output at rest was near normal in patients with BS prostheses and normal in patients with BC and PX valves. However, the cardiac output response to exercise was abnormal in all patients studied. The mean
0.5-
Haptoglobin g'L (0.5-1.5)
i+i
i
(150-540)
1.0
Fig. 4. Serum lactic acid dehyrogenase (L.D.H.), hemopexin, and haptoglobin levels (mean values ± S.E.M.) in 32 patients with mitral valve replacement. pulmonary wedge pressure at rest was almost normal in patients with the mitral PX valve but was elevated in patients with BS and BC prostheses. The pulmonary artery and wedge pressures rose abnormally during exercise, the minimal rise being in patients with mitral PX valves. The mean diastolic gradients at rest were 6.2 ± 1.0, 8.3 ± 1.1, and 6.4 ± 0.5 mm. Hg in patients with BS, BC, and PX valves, respectively. The corresponding figures during exercise were 17.6 ± 0.6, 18.1 ± 1.3, and 15.3 ± 0.9 mm. Hg. The calculated valve surface areas at rest were 1.8 ± 0.3, 1.6 =t 0.1, and 2.0 ± T 0.1 sq. cm. in patients with the BS, BC, and PX valves, respectively. Although the surface area increased significantly during exercise in patients with BS prostheses and PX valves, only a marginal increase was noted in patients with BC prostheses (Fig. 5). The pulmonary vascular resistance was abnormally elevated both at rest and during exercise in all three groups of patients. Left ventricular angiography demonstrated mild incompetence in two patients with BS valves, but the rest of the prostheses and all 29 PX valves were competent.
The Journal of Thoracic and Cardiovascular Surgery
7 6 8 Tandon et al.
Table VI. Postoperative hemodynamic data (mean values ± S.E.M.) at rest in 41 patients with mitral valve replacement Variable Oxygen uptake (ml./min/sq. M.) Cardiac index (L./min/sq. M.) Mean pulmonary artery pressure (mm. Hg) Mean pulmonary wedge pressure (mm. Hg) Pulmonary vascular resistance (dynes sec. cm."' sq. M.) Mean diastolic gradient (mm. Hg) Calculated valve surface area (sq. cm.) No. of patients Mean duration postop. (mo.)
BjorkShiley
BraunwaldCutter
Pericardial xenograft
110.7 ±8.5 2.2 ±0.3 26.0 ±1.5 16.4 ±1.2 348.8 ±24.6 6.2 ±1.0 1.8 ±0.3
116.8 ±7.1 2.5 ±0.2 26.8 ±3.7 16.7 ±2.4 323.2 ±28.2 8.3 ±1.1 1.6 ±0.1
136.3 ±4.0 2.6 ±0.1 22.8 ±1.0 13.8 ±0.8 285.0 ±20.7 6.4 ±0.5 2.0 ±0.1
Discussion Three different mitral valve substitutes were used in the present series of 220 patients and follow-up data to 7 years are presented. Although this was not a randomized study, comparison of the preoperative clinical variables of the three groups of patients did not reveal any statistically significant differences. All the valve replacement procedures were undertaken by the same surgical team employing identical operative techniques. It is generally agreed that the hospital mortality rate is independent of the valve substitute employed. 9-11 None of the early deaths in the present series could be ascribed to the type of valve substitute used. Although the long-term survival following mitral valve replacement is related to a number of factors, such as age, the extent and duration of preoperative hemodynamic impairment, and coexistent myocardial and coronary artery disease, the complications directly related to various valve substitutes continue to contribute significantly to the over-all mortality rate." - 1 3 In the actuarial analysis of the long-term results of this series, the hospital mortality data have been excluded.14 The expected survival rate 7 years following valve replacement is 81.9 ± 12.8 percent for patients with BS prostheses and 89.0 ± 9.3 percent for patients with the PX valve, whereas with the BC prostheses the survival rate at 6 years was only 41.7 ± 22.9 percent. Valve-related complications were responsible for 62.5 percent of late deaths in patients with BC prostheses, whereas 40 percent and 33.3 percent of late deaths in
Table VII. Postoperative hemodynamic data (mean values ± S.E.M.) during exercise in 41 patients with mitral valve replacement Variable Oxygen uptake (ml./min./sq. M.) Cardiac index (L./min./sq. M.) Mean pulmonary artery pressure (mm. Hg) Mean pulmonary wedge pressure (mm. Hg) Pulmonary vascular resistance (dynes sec. cm. - 5 sq. M.) Mean diastolic gradient (mm. Hg) Calculated valve surface area (sq. cm.) No. of patients Mean duration postop (mo.)
BjorkShiley
BraunwaldCutter
Pericardial xenograft
390.5 ±24.3 4.3 ±0.6 48.4 ±4.5 34.2 ±2.1 264.2 ±32.3 17.6 ±0.6 2.2 ±0.4 6 22.7 ±2.3
365.7 ±20.7 4.3 ±0.3 48.5 ±6.2 32.5 ±3.6 297.7 ±24.6 18.1 ±1.3 1.7 ±0.1 6 43.0 ±2.8
406.5 ±14.4 4.1 ±0.2 42.0 ±1.9 29.4 ±1.5 260.3 ±26.3 15.3 ±0.9 2.3 ±0.1 29 40.3 ±2.8
patients with BS and PX valves, respectively, were valve related. Mechanical dysfunction of the prosthesis was responsible for three of eight late deaths in patients with the BC valve, whereas thrombotic occlusion was the cause of one of five late deaths in patients with BS prostheses. None of the late deaths in patients with PX valves was due to valvular dysfunction. Bacterial endocarditis was responsible for two valve-related deaths occurring in this group of patients. Systemic thromboembolism still remains the most significant complication associated with mitral valve replacement. The embolic rates for the three groups of patients in this series ranged from 3.2 for the PX group to 11.9 percent for patients with the BS prosthesis. The predicted fhromboembolism-free rate was 95.8 ± 2.1 percent in patients with PX valves and 93.2 ± 5.3 percent in patients with BC prostheses. Only 84.8 ± 8.4 percent of patients with the BS prosthesis are expected to be free of thromboembolism at 7 years' follow-up. Fatal valve thrombosis occurred in one patient with a BS prosthesis. Of the five embolic episodes in the BS group, three were associated with considerable morbidity; one patient underwent bowel resection and two continue to have residual neurologic deficit. All these embolic episodes occurred in the presence of a correct level of anticoagulation. All the embolic episodes in the PX group occurred in patients without any anticoagulant treatment and during the first 6 weeks following mitral valve replacement. Among the 56 patients operated upon since May, 1976,
Volume 76 Number 6 December, 1978
769
Mitral valve replacement
3L./min.Vm! X
g
T _
4"
mm. Hg. 15-
/A
-■'/
■■" /
O 2-
f R
T<
S
S
10-
5E
<
Cm.'
UJ
1
<25-
r^
UJ
5? S
UJ
I
R
E
_j
R
E
1-
* — - * Bjork-Shiley valve,(valve size 27,29), 6 patients 22.7*2.3 mths. m-—■■ Braunwald-Cutter valve,(valve size 32,34), 6 patients 4 3 . 0 * 2 . 8 mths. • — • l - S Pericardial Xenograft valve, (valve size 25,27,29), 29 patients 40.3±2.8 mths.
Fig. 5. Postoperative hemodynamic data (mean values valve replacement. R, Rest. E, Exercise. and receiving prothrombin depressants for 4 to 6 weeks, there have been no thromboembolic complications despite the suboptimal levels of anticoagulation. Valve thrombosis has not been encountered in patients with PX valves.15 Thus the PX valve is definitely superior to prosthetic valves in reducing the incidence of thromboembolism even without use of long-term anticoagulation. The benefits of a low embolic rate and the absence of valve thrombosis, without long-term anticoagulation, become even more significant when considering the life-threatening complications resulting from the use of prothrombin depressants per se.16, 17 In the present series fatal hemorrhage occurred in one patient each with BC (1.9 percent) and BS (2.4 percent) prostheses (Table VIII). The latest follow-up assessment of long-term survivors showed that 91 percent of patients with PX, 80 percent with BS, and 56.4 percent with BC prostheses were in N.Y.H.A. Class I. The hemodynamic investigations demonstrated significant circulatory improvement following mitral valve replacement in all patients studied. However, the investigations failed to reveal any significant difference among the results obtained with the three types of valve substitutes. A certain degree of obstruction to forward flow across the mitral valve was found with all three valve substitutes, and this was significant during exercise. The mean diastolic gradients were identical with BS and PX valves at rest, but during exercise the PX had a marginal advantage over the BS prosthesis. The most significant single factor in producing obstruction to forward flow is the presence of rigid rings in all currently used valve substitutes that do not expand during diastole and therefore encroach upon the normal functional mechanism of the mitral valve anulus.10
standard error of the mean) of 41 patients with mitral Table VLU. Long-term complication rate of the valve substitutes used for mitral valve replacement Valve substitute
Thrombosed Embolism (%) valve (%)
BjorkShiley BraunwaldCutter Pericardial xenograft
Fatal hemorrhage (%)
Late mortality (%)
2.4
11.9
2.4
11.9
-
6.8
1.9
15.4
-
3.2
-
4.8
In practical terms the most important hemodynamic determinant of a valve substitute is the effective surface area as measured in vivo. In the present study the largest surface areas were obtained in patients with PX valves, both at rest and during exercise.18 The BS prosthesis was found to have a surface area comparable to that of the PX valve, but that of the BC prosthesis was obviously suboptimal. Conclusions The BC prosthesis is associated with a comparatively low survival rate and with serious mechanical complications owing to wear of the cloth fabric. We have discontinued its use and are now in the process of electively replacing the remaining valves. Despite subjective impressions, there was no statistically significant difference between the clinical and hemodynamic performance of the BS prosthesis and the Ionescu-Shiley PX valve. The BS prosthesis and the PX valve offer similar mechanical durability and hemodynamic performance for an almost identical duration of follow-up. Use of the BS valve is associated with increased
770
Tandon et al.
morbidity despite the imperative need for permanent anticoagulation. Since the PX valve is associated with an improved quality of life and minimal thromboembolic complications in the absence of long-term anticoagulation, we are using this preparation exclusively for all heart valve replacements. We gratefully acknowledge the help given by Miss Ann Tunnicliffe in the preparation of this manuscript. REFERENCES 1 Starr A, Edwards ML: Mitral replacement. Clinical experience with a ball valve prosthesis. Ann Surg 154:726740, 1961 2 Technical information on Bjbrk-Shiley heart valves, Shiley Laboratories, Irvine, Calif., 1972 3 Braunwald NS, Tatooles C, Turina M, Detmer D: New development in the design of fabric-covered prosthetic heart valves. J THORAC CARDIOVASC SURG 62:673-682,
1971 4 Technical information on the Ionescu-Shiley pericardial xenograft heart valve, Shiley Laboratories, Irvine, Calif., 1977 5 Gorlin R, Gorlin SG: Hydraulic formula for calculation of the area of the stenotic mitral valve, other cardiac valves, and central circulatory shunt. Am Heart J 41:1-29, 1951 6 Ionescu MI, Pakrashi BC, Holden MP, Mary DAS, Wooler GH: Results of aortic valve replacement with frame-supported fascia lata and pericardidal grafts. J THORAC CARDIOVASC SURG 64:340-353, 1972
7 Mary DAS, Pakrashi BC, Catchpole RW, Ionescu MI: Tissue valves in the mitral position. Five year's experience. Br Heart J 37:1123-1132, 1975 8 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
The Journal of Thoracic and Cardiovascular Surgery
9 Simonsen S, Forfang K, Anderson A, Efskind L: Hospital mortality after mitral valve replacement. Prognostic significance of pre-operative clinical and haemodynmaic factors. Acta Med Scand 195:243-246, 1974 10 Brawley RK, Donahoo RK, Gott VL: Current status of the Beall, Bjbrk-Shiley, Braunwald-Cutter, LilleheiKasterand Smeloff-Cutter cardiac valve prostheses. Am J Cardiol 35:855-865, 1975 11 Salomon NW, Stinson EB, Griepp RB, Shumway NE: Mitral valve replacement. Long-term evaluation of prosthesis related mortality and morbidity. Circulation 56: (Suppl.):94-101, 1977 12 Angell WW, Angell JD, Sywak A: Selection of tissue or prosthetic valve. A five-year prospective, randomized comparison. J THORAC CARDIOVASC SURG 73:43-53,
1977 13 Leading article: Which prosthetic valve? Lancet 1:619620, 1976 14 Rahimtoola SH: Valve replacement—a perspective. Am J Cardiol 35:711-715, 1975 15 Ionescu MI, Tandon AP, Smith DR: Long-term clinical and hemodynamic evaluation of the Ionescu-Shiley pericardial xenograft heart valve. Presented at the Fourteenth Congress of the Pan Pacific Surgical Association, Honolulu, Hawaii, April 1-7, 1978 16 Hume M, Sevitt S, Thomas DP: Venous Thrombosis and Pulmonary Embolism. Cambridge, Mass., 1970, Harvard University Press, p 455 17 Isom OW, William CD, Falk EA, Spencer FL, Glassman E: Evaluation of anticoagulant therapy in cloth covered prosthetic valves. Circulation 47:Suppl 3:48-50, 1973 18 Tandon AP, Smith DR, Ionescu Ml: Hemodynamic evaluation of the Ionescu-Shiley pericardial xenograft in the mitral position. Am Heart J 95:595-601, 1978 (For discussion see page 785)