Venovenous modified ultrafiltration after cardiopulmonary bypass in children: a prospective randomized study.

BACKGROUND: Cardiopulmonary bypass is associated with the production of both proinflammatory and anti-inflammatory cytokines, the balance of which leads to varying degrees of postoperative systemic inflammation. Arteriovenous modified ultrafiltration effectively reduces total body water and improves postoperative hemodynamic and homeostatic functions. Venovenous modified ultrafiltration is a modification of this technique, which has the potentially added advantage of eliminating the obligatory left-to-right shunt associated with arteriovenous modified ultrafiltration. We tested the hypothesis that venovenous modified ultrafiltration is a safe and effective method of achieving ultrafiltration in children after cardiopulmonary bypass. METHODS: Thirty-eight pediatric patients were randomly assigned to undergo conventional, venovenous (n = 13), or no ultrafiltration venovenous (n = 13), and controls (n = 12). Perioperative, cardiopulmonary, and cytokine (tumor necrosis factor-alpha, interleukin-1beta, interleukin-6, interleukin-8, and interleukin-10) data were collected for statistical analysis. RESULTS: Compared with patients in the conventional ultrafiltration and control groups, patients undergoing venovenous modified ultrafiltration had the greatest volume of ultrafiltrate removed (46. 9 +/- 8.4 mL/kg vs 20.1 +/- 5.0 mL/kg and 0 mL/kg for conventional ultrafiltration and control groups, respectively; P =.0001), least increase in total body water (1.91% +/- 1.49% vs 3.90% +/- 1.86% and 8.24% +/- 3.41%; P =.05), greatest rise in hematocrit (39.7% +/- 1. 7% vs 33.8% +/- 2.1% and 29.6% +/- 2.3%; P =.006), and shortest length of hospital stay (4.41 +/- 0.28 days vs 6.69 +/- 1.47 days and 8.38 +/- 1.11 days; P =.03, P =.03). CONCLUSIONS: Venovenous modified ultrafiltration is a safe and effective method of reducing the increase in total body water and duration of postoperative convalescence after cardiopulmonary bypass.

Modified Norwood operation for single left ventricle and ventriculoarterial discordance: an improved surgical technique.

BACKGROUND: Patients with univentricular hearts and ventriculoarterial discordance with potentially obstructed systemic blood flow continue to pose difficult management problems. The goals of neonatal palliative operations are to control pulmonary blood flow while avoiding pulmonary artery distortion, to relieve systemic outflow tract obstruction, and to avoid heart block. METHODS: Between January 1987 and December 1996, 38 patients with either tricuspid atresia or a double-inlet left ventricle and ventriculoarterial discordance underwent a modified Norwood procedure. Their mean age was 15 days, and their mean weight was 3.4 kg. Aortic arch anomalies were present in 92% of the patients. Morbidity and mortality statistics, intraoperative data, and postoperative echocardiograms were reviewed. RESULTS: There were 3 early deaths (7.8%) and 5 late deaths (13.1%). The actuarial survival rates at 1 month, 1 year, and 5 years were 89%, 82%, and 71%, respectively. Follow-up was complete in all children at a mean interval of 30 +/- 9 months. None of the patients had significant neoaortic valve insufficiency, and 1 patient required therapy for residual aortic arch obstruction. Nine patients (30% of the survivors) have undergone the hemi-Fontan procedure, and 18 patients (60%) successfully have undergone the Fontan procedure. CONCLUSIONS: In this patient population, we recommend the modified Norwood procedure as the neonatal palliative treatment of choice. It can be performed with acceptable early morbidity and mortality, and it improves suitability for the Fontan procedure. It reliably relieves all levels of systemic outflow tract obstruction, controls pulmonary blood flow, and avoids heart block.

Bidirectional superior cavopulmonary connection in young infants.

BACKGROUND: Bidirectional superior cavopulmonary connection (BSCC) has become widely used in patients with univentricular AV connections. However, concerns remain about perioperative morbidity and mortality and about the adequacy of oxygenation after cavopulmonary connection in very young patients. This report examines our experience with BSCC in young infants to evaluate whether young age affects operative outcome, to examine the effect of young age on postoperative oxygenation, and to define the lower age limit for successful use of the procedure. METHODS AND RESULTS: The records of the 85 consecutive patients < 6.5 months old who underwent BSCC from December 1990 through February 1995 were reviewed. The average patient age was 4.8 +/- 1.4 months (range, 5 weeks to 6.5 months), with 13 patients being < 3 months old. There were 5 hospital deaths (6%; 70% confidence limits, 3% to 10%). Pulmonary artery thrombosis occurred in 3 patients (4%; 70% confidence limits, 2% to 7%). Younger age was significantly associated with pulmonary artery thrombosis but not with operative death. Oxygenation (arterial PO2, and oxygen saturation) improved significantly and spontaneously over the first 48 hours after BSCC. Younger age had a significant adverse effect on oxygenation in the early postoperative period (first 48 hours). CONCLUSIONS: BSCC can be performed successfully in infants < 6 months old and as young as 5 weeks old. Within this patient population, younger age is not associated with perioperative death but is associated with pulmonary artery thrombosis and postoperative hypoxemia. We suggest that BSCC may be performed any time beyond the neonatal period in symptomatic patients and may be delayed until 4 to 6 months of age if completely elective.

Conversion of the hemi-Fontan procedure to fenestrated total extracardiac cavopulmonary bypass.

BACKGROUND: Arrhythmias occur frequently after Fontan operations, and are related in part to high atrial pressure, wall distention, and scarring caused by extensive suture lines. These arrhythmic factors may be avoided by an extracardiac total cavopulmonary anastomosis. We have embarked on a program of conversion of the hemi-Fontan operation to a fenestrated extracardiac Fontan procedure with a relatively simple operation. METHODS: In a 4-month period ending in December 1994, 4 consecutive patients underwent this procedure. The inferior vena cava was divided and the cardiac end was oversewn. A large (20 to 25 mm) ascending aortic homograft, from which the inlet portion and valve had previously been excised, was interposed between the divided distal end of the inferior vena cava and the hood of the superior cavopulmonary anastomosis. A 4-mm fenestration with a pursestring snare mechanism was placed within the cavoatrial patch that had been implanted at the time of the hemi-Fontan procedure. RESULTS: There were no deaths, and the average length of stay was 12 +/- 4 days (range, 8 to 18 days). In early follow-up, there have been no atrial arrhythmias, and three of the four fenestrations have been documented to be patent. CONCLUSIONS: An extracardiac fenestrated Fontan procedure can safely and successfully be performed after a hemi-Fontan operation, and may have both hemodynamic and arrhythmic benefits.

Intermediate results after complete repair of tetralogy of Fallot in neonates.

From July 1988 through September 1993, 30 neonates with symptomatic tetralogy of Fallot underwent complete repair. Sixteen patients had tetralogy and pulmonary stenosis, 9 had pulmonary atresia, 3 had nonconfluent pulmonary arteries, and 2 had both pulmonary atresia and nonconfluent pulmonary arteries. The median age at operation was 11 days (mean +/- standard error of the mean, 12.6 +/- 2.9 days), with a mean weight of 3.1 +/- 0.1 kg (range 1.5 to 4.4 kg). Preoperatively, 14 patients were receiving an infusion of prostaglandin, 13 were mechanically ventilated, and 6 required inotropic support. Right ventricular outflow tract obstruction was managed by a limited transannular patch in 25 patients, infundibular muscle division with limited resection in 15, and insertion of a right ventricle-pulmonary artery valved aortic homograft conduit in 5 patients. Follow-up was complete at a median interval of 24 months (range 1 to 62 months). There were no hospital deaths and two late deaths, for 1-month, 1-year, and 5-year actuarial survivals of 100%, 93%, and 93%, respectively. The hazard function for death had a rapidly declining single phase that approached zero by 6 months after the operation. Both late deaths occurred in patients with tetralogy of Fallot and pulmonary atresia who had undergone aortic homograft conduit reconstruction, so that the only independent risk factor for death was the use of a valved homograft conduit (p < or = 0.005). Eight patients required reoperation, resulting in 1-month, 1-year, and 5-year freedom from reoperation rates of 100%, 93%, and 66%, respectively. Indications for reoperation were branch left pulmonary artery stenosis in 5 patients, residual right ventricular outflow tract obstruction in 2 patients, and severe pulmonary insufficiency in 1 patient. Independent risk factors for reoperation included an intraoperative pressure ratio between the right and left ventricles of 0.75 or greater (p = 0.01), Doppler residual left pulmonary artery stenosis of 15 mm Hg or more, or Doppler right ventricular outflow tract obstruction gradient of 40 mm Hg or more at hospital discharge (p = 0.002 and 0.02, respectively). This series demonstrates the safety of early hemodynamic repair of symptomatic tetralogy of Fallot in neonates. It also emphasizes the importance of relieving all sources of right ventricular outflow tract obstruction at the initial operation, particularly that located at the site of insertion of the ductus arteriosus, which may be difficult to diagnose in the neonate before ductal closure occurs. The safety and efficacy of valved aortic homograft conduits in neonates requires further investigation.

Relationship of the proinflammatory cytokines to myocardial ischemia and dysfunction after uncomplicated coronary revascularization.

The proinflammatory cytokines have been implicated in mediating myocardial dysfunction associated with myocardial infarction, severe congestive heart failure, and sepsis. We tested the hypothesis that cytokine levels are elevated after uncomplicated coronary artery bypass grafting and associated with episodes of postoperative myocardial ischemia and dysfunction. Coronary artery bypass grafting was performed under general anesthesia with moderate systemic hypothermia and cold-blood potassium cardioplegic solution. Tumor necrosis factor-alpha and interleukin-6 levels were determined by bioassays, and interleukin-8 levels were measured by a sandwich enzyme-linked immunosorbent assay. Myocardial function and ischemic episodes were assessed by intraoperative transesophageal echocardiography and perioperative 12-channel Holter monitoring. A total of 22 patients were studied, with no deaths or complications. Arterial tumor necrosis factor-alpha rose in a bimodal distribution, peaking at 2 and 18 to 24 hours after the operation (at 20.2 +/- 6.4 pg/ml, [mean +/- standard error of the mean]) and 5.8 +/- 1.6 pg/ml, respectively; before cardiopulmonary bypass: 0.90 +/- 0.20 pg/ml, p < 0.001 for both peaks) then progressively declined to levels before bypass. Arterial interleukin-6 was maximally elevated immediately on termination of cardiopulmonary bypass and peaked again 12 to 18 hours after cardiopulmonary bypass (at 7520 +/- 2439 pg/ml and 6216 +/- 1928 pg/ml, respectively; before bypass: 746 +/- 187 pg/ml, p < 0.0001 for both peaks). Arterial interleukin-8 levels were more variable but followed a similar pattern, peaking in the early period after cardiopulmonary bypass and again at 16 to 18 hours after the operation (at 4110 +/- 1403 pg/ml and 1760 +/- 1145 pg/ml, respectively; before bypass: 461 +/- 158, p < 0.05 for both peaks). By multivariate analysis, the aortic crossclamp time was independently predictive of postoperative cytokine levels. Left ventricular wall motion abnormalities were associated with both interleukin-6 and interleukin-8 levels, worsening scores being associated with increasing levels (for interleukin-6, p = 0.003; for interleukin-8, p = 0.05). Postoperative myocardial ischemic episodes were associated with interleukin-6 levels, six of seven (85%) patients with episodes of myocardial ischemia after a peak in interleukin-6 concentrations (p < 0.01). We conclude that proinflammatory cytokines are elevated after uncomplicated coronary revascularization and may contribute to postoperative myocardial ischemia and segmental wall motion abnormalities.

Predictors of postoperative outcome after general surgical procedures in patients with congenital heart disease.

This study was undertaken to evaluate factors predictive of postoperative outcome after general surgical (GS) procedures in patients with congenital heart disease (CHD). All patients with a diagnosis of CHD who underwent a GS procedure under general anesthesia during a consecutive 10-year period were considered eligible for the study The congenital heart defect was classified as either simple (ASD, VSD, PDA) or complex (endocardial cushion defects, transposition of the great vessels, tetralogy of Fallot), and the GS procedure as either major (intraperitoneal, intrathoracic, or vascular reconstructive) or minor (inguinal herniorrhaphy, vascular access). The overall mortality rate for the patient population was 12% (27 deaths among 226 procedures), minor procedures being associated with a 3% mortality rate (2 of 70 procedures), and major procedures with a 16% mortality rate (25 of 156 procedures). Incremental risk factors for mortality included a preoperative American Society of Anesthesiologists' (ASA) physical status class of IV or higher (P = .0003), a preoperative in-hospital stay of 10 or more days (P = .004), birth at a tertiary care center (P = .04), and emergency operations (P = .05). In the subgroup of patients less than 6 months old, weight of less than 2.4 kg at the time of surgery and a 1-minute Apgar score of less than 4 were additional independent risk factors (P = .04 and .01, respectively). By logistic analysis, previous corrective cardiac procedures, whether complete or palliative, did not significantly alter the postoperative outcome. The authors conclude that physiologically well-compensated patients with CHD can undergo elective operations at a low operative risk; however, poorly compensated patients undergoing urgent or emergent operations are at high risk.(ABSTRACT TRUNCATED AT 250 WORDS)

Operative treatment of pediatric obstructive hypertrophic cardiomyopathy: a 26-year experience.

We retrospectively reviewed the 26-year National Institutes of Health experience with operative treatment of obstructive hypertrophic cardiomyopathy in pediatric patients. Operative criteria were either severe obstructive symptoms (New York Heart Association functional class III or IV) or sudden death. Seventeen patients underwent 19 open procedures, of which the present study is comprised. Complete follow-up was available 10.1 +/- 1.4 years (mean +/- standard error; range, 0.8 to 26.2 years) after operation. The mean ages at diagnosis and operation were 11.9 +/- 1.3 years (range, 1 to 17 years) and 14.8 +/- 0.7 years (range, 9 to 17 years), respectively. The preoperative intraventricular septum mean dimension was 23.2 +/- 1.3 mm (range, 11 to 36 mm). The left ventricular outflow tract gradient was 74 +/- 9 mm Hg (range, 20 to 175 mm Hg) at rest and 94 +/- 7 mm Hg (range, 55 to 175 mm Hg) with provocation. Fifteen patients (88%) underwent left ventricular myotomy and myectomy, and 2 underwent mitral valve replacement. Two patients who initially received left ventricular myotomy and myectomy later underwent mitral valve replacement. There were one perioperative death (6%) and five late sudden deaths (31%) at 3.8, 8.7, 9.6, 14.1, and 21 years postoperatively. Kaplan-Meier survival was 86% +/- 8% at 5 years and 77% +/- 12% at 10 years. After operation, the left ventricular outflow tract gradient decreased almost 80% to 21 +/- 15 mm Hg (p = 0.0001). In 8 patients, the left ventricular outflow tract gradient completely resolved.(ABSTRACT TRUNCATED AT 250 WORDS)

Left ventricular function in experimental mitral regurgitation with intact chordae tendineae.

Left ventricular function in mitral regurgitation has typically been studied in models that either sever the chordae tendineae or create a ventriculoatrial shunt. These methods may have adverse effects on left ventricular function independent of the regurgitant lesion. An animal model of chronic mitral regurgitation was therefore developed that both preserves annuloventricular continuity and avoids the use of external shunts. A circular 0.16 to 0.24 mm/kg defect was created in the anterior mitral valve leaflet of weanling sheep under direct vision with the aid of cardiopulmonary bypass. Six animals were studied preoperatively and immediately postoperatively (acute regurgitation group), and 20 animals were studied 8.1 +/- 0.2 (mean +/- one standard deviation) months postoperatively (chronic regurgitation group). Animals with chronic mitral regurgitation were compared with an age- and weight-matched control group that was not operated on (n = 7). Volumetric data and ejection fraction were derived from digitalized cineangiographic images. Maximal elastance was calculated from pressure-volume loops obtained from the simultaneous recording of left ventricular pressure by micromanometer-tipped left ventricular catheters, and volumes were obtained from digitalized images of epicardial echocardiographic recordings. Mitral valve perforation resulted in 3+ to 4+ mitral regurgitation and a calculated regurgitant fraction of 37% +/- 7% (mean +/- one standard deviation). Acute mitral regurgitation was associated with an increase in left ventricular end-diastolic volume from 110 +/- 17 to 121 +/- 23 ml (p < or = 0.05) and no change in end-systolic volume. These changes were associated with an increase in fractional shortening, from 29% +/- 11% to 40% +/- 10% (p < or = 0.05), and an increase in velocity of circumferential shortening, from 1.5 +/- 0.7 to 2.9 +/- 0.7 circ/sec (p < or = 0.05). However, there was no change in maximal elastance, a load-independent index of left ventricular function. Conversely, animals with chronic mitral regurgitation exhibited an elevated end-diastolic volume (202 +/- 32 versus 145 +/- 34, p < or = 0.05), an elevated end-systolic volume (104 +/- 17 versus 63 +/- 20 ml, p < or = 0.05), and a reduced ejection fraction (48% +/- 6% versus 57% +/- 9%, p < or = 0.05) compared with controls.(ABSTRACT TRUNCATED AT 400 WORDS)

Aortic revascularization of orthotopic liver allografts: indications and long-term follow-up.

BACKGROUND: Occasionally because of certain conditions that would imperil arterial supply to a hepatic graft, it is necessary to base the arterial supply on the aorta. METHODS: Twenty orthotopic liver transplants (OLTs) in 19 patients were performed with arterial revascularization based on the aorta (Ao-OLT). In two patients the donor celiac axis was anastomosed directly to the aorta and in 18 to a conduit anastomosed to either the supraceliac (n = 10) or infrarenal (n = 8) aorta. RESULTS: One thrombosis occurred 2 months after the placement of a supraceliac conduit in an adult patients, accounting for a cumulative 1-year hepatic artery patency rate of 91.7% +/- 8.0% for 16 grafts placed in 15 adults and 100% for four grafts placed in four children. In comparison, hepatic artery-based transplantation was associated with a 1-year patency rate of 92.6% +/- 1.9% for 245 adults and 94.7% +/- 5.1% for 19 children (difference not significant compared with Ao-OLT). By logistic regression analysis, the only factor independently associated with hepatic artery thrombosis was retransplantation (1-year hepatic artery patency rate 85.7% +/- 5.9% [n = 38] for retransplants vs 93.9% +/- 1.7% for primary transplantation [n = 246]; p < 0.05). For retransplantation, Ao-OLT revascularizations were superior to those based on the hepatic artery (1-year patency rate 100% [n = 11] vs 79.9% +/- 8.1% [n = 27]; p < 0.05). CONCLUSIONS: Indications for Ao-OLT include poor hepatic arterial inflow, small or anomalous recipient hepatic arteries, friable or attenuated native hepatic arteries as may be present during retransplantation, and recipient age less than 1 year, especially those less than 15 kg or in whom the recipient artery is less than 3 mm in diameter. In these more precarious situations, Ao-OLT achieves patency rates similar to those of primary, uncomplicated OLT, is superior for retransplantation, and has technical advantages in the small pediatric liver recipient.

Lower extremity arterial disease in systemic lupus erythematosus.

Lupus vasculitis primarily affects microvascular circulation, and large-vessel thrombosis is a rare complication of this disease. Large-vessel occlusive disease in systemic lupus erythematosus is most likely related to hypercoagulability in addition to immune complex-mediated endothelial damage. We describe the 11th and 12th patients reported to have systemic lupus erythematosus and macrovascular occlusive disease of the lower extremities. Our experience and a review of the literature suggest that, while aortoiliac disease is amenable to bypass or endarterectomy, infrainguinal disease is rarely correctable surgically, and amputation becomes necessary in most of these patients.

The results of operation in patients with hypertrophic cardiomyopathy and pulmonary hypertension.

The clinical course and hemodynamic results in patients undergoing operation for obstructive hypertrophic cardiomyopathy with preoperative pulmonary arterial hypertension were unknown. The hypothesis tested in this retrospective study was that operative relief of left ventricular outflow tract obstruction resulted in a substantial reduction in pulmonary artery pressures and mitral regurgitation without necessitating mitral valve replacement. Patients were included if their preoperative pulmonary systolic pressure was greater than 35 mm Hg and if they were without concomitant cardiac disease, with the exception of mitral regurgitation. Since 1962, 49 patients who fit our criteria underwent left ventricular myotomy and myectomy with 98% follow-up. Mean follow-up was 7.9 +/- 0.7 (mean +/- standard error of the mean) years with a range of 0.8 to 18.4 years. Early hospital mortality rate was 12% (n = 6); two deaths from low cardiac output and four from arrhythmia. There were 43 (88%) hospital survivors and 18 late deaths. Actuarial survival rate after operation was 87% +/- 5% (n = 31) at 5 years and 55% +/- 8% (n = 9) at 10 years. Thirty-nine of 43 survivors (91%) returned 9 +/- 1 months postoperatively for follow-up evaluation including cardiac catheterization. The majority (83%) were in New York Heart Association functional class I or II postoperatively. Cardiac catheterizations indicated a fall in pulmonary arterial systolic pressure from 62 +/- 3 (range = 36 to 105) to 38 +/- 2 (range = 21 to 65) mm Hg (p = 0.0001) with no difference in right atrial pressure or cardiac output. Pulmonary arterial wedge mean pressure decreased from 24 +/- 1 to 16 +/- 5 mm Hg (p = 0.0002) and preoperative mitral regurgitation improved or was abolished in 85% of patients studied (n = 13). Rest and maximal provocable left ventricular outflow tract gradients decreased from 81 +/- 7 and 103 +/- 5 to 14 +/- 3 and 45 +/- 8 mm Hg, respectively (p = 0.0001). Comparison of the above-mentioned patients, operated on since 1982, with a preoperatively matched group who underwent mitral valve replacement in the same interval showed no statistically significant difference in mortality, morbidity, hemodynamic outcome, or functional outcome with a mean follow-up of 2 years. We conclude that a consistent, significant reduction (mean = 40%) in preoperative pulmonary arterial systolic pressure, clinical symptoms, and mitral regurgitation occurs with relief of outflow tract obstruction by left ventricular myotomy and myectomy and that pulmonary hypertension and mitral regurgitation are not indications for mitral valve replacement in these patients.

Prostaglandin E2 inhibits in vitro and in vivo lymphocyte responses in allogeneic transplantation.

Prostaglandin E2 (PGE2) has been shown to a clear role in the suppression of immune responses after burn and trauma injury. This probably results from inhibition of interleukin-2 production. This study examined the effects of PGE2 in vivo on the survival of solid-organ allografts and in vitro on the rat allogeneic mixed lymphocyte response. Administration of 16,16-dimethyl prostaglandin E2 (DMPGE2), a stable analogue of PGE2, significantly prolonged the survival of heterotopic cardiac allografts from ACI to LBN rats: 10.4 +/- 1.7 days versus 5.7 +/- 1.1 days (mean +/- standard error of the mean) (p less than or equal to 0.001). In 1 animal, DMPGE2 apparently led to the induction of long-term tolerance. Mixed lymphocyte cultures using splenocytes from naive LBN and ACI rats to which DMPGE2 was added showed a dose-dependent suppression of the mixed lymphocyte response with concentrations as low as 1 x 10(-7) mol/L. Splenocytes harvested from treated animals with functioning but histologically rejecting hearts demonstrated a marked decrease in mixed lymphocyte response to donor (ACI) stimulators compared with naive LBN controls (3,804 +/- 603 versus 27,395 +/- 2,668 cpm, n = 4), but maintained a normal response to third-party (Wistar Furth) stimulators. We conclude that DMPGE2 suppressed solid-organ allograft rejection, inhibited the allogeneic mixed lymphocyte response, and induced donor-specific in vitro hyporesponsiveness.

Comparative assessment of chordal preservation versus chordal resection during mitral valve replacement.

Left ventricular function often deteriorates after mitral valve replacement for mitral regurgitation. It has been postulated that disruption of the mitral valve apparatus at operation is a major mechanism of postoperative dysfunction. The hypothesis tested in this investigation was that chordal preservation results in more favorable left ventricular function. Sixty-nine patients with isolated mitral regurgitation who underwent mitral valve replacement were studied before and 6 months after operation by treadmill exercise testing, catheterization, echocardiography, and radionuclide angiography. Nine patients underwent mitral valve replacement with preservation of the entire mitral apparatus and five with preservation of the posterior leaflet and attached chordae. The remaining 55 had mitral valve replacement with complete excision of the native valve. Preoperatively, there were no differences among groups in age, gender, exercise capacity, cardiac index, rest or exercise ejection fraction, fractional shortening, or pulmonary artery pressures. There were four perioperative deaths (7%) and eight late deaths among the 55 patients with chordal resection but no early or late deaths of patients whose chordae were preserved (p = 0.05). In patients in whom the chordae were excised, exercise capacity, left ventricular systolic dimensions, and cardiac index did not improve after mitral valve replacement, and left ventricular function deteriorated, as evidenced by a reduction of both the resting and exercise ejection fractions (from 46% +/- 13% to 31% +/- 13%, p = 0.0001, and from 49% +/- 12% to 37% +/- 14%, p = 0.0007, respectively) and fractional shortening (from 34% +/- 10% to 26% +/- 14%, p = 0.0001). In contrast, exercise capacity improved after mitral valve replacement in patients in whom the entire apparatus was spared (by 4 +/- 3 minutes, p = 0.05), left ventricular systolic dimensions decreased (from 44 +/- 8 to 36 +/- 9 mm, p = 0.03), and left ventricular function was maintained or improved, as evidenced by preservation of the resting ejection fraction (preoperative, 50% +/- 14%; postoperative, 54% +/- 11%; p = no significant difference), exercise ejection fraction (46% +/- 16% versus 52% +/- 9%, p = no significant difference), fractional shortening (from 31% +/- 9% to 28% +/- 9%, p = no significant difference), and an increase in the cardiac index (from 2.0 +/- 0.3 to 2.7 +/- 0.5 L/min/m2, p = 0.05). No statistically significant differences between posterior chordal resection only and preservation of the entire apparatus were found.(ABSTRACT TRUNCATED AT 250 WORDS)