The Early Evaluation of Survivors After Extracorporeal Membrane Oxygenation for Neonatal Pulmonary Failure

By Thomas M. Krummel, Lazar J. Greenfield, Barry V. Kirkpatrick, Dawn G. Mueller, Kathryn W. Kerkering, Miguel Ormazabal, Edwin C. Myer, Robert W. Barnes, and Arnold M. Salzberg

Richmond, Virginia

�9 Excluding mortality data, there is little informa- tion regarding patients' development after extra corporeal membrane oxygenation (ECMO). In six of nine neonates surviving ECMO for predictably fatal pulmonary failure, examination 15 to 21 months afterward showed (1) physical growth and develop- ment, normal in six; (2~ chest x-ray, normal pulmo- nary parenchyma; (3) average arterial blood gases, PO 2 80, Pco z 35, pH 7.35; (4) echocardiogram, nor- mal, without evidence of pulmonary hypertension; (5) cerebrovascular dopplers, normal ophthalmic artery flow in five patients, retrograde in one; {6) CT scan, EEG, neurologic survey, normal in five, cerebral atrophy in one patient who had an air embolus during decannulation; (7) psychologic examination, normal in all. This early evaluation of ECMO survivors should encourage its further application in those newborns who would otherwise die.

INDEX WORDS: Extracorporeal membrane oxygena- tion (ECMO).

S INCE 1977, extracorporeal membrane oxy- genation (ECMO) has been available as a

life-saving supportive measure for neonates with severe hypoxia due to either pulmonary paren- chymal dysfunction or persistent fetal circula- tion. Thus far however, only three teams in four institutions have accumulated experience with neonatal ECMO. 2~ Widespread reluctance to adopt this technique may be due, in part, to the paucity of data documenting growth, develop- ment, and eventual neurologic performance of infants surviving ECMO support. Thus, a critical scrutiny of survivors is essential to properly assess the value of ECMO in neonatal hypoxia.

Nine newborns with predictably fatal hypoxia have been managed with ECMO, six have sur- vived. Five of the six are normal 15 to 21 months after ECMO. The sixth infant has had satisfac- tory somatic growth and development compli- cated by neurologic impairment with cortical atrophy secondary to an air embolus during venous decannulation.

MATERIALS AND METHODS

Nine neonates with predictably fatal pulmonary failure have been supported with ECMO at the Medical College of

Virginia since October 1980 (Table 1). All were judged to have 100% mortality despite optimal medical management, and ECMO was utilized as a final resuscitative measure. Selection criteria based on alveolar-arterial oxygen gradient (A-aDO2) have been previously reported. 4 Partial venoarter- ial bypass was achieved using the right common-carotid artery and internal jugular vein to position the arterial cannula in the proximal aortic arch and the venous catheter in the right atrium (Fig 1).

Throughout the procedure, systemic heparinization was maintained, titrating the dose to achieve an activated clotting time between 240 and 360 seconds. 5 ECMO was continued for up to 313 hours, with progressive pulmonary improvement in all cases; ultimately six patients were successfully weaned. Following discontinuation of ECMO, both the artery and vein were ligated during decannulation. Thereafter, the survi- vors were routinely weaned from the ventilator and eventu- ally discharged.

Since discharge, these infants have been followed at regu- lar intervals, with special attention to physical growth and development, cardiopulmonary function, cerebrovascular status, and neurologic competence.

In addition to careful routine examinations, all patients have had chest x-rays, arterial blood gases, electrocardio- grams, and 2-D echocardiograms. Adequacy of cerebral blood flows have been assessed using periorbital doppler examination. Interval examinations by the pediatric neurolo- gist have documented neurologic function and development, confirmed by EEG and cerebral computed tomograph (CT) scan and psychologic and psychometric testing.

RESULTS

Six survivors were weaned from ECMO after 77 to 200 hours (Table 2). The remainder of their hospitalization was unremarkable and all were discharged without apparent complication. Since discharge they have been followed for 15 to 21 months.

From the Department of Surgery and Pediatrics, Medical College of Virginia, Richmond, Virginia.

Presented at the British Association of Pediatric Sur- geons, Madrid, Spain, 1982.

Supported in part by a VCU Faculty Grant-in-Aid. Address reprint requests to Arnold M. Salzberg, MD,

Department of Surgery, Medical College of Virginia, Rich- mond, VA 23298.

�9 1984 by Grune & Stratton, Inc. 0022-3468/84/1905-0021503.00/0

Journal of Pediatric Surgery, VoI. 19, No. 5 (October), 1984 585

586

Table 1. Results in Neonates Supported With ECMO

Diagnosis Number Survivors

Meconium aspiration syndrome (MAS) 5 3 Congenital diaphragmatic hernia (CDH) 2 2

Persistent fetal circulation (PFC) 2 1

Duration of ECMO support averaged 167 hours (77 to 313).

Growth and Development

Physical growth and development are depicted in Figs 2 and 3. Weight gain has been uniform and steady, and all are above the 20th percentile for their age. Head circumference is normal in five, but below the fifth percentile in the infant with the air embolus (Case 6).

Cardiorespiratory Development

Pulmonary function is excellent in all infants; no resting or exertional cyanosis has been noted. Arterial blood gases on room air are normal in all, with an average PO2 of 80 torr. This can be contrasted with pre-ECMO values as seen in Table 3.

Evidence of radiologic improvement on chest x-ray lagged behind clinical improvement during the ECMO perfusion, but by discharge was obvious in all. In Case 5 a small recurrent diaphragmatic hernia became apparent at 9 months of age and was repaired without incident. At present, all survivors have unremarkable pul- monary parenchyma and cardiac silhouettes on chest x-ray. EKGs are similarly normal.

2-D echocardiogram demonstrates normal size and relationship of all four cardic chambers and the great vessels in each infant. There is no direct

OXYGE

Fig 1. ECMO circuit.

KRUMMEL ET AL

or indirect evidence of persistent pulmonary hypertension or patent ductus arteriosus in any child.

Cerebrovascular Assessment

Cerebral circulation was evaluated using the periorbital doppler examination. 6 All infants had evidence of adequate cerebral blood flow to both hemispheres despite the necessity of right com- mon carotid artery ligation for satisfactory ECMO support, v Blood flow in the right ophthal- mic artery was antegrade in five survivors and retrograde in one (Case 4). More sophisticated noninvasive doppler studies will not be possible until the infants attain enough maturity to coop- erate during the exam.

Neurologic Development

All survivors have been examined at intervals by the pediatric neurologist (ECM). Findings on clinical examination have been correlated with EEGs and CT scans. Though all had some seizure activity before or during ECMO, none have had seizures since discharge and all have discontinued prophylactic medications.

Five infants have passed developmental mile- stones on schedule and have no significant neuro- logic deficit. Mentation is appropriate for age in all. Many of these survivors do show right-hand preference, and mild facial asymmetry. Reflexes are often more brisk on the left side. The signifi- cance of these findings is uncertain, but clearly there is no functional impairment.

EEGs in these five show nonspecific changes, but no discernible abnormality. Similarly, no pathology is seen on CT scans.

The one survivor with the air embolus (Case 6) has significant neurologic deficit at 15 months age. She has decreased movement and power with hypertonic reflexes on the left side, consis- tent with cortical atrophy on the right. The slowed velocity of head circumference growth also suggests some failure of hemispheric growth. This dysfunction appears to be static, with no primary treatable condition.

These clinical findings correlate with the EEG, which demonstrates diffuse slowing on the right side. CT scan is remarkable for bilateral cortical atrophy, greater on the right.

This infant's neurologic deficit is significant at present; her eventual level of function is unclear.

ECMO SURVIVOR FOLLOW-UP 587

Table 2, ECMO Survivor Data

Age at Start ECMO Follow-up Birth Weight ECMO Duration Duration

Case No. Sex (g) Diagnosis (hr) (hr) (rno)

1 F 3440 MAS 40 177 21

2 M 4440 PFC 36 154 20

3 F 3100 MAS/PFC 26 141 19

4 M 3800 MAS 28 111 18

5 M 3110 CDH/PFC 11 77 16

6 F 3300 CDH/PFC 18 200 15

Brain growth and development appears to have been retarded by cerebral hypoxia and ischemia, presumably secondary to air embolization during decannulation and the ensuing cardiorespiratory collapse.

Psychologic Evaluation

In five patients the intelligence quotient using the Bailey infant psychometric scores varied between 98 and 105 (normal = 100). The score was similarly normal (100) in the patient with the air embolus.

DISCUSSION

Though neonatal ECMO has been feasible for 5 years, ~ few institutions have acquired ECMO capabilities. Certainly hypoxia due to lung dys- function or pulmonary hypertension is the major cause of neonatal morbidity and mortality. Any therapeutic advance capable of blunting this effect would ordinarily be enthusiastically applied. ECMO support entails some potentially morbid manuevers, however, including ligation of the right common carotid artery and the inherent risks of prolonged extracorporeal circu- lation. These hazards must necessarily be super- imposed on an infant already severely stressed with the hypoxia, acidosis, and hypotension of the underlying disease.

Without documentation of eventual somatic and especially neurologic development, ECMO survivors could potentially develop crippling sequelae with a resultant quality of life that would be unacceptable.

Those few infants surviving ECMO at the Medical College of Virginia have been observed for a short interval since discharge, but early evaluation is encouraging. Somatic growth is progressing normally at this early date and there is no reason to doubt that this pattern will continue.

Lung development and function, crudely assessed by chest x-ray and resting arterial blood gases, is certainly adequate in all. In the future, pulmonary function tests and blood gases follow- ing exercise will better identify any deficits, but this preliminary data suggests that eventual pul- monary function will not be restrictive.

Cardiac performance is also normal. Of par- ticular interest is the absence of evidence of persistent pulmonary hypertension on EKG or 2-D echocardiogram. PFC has been a significant factor in the initial hypoxia of four of the six survivors. This was documented with 2-D echo- cardiograms prior to ECMO showing increased right heart size, and a patent foramen ovale and patent ductus arteriosus. These findings have completely resolved. This is consistent with the observations of others 8'9 that if elevated pulmo- nary vascular resistance is reversible pharmaco- logically, it eventually resolves.

Obvious concern centers around the effects of carotid artery ligation on cerebral perfusion. Ligation may not be necessary for all ECMO perfusions if venovenous bypass proves work- able, 2''0 however, venoarterial capability provides an essential margin of safety if venovenous bypass becomes inadequate. Thus, the nagging question regarding the consequences of carotid ligation must be addressed. ECMO survivors provide the best documentation of those conse- quences, if any exist.

Interruption of the right common carotid artery by ECMO arterial eannulation leaves three major vessels to perfuse the brain via the circle of Willis. The circle is a unique anatomic arrangement: it functions as an arterial mani- fold, balancing and distributing inflow to the anterior posterior circulation. If the circle of Willis is intact, no deficits in cerebral circulation will occur with unilateral carotid ligation. 11

Unilateral carotid ligation is thus apt to be

~b

40

36

34

32

3O

28

26

24

_ 22 ~ -

~ 20-

1 8 -

1 6 -

1 4 - ~

1 2 "

1 0 - t

-i 8" /

6 - }

o

in. 0 21

20,

19-

18-

17-

14-

~8 24 3

Aqe {months)

Age (months)

12 18 24 I, 1 I

35

95tn

90t~

75~.

501~

2~t~

XOln

Fig 2. for age.

Growth chert: weight

3O 36 l I

. _ . . . . . . 9 O r .

7 $ t h

5 0 t n

2 5 t h

Fig 3. Growth chart: head circumference for age.

ECMO SURVIVOR FOLLOW-UP

Table 3. Arterial Blood Gas Studies

Pre-ECMO Follow-Up

PO 2 26 PO z 80 PCO 2 46 PCO 2 35

pH 7.10 pH 7.35

Mechanical ventilation Spontaneous ventilation

100% 02 Room Air

well tolerated if the remainder of the inflow to the circle is intact and if the circle is complete. Aplasia of one or both carotids 12-~4 as well as hypoplasia of the entire internal carotid sys- tem 15-~7 have been reported; a total of 21 cases have been found in the world literature.

Though variation in the circle of Willis are common, anomalies leading to a truly incomplete circle are also rare; variously reported from 0% to 5%.~8 22 The only large study of the incidence of cerebrovascular lesions in children reported no vascular lesion affecting the central nervous sys- tem in 8948 children under 14.9 years of age. 23

Thus, although anomalies that would make carotid ligation extremely morbid do occur, they are rare. It would appear that though carotid ligation is emotionally bothersome, it is much less morbid than the underlying disease process left untreated.

Noninvasive doppler studies can be used to evaluate cerebral blood flow and adequacy of collaterals. This would be reassuring if used prior to ECMO cannulation, but to date this applica- tion has been logistically difficult.

In these survivors with ligation, flow was ante- grade in the ophthalmic artery in five and retro- grade in one. Both findings are consistent with right carotid ligation and adequate collateral flow. More sophisticated analyses of cerebral blood flow will be possible in the future, but cerebral blood flow appears satisfactory in all survivors.

Neurologic competence appropriate for age has developed in five (83%) infants surviving

589

ECMO. Mentation, motor, and sensory func- tions are all without significant impairment; all are free of seizures. Some mild degree of asym- metry is noted in many by right-handed prefer- ence, and slightly increased left-sided reflexes, but there is no functional handicap. EEG and CT scan support this finding.

One survivor has clinical findings and neuro- logic testing demonstrative of intracranial insult. Developmental progress has been slow, and CT scan has imaged prominent ventricles and corti- cal atrophy, greater on the right. EEG showed diffuse slowing with right-sided amplitude depression consistent with the structural damage observed on CT. These deficits are presumed secondary to air embolus and the accompanying cardiovascular collapse at decannulation, though her unstable pre-ECMO course was also poten- tially damaging to the brain. It is, however, encouraging to note her satisfactory performance on intelligence testing studies.

In all patients referred for ECMO, the sub- strate for cortical injury exists as a result of the basic disease process; carotid ligation gives cause for additional concern. The clinical or subclinical deficits observed in all survivors could be due to ligation, but the cortical insults of hypoxia, aci- dosis and hypotension necessitating ECMO may have been causative as well. Differentiation between pre-existing deficits and those secon- dary to ECMO manipulations is difficult; how- ever, the functional normalcy of five survivors is encouraging when one considers their pre- ECMO distress.

Further follow-up is required to ascertain ulti- mate neurologic performance as these infants grow, but this early assessment suggests that ECMO support can be accomplished safely and that normal development is possible. These results should encourage the further application of ECMO in those newborns who would other- wise certainly die.

REFERENCES

1. Bartlett RH, Gazzaniga AD, Huxtable RF, et al: Extracorporeal circulation (ECMO) in neonatal respiratory failure. J Thorac Cardiovasc Surg 74:826-833, 1977

2. Bartlett RH, Andrews AF, Toomasian JM, et al: Extra- corporeal membrane oxygenation (ECMO) for newborn res- piratory failure: 45 cases. Surgery 92:425-433, 1982

3. Hardesty RL, Griffith BP, Debski RF, et al: Extracor- poreal membrane oxygenation: successful treatment of per-

sistent fetal circulation following repair of congenital dia- phragmatic hernia. J Thorac Cardiovasc Surg 81:556, 1981

4. Krummel TM, Greenfield L J, Kirkpatrick BV, et al: Clinical use of an extracorporeal membrane oxygenator in neonatal pulmonary failure. J Pediatr Surg 17:525-531, 1982

5. Baden JP, Jonnefield M, Ferlie RM, et al: The Bason

590 KRUMMEL ET AL

test: a rapid bedside test for control of heparin therapy. Surg Forum 32:172, 1971

6. Barnes RW, Russell HE, Bone EG, et al: The doppler cerebrovascular examination: improved results with refine- ments in technique, Stroke 8:468-471, 1977

7. Bartlett RH, Gassaniga, AB: Extracorporeal circula- tion of cardiopulmonary failure, in Current Problems in Surgery, Vol 15. Chicago, Yearbook Medical Publishers, 1978

8. Moodie DS, Telander RL, Kleinberg F: Use of tolazol- ine in newborn infants with diaphragmatic hernia and severe cardiopulmonary disease. J Thorac Cardiovasc Surg 75:725- 729, 1978

9. Bloss RS, Turmen T, Beardmore HE: Tolazoline ther- apy for persistent pulmonary hypertension after congenital diaphragmatic hernia repair. J Pediatr 97:984-988, 1980

10. Bartlett RH: Personal Communication, University of Michigan Ann Arbor, MI 1982

11. Moore W: Anatomy of the arterial cerebral circula- tion, in Rutherford RB, (ed): Vascular Surgery. Philadel- phia, W.B. Saunders Co. 1977

12. Fischer AGT: A case of complete absence of both internal cartoid arteries with a preliminary note on the development history of the stapedial artery. J Anat Physiol 48:37-46, 1913

13. Turnbull I: Agenesis of the internal carotid artery. Neurology 12:588-590, 1962

14. Hills J, Sament S: Bilateral agenesis of the internal

carotid artery associated with cardiac and other anomalies. Neurology 18:142-146, 1968

15. Lie TA: Congenital Anomalies of the Carotid Arteries. Amsterdam, Excertpa Medica Foundation, 1968

16. Smith DR Jr, Nelson JS, Dooley JM: Bilateral "hypo- plasia" of the internal carotid arteries. Neurology 18:1149- 1156, 1968

17. L'hermitte F, Gautier JD, Poirier J, et al: Hypoplasia of the internal carotid artery. Neurology 18:439-446, 1968

18. Alpers B J, Barry RG, Paddison RM: Anatomical studies of the circle of Willis in normal brain. AMA Arch Neurol Psychol 81:409-418, 1959

19. Hodes PJ, Campoy F, Riggs HE, et al: Cerebral angiography: fundamentals in anatomy and physiology. Am J Roengeno170:61-81, 1953

20. Godinov VM: Arterial system of the brain. Am J Phys Anthropol 13:359-388, 1929

21. Fawcell E, Blachford JV: The circle of Willis: an examination of 700 specimens. J Anat Physiol 40:63-70, 1906

22. Fetterman GH, Moran T J: Anomalies of the circle of Willis in relation to cerebral softening. Arch Pathol 32:251- 257, 1941

23. General Practitioners' Record: An analysis of the clinical records of some general practices during the period April 1952 to March 1954 General Register Office Studies on Medical and Population Subjects No. 9 London, H.M. Stationery Office, 1956