Mitral Valve Replacement: Randomized Trial of St. Jude and Medtronic-Hall Prostheses Andrew C. Fiore, MD, Keith S. Naunheim, MD, Stephen DOrazio, MD, George C. Kaiser, MD, Lawrence R. McBride, MD, D. Glenn Pennington, MD, Pamela S. Peigh, MD, Vallee L. Willman, MD, Arthur J. Labovitz, MD, and Hendrick B. Barner, MD Division of Cardiothoracic Surgery and Cardiology, St. Louis University Medical Center and St. Mary’s Health Center, St. Louis, Missouri
To better define the merits of the bileaflet and tilting-disc valves, we prospectively randomized 102 patients (mean age, 57 years; range, 11 to 85 years) to receive either the St. Jude (n = 55) or the Medtronic-Hall (n = 47) mitral valve prosthesis between September 1986 and May 1991. The two groups were not different with respect to preoperative New York Heart Association class, incidence of mitral stenosis and insufficiency, angina score, extent of coronary artery disease, ventricular function, completeness of revascularization, or cross-clamp or bypass time. The hospital mortality (14.5% versus 10.6%, St. Jude versus Medtronic-Hall) and late mortality (7.3% versus 2.1%) were not significantly different. Follow-up was complete in 84 of 89 hospital survivors (94%)with a mean of 26 months (range, 1 to 60 months). The linearized rates
of valve-related events and the 3-year actuarial survival demonstrated no significant differences between both cohorts. Comparison of the clinical outcome and echocardiographic parameters obtained at the time of follow-up demonstrated no significant differences between the two prostheses. These data indicate that the Medtronic-Hall and St. Jude mitral prostheses are similar with respect to their rates of valve-related complications and hemodynamic profiles. This study suggests that there is no difference between the St. Jude and Medtronic-Hall prostheses with regard to early clinical performance or hemodynamic results and therefore does not support the preferential selection of either prosthesis. (Ann Thorac Surg 1992;54:68-73)
A
tion were prospectively randomized to receive either the Medtronic-Hall (MH, 47 patients) or the St. Jude (SJ, 55 patients) mitral prosthesis. Patients requiring emergency operations and those requiring double aortic and mitral valve replacement were excluded from this study. Randomization was accomplished by a blind drawing from a mixture of identical appearing cards. The clinical data recorded included age, sex, admitting diagnosis, cardiac risk factors, procedure performed, and postoperative complications. A New York Heart Association (NYHA) functional classification was determined for each patient. Angina was graded using the Canadian Cardiovascular Society Classification. Cardiac catheterization data noted include left ventricular end-diastolic pressure, pulmonary artery pressure, cardiac index, regional left ventricular score (regional wall motion scoring as defined by Coronary Artery Surgery Study) [l], mitral valve area (calculated using the Gorlin formula), and the presence and severity (grade 0 to 4) of angiographically determined rnitral insufficiency, number of major coronary vessels diseased (defined as 270% luminal obstruction in any angiographic plane), presence of left main coronary stenosis (defined as 250% luminal obstruction). M mode, two-dimensional, and Doppler echocardiography was performed preoperatively and 1 to 6 months postoperatively. Ventricular end-systolic and end-diastol-
lthough heart valve replacement is a safe and commonly performed procedure, all available prostheses are associated with valve-related complications that influence their clinical use. The most widely used prosthetic valves in the United States are the bileaflet St. Jude and the tilting-disc Medtronic-Hall. Both prostheses possess specific mechanical features that appear advantageous. The wide opening angle (70 degrees), upward gliding disc, and monostrut construction are attractive properties of the MedtronicHall valve, whereas the central flow design of the lowprofile bileaflet St. Jude prosthesis may be hemodynamically advantageous. To determine if one prosthesis has clinical or hemodynamic superiority, we designed a prospective, randomized trial to compare these two valves in patients requiring mitral valve replacement.
Material and Methods From September 1986 until May 1991, 102 patients undergoing elective valve replacement for mitral valve dysfuncPresented at the Thirty-eighth Annual Meeting of the Southern Thoracic Surgical Association, Orlando, FL, Nov 7-9, 1991. Address reprint requests to Dr Fiore, Cardiothoracic Division, Department of Surgery, St. Louis University Medical Center, 3635 Vista Ave at Grand Blvd, PO Box 15250, St. Louis, MO 63110-0250.
0 1992 by The Society of Thoracic Surgeons
0003-4975/92/$5.00
Ann Thorac Surg 1992;54:6&73
ic dimensions were measured, and ejection fraction and prosthetic valve gradient were calculated. In patients with atrial fibrillation, the mean of ten consecutive cardiac cycles was calculated. The ejection fraction was calculated by Simpson’s rule as previously described [2].Mitral valve areas were calculated by the pressure half-time method with either continuous or pulsed wave recordings [3,41. The presence and degree of prosthetic valve regurgitation was qualitatively assessed by pulse wave, continuous wave, and color Doppler echocardiography from multiple windows. Cardiopulmonary bypass was established using ascending aortic and bicaval cannulation employing systemic hypothermia and hemodilution. Myocardial protection included the use of cold hyperkalemic blood cardioplegia (6°C). The infusate consisted of a 4:l dilution of blood with 0.9% normal saline solution which contained 20 mEq/L of sodium bicarbonate and 0.1 mg/mL of xylocaine hydrochloride. The initial infusion contained 14 mEq/L of potassium, and all subsequent infusions contained 7 mEq/L of potassium. In patients in whom the aortic valve was competent, an initial dose (15mL/kg) was given through the aortic root and subsequent doses (7.5mL/kg) were infused through the coronary sinus. In patients with predominantly aortic insufficiency, all doses of cardioplegia were given through the coronary sinus. Topical myocardial cooling with Ringer’s lactate slush and a cardiac insulation pad for phrenic nerve protection were employed in all patients. Proximal vein graft anastomoses were performed using a partially occlusive aortic clamp while distal anastomoses were constructed with the aorta cross-clamped. Myocardial revascularization was performed before mitral valve replacement to avoid raising the heart after the mitral prosthesis had been inserted, thus decreasing the risk of myocardial rupture. The operative procedure was performed by the same group of surgeons using similar techniques. The mitral prosthesis was inserted with interrupted horizontal mattress sutures of 2-0 Ethibond (Ethicon, Somerville, NJ) placed through the left atrial side so as to evert the mitral annulus. When feasible, the posterior mitral leaflet was left intact. The St. Jude prosthesis was oriented either in the antianatomic axis perpendicular to the mitral commissures or in the anatomic plane parallel to the mitral commissures. The Medtronic-Hall valve was generally oriented with the larger orifice opening posteriorly. This prosthesis could be rotated according to the surgeon’s preference to prevent entrapment of the disk by the retained posterior mitral leaflet. Recorded operative variables included cardiopulmonary bypass and cross-clamp times. Prophylactic anticoagulation with sodium warfarin was carried out in all patients. Anticoagulation was instituted within 72 hours after the operation unless clinical contraindications existed. The control of therapy was made by interval analysis of the prothrombin index, which was kept at 1.5times the control value. Antibiotic prophylaxis consisted of cefazolin sodium (Ancef; Smith Mine & French Laboratories, Philadelphia, PA) administered preoperatively and for 48 hours postoperatively. Follow-up data were collected by a registered nurse
FIOREETAL PROSTHETIC MITRAL VALVE REPLACEMENT
69
using written questionnaires and telephone interviews with the patient, the patient‘s physician, or both. Patients surviving the operative period were followed up for a minimum of 1 month and a maximum of 60 months (mean, 26 months). The cumulative follow-up was 105 patient-years for the SJ and 121 patient-years for the MH group. Five patients were lost to follow-up at 1 (SJ), 1 (SJ), 10 (MH), 10 (MH), and 35 (SJ) months. The definitions of events, their classification, and the analysis of the data follow the recently suggested guidelines for reporting morbidity and mortality after cardiac valve prosthesis insertion [5].Data were analyzed using the Statview I1 statistical software package (Brainpower, Inc, Calabasas, CA). Univariate analysis of discrete variables was performed using the ,$ analysis or Fisher’s exact test where appropriate. Paired and unpaired Student’s t tests were employed for continuous variables where appropriate. Mean values are expressed as plus or minus the standard deviation except as noted. The linearized rates were computed using an incidence density model. All events were considered in the calculation of the linearized incidences, which are expressed as events per 100 patient-years of follow-up. These rates were compared using a maximum likelihood ratio. The actuarial curves were constructed using the methods described by Berkson and Gage and compared using a log rank analysis [6,71.A p value of less than 0.05 was considered significant.
Results Clinical Profile The cohorts were similar with respect to age, sex, preoperative NYHA functional class, hemodynamic and operative parameters, and the degree of mitral stenosis or insufficiency. The majority of patients had mild left ventricular dysfunction with moderate pulmonary hypertension. Twenty percent of patients in both groups had had prior valve replacement, and a similar number required concomitant myocardial revascularization (Table 1).
Operative Mortality A total of 13 patients (12%) died in the hospital. The overall operative mortality was 14.5% for the SJgroup and 10.6% for the MH group (not significant [NS]) (Table 2). The operative mortality for isolated mitral valve replacement was 16.3% for the SJ group and 11.1% for the MH group (NS), whereas that for mitral valve replacement with myocardial revascularization was 8.3% and 12.5%, respectively (NS). The most common cause of operative death was low cardiac output syndrome. Six patients died as a result of low cardiac output associated with biventricular failure. In two instances (1 MH; 1 SJ) operative death was caused by atrioventricular sulcus rupture. Fatal dysrhythmias developed in 1 patient in the MH group, whereas 1 patient receiving the St. Jude prosthesis died of mediastinitis and generalized sepsis 33 days postoperatively. Three additional patients, all in the SJ group, died of multiorgan system failure. One patient with a history of
70
FIOREETAL PROSTHETIC MITRAL VALVE REPLACEMENT
Ann Thorac Surg 1992;54:6%73
100%
Table 1. Preoperative Clinical Profile"
Variable Age (Y) Male sex (%) NYHA class LVEDP (mm Hg) Left ventricular score Pulmonary artery systolic (mm Hg) Cardiac index (L * min-' . m-') Mitral area (cm') Mitral insufficiency Coronary artery disease (%) CABG (%)
Prior valve replacement
(a)
Duration of CPB (min) Cross-clamp time (min) a
SJ
MH
P Value
58.00 +. 15.4 43.0 3.00 2 0.7 14.40 ? 6.8 7.00 ? 3.0 50.50 2 20.3
57.00 2 10.0 27.7 3.02 2 0.5 13.90 ? 7.1 7.60 2 3.6 52.10 t 16.0
NS NS NS NS NS NS
2.30 2 0.6
2.30 2 0.5
NS
401
ST.JUDE
20
0
0 PATIENTS 47 ATRISK 55
-
1
1 31 37
,
2 23
32
3 17 22
I
,
4 9 8
I
,
5 YEARS 0
Medtronic
0 St.Jude
NS NS NS
Fig 1. Actuarial survival rates after mitral valve replacement with the St. Jude and Medtronic-Hall valves. Hospital and late deaths are included. (NS = not significant.)
21.8 21.8
17 21.3
NS NS
Late Mortality
NS NS
138.50 It 45.0 147.50 2 63.0 81.00 ? 28.0 86.00 2 31.0
chronic hepatitis died 13 days postoperatively of liver and renal failure. A second patient remained comatose postoperatively; renal and respiratory failure developed, and the patient died on the fourth postoperative day. The remaining patient had undergone urgent third-time mitral valve replacement. He died of multisystem organ failure after several hypotensive episodes on postoperative day 5. The patients who died in the hospital had a mean NYHA functional class of 3.3, and 77% of hospital deaths were in patients with mitral regurgitation as the primary pathologic process. Univariate analysis demonstrated that the presence of mitral insufficiency and a left ventricular end-diastolic pressure of 20.4 k 0.7 mm Hg correlated with operative mortality (p < 0.05). Sex, pulmonary artery systolic pressure, the presence of coronary artery disease, and the global ischemic time were not correlated with hospital death. Table 2. Causes of Operative Mortality
Cause
SJ
Left ventricular failure Arrhythmia Septicemia Multiple-organ failure
MH
3 0 1 3 1
Left ventricular tear Total
8 14.5%
Operative mortality Medtronic-Hall;
Fi
0.96 2 0.42 2.70 2 1.5 23.4
CPB = cardiopulmonary CABG = coronary artery bypass grafting; bypass; LVEDP = left ventricular end-diastolic pressure; MH = Medtronic-Hall; NS = not significant; NYHA = New York Heart SJ = St. Jude. Association;
=
T
1
0.97 2 0.37 3.00 2 1.3 30.9
Values expressed as percent or mean 2 the standard deviation.
MH
C
SJ = St. Jude.
Late deaths occurred in 7 patients (6 SJ and 1 MH; 7%). The mortality rates for the SJ and MH group were 7.3% and 2.1% (NS), respectively. Valve-related complicationswere the cause of late death in 6 patients (5 SJ, 9%; 1 MH, 2%). The most common cause of valve-related death was sudden death, which occurred in 3 patients (SJ group) at 6, 37, and 39 months postoperatively. Ventricular fibrillation developed at home in 2 patients who could not be resuscitated. The remaining patient died suddenly at home in association with severe dyspnea. None of these received postmortem examination. One death each resulted from cerebral hemorrhage (MH), valve thrombosis (SJ), and cerebral embolus (SJ)at 1 month, 2 months, and 4.5 years, respectively. One valve-unrelated death (SJ) occurred at 6 months postoperatively in a patient with an ischemic cardiomyopathy.
Patient Survival The actuarial survival curve for the population is shown on Figure 1. At 3 years, the survival rates were slightly better for the patients receiving the Medtronic-Hall valve, but the difference was not significant (87 t 4.9 versus 79 +- 5.3; NS). The postoperative NYHA class in the 84 survivors was equally improved in both cohorts (MH, 1.8 Ifr. 1.0; SJ, 1.7 k 0.9; NS).
Valve-Related Complications Nineteen valve-related complications developed in 18 of the 89 hospital survivors (20%). The linearized rates of valve-related complications and the actuarial freedom from these morbid events are shown in Tables 3 and 4, respectively. There were a total of seven episodes of thromboembolism occurring in 6 patients. Two were fatal. Valve thrombosis occurred in 2 patients, 1 in each group. The patient in the SJ group died suddenly before reoperation, whereas the patient in the MH group underwent successful reoperation and is a long-term survivor.
THROMBOEMBOLISM.
5 10.6%
71
FIOREETAL PROSTHETIC MITRAL VALVE REPLACEMENT
Table 3 . Linearized Rate of Valve-Related Events" Event
Thromboembolism Sudden death Valve thrombosis Hemorrhage Endocarditis
1.6 f 1.2 2.5 f 1.4 0.8 f 0.8 2.5 f 1.4 0 0 0
Paravalvular leak
Reoperation a
MH
SJ
Expressed as events per 100 patient-years
MH
=
Medtronic-Hall;
NS
4.9
Value
NS NS NS NS NS NS NS
2.2
0
1.0 f 1.0 2.9 f 1.7 2.0 f 1.4 1.0 f 1.0 2.0 & 1.4 ?
standard error of mean.
SJ = St. Jude.
not significant;
=
f
P
Two patients in the MH group had transient ischemic attacks, and 3 patients (2 MH; 1 SJ)had presumed embolic strokes, one of which was fatal (SJ).The linearized rate of thromboembolism was higher in the MH group, but the difference did not achieve statistical significance. There were six instances of hemorrhage related to anticoagulant treatment (3 MH; 3 SJ). Bleeding occurred intracerebrally in 2 patients and was fatal in 1 at 4 weeks postoperatively (MH group). The remaining 4 patients had nonfatal gastrointestinal bleeding that did not require operation. The linearized rates were similar in both groups. HEMORRHAGE.
Bacterial endocarditis developed in 2 patients with Medtronic-Hall valves. A paravalvular leak developed in 1 patient, who underwent replacement of the prosthesis. The remaining patient was successfully treated nonoperatively. Both are long-term survivors. The linearized rate of endocarditis was not significantly different in both cohorts. ENDOCARDITIS.
PARAVALVULAR LEAK. A single patient had a paravalvular leak as noted in the previous section.
Postoperative Echocardiography Doppler echocardiographic evaluation was performed postoperatively in 33 patients in the SJ group and 32 patients in the MH group. Thirty patients (19 SJ, 11 MH) underwent evaluation during the first 3 months after operation. The remaining 35 patients (20 SJ and 15 MH) underwent Doppler echocardiography between 3 and 6 months postoperatively. The mean transvalvular gradient at rest was not significantly different between groups at any valve size (Table 5).
Table 4 . Actuarial 3-Year Freedom From Events" MH
SJ
Event
Survival Thromboembolism
98 2 2.0
88 f 5.5
Hemorrhage
91
94
a
79
Values expressed as percent
MH
=
Medtronic-Hall;
NS
f
f
5.3 5.3
87
f 4.9 f 4.4
?
standard error of mean.
=
not significant;
Table 5. Postoperative Transvalvular Echocardiographic Mean Gradient
P
Value
NS NS NS
SJ = St. Jude.
MH
SJ
Valve Size
Gradient
Gradient
(mm)
(mmHg)
N
(mmHg)
25 27 29 31 33
4.2 2.7 2.0 2.0 3.8
7 6 8 7 5
4.0 4.0 3.0 2.7 3.4
f 3.3
1.3 f 1.0 f 0.7 f 2.3 f
MH = Medtronic-Hall; SJ = St. Jude. icant;
N
=
f 2.5
1.5 f 1.2 f 0.6 f
f
1.1
number of patients;
P
N
Value
7 6 6 3 10
NS NS NS NS NS
NS
=
not signif-
Reoperations Two patients, both in the MH group, underwent successful reoperation at 6 weeks and 7 months postoperatively. The indications for reoperation were prosthetic valve endocarditis in 1 and valve thrombosis secondary to sodium warfarin noncompliance in the other. Both patients are long-term survivors.
Comment The two most common cardiac valve prostheses used in the United States are the bileaflet St. Jude and the Medtronic-Hall pivoting-disc prosthesis. These devices have undergone virtually no alteration in design since their development in 1977 and 1978, and had remained unmodified since this study commenced in 1986. Several investigators have reported the 5-year and 10-year clinical and hemodynamic results of both valves (8-101; however, a prospective, randomized trial comparing both prostheses in the mitral position has not been reported. As in other series, the leading cause of hospital mortality in our series was perioperative heart failure. The overall operative mortality was similar to that reported by other investigators [ l l , 121. Patient survival based on 3-year actuarial estimates was not significantly different between groups. Our mediumterm survival is comparable with those of published reports for the St. Jude and Medtronic valves [13, 141. There were 3 cases of unobserved or sudden death in our series. Although autopsy information is not available, these 3 patients had end-stage cardiac failure and had no evidence in their postoperative evaluation of prosthetic valve dysfunction. At the time of their death, all 3 patients had therapeutic anticoagulation levels. Survival itself is usually regarded as an insensitive measure of overall valve performance and relates mainly to the underlying disease and the perioperative status of the myocardium. A more specific measure of valve performance is thromboembolism, which is a major risk factor for patients with prosthetic valves. The incidence of thromboembolic complications in the third-generation valvular devices has decreased dramatically. This has been due to improved hemodynamic design resulting in a better flow velocity profile, less turbulence, and reduction
72
FIOREETAL PROSTHETIC MITRAL VALVE REPLACEMENT
R C
T
I
C
I
20
0 0 PATIENTS 47 ATRISK 55
Ann Thorac Surg 199254 :6a73
1 28 37
2 20 32
3 16 22
p=NS
4 9 8
I 1
5 YEARS 0 Medtronic 0 St. Jude
Fig 2. Actuarial freedom from thromboembolism after mitral valve replacement with the St. Jude and Medtronic-Hal1 valves. (NS = not significant.)
of stagnant areas, all of which are known to affect the thrombogenicity of the prosthesis. In the Medtronic-Hall valve, a more central position of the disc in the open position enlarged the minor orifice, thus reducing stagnation of blood and turbulence in this area. The central flow design of the bileaflet St. Jude valve may contribute to its low incidence of thromboembolic events. In this report, the 3-year actuarial freedom from thromboembolism was slightly better for the St. Jude prosthesis, but the difference was not significant (Fig 2). The 4 patients in whom nonfatal thromboembolic events developed were adequately anticoagulated. Dipyridamole (300 mg daily) was added to their anticoagulation regimen to reduce the recurrent risk of thromboembolism. There were 2 cases of valve thrombosis, 1 in each group. Noncompliance with sodium warfarin anticoagulation in the presence of atrial fibrillation was believed to have contributed to thrombus formation in both patients. In this series, the overall rates of thromboembolic complications compared favorably with those reported by Arom and associates 1151 for the St. Jude valve and Keenan and co-workers [16] for the Medtronic-Hall valve. Hemorrhage was the second most common valverelated complication. This complication was fatal in 1 hypertensive patient in whom an intracerebral hemorrhage developed. The actuarial freedom from hemorrhage was similarly low in both cohorts and compared favorably with that reported in other series for the Medtronic and St. Jude valves (Fig 3 ) 117, 181. The hemodynamic advantages of the St. Jude prosthesis, especially in smaller valve sizes, have been documented in the literature 119, 201. In this report, we measured the mean transvalvular gradient between 1 and 6 months postoperatively using Doppler echocardiography. No significant difference in the resting mean gradient could be demonstrated when both prostheses were compared at any valve size. Improved hemodynamic performance has been associated with symptomatic improvement. In this series, the St. Jude and Medtronic valves achieved equal improvement in patient symptoms as evidenced by the low postoperative NYHA functional score observed in both cohorts. The results of this study indicate that in the first years after implantation, the performance of the St. Jude and
0
T
20
MEDTRONIC ST.JUDE
0 0 PATIENTS 47 ATRISK 55
1 1 2 3 4 5 YEARS 31 22 17 8 0 Medtronic 36 30 20 7 0 St.Jude Fig 3. Actuarial freedom from hemorrhage after mitral valve replacement with the St. Iude and Medtronic-Hall valves. (NS = not significant.)
the Medtronic-Hall prostheses are comparable in this patient population. These two valves are among the best of the modern generation of prostheses, and either may be confidently recommended as valve substitutes. The choice of one or the other should be based on the surgeon’s experience and preference, which may be linked to other factors such as ease of insertion, availability, and cost. Clearly, this study suggests that neither valve can claim superior performance or safety. We thank Jan Castanis, RN, and Jan St. Vrain, RN, for special technical assistance. We gratefully acknowledge Jean Wamhoff, Scott Langreder, Mary Nonte, and Karen Kelly for their assistance with preparation of the manuscript and Gary Grunkemeier for his assistance with the statistical analysis.
References 1. Principal investigators of CASS and their associates. The National Heart, Lung, and Blood Institute Coronary Artery Surgery Study. Circulation 1981;63(Suppl 1):1-81. 2. Mehdirad AA, Williams GA, Bryg RJ, et al. Evaluation of left ventricular function with upright exercise: correlation of exercise Doppler with post-exercise wall motion and single plane echocardiographic ejection fraction. Circulation 1987; 75:413-9. 3. Hatle L, Angelsen B, Tromsdal A. Non-invasive assessment of atrioventricular pressure half-time by Doppler ultrasound. Circulation 1969;60:109&104. 4. Bryg RB, Williams GA, Labovitz AJ, Aker U, Kennedy HL. Effect of atrial fibrillation and mitral regurgitation on calculated mitral valve area: comparison of Doppler, twodimensional echocardiography and cardiac catheterization. Am J Cardiol 1986;57:63&8. 5. Edmunds LH Jr, Clark RE, Cohn LH, Miller DC, Weisel RD. Guidelines for reporting morbidity and mortality after cardiac valvular operations. Ann Thorac Surg 1988;46:257-9. 6. Berkson J, Gage RP. Calculation of survival rates for cancer. Proc Meet Mayo Clin 1950;25:270-86. 7. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959;22:71943. 8. Gray R, Chaux A, Matloff J, Raymond M. Early postoperative hemodynamic comparison of !St. Jude cardiac prosthesis and porcine xenografts at rest and exercise [Abstract]. Circulation 1979;6O(Suppl 2):222. 9. Antunes MJ, Wessels A, Sadowski RJ, et al. Medtronic-Hall valve replacement in a third-world population group: a
FIOREETAL PROSTHETIC MITRAL VALVE REPLACEMENT
10. 11. 12.
13.
14.
review of the performance of 1,000 prostheses. J Thorac Cardiovasc Surg 1988;95:980-93. Antunes MJ. Clinical performance of St. Jude and MedtronicHall prostheses: a randomized comparative study. Ann Thorac Surg 1990;50:743-7. Nicoloff DM, Emery RW, Arom KV, et al. Clinical and hemodynamic results with the St. Jude Medical cardiac valve prosthesis. J Thorac Cardiovasc Surg 1981;82:674-83. Butchart EG, Lewis PA, Grunkemeier GL, Kulatilake N, Breckenridge IM. Low risk of thrombosis and serious embolic events despite low-intensity anticoagulation. Experience with 1,004 Medtronic-Hall valves. Circulation 1988;78(Suppl 1):66-71. Arom KV, Nicoloff DM, Kersten TE, Northrup WF 111, Lindsay WG. Six years of experience with the St. Jude Medical valvular prosthesis. Circulation 1985;72(Suppl 2): 153-8. Hall KV, Nitter-Hauge S, Abdelnoor M. Seven and one-half years‘ experience with the Medtronic-Hall valve. J Am Coll Cardiol 1985;6:1417-21.
73
15. Arom KV, Nicoloff DM, Kersten TE, et al. Ten years’ experience with the St. Jude Medical valve prosthesis. Ann Thorac Surg 1989;47:831-7. 16. Keenan RJ, Armitage JM, Trento A, et al. Clinical experience with the Medtronic-Hall valve prosthesis. Ann Thorac Surg 1990;50:748-53. 17. Nitter-Hauge S, Semb 8, Abdelnoor M, Hall KV. A 5 year experience with the Medtronic Hall disc valve prosthesis. Circulation 1983;68(Suppl 2):169-74. 18. Chaux A, Gray RJ, Matloff JM, Feldman H, Sustaita H. An appreciation of the new St. Jude valvular prosthesis. J Thorac Cardiovasc Surg 1981;81:202-11. 19. Harstkotte E, Haerten K, Seipel L, et al. Central hemodynamics at rest and during exercise after mitral valve replacement with different prostheses. Circulation 1983;68(Suppl 2):161-8. 20. Gil CC, Ing HC, Lytle BW, Cosgrove DM, Golding LAR, Loop FD. Early clinical evaluation after aortic valve replacement with the St. Jude Medical valve in patients with a small aortic root. Circulation 1982;66(Suppl 1):147-9.
DISCUSSION DR OSCAR R. BAEZA (Newark, NJ): I congratulate you for your nice presentation. I have a question, though, and that relates to the preservation of the posterior leaflet of the mitral valve or the subvalvular apparatus, vis a vis the profile of both the St. Jude and Medtronic-Hall prostheses. I participated in some of the clinical evaluations when the Medtronic-Hall was called the Hall-Kaster valve, and we had problems with the occluder being stuck in a closed position by interference with suture tails or other subvalvular structures. My question to you is, do you see an advantage of using a lower profile valve, like the St. Jude, in patients who have small ventricles and in whom you would like to preserve the subvalvular apparatus? I stopped using the Medtronic-Hall valve many years ago because the occluder enters the housing like a door that closes inside its frame. Like a bathroom door that is slammed stuck with a towel hanging from the knob, the occluder of the MedtronicHall may also get stuck in a closed position. In my opinion, the St. Jude prosthesis offers the advantage of a very low profile and very little chance to interfere with sutures or the mitral subvalvular apparatus during mitral valve replacement. D R KIT V. A R O M (Minneapolis, MN): I would like to congratulate Dr Fiore for a very interesting presentation. This communication is one of the very few prospective, randomized studies, of which we need more. I think he has confirmed the previous studies that the St. Jude Medical and the Medtronic-Hall are probably the better two mechanical valves available at the present time. However, in the previous in vitro study, the St. Jude valve had some advantage because of bileaflet design, low pressure gradient, and better effective orifice areas. You also confirm that the incidence of thromboembolism is low in both of these prostheses. There is a cumulative report, about 4,000 cases in each group, showing that the incidence of thromboembolism for both St. Jude and Medtronic valves is very low, being 1.6 and 2.6 episodes/100 patient-years, respectively. I have one comment about using Doppler flow study to obtain the gradient. As you know, Doppler picks up the velocity and the gradient is derived from this. Because of the design, the St. Jude has a higher velocity in the central opening of the valve, which could give you a false high gradient.
I also have a few questions. First, how many of the patients in the Medtronic-Hall group had chordae or supporting structure of the mitral valve preserved, and did you have any trouble from the high profile of the Medtronic-Hall occluder relating to this? Second, is the incidence of bleeding higher than what we usually see with the St. Jude valve? What is the prothrombin level in this group? We now keep the prothrombin time between 18 and 20 seconds for the St. Jude valve, and by doing so we are able to cut down the incidence of bleeding significantly. The other question is, why do you see a higher gradient in the 33-mm prostheses of both groups? Can you explain why the gradient in some of the larger valves is higher than the smaller ones? The last question is, have you done a similar study in the aortic position? D R FIORE: I will answer Dr Arom’s questions first. We did try to retain the posterior leaflet whenever possible. In patients with mitral stenosis it was not usually possible to accomplish this safely. In patients with mitral insufficiency secondary to myxomatous degeneration, great care was taken to be certain that the posterior leaflet was carefully reefed under tension when inserting the prosthesis. To best accomplish this, sutures were placed from the left atrial side so as to evert the posterior leaflet and the annulus. This avoids any potential disruption of the disk with the subvalvular apparatus. We had one fatal episode of anticoagulant-related hemorrhage. This patient had atrial fibrillation and known left atrial thrombus, and for these reasons, the prothrombin time was kept at approximately two times the control. This may have accounted for the bleeding episode. In general, we try to keep our prothrombin times 1% times the control. The Doppler-derived transvalvular gradients did show slightly elevated values in the larger size valves, although all the numbers were quite low. With respect to your last question, we have performed a prospective, randomized trial comparing both valves in the aortic position. The results are currently being analyzed. Dr Baeza, we agree that in patients with a small left ventricle in whom you wish to preserve the subvalvular apparatus, there may be an advantage of using the lower profile St. Jude valve. Our results suggest that this advantage would be technical and not hemodynamic.
To better define the merits of the bileaflet and tilting-disc valves, we prospectively randomized 102 patients (mean age, 57 years; range, 11 to 85 years) to receive either the St. Jude (n = 55) or the Medtronic-Hall (n = 47) mitral valve prosthesis between September 1986 and May 1991. The two groups were not different with respect to preoperative New York Heart Association class, incidence of mitral stenosis and insufficiency, angina score, extent of coronary artery disease, ventricular function, completeness of revascularization, or cross-clamp or bypass time. The hospital mortality (14.5% versus 10.6%, St. Jude versus Medtronic-Hall) and late mortality (7.3% versus 2.1%) were not significantly different. Follow-up was complete in 84 of 89 hospital survivors (94%)with a mean of 26 months (range, 1 to 60 months). The linearized rates
of valve-related events and the 3-year actuarial survival demonstrated no significant differences between both cohorts. Comparison of the clinical outcome and echocardiographic parameters obtained at the time of follow-up demonstrated no significant differences between the two prostheses. These data indicate that the Medtronic-Hall and St. Jude mitral prostheses are similar with respect to their rates of valve-related complications and hemodynamic profiles. This study suggests that there is no difference between the St. Jude and Medtronic-Hall prostheses with regard to early clinical performance or hemodynamic results and therefore does not support the preferential selection of either prosthesis. (Ann Thorac Surg 1992;54:68-73)
A
tion were prospectively randomized to receive either the Medtronic-Hall (MH, 47 patients) or the St. Jude (SJ, 55 patients) mitral prosthesis. Patients requiring emergency operations and those requiring double aortic and mitral valve replacement were excluded from this study. Randomization was accomplished by a blind drawing from a mixture of identical appearing cards. The clinical data recorded included age, sex, admitting diagnosis, cardiac risk factors, procedure performed, and postoperative complications. A New York Heart Association (NYHA) functional classification was determined for each patient. Angina was graded using the Canadian Cardiovascular Society Classification. Cardiac catheterization data noted include left ventricular end-diastolic pressure, pulmonary artery pressure, cardiac index, regional left ventricular score (regional wall motion scoring as defined by Coronary Artery Surgery Study) [l], mitral valve area (calculated using the Gorlin formula), and the presence and severity (grade 0 to 4) of angiographically determined rnitral insufficiency, number of major coronary vessels diseased (defined as 270% luminal obstruction in any angiographic plane), presence of left main coronary stenosis (defined as 250% luminal obstruction). M mode, two-dimensional, and Doppler echocardiography was performed preoperatively and 1 to 6 months postoperatively. Ventricular end-systolic and end-diastol-
lthough heart valve replacement is a safe and commonly performed procedure, all available prostheses are associated with valve-related complications that influence their clinical use. The most widely used prosthetic valves in the United States are the bileaflet St. Jude and the tilting-disc Medtronic-Hall. Both prostheses possess specific mechanical features that appear advantageous. The wide opening angle (70 degrees), upward gliding disc, and monostrut construction are attractive properties of the MedtronicHall valve, whereas the central flow design of the lowprofile bileaflet St. Jude prosthesis may be hemodynamically advantageous. To determine if one prosthesis has clinical or hemodynamic superiority, we designed a prospective, randomized trial to compare these two valves in patients requiring mitral valve replacement.
Material and Methods From September 1986 until May 1991, 102 patients undergoing elective valve replacement for mitral valve dysfuncPresented at the Thirty-eighth Annual Meeting of the Southern Thoracic Surgical Association, Orlando, FL, Nov 7-9, 1991. Address reprint requests to Dr Fiore, Cardiothoracic Division, Department of Surgery, St. Louis University Medical Center, 3635 Vista Ave at Grand Blvd, PO Box 15250, St. Louis, MO 63110-0250.
0 1992 by The Society of Thoracic Surgeons
0003-4975/92/$5.00
Ann Thorac Surg 1992;54:6&73
ic dimensions were measured, and ejection fraction and prosthetic valve gradient were calculated. In patients with atrial fibrillation, the mean of ten consecutive cardiac cycles was calculated. The ejection fraction was calculated by Simpson’s rule as previously described [2].Mitral valve areas were calculated by the pressure half-time method with either continuous or pulsed wave recordings [3,41. The presence and degree of prosthetic valve regurgitation was qualitatively assessed by pulse wave, continuous wave, and color Doppler echocardiography from multiple windows. Cardiopulmonary bypass was established using ascending aortic and bicaval cannulation employing systemic hypothermia and hemodilution. Myocardial protection included the use of cold hyperkalemic blood cardioplegia (6°C). The infusate consisted of a 4:l dilution of blood with 0.9% normal saline solution which contained 20 mEq/L of sodium bicarbonate and 0.1 mg/mL of xylocaine hydrochloride. The initial infusion contained 14 mEq/L of potassium, and all subsequent infusions contained 7 mEq/L of potassium. In patients in whom the aortic valve was competent, an initial dose (15mL/kg) was given through the aortic root and subsequent doses (7.5mL/kg) were infused through the coronary sinus. In patients with predominantly aortic insufficiency, all doses of cardioplegia were given through the coronary sinus. Topical myocardial cooling with Ringer’s lactate slush and a cardiac insulation pad for phrenic nerve protection were employed in all patients. Proximal vein graft anastomoses were performed using a partially occlusive aortic clamp while distal anastomoses were constructed with the aorta cross-clamped. Myocardial revascularization was performed before mitral valve replacement to avoid raising the heart after the mitral prosthesis had been inserted, thus decreasing the risk of myocardial rupture. The operative procedure was performed by the same group of surgeons using similar techniques. The mitral prosthesis was inserted with interrupted horizontal mattress sutures of 2-0 Ethibond (Ethicon, Somerville, NJ) placed through the left atrial side so as to evert the mitral annulus. When feasible, the posterior mitral leaflet was left intact. The St. Jude prosthesis was oriented either in the antianatomic axis perpendicular to the mitral commissures or in the anatomic plane parallel to the mitral commissures. The Medtronic-Hall valve was generally oriented with the larger orifice opening posteriorly. This prosthesis could be rotated according to the surgeon’s preference to prevent entrapment of the disk by the retained posterior mitral leaflet. Recorded operative variables included cardiopulmonary bypass and cross-clamp times. Prophylactic anticoagulation with sodium warfarin was carried out in all patients. Anticoagulation was instituted within 72 hours after the operation unless clinical contraindications existed. The control of therapy was made by interval analysis of the prothrombin index, which was kept at 1.5times the control value. Antibiotic prophylaxis consisted of cefazolin sodium (Ancef; Smith Mine & French Laboratories, Philadelphia, PA) administered preoperatively and for 48 hours postoperatively. Follow-up data were collected by a registered nurse
FIOREETAL PROSTHETIC MITRAL VALVE REPLACEMENT
69
using written questionnaires and telephone interviews with the patient, the patient‘s physician, or both. Patients surviving the operative period were followed up for a minimum of 1 month and a maximum of 60 months (mean, 26 months). The cumulative follow-up was 105 patient-years for the SJ and 121 patient-years for the MH group. Five patients were lost to follow-up at 1 (SJ), 1 (SJ), 10 (MH), 10 (MH), and 35 (SJ) months. The definitions of events, their classification, and the analysis of the data follow the recently suggested guidelines for reporting morbidity and mortality after cardiac valve prosthesis insertion [5].Data were analyzed using the Statview I1 statistical software package (Brainpower, Inc, Calabasas, CA). Univariate analysis of discrete variables was performed using the ,$ analysis or Fisher’s exact test where appropriate. Paired and unpaired Student’s t tests were employed for continuous variables where appropriate. Mean values are expressed as plus or minus the standard deviation except as noted. The linearized rates were computed using an incidence density model. All events were considered in the calculation of the linearized incidences, which are expressed as events per 100 patient-years of follow-up. These rates were compared using a maximum likelihood ratio. The actuarial curves were constructed using the methods described by Berkson and Gage and compared using a log rank analysis [6,71.A p value of less than 0.05 was considered significant.
Results Clinical Profile The cohorts were similar with respect to age, sex, preoperative NYHA functional class, hemodynamic and operative parameters, and the degree of mitral stenosis or insufficiency. The majority of patients had mild left ventricular dysfunction with moderate pulmonary hypertension. Twenty percent of patients in both groups had had prior valve replacement, and a similar number required concomitant myocardial revascularization (Table 1).
Operative Mortality A total of 13 patients (12%) died in the hospital. The overall operative mortality was 14.5% for the SJgroup and 10.6% for the MH group (not significant [NS]) (Table 2). The operative mortality for isolated mitral valve replacement was 16.3% for the SJ group and 11.1% for the MH group (NS), whereas that for mitral valve replacement with myocardial revascularization was 8.3% and 12.5%, respectively (NS). The most common cause of operative death was low cardiac output syndrome. Six patients died as a result of low cardiac output associated with biventricular failure. In two instances (1 MH; 1 SJ) operative death was caused by atrioventricular sulcus rupture. Fatal dysrhythmias developed in 1 patient in the MH group, whereas 1 patient receiving the St. Jude prosthesis died of mediastinitis and generalized sepsis 33 days postoperatively. Three additional patients, all in the SJ group, died of multiorgan system failure. One patient with a history of
70
FIOREETAL PROSTHETIC MITRAL VALVE REPLACEMENT
Ann Thorac Surg 1992;54:6%73
100%
Table 1. Preoperative Clinical Profile"
Variable Age (Y) Male sex (%) NYHA class LVEDP (mm Hg) Left ventricular score Pulmonary artery systolic (mm Hg) Cardiac index (L * min-' . m-') Mitral area (cm') Mitral insufficiency Coronary artery disease (%) CABG (%)
Prior valve replacement
(a)
Duration of CPB (min) Cross-clamp time (min) a
SJ
MH
P Value
58.00 +. 15.4 43.0 3.00 2 0.7 14.40 ? 6.8 7.00 ? 3.0 50.50 2 20.3
57.00 2 10.0 27.7 3.02 2 0.5 13.90 ? 7.1 7.60 2 3.6 52.10 t 16.0
NS NS NS NS NS NS
2.30 2 0.6
2.30 2 0.5
NS
401
ST.JUDE
20
0
0 PATIENTS 47 ATRISK 55
-
1
1 31 37
,
2 23
32
3 17 22
I
,
4 9 8
I
,
5 YEARS 0
Medtronic
0 St.Jude
NS NS NS
Fig 1. Actuarial survival rates after mitral valve replacement with the St. Jude and Medtronic-Hall valves. Hospital and late deaths are included. (NS = not significant.)
21.8 21.8
17 21.3
NS NS
Late Mortality
NS NS
138.50 It 45.0 147.50 2 63.0 81.00 ? 28.0 86.00 2 31.0
chronic hepatitis died 13 days postoperatively of liver and renal failure. A second patient remained comatose postoperatively; renal and respiratory failure developed, and the patient died on the fourth postoperative day. The remaining patient had undergone urgent third-time mitral valve replacement. He died of multisystem organ failure after several hypotensive episodes on postoperative day 5. The patients who died in the hospital had a mean NYHA functional class of 3.3, and 77% of hospital deaths were in patients with mitral regurgitation as the primary pathologic process. Univariate analysis demonstrated that the presence of mitral insufficiency and a left ventricular end-diastolic pressure of 20.4 k 0.7 mm Hg correlated with operative mortality (p < 0.05). Sex, pulmonary artery systolic pressure, the presence of coronary artery disease, and the global ischemic time were not correlated with hospital death. Table 2. Causes of Operative Mortality
Cause
SJ
Left ventricular failure Arrhythmia Septicemia Multiple-organ failure
MH
3 0 1 3 1
Left ventricular tear Total
8 14.5%
Operative mortality Medtronic-Hall;
Fi
0.96 2 0.42 2.70 2 1.5 23.4
CPB = cardiopulmonary CABG = coronary artery bypass grafting; bypass; LVEDP = left ventricular end-diastolic pressure; MH = Medtronic-Hall; NS = not significant; NYHA = New York Heart SJ = St. Jude. Association;
=
T
1
0.97 2 0.37 3.00 2 1.3 30.9
Values expressed as percent or mean 2 the standard deviation.
MH
C
SJ = St. Jude.
Late deaths occurred in 7 patients (6 SJ and 1 MH; 7%). The mortality rates for the SJ and MH group were 7.3% and 2.1% (NS), respectively. Valve-related complicationswere the cause of late death in 6 patients (5 SJ, 9%; 1 MH, 2%). The most common cause of valve-related death was sudden death, which occurred in 3 patients (SJ group) at 6, 37, and 39 months postoperatively. Ventricular fibrillation developed at home in 2 patients who could not be resuscitated. The remaining patient died suddenly at home in association with severe dyspnea. None of these received postmortem examination. One death each resulted from cerebral hemorrhage (MH), valve thrombosis (SJ), and cerebral embolus (SJ)at 1 month, 2 months, and 4.5 years, respectively. One valve-unrelated death (SJ) occurred at 6 months postoperatively in a patient with an ischemic cardiomyopathy.
Patient Survival The actuarial survival curve for the population is shown on Figure 1. At 3 years, the survival rates were slightly better for the patients receiving the Medtronic-Hall valve, but the difference was not significant (87 t 4.9 versus 79 +- 5.3; NS). The postoperative NYHA class in the 84 survivors was equally improved in both cohorts (MH, 1.8 Ifr. 1.0; SJ, 1.7 k 0.9; NS).
Valve-Related Complications Nineteen valve-related complications developed in 18 of the 89 hospital survivors (20%). The linearized rates of valve-related complications and the actuarial freedom from these morbid events are shown in Tables 3 and 4, respectively. There were a total of seven episodes of thromboembolism occurring in 6 patients. Two were fatal. Valve thrombosis occurred in 2 patients, 1 in each group. The patient in the SJ group died suddenly before reoperation, whereas the patient in the MH group underwent successful reoperation and is a long-term survivor.
THROMBOEMBOLISM.
5 10.6%
71
FIOREETAL PROSTHETIC MITRAL VALVE REPLACEMENT
Table 3 . Linearized Rate of Valve-Related Events" Event
Thromboembolism Sudden death Valve thrombosis Hemorrhage Endocarditis
1.6 f 1.2 2.5 f 1.4 0.8 f 0.8 2.5 f 1.4 0 0 0
Paravalvular leak
Reoperation a
MH
SJ
Expressed as events per 100 patient-years
MH
=
Medtronic-Hall;
NS
4.9
Value
NS NS NS NS NS NS NS
2.2
0
1.0 f 1.0 2.9 f 1.7 2.0 f 1.4 1.0 f 1.0 2.0 & 1.4 ?
standard error of mean.
SJ = St. Jude.
not significant;
=
f
P
Two patients in the MH group had transient ischemic attacks, and 3 patients (2 MH; 1 SJ)had presumed embolic strokes, one of which was fatal (SJ).The linearized rate of thromboembolism was higher in the MH group, but the difference did not achieve statistical significance. There were six instances of hemorrhage related to anticoagulant treatment (3 MH; 3 SJ). Bleeding occurred intracerebrally in 2 patients and was fatal in 1 at 4 weeks postoperatively (MH group). The remaining 4 patients had nonfatal gastrointestinal bleeding that did not require operation. The linearized rates were similar in both groups. HEMORRHAGE.
Bacterial endocarditis developed in 2 patients with Medtronic-Hall valves. A paravalvular leak developed in 1 patient, who underwent replacement of the prosthesis. The remaining patient was successfully treated nonoperatively. Both are long-term survivors. The linearized rate of endocarditis was not significantly different in both cohorts. ENDOCARDITIS.
PARAVALVULAR LEAK. A single patient had a paravalvular leak as noted in the previous section.
Postoperative Echocardiography Doppler echocardiographic evaluation was performed postoperatively in 33 patients in the SJ group and 32 patients in the MH group. Thirty patients (19 SJ, 11 MH) underwent evaluation during the first 3 months after operation. The remaining 35 patients (20 SJ and 15 MH) underwent Doppler echocardiography between 3 and 6 months postoperatively. The mean transvalvular gradient at rest was not significantly different between groups at any valve size (Table 5).
Table 4 . Actuarial 3-Year Freedom From Events" MH
SJ
Event
Survival Thromboembolism
98 2 2.0
88 f 5.5
Hemorrhage
91
94
a
79
Values expressed as percent
MH
=
Medtronic-Hall;
NS
f
f
5.3 5.3
87
f 4.9 f 4.4
?
standard error of mean.
=
not significant;
Table 5. Postoperative Transvalvular Echocardiographic Mean Gradient
P
Value
NS NS NS
SJ = St. Jude.
MH
SJ
Valve Size
Gradient
Gradient
(mm)
(mmHg)
N
(mmHg)
25 27 29 31 33
4.2 2.7 2.0 2.0 3.8
7 6 8 7 5
4.0 4.0 3.0 2.7 3.4
f 3.3
1.3 f 1.0 f 0.7 f 2.3 f
MH = Medtronic-Hall; SJ = St. Jude. icant;
N
=
f 2.5
1.5 f 1.2 f 0.6 f
f
1.1
number of patients;
P
N
Value
7 6 6 3 10
NS NS NS NS NS
NS
=
not signif-
Reoperations Two patients, both in the MH group, underwent successful reoperation at 6 weeks and 7 months postoperatively. The indications for reoperation were prosthetic valve endocarditis in 1 and valve thrombosis secondary to sodium warfarin noncompliance in the other. Both patients are long-term survivors.
Comment The two most common cardiac valve prostheses used in the United States are the bileaflet St. Jude and the Medtronic-Hall pivoting-disc prosthesis. These devices have undergone virtually no alteration in design since their development in 1977 and 1978, and had remained unmodified since this study commenced in 1986. Several investigators have reported the 5-year and 10-year clinical and hemodynamic results of both valves (8-101; however, a prospective, randomized trial comparing both prostheses in the mitral position has not been reported. As in other series, the leading cause of hospital mortality in our series was perioperative heart failure. The overall operative mortality was similar to that reported by other investigators [ l l , 121. Patient survival based on 3-year actuarial estimates was not significantly different between groups. Our mediumterm survival is comparable with those of published reports for the St. Jude and Medtronic valves [13, 141. There were 3 cases of unobserved or sudden death in our series. Although autopsy information is not available, these 3 patients had end-stage cardiac failure and had no evidence in their postoperative evaluation of prosthetic valve dysfunction. At the time of their death, all 3 patients had therapeutic anticoagulation levels. Survival itself is usually regarded as an insensitive measure of overall valve performance and relates mainly to the underlying disease and the perioperative status of the myocardium. A more specific measure of valve performance is thromboembolism, which is a major risk factor for patients with prosthetic valves. The incidence of thromboembolic complications in the third-generation valvular devices has decreased dramatically. This has been due to improved hemodynamic design resulting in a better flow velocity profile, less turbulence, and reduction
72
FIOREETAL PROSTHETIC MITRAL VALVE REPLACEMENT
R C
T
I
C
I
20
0 0 PATIENTS 47 ATRISK 55
Ann Thorac Surg 199254 :6a73
1 28 37
2 20 32
3 16 22
p=NS
4 9 8
I 1
5 YEARS 0 Medtronic 0 St. Jude
Fig 2. Actuarial freedom from thromboembolism after mitral valve replacement with the St. Jude and Medtronic-Hal1 valves. (NS = not significant.)
of stagnant areas, all of which are known to affect the thrombogenicity of the prosthesis. In the Medtronic-Hall valve, a more central position of the disc in the open position enlarged the minor orifice, thus reducing stagnation of blood and turbulence in this area. The central flow design of the bileaflet St. Jude valve may contribute to its low incidence of thromboembolic events. In this report, the 3-year actuarial freedom from thromboembolism was slightly better for the St. Jude prosthesis, but the difference was not significant (Fig 2). The 4 patients in whom nonfatal thromboembolic events developed were adequately anticoagulated. Dipyridamole (300 mg daily) was added to their anticoagulation regimen to reduce the recurrent risk of thromboembolism. There were 2 cases of valve thrombosis, 1 in each group. Noncompliance with sodium warfarin anticoagulation in the presence of atrial fibrillation was believed to have contributed to thrombus formation in both patients. In this series, the overall rates of thromboembolic complications compared favorably with those reported by Arom and associates 1151 for the St. Jude valve and Keenan and co-workers [16] for the Medtronic-Hall valve. Hemorrhage was the second most common valverelated complication. This complication was fatal in 1 hypertensive patient in whom an intracerebral hemorrhage developed. The actuarial freedom from hemorrhage was similarly low in both cohorts and compared favorably with that reported in other series for the Medtronic and St. Jude valves (Fig 3 ) 117, 181. The hemodynamic advantages of the St. Jude prosthesis, especially in smaller valve sizes, have been documented in the literature 119, 201. In this report, we measured the mean transvalvular gradient between 1 and 6 months postoperatively using Doppler echocardiography. No significant difference in the resting mean gradient could be demonstrated when both prostheses were compared at any valve size. Improved hemodynamic performance has been associated with symptomatic improvement. In this series, the St. Jude and Medtronic valves achieved equal improvement in patient symptoms as evidenced by the low postoperative NYHA functional score observed in both cohorts. The results of this study indicate that in the first years after implantation, the performance of the St. Jude and
0
T
20
MEDTRONIC ST.JUDE
0 0 PATIENTS 47 ATRISK 55
1 1 2 3 4 5 YEARS 31 22 17 8 0 Medtronic 36 30 20 7 0 St.Jude Fig 3. Actuarial freedom from hemorrhage after mitral valve replacement with the St. Iude and Medtronic-Hall valves. (NS = not significant.)
the Medtronic-Hall prostheses are comparable in this patient population. These two valves are among the best of the modern generation of prostheses, and either may be confidently recommended as valve substitutes. The choice of one or the other should be based on the surgeon’s experience and preference, which may be linked to other factors such as ease of insertion, availability, and cost. Clearly, this study suggests that neither valve can claim superior performance or safety. We thank Jan Castanis, RN, and Jan St. Vrain, RN, for special technical assistance. We gratefully acknowledge Jean Wamhoff, Scott Langreder, Mary Nonte, and Karen Kelly for their assistance with preparation of the manuscript and Gary Grunkemeier for his assistance with the statistical analysis.
References 1. Principal investigators of CASS and their associates. The National Heart, Lung, and Blood Institute Coronary Artery Surgery Study. Circulation 1981;63(Suppl 1):1-81. 2. Mehdirad AA, Williams GA, Bryg RJ, et al. Evaluation of left ventricular function with upright exercise: correlation of exercise Doppler with post-exercise wall motion and single plane echocardiographic ejection fraction. Circulation 1987; 75:413-9. 3. Hatle L, Angelsen B, Tromsdal A. Non-invasive assessment of atrioventricular pressure half-time by Doppler ultrasound. Circulation 1969;60:109&104. 4. Bryg RB, Williams GA, Labovitz AJ, Aker U, Kennedy HL. Effect of atrial fibrillation and mitral regurgitation on calculated mitral valve area: comparison of Doppler, twodimensional echocardiography and cardiac catheterization. Am J Cardiol 1986;57:63&8. 5. Edmunds LH Jr, Clark RE, Cohn LH, Miller DC, Weisel RD. Guidelines for reporting morbidity and mortality after cardiac valvular operations. Ann Thorac Surg 1988;46:257-9. 6. Berkson J, Gage RP. Calculation of survival rates for cancer. Proc Meet Mayo Clin 1950;25:270-86. 7. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959;22:71943. 8. Gray R, Chaux A, Matloff J, Raymond M. Early postoperative hemodynamic comparison of !St. Jude cardiac prosthesis and porcine xenografts at rest and exercise [Abstract]. Circulation 1979;6O(Suppl 2):222. 9. Antunes MJ, Wessels A, Sadowski RJ, et al. Medtronic-Hall valve replacement in a third-world population group: a
FIOREETAL PROSTHETIC MITRAL VALVE REPLACEMENT
10. 11. 12.
13.
14.
review of the performance of 1,000 prostheses. J Thorac Cardiovasc Surg 1988;95:980-93. Antunes MJ. Clinical performance of St. Jude and MedtronicHall prostheses: a randomized comparative study. Ann Thorac Surg 1990;50:743-7. Nicoloff DM, Emery RW, Arom KV, et al. Clinical and hemodynamic results with the St. Jude Medical cardiac valve prosthesis. J Thorac Cardiovasc Surg 1981;82:674-83. Butchart EG, Lewis PA, Grunkemeier GL, Kulatilake N, Breckenridge IM. Low risk of thrombosis and serious embolic events despite low-intensity anticoagulation. Experience with 1,004 Medtronic-Hall valves. Circulation 1988;78(Suppl 1):66-71. Arom KV, Nicoloff DM, Kersten TE, Northrup WF 111, Lindsay WG. Six years of experience with the St. Jude Medical valvular prosthesis. Circulation 1985;72(Suppl 2): 153-8. Hall KV, Nitter-Hauge S, Abdelnoor M. Seven and one-half years‘ experience with the Medtronic-Hall valve. J Am Coll Cardiol 1985;6:1417-21.
73
15. Arom KV, Nicoloff DM, Kersten TE, et al. Ten years’ experience with the St. Jude Medical valve prosthesis. Ann Thorac Surg 1989;47:831-7. 16. Keenan RJ, Armitage JM, Trento A, et al. Clinical experience with the Medtronic-Hall valve prosthesis. Ann Thorac Surg 1990;50:748-53. 17. Nitter-Hauge S, Semb 8, Abdelnoor M, Hall KV. A 5 year experience with the Medtronic Hall disc valve prosthesis. Circulation 1983;68(Suppl 2):169-74. 18. Chaux A, Gray RJ, Matloff JM, Feldman H, Sustaita H. An appreciation of the new St. Jude valvular prosthesis. J Thorac Cardiovasc Surg 1981;81:202-11. 19. Harstkotte E, Haerten K, Seipel L, et al. Central hemodynamics at rest and during exercise after mitral valve replacement with different prostheses. Circulation 1983;68(Suppl 2):161-8. 20. Gil CC, Ing HC, Lytle BW, Cosgrove DM, Golding LAR, Loop FD. Early clinical evaluation after aortic valve replacement with the St. Jude Medical valve in patients with a small aortic root. Circulation 1982;66(Suppl 1):147-9.
DISCUSSION DR OSCAR R. BAEZA (Newark, NJ): I congratulate you for your nice presentation. I have a question, though, and that relates to the preservation of the posterior leaflet of the mitral valve or the subvalvular apparatus, vis a vis the profile of both the St. Jude and Medtronic-Hall prostheses. I participated in some of the clinical evaluations when the Medtronic-Hall was called the Hall-Kaster valve, and we had problems with the occluder being stuck in a closed position by interference with suture tails or other subvalvular structures. My question to you is, do you see an advantage of using a lower profile valve, like the St. Jude, in patients who have small ventricles and in whom you would like to preserve the subvalvular apparatus? I stopped using the Medtronic-Hall valve many years ago because the occluder enters the housing like a door that closes inside its frame. Like a bathroom door that is slammed stuck with a towel hanging from the knob, the occluder of the MedtronicHall may also get stuck in a closed position. In my opinion, the St. Jude prosthesis offers the advantage of a very low profile and very little chance to interfere with sutures or the mitral subvalvular apparatus during mitral valve replacement. D R KIT V. A R O M (Minneapolis, MN): I would like to congratulate Dr Fiore for a very interesting presentation. This communication is one of the very few prospective, randomized studies, of which we need more. I think he has confirmed the previous studies that the St. Jude Medical and the Medtronic-Hall are probably the better two mechanical valves available at the present time. However, in the previous in vitro study, the St. Jude valve had some advantage because of bileaflet design, low pressure gradient, and better effective orifice areas. You also confirm that the incidence of thromboembolism is low in both of these prostheses. There is a cumulative report, about 4,000 cases in each group, showing that the incidence of thromboembolism for both St. Jude and Medtronic valves is very low, being 1.6 and 2.6 episodes/100 patient-years, respectively. I have one comment about using Doppler flow study to obtain the gradient. As you know, Doppler picks up the velocity and the gradient is derived from this. Because of the design, the St. Jude has a higher velocity in the central opening of the valve, which could give you a false high gradient.
I also have a few questions. First, how many of the patients in the Medtronic-Hall group had chordae or supporting structure of the mitral valve preserved, and did you have any trouble from the high profile of the Medtronic-Hall occluder relating to this? Second, is the incidence of bleeding higher than what we usually see with the St. Jude valve? What is the prothrombin level in this group? We now keep the prothrombin time between 18 and 20 seconds for the St. Jude valve, and by doing so we are able to cut down the incidence of bleeding significantly. The other question is, why do you see a higher gradient in the 33-mm prostheses of both groups? Can you explain why the gradient in some of the larger valves is higher than the smaller ones? The last question is, have you done a similar study in the aortic position? D R FIORE: I will answer Dr Arom’s questions first. We did try to retain the posterior leaflet whenever possible. In patients with mitral stenosis it was not usually possible to accomplish this safely. In patients with mitral insufficiency secondary to myxomatous degeneration, great care was taken to be certain that the posterior leaflet was carefully reefed under tension when inserting the prosthesis. To best accomplish this, sutures were placed from the left atrial side so as to evert the posterior leaflet and the annulus. This avoids any potential disruption of the disk with the subvalvular apparatus. We had one fatal episode of anticoagulant-related hemorrhage. This patient had atrial fibrillation and known left atrial thrombus, and for these reasons, the prothrombin time was kept at approximately two times the control. This may have accounted for the bleeding episode. In general, we try to keep our prothrombin times 1% times the control. The Doppler-derived transvalvular gradients did show slightly elevated values in the larger size valves, although all the numbers were quite low. With respect to your last question, we have performed a prospective, randomized trial comparing both valves in the aortic position. The results are currently being analyzed. Dr Baeza, we agree that in patients with a small left ventricle in whom you wish to preserve the subvalvular apparatus, there may be an advantage of using the lower profile St. Jude valve. Our results suggest that this advantage would be technical and not hemodynamic.
