Biventricular repair of transposition of the great arteries and unbalanced ventricles.

BACKGROUND: It is well established that the arterial switch operation is the surgical procedure of choice in patients with transposition of the great arteries and balanced ventricular anatomy. The surgical approach of choice in patients with transposition but unbalanced ventricular size is unknown. OBJECTIVES: Since the beginning of the arterial switch operation program, patients with transposition of the great arteries and unbalanced ventricles underwent biventricular repair by means of the arterial switch operation and repair of any associated lesions, either through a single or staged surgical procedure. The aim of this retrospective study is to analyze whether this approach can be proposed to such patients. METHODS: Forty-four patients with transposition of the great arteries and unbalanced ventricles underwent this surgical approach since 1984. Two groups were defined: group I had transposition with a dominant right ventricle (n = 28), and group II had transposition with a dominant left ventricle (n = 16). In group I the median age and weight at the arterial switch operation were 8.5 days (range, 5-70 days) and 3.1 kg (range, 1.5-3.7 kg), respectively. The median end-diastolic left ventricular volume, mass, and long-axis ratio were 15 mL/m2 (range, 11-16 mL/m2), 31.5 g/m2 (range, 20-66 g/m2), and 0.85 (range, 0.9-0.7), respectively. The mitral valve diameter was slightly hypoplastic, with a median z value of -1.22 (range, -0.3 to 3.7). In group 2 the median age and weight at the arterial switch operation were 42 days (range, 8 days-15 years) and 3.5 kg (range, 2.8-35 kg), respectively. Associated lesions in this group were coarctation in 9 and single (n = 12) or multiple (n = 4) ventricular septal defects. The median long-axis ratio and tricuspid z value were 0.6 (range, 0.3-0.8) and -0.9 (range, -0.5 to 3.3), respectively. In this group 9 patients had a single-stage procedure with fenestrated ventricular defect patches, atrial septal defect patches, or both; 7 patients underwent the staged approach. RESULTS: In group I there was 1 early death from sepsis after weaning from postoperative extracorporeal membrane oxygenation. Three patients had severe pulmonary hypertension, one of whom died 1 year later. All survivors demonstrated, at discharge from the hospital, equilibrated ventricular size, with a median left ventricular end-diastolic volume of 25 mL/m2 (range, 21-30 mL/m2). In group II there were 2 early and 1 late deaths. All early deaths occurred in patients without voluntary residual intracardiac shunts. Median early postoperative long-axis ratio and tricuspid z value were 0.8 (range, 0.7-1) and -0.2 (range, 0.74 to 1.2), respectively. CONCLUSION: This study demonstrates that the arterial switch operation in patients with transposition of the great arteries and unbalanced ventricles remains a good surgical option.

Surgical management and indication of left ventricular retraining in arterial switch for transposition of the great arteries with intact ventricular septum.

OBJECTIVE: Arterial switch is the operation of reference for the surgical treatment of transposition of the great arteries. In cases of late referral, perinatal complications or early left ventricular (LV) dysfunction, the one stage arterial switch is contra indicated. Anatomical repair remains possible in these patients following a LV retraining. METHODS: From January 1992 to January 2000, a LV retraining was attempted in 22 patients with transposition of the great arteries with intact ventricular septum (TGA IVS), whereas 470 direct arterial switch and 2 Senning were performed. Indication for LV retraining was based on a combination of factors including: an age older than 3 weeks, a "banana shape" aspect of the inter-ventricular septum and mainly a LV mass <35G/m(2). RESULTS: The mean age at LV retraining was 3.2 months ranging from 9 days to 8 months. Usually conducted by sterntomy, it associated a loose PA banding with a LV/RV at 65% with a systemico-pulmonary shunt. The first stage was associated with frequent LV dysfunction and the LV retraining was discontinued in two patients in favor of one Senning and one early switch followed by ECMO. One patient died at first stage from a mediastinitis. Nineteen patients underwent a second stage arterial switch that was performed when the LV mass had reached 50 G/m(2) after a mean delay of 10 days, ranging from 5 days to 6 weeks. After a mean follow up of 25 months, there was one non-cardiac late death. The 17 patients followed and leaving with an arterial switch are in NYHA class I, with a mean LV shortening fraction of 39%. CONCLUSIONS: Arterial switch following LV retraining in TGA IVS is a satisfactory option. The inferior limit of 35 G/m(2) adopted, to indicate LV retraining, seems a safe landmark. The quality of the myocardium generated and the respective roles played by the LV afterload, LV wall shear stress, LV inflow and outflow to induce the LV remodeling remain under debate.

Surgically created double-orifice left atrioventricular valve: a valve-sparing repair in selected atrioventricular septal defects.

OBJECTIVES: Some features of the left atrioventricular valve (large mural leaflet, dystrophic tissue) represent a challenge for repair of atrioventricular septal defects without postoperative regurgitation. A retrospective study was conducted to evaluate the results of surgically creating a double-orifice left atrioventricular valve in such circumstances. Clinical results were analyzed according to valvular and subvalvular left atrioventricular valve measurements in pathologic specimens with atrioventricular septal defects. METHODS: Among 157 patients operated on for atrioventricular septal defect since October 1989, 10 patients underwent primary repair (n = 8) or reoperation (n = 2) by this procedure. Median age at repair was 3.3 years (0.1-33 years). Anatomic types were complete (n = 3), intermediate (n = 5), and partial (n = 2). Preoperative moderate to severe left atrioventricular valve regurgitation was present in 6 patients. After the repair (two-patch technique in complete atrioventricular septal defect, cleft closed in each case), these 10 patients were found to have moderate to severe residual regurgitation not amenable to repair by annuloplasty. The top edge of the mural leaflet was anchored to the facing free edge of the cleft. RESULTS: No hospital death or morbidity was observed. Left atrioventricular valve regurgitation was absent or trivial (8 patients) and mild (2 patients). Color-coded echocardiography did not show significant left atrioventricular valve stenosis. The mean diastolic pressure gradient across the left atrioventricular valve was 3.2 +/- 1.1 mm Hg (1.4-4.5 mm Hg). At a median follow-up of 72 months (6-91 months), there was 1 late death, unrelated to left atrioventricular valve malfunction, due to pulmonary vascular obstructive disease. Left atrioventricular valve regurgitation did not increase over time, except in 1 patient in whom regurgitation recently progressed from mild to moderate. At rest, the mean diastolic pressure gradient across the left atrioventricular valve was 3.8 +/- 2.9 mm Hg (1.5-11.2 mm Hg). One child had an early moderate stenosis without pulmonary hypertension. Studies on pathologic specimens (n = 34) indicated that long chordal lengths and large mural leaflet size are essential independent anatomic features to assess its feasibility. CONCLUSIONS: Surgical creation of a double-orifice left atrioventricular valve is an effective additional procedure for repair of atypical cases of atrioventricular septal defect. The operation may decrease the need for reoperation or left atrioventricular valve replacement.

Congenital mitral stenosis with or without associated defects: An evolving surgical strategy.

BACKGROUND: Congenital mitral stenosis (CMS) remains a surgical challenge, particularly when it is associated with other heart defects. As in other groups of heart defects, there is a trend toward early single-stage complete repair, but the optimal surgical approach remains unanswered. METHODS AND RESULTS: This study was designed to analyze the evolution of surgical strategies in patients with CMS and associated defects through single-stage and staged repair. Between 1980 and 1999, 72 children were operated on for congenital heart defects, including CMS. Preoperative transmitral gradient was 12.6+/-7 mm Hg. Preoperatively, all the patients were NYHA class III to IV. Thirteen patients had an isolated CMS; in 59, it was associated with other heart defects, mainly ventricular septal defect (n=28) or multilevel left ventricular obstruction (n=41). In this group of patients, 33 had a staged approach, and 26 had a single-stage approach. Early mortality was 12.5% (9 patients). There were no deaths in the isolated CMS and single-stage repair groups. Logistic regression revealed that early mortality was influenced by association with left ventricular outflow tract obstruction (P:<0.001) and by use of a staged approach (P:<0.01). There was no late mortality in isolated CMS; there were 2 late deaths in the group of single-stage repair and 6 late deaths in the staged approach group (P:<0.01). Reoperation was required in 24 patients, mainly for residual mitral valve dysfunction or residual left ventricular outflow tract obstruction. Including the reoperations, 10 patients received a prosthetic mitral valve. At 15 years after surgery, survival was 69.6+/-7.5%, freedom from reoperation was 70.8+/-6.3%, and freedom from mitral valve replacement was 69+/-6%. CONCLUSIONS: Surgery for isolated CMS gives excellent early and long-term results. In patients with associated heart defects, a single-stage operation seems superior to a staged approach. Mitral valve replacement in this category of patients should be reserved as a salvage procedure.

Surgical management of progressive pulmonary venous obstruction after repair of total anomalous pulmonary venous connection.

BACKGROUND: The occurrence of a progressive pulmonary venous obstruction after the repair of the total anomalous pulmonary venous connection is a severe complication. OBJECTIVES: The objectives of this study were to retrospectively review the patients with this condition and to report our experience with a new surgical technique with a sutureless in situ pericardium repair. METHODS: Of 178 patients who underwent correction of total anomalous pulmonary venous connection, 16 patients (9%) experienced the development of a progressive pulmonary venous obstruction in a median interval of 4 months (5 weeks-12 years). Three patients had isolated anastomotic stenosis, 4 patients had isolated pulmonary venous ostial stenosis, and 9 patients had both. Pulmonary venous obstruction was bilateral in 7 patients. The surgical procedures used at reoperation included 8 patch enlargements, 5 ostial endarterectomies, 1 intraoperative stenting, and 7 sutureless in situ pericardium repairs. RESULTS: There were 4 deaths after reoperation (4 of 15 patients; 27%). The only significant mortality risk factor was the bilateral location of the pulmonary venous obstruction (P =.045). In patients with isolated anastomotic stenosis or with only 1 pulmonary venous ostial stenosis (n = 5), there was no death, except the patient presenting with a single ventricle. In patients with 2 or more pulmonary venous ostial stenoses (n = 10), there were 3 deaths; 5 of the 7 survivors were successfully treated with the in situ pericardial technique, with normalized pulmonary artery pressure at a mean follow-up of 26 months. CONCLUSION: Progressive pulmonary venous stenosis after repair of total anomalous pulmonary venous connection remains a severe complication when bilateral. The sutureless in situ pericardial repair offers a satisfactory solution, particularly on the right side.

Biventricular repair for double-outlet right ventricle. Results and long-term follow-up.

BACKGROUND: The purpose of the present study was to define the optimal management and to identify the risk factors for death and repeat operation in patients with double-outlet right ventricle. METHODS AND RESULTS: From 1985 through 1996, 154 consecutive patients underwent biventricular repair for double-outlet right ventricle. The presence of bilateral infundibular structures was the major inclusion criteria (142 patients). According to the relationship of the ventricular septal defect (VSD) to the great arteries, there were 86 patients with a subaortic VSD (56%), 45 patients with a subpulmonary VSD (29%), 18 patients with a noncommitted VSD (12%), and 5 patients with a doubly committed VSD (3%). Sixty-five patients (42%) had undergone previous palliative procedures. At repair, the median age was 10 months, and the median weight was 6.5 kg. Two main types of repair were used: intraventricular baffle repair (n = 115) and arterial switch operation with VSD-to-pulmonary artery baffle (n = 39). There were 14 hospital deaths (9%; 70% confidence limit [CL], 7% to 12%). The only significant risk factor for early death was the presence of congenital mitral valve anomalies (P = 0.02). Twenty-eight patients (18%) required 39 repeat operations. The repeat operation rate was higher in patients with associated VSD enlargement at baffle construction (n = 29; 19%) (P = 0.01). There were 6 late deaths (4%; 70% CL, 2% to 7%). Patients presenting with pulmonary stenosis constituted a low-risk group for global death (P = 0.008). The median follow-up was 52 months. Ten-year actuarial survival and survival with freedom from repeat operation rates were 86% and 62% (70% CL, 83% to 89% and 54% to 70%), respectively. CONCLUSIONS: Long-term survival with good quality of life can be achieved after either 1- or 2-stage repair of this complex anomaly.

Early liver transplantation is crucial in children with liver disease and pulmonary artery hypertension.

BACKGROUND/AIMS: Early liver transplantation is crucial in children with liver disease and pulmonary artery hypertension. Some severe pulmonary vascular anomalies associated with portal hypertension disappear after isolated liver transplantation. Evolution of pulmonary artery hypertension due to plexogenic arteriopathy is controversial, as this association is still considered a contraindication to isolated liver transplantation. Outcome of pulmonary hypertension after isolated liver transplantation is reported in three patients with portal hypertension. METHODS: After echocardiographic diagnosis, the patients had a complete hemodynamic exploration, and two had a lung biopsy. After liver transplantation, the survivors had echocardiographic follow up and a second hemodynamic exploration. RESULTS: In two children, pulmonary pressures and resistances returned to near-normal values 1 and 6 years after successful isolated liver transplantation. The third patient, with the most severe arteriopathy, had to wait 1 year for a donor, and the attempted transplantation was complicated by ventricular tachycardia; death occurred 2 days after surgery. CONCLUSIONS: Liver transplantation can reverse pulmonary artery hypertension due to high pulmonary resistances complicating liver disease with portal hypertension, provided it is carried out at an early stage. Early detection of pulmonary hypertension by systematic echocardiography may thus be crucial in these children with portal hypertension.

[Anatomical repair in transposition of great vessels]. / La réparation anatomique de la transposition des gros vaisseaux.

Transposition of the great arteries (TGA) covers a wide range of anatomic variants involving not only the origin of the great arteries, but the intracardiac structures and the aortic arch. TGA can now benefit from complete anatomic repair, namely switch operation, performed in simple forms during the neonatal period. Accurate pre and intraoperative assessment must take into account the main anatomic features: relationship between the aorta and the pulmonary artery and coronary arteries distribution. Over the last sixteen years, 1,063 patients with TGA underwent anatomic repair in our institution. Among 728 patients with simple TGA, 92% were operated on within the first 2 weeks of life. Among 335 patients with complex TGA, all had a large VSD and 154 had additional anomalies. Among these 335 patients, 264 had one-stage complete repair and 71 had two-stage repair, at the beginning of our experience. In complex forms the main associated anomalies involved the ventricles (right ventricle hypoplasia (11), malaligned VSD (90), restrictive VSD (3), multiple VSD (43), the atrioventricular valves (straddling of the mitral or tricuspid valve (15), mitral valve anomalies (cleft or tissue tag, 4) and the aortic arch (coarctation [88], interrupted aortic arch (8)). The hospital mortality was 6.9% for simple TGA, 14.6% for complex TGA. Main causes of early death were failure in coronary artery relocation and pulmonary hypertensive crisis. Mean follow-up (95.5% of patients) was 49 months. Overall 16-year survival was 87.9%, 91.5% for simple forms and 80.4% for complex forms. Anatomic repair is nowadays the treatment of choice for TGA. Long-term results will depend on the evolution of the relocated coronary arteries and of the neo-aortic valve.

Biventricular repair of conotruncal anomalies associated with aortic arch obstruction: 103 patients.

BACKGROUND: Biventricular repair of conotruncal anomalies associated with aortic arch obstruction is a complex surgical procedure that combines a cardiac repair and a aortic arch reconstruction. METHODS AND RESULTS: From January 1984 to April 1996, such a repair was performed in 103 patients. The conotruncal anomalies included: 15 transpositions of the great arteries (TGAs) with intact ventricular septum, 44 TGAs with ventricular septal defect, 32 double outlet right ventricle with subpulmonary ventricular septal defect, 10 truncus arteriosus, one double outlet left ventricle, and one tetralogy of Fallot. The arch obstruction included 88 coarctation and 15 interrupted aortic arch. One-stage repair has been the favored technique since 1990 and was performed in 58 neonates, including 38 TGAs or double outlet right ventricle and ventricular septal defect, 10 TGAs with intact ventricular septum, and all of the 10 truncus arteriosus. The cardiac repair included 89 arterial switch operations, 2 Kawashima rerouting, 10 truncus arteriosus repairs, and one double-outlet left ventricle repair and one tetralogy of Fallot repair. The aortic arch was reconstructed by direct anastomosis in 85 patients, with a Gore-Tex conduit in three patients and more recently by an ascending aortic patch augmentation in 15 patients. The hospital mortality was 12% (7 of 58) for the one-stage repair and 20% (9 of 45) for the two-stage repair. There were six late deaths. Reoperations or angioplasties were mandatory for 12 right ventricle outflow tract obstructions after arterial switch, involving 10 patients with double outlet right ventricle (P=.02), 10 recurrent arch obstruction, and six miscellaneous lesions. CONCLUSIONS: One-stage biventricular repair of conotruncal anomalies associated with aortic arch obstruction can be achieved in selected patients with an 83% survival rate at 7 years.

Surgical management of atrio ventricular septal defects with normal caryotype.

OBJECTIVE: Atrio ventricular septal defects (AVSD) with normal caryotype represent in average 25% of AVSD. They constitute a more complex group of patients characterized by frequent left sided heart obstructive lesions, raising the problem of the appropriate indications between biventricular and univentricular procedures. METHODS: Sixty-nine consecutive patients, who had AVSD with normal caryotype underwent surgery. According to the anatomical complexity there were 22 intermediate AVSD, 36 complete AVSD and 11 complex AVSD. Associated lesions were present in 68% of the patients including left sided heart anomalies in 57%. According to the size of the left ventricle (LV) evaluated on the LV/RV end diastolic diameter ratio calculated at 2D echocardiogram: right ventricular (RV) dominance was found in 29%; with border line LV in 13 patients and truly hypoplastic LV in 7 patients. Biventricular repair was always favored in case of border line LV and precluded when the LV/RV ratio was less than 0.33. RESULTS: There were 57 biventricular repairs with 10 years actuarial survival of 70 +/- 8% and respectively, 100% in the complex AVSD, 86% in the intermediate AVSD and 51% in the complete AVSD (P = 0.006). The risk factors for biventricular repair were the association to a subaortic stenosis (P = 0.01) and the severity of the mitral valve lesions (P = 0.03) that led to 38% reoperation. There were 12 univentricular palliation with 10 years survival of 66 +/- 14%. The risk factor for univentricular palliation was the association to a severe pre-operative mitral regurgitation (P = 0.005). CONCLUSIONS: Biventricular repair should be precluded in patients presenting with subaortic stenosis. Severe mitral valve anomalies lead to elevated mortality and morbidity with frequent reoperations. Univentricular repair might have larger indications and cardiac transplantation might be considered in patients with truly hypoplastic LV presenting with severe pre-operative AV valve regurgitation.

Aortico-left ventricular tunnel in fetuses and infants.

BACKGROUND: Aortico-left ventricular tunnel is a rare congenital abnormal communication between the aorta and the left ventricle presenting in early childhood as aortic regurgitation and cardiac failure. This condition has rarely been reported in fetuses. Operation is the only treatment, and postoperative aortic incompetence could be related to the age or the type of repair. METHODS: We conducted a retrospective, two-institution review, from 1983 to 1995, of aortico-left ventricular tunnel diagnosed in utero and before 6 months of age. RESULTS: Three cases of aortico-left ventricular tunnel were diagnosed in utero by Doppler echocardiography between 22 and 24 weeks' gestation. Prenatal aortico-left ventricular tunnel was associated with severe left ventricular dysfunction, aortic valve anomalies, and fetal hydrops. One death occurred in utero and one immediately after birth, and in 1 case pregnancy was interrupted. In these 3 cases the diagnosis was confirmed by autopsy. Three neonates and 2 infants had the diagnosis of aortico-left ventricular tunnel made after birth and underwent successful surgical repair. At short and midterm follow-up all patients are alive and aortic valve regurgitation is absent or trivial. CONCLUSIONS: This series shows that aortico-left ventricular tunnel covers an anatomic spectrum of lesions. Cases diagnosed in utero by Doppler echocardiography are characterized by severe ventricular dysfunction, associated aortic valve lesions, and poor outcome. Postnatal cases represent the more favorable end of the spectrum, with no associated lesions, and can be repaired without mortality and with good functional results.

Transposition of the great arteries complicated by tricuspid valve incompetence.

BACKGROUND: Tricuspid valve insufficiency secondary to structural anomalies of the valve itself or to an iatrogenic complication of the Rashkind procedure is very rarely associated with transposition of the the great arteries. This condition represents an interesting perioperative challenge. Rapid restoration of the tricuspid valve to a low-pressure system by arterial switch operation associated with tricuspid repair should theoretically improve the outcome in terms of myocardial and valve function. METHODS: Thirteen of 839 patients who underwent an arterial switch operation for various forms of transposition of the great arteries presented with moderate to severe tricuspid insufficiency. Three of them had a ventricular septum defect. Nine experienced severe cardiac failure with profound hypoxemia. Ventilatory support was necessary in 7, 6 had renal or hepatic dysfunction, and 5 had coagulation disorders. Inotropic support was started preoperatively in 8 patients. RESULTS: Tricuspid lesions were as follows: primary annular dilatation and lack of coaptation at the commissural level (n = 1), straddling tricuspid valve (n = 1) redundant tricuspid valve tissue leading to left ventricular outflow tract obstruction (n = 1), small cleft of the septal leaflet (n = 1), and dysplastic valve tissue with juxtacommissural regurgitation (n = 1). In 8 patients, the cause of the tricuspid valve insufficiency was most probably an iatrogenic lesion, with rupture of the papillary muscle (n = 2), rupture of the chordae (n = 1), or tear of the anterior leaflet (n = 5), whereas no clear cause could be found in 1 patient. Repair consisted of the arterial switch operation associated with tricuspid valve repair in 10 patients. In 2 patients with only discrete anomaly and in 1 without a clear cause of tricuspid regurgitation, no valve repair was performed. Three patients had their ventricular septal defect closed. There were only one early and one late death, both not related to the tricuspid lesions. Late postoperative (mean, 6.5 years) evaluation revealed normal left ventricular function in 10, with no tricuspid incompetence in 7 and trivial tricuspid insufficiency in 3. CONCLUSIONS: Restoration of an incompetent tricuspid valve in a low-pressure system by the arterial switch operation combined with valve repair provides good ventricular and valvar results. Preoperative management and appropriate timing of operation seem to be of utmost importance.

Surgical approaches for double-outlet right ventricle or transposition of the great arteries associated with straddling atrioventricular valves.

The surgical management of patients with double-outlet right ventricle or transposition of the great arteries and straddling atrioventricular valves remains a subject of controversy. Biventricular repair has theoretic advantages because it establishes normal anatomy and physiology. In some instances, however, it seems to carry too high operative risk, and a univentricular heart repair is preferred. Since 1984, we have operated on 34 patients with double-outlet right ventricle (n = 15) or transposition of the great arteries (n = 19) with isolated straddling tricuspid valve (n = 17), isolated straddling mitral valve (n = 9), both mitral and tricuspid straddling (n = 2), or abnormal insertion of tricuspid (n = 7) or mitral (n = 2) chordae in the left ventricular outlet, precluding an adequate tunnel construction. Straddling was categorized according to the location of the papillary muscle insertion in the opposite ventricular chamber: type A, on the edge of the ventricular septal defect (n = 14); type B, on the opposite side of the ventricular septum away from the edge of the defect (n = 8); type C, on the free wall of the opposite ventricular chamber (n = 8). Abnormal chordal insertions were classified according to the location of their attachments around the edges of the defect. Three types of chordal distribution were identified: on the aortic conus, on the pulmonary conus crossing the ventricular septal defect, or around the defect closing it like a curtain. All but three patients had two ventricles of adequate size. Sixteen patients underwent palliation. Median age at the definitive operation was 6.5 months (range 1 to 130 months). Thirty patients underwent a biventricular repair and four had a univentricular repair. Biventricular repair was achieved by an arterial switch operation in 18 patients and by tunnel construction from the left ventricle to the aorta in 12. In isolated straddling of types A and B, the ventricular septal defect was closed by adjusting the septal patch on the ventricular side above the straddled papillary muscle. In type C, the patch was sewn over the papillary muscle by applying it on the septum. In double straddling, the ventricular septum was incised between the two papillary muscles, and an ellipsoid patch was used to reconstruct the septal defect, directing each subvalvular apparatus into its own ventricular chamber. When the abnormal chordae in the left outflow tract inserted on the aortic or pulmonary conus, the conus was incised and tailored to make a flap, leaving an unobstructed left ventricular outflow tract. In two patients the subvalvular apparatus was resected and reattached to the patch. Curtainlike chordae were a contraindication to biventricular repair in double-outlet right ventricle but not in transposition. There were four early deaths and one late death, all occurring in the group having biventricular repair. Death was due to myocardial ischemia (n = 1), right ventricular hypoplasia (n = 1), pulmonary hypertension (n = 1), and residual subaortic stenosis (n = 1). Two patients had moderate to severe postoperative atrioventricular valve incompetence, caused by a cleft in the mitral valve in one patient. Three patients were reoperated on for subaortic stenosis (n = 1), pulmonary stenosis (n = 1), and mitral regurgitation (n = 1). Mean follow-up of 30.7 +/- 19.4 months was achieved in the survivors. All but one patient (univentricular repair) were in New York Heart Association class I, without atrioventricular valve incompetence. Actuarial survival at 4 years was 85.3% +/- 3%. We conclude that straddling or abnormal distribution of chordae tendineae of the atrioventricular valves does not preclude biventricular repair in double-outlet right ventricle or transposition of the great arteries provided that the ventricles are of adequate size. Curtainlike abnormal tricuspid chordae remain a contraindication to biventricular repair in double-outlet right ventricle.

Surgery for congenital mitral valve disease in the first year of life.

Between 1980 and 1993, 20 patients less than 1 year of age underwent operations for congenital mitral valve disease. Ten patients had congenital mitral incompetence and 10 had congenital mitral stenosis. Mean age was 6.6 +/- 3.4 months and mean weight was 5.6 +/- 1.5 kg. Atrioventricular canal defects, univentricular heart, class III/IV hypoplastic left heart syndrome, discordant atrioventricular and ventriculoarterial connections, and acquired mitral valve disease were excluded. Indications for operation were intractable heart failure or severe pulmonary hypertension, or both. Associated lesions, present in 90% of the patients, had been corrected by a previous operation in seven. In congenital mitral incompetence there was normal leaflet motion (n = 3), leaflet prolapse (n = 2), and restricted leaflet motion (n = 5). In congenital mitral stenosis anatomic abnormalities were parachute mitral valve (n = 4), typical mitral stenosis (n = 3), hammock mitral valve (n = 2), and supramitral ring (n = 1). Mitral valve repair was initially performed in 19 patients and valve replacement in one with hammock valve. Concurrent repair of associated lesions was performed in 12 patients. The operative mortality rate was zero. There were six early reoperations in five patients for mitral valve replacement (n = 4), a second repair (n = 1), and prosthetic valve thrombectomy (n = 1). One late death occurred 9 months after valve replacement. Late reoperations for mitral valve replacement (n = 2), aortic valve replacement (n = 1), mitral valve repair (n = 2), subaortic stenosis resection (n = 1), and second mitral valve replacement (n = 1) were performed in five patients. Actuarial freedom from reoperation is 58.0% +/- 11.3% (70% confidence limits 46.9% to 68.9%) at 7 years. After a mean follow-up of 67.6 +/- 42.8 months, 94% of living patients are in New York Heart Association class I. Doppler echocardiographic studies among the 13 patients with a native mitral valve show mitral incompetence of greater than moderate degree in one patient and no significant residual mitral stenosis. Overall, six patients have mitral prosthetic valves with a mean transprosthetic gradient of 6.2 +/- 3.7 mm Hg. These results show that surgical treatment for congenital mitral valve disease in the first year of life can be performed with low mortality. Valve repair is a realistic goal in about 70% of patients and possibly more with increased experience. Reoperation rate is still high and is related to complexity of mitral lesions and associated anomalies, but late functional results are encouraging.

Subaortic obstruction in double outlet right ventricles. Surgical considerations for anatomic repair.

BACKGROUND: Subaortic obstruction is one of the risk factors for anatomic repair of double outlet right ventricles (DORV). A comprehensive approach to such lesions has been developed in our institution since 1981. This retrospective work analyzes the results of this approach. METHODS AND RESULTS: Between January 1981 and September 1992, 30 patients aged 15 days to 15 years (mean, 44.8 months) underwent repair of a DORV associated with subaortic obstruction. Eighteen patients had a palliative procedure before complete repair. The ventricular septal defect (VSD) was subaortic in 15 patients, doubly committed in 1, noncommitted in 9, and subpulmonary in 5. The subaortic obstruction was a result of restrictive VSD in 29 patients and of double straddling of mitral and tricuspid valves once. The preoperative peak systolic pressure gradient between the left ventricle and the aorta (LV-Ao) was 68.7 +/- 23 mm Hg. Reconstruction of the left ventricular outflow tract comprised a ventral enlargement of the VSD in subaortic, doubly committed, and those subpulmonary VSDs scheduled for an arterial switch operation or a conal resection in noncommitted and other subpulmonary forms. Reconstruction of the right ventricular outflow tract included primary closure of the right ventricle in 12 patients, an infundibular patch in 9, a transannular patch in 4, and insertion of a right ventricular pulmonary valved conduit in 5. There were two early (6.6%) and two late (7.1%) deaths. Three patients required reoperation. A mean follow-up of 60.5 +/- 46.8 months was achieved in all the survivors. They were all in New York Heart Association class I or II, in sinus rhythm. At last follow-up, the mean LV-Ao gradient was 7.5 +/- 6.2 mm Hg, and LV function indices were within normal ranges. Actuarial survival and freedom from reoperation rates at 8 years were 86.6% and 87.0%, respectively. CONCLUSIONS: Surgical relief of subaortic obstruction in DORV has to be adapted to VSD location and spatial arrangement of atrioventricular valves and great vessels.

[Percutaneous mitral valvulotomy: comparison of 2 techniques in 100 matched-pair patients]. / Valvulotomie mitrale percutanée: comparaison de deux techniques chez 100 patients appariés.

The results of percutaneous mitral valvulotomy by the Double Balloon and the Inoue Balloon were compared in 100 patients referred to Marie-Lannelongue Hospital between 7/11/86 and 8/3/91, paired for age, sex and echocardiographic features of the mitral valve. The transvalvular diastolic pressure gradient at catheterisation or Doppler echocardiography decreased significantly and similarly with the two techniques: -65 +/- 19% versus -64 +/- 15% (p = 0.10) and -66 +/- 22% versus -58 +/- 23% (p = 0.20) for the Double-Balloon and the Inoue Balloon techniques respectively. Similarly, the mitral valve surface area measured by planimetry during echocardiography or calculated from the Gorlin formula increased by 0.9 +/- 0.39 cm2 with the Double Balloon and by 0.88 +/- 0.32 cm2 with the Inoué Balloon (p = 0.91) or by 0.98 +/- 0.62 cm2 versus 0.87 +/- 0.45 cm2 (p = 0.42). The number of commissures opened was the same with both techniques. The evolution of mitral regurgitation after dilatation was similar: stable or less severe in 61% dilated by the Double Balloon and 68% by the Inoue Balloon; aggravation by more than one grade in 7 and 5% respectively (p = 0.17). Complications were rare and comparable with both techniques. The results with the two methods were the same if patients with an optimal indication and those with more severe mitral valve disease were considered separately. The rapidity and simplicity of percutaneous mitral valvulotomy with the Inoué Balloon make it the technique of choice.

[Treatment of rheumatic mitral stenosis in children and adolescents by a balloon catheter]. / Cure du rétrécissement mitral rhumatismal chez l'enfant et l'adolescent par un cathéter à ballonnet.

Between April 1986 and August 1987, eight children, 9 to 18 years old, with rheumatic mitral valve stenosis underwent a percutaneous balloon valvotomy. Immediate results were good. By echocardiographic measurements, mean transmitral gradients went from 23.7 +/- 7.1 mmHg to 5.8 +/- 4.7 mmHg (p less than 0.001), diastolic pressure half-times went from 299.0 +/- 15.6 msec to 123.3 +/- 42.3 msec (p less than 0.001) and mitral valve areas measured by planimetry went from 1.00 +/- 0.15 cm2/1.73 m2 to 2.68 +/- 0.71 cm2/1.73 m2 (p less than 0.001). Hemodynamic measurements showed a decrease in mean pulmonary artery pressure from 48.3 +/- 13.2 mmHg to 32.5 +/- 5.7 mmHg (p less than 0.01), a decrease in the mean left atrial pressure from 26.1 +/- 6.3 mmHg to 14.6 +/- 4.7 mmHg (p less than 0.02) and an increase in calculated mitral valve area (Gorlin formula) from 1.07 +/- 0.31 cm2/1.73 m2 to 2.73 +/- 1.17 cm2/1.73 m2 (p less than 0.02). The only complication observed was a grade III mitral insufficiency in a child with marked subvalvular fibrosis. Six patients had a very good result and one kept a moderate mitral stenosis. Percutaneous balloon valvotomy gives probably better results in children than in adults and obtains at least as good results as surgical closed mitral commissurotomy. Presently, this technique is the treatment of choice for rheumatic mitral stenosis in children. Contra-indications are intra-atrial thrombus and shortened, thickened subvalvular apparatus.

Comparison of three methods for recovery of Mycobacterium avium complex from blood specimens.

Three methods were used for the recovery of mycobacteria from blood specimens obtained from acquired immunodeficiency syndrome patients: (i) inoculation of 7H11 agar plates with a concentrated specimen, (ii) inoculation of 7H12 BACTEC vials with a concentrated specimen, and (iii) inoculation of 7H13 BACTEC vials with a nonconcentrated specimen. In this study, we examined 255 specimens and obtained positive mycobacterial growth in 47 of them. Among these 47 cultures, 40 were found to be positive by all three methods, and the total recovery rates in relation to these culture-positive specimens were 94% for method 1, 89% for method 2, and 96% for method 3. The advantages and disadvantages of these three methods are discussed.

Intravitreal ganciclovir in the treatment of AIDS-associated cytomegalovirus retinitis.

Ganciclovir (BW B759U, DHPG, dihydroxy propoxymethyl guanine) was injected directly into the vitreous in 14 eyes of 11 patients with severe acquired immune deficiency syndrome (AIDS)-associated cytomegalovirus (CMV) retinitis. All 11 patients either demonstrated serious myelosuppression which precluded the continuation of intravenous ganciclovir therapy (5 patients) or were experiencing progressive CMV retinitis despite therapy with maximum-tolerable dosages of intravenous ganciclovir (6 patients). Suppression of the retinitis was observed in 11 (78%) of the 14 treated eyes. Three eyes (22%) showed no improvement after the initial intravitreal injection. One rhegmatogenous retinal detachment (RD) occurred during an injection. There were no other complications, and no intraocular drug toxicity was observed. Reactivation of CMV retinitis necessitated repeated injections in 9 (64%) of the 14 eyes. The authors' experience with these 30 intravitreal injections indicates that the procedure is safe and effective both as an alternative to intravenous ganciclovir therapy in myelosuppressed patients and as a supplement to intravenous therapy in uncontrolled CMV retinitis.