Neonatal And Paediatric Extracorporeal
Andrew I>.Cochrane, FRACS, Allison M.Horton, CCP, Warwick W. Butt, FRACP, Peter D. Skillington, FRACS, Tom R. Karl, MD, Roger B. B. Mee, FRACS Victorian
Paediatric Cardiac Surgery Unit and Paediatric Intensive Care Unit, Royal Children’s Hospital, Parkville, Victoria, Australia
ost types of respiratory failure in the newborn are reversible, but supportive treatment can damage the lungs and airways. In the last decade prolonged extracorporeal membrane oxygenation (ECMO) has become an accepted technique, particularly in the neonatal group. We have placed 46 patients (26 neonates, 20 children) on ECMO for respiratory or cardiac support. The major indications in neonatal patients were congenital diaphragmatic hernia, persistent foetal circulation and meconium aspiration. The major indications in children were bacterial and viral pneumonias and systemic sepsis with respiratory failure. The hospital survival was 65% in the neonatal group and 40% in the paediatric group. The majority of neonatal deaths occurred in the group with congenital diaphragmatic hernia. In the other neonatal patients hospital survival was 74%. Carotid and jugular cannulation was the preferred technique, with vascular repair at the time of decannulation. Only one neonate and one child who have survived have a neurological or developmental deficit. We conclude that ECMO is most suitable for neonates with reversible lung disease, and congenital diaphragmatic hernia remains the major cause of death, due to pulmonary hypoplasia. The poorer results in children reflect the greater incidence of destructive and irreversible lung disease in this group, with associated systemic sepsis, multiorgan failure and coagulopathy from the primary disease. (AustralAs J Cardiac Thorac Surg 1992; l(2): pp.17-22)
Patients and Methods
Respiratory failure is a common cause of death in neonates. The treatment itself (oxygen and mechanical ventilation) can cause further lung injury, progressing to chronic lung disease known as bronchopulmonary dysplasia (BPD). However unlike that occurring in older children and adults, respiratory failure in neonates is usually caused by immaturity, abnormalities of the airways, or temporary abnormalities of the pulmonary circulation, rather than pulmonary parenchymal disease.
Patient population Over 45 months from May, 1988 to February, 1992,46 patients were placed on ECMO. There were 26 neonates and 20 children. The gestational age of the neonates ranged from 35 to 42 weeks (mean of 38.6 weeks). The paediatric group ranged in age from 1 month to 15 years (mean 6 yrs 2 months). The primary diagnoses are listed in Table 1. The most common neonatal condition was congenital diaphragmatic hernia (CDH). In 5 of these nonates, repair of the diaphragm was performed while supported on ECMO, with the surgery done in the Intensive Care Unit.
Extracorporeal membrane oxygenation (ECMO) is a means of prolonged cardiopulmonary bypass achieved by extra-thoracic cannulation, in patients with pulmonary or cardiac failure who have not responded to conventional therapy. In the last decade it has become an accepted technique for treatment of the newborn infant with respiratory failure’T2. However, it is a supportive and not a curative intervention.
In the paediatric group, the most common indication for ECMO was pneumonia, either bacterial or viral in origin. Five patients were placed on ECMO primarily for cardiac support. Patient selection The general selection criteria for institution ECMO were:
At the Royal Children’s Hospital the first paediatric ECMO case was performed in May, 1988, and the program was extended to neonates in April, 1989. We present here the details of 46 patients treated up to the end of February, 1992.
(1) the presence of a reversible or treatable disease, with no major, persisting sequelae; (2) a predicted 80% mortality despite maximal support with conventional techniques;
Presentedat the Gold CoastRACS-ASCTSCardiothoracicSurgical Meeting, October 1991.
AustralAs J Cardiac Thorac Surg 1992; l(2): pp.17-22
Cochrane, Horton, Butt, Skillington, Karl, Mee Neonatal
(3) the absence of a major pre-existing a poor prognosis. Table 1: Experience ing to diagnosis.
PFC: Persistent foetal circulation; hypertension. Age Group
The contraindications were as follows.
PPH: persistent pulmonary Deaths
4 2 0 1
in our practice
1. Profound neurologic impairment 2.
Congenital defects not compatible a meaningful life
Congenital diaphragmatic hernia PFC/PPH Meconium aspirate Pneumonia/sepsis Hyaline membrane disease Post-cardiac surgery
8 5 3 2
1. Gestational age below 34 weeks, and weight below 2,000 gm
Bacterial pneumonia Viral pneumonia Sepsis Aspiration pneumonia Adult respiratory distress syndrome Post-cardiac surgery Myocarditis Cardiomyopathy - transplant bridge Near drowning Total
5 1 1 1 0
3 2 2
for longer than
Cardiac and cranial ultrasound examinations were performed on all neonatal patients under consideration for ECMO.
Our technique involved neck cannulation, venoarterial perfusion with a Biomedicus constrained vortex pump, and a Scimed or Minimax oxygenator. The total priming volume of the circuit for a neonate was 350 mL, using one unit of leucocytefiltered blood, albumin, 5% dextrose, calcium chloride and sodium bicarbonate. A small amount of heparin was added (1,000 Units). This volume is equal to the blood volume of a baby, so that it is important that the priming volume be adjusted biochemically to be as similar as possible to the patient’s blood before commencing ECMO perfusion. This prevents problems of major fluid shifts, hypocalcaemia, acidosis and large fluctuations in glucose.
where PIP is the peak inspiratory pressure in mmHg, PEEP is the positive end-expiratory pressure in mmHg and paC02 is the arterial partial pressure of carbon dioxide in mmHg. Oxygenation Index =
Mechanical ventilation 7 to 10 days
The major indication was reversible (within 2 weeks) of treatable pulmonary failure, provided the disease process was not expected to lead to such a degree of irreversible lung damage to prevent weaning from a ventilator or residual severe respiratory impairment. We used the following criteria12-14 according to the age of the patient and disease process present, to define an 80% mortality risk: Ventilation Index
100 x FI02 x Mean Insp. Pressure
The cannulation procedure was done in the intensive care unit. The right common carotid artery and internal jugular vein were generally used, and after exposure and securing haemostasis a small loading dose of heparin (100 U/kg) was given. Longitudinal vascular incisions were used. Thin-walled cannulae with maximal internal diameter were utilized (Medtronic Bio-medicus), usually 8,lO or 12 Fr. in the artery and 12 or 14 Fr. in the jugular vein for a neonate. The cannula position was estimated externally before insertion and then checked later on chest x-ray examination. (The venous cannula tip should lie in the mid-right atrium, and the arterial cannula tip should lie just within the ascending aorta.)
where FI02 is the fractional inspired oxygen concentration and paOz is the arterial pressure of oxygen in mmHg”. A neonate or child was considered for ECMO if these values were exceeded for 4 or more hours: In children: VI > 40 and 01 > 40. In neonates without CDH: VI > 70, 01 > 25. In neonates with CDH: VI > 90, 01 > 40. 18
The cannulae were secured in the vessels by the application of haemoclips to the soft vascular slings, and the excess length of sling was cut off (Fig. 1). This provided haemostasis around the cannula, minimal vascular trauma, and allowed reconstruction at the time of decannulation”‘. The cannulae were then further secured to the skin of the neck by sutures, and then by tape to the head. Secure fixation helped to maintain haemostasis, because movement of the cannulae in the vessels predisposes to bleeding around the cannulae.
Horton, Butt, Skillington, Karl, Mee Neonatal and paediatric ECMO
of thrombosis in the circuit and the risk of bleeding, in particular the risk of intracranial haemorrhage (ICH) in the neonate. The platelet count was maintained above 100,000 per cu.mm, in particular to prevent ICH. Platelet consumption varies and is usually a function of oxygenator size, bleeding, presence or absence of sepsis, and the need for previous exchange transfusion’s. To reduce platelet consumption, the oxygenator with the smallest surface area able to provide adequate gas transfer was selected. Regular platelet transfusions were often necessary. Haemolysis was checked for by daily measurement of the free plasma haemoglobin. A regular cranial ultrasound examination was performed to monitor for the development of intracranial haemorrhage. The ventilator rate was reduced to minimise lung barotrauma and the inspired oxygen concentration decreased. Typical settings were: 5 breaths per minute; inspired O2 of 21% (air); positive endexpiratory pressure of 8 to 10 mm Hg. Adequate tracheal toilet was maintained and measures taken to minimise atelectasis. The tidal volume was measured regularly monitor improvement in lung compliance.
INTERNAL JUGULAR VEIN
Renal function was often poor on ECMO, and haemofiltration when necessary was incorporated into the circuit by means of a small shunt, placed after the pump and in parallel to the oxygenator. Parentera1 nutrition was usually necessary.
Statistical methods. For the analysis of differences in status prior to ECMO, Fisher’s exact test was used.
Fig. 1. Method of neck cannulation, with vessels well secured by soft vascular loops, but exposed to minimal trauma, to allow later reconstruction. The purse string suture at the site of cannulation is sometimes not used.
Survival Within the neonatal group, there were 17 survivors in 26 infants (65%). The majority of deaths (4 of 7) were in neonates with congenital diaphragmatic hernia, and these neonates possessed insufficient lung tissue to achieve independent survival. In neonates without diaphragmatic hernia, there were 5 deaths out of 19 (74 % survival, Table 2).
At decannulation, permanent ligation or vascular repair was performed. This was again done in the Intensive Care Unit, using a continuous 7/O prolene suture after clearing the vessels of thrombus in both directions. Management of ECMO When cannulated, the flow was gradually increased to the required level, usually 120-150 mL/kg/ min. Heparin therapy was given by continuous infusion and controlled by hourly measurement (Hemotech Inc, Englewood, Colorado, USA) of the activated clotting time, which was kept at 160-170 seconds. This is significantly lower than the minimal level of 400 set maintained during a surgical bypass procedure. A balance must be found between the risk
Table 2: Hospital survival of patients placed on ECMO. centages given are the mean and 70% confidence limits. Patient
Neonates Without diaphragmatic With diaphragmatic Children
19 7 20
14 74% 3 43% 8 40%
(64~84%) (24-62%) (29-51%)
In the paediatric group, there were 8 survivors in 20 patients (40%). Of those placed on ECMO for primary respiratory support, 8 of 16 survived (50%), but in the largest group treated, children with severe bacterial pneumonia, there were no survivors. When ECMO was instituted for primary cardiac support, there was only one survivor among 5 children. This patient was a lZyear-old girl who underwent urgent aortic valve replacement for staphylococcal endocarditis and septicaemia. At 24 hours after operation she had a cardiac arrest, and after ressuscitation was placed on ECMO for 5 days, with a successful outcome. Pre-ECMO
The pre-ECMO status of the children was significantly different from that of the neonates for the incidence of cardiac arrest, renal and liver failure, and coagulopathy (Table 3). There was no difference in the incidence of seizures prior to ECMO.
Cardiac arrest Renal failure Liver failure Seizures Disseminated microvascular coagulation or sepsis
8 7 8 5
32 6 3 6
8 on ECMO
5 5 7 2 2 3 1 I 1
Significant haemolysis (plasma Hb > 1.0 g/L) occurred in 8 patients, 6 of these being neonates. Seven episodes of severe bleeding occurred, 5 of these in the paediatric group and only 2 in neonates. Only 1 episode of intracranial haemorrhage was found in the 26 neonatal patients. Bleeding was a more serious problem in the paediatric group, with bleeding from the neck wound, surgical wound, and in 1 case a pulmonary haemorrhage.
and children of the two
1 1 0 6
Mechanical Oxygenator failure Pump head failure Air in circuit Clots in circuit Patient related Haemolysis Infection lnotropic support required Electrolyte disturbance Bleeding at cannula site Surgical wound bleeding Seizures Pulmonary haemorrhage Intracranial haemorrhage Gastrointestinal haemorrhage
Table 3: The pre-ECMO status of neonatal patients placed on ECMO. P values are for the comparison groups.
4: Complications occurring while mechanical or patient related.
Skillington, and paediatric
Among 17 surviving neonatal patients, 16 were ‘normal’ (at an early stage of development) and only 1 has a moderate handicap. Among 8 surviving children, only 1 has a mild deficit.
Vascular patency after repair
Follow-up by Doppler ultrasound of 10 babies demonstrated a patent carotid artery in all, and a patent jugular vein in 7 of 10 infants. In particular, no case of infection with either vascular rupture or false aneurysm occurred after repair. There is no definite evidence in favour of vessel repair over ligation, but we believe that repair of both vessels, in particular the carotid artery, should be done whenever possible.
Within a few years of the introduction of the heart-lung machine by Gibbon in 19533, work on pro-’ longed support with membrane oxygenators was in progress, and Kolobow and Bartlett separately supported animals for up to 1 week without excessive haemolysis 4*5. Bartlett demonstrated that much lower doses of heparin could be used, in order to minimise the bleeding complications of full heparin therapy6. The first successful human ECMO support was reported by Hill’. The first successful neonatal patient was reported by Bartlett in 1975’, and by 1981 his group had reported the treatment of 45 neonates with 25 survivors’. Over the next decade the technique became widely used, until by 1991 over 5,000 neonates had been supported on ECMO, with an 83% survival”.
Complications can be categorised as mechanical problems, related to the pump, circuit and oxygenator, or patient-related complications due to adverse effects of this invasive procedure on the patient. The complications occurring in over 5,000 hours of ECMO support are listed in Table 4. Oxygenator deterioration requiring change was relatively common, and usually a deterioration in the patient’s arterial blood gas levels and the postoxygenator blood gas levels provided warning, and allowed a semi-elective change of oxygenator.
Neonatal ECMO has become an established technique, with good results from many centres. Our results confirm this and are similar to the neonatal ECMO Registry” which reports an overall 83% 20
AustralAs J CardiacThorac Surg 1992; l(2):
Cochrane,Horton, Butt, Skillington, Karl, Mee
It is important to recognise that selection criteria derived from the literature may not reflect the individual hospital’s experience, and each unit must determine its own clinical predictors of 80% mortality, and therefore its own criteria for institution of ECMO. As the complications of ECMO decrease and further advances occur, ECMO might be considered at a lower mortality2”.
survival in 5,162 neonates, treated at 78 centres. The Registry reports a 93% survival for meconiumaspiration, (all 5 survived in our experience), falling to 61% in congenital diaphragmatic hernias (only 3 of 7 survived in our experience). In the neonate, persistent pulmonary hypertension is the major pathophysiologic mechanism of respiratory failure in the full-term baby, regardless of the underlying disease process2’. The pulmonary vasculature of the neonate responds to factors such as hypoxia, hypercapnia and acidosis by vasoconstriction, resulting in increased pulmonary vascular resistance with right to left shunting through the ductus arteriosus and the foramen ovale. This exacerbates the hypoxia, resulting in further elevation of pulmonary vascular resistance21. This vicious cycle can usually be broken by paralysis, ventilation and pharmacologic therapy, but 2% to 5% of neonates fail to respond 22. The beneficial physiologic effects of ECMO in this situation may be summarised as follows.
It is also important to identify those neonates with no reasonable chance of recovery or suffering from another severe disease process. The initial trials of neonatal ECMO resulted in many episodes of ICH, in up to 35% of neonates and associated with a 94% mortality16. Gestational age greater than 34 weeks was the single most important factor in avoiding ICH. The successful management of a patient on ECMO requires a team approach, including the neonatal intensivist , perfusion technicians, nursing staff trained to run the equipment, and cardiac surgeon. Detailed attention to numerous factors is necessary: the patient’s haemodynamics, circulating volume, oxygenation, anticoagulation, and observation for thrombus or air in the circuit and pump-head deterioration. Close attention to haematological and biochemical parameters is required.
Reversal of the cycle of hypoxia, acidosis and increasing pulmonary vascular resistance. Reduction in the complications from highpressure ventilation and O2 toxicity. Provision of time for the pulmonary to resolve and for the iung to heal.
The good results for neonatal ECMO, with the exception of congenital diaphragmatic hernia, probably reflect the presence of single organ failure alone, with the pulmonary problem being reversible or a matter of immaturity, and major lung damage absent. However, in the paediatric age-group, the results are worse than in the neonatal group, but are similar to the results for adult ECMO. Our results (40% survival) are again quite similar to the paediatric experience of the ECMO Registry, which reports a 45% survival in 253 children treated for respiratory failure in 48 centres” . The poorer results in children appear to be due to the same general causes as found in adult ECMO: a higher incidence of destructive and irreversible lung disease with systemic sepsis frequently present, coagulopathy with bleeding problems, and often multi-organ failure. In those supported for pneumonia, the underlying destructive lung infection was clearly too severe, and ECMO did not alter the natural history of this illness.
Inotropic and vasoactive drugs can be ceased or reduced. Despite the increasingly widespread use of ECMO in these age-groups, only two prospective studies comparing ECMO with conventional ventilation have been performed showing a benefit from this procedure14,24, and in the case of Bartlett’s study14 significant criticisms of the methods of treatment randomisation were made25. In O’Rourke’s study24, infants who met a criterion for 85% chance of death, as defined by a ratio of arterial p02 to alveolar p02 of < 0.15, were included. In phase 1 of the study, 6 of 10 babies in the conventional therapy group survived while all of 9 babies in the ECMO group survived. Randomisation was halted, and the next 20 babies received ECMO, with 19 surviving. The overall survival of ECMO-treated neonates was 97% (28 of 29) compared with 60% (6 of 10) in the conventional treatment group (~~0.05). There are clearly serious ethical problems in conducting trials where death is the end-point, because of both anecdotal evidence in support of ECMO and studies comparing ECMO with historical controls that suggest ECMO improves survival.
Although our numbers are limited, the rather poor result for ECMO in patients requiring cardiac support warrants some consideration. The ECMO Registry experience reports a 33% survival in 385 children with primary cardiac disease”. Firstly, in the patients whom we placed on ECMO, the decision to commence ECMO may have sometimes been taken too late. 21
AustralAs J Cardiac Thorac Surg 1992; l(2): pp.17-22
Cochrane, Horton, Butt, Skillington, Karl, Mee Neonatal
Secondly, in relation to the mode of assistance, in some institutions all paediatric patients requiring post-operative cardiac support have been placed on ECMO, whereas for post-operative support we have favoured the left ventricular assist device; we have had experience of this in 20 patients, for a 50% discharge rate and 40% long-term survival (unpublished data). Therefore the cardiac patients receiving ECMO may have been a selected and particularly high-risk group in this experience.
Bartlett RH, Kittredge D, Noyes BS Jr, et al. Development of a membrane oxygenator: Overcoming blood diffusion limitation. J Thorac Cardiovasc Surg 1969; 58:795. 7. Hill JD, O’Brien TG, Murray JJ, et al. Extracorporeal oxygenation for acute post-traumatic respiratory failure (shock-lung syndrome): Use of the Bramson Membrane Lung. N Engl J Med 1972; 286:629-34. 8. Bartlett RH, Gazzaniga AB, Jefferies R, et al. Extracorporeal membrane oxygenation (ECMO) cardio-pulmonary support in infancy. ASAIO Trans 1976; 22:80-93. 9. Bartlett RH, Gazzaniga AB, Huxtable RF, et al. Extracorporeal circulation (ECMO) in neonatal respiratory failure. J.Thorac Cardiovasc Surg 1977; 74:826-33. IO. The ECMO Registry report of the Extracorporeal Life Support Organisation, July 1991. 11 Krummel TM, Greenfield LJ, Kirkpatrick BV, et al. Alveolararterial oxygen gradients versus the neonatal pulmonary insufficiency index for prediction of mortality in ECMO candidates. J Pediatr Surg 1984; 19: 380-4. 12. Beck R, Anderson KD, Pearson GD, et al. Criteria for extracorporeal membrane oxygenation in a population of infants with persistent pulmonary hypertension of the newborn. J Pediatr Surg 1986; 21:297-302. 13. Bartlett RH, Roloff DW, Cornell RG, et al. Extracorporeal circulation in neonatal respiratory failure: A prospective randomised study. Pediatrics 1985; 76:479-87. 14. Hallman M, Merritt A, Jarvenpaa AL. Exogenous human surfactant for treatment of severe respiratory distress syndrome: A randomised prospective clinical trial. J Pediatr 1985; 106:9639. 1.5. Cilley RE, Zwischenberger JB, Andrews AF, et al. Intracranial hemorrhage during extracorporeal membrane oxygenation in neonates. Pediatrics 1986; 78:699-704. 16. Karl TR, Iyer KS, Sano S, Mee RBB. Infant ECMO cannulation technique allowing preservation of carotid and jugular vessels. Ann Thor Surg 1990; 50:488-9. 17. Anderson HL, Cilley RE, Zwischenberger JB, et al. Thrombocytopenia in neonates after extracorporeal membrane oxygenation. ASAIO Trans 1986; 32:534-7. 18. Andrews AF, Roloff DW, Bartlett RH. Use of extracorporeal membrane oxygenators in persistent pulmonary hypertension of the newborn. Clin Perinatol 1984; ll:729-35. 19. O’Rourke PP, Crone RK, Vacanti JP, et al. Extracorporeal membrane oxygenation and conventional medical therapy in neonates with persistent pulmonary hypertension of the newborn: A prospective randomised study. Pediatrics 1989; 84:957-63. 20. Ware JH, Epstein MF. Extracorporeal circulation in respiratory failure (Commentaries). Pediatrics 1985; 76849-51. 21. Butt W. Paediatric extracorporeal membrane oxygenation, Current Opinion in Anaesthesiology 1990; 3:394-8. 22. Starnes VA, Oyer PE, Bernstein D, Baum D, Gamberg P, Miller J, Shumway NE. Heart, Heart - Lung, and Lung Transplantation in the first year of life. Ann Thor Surg 1992; 53:306-10.
New advances are occurring in two particular areas, which offer the prospect of some modest benefits. The use of heparin-bonded tubing and oxygenators should minimise the need for heparin and improve the safety of ECMO in terms of bleeding complications. The Carmeda method of heparin ‘end point attachment’ provides a long-term thromboresistant circuit, compared to earlier, temporary methods of heparin-bonding. Secondly, the application of neonatal lung transplantation will slowly occur, in particular the use of a ‘cut-down’ paediatric or adult lung in infants with diaphragmatic hernia and severe lung hypoplasia. In these infants the role of ECMO will therefore become better defined, as a bridge to lung transplantation. This surgery has started in a limited number of centres2’. References
1. Bartlett RH, Gazzaniga AB, Toomasian J, Corwin AG, Roloff D, Rucker R. Extracorporeal membrane oxygenation (ECMO) in neonatal respiratory failure : 100 cases. Ann Surg 1986; 204: 236-45. 2. Stolar CJ, Snedecor SM, Bartlett RH. Extracorporeal membrane oxygenation and neonatal respiratory failure: Experience from the Extracorporeal Life Support Organisation. J Paed Surg 1991; 26: 563-71. 3. Gibbon JH: Application of a mechanical heart and lung apparatus to cardiac surgery. Minn Med 1954; 37: 171. 4. Kolobow T, Zapol W, Pierce JE, Keeley AF, Replogle RL, Haller A. Partial extracorporeal gas exchange in alert newborn lambs with a membrane artificial lung perfused via an AV shunt for periods up to 96 hours. ASAIO Trans 1968; 14:32834. 5. Bartlett RH, Isherwood J, Moss RA, et al. A toroidal flow membrane oxygenator: Four day partial bypass in dogs. Surg Forum 1969; 20:152-3.
In 1991 the Tag Medical company, distributors of Snowden Pincer surgical instruments and agents for the Carbomedics valve, generously offered an annual prize for excellence in endeavour by a surgical trainee. The competition was organised by the Australasian Society of Cardiac and Thoracic Surgeons.
the winner and the author will receive Snowden Pincer instruments to the value of $5,000. The winning manuscript, entitled Neonatal and Paediatric Extracorporeal Membrane Oxygenation, was written by Andrew Cochrane, based on a series of clinical cases collected at the Royal Children’s Hospital. The Australasian Society of Cardiac and Thoracic Surgeons takes this opportunity to thank Gary and Maree Pierce of Tag Medical for their farsighted and generous gesture. The Society congratulates Andrew on his success and wishes him and the other entrants well in their further training. Tag Medical have also offered a prize to be won in 1992. The competition will be based on a presentation by a trainee at the Gold Coast Meeting November 26-28, 1992.
It was decided to award the prize for the best manuscript prepared on the basis of a talk given at the October 1991 Gold Coast (Inter-GSM) Meeting and submitted to The AustralAsian Journal of Cardiac and Thoracic Surgery. Accordingly the ASCTS appointed a panel of judges well respected in the profession. Three manuscripts were received, one by Andrew Cochrane, one by George Matalanis, and one by Eli Khouri. One was rated as