Elevated serum lactate correlates with intracranial hemorrhage in neonates treated with extracorporeal life support

OBJECTIVES

To correlate the initial and maximal lactate levels with the occurrence of intracranial hemorrhage (ICH) and survival in patients treated with extracorporeal life support (ECLS).

DESIGN

Retrospective chart review.

SETTING

Pediatric intensive care unit.

PATIENTS

Eighty-two neonatal patients placed on ECLS for respiratory failure due to sepsis, meconium aspiration, or persistent pulmonary hypertension of the newborn.

MEASUREMENTS

The initial lactate level measured within 6 hours of initiating ECLS and the maximal lactate level measured throughout the ECLS course were collected. Lactate levels were described as mean lactate +/- SE (mM). Head ultrasound reports and survival were reviewed. Platelet counts and activated clotting times (ACTs) were examined.

RESULTS

The mean initial and maximal lactate levels were higher in ECLS patients who developed ICH (initial: 10 +/- 1.7 mM vs 6.4 +/- 0.8 mM, p = .05 and maximal: 12.4 +/- 2.5 mM vs 7.9 +/- 0.8 mM, p = .04). Initial and maximal lactate levels were also elevated in nonsurvivors (initial: 11.7 +/- 3 mM vs 6.4 +/- 0.7 mM, p = .01 and maximal: 14.8 +/- 3.3 mM vs 7.8 +/- 0.8 mM, P < .01). Platelet counts and ACT did not differ in patients with and without ICH.

CONCLUSIONS

Lactate is a useful marker for the development of ICH in ECLS patients. In addition, elevated lactates during ECLS identify a subgroup of patients with poor outcome. Prospective studies are needed to determine whether the incorporation of this information into pre-ECLS and ECLS management will decrease the occurrence of ICH and improve survival.

Comparing two strategies of cardiopulmonary bypass cooling on jugular venous oxygen saturation in neonates and infants

BACKGROUND

Cerebral protection during deep hypothermic circulatory arrest is predicted on efficient and complete cerebral cooling. Institutions approach cooling quite differently. We compared two different cooling strategies in terms of measured jugular venous bulb saturations in 39 infants undergoing deep hypothermic cardiopulmonary bypass to evaluate the effect of institutional cooling practices on jugular venous bulb saturation, an indirect measure of cerebral cooling efficiency.

METHODS

The patients were grouped based on the method of core cooling. In group A (n = 17), core cooling was achieved rapidly by setting the water bath temperature of the heat exchanger at 4 degrees to 5 degrees C, and the patient was cooled until rectal temperature and nasopharyngeal temperature were 15 degrees C or lower. In group B (n = 22), the heat exchanger was initially set at 18 degrees C and slowly lowered to 12 degrees C. Hypothermic temperatures of 12 degrees C were maintained until the nasopharyngeal temperature was 18 degrees C or less and the rectal temperature was 20 degrees C or lower. Once cooling was complete, blood samples were analyzed by cooximetry for determination of arterial oxygen saturation and jugular venous bulb saturation.

RESULTS

In group A, the measured jugular venous bulb saturation was 98.0% +/- 0.9% and the oxygen saturation to jugular venous bulb saturation difference was 0.3% +/- 0.5%, measured at the time that institutional cooling objectives were achieved (total cooling time, 15.0 +/- 0.45 minutes). In group B, jugular venous bulb saturation was 86.2% +/- 12% and the oxygen saturation to jugular venous bulb saturation difference was 10.8% +/- 12.2%, measured at the time that institutional cooling objectives were achieved (total cooling time, 17.5 +/- 1.1 minutes (p < 0.01).

CONCLUSIONS

Differences in cardiopulmonary bypass cooling techniques may alter the rate at which jugular bulb saturations rise. We believe this represents an indirect measure of the efficiency of brain cooling and therefore of cerebral protection.

Jugular ligation does not increase intracranial pressure but does increase bihemispheric cerebral blood flow and metabolism

OBJECTIVES

To answer the following questions: a) Does jugular venous ligation (simulating venovenous extracorporeal life support) alter proximal jugular venous pressure, intracranial pressure, hemispheric cerebral blood flow, or cerebral metabolism? b) Does release of ligation reverse these effects? and c) What are the comparative effects of venous ligation alone vs. venous ligation in combination with arterial ligation?

DESIGN

Prospective, randomized, laboratory investigation.

SETTING

Multidisciplinary laboratory setting.

SUBJECTS

Sixteen swine, weighing 8.1 to 12.1 kg, 3 to 4 wks of age.

INTERVENTIONS

Sixteen swine were randomly assigned to two groups, utilizing a random sequence of vessel ligation. Nine swine underwent occlusion of the right internal and external jugular veins alone (venovenous ligation) followed by release of the occlusion and then occlusion of the right common carotid artery and the right internal and external jugular veins together (venoarterial ligation). The remaining seven swine underwent venoarterial ligation, followed by release of the occlusion and then venovenous ligation. In the experimental group in which venovenous ligation was performed first, the 5, and 30-min release periods after ligation were taken to represent the effects of draining the right jugular vein during venovenous extracorporeal life support.

MEASUREMENTS AND MAIN RESULTS

Data were obtained at baseline, 5, and 30 mins after each ligation/release period. Intracranial pressure, right and left internal jugular pressures/flow rates, and cerebral sinus lactate concentrations were measured. Cerebral blood flow was determined using 133Xe clearance methodology, and the cerebral metabolic rate was calculated. There were no significant differences between the ipsilateral internal jugular pressure or extracorporeal life support at 5 or 30 mins after venovenous or venoarterial ligation compared with baseline values or compared with the release of the ligation at 5 or 30 mins. There was a significant increase in right-side (44.7 +/- 2.0 vs. 38.8 +/- 2.4 mL/kg/min; p < .05) and left-side (42.9 +/- 2.3 vs. 38.7 +/- 1.9 mL/kg/min; p < .05) cerebral blood flow 5 mins after venovenous ligation when compared with baseline values. Similarly, after venoarterial ligation, there was a significant increase in right-side (44.6 +/- 2.2 vs. 38.8 +/- 2.4 mL/kg/min; p < .05) and left-side (43.9 +/- 1.5 vs. 38.7 +/- 1.9 mL/kg/min; p < .05) and cerebral blood flow. Cerebral oxygen consumption was significantly increased after venovenous (2.7 +/- 0.2 to 3.2 +/- 0.2 mL/kg/min; p < .05) and venoarterial (2.7 +/- 0.2 to 3.1 +/- 0.2 mL/kg/min; p < .05) ligation at 5 mins after ligation. This increase persisted at the 30-min period and after release of ligation.

CONCLUSIONS

Ligation of the right jugular veins alone (venovenous ligation) or jugular veins and right carotid artery (venoarterial ligation) does not increase jugular venous pressures or intracranial pressure. However, this procedure does increase cerebral blood flow and cerebral oxygen consumption. These findings demonstrate that there is adequate decompression of the venous system by the cerebrovascular system and retrograde decompression during extracorporeal life support appears unwarranted.

Normalization of priming solution ionized calcium concentration improves hemodynamic stability of neonates receiving venovenous ECMO

The authors' objectives in this investigation were: 1) to prospectively determine whether a normocalcemic priming solution would result in elimination of hypocalcemia after the initiation of extracorporeal membrane oxygenation (ECMO); 2) to investigate whether normocalcemia would result in improvements in the patient's hemodynamics during the initiation of ECMO; and 3) to further define the relationship between ionized calcium measurements and total calcium, serum total protein, serum albumin, and total magnesium. This was a prospective study done in our neonatal intensive care unit, and included nine neonatal patients placed on ECMO for cardiopulmonary support. The bypass circuit was primed in the standardized manner with 100 mg calcium chloride. Circuit ionized calcium measurements were performed, and additional calcium chloride was added to normalize the ionized calcium in the priming solution. Ionized calcium was measured from the circuit and the patient before the initiation of ECMO, and then again from the patient 5, 10, 15, 30, 60, 120, and 240 minutes after the initiation of ECMO. The patients' mean arterial pressure was measured simultaneously with each ionized calcium measurement. Ionized calcium, serum total calcium, total protein, serum albumin, and total magnesium were measured from blood samples simultaneously collected four times daily. There was no significant change in the ionized calcium measured in the patients after the initiation of ECMO. There was, however, a significant increase in blood pressure 5 min after the initiation of ECMO (62 +/- 7 mmHg vs 53 +/- 6 mmHg, p = 0.01). Thereafter, there was no difference in blood pressure measured when compared with pre ECMO values. A poor correlation was demonstrated between ionized calcium and total calcium (r2 = 0.35), serum total protein (r2 = 0.26), serum albumin (r2 = 0.27), and total magnesium (r2 = 0.05). On the basis of the authors' data, the initiation of ECMO with a normocalcemic prime results in a minimal change in patient ionized calcium and resolution of the hypotension previously observed. In addition, there was poor correction between ionized calcium, total calcium, and other indirect measures of ionized calcium. Ionized calcium measurements are critical for patient hemodynamic stability before bypass and should be normalized in both the patient and priming solution before the initiation of bypass.

In acute lung injury, inhaled nitric oxide improves ventilation-perfusion matching, pulmonary vascular mechanics, and transpulmonary vascular efficiency

Acute respiratory distress syndrome continues to be associated with significant morbidity and mortality related to ventilation-perfusion mismatch, pulmonary hypertension, and right ventricular failure. It has been suggested that inhaled nitric oxide, which is a selective pulmonary vasodilator, may be effective in the treatment of acute respiratory distress syndrome; however, the effects of nitric oxide on cardiopulmonary interactions are poorly understood. We therefore developed a model of acute lung injury that mimics the clinical syndrome of acute respiratory distress syndrome. In our model, inhaled nitric oxide significantly reduced pulmonary artery pressure, pulmonary vascular resistance, and pulmonary vascular impedance. In addition, inhaled nitric oxide improved transpulmonary vascular efficiency and ventilation-perfusion matching, which resulted in increased arterial oxygen tension. Although arterial oxygen tension increased, oxygen delivery did not improve significantly. These data suggest that by improving ventilation-perfusion matching and arterial oxygen tension while lowering pulmonary vascular resistance and impedance, nitric oxide may be beneficial in patients with acute respiratory distress syndrome. However, additional measures to enhance cardiac performance may be required.

Blockade of endothelin-converting enzyme reduces pulmonary hypertension after cardiopulmonary bypass and circulatory arrest

BACKGROUND

Pulmonary dysfunction associated with elevated pulmonary vascular resistance is a significant problem after cardiopulmonary bypass (CPB) and circulatory arrest. Mediators of the pulmonary hypertensive response to CPB have not been fully elucidated. The purpose of this study was to examine the contribution of the endothelium-derived vasoconstrictor endothelin-1 to postbypass pulmonary hypertension.

METHODS

Twelve 1-month-old piglets were instrumented with left atrial and pulmonary artery (PA) micromanometers and a PA flow probe. Phosphoramidon (Phos, n = 6) pigs received a 30 mg/kg bolus of Phos, an endothelin converting enzyme inhibitor. Controls (n = 6) received saline solution. All animals were placed on CPB and underwent a 60-minute period of circulatory arrest. The indexed pulmonary vascular resistance (PVRI) was calculated at baseline for controls, both before and 10 minutes after drug infusion in the Phos group, and 15 minutes after separation from CPB in both groups.

RESULTS

Pre-CPB, mean PA pressure, and PVRI were not different between the control and Phos groups (14.6 +/- 1.1 versus 14.5 +/- 1.1 mm Hg and 7322 +/- 1269 versus 7260 +/- 947 dyne/sec/kg/cm-5, respectively). After CPB mean PA pressure was significantly higher in control than Phos animals (32.1 +/- 1.1 versus 22.5 +/- 1.3 mm Hg, p = 0.0003). PVRI was also significantly higher in the controls (30896 +/- 4714 versus 14972 +/- 1710, dyne/sec/kg/cm-5, p = 0.02).

CONCLUSIONS

Production of endothelin-1 during CPB and circulatory arrest is a mediator of postbypass pulmonary hypertension.

Superior vena cava obstruction after extracorporeal membrane oxygenation

Of 60 neonates who survived extracorporeal membrane oxygenation (ECMO) in our institution between June 1992 and March 1994, seven had either complete or partial superior vena cava (SVC) obstruction. When the patients with SVC obstruction were compared with those who had an echocardiogram after ECMO, no predisposing factors for the development of SVC thrombus could be found. Our data show that SVC thrombus may be a significant complication after ECMO.

Alveolar-arterial oxygen gradients before extracorporeal life support for severe pediatric respiratory failure: improved outcome for extracorporeal life support-managed patients?

OBJECTIVE

Recent reports have described the usefulness of the alveolar-arterial oxygen tension difference (P[A-a]O2) in predicting mortality in children with acute respiratory failure managed with mechanical ventilation. We reviewed our experience with extracorporeal life support for acute pediatric respiratory failure and specifically examined P(A-a)O2 measurements during the 24 hrs before extracorporeal life support to determine if defined cutoffs established with conventional mechanical ventilation were applicable to extracorporeal life-support survival.

DESIGN

Retrospective, case-series chart review.

SETTING

A university tertiary medical center.

PATIENTS

Infants and children (n = 36), one month to 18 yrs of age, with severe life-threatening respiratory failure who were believed to have failed conventional mechanical ventilatory support.

INTERVENTIONS

Veno-venous or veno-arterial extracorporeal life support.

MEASUREMENTS AND MAIN RESULTS

From 1982 to 1992, we managed 36 pediatric patients with severe respiratory failure using extracorporeal life support. We identified 28 patients who had P(A-a)O2 values of > 400 torr (> 53.3 kPa) for the 24-hr time period before placement on bypass. At the time of bypass initiation, all blood gas and mechanical ventilator parameters except PaCO2 showed trends of worsening pulmonary function, compared with measurements done 24 hrs before bypass initiation. Oxygenation-related variables showed statistically significant worsening trends when measured 24 hrs before bypass, compared with the time of bypass: P(A-a)O2 539 vs. 582 torr (71.9 vs. 77.6 kPa), p < .01; PaO2/FIO2 ratio 70 vs. 57 torr (9.3 vs. 7.6 kPa), p < .05; oxygenation index 32 vs. 47 cm H2O/torr, p < .01; and FIO2 0.94 vs. 0.98, p < .05. Sixty-one percent of extracorporeal life support-managed patients (17 of 28) survived their life-threatening respiratory illness to be discharged home.

CONCLUSIONS

Based on previous reports of the utility of P(A-a)O2 measurements to predict mortality, our preliminary evidence suggests that extracorporeal life support results in 62% survival for pediatric respiratory failure patients predicted to have no chance of survival using conventional mechanical ventilation. Prospective, randomized trials of children with severe acute respiratory failure managed with mechanical ventilation vs. extracorporeal life support may be indicated.

Effect of deep hypothermia and circulatory arrest on cerebral blood flow and metabolism

The primary goal of monitoring cerebral blood flow and metabolism is to improve our understanding of the association with cardiopulmonary bypass and deep hypothermic circulatory arrest so that effective brain protection strategies can be developed and employed. A review of our cerebral blood flow/cardiopulmonary bypass database, presently totaling 275 neonates and infants, for the purposes of this publication, reveals certain trends and some conclusions that can be drawn. Deep hypothermic circulatory arrest continues to be a factor in the delayed recovery of cerebral blood flow and metabolism in these patients. Examining flow and metabolism serially in the postoperative period shows that in the majority of patients, flow, metabolism and autoregulation return to normal within 24 hours after operation. Some patients' cerebral oxygen metabolism is unable to exert a protective response of increasing extraction in the setting of low cerebral blood flow. We have also observed that in the setting of low cardiac output after cardiac repair, cerebral blood flow is low. It is therefore likely that low cardiac output and pressure-passive cerebral blood flow potentiate brain ischemia after cardiopulmonary bypass and operation in some patients. We have also examined in our series of 275 patients selective neuroprotection strategies for their potential for improving recovery of cerebral blood flow and cerebral metabolism. Duration of cooling on cardiopulmonary bypass correlates directly with suppression of metabolism due to hypothermia. Low-flow cardiopulmonary bypass instead of deep hypothermic circulatory arrest, and topical brain cooling with ice during deep hypothermic circulatory arrest, improve cerebral blood flow and cerebral metabolic recovery.

In vitro evaluation of volumetric flow from Doppler power-weighted and amplitude-weighted mean velocities

Ultrasound theory suggests that the volume of flow is directly related to the power and amplitude of the backscattered Doppler signals. To evaluate the accuracy of volume flow calculated with power-weighted and amplitude-weighted mean velocities (PWMV and AWMV), volume flows were measured in a pulsatile flow-tank system equipped with a 1.25 cm diameter simulated femoral artery. Analyses were performed throughout a range of physiologic flows, mean driving pressures, and pulse rates. At each hemodynamic setting, volume flow in the simulated artery was measured with an electromagnetic flow probe and with pulsed Doppler echocardiography by use of 7.0 and 3.5 MHz transducers. In addition, to determine the effects of vessel size and parabolic flow on the accuracy of the Doppler volumes, volume flow was evaluated in several differently sized vessels at sampling distances of 20 times the vessel diameter downstream from the orifice. On the ultrasound system, PWMV was calculated as the sum of the individual velocities multiplied by their respective power fractions (the fraction of the total instantaneous power represented by the individual signal power). The instantaneous PWMV was plotted continuously in time and superimposed on the spectral recording. Similarly, AWMV was calculated with amplitudes measured as the square root of the signal power. The PWMV and AWMV were integrated over the flow period and multiplied by the known cross-sectional area of flow to obtain the Doppler volume. In all analyses performed, volumetric flows calculated with Doppler echocardiography with PWMV and AWMV correlated extremely well with those measured with the electromagnetic flow probe. Thus, over a wide range of physiologic conditions, transducers frequencies, and vessel sizes, volume flow can be accurately calculated from PWMV and AWMV Doppler data. This technique provides an accurate, automatic method for on-line determination of volumetric flow.

Doppler evaluation of normalized peak filling rate in normal children and children with left ventricular outflow obstruction

To evaluate the early diastolic peak filling rate of the left ventricle, three groups of children (normal children, patients with aortic valvular stenosis, and patients with aortic coarctation) were examined with the peak filling rate normalized to stroke volume calculated from the mitral valve inflow Doppler recording as the peak E velocity divided by the velocity time integral. The normal value for this index in children was 6.78 +/- 0.99 SV/sec and did not vary with age, weight, body surface area, or heart rate. Compared with normal subjects, both patients with aortic stenosis and patients with coarctation had increased left ventricular mass, but patients with aortic stenosis had decreased normalized peak filling rates (5.3 +/- 0.84 SV/sec, p < 0.01), while patients with coarctation had normal rates (6.79 +/- 0.98 SV/sec, p = 0.97). Compared with patients with aortic coarctation, patients with aortic stenosis had higher Doppler gradients. Thus the Doppler index of peak filling rate normalized to stroke volume is particularly useful in children because it is independent of heart rate, age, weight, and body surface area. Patients with coarctation may have normal peak filling rates normalized to stroke volume despite increased left ventricular mass because of milder obstruction or better coronary artery perfusion compared with that of patients with aortic stenosis.

High-frequency jet ventilation for respiratory failure after congenital heart surgery

BACKGROUND

Extracorporeal membrane oxygenation (ECMO) is considered when respiratory failure (RF) persists despite increasing conventional mechanical ventilation (CMV). High-frequency jet ventilation (HFJV) can improve ventilation with comparable mean airway pressure (PAW) to that found on CMV. This study was undertaken to determine whether HFJV is an effective treatment and alternative to ECMO for RF after congenital heart surgery.

METHODS AND RESULTS

HFJV was instituted in nine patients ranging in age from 6 days to 3.3 years with congenital heart disease meeting pulmonary criteria for ECMO. Indications for HFJV were pulmonary hypertension (six), adult-type respiratory distress syndrome (two), and pneumonitis (one). Seven patients (77%) were placed on HFJV within 24 hours of operation, and two patients required HFJV 2 weeks after operation. HFJV resulted in resolution of RF in eight of nine patients (89%). After 1 hour of HFJV, the arterial pH increased from 7.40 +/- 0.1 to 7.56 +/- 0.1 (p < 0.05) and the PaCO2 decreased from 44 +/- 15 to 29 +/- 12 mm Hg (p < 0.05). During HFJV there was no change in PaO2, although the FIO2 decreased from 0.99 +/- 0.0 to 0.73 +/- 0.2 (p < 0.05). There was no change in PAW, peak inspiratory pressures, positive end-expiratory pressures, heart rate, or mean arterial blood pressure during HFJV when compared with CMV. Mean duration of HFJV was 43 hours. Four patients were extubated and discharged from the hospital. Two patients were extubated but died from sepsis. Two patients had resolution of RF, but one died at reoperation and one from multisystem organ failure. The patient who failed HFJV therapy was placed on ECMO and died.

CONCLUSIONS

This study suggests that HFJV improves ventilation and is an alternative to ECMO in patients with RF after surgery for congenital heart disease.

Respiratory syncytial virus morbidity and mortality estimates in congenital heart disease patients: a recent experience

OBJECTIVE

To determine recent morbidity and mortality rates from respiratory syncytial virus infection in a pediatric congenital heart disease population.

DESIGN

Retrospective cohort study design.

SETTING

The C. S. Mott Children's Hospital, University of Michigan Medical Center.

PATIENTS

A total of 740 pediatric patients hospitalized at the University of Michigan Medical Center for symptomatic respiratory syncytial virus infection, of whom, 79 patients had clinically important congenital heart disease.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

We retrospectively examined the charts of 740 patients hospitalized at our children's hospital from July 1, 1983 to June 30, 1990 with symptomatic respiratory syncytial virus infection to assess morbidity and mortality outcomes. Seventy-nine patients had congenital heart disease and 40 of these patients had pulmonary hypertension. For the entire cohort and a subset of patients with community-acquired infection, those patients with congenital heart disease had longer durations of hospitalization and greater need for, and days of, both intensive care and mechanical ventilation than patients without congenital heart disease. Mortality risk for respiratory syncytial virus community-acquired infection was not different for congenital heart disease vs. noncongenital heart disease patients (0.0% vs. 0.2%; p = 1.00). When examining only patients with congenital heart disease, those patients with pulmonary hypertension had increased hospital days and greater intensive care and mechanical ventilation durations compared with patients without this diagnosis. The overall mortality rate was low and was equally low for congenital heart disease groups with or without pulmonary hypertension (2.5 vs. 2.6). For community-acquired illness, no mortality was found in either congenital heart disease group. When the cohort of congenital heart disease patients was divided into pre- and postribavirin administration eras, no differences in mean hospital duration, ICU days, and mechanical ventilation days were noted. Of the 79 congenital heart disease patients, only two died during their hospitalization in which respiratory syncytial virus infection occurred. Both patients had nosocomial-acquired respiratory syncytial virus and both were from the postribavirin administration cohort. One of these two patients had received antiviral therapy. Neither death was secondary to respiratory syncytial virus respiratory failure (based on pathologic examination).

CONCLUSIONS

We conclude that respiratory syncytial virus mortality risk in pediatric patients with congenital heart disease is less than the risk reported a decade ago. Respiratory syncytial virus infection in congenital heart disease patients with pulmonary hypertension is associated with increased morbidity but not increased mortality rates. The markedly decreased respiratory syncytial virus mortality risk in patients with congenital heart disease currently experienced is likely secondary to improvements in intensive care management and advances in the surgical correction in this population rather than antiviral therapy.

Use of extracorporeal life support in patients with congenital heart disease

OBJECTIVES

To review a large experience with extracorporeal life support in patients with congenital heart disease. To determine the major causes of mortality and morbidity in order to improve the results of using this technology in this patient population.

DESIGN

Retrospective chart review.

PATIENTS

Twenty-five patients between the ages of 1 day and 8 yrs. These patients had congenital heart disease and were clinically felt to be at high risk for death caused by cardiac failure or by respiratory failure complicated by congenital heart disease.

INTERVENTIONS

All patients in this report were placed on extracorporeal life support to allow recovery of myocardial or pulmonary function.

MEASUREMENTS AND MAIN RESULTS

Of these 25 patients, 52% were weaned from bypass support and 40% survived to discharge. Patients who were not weaned from extracorporeal life support characteristically suffered from irreversible neurologic injury, multiple organ failure, or bleeding complications. Only one patient died of irreversible cardiac failure.

CONCLUSIONS

Extracorporeal life support can be useful in supporting patients with congenital heart disease with life-threatening cardiac or pulmonary failure. Improvements in limiting neurologic and bleeding complications may lead to improvements in the use of extracorporeal life support for this indication. However, prospective, randomized studies are needed to appreciate the role of extracorporeal life support in these patients.

Extracorporeal life support for pediatric respiratory failure

OBJECTIVES

The purposes of this report are to a) describe the University of Michigan experience with venoarterial or venovenous extracorporeal life support for severe pediatric pulmonary rescue therapy, and b) examine survivors and nonsurvivors for differences that might be useful for examination in future, prospective studies.

DESIGN

Case series report. Phase I study of safety and effectiveness of extracorporeal life support for pediatric respiratory failure.

SETTING

University of Michigan Medical Center.

PATIENTS

Non-neonatal pediatric patients treated with extracorporeal life support for severe respiratory failure at the University of Michigan.

INTERVENTIONS

Extracorporeal life support for pulmonary failure.

MEASUREMENTS AND MAIN RESULTS

From November 1982 until May 1991, 25 pediatric patients underwent extracorporeal life support for severe pulmonary failure. Twenty patients were treated in the last 36 months. Sixty percent (15/25 patients) survived their life-threatening respiratory illness, were weaned from mechanical ventilation, and were discharged home. The mean patient age was 4.1 yrs, and mechanical ventilation duration before extracorporeal life support was 5.9 days. Mean blood gas data and mechanical ventilation pressures before extracorporeal life support were: peak inspiratory pressure of 48.6 cm H2O, mean airway pressure of 21.9 cm H2O, positive end-expiratory pressure of 9.7 cm H2O, PaCO2 of 43 torr (5.7 kPa), PaO2 of 69 torr (9.1 kPa), estimated alveolar-arterial oxygen gradient of 563 torr (75 kPa), and FIO2 of 0.98. Variables associated with survival included: age of survivors vs. nonsurvivors, 2.1 vs. 7.1 yrs (p less than .02); peak inspiratory pressure of survivors vs. nonsurvivors, 43.1 vs. 57.9 cm H2O (p less than .03); mean airway pressure of survivors vs. nonsurvivors, 18.4 vs. 27.2 cm H2O (p less than .03); and positive end-expiratory pressure of survivors vs. nonsurvivors, 8.1 vs. 12.1 cm H2O (p less than .01). There were no differences detectable in the blood gas values (PaO2, PaCO2, P[A-a]O2) in survivors and nonsurvivors before extracorporeal life support. The number of days mechanical ventilation was used before extracorporeal life support in survivors and in nonsurvivors was similar.

CONCLUSION

Extracorporeal life support is an effective rescue therapy for pediatric patients with severe respiratory failure (University of Michigan survival rate of 60%).

Unsuspected congenital heart disease in neonates receiving extracorporeal life support: a review of ninety-five cases from the Extracorporeal Life Support Organization Registry

The purpose of this study was to determine the frequency of patients with congenital heart disease who were given extracorporeal life support (ECLS) for respiratory failure. Underlying congenital heart disease "masked" by respiratory failure occurred in 2%. The most frequent pre-ECLS diagnosis that "masked" congenital heart disease was persistent fetal circulation. Of neonates with a pre-ECLS diagnosis of persistent fetal circulation, congenital heart disease was found in 56 (9%) of 623 patients.

Use of balloon-expandable stents to treat experimental peripheral pulmonary artery and superior vena caval stenosis: preliminary experience

Current therapy of congenital or acquired stenoses of the peripheral pulmonary arteries and superior vena cava are frequently ineffective. This report describes our initial experience with the use of a balloon-expandable stainless steel stent to treat experimentally created branch pulmonary artery and superior vena cava stenosis. Fifteen adult mongrel dogs had surgically created stenoses of either a branch pulmonary artery and/or superior vena cava. A balloon-expandable stainless steel (0.076 mm), 3 cm long, intravascular stent was used in all animals. Stents were successfully placed in 13 of 15 dogs (nine with branch pulmonary stenosis and four with superior vena caval stenosis) with hemodynamic and angiographic relief of the stenoses in all. In three animals, successful stent placement was not accomplished because the distal right pulmonary artery was found to be totally obstructed in two and in one dog with combined vena cava and pulmonary stenosis the distal right pulmonary artery was so severely stenotic that the stenosis could not be crossed. Repeat catheterization performed 6 months following stent placement documented persistent gradient relief and angiographic evidence of unobstructed flow through the stent without thrombus formation and with patent side branch vessels. Our preliminary results suggests that balloon-expandable stents are a potential therapy for the treatment of branch pulmonary artery and superior vena cava stenoses.

Intermediate-term survival and functional results after arterial repair for transposition of the great arteries

An assessment of late morbidity and mortality is essential before arterial repair can be considered truly corrective for patients with transposition of the great arteries. We describe the early and intermediate-term results in 126 patients who underwent arterial repair. Operation was performed at a median age of 6 days, with 76 patients operated on within the first 7 days of life. Coronary artery anatomy differed from the usual arrangement in 37 patients. Simultaneous procedures included ventricular septal defect closure (35) and repair of interrupted aortic arch (2) or coarctation (5). Hospital mortality was seven of 126 (5.5%), with three deaths among the most recent 100 patients (3%). There were one late, noncardiac death and one late death after reoperation. Reoperation for pulmonary artery stenosis was required in 10 of the first 63 patients (16%), all of whom underwent pulmonary artery reconstruction with separate patches for closure of the coronary excision sites. Of the last 63 patients, all of whom underwent pulmonary artery reconstruction with a single pantaloon-shaped pericardial patch, one (2%) required reoperation for pulmonary artery stenosis. Doppler flow studies and echocardiography performed in 115 of 119 surviving patients at a mean of 12 months after repair demonstrated normal left ventricular function, minimal left ventricular outflow gradients, and no more than trivial aortic regurgitation. Peak gradient across the right ventricular outflow tract was 19 +/- 3 mm Hg in patients with separate pulmonary artery patches and 5 +/- 2 mm Hg in those with a single pantaloon patch (p = 0.0001). Follow-up is 96% complete from 1 month to 8 years after operation (mean 2.5 years). The actuarial survival rate at 5 years, including operative mortality, was 92%. All patients are in sinus rhythm, and none requires antiarrhythmic medications. These data suggest that pulmonary artery reconstruction with a single pantaloon patch may be associated with a decreased requirement for reoperation. Intermediate-term survival and functional results are excellent after arterial repair for transposition of the great arteries.

Echocardiographic indexes of allograft rejection in pediatric cardiac transplant recipients

To determine the usefulness of echocardiographic indexes of left ventricular (LV) function as possible predictors of cardiac rejection, 12 transplant recipients (ages 3 to 17 years) underwent a total of 52 serial echocardiographic examinations and cardiac biopsies. The results were compared to those of 12 normal children (ages 2 to 17 years). Biopsies were graded as no rejection (n = 23), mild rejection (cellular infiltrate, n = 13), and moderate rejection (myocyte necrosis, n = 16). LV dimensions, percent shortening fraction, indexed LV mass, and ejection fraction were measured from M-mode and two-dimensional echocardiography. From the mitral valve Doppler tracing, the following measurements were made: isovolumic relaxation time, peak E and peak A velocities, and the fraction of filling under the E and A waves as well as in the first third of diastole. Compared with normal subjects, transplant recipients with no rejection had higher heart rates (95 +/- 15 vs 80 +/- 17 beats/min), longer isovolumic relaxation time (68.8 +/- 11.2 vs 51.5 +/- 13.6 msec), decreased first third area fraction (0.48 +/- 0.10 vs 0.57 +/- 0.10), and similar shortening fraction, LV mass, and peak E and A velocities (p less than 0.03). Compared with transplant recipients with no rejection, patients in whom mild rejection developed also had decreased shortening fraction (31% +/- 10% vs 37% +/- 8%) and decreased peak E velocity (0.68 +/- 0.19 vs 0.88 +/- 0.15 m/s) (p less than 0.03). From mild to moderate rejection, no further changes were noted in any echocardiographic indexes measured.(ABSTRACT TRUNCATED AT 250 WORDS)

Doppler evaluation of aortic valve area in children with aortic stenosis

To evaluate the usefulness of the Doppler-derived aortic valve area calculated from the continuity equation in assessing the hemodynamic severity of aortic valve stenosis in infants and children, two-dimensional and Doppler echocardiographic examinations were performed on 42 patients (aged 1 day to 24 years) a median of 1 day before or after cardiac catheterization. The left ventricular outflow tract diameter was measured from the parasternal long-axis view at the base of the aortic cusps from inner edge to inner edge in early systole. The flow velocities proximal to the aortic valve were measured from the apical view with use of pulsed Doppler echocardiography; the jet velocities were recorded from the apical, right parasternal and suprasternal views by using continuous wave Doppler echocardiography. The velocity-time integral, mean velocity and peak velocity were measured by tracing the Doppler waveforms along their outermost margins. Seventeen patients (all less than or equal to 6 years old) had a very small left ventricular outflow tract diameter (less than or equal to 1.4 cm) and cross-sectional area (less than or equal to 1.5 cm2). The Doppler aortic valve area calculated with use of velocity-time integrals in the continuity equation (0.57 +/- 0.25 cm2/m2, mean value +/- SD) correlated well with the Doppler aortic valve area calculated by using mean (0.55 +/- 0.25 cm2/m2) and peak (0.54 +/- 0.24 cm2/m2) velocities, with correlations of r = 0.97 and 0.95, respectively. Thirty-four patients had sufficient catheterization data to calculate aortic valve area from the Gorlin formula.(ABSTRACT TRUNCATED AT 250 WORDS)

High-frequency jet ventilation improves cardiac function after the Fontan procedure

Pulmonary vascular resistance is an important determinant of cardiac output after the Fontan procedure and is adversely affected by elevated mean airway pressure. High-frequency jet ventilation (HFJV) is an alternate form of mechanical ventilation which supports gas exchange at lower mean airway pressure. This study was performed to determine if HFJV could lower mean airway pressure and pulmonary vascular resistance and result in an increase in cardiac output after the Fontan procedure. We prospectively evaluated 13 patients ranging in age from 0.9 to 8.5 years (mean, 3.9 years) and in weight from 6.2 to 20.1 kg (mean, 13.9 kg). Right atrial, left atrial and pulmonary artery catheters were used to measure hemodynamic parameters. Cardiac index was measured by dye dilution technique, and pulmonary vascular resistance was calculated. The patients were stabilized on mechanical ventilation to achieve a PaCO2 = 30 +/- 5 mm Hg, and baseline hemodynamic and respiratory measurements were made. HFJV was begun at settings adjusted to achieve similar gas exchange. Respiratory and hemodynamic measurements were repeated after 30-60 minutes of HFJV. Mechanical ventilation was then resumed at baseline settings, and measurements were repeated 0.5-1 hour later. There was no significant change in gas exchange. HFJV resulted in a 50% reduction in mean airway pressure (9.2 +/- 0.2 cm H2O to 4.6 +/- 0.1 cm H2O, p less than 0.001), a 59% reduction in pulmonary vascular resistance (3.82 +/- 0.36 to 1.52 +/- 0.16 Woods units, p less than 0.001), and a 25% increase in cardiac index (2.32 +/- 0.12 l/min/m2 to 2.91 +/- 0.12 l/min/m2, p less than 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)

Repair of tetralogy of Fallot in infancy. Effect of pulmonary artery size on outcome

The possible influence of small pulmonary artery size on the results following complete repair of tetralogy of Fallot has not been investigated in patients less than 1 year of age. We, therefore, reviewed our recent experience with infant repair to assess this. Between January 87 and October 90, repair was performed in 58 infants less than 1 year old (median, 6 months) with tetralogy of Fallot and pulmonary stenosis (n = 53) or pulmonary atresia without important aortopulmonary collaterals (n = 5). The McGoon ratio (diameter of the right pulmonary artery at the prebranching point plus that of the left divided by the diameter of the aorta at the diaphragm) by cineangiography (n = 48) was between 1.2 and 1.5 in 15 patients, 1.6 and 2.0 in 20, and greater than 2.0 in 13. Transannular patching was performed in 19 of the patients with pulmonary stenosis (36%). Muscle resection was avoided in 29 infants and was minimal in the remaining 29. A weak correlation, with considerable scatter, was found between the McGoon ratio and postrepair right ventricular/left ventricular pressure measured in the operating room (r = 0.4), and the relations were similar to those previously reported for children. Right ventricular/left ventricular pressure was 0.55 +/- 0.03 in infants with McGoon ratios of 1.2-1.5, 0.46 +/- 0.03 with ratios of 1.6-2.0, and 0.47 +/- 0.03 with ratios greater than 2.0 (p = 0.01). There were no hospital deaths, and the 3-year survivorship was 94%. Two infants underwent successful reoperation for residual right ventricular outflow tract obstruction. All other patients are free of significant hemodynamic residua.(ABSTRACT TRUNCATED AT 250 WORDS)

A balloon-dilatable pulmonary artery band in the dog. Results at one year

Increased pulmonary blood flow and pulmonary hypertension are frequent problems in infants with congenital heart disease. Although the use of pulmonary artery banding to limit pulmonary blood flow has decreased, the procedure may still be beneficial in certain forms of complex heart disease. The ability to noninvasively relieve the obstruction caused by the band may significantly reduce later operative complexity or even avoid reoperation entirely. The present study evaluated the effectiveness of a balloon-dilatable pulmonary artery band. Twenty 1-week-old dogs had a band of an absorbable suture material (Vicryl; Ethicon, Inc., Somerville, N.J.) placed around the main pulmonary artery. Eight dogs underwent angioplasty 6 months after band placement and also underwent follow-up catheterizations 3 and 6 months after angioplasty. Balloon angioplasty acutely reduced both the right ventricle-pulmonary artery pressure gradient (from 37 +/- 7 mm Hg to 3 +/- 1 mm Hg, p less than 0.001) and the right ventricular systolic pressure (from 62 +/- 8 mm Hg to 32 +/- 2 mm Hg, p less than 0.01). At follow-up the gradient remained low, measuring 4 +/- 1 mm Hg at 3 months and 3 +/- 1 mm Hg at 6 months. Twelve dogs did not undergo balloon dilatation until 12 months after band placement to determine whether any obstruction persisted and whether the band could be relieved after long-term placement. These 12 dogs had progressive increases in right ventricle-pulmonary artery gradient, from 27 +/- 3 mm Hg at 6 months to 43 +/- 4 mm Hg at 12 months. Ten of these dogs underwent dilation 1 year after pulmonary artery band placement. This dilation significantly reduced the right ventricular outflow tract gradient (from 43 +/- 4 mm Hg to 1 +/- 1 mm Hg, p less than 0.001). The remaining two dogs underwent successful partial dilation of the band 12 months after placement. This study demonstrated that a pulmonary artery band of absorbable suture material maintains effective right ventricular outflow tract obstruction for at least 1 year. Additionally, the effect of the pulmonary band can be successfully and persistently relieved.

Doppler assessment of pulmonary artery flow patterns and ventricular function after the Fontan operation

To assess the relation between ventricular systolic and diastolic function and pulmonary artery (PA) flow patterns after the Fontan operation, 15 postoperative patients were prospectively evaluated with echocardiography. Blood flow velocities in the PA were recorded with pulsed Doppler echocardiography. Ejection fraction was measured by 2-dimensional echocardiography using Simpson's rule. Indexes of diastolic function were measured from the systemic atrioventricular valve inflow Doppler and included peak E and A velocities, peak filling rate normalized for stroke volume, the fractions of filling in early and late diastole (E and A area fractions), and the E/A velocity and area ratios. Compared with 15 age-matched control subjects, the 15 patients who had undergone the Fontan procedure had decreased peak E velocity (0.65 +/- 0.20 vs 0.87 +/- 0.10 m/s), decreased E/A velocity ratio (1.29 +/- 0.23 vs 1.98 +/- 0.46), decreased normalized peak filling rate (6.09 +/- 0.90 vs 6.81 +/- 0.83 s-1), decreased E area fraction (0.63 +/- 0.09 vs 0.72 +/- 0.07), increased A area fraction (0.37 +/- 0.07 vs 0.24 +/- 0.06), and decreased E/A area ratio (1.77 +/- 0.45 vs 3.33 +/- 1.15) (p less than 0.05). These diastolic filling abnormalities are consistent with impaired ventricular relaxation and decreased early diastolic transvalvular pressure gradient. PA Doppler recordings showed 2 distinct patterns of flow. Pattern I, observed in 9 patients, showed biphasic forward flow with peak velocities in mid to late systole and mid-diastole. Pattern II, observed in the remaining 6 patients, showed decreased systolic forward flow, a late systolic to early diastolic flow reversal, and delayed onset of diastolic forward flow.(ABSTRACT TRUNCATED AT 250 WORDS)

Extracorporeal life support for cardiac assist in pediatric patients. Review of ELSO Registry data

The collected data on extracorporeal membrane oxygenation (ECMO), now referred to as extracorporeal life support (ECLS), for pediatric cardiac support has not been analyzed. The purpose of this study was to review the Extracorporeal Life Support (ELSO) Registry data to evaluate the results, identify possible predictors of outcome, and attempt to establish criteria. From 1981 to June of 1990, 189 patients were placed on ECLS for cardiac assist. The age range was 0-204 months (median, 7 months). Mean time on ECLS was 115 +/- 75 hours. Fourteen patients were placed on ECLS as a bridge to transplant or for management of transplant rejection. All of the remaining 175 patients were treated in the postoperative period. The causes of mortality included lack of improvement in cardiovascular function in 69 (37%) of the patients, major central nervous system damage in 28 (15%), uncontrollable hemorrhage in three (2%), sepsis in three (2%), and pulmonary interstitial disease in two (1%). The Registry data were examined for predictors of outcome. There was no significant difference between survivors and nonsurvivors when compared for duration of ECLS, mechanical complications, arterial or venous blood gases, ventilation settings, or hemodynamics. Forty-three percent of 189 pediatric patients treated with ECLS for cardiac failure survived. The highest survival, 61%, occurred in right-sided lesions and the lowest, 18%, in post-Fontan. Mediastinal bleeding, cardiac arrest, renal failure, and prolonged intubation were all associated with a poor outcome. Most deaths were attributed to irreversible cardiac or brain injury, suggesting that results could be improved by earlier identification of high-risk patients and earlier institution of ECLS.