Mitral valve replacement in the first year of life

J

THORAC CARDIOVASC SURG

1990;100:762-8

Mitral valve replacement in the first year of life From 1973 through 1987 25 patients underwent mitral valve replacement in the first year of life for mitral stenosis and mitral regurgitation. The patients with mitral stenosis included two with mitral arcade, two with supravalvular mitral stenosis with hypoplastic mitral valve, and one with parachute mitral valve. Included in the group of patients with mitral regurgitation were 12 with atrioventricular canal defect, six with chordal and leaflet defects, one with Marfan's syndrome, and one with bacterial endocarditis. Prostheses included 12 Bjork-Shiley (17 mm), seven St. Jude Medical (19 mm in four, 21 mm in three), five stent-mounted dura mater valves (12 mm to 16 mm), and one porcine xenograft (19 mm). In four patients the valves were placed in the left atrium in a supraannular location. There were nine operative (atrioventricular canal defect seven, mitral regurgitation two) and five late (atrioventricular canal defect four, mitral stenosis one) deaths, giving actuarial I- and 5-year survival rates of 52% and 43%, respectively. All 6 patients with tissue valves died; the four with supraannular mitral valve replacement survived. Since 1983 operative mortality has been reduced to 0% (70% confidence limits 0% to 24% ~ Nine patients required a second mitral valve replacement for prosthetic stenosis 5 to 69 (mean 30) months after the original mitral valve replacement (one operative death~ Because of improvements in repair of atrioventricular canal defect in infancy, the need for mitral valve replacement at atrioventricular canal defect repair has decreased. Although valvuloplasty has been advocated for repair of congenital mitral valve disease and is applicable in some infants with mitral regurgitation, mitral valve replacement is frequently unavoidable for congenital mitral disease and can now be accomplished at a low operative risk, even when the prosthesis has to be positioned supraannularly.

Keishi Kadoba, MD, Richard A. Jonas, MD, John E. Mayer, MD, and Aldo R. Castaneda, MD, Boston, Mass.

A

small percentage of patients with mitral valve stenosis or regurgitation become severely symptomatic during early infancy and require surgical intervention. 1• 2 Reconstructive procedures are obviously preferable and should be considered the procedure of choice; however, the complex pathology of the mitral apparatus in this age group in fact seldom lends itself to a reparative operation. Mitral valve replacement (MVR) is a well-established and safe procedure in adults and older children; however, it poses special clinical and technical difficulties in infants, including (1) the small size of the mitral anulus, 3· 6 left atrium, and left ventricle; (2) great variability in valve From the Department of Cardiac Surgery, Children's Hospital, and the Department of Surgery, Harvard Medical School, Boston, Mass. Received for publication Aug. 22, 1989. Accepted for publication Jan. 5, 1990. Address for reprints: Richard A. Jonas, MD, Department of Cardiac Surgery, The Children's Hospital, 300 Longwood Ave., Boston, MA 02115. 12/1/19306

762

anomalies 7• 8; (3) frequent association with other cardiovascular anomalies8• 9 ; (4) the problem of a fixed prosthetic size in a growing patient 10• 11 ; and ( 5) the difficulty in maintaining adequate anticoagulant therapy after operation. 12• 13 we therefore reviewed retrospectively our experience with MVR in infants (first year of life). Patients and methods The hospital charts of all infants who underwent MVR between 1973 and 1987 at the Children's Hospital, Boston, were reviewed. The status of all patients as of January 1989 was obtained by contacting either families of patients directly or the referring cardiologist or pediatrician. Statistical examination of the data was performed with the unpaired Student's t test, Fisher's exact test (two-tailed), or actuarial method for events in the postoperative course. Patient profile. From 1973 through 1987,25 infants underwent MVR. This represents 20.7% (25/121) of the children (under 19 years old) who underwent MVR during the same period at the Children's Hospital, Boston. Age of the patients at the initial MVR ranged from 2 days to 11 months (mean 7 months); body weight ranged from 3.0

Volume 100 Number 5 November 1990

to 7.8 (mean 5.5 ± 1.1) kg. Eleven were male and 14 were female. Diagnoses (Fig. I) included mitral regurgitation (MR) in 20 patients (including 12 with atrioventricular canal defect [A VC] and eight non-A VC patients) and congenital mitral stenosis (MS) in five. Seven patients with AVC had Down's syndrome; two with MS had the characteristics of Shone's syndrome. Situs solitus and atrioventricular and ventriculoarterial concordance were present in all patients; that is, the mitral valve was the systemic atrioventricular valve. Of 12 A VC patients, nine had complete A VC (Rastelli type A in six, type C in three) and three had incomplete A VC. Four patients required MVR as part of initial A VC repair because of lack of sufficient valve tissue in three and failed repair in one; in the remaining eight patients it was necessary to replace the mitral valve because of progressive MR 5 days to 2 months after the A VC operation. In these eight patients, operative findings at MVR included a perforation in the leaflet (possibly infectious, one patient), disruption of the leaflet near the suture line (one patient), distortion of the pericardium used to augment the deficient septal leaflet at A VC operation (one patient), and lack of sufficient valve tissue in the other five patients. The number of patients who required MVR during the perioperative period (within two months) of A VC operation were three of90 infants (3.3%) during the last 5 years compared with nine of 54 infants (16.7%) during the previous period (p < 0.01). Of the eight patients with non-A VC mitral regurgitation, four had similar valve abnormalities, which included dysplastic and retracted leaflets, thick and short chordae tendineae, obliterated interchordal spaces, and hypertrophied and calcified papillary muscles. Of the other four patients, one had a cleft mitral valve with an idiopathic left atrial aneurysm, one had a prolapsed mitral valve with myxogenous degeneration associated with Marfan's syndrome, one had a vegetative lesion of the mitral valve caused by acute streptococcal bacterial endocarditis, and the youngest patient (2 days old) had immature and deficient valve tissue with three hypertrophied papillary muscles. Of the five patients with mitral stenosis, two had a mitral arcade, two supramitral stenosing rings associated with severe hypoplasia of the mitral valve anulus, and one parachute mitral valve. In 13 patients other than those with A VC, eight had associated cardiovascular anomalies, including aortic or subaortic stenosis in five, coarctation of the aorta in three, ventricular septal defect (VSD) in two, and double aortic arch in one. Previous operations. Fourteen patients had undergone at least one operation before the MVR: A VC operation in eight (six complete, two partial), aortic valvotomy in two, coarctectomy in three, ligation of patent ductus arteriosus in two, and resection of the supramitral stenosing ring with closure of a VSD in one. Indication for mitral operation Clinical status. Uncontrollable congestive heart failure was the principal indication in all patients. Operations were performed on an emergency basis in 13 patients. Catheterization data. Most patients showed some degree of pulmonary hypertension, with a pulmonary-systemic mean pressure ratio of 0.70 ± 0.26 (0.26 to 1.05, n = 19). In the patients with MS, transmitral valve gradients measured between left atrial mean pressure (or pulmonary wedge mean

MVR in first year of life

IIIII

rn 12 1-

0

zw

763

Hospital Death Total Patients

i= 8

~

u.

0 4 0 z AVC

MR DIAGNOSIS

Fig. I. Diagnostic groups requiring M VR in infancy. A VC, Atrioventricular canal defects; MR, mitral regurgitation; MS, mitral stenosis.

pressure) and left ventricular end-diastolic pressure were 14.0 ± 4.3 mm Hg (9.0 to 19.0, n = 5). Left atrial mean pressure (or pulmonary wedge mean pressure) for the whole MR group was 18.6 ± 8.5 mm Hg (6.0 to 30.0, n = 16); for the patients who required emergent MVR, the left atrial mean pressure. was 23.5 ± 8.0 mm Hg (n = 8), whereas it was only 13.6 ± 5.6 mm Hg (n= 8) (p < 0.05) in the nonemergent group. MVR. A median sternotomy was used in all patients. Deep hypothermic circulatory arrest was used in 19 of the 25 patients. The mitral valve was exposed through the right atrium and the atrial septum in 14 patients (including 12 with AVC) and through the left atrium in the remaining 11 patients. In three patients with congenital MS, a prosthesis was sutured above the native mitral valve (supraannular MVR) because the mitral anulus was hypoplastic and did not accept the smallest available prosthetic valve. A supraannular MVR was also done for the patient with bacterial endocarditis and an annular abscess. In these patients the prosthetic valve was sutured to the left atrial wall approximately I em above the original anulus. Unhindered excursion of the prosthetic valve disc was checked meticulously, and, if necessary, the valve was rotated to achieve optimal position. In all other patients the prosthesis was placed in a conventional manner within the native anulus. Three patients had concomitant resection of subaortic stenosis and one had aortic valvotomy. Prosthetic valves implanted (Fig. 2) included 12 Bjork-Shiley (17 mm) (Shiley, Inc., Irvine, Calif.), seven St. Jude Medical (19 mm in four, 21 mm in three) (St. Jude Medical, Inc., St. Paul, Minn.), five stent-mounted dura mater (12 mm in two, 16 mm in three), and one porcine xenograft (19 mm). All patients with supraannular MVR received a 17 mm Bjork-Shiley valve.

Results Operative mortality. Nine patients died within 30 days (up to 9 days) of the operation (36%) (Fig. 3); seven had A VC, one Marfan's syndrome, and one congenital MR. Six operative deaths (46%) occurred in the 13 patients who underwent operation on an emergent basis, and five occurred in the six patients (83%) with tissue valves (four dura mater, one Hancock valve; Johnson &

The Journal of Thoracic and Cardiovascular Surgery

7 6 4 Kadoba et al.

Dura Mater

5

16mm

3

SJM

21 mm

7

3

Size

Prosthesis

Fig. 2. Valve types and sizes used for MVR. SJM, St. Jude Medical.

1111

Hospital Death [ ] Total Patients .... Mortality

12

u;

1Q.

u.

8

80 60

> !::: ..J

<

1-

a:

0

d z

100

40 0

::E

4

20 '73-'77

'78-'82

'83-'87

te

0

YEAR

Fig. 3. Hospital deaths after MVR according to year of operation.

Johnson Cardiovascular, King of Prussia, Pa.). The seven patients operated on since 1983 (including three A VC patients) all survived (0% mortality, 70% confidence limits 0% to 24%). Causes of hospital deaths. Causes of the hospital deaths were multifactorial (Table 1). Major factors that contributed to hospital death included left ventricular outflow obstruction in four patients, myocardial failure after cardiopulmonary bypass in two, Candida sepsis leading to multiorgan failure in one, low cardiac output associated with pulmonary hypertension in one, and unexplained sudden cardiac arrest in one. Postmortem examinations were performed in six of the nine patients. Of these six patients, four (all with A VC) had significant left ventricular outflow tract obstruction caused by the prosthesis, or by a combination of prosthesis and inherent anatomy. Three of those prostheses were dura mater valves and one a Bjork-Shiley valve. Postoperative course and complications. Three of the four patients with supraannular MVR did well postoperatively. The fourth patient had a preoperative pulmonary-systemic resistance ratio (Rp/Rs) of 0.77 that persisted postoperatively (Rp/Rs = 0.5) without a significant transprosthetic gradient (2 to 4 mm Hg). Pulmonary vascular resistance had decreased (Rp/Rs = 0.3) at

the second postoperative catheterization 15 days after MVR. In the operative survivors, the clinical improvement was generally rapid and impressive, except in two AVC patients with Down's syndrome. One fed poorly and required a gastrostomy; the other required diuretics, although postoperative catheterization revealed only a 1 mm Hg transprosthetic gradient. Permanent complete heart block occurred in four patients, all with A VC (one at initial A VC operation, two at MVR, and one at resection of the subaortic membrane on the same day of MVR); two of these patients died early, and the other two died late after operation. Infectious complications occurred in three patients (one case of Salmonella enteritis and sepsis, one case of Candida. ~epsis with meningeal involvement, and one wound infection with Staphylococcus aureus infection necessitating debridement. A left-to-right shunt (Qp/Qs = 1.8) persisted in the A VC patient with multiple muscular VSDs. A mild periprosthetic leak occurred in one patient. The other perioperative morbidities are listed in Table I. Anticoagulation. All surviving patients· except one who received a tissue valve were placed on a regimen of sodium warfarin (Coumadin), aiming to maintain the prothrombin time between 1.5 to 2.0 times control values. Bloody vomitus developed in one patient, with the prothrombin time exceeding the therapeutic limit 2 years after operation; this episode was controlled ~ithout incident. Another patient who lived overseas was found to have a prolonged prothrombin time (6%). The condition was treated with blood products including cryoprecipitate, and thrombosis of the prosthesis subsequently developed, successfully managed by valve replacement. Follow-up. Sixteen operative survivors were followed up a mean of 49 months (2 months to 115 months). Two patients were lost to follow-up at 24 months and 35 months. There were 5 late deaths, including four patients with A VC repair 2 months to 2 years after operation. One

Volume 100 Number 5 November 1990

MVR in first year of life

• Total Pts. (25) Survivors ( 16)

E

o Op.

765

100

_g~ 80 CD::!:

~a;

60

~:g_ .!!JCD 40 c.. a: ~

20

20

0 2

3

4

5

6

7

8

9

0

2 4 3 Years after Original MVR

10

Years after MVR

Fig. 4. Actuarial survival (including hospital deaths) after

5

6

Fig. 5. Actuarial freedom from repeat MVRs in 16 hospital survivors.

MVR.

patient died postoperatively at repeat MVR because of the sudden occurrence of unexplained electromechanical dissociation. On postmortem examination, the implanted I7 mm Bjork-Shiley valve was intact and no left ventricular outflow obstruction was noted, but two residual VSDs were found. One patient died of Clostridium sepsis 2 months after MVR. The cause of the other three late deaths was not specified, although two patients died in the course of an infectious illness. As of January I989, II patients were alive and doing well. Thus the actuarial survival rates for all patients at I and 5 years are 52% and 43%, respectively. Actuarial survival rates at 1 and 5 years for operative survivors of the original MVR are 8I% and 68%, respectively (Fig. 4 ). Repeat MVR. Among the 16 operative survivors, nine required repeat MVRs 5 to 69 (mean 30) months after the original MVR, giving an actuarial freedom from repeat MVR at 3 years of 45% (Fig. 5). Indications for repeat MVR were acute pulmonary edema in two patients. One of these (mentioned previously) had acute circulatory collapse resulting from a thrombosed prosthesis. The other patient, who was symptomatic for several days before repeat MVR, had extensive pannus ingrowth over the sewing ring of the prosthesis. In the other eight patients symptoms of congestive heart failure gradually developed, and at repeat MVR all had exuberant pannus over the left atrial aspect of the prosthesis. However, disc motion was not impaired. Bodyweightofth epatientsdouble dfrom5.5 ± 1.1 kg atthe original MVR to II.O ± 4.0 kg atthe time of repeat MVR. Two patients had replacement of a 17 mm BjorkShiley valve with another I7 mm Bjork-Shiley valve, whereas the patient with a I6 mm dura mater valve accepted a I7 mm Bjork-Shiley valve. In five patients a valve two sizes larger and in one a valve three sizes larger than the original prostheses could be inserted (Fig. 6). Three of the four patients with supraannular MVR

'E

.§

30

0

5

10 15 20 Age· (year) at Surgery

25

30

Fig. 6. Increment in valve size achieved at repeat MVR in 37 children, including 10 patients who were 1 year old or less (within dashed rectangle labeled infant group) at the time of initial MVR. Asterisks indicate ·operative deaths at repeat MVR.

required repeat MVR 5 months, II months, and 6 years after the initial MVR. In one patient the new valve was repositioned supraannularly (21 mm St. Jude Medical), while in the other two patients the anulus had widened sufficiently to accept a 17 mm Bjork-Shiley and a 21 mm St. Jude Medical valve, respectively. The fourth patient with a supraannular MVR was doing well I4 months after the initial MVR. An additional patient who had a I7 mm Bjork-Shiley valve placed twice in the annular position had a I9 mm St. Jude Medical valve placed supraannularly at the third operation. One patient who had an A VC operation and a 16 mm dura mater valve died at the second MVR because of Salmonella sepsis. In one other patient complete heart block developed at repeat MVR and a permanent pacemaker was implanted. Discussion Although there have been several reports of prosthetic valve replacement in children, 14- 18 except for some iso-

The Journal of Thoracic and Cardiovascular Surgery

7 6 6 Kadoba et a/.

Table I. Causes of hospital death Patient

Diagnosis

Prosthesis

Died on

Findings/events LVOTO, renal failure, hyperkalemia, moribund preoperatively Myocardial failure, died in OR Myocardial failure, died in OR

I*

IAVC

160M

PODO

2 3

IAVC MR, Marfan CAVC CAVC CAVC IAVC CAVC

17 BS 19 HC

PODO PODO

19SJM 120M 160M 17 BS 21 SJM

POD! POD2 POD 3 PODS POD6

160M

POD9

4 5* 6* 7* 8* 9*

MR, congenital

Sudden cardiac arrest, moribund preoperatively L VOTO, PH, bleeding from PA line site LVOTO, respiratory failure with pulmonary infiltrates LVOTO,CHB Low cardiac output, PH, subaortic resection on day of MVR,CHB Candida sepsis, multiorgan failure, moribund preoperatively, moderate L VOTO

lA YC and CA VC, Incomplete and complete A VC; POD; postoperative day; LVOTO, left ventricular outflow obstruction; OR, operating room; PH, documented pulmonary hypertension to systemic level; PA, pulmonary artery; CHB, complete heart block; DM, dura mater valve; BS, Bjork-Shiley; HC, Hancock; SJM, St. Jude Medical. *Postmortem done.

lated case reports little information is available specifically on MVR in infants. Although the overall mortality in this group of patients has been high, hospital mortality has decreased significantly since 1983. During the past 5 years, all seven infants who required MVR, including three A VC patients, survived (Fig. 3). Furthermore, the two operative deaths in patients without A VC involved the first two patients in this series; since then there were no operative deaths in eleven consecutive non-AVC patients. Clearly, MVR can be performed safely, particularly in non-AVC patients in this age group. There were five late deaths, including one operative death at repeat MVR. Two of the five late deaths occurred during the last 5-year period; both patients had Down's syndrome with A VC and a permanent pacemaker. Patients with AVC. Not surprisingly, the patients with AVC requiring primary or secondary MVR incurred the highest mortality among the anatomic subgroups, and long-term prognosis has been poor. In fact, only one of the original12 patients is a long-term survivor. The Toronto experience is similarly discouraging; seven of nine patients (78%) with A VC and MVR died. 14 The mortality in our patients with A VC was mostly attributable to ( 1) a seriously compromised preoperative state, including pulmonary edema and persistence of high left atrial pressures after A VC operation; (2) use of a high-profile tissue prosthesis (mostly dura mater valves), leading to left ventricular outflow obstruction postoperatively; and (3) postoperative complete heart block. More recent improved understanding of the varied anomalies of

complete A VC 19-22 and consequent refinements in reparative techniques, mostly of the left atrioventricular valve, 22-24 have significantly reduced the need for MVR at AVC operation. Interestingly, similar to the aortic valve, there seems to occur continuous postnatal maturation of the mitral valve apparatus in this lesion; therefore we try to avoid repair of A VC in neonates. Patients without AVC. Prognosis of non-A VC patients after MVR has been significantly better than that of patients with A VC. Clearly, reparative techniques for regurgitant mitral lesions are applicable to children 25 - 29 and should be favored over replacement. In our experience with infants, however, and in contrast to the report by Carpentier and colleagues, 28 we encountered many valves significantly lacking in valve tissue and also having other severe deformation of the subvalvular apparatus that rendered reconstruction impossible. Especially in congenital MS, a significant spectrum of anomalies with severe obstruction, including at the subvalvular level, was frequently encountered. Moreover, the valve anulus in these patients is often hypoplastic, making conventional placement of prosthetic valves impossible. To complicate matters further, four of the five patients (80%) also had obstructive lesions within the left ventricular outflow tract and aortic arch. Despite this additional risk, however, all five patients survived operation, and four had continued to do well to the time of writing. Supraannular MVR. There were six supraannular MVRs in this series: four at the first MVRs, one at the second, and one at the third. The most common indica-

Volume 100 Number 5 November 1990

tion for supraannular placement of a mitral prosthesis was significant hypoplasia of the anulus (three patients), involvement of the mitral apparatus by bacterial endocarditis (one patient), and also to replace a larger prosthesis for age and body weight, particularly after a previous conventional annular placement of a small-sized prosthesis. Schaffer and co-workers 30 reported on eight pediatric patients with supraannular MVR with St. Jude Medical prostheses. The smallest size of valve in their series was 19 mm. Seven of their eight patients died (2 early and five late). They therefore concluded that annular placement of a mitral valve prosthesis is critical to good outcome. Our experience is clearly different; all our patients survived the operation without postoperative difficulties, and most were doing well at follow-up. In this particular position, a valve that can be readjusted after implantation(i.e., Bjork-Shiley) seems preferable. All our patients had Bjork-Shiley valves, whereas Schaffer and co-workers 30 used St. Jude Medical prostheses. Whether this difference in valve design is critical under these conditions remains speculative. We believe that this technique provides an important option in an infant with a hypoplastic p1itral anulus, in the presence of an infected mitral valve apparatus, or in the presence of a narrow left ventricular outflow tract. 31 MVR in infants carries the obvious disadvantage of requiring repeat MVR; indeed, nine of the 11 long-term survivors so far required ten repeat MVRs. The greater prevalence ofreoperation (55% at 3 years after the initial MVR) compared with the previous report 10• 11 may simply reflect the younger age in our patients at the initial MVR. In one patient the prosthesis became thrombosed, but the other eight patients simply outgrew their prostheses. The largest prosthesis implanted at repeat MVR was a 23 mm St. Jude Medical valve. We anticipate that a patient who requires MVR in infancy will require at least two additional MVRs during a lifetime. Surgical alternatives. There are some additional surgical options for dealing with a severely hypoplastic mitral valve apparatus other than MVR. For example, a left atrial-left ventricular valved conduit can be placed to bypass a hypoplastic mitral valve. 3- 5 We have used this technique successfully in two patients. Obviously repeat operation will be necessary also in these patients. Another option is an initial Norwood type procedure, 32 followed later by a Fontan type operation; we have used this option in the presence of a hypoplastic left ventricle. Prosthesis size. The 17 mm Bjork-Shiley prosthesis has been withdrawn from the market related to new Food and Drug Administration (FDA) requirements. This

MVR in first year of life 7 6 7

particular prosthesis accounted for 48% (12/25) of the prostheses used in this series of patients. As mentioned previously, all four patients with supraannular MVRs received 17 mm Bjork-Shiley valves, not only because it was the smallest prosthesis available but also because an important feature of this valve allows rotation of the valve disc within its housing. This is particularly desirable in this location, where the excursion of the valve disc can be limited by atrial wall tissue. It will be important to encourage valve manufacturers and the FDA to allow continuing production of the 17 mm mitral prosthesis.

REFERENCES Duncan WJ, Bharadwaj B, Tyrrell MJ. Mitral valve replacement in the neonate: a report of two cases. Pediatr Cardiol 1984;5:307-12. 2. Pollock JC, Shawkat S, Houston A. Mitral valve replacement in the first three months of life. Br Heart J 1984; 1.

52:549-51.

3. Laks H, Hellenbrand WE, Kleinman C, Talner NS. Left atrial-left ventricular conduit for relief of congenital mitral stenosis in infancy. J THORAC CARDIOVASC SURG 1980; 80:782-7. 4. Midgley FM, Perry LW, Potter BM. Conduit bypass of

5.

6.

7. 8.

9. 10.

mitral valve: a palliative approach to congenital mitral stenosis. Am J Cardia! 1985;56:493-4. Como A, Giannico S, Leibovich S, Mazzera E, Marcelletti C. The hypoplastic mitral valve: When should a left atrialleft ventricular extracardiac valved conduit be used? J THORAC CARDIOVASC SURG 1986;91:848-51. Westerman GR, VanDevanter SH, Norton JB Jr, Readinger Rl. Congenital m'itral valve stenosis in infancy: a different approach to a difficult problem. J THORAC CARDIOVASC SURG 1987;94:305-7. Davachi F, Moller JH, Edwards JE. Diseases of the mitral valve in infancy-an anatomic analysis of 55 cases. Circulation 1971;43:565-79. Ruckman RN, Van Praagh R. Anatomic types of congenital mitral stenosis: report of 49 autopsy cases with consideration of diagnosis and surgical implications. Am J Cardial 1978;42:592-60 1. Collins-Nakai RL, Rosenthal A, Castaneda AR, Bernhard WF, Nadas AS. Congenital mitral stenosis: a review of 20 years' experience. Circulation 1977;56:1039-47. Friedman S, Edmunds LH Jr, Cuaso CC. Long-term mitral valve replacement in young children: influence of somatic growth on prosthetic valve adequacy. Circulation 1978;57:981-6.

11.

Nudelman I, Schachner A, Levy MJ. Repeated mitral valve replacement in the growing child with congenital mitral valve disease. J THORAC CARDIOVASC SURG 1980;

12.

Bradley LM, Midgley FM, Watson DC, Getson PR, Scott

79:765-9.

7 6 8 Kadoba et a/.

13.

14.

15.

16. 17.

18.

19.

20.

21.

22.

LP III. Anticoagulation in children with mechanical prosthetic cardiac valve. Am J Cardiol 1985;56:533-5. Sade RM, Ballenger JF, Hohn AR, Arrants JE, Riopel DA, Taylor AB. Cardiac valve replacement in children: comparison of tissue with mechanical prostheses. J THORAC CARDIOVASC SURG 1979;78: 123-7. Williams WG, Pollock JC, Geiss OM, Trusler GA, Fowler RS. Experience with aortic and mitral valve replacement in children. J THORAC CARDIOVASC SURG 1981;81:326-33. Gardner TJ, Roland JA, Neill CA, Donahoo JS. Valve replacement in children: a fifteen-year perspective. J THORAe CARDIOVASC SURG 1982;83:178-85. Elliot MJ, de Leva! M. Valve replacement in children. World J Surg 1985;9:568-78. Spevak PJ, Freed MD, Castaneda AR, Norwood WI, Pollack P. Valve replacement in children less than 5 years of age. JAm Coli Cardiol 1986;8:901-8. Brown JW, Cooper JL, Deshner WR, King H. Mitral valve replacement in children: a twenty-one year experience. In: Dunn JM, ed. Cardiac valve disease in children. Elsevier, 1988:236-53. Rastelli GC, Ongley PA, Kirklin JW, McGoon DC. Surgical repair of the complete form of persistent common atrioventricular canal. J THORAC CARDIOVASC SuRG 1968; 55:299. Goor D, Lillehei CW, Edwards JE. Further observations on the pathology of the atrioventricular canal malformation. Arch Surg 1962;97:954. David I, Castaneda AR, Van Praagh R. Potentially parachute mitral valve in common atrioventricular canal: pathological anatomy and surgical importance. J THORAC CARDIOVASC SURG 1982;84:178-86. Carpentier A. Surgical anatomy and management of the mitral component of atrioventricular canal defects. In: Anderson RB, Shinebourne EA, ed. London, Churchill Livingstone, 1978;477-90.

The Journal of Thoracic and Cardiovascular Surgery

23. Castaneda AR, Mayer JE Jr, Jonas RA. Repair of complete atrioventricular canal in infancy. World J Surg 1985;9:590-7. 24. Katz NM, Blackstone EH, Kirklin JW, Bradley EL, Lemons JE. Suture techniques for atrioventricular valves: experimental study. J THORAC CARDIOV ASC SURG 1981 ;81 :52836. 25. McGoon DC. Repair of mitral insufficiency due to ruptured chordae tendineae. J THORAC CARDIOV ASC SURG 1960; 39:357-62. 26. Kay JH, Maselli Campagna G, Tsuji HK. Surgical treatment of tricuspid insufficiency. Ann Surg 1965;162:53-8. 27. Reed GE, Tice DA, Clauss RH. Asymmetric exaggerated mitral annuloplasty: repair of mitral insufficiency with hemodynamic predictability. J THORAC CARDIOVASC SURG 1965;49:752-61. 28. Carpentier A, Branchini B, Cour JC, et al. Congenital malformations of the mitral valve: pathology and surgical treatment. J THORAC CARDIOVASC SURG 1976;72:854-66. 29. Carpentier A, Reiland J, Deloche A, et al. Conservative management of the prolapsed mitral valve. Ann Thorac Surg 1978;26:294-302. 30. Schaffer MS, Clarke DR, Campbell ON, Madigan CK, Wiggins JW, Wolfe RR. The St. Jude Medical cardiac valve in infants and children: role of anticoagulant therapy. JAm Coli Cardiol 1987;9:235-9. 31. McGrath LB, Kirklin JW, Soto B, Bargeron LM Jr. Secondary left atrioventricular valve replacement in atrioventricular septal (A V canal) defect: a method to avoid left ventricular outflow tract obstruction. J THORAC CARDIOVASC SURG 1985;89:632-5. . 32. Norwood WI, Lang P, Castaneda AR, Campbell DN. Experience with operation for hypoplastic left heart syndrome. J THORAC CARDIOVASC SURG 1981 ;82:511-79.