Valve replacement in the young patient with rheumatic heart disease

J

THORAC CARDIOVASC SURG

1990;99:631-8

Valve replacement in the young patient with rheumatic heart disease Review of a twenty-year experience During a 20-year period 303 young subjects between 9 and 20 years of age (mean, 16.2 ± 2.72 years) with rapid and relentlessly progressive valvular disease from rheumatic fever underwent valve replacements. The Starr-Edwards ball valve prosthesis remains the device of choice, although other valves have been implanted. The overall hospital mortality rate was 9.6% in the mitral valve, 3.5% in the aortic valve, and 4.2% in the double valve replacement groups. Actuarial survival at 10, 15, and 20 years was 78.4 % (± 3.3 % ~ 70.0 % (± 5.8 % ~ and 59.3 % (± 11.1 % ~ respectively, for patients with mitral valve replacement The rates for aortic valve replacement were 85.9% (±4.6%) at 10 and 15 years and 72.7% (12~8) at 20 years. In the double valve replacement group the survival rates after 5 and 10 years were 79.9% (±5.1 %). The incidence of thromboembolism was 0.41, 0.59, and 1.04 per 100 patientyears for the mitral, aortic, and double-valve prostheses, respectively. The prospect of childbearing seems promising in those young women who were subsequently married. Our favorable and gratifying experience in this review bears testimony to the physiologic advantages of the Starr-Edwards valve as the device of choice in the rehabilitation of patients with advanced and severe valvular disease after rheumatic fever.

Stanley John, MS, MS(Thor), FACC,a Edwin Ravikumar, MS, MNAMS, MCh,a P. S. Jairaj, MS, FRACS,a Ujjwal Chowdhury, MS, and S. Krishnaswami, MD, DM(CRD), FACC, FRCP,b Vel/ore, India

In many areas of the Third World the valvular disease that follows rheumatic fever results in profound cardiac disability at a very young age. I, 2 We have previously shown that closed transventricular mitral valvotomy offers excellent results for most of the young patients with mitral stenosis.' Consequently, we continue to be advocates of this operative approach for most cases of mitral stenosis. However, for the last two decades we have done valve replacements in patients with significant mitral or aortic regurgitation or a combination of both. In the

younger age group in particular, selection of a cardiac valve prosthesis remains difficult for the surgeon. In view of the acknowledged advantage of long-term durability and increased thromboresistance, we have relied heavily on the Starr-Edwards prosthesis (Baxter Healthcare Corp., Edwards Division, Santa Ana, Calif.) as the device of choice. In a previous series from our institution we reported the early and late results in 118 young subjects who underwent mitral valve replacement." This article describes our further experience with similar subjects as well as others in whom aortic and multiple valves were replaced.

From the Department of Thoracic and Cardiovascular Surgery" and the Department of Cardiology," Christian Medical College Hospital, Vellore, India.

Methods

Received for publication Nov. 9, 1988. Accepted for publication June 9, 1989. Address for reprints: Stanley John, MS, MS, FACC, Department of Cardiothoracic Surgery, Christian Medical College Hospital, Vellore 632 004, India.

12/1/15523

Patients. Between 1967 and June 1988, 303 young subjects ranging from 9 to 20 years of age (mean, 16.2 ± 2.72 years) underwent mitral (MVR), aortic (A VR), or combined mitral and aortic valve replacement (DVR) with or without concomitant tricuspid valve repair at the Christian Medical College Hospital in Vellore, India. The MVRs and AVRs were performed over a 20-year period and the DVRs were done since 1977.There were 169 male and 134 female patients; 68 (22.4%)

631

6 32

The Journal of Thoracic and Cardiovascular

John et al.

Surgery

Table I. Valvular profile with indications for replacement Profile

Mitral valve disease' Predominantmitral regurgitation Mitral stenosis Previous closed valvotomy Closed revalvotomy Traumatic postoperative regurgitation during transventricular valvotomy Aortic valve disease Predominantregurgitation (including 11 cases of associated mitral stenosis) Stenosis and regurgitation Predominantaortic stenosis Mitral and aortic valve disease' Dominant mitral regurgitation and mixed aortic valve disease Dominantmitral stenosis and dominant aortic stenosis Total

No. of cases

198 184 14 10 I 3

57 47 4 6

48 47

303

'Includes 19 cases of Lutembacher's syndrome (mitral = 18; combined = 1).

were younger than 15 years of age. One hundred twenty-two patients (40.2%) weighed less than 70% of ideal body weight for their age (based on weight for age by American standards). In the MVR, AVR, and DVR groups, 72.0%, 72.7%, and 84.8%, respectively, belonged to New York Heart Association (NYHA) functional class IV, and the remainder were in class Ill. Atrial fibrillation was present in 44% of those who underwent MVR, in 30% of those with DVR, and in only 5% of the A VR group. The valvular profile of the patients studied who had indication for replacement in each situation is depicted in Table I. Hemodynamic features. Cardiac catheterization and angiocardiography were performed in all except two patients. The cardinal hemodynamic features are shown in Table II. The pulmonary artery peak systolic pressures ranged from 20 to 140 mm Hg in MVR, 14 to 75 mm Hg in the AVR, and 20 to 90 mm Hg in the DVR groups. The systemic index ranged from 1.2 to 4.8 Ljminjm 2 . In 19 patients a significant left-to-right shunt was documented at the atrial level. The left ventricular end-diastolic pressure ranged from 3 to 30 mm Hg (mean, 13.3 ± 6.7 mm Hg) in the A VR group. Technical considerations. Cardiopulmonary bypass was accomplished by a disc oxygenator earlier in the series and subsequently by the R ygg-Kyvsgaard (Cardiovascular Specialties Ltd., Scarborough, Ontario, Canada), Shiley (Shiley, Inc., Irvine, Calif.), or Bentley (Bentley Laboratories, Inc., Baxter Healthcare Corporation, Irvine, Calif.) oxygenators. Dextran 40 (Rheomacrodex) and Ringer's lactate solution have been used as diluents in the prime. Continuous electrical fibrillation was used early in our experience, and cold potassium cardioplegia techniques were used since 1977. In addition to cold potassium cardioplegia, added protection is offered by topical hypothermia. Forty-eight patients in the MVR group had significant tricuspid regurgitation, whereas this association was seen in only one patient who underwent DVR. Tricuspid regurgitation was evaluated before cardiopulmonary bypass was begun, and con-

comitant tricuspid annuloplasty with the De Vega or the KayMendez-Zubiate technique was accomplished. In those with Lutembacher's complex, a large atrial septal defect was closed with a patch in 19 subjects (18 were in the MVR group and one in the DVR group). Associated mitral stenosis that was evident in 19.3% of patients required open valvotomy and aortic valve replacement. The types of prostheses and the sizes implanted are shown in Table III. The Starr-Edwards ball valve prosthesis (model 6120) was the commonly used implant throughout the study. However, in 24 subjects before 1975, cloth-covered composite seat prosthesis models 6300, 6310, and 6320 were used. In the aortic valve study, model 1260 was always used. However, other valve implants were inserted in a small number of patients. Follow-up method. Every survivor was advised to report to our follow-up clinic at the end of the first year after operation. At that time all important events such as thromboembolic manifestations, NYHA functional status, recurrent rheumatic activity, and infective endocarditis were documented. Survivors were subsequently advised to visit the clinic after every 3 years. Absentees were contacted by mail and questionnaires sent to their personal physicians or to the patients themselves to obtain the follow-up information. When all efforts failed, follow-up was accomplished by a social worker who was sent to visit these patients. Statistical analysis. Mortality rates were calculated on the basis of the total number of years of follow-up for each patient. Survival curves were drawn based on the rates of survival and standard error computed according to the methods of Cutler and Ederer.P

Results The number of patients who underwent various valve replacements between 1967 and 1988, the. number of early and late deaths, and the number lost to follow-up are shown in Table IV. The operative mortality rate (deaths within 30 days of operation) in the MVR group was 9.6%, which has declined to 2.4% (3 of 129) in the last 5 years. It was 3.5% in the AVR group and 4.2% in DVR group. The follow-up period ranged from 1 to 20 years in the MVR and AVR groups (mean, 6.2 and 6.3 years), whereas it was 1 to 10 years in the DVR group (mean, 4.0 years). Only 2% (four in MVR and three in A VR series) were lost to follow-up, for an overall successful follow-up rate of 98% of patients. The overall mortality rates including hospital mortality rate for MVR, A VR, and DVR groups were 34, 25, and 36 per 1000 per year, respectively. If only those who survived the first 30 days are considered, the mortality rates (per 1000 per year) in the MVR, AVR, and DVR groups are reduced to 18.9, 19.7, and 26.3, respectively. The rates in children (9 to 14 years) were significantly lower compared with those among teenagers. The survival curves for the three groups of patients are shown in Fig. 1. This portrayed the estimated probability (± standard error) of survival at 10, 15, and 20 years as

Volume 99

Valve replacement in rheumatic heart disease 633

Number 4 April 1990

Table II. Cardinal hemodynamicfeatures Systemic index (L/min/m 2)

Pulmonary artery (systolic) pressure (mm Hg)

Mean

Mean Group

Range

(SD)

Range

(SD)

MVR AVR DVR

20-140 14-75 20-90

57.6 (20.6) 26.6 (12.0) 49.0 (20.5)

1.2-3.8 1.7-4.8 1.6-3.8

2.1 (0.8) 3.0 (0.2) 2.6 (0.9)

SD. Standard deviation.

78.4% (± 3.3%),70.0% (± 5.8%), and 59.3% (± ILl %), respectively, for patients with mitral valve replacement. For patients with aortic valve replacement, the rate at 10 and 15 years was 85.9% (± 4.6%) and at 20 years, 72.7% (± 12.8%). Forthosewith DVR, the survival rate after 5 and 10 years was 79.9% (±5.l%). There were 35 late deaths among the 280 survivors (MVR, 23; AVR, 7; DVR, 5). In five patients death was sudden, presumably the result of arrhythmias; in five patients hemorrhage from overdosage ofanticoagulants was the causative factor. Valve thrombosis in one boy in the DVR group represents the only proved instance of such a complication in our series. A young woman died 19 years after valve replacement after a cesarean section. Another died suddenly after delivery of a normal child. Hemiparesis with homologous serum hepatitis resulted in death in one patient. Congestive cardiac failure occurred in eight patients. Four subjects died after infective endocarditis; one of these patients died as a result of a fungal infection. Bronchopneumonia caused death in two subjects. Unrelated noncardiac causes accounted for the deaths in the remaining seven subjects. Ninety-three percent of the survivors were in NYHA functional class I, and the remaining subjects were in class II when last followed up. Radiologic changes paralleled clinical improvement (Figs. 2 and 3). In 36 subjects for whom hemodynamic data were obtained in the MVR group 1 to 4 years after operation, a pronounced decrease in the pulmonary artery pressure to normal or near normal levelswas noted. During follow-up there was no documented evidence of rheumatic reactivity in any of our subjects because they were receiving long-term chemoprophylaxis with penicillin (compliance rate, 95%). All subjects were taking anticoagulants with coumarin or one of its derivatives with the objective of maintaining a prothrombin time 1.5 times that of the control level; this regimen was in effect since 1975. Patients living in remote areas of the country or in neighboring countries have had problems related to the standardization of anticoagulant regimen. This is because of the variability in the standards of laboratory controls. Nevertheless, this was overcome

Table III. Types ofprostheses Prosthesis

Mitral

Aortic

Multiple

Starr-Edwards ball Bjork-Shiley disc Biologic Others

175* 12

46 8

80 16

8

3

'Size 1M = 53; 2M

3

= lIS; 3M = 7.

by adopting the 1:1.5 optimal prothrombin control, which has been successful. Five percent of our patients among 280 survivors were noncompliant with regard to their anticoagulant therapy, but they had no embolic complications. A thromboembolic event is defined as an acute peripheral embolic episode or an acute neurologic event with localizing signs. Furthermore, we classify cerebral embolism with reversible ischemic neurologic deficits. In this study group there was not a single instance of permanent neurologic sequelae persisting after 2 weeks. There were fivethromboembolic occurrences in the MVR group, and two each in AVR and DVR groups. The rates per 100 patient-years are therefore 0.41, 0.56, and 1.04 in the MVR, AVR, and DVR groups, respectively. Of the two instances of late major embolic phenomena in the AVR group, one occurred in a boy who had peripheral emboli after a fungal endocarditis and resulted in his death. Another patient had hemiparesis and serum hepatitis; this patient also died. If the former were to be excluded, there is only the latter instance of thromboembolism, denoting an incidence of 0.28 per 100 patient-years in the AVR group. Late infective endocarditis developed in four patients; only one patient could be successfully treated with appropriate antibiotics. Among five subjects who had an lonescu-Shiley pericardial valve (Shiley, Inc.), two are awaiting rereplacements because of fibrocalcific prosthetic obstruction. It is pertinent to mention that a young boy who had two closed mitral valvotomies and a placement of a Lillehei-Kaster valve (Medical Inc., Inver Grove Heights, Minn.) underwent rereplacement of this device with a Starr-Edwards valve prosthesis 7 years later.

The Journal of Thoracic and Cardiovascular

John et al.

634

Surgery

100

90

12 2

G3 80

.=:

45

t

~

iZl

+oJ

18

4

AVf<

70

C

.. 60
.:.;

~I

0..

AVR

50

o

f

.

i

....

30d

4

0

OVR

•

I'IVR

A

HVR

i

5

4

3

7

6

8

'I}

10 11 12 13 14 15 16 Follow up Years

17 18 1:l

20

Fig. 1. Actuarial survival of patients with mitral, aortic, and double valve replacements (± SE). Time zero is the time of the operation.

Table IV. Patient data from 1967 to 1988 MVR

AVR

Died

Died Operation year

No. of operations

1967-1971 1972-1976 1977-1981 1982-1986 1987-1988 Total

22 30 66 69 11 198

* Early,

DVR

Lost to follow-up

No. of operations

Died

Early *

Latei

6 6

9

8

3

5

2 5

7 2

3 1

4 17 23

1 2 1

19

23

4

57

Early*

Later

Lost to follow-up

No. of operations

3

1 12 32 3 48

5 2

7

Early *

Latei

Lost to follow-up

3

2 2

5

<30 days.

t l.atc, =":30 days.

Discussion The replacement of aortic and mitral valves in adults with advanced rheumatic heart disease has long been considered standard treatment. 6-8 In Third World countries the severity and rapid progression of the valvular disease subsequent to rheumatic heart disease in the pediatric age group and in teenagers pose a challenging surgical problem. There have been few reports with longterm' follow-up of young subjects undergoing valve

replacementsj" 10 and controversy exists whether valve replacement or repair is the therapy of choice." II Functional class IV symptoms often constitute compelling indications, particularly when associated with isolated mitral or combined mitral and aortic valvular lesions. In this study 72% of subjects who underwent MVR and 84.8% of the DVR group were in NYHA functional class IV. It is noteworthy that 7.5% of subjects undergoing MVR in this study had a systemic index less

Volume 99 Number 4 April 1990

Valve replacement in rheumatic heart disease

635

Fig. 2. A, Preoperative chest x-ray film with cardiomegaly of severe mitral incompetence. B, Postoperative film of the same patient 2 years after Starr-Edwards valve implantation.

Fig. 3. A, Preoperative chest x-ray film of a l2-year-old boy with evidence of severe mitral and aortic incompetence. B, Two years after Starr-Edwards and Bjork-Shiley valve implants.

than or equal to 1.5 L/min/m 2• Bryant and Trinkle'? cited a correlation between operative death and the systemic index and indicated a mortality rate of 46% in those patients in whom the preoperative systemic index was 1.6 Lzrnin/rn? or less. Forty percent of the patients showed evidence of malnutrition, this association was reported

earlier.P Nonetheless, operation was not denied to any on the basis of advanced cardiac disease or poor nutritional state. The major indications for valve replacement in children reported previously have not been related to sequelae of rheumatic heart disease but to mitral regurgitation caused by endocardial cushion defects.!"

The Journal of

636 John et al.

The severityof anatomic changes seen in the valves in our patients precludes reparative procedures; hence, we prefer the certainty of hemodynamic correction with valve replacement to the possibility of late recurrence of the disease with valvuloplastic techniques.v? Chauvaud and colleagues!' indicated the need for reoperation in 21.5%after 10 years of follow-up of subjects undergoing valve repair, despite a large number of patients in their series belongingto NYHA functional class II. However, Reed and co-workers'" and Stevenson and colleagues.l" in a small series, reported encouraging experience with mitral annuloplasty for rheumatic mitral regurgitation. By contrast, in our study with all patients in functional classesIII and IV, the reoperationrate for prostheticvalve dysfunctionhas been negligible; onlyone patient required repeat replacement. This was a young boy who had previously had two closedvalvotomies and a LiIIehei-Kaster valveon whichthe thrombus that occurred preventedthe normal excursionof the disc. Seven years later he underwent successful repeat replacementwith a Starr- Edwards ball valve. It is pertinent to state that all except three valveswere the adult size;this was possible becauseof the presence of cardiomegaly from predominant mitral regurgitation. The overall hospital mortality rate in our study was 9.7% for those patients undergoingMVR (only 2.4% in the last 5 years) and 4.2%for those in the DVR group. In a recent review Robbins, Bowman,and Maim 17 reported that valvereplacement in children continuesto be a highrisk venture, with an overall operative mortality rate of 12%and a mortality rate as high as 23% whenboth mitral and aortic valves require replacement. Klint and colleagues 18 cited an early death rate of 17.1 % after mitral valve replacement, whereas Mathews and coworkers'? noted a 21% hospital mortality rate and concluded that valvereplacement is a serious undertaking in the pediatric age group.Other investigators10,20 indicated slightly lower mortality rates. Optimal perioperativemyocardial protectioniscritical to the preservation of the hypertrophied or decompensated ventricle.This was alluded to in our earlier report." Cardioplegictechnique was used from January 1977and additive protection was offered by topical hypothermia. The judicious use of inotropic agents both intraoperatively and postoperatively provided further therapeutic support to several of these subjects. Tricuspid regurgitation was a powerfulpredictor of postoperative outcomein this study and in previous reports.21-23 This factor was reinforcedin a previous study from our center'": 48 patients in the MVR group and one in the DVR group with associated tricuspid incompetence underwent concomitant annuloplasty with either De Vega or Kay techniques.

Thoracic and Cardiovascular Surgery

We used the Starr-Edwards ball valve in most of our young subjects because of our previous experience of excellent durability and low thrombogenicity with this device. The enlarged hearts resulting from valvularrgurgitation have accepted an adult-size prosthesis in almost everysubject for whom this implant waschosen. In a few instances, however, when the stenotic element was predominant, we chose a low-profile disc valve, such as the Bjork-Shiley valve (Shiley, Inc.). Furthermore, in those subjects in whom aortic valvereplacement was necessary in the narrow roots, we chose a Bjork-Shiley valve. We avoided the use of biologic valves in our young patients becauseof acceleratedfibrocalcific obstructionassociated with these valves when implanted in patients under 20 years of age.25 Early in this experience we favoredthe 10nescu-Shiley pericardial valve as an alternative in five subjects in whom long-term safe anticoagulant control was not considered possible. Two of these are awaiting repeat replacement because of intervening fibrocalcific obstruction. The functional classification has improved remarkably, and at the time of a recent survey(mean follow-up, 5.9 years) on whichthis report is based, 92.9% of our patients were in functionalclass I and the remainder were in class II. The functional NYHA classification provides a subjective measure of hemodynamic performance of prosthetic valves after operation. Sixty-five to seventypercent of valve-related events are said to be caused by thromboembolic phenomena. 26- 28 The occurrence rate of thromboembolism in .our study was 0.41 per 100 patient-years in the MVR group. Our experience thus contrasts strikingly that of Vidne and Levy,29 who reported an incidence as high as 20% in young patients. Other authors30,31 have also reported a high incidenceof systemic thromboembolism. However, another report from India indicated a low incidence.'? Earlier reviews from this center in 1973and 1983alluded to a similar experience.v? We used anticoagulant therapy with coumarin or one of its derivatives in all survivors. The brochures that are given to each patient have underlined the importance of adhering to a constant and optimal dose of anticoagulant rather than therapeutic anticoagulation.An analogymay be drawn between the low incidenceof thromboembolic complications seen heretoforeand the very lowincidence (<1%) of pulmonary thromboembolism in a random necropsy series in this institution. The postmortem incidence of pulmonaryembolismin a United Kingdomhospital population reported by Morrell and Dunnillf in 1968was 51.7%. These data lend support to the observations in the follow-up of our patients after valvereplacement but do not explain the mechanism that seems to protect the Indian patient from the development of

Volume 99 Number 4 April 1990

thromboembolic disease. Arom and colleagues':' reported an incidence of 2.4 to 4.6 per 100 patient-years of anticoagulant-related hemorrhage during the follow-up period in their review. Bradley and colleagues'" cited 22 episodes of hemorrhage per 100 patient-years in young children in the warfarin group. Our less stringent anticoagulant control and education of the patient taking anticoagulants afford protection against systemic thromboembolism and anticoagulant-related hemorrhage. Over the last 15 years 23 young women undergoing mitral or aortic valve replacement who were subsequently married carried 27 pregnancies to term and were delivered of 26 normal children. Oakley and Doherty" and Pavankumar and colleagues'" reported a slightly higher incidence of fetal death in their series of patients with valve replacements who carried a pregnancy to term. Salazar and co-woriers'? concluded that women with cardiac valve prostheses should be counseled against becoming pregnant. It is conceivable that our low rate of complications both in the expectant mother and in the newborn infant may be the result of our policy of substituting heparin administered intramuscularly for the coumarin derivatives during the last trimester. Heparin is administered as 5000 units twice daily by deep intramuscular injection. No complications of hematoma have been reported to date. It is therefore apparent that precise anticoagulation is less critical when the Starr-Edwards ball valve prostheses are used. There was only a negligible incidence of rheumatic reactivity during the period of follow-up that ranged from 1 to 20 years in the MVR and AVR groups. Hemodynamic data revealed gratifying results in those subjects who had recatheterization after mitral valve replacement. There was only one instance of paravalvar leak in the AVR group. Life-table survival rates (± standard error) revealed 59.3%(± 11.1%) and 72.7% (± 12.8%) at20yearsforthe MVR and AVR groups, respectively. It was 79.9% (±5.1%) in the DVR group at 10 years of follow-up. Comparable survival curves cited by Robbins, Bowman, and Malm'? in a recent review were less favorable, with only 44% and 70% 5-year survival for MVRs and AVRs, respectively. Tepley and co-workers'f mentioned a 10year actuarial survival (± standard error) of 56% ( ± 2%) after valve replacements. The remarkably low incidence of thromboembolic complications combined with the long-term durability of Starr-Edwards prostheses remain the foremost factors supporting our policy of valve replacement, especially when considered in light of the advanced nature of the disease encountered at the time of operation. With our patients at low risk of thromboembolism, we recently have been using aspirin, 300 mg once daily, as the

Valve replacement in rheumatic heart disease

637

sole anticoagulant agent in selected subjects; our results and the experience of others''? have been promising. We conclude that valve replacement in young patients is certainly an attractive and valid alternative to reparative procedures in severe valvular dysfunction of rheumatic origin involving the mitral and aortic valves. We appreciate the statistical expertise offered by Professor P. S. S. Sundar Rao. We are grateful to K. Mani for statistical assistance and R. Mohanavelu and V. Nagarajan for secretarial help. REFERENCES I. Paul ATS. The problem of mitral stenosis in childhood. Ann R Coll Surg Engl 1967;41:387-402. 2. Cherian G, Vytilingam Kl, Sukumar IP, Gopinath N. Mitral valvotomy in young patients. Br Heart J 1964;26:15766. 3. John S, Krishnaswami S, Jairaj PS, et al. The profile and surgical management of mitral stenosis in young patients. J THORAC CARDIOVASC SURa 1975;69:631-8. 4. John S, Jayaraj PS, Muralidharan S, et al. Mitral valve replacement in the young patient with rheumatic heart disease. Early and late results in 118 subjects. J THORAC CARDIOVASC SURa 1983;86:209-16. 5. Cutler SJ, Ederer F, Maximum utilisation of the lifetable method in analysing survival. J Chron Dis 1968;8:699-712. 6. McGoon MD, Fuster V, McGoon DC, et al. Aortic and mitral valve incompetence: long term follow up (10-19 years) of patients treated with the Starr-Edwards prosthesis. J Am Coll Cardiol 1984;3:930-8. 7. John S, Faruqi A, Jairaj PS, et al. Mitral valve replacement: experience with 623 consecutive patients. Indian Heart J 1984;36:352-7. 8. John S, Jairaj PS, Muralidharan S, et al. Aortic valve replacement in India. Early and long term results. J Cardiovase Surg 1986;27:207-12. 9. John S, Munsi SC, Gupta RP, et al. Results of mitral valve replacement in young patients with rheumatic heart disease. J THORAC CARDIOVASC SURa 1973;66:255-64. 10. Iyer KS, Reddy KS, Rao 1M, Bhatia ML, Gopinath N, Venugopal P. Valve replacement in children under 20 years of age---experience with Bjork-Shiley valve. J THORAC CARDIOVASC SURa 1984;88:217-24. 11. Chauvaud S, Perier P, Tonati G, et al. Long term results of valve repair in children with acquired mitral valve incompetence. Circulation 1986;74(Pt 2):1104-9. 12. Bryant LR, Trinkle JK. Cardiac valve replacements in patients with severely reduced cardiac output. Ann Thorac Surg 1971;11:517-22. 13. John S, Bashi VV, Jairaj PS, et al. Closed mitral valvotomy: early results and long term follow-up of 3724 consecutive patients. Circulation 1983;68:891-6. 14. Freed MD, Bernhard WF. Prosthetic valve replacement in children. Prog Cardiovasc Dis 1975;17:474-87. 15. Reed GE, Kloth HH, Kiely B, Danilowicz DA, Rader B,

The Journal of

638

16.

17.

18.

19.

20.

21.

22.

23.

24.

25. 26.

John et al.

Doyle EF. Long term results of mitral annuloplasty in children with rheumatic mitral regurgitation. Circulation 1974;59(Pt 2):11189-92. Stevenson JG, Kawabori I, Morgan BC, Dillard DH, Meredino KA, Guntheroth WG. Rheumatic mitral regurgitation. The case for annuloplasty in the pediatric age group. Circulation 1975;51(Pt 2):149-51. Robbins RC, Bowman FO Jr, Maim JR. Cardiac valve replacement: a 20 year series. Ann Thorac Surg 1988;45:5661. Klint R, Hernandez A, Weldon C, Harmann AF, Goldring D. Replacement of cardiac valves in children. J Pediatr 1972;80:980-7. Mathews RA, Neckes WH, Lenox CC, et al. Valve replacement in children and adolescents. J THORAC CARDIOVASC SURG 1977;73:872-6. Gardner TJ, Roland JMA, Neill CA, Donahoo JA. Valve replacement in children. A fifteen year perspective. J THORAC CARDIOVASC SURG 1982;83:178-85. Baugman KL, Kallmann CH, Yurchat PM, et al. Predictors of survival after tricuspid valve surgery. Am J Cardiol 1984;54:137-41. Breyer RH, McClenathan JH, Michaelis LL, Mcintosh CL, Morrow AG. Tricuspid regurgitation: a comparison of nonoperative management, tricuspid annuloplasty and tricuspid valve replacement. J THoRAc CARDIOVASC SURG 1976;72:867-74. King LM, Schaff AV, Danielson GK, et al. Surgery for tricuspid regurgitation late after mitral replacement. Circulation 1984;70(Pt 2):1193-7. John S, Munsi SC, Bhati BS, Gupta RP, Sukumar IP, Cherian G. Coexistent mitral valve disease with left to right shunt at the atrial level. Results of surgical treatment in 15 cases. Am Heart J 1987;114:1406-15. Dunn JM. Porcine valve durability in children. Ann Thorae Surg 1981;32:357-67. Edmunds LH Jr. Thrombotic and bleeding complications of prosthetic heart valves. Ann Thorac Surg 1987;44:43045.

Thoracic and Cardiovascular Surgery

27. Sethia B, Turner MA, Lewis S, et al. Fourteen years' experience with the Bjork-Shiley tilting disc prosthesis. J THORAC CARDIOVASC SURG 1986;91:350-61. 28. Czer LSC, Matloff J, Chaux A, et al. A six year experience with a St. Jude Medical valve: haemodynamic performance, surgical results, biocompatibility and follow up. J Am Coli Cardiol 1985;6:904-12. 29. Vidne B, Levy MJ. Heart valve replacement in children. Thorax 1970;25:57-61. 30. Bloodwell RD, Hallman GL, Cooley DA. Cardiac valve replacement in children. Surgery 1968;63:77-89. 31. Sade RM, Ballenger JF, Hohn AR, Arranta JE, Riopel DO, Taylor AB. Cardiac valve replacement in children. J THORAC CARDIOVASC SURG 1979;78:123-7. 32. Morrell MT, Dunnill MS. The postmortem incidence of pulmonary embolism in a hospital population. Br J Surg 1968;55:347-52. 33. Arom KV, Nicoloff OM, Kersum TE, et al. Six years of experience with the St. Jude medical valvular prosthesis. Circulation 1985;72(Pt 2):11153-8. 34. Bradley LM, Midgley FM, Watson DC, Getson PR, Scott IP III. Anticoagulation therapy in children in mechanical prosthetic cardiac valves. Am J Cardiol 1955;56:533-5. 35. Oakley C, Doherty P. Pregnancy in patients after valve replacements. Br Heart J 1976;38:1140-8. 36. Pavankumar P, Venugpopal P, Kaul U, et al. Pregnancy in patients with prosthetic cardiac valve:a I O-yearexperience. Scand J Cardiovasc Surg 1988;22:19-22. 37. Salazar E, Zajarias A, Gutierrez N, Iturbe I. The problem of cardiac valve prostheses, anticoagulants and pregnancy. Circulation 1984;70(Pt 2):1169-77. 38. Tepley JF, Grunkemeier GL, Sutherland HD, Lambert LE, Johnson VA, Starr A. The ultimate prognosis after valve replacement: an assessment at 20 years. Ann Thorac Surg 1981;32:111-9. 39. Weinstein GS, Mavroudis C, Ebert PA. Preliminary experience with aspirin for anticoagulation in children with prosthetic cardiac valves. Ann Thorac Surg 1981;33:54953.