A reversible pulmonary artery band: preliminary experience.

Pulmonary artery banding has become an infrequently used surgical technique. However, if a band was developed that could be relieved without the need for open heart surgery, it is likely that pulmonary artery banding would be used more frequently in the management of infants with congenital heart disease. Such a pulmonary artery band was placed in seven 1 week old mongrel puppies by using a loop of an absorbable suture material (Vicryl). One dog died at 2 months as a result of right ventricular failure. The remaining six dogs underwent cardiac catheterization and pulmonary balloon angioplasty at 6 months of age. After measuring pulmonary artery, right ventricular and aortic pressures and performing a right ventricular angiogram, balloon angioplasty of the band site was performed. A 20 mm balloon angioplasty catheter (Medi-Tech) was used in all dogs. Balloon angioplasty decreased right ventricular pressure from 101 +/- 19 to 42 +/- 3 mm Hg (p less than 0.05) and right ventricular systolic outflow tract gradient from 59 +/- 14 to 7 +/- 2 mm Hg (p less than 0.03), and increased the size of the band site from 8.7 +/- 0.03 to 14.9 +/- 0.5 mm (p less than 0.01). All dogs were recatheterized 2 months after angioplasty and were then killed for pathologic evaluation. At follow-up catheterization, right ventricular pressure, right ventricular outflow tract gradient and pulmonary artery size at the band site remained at the values obtained immediately after angioplasty. Postmortem examination demonstrated that there was no evidence of pulmonary artery damage. Although these studies are preliminary, they suggest that a reversible pulmonary artery band can be performed.

Posttransplant recoarctation of the aorta: a twelve year experience.

OBJECTIVES: This study was undertaken to investigate the incidence of posttransplant recoarctation of the aorta, delineate the mode of presentation, identify risk factors that predict recoarctation and examine the results of intervention for posttransplant recoarctation. BACKGROUND: Patients with aortic arch hypoplasia require extended arch reconstruction at transplant, with an inherent possibility of subsequent recoarctation of the aorta. METHODS: This was a retrospective review of all children (age <18 years) who underwent cardiac transplantation over a 10-year period. Collected data included pretransplant diagnosis, details of the transplant procedure and posttransplant data including development of recoarctation of the aorta, interventions for recoarctation and the most recent follow-up assessment of the aortic arch. RESULTS: Two hundred eighty-eight transplants were performed on 279 children (follow-up = 1,075 patient-years; range 0 to 133 months, median 43.7). Thirty-two of 152 patients (21%) who underwent extended aortic arch reconstruction subsequently developed recoarctation. All but one patient developed recoarctation within 2 years after transplant; 87% were hypertensive at presentation. Of 30 patients who underwent intervention for recoarctation (balloon angioplasty [n = 26] and surgical repair of recoarctation [n = 4]), 26 (87%) have remained recurrence-free (follow-up = 133 patient-years; range 8 to 106 months, median 47). CONCLUSIONS: The high frequency of recoarctation after cardiac transplantation with extended aortic arch reconstruction mandates serial echocardiographic evaluation of the aortic arch. Patients typically present with systemic hypertension within the first two years after transplantation. Balloon angioplasty is a safe, effective and durable method of treatment.

The role of anti-pig antibody in pig-to-baboon cardiac xenotransplant rejection.

The role of naturally produced antibody in discordant xenograft rejection is still uncertain. Twelve orthotopic pig-to-baboon heart transplants (HTx) were performed. In 2 baboons, no antibody adsorption (AbA) was performed. In 5 baboons, AbA with a pig lung was performed during circulatory arrest. In 5 baboons, AbA and blood exsanguination at the beginning of cardiopulmonary bypass (CPB) were performed. Baboons were divided into 2 groups; group 1 (n = 4) died within 24 hr of HTx and group 2 (n = 8) survived more than 24 hr. Mean survival period was 9.8 +/- 3.0 hr in group 1 and 151 +/- 33 hr in group 2. Baboon anti-pig antibody (Ab) was measured before CPB, before circulatory arrest, during AbA, at the end of CPB, and daily after HTx. Anti-RBC Ab was measured by the titration method at temperatures of 4 degrees C and 37 degrees C (RAb-4 and RAb-37). Anti-endothelial cell Ab (EAb) and anti-white blood cell Ab (WAb) titers were measured with ELISA. RAb titration > or = 1/4 and EAB and WAb > or = 1/256 were determined to be seropositive (S(+)). S(+) rate of RAb-37 at the end of CPB (endCPB) in group 2 was significantly higher than that in group 1 (8/8 vs. 1/4; P < 0.05). The seronegative (S(-)) rates of RBC-4 and EAb (endCPB) in group 2 were higher than those in group 1 (7/8 vs. 1/4 and 6/8 vs. 1/4, respectively), but not significantly. There was no difference in S(-) rate of WAb (endCPB) between group 1 and group 2. More than 4-fold decrease in RAb-4 and RAb-37 by AbA with a pig lung was observed in 5 and 7 of 8 baboons, while EAb and WAb did not change by AbA. In all of group 2, RAb-4 reverted to S(+) within 3 days after HTx. One baboon had no rejection episode and died of infection 16 days after HTx (baboon 16); however, it also became S(+) for RAb-4 a day after HTx until death. In 4 of group 2, RAb-37 became S(+) 1 or 2 days before death by rejection. Baboon 16, however, became S(+) for RAb-37 7 days after HTx and S(-) again 9 days after HTx until death. EAb became S(+) in all of group 2, but 5 of them survived more than 5 days after seroconversion.(ABSTRACT TRUNCATED AT 400 WORDS)

Ventricular remodeling following infant-pediatric cardiac transplantation. Does age at transplantation or size disparity matter?

Early left ventricular (LV) remodeling following pediatric cardiac transplantation has not been described. To identify patterns and determinants of change in left ventricular mass and volume posttransplant, we studied 125 consecutive children who underwent cardiac transplantation between January 1, 1989 and July 31, 1993. Two-dimensional imaging-directed M-mode echocardiograms were studied weekly until 26 weeks post-transplant. LV mass and volume (indexed to BSA1.5) were measured. LV mass index increased until 3 weeks post-transplant, and then decreased. The mean decrement in LV mass index after 8 weeks post-transplant (relative to baseline) was significantly larger in patients with donor-recipient weight ratio > 1.5 compared with patients with donor-recipient weight ratio < or = 1.5 (-2.2 g/m3 compared with 33.4 g/m3, respectively, P < 0.01). Multiple linear regression was performed employing donor-recipient weight ratio, time since transplantation, ischemic time, and age at transplant as prognostic variables. Donor-recipient weight ratio (P < 0.0001), time since transplant (P < 0.01), and age at transplant (P = 0.02) were identified as independent predictors of change in LV mass index. Donor-recipient weight ratio (P = 0.001) and time since transplantation (P = 0.02) were independent predictors of change in LV volume index. There was an interaction between donor-recipient weight ratio and time since transplantation, suggesting that donor-recipient weight ratio has an independent effect as well as a time-dependent effect on change in LV mass and volume indices. LV mass and volume indices increased early posttransplant and then decreased; this pattern was temporally predictable, and dependent on donor-recipient weight ratio and age at transplant.

Modification of myocardial ischemia in normal and hypertrophied hearts utilizing diastolic retroperfusion of the coronary veins.

Myocardial infarction following coronary occlusion limits the effectiveness of emergency coronary artery bypass operations. We designed and evaluated a transvenously introduced balloon-tipped catheter and an electrocardiogram-activated pumping system for perfusing ischemic myocardium by retrograde pulsation of oxygenated blood into the coronary veins during diastole. Balloon deflation during systole allowed for normal venous drainage. Four groups of dogs (n = 26) were instrumented with sonomicrometry crystals and catheters to measure regional and global left ventricular function. Two groups of dogs had chronic left ventricular (LVH) produced by prior aortic banding (left ventricular mass 174 gm versus 115 gm for control dogs of equal body weight, p less than or equal to 0.05). The left anterior descending coronary artery (LAD) was occluded for 40 minutes; after 10 minutes left ventricular function was severely depressed in all groups (less than or equal to 0.05 compared to baseline). Groups 1 (normal left ventricle, n = 8) and 2 (LVH, n = 5) had no further therapy for the following 30 minutes. Groups 3 (normal left ventricle, n = 8) and 4 (LVH, n = 5) received 30 minutes of coronary vein retroperfusion (CVRP) 10 minutes following the LAD occlusion. CVRP restored 37% of systolic shortening, whereas there was no restoration of systolic shortening in control dogs (p less than or equal to 0.001). All other physiological and hemodynamic parameters including heart rate, cardiac output, aortic pressure, dP/dt, and left ventricular dilatation were normalized during CVRP while remaining severely depressed in control dogs (p less than or equal to 0.05). Following restoration of arterial flow at 40 minutes, 10 of 13 CVRP-treated dogs recovered normal left ventricular function while only two of 13 untreated dogs survived. CVRP offers a transvenous approach for modifying myocardial ischemia prior to emergency coronary artery bypass grafting.

Prognostic factors in valvotomy for critical aortic stenosis in infancy.

Aortic valvotomy for critical aortic stenosis in infancy has had a high mortality. To determine the factors that influence survival, we reviewed the cases of 24 infants who underwent aortic valvotomy in the first 6 months of life (mean 4 1/2 weeks) for aortic stenosis from 1978 to 1984. Cardiopulmonary bypass was used in all patients. Operative mortality was 21% (5/24), four of the five deaths occurring from low cardiac output. Analysis of preoperative factors affecting survival versus nonsurvival revealed that low ejection fraction (60% +/- 17% in survivors versus 36% +/- 2% in nonsurvivors), high left ventricular end-diastolic pressure (16 +/- 7 mm Hg in survivors versus 30 +/- 14 mm Hg in nonsurvivors), and presence of endocardial fibroelastosis (25% in survivors versus 100% in nonsurvivors) all were predictive of a poor outcome, although the small sample size indicated caution in interpreting results. Factors that did not appear to influence survival included peak systolic gradient (79 +/- 30 mm Hg in survivors versus 60 +/- 15 mm Hg in nonsurvivors) and left ventricular end-diastolic volume (37 +/- 17 cm3/m2 in survivors versus 36 +/- 7 cm3/m2 in nonsurvivors). Four patients with a left ventricular end-diastolic volume below 26 cm3/m2 survived. Postoperative gradients averaged 25 +/- 21 mm Hg at 3.4 +/- 2 years' follow-up in nine recatheterized patients. Ejection fraction of these patients increased from 45% +/- 10% to 70% +/- 11% and left ventricular end-diastolic volume increased from 37 +/- 17 to 58 +/- 5 cm3/m2. Two of 17 patients have required apical-aortic conduits; all other patients are asymptomatic. We conclude that infants with critical aortic stenosis benefit from valvotomy even with impaired left ventricular function and severely reduced left ventricular dimensions and many have nearly normal hemodynamics on late follow-up.

Sutureless coronary artery bypass with biologic glued anastomoses: preliminary in vivo and in vitro results.

OBJECTIVE: As heart surgery becomes increasingly focused on minimally invasive techniques, it has become apparent that conventional techniques of anastomosis will need to be severely altered or abandoned. Toward that end, we developed and tested in vitro and in vivo coronary artery bypass graft anastomoses using a biologic glue formulated from bovine albumin and glutaraldehyde. We used a double-balloon catheter as a temporary internal stent to create and seal the anastomosis during gluing. METHODS: Initially, anastomoses were made between cryopreserved human saphenous vein segments and coronary arteries in vitro on 12 intact bovine hearts. A total of 42 anastomoses were created with the catheter system introduced into the distal end of the graft, exiting the back wall, and entering the anterior wall of the coronary artery. Two balloons (one in the graft and one in the coronary artery) held the anastomosis stable while the biologic glue was applied externally and allowed to set for 2 minutes. The balloon catheter was then removed from the end of the graft simulating a side-to-side internal thoracic artery anastomosis. After the graft had been flushed to assure distal end patency, the open end of the graft was clipped, turning the anastomosis into an end-to-side graft. A pressure transducer was then attached to the graft and saline solution forcefully infused. RESULTS: All grafts easily held a pressure of 300 mm Hg; 10 grafts were tested up to 560 mm Hg without leaks. Distal and proximal coronary artery patency was checked by examining flow out of the coronary ostia and by cutting arteries distal to the grafts. All anastomoses were patent on being opened and no glue was seen intraluminally. Subsequently, 3 anastomoses of the left internal thoracic artery to the left anterior descending artery have been constructed in goats, with autopsies at 24 hours, 10 months, and 1 year revealing patent anastomoses. CONCLUSION: A biologic glue and catheter system has been developed that allows a coronary anastomosis with a high bursting strength to be performed. When the system has been further developed and tested, truly minimally invasive heart surgery may be possible.

Risk factors for graft failure associated with pulmonary hypertension after pediatric heart transplantation.

Postoperative pulmonary hypertension can be a major cause of early death after heart transplantation in children. To identify predictive risk factors of pulmonary hypertension after heart transplantation, we performed a retrospective analysis of our 194 infant and pediatric recipients who underwent heart transplantation between 1987 and 1992. Because the response of pulmonary vasculature may change during growth, the patients were divided into two groups: age less than 1 year in group I (n = 152) and 1 year or older in group C (n = 43). The following risk factors were evaluated: cardiomyopathy, congenital heart disease and hypoplastic left heart syndrome, pretransplant pulmonary hypertension, history of operation, oversized donor (donor/recipient weight ratio greater than 2), donor's history of cardiopulmonary resuscitation, and prolonged graft ischemic time (graft ischemic time 360 minutes or longer). Though there was no significant difference between group I and group C in overall early mortality including early graft loss (19 of 152 versus 5 of 42), the mortality rate from pulmonary hypertension in group I was significantly lower than that in group C (2 of 152 versus 4 of 42; p < 0.05). The mortality rate from pulmonary hypertension in patients with congenital heart disease in group I was significantly lower than that in group C (0 of 44 versus 4 of 24; p < 0.05). In group I, there was no significant difference in the early mortality rate or the mortality rate from pulmonary hypertension from any factors studied. The mortality rate from pulmonary hypertension in association with prolonged graft ischemic time in group C was significantly higher than when no prolonged graft ischemic time was present in group C and with either prolonged graft ischemic time or no prolonged graft ischemic time in group I (4 of 16 versus 0 of 26, 0 of 37, and 2 of 115). In conclusion, older patients had a higher mortality rate from pulmonary hypertension after heart transplantation, especially in patients with congenital heart disease who received a graft preserved more than 6 hours. This study demonstrates another benefit of early heart transplantation in infancy, that is, prevention of death from pulmonary hypertension.

Bless the babies: one hundred fifteen late survivors of heart transplantation during the first year of life. The Loma Linda University Pediatric Heart Transplant Group.

There is a rapid growth of interest in heart transplantation therapy during early infancy. From 10% to 25% of the infants who are listed for transplantation annually have died while awaiting a donor heart. There has been no significant trend in this variable. Since November 1985, 140 consecutive orthotopic transplantation procedures were performed in 139 infants who were from 3 hours to 12 months of age. Indications for transplantation included hypoplastic left heart syndrome (63%), other complex structural anomalies (29%), myopathy (6.5%), and tumors (1.5%). Most recipients had ductus-dependent circulation and received continuous infusion of prostaglandin E1. Heart donors were usually victims of trauma, sudden infant death, or birth asphyxia. A donor-recipient weight ratio of 4.0 or less was found to be acceptable. The amount of time the graft underwent cold ischemia, ranged from 64 to 576 minutes. The procurement process was facilitated by a single dose of cold crystalloid cardioplegic solution and cold immersion transport. Profound hypothermic circulatory arrest was used for graft implantation. One hundred twenty-four (89%) recipients survived transplantation and were discharged from the hospital. There were 9 late deaths, which resulted in an 83% overall survival. The 5-year actuarial survival is 80%. The survival among newborn recipients (n = 60) at 5 years is 84%. Chronic immunomodulation was cyclosporine-based and steroid-free. Surveillance was noninvasive and relied heavily on echocardiography, electrocardiography, and clinical intuition. There was one documented late lethal infection, tumor was not encountered, and coronary occlusive disease was known to exist in only one long-term survivor. We concluded that transplantation results in excellent life quality and is a highly effective and durable therapy when applied during early infancy.

Successful survival of primates receiving transplantation with "dead," nonbeating donor hearts.

A paucity of donor organs is the principal limitation in human heart transplantation. Prompted by our short-term studies of reanimating "dead" donor hearts in sheep, we applied the same reperfusion modifications in juvenile baboons to determine human applications in an anoxic arrest model (as occurs when non-brain-dead patients are extubated and allowed to die). Ten juvenile baboons (mean weight 3.6 kg) were studied. Five baboons were used as donors. After being anesthetized, donors were pretreated with methylprednisolone (Solu-Medrol), 50% dextrose, nifedipine, and prostaglandin E1 and then paralyzed and extubated. Donors became pulseless at 7 +/- 1 minutes and had electric arrest 9 to 18 minutes after paralysis. The five donors were left undisturbed and warm for 15, 22, 30, 30, and 31 minutes, respectively, after asystole. They were then given 250 ml of 4 degrees C Roe's crystalloid cardioplegic solution via the aortic root and the hearts were explanted into iced Euro-Collins solution. Five baboons served as recipients. After donor harvest, recipients were placed on cardiopulmonary bypass, given prostaglandin E1, and cooled to 18 degrees C; circulatory arrest was instituted and the recipient's heart excised. The donor heart was transplanted in an orthotopic position. Before reinstitution of bypass, 250 ml of terminal leukocyte-depleted blood cardioplegic solution was given, then bypass was restarted and the hearts were reperfused for 60 minutes. All animals were weaned from bypass without the use of inotropic agents. All animals were extubated within 2 to 4 hours after bypass and received standard immunosuppression. Peak creatine kinase MB/total creatine kinase ratio was 0.2% +/- 0.2%. Postoperative ejection fractions by echocardiography were 75% to 80% (mean 76%). Animals survived 1, 9, 13, 16, and 34 days, with three deaths caused by acute rejection and one each by stroke and diarrhea/dehydration. Pathologic findings showed no areas of fibrosis or ischemic damage. We conclude that successful reanimation and engraftment can be achieved with the use of the asystolic primate heart; this work suggests that human application is realistic and could greatly expand the donor pool.

Myosin light chain efflux after heart transplantation in infants and children and its correlation with ischemic preservation time.

Serum levels of cardiac myosin light chain 1 after heart transplantation were studied in 24 infants and children who underwent heart transplantation between June 1990 and April 1991. The ages of the patients ranged from 4 days to 6 years 7 months (mean, 9.9 months), and their body weights ranged from 2.2 to 20 kg (mean, 5.6 kg). The ages of the donors ranged from 2 days to 8 years, 7 months (mean, 26.6 months), and their body weights ranged from 2.5 to 26 kg (mean, 11.4 kg). The donor heart ischemic time ranged from 90 minutes to 482 minutes (mean, 279 minutes). Peak myosin levels after heart transplantation showed significant correlation with the duration of graft ischemia (p < 0.01) and with diastolic cardiac function in the first posttransplant week (p < 0.05). Peak myosin levels did not correlate with systolic cardiac function, age of the donor, or age of the recipient. Myosin levels of the 15 patients with graft ischemic times exceeding 4 hours averaged 6.30 +/- 3.50 ng/ml. These levels were significantly higher than those of patients with graft ischemia lasting less than 4 hours (2.60 +/- 1.20 ng/ml; p < 0.01). Both of the values are higher than previously reported values of normal controls but lower than previously reported values of patients with myocardial infarction. Preservation techniques used for this series of transplant operations provided good clinical protection of the donor heart for up to 8 hours, although release of the cardiac myosin light chain fragment correlated with duration of graft ischemia. Cardiac myosin levels appeared to be a good indicator of heart graft damage during ischemic preservation. It remains to be determined at what level of myosin release (and, hence, at what duration of graft ischemia) irreversible myocardial damage, which might result in permanent functional compromise, occurs.

Optimal delivery of cardioplegic solution for "redo" operations.

Increasing experience suggests that retrograde cardioplegia offers several benefits during cardiac reoperations. However, the need for dissection to allow caval snares for open coronary sinus intubation or to palpate the atrioventricular groove for transatrial coronary sinus intubation may disturb diseased vein grafts or require more dissection than necessary. Although antegrade-retrograde techniques can be used, antegrade cardioplegia risks atheromatous embolization from old vein grafts. To optimize delivery of cardioplegic solution, we designed and used "no touch" transatrial intubation of the coronary sinus for retrograde delivery of cardioplegic solution in 63 consecutive patients aged 20 to 87 years (mean 68 years) undergoing 36 redo coronary bypass operations, 7 combined redo coronary bypass/valve replacements, 6 redo aortic valve repairs/replacements, 6 redo mitral valve repairs/replacements, 4 redo double valve repairs/replacements, 2 redo triple valve repairs/replacements, and 2 redo composite aortic valve and arch replacements. "No touch" coronary sinus cannulation was achieved by minimally dissecting the aorta and high right atrium enough for two purse-string sutures. No attempt was made to dissect the junction of the inferior vena cava and atrioventricular groove if old vein grafts were present. The distal pressure line of the Gundry DLP RCSP retrograde cardioplegia cannula (DPL, Inc., Grand Rapids, Mich.) was connected to a transducer, flushed, and then introduced into the right atrium. The pressure tracing thus obtained was observed while the catheter was advanced, using its curved stylet, "blindly" without touching the heart, through the right atrium into the coronary sinus until a coronary sinus waveform was obtained (similar to floating a thermodilution catheter). The catheter's distal balloon was then inflated to occlude the coronary sinus momentarily. A rise in sinus pressure confirmed placement. If pressure did not rise, the cannula was usually in the right ventricle and was repositioned. All coronary sinuses were successfully intubated blindly. Bypass was then instituted, the aorta crossclamped, and the proximal aorta vented. Old vein grafts were cut at the aorta before retrograde cardioplegia was begun; atheromatous material was routinely flushed retrogradely from vein grafts. Only after arrest were hearts dissected as needed. Antegrade cardioplegia was not used. There were two (3%) deaths, both from hospital-acquired pneumonia, no perioperative myocardial infarctions, and no episodes of heart block. Inotropic agents were used in six of 63 patients (10%). We conclude that "no touch" transatrial retrograde cardioplegia offers optimal, simplified myocardial protection for cardiac reoperations, permits arrest of the heart before cardiac manipulations, and expands the use of retrograde cardioplegia by obviating cardiac dissection.

The optimal Fontan connection: a growing extracardiac lateral tunnel with pedicled pericardium.

OBJECTIVE: The concept of a lateral tunnel for the Fontan operation is now widely accepted. Most lateral tunnels are constructed intraatrially with the use of aortic crossclamping. Construction of extracardiac lateral tunnels with the use of homografts or other nonviable tubes eliminates aortic crossclamping but lacks growth potential in length or width. The native pericardium, which is "sealed" posteriorly along the pulmonary artery, atrium, and inferior vena cava, could be turned down onto the right atrium to form a viable extracardiac lateral tunnel. METHODS: We designed and successfully constructed extracardiac lateral tunnels using viable autologous pericardium, pedicled on its lateral blood supply, in 19 patients aged 9 months to 5 years. All patients had a previous Glenn shunt; five patients had dextrocardia and a midline inferior vena cava. The patients' inferior vena cava-right atrial connection was opened transversely and the right atrial opening was sutured to its back wall, keeping the eustachian valve in the inferior vena cava. The underside of the right pulmonary artery was opened longitudinally; its inferior edge was sewn to the adjacent pericardial reflection. Any "pocket" or depressions in the posterior pericardium along the pulmonary veins were closed with running suture. Two incisions were made in the right pericardium down to the phrenic nerve parallel to the inferior vena caval and pulmonary arterial openings. This pedicled pericardium was trimmed and sewn as a roof to the upper edges of the inferior vena cava and pulmonary artery openings and then sewn longitudinally along the unopened right atrial wall, completing the viable extracardiac lateral tunnel. Although no fenestrations were used, these could be made during construction, or more significantly, owing to the lack of thick walled structures, in the catheterization laboratory in the postoperative period. RESULTS: All 19 patients had respiratory/cardiac pulsations in the pulmonary arteries owing to the compressible lateral tunnel. At follow-up of up to 2 1/2 years, all tunnels are growing and no obstructions have occurred. CONCLUSION: The viable autologous pericardial extracardiac lateral tunnel can be constructed without cardiac ischemia, can be fenestrated in the postoperative period, and forms a compressible, nonthrombogenic conduit capable of growth, which can be constructed early in infancy.

Candida esophagitis following cardiac operation and short-term antibiotic prophylaxis.

Dysphagia and retrosternal pain are common complaints in patients after cardiac operations, and most often they result from the median sternotomy and/or endotracheal intubation. Although Candida esophagitis is a recognized cause of similar symptoms, it is usually not suspected except in immunologically compromised hosts. This report describes the case histories of five patients, not immunosuppressed or cachectic, who developed persistent dysphagia during recovery from cardiac operations; four patients received only 4 days of preoperative and postoperative prophylactic antibiotic treatment with cefazolin (Kefzol) and cephalexin (Keflex). A nasogastric tube had been used for less than 24 hours in the postoperative period. The fifth patient developed symptoms following prolonged and varied antibiotic therapy. Candida esophagitis was diagnosed by a combination of coexisting oral candidiasis (5/5), roentgenographic appearance on barium swallow (5/5), endoscopy (4/4), and biopsy or culture (2/4). Initial therapy consisted of antireflux measures and antacids (4/5), cimetidine (4/5), oral nystatin in methylcellulose base (1,000,000 units every 4 hours) (4/5), and termination of other antibiotic therapy (1/5). These measures were effective in clearing the infection in only two patients. A third patient required prolonged massive oral nystatin therapy, and in two patients intravenous Amphotericin B was necessary to control infection. Two patients subsequently developed strictures which necessitated multiple esophageal dilatations. One of these patients developed endocarditis during home dilatation therapy. All patients are currently free of disease. Current measures utilized to recognize and treat the disease are discussed.

Chest roentgenograms in diagnosis of traumatic rupture of the aorta. Observer variation in interpretation.

A significant p value is not always a good measure of the clinical value of a diagnostic test. By examining interobserver agreement with respect to chest roentgenogram interpretation, we determined which of seven roentgenographic signs statistically associated with traumatic rupture of the aorta (TRA) are most likely to be useful in clinical practice. Four surgeons and two radiologists were asked to interpret individually, in blinded fashion, the initial chest films of 149 trauma patients who had undergone aortography to rule out TRA. Agreement between all observers, as well as between specialty groups, was examined by chi-square and by calculation of Cohen's kappa statistic, which estimates the extent of agreement. Statistically significant agreement (p = .0000) was found between all observers with regard to mediastinal widening and obscuration of the aortic knob, but other comparisons showed no better than random agreement. Of the seven roentgenographic signs associated with TRA, identification of mediastinal widening and obscuration of the aortic knob show the most consistent interobserver agreement and are the most likely to be useful in clinical practice.

Seven-year follow-up of coronary artery bypasses performed with and without cardiopulmonary bypass.

BACKGROUND: There has been resurgent interest in coronary revascularization performed on the beating heart. Heretofore, there has been no long-term comparison of this technique to traditional coronary artery bypass with cardioplegia. OBJECTIVE: The purpose of this study was to provide a comparison of long-term survival and intervention-free outcome between patient groups subjected to coronary bypass accomplished with or without the use of cardiopulmonary bypass. METHOD: From June 1989 to July 1990, all patients treated for coronary revascularization by three surgeons were considered for coronary revascularization with the heart beating: 107 patients underwent coronary bypass on the beating heart, and 112 patients underwent revascularization with the aid of bypass with cardioplegia. Mean ages (65 +/- 10 years) and risk factors were identical. Patients operated on with the heart beating had 2.4 +/- 0.9 grafts versus 3.2 +/- 1.1 grafts for patients having cardiopulmonary bypass with cardioplegia. RESULTS: At 7-year follow-up, 86 of 107 (80%) patients operated on with the heart beating were alive versus 88 of 112 (79%) patients in whom cardiopulmonary bypass with cardioplegia was used. Cardiac deaths occurred in 13 of 107 (12%) patients in the former group versus 10 of 112 (9%) patients in the latter group. However, 32 of 107 patients operated on with the heart beating (30%) needed catheterization for their symptoms versus 18 of 112 (16%) patients in the bypass with cardioplegia group (p = 0.01). This results in 21 of 107 (20%) patients in the beating heart group needing angioplasty or a second coronary bypass versus only 8 of 112 (7%) patients in the bypass with cardioplegia group. No patient in the bypass with cardioplegia group required reoperation. Most of the reinterventions for the beating heart group were percutaneous transluminal coronary angioplasty (15 of 21 [71%] patients). CONCLUSION: Despite one less graft per patient, survival and cardiac death rates were similar for the two groups. However, twice as many patients in the beating heart group required recatheterization (30% versus 16%), and 20% needed a second intervention. Only 7% of the bypass with cardioplegia group required reintervention. Limited revascularization of the beating heart provides long-term results comparable to full revascularization with cardiopulmonary bypass, but at the cost of a threefold increase in reinterventions.

Fate of the pericardial monocusp pulmonary valve for right ventricular outflow tract reconstruction. Early function, late failure without obstruction.

Sudden change from a pressure-loaded to a volume-loaded right ventricle as a result of transannular patch reconstruction of the right ventricular outflow tract may result in early hemodynamic compromise. Tri-leaflet porcine valves in conduits function well early but typically create late obstruction. We studied the fate of a pericardial monocusp valve constructed during transannular patch reconstruction for right ventricular outflow tract obstruction in 19 patients, 2 weeks to 27 years of age (mean age 61 months). Patients had the monocusp constructed of autologous (n = 16) or bovine pericardium (n = 3) when the former was not available during transannular patch reconstruction associated with repair of tetralogy of Fallot (n = 12), pulmonary stenosis/atresia (n = 4), and truncus arteriosus (n = 3). Function of the monocusp was assessed by presence of a split-second heart sound, echocardiographic assessment of right ventricular dilatation, monocusp competence, and fluoroscopic evaluation of monocusp motion. Functional assessments were accomplished immediately after the operation and at 2, 6, 12, and 24 months after the operation. There were no operative deaths, but there was one late hospital death. Sixteen of nineteen patients (84%) had competent pulmonary monocusp valves immediately after the operation, but, by 24 months, only one of seven patients (14%) had a competent valve. No patient had monocusp stenosis. We conclude that a pericardial monocusp valve for right ventricular outflow tract reconstruction provides excellent early hemodynamic function but that these effects are limited in duration. Because late stenosis has not been seen, this inexpensive and easily constructed valve can be used as an excellent short-term adjunct to right ventricular outflow tract reconstruction.

Transplantation of hearts after arrest and resuscitation. Early and long-term results.

Transplant surgeons are reluctant to use hearts that have undergone cardiopulmonary resuscitation for cardiac arrest because of the fear of poor early and late cardiac function. A policy of minimizing contraindications to use of donor hearts has led to the unique opportunity of assessing the effects of donor arrest and successful cardiopulmonary resuscitation on early and late cardiac function in pediatric heart transplantation. A number of 140 infants and children undergoing transplantation from birth to 17 years of age were studied retrospectively and divided into two groups on the basis of cardiopulmonary resuscitation status. Group 1 (72 patients) received donor hearts that were not subjected to cardiopulmonary resuscitation; group 2 (68 patients) received donor hearts that had cardiopulmonary resuscitation for a mean of 18.8 +/- 14.6 minutes, the longest period of time being 60 minutes. Mean ischemic times were almost identical in the two groups: 4.43 +/- 2.0 hours (cardiopulmonary resuscitation) versus 4.5 +/- 2.1 hours (no cardiopulmonary resuscitation). Early cardiac function was assessed on the basis of the number of days the recipient was supported by the ventilator, days receiving dopamine, days receiving isoproterenol, and the amount of inotropic agents required after the operation. The groups did not differ. Parameters of systolic function included fractional shortening, posterior wall thickening, and maximum velocity of change in left ventricular posterior wall dimension during systole. Diastolic function was measured on the basis of left ventricular end-diastolic volume, left ventricular mass, and maximum velocity of change in left ventricular posterior wall dimension during diastole. Both systolic and diastolic function were measured and analyzed from M-mode echocardiography at 1 week, 1 month, 6 months, 1 year, and 2 years after the operation. There were no statistically significant differences in graft function between the two groups in any of the echocardiographic parameters studied, even at 2 years. No group differed from ranges of normal. Our results suggest that hearts undergoing cardiopulmonary resuscitation for periods of up to 60 minutes can be used safely without evidence of deterioration of early or late cardiac function.

The Mustard procedure for correction of simple transposition of the great arteries before 1 month of age.

Since April 1976, 34 infants (25 male and 9 female) less than 1 month of age underwent a Mustard intraatrial baffle procedure for repair of simple transposition of the great arteries. Thirty patients were less than 2 weeks old and 19 patients less than 1 week (mean 7.8 +/- 6 days). The weights ranged from 2.6 to 4.4 kg (mean 3.4 +/- 0.4 kg). Rashkind balloon atrial septostomy was performed in the first hours or days of life in 29 patients. The average interval from balloon atrial septostomy to baffle repair was 3.9 days (range 2 hours to 14 days). Mechanical ventilation was required in eight patients preoperatively and prostaglandin E1 was infused in 17 patients to maintain ductal patency. In all patients, the Mustard procedure was performed with the use of deep hypothermic circulatory arrest, averaging 53 minutes (range 37 to 82 minutes). The duration of postoperative intubation and ventilatory support averaged 1.7 +/- 1.0 days (range 1 to 5 days). Inotropic drugs were used in 24 patients during a period of 1.4 +/- 1.3 days (range 1 to 6 days) postoperatively. There were no hospital deaths. Follow-up evaluation has extended from 1 month to 14 years (mean 3 +/- 3 years). One infant died 2 months postoperatively as a result of milk aspiration; no cardiac defects were found at the autopsy. A second infant died at 1 year with right ventricular and tricuspid valve dysfunction. Baffle complications occurred in 6 of the 32 survivors, including superior vena caval stenosis in 4, inferior vena caval stenosis in 1, and pulmonary venous obstruction in 3. Reoperations for baffle obstructions were performed in three patients (8.8%) and balloon angioplasties in two. One patient required permanent pacemaker implantation. Results with the Mustard procedure before 1 month of age show that it can be performed with negligible mortality and a low incidence of late complications at an age comparable to when arterial switching would be performed. Until long-term studies demonstrate superiority of arterial operations, the low operative mortality favors continued evaluation of the neonatal Mustard repair as a valid alternative to the arterial switch.

The effects of donor-recipient size disparity in infant and pediatric heart transplantation.

To determine the effect of heart donor and recipient size mismatches in infant and pediatric heart transplantation, we studied all 69 patients (age 1 day to 11 years) having 71 orthotopic heart transplants from 1985 to 1989. Patients were divided into three groups based on donor to recipient weight ratios. Group I comprised 13 heart transplants with a donor to recipient weight ratio less than 0.95 (mean 0.81, range 0.48 to 0.94); group II comprised 29 heart transplants with a weight ratio between 0.95 and 1.60 (mean 1.28); and group III had 27 heart transplants with weight ratios greater than 1.60 (mean 2.2, range 1.61 to 3.09). All chests were closed primarily. The cardiothoracic ratio by chest radiography was significantly larger in group III (p = 0.0002); 75% of group III patients had periods of lobar or complete lung collapse by chest radiography compared with 28% of group II and 19% of group I patients (p < 0.05). Despite this, there was no difference in the number of days of ventilator support for any group (p = 0.92). There was no difference in graft ischemic time or inotropic drug use among groups, nor were differences found in the cardiac systolic function parameters of left ventricular preejection time (p = 0.975), left ventricular ejection time (p = 0.975), left ventricular fiber shortening (p = 0.97), and left ventricular fractional shortening (p = 0.596). Thus despite a high incidence of transient lobar or complete lung collapse in high donor to recipient weight ratio transplants, large donor heart size produces very little clinical impairment in recipient lung function. Size mismatches do not influence cardiac systolic function. Overall, large size mismatches appear to be very well tolerated in infant and pediatric heart transplantation.