Congenital
Diaphragmatic
Hernia Repair on ECMO
By Robert H. Connors, Tom Tracy, Jr, Patrick V. Bailey, Barbara Kountzman, and Thomas R. Weber St Louis, Missouri 0 Congenital diaphragmatic hernia (CDH) with severe respiratory failure in the first few hours of life continues to be associated with significant mortality. Extracorporeal membrane oxygenation (ECMO) has been successfully used postoperatively to reverse the effects of severe pulmonary hypertension. Since 1964, ECMO has been required in 27 of the patients we treated with CDH. This report describes our experience with six very high-risk patients placed on ECMO prior to the operation who subsequently underwent repair of their diaphragmatic hernias while on ECMO. Two patients presented in extremis, unlikely to survive initial operative repair, and were placed on ECMO prior to the operation. All six patients had immediate respiratory distress after birth with mean Apgars of 2.3 and 3.7. The best pre-ECMO arterial blood gas (postductal) showed mean + SEM values of 6.97 + 0.1; PO, = 54.6 c 5.9; PCO, = 79.5 f 16.9. Immediately prior to ECMO. the mean f SEM ventilatory index (VI = rate x mean airway pressure) was 1,233 * 44, with a mean pH of 7.17 f 0.05; PO, = 32 + 2.9; PCO, = 59 * 5.3 and a mean AaDO, of 622 + 4.6. The timing of the operative repair averaged 25 hours following initiation of ECMO. Three right-sided and three left-sided hernias were treated. Four were repaired through an abdominal approach, and two via thoracotomy; four required a Gortex patch closure. Postoperative bleeding was not a major problem in these heparinized patients. Four of these six patients survived, and follow-up of 2 months to 3 years shows no significant respiratory compromise. Based on this initial experience, we feel that repair of CDH can be done safely while the patient is on ECMO. This approach avoids perioperative barotrauma and protects the patient from the pulmonary function deterioration and worsening pulmonary hypertension that often is associated with operative repair. This option should be considered for high-risk patients with CDH who deteriorate during prolonged preoperative treatment and for those who present in extremis. @ 1990 by W.B. Saunders Company. INDEX WORDS: Congenital diaphragmatic hernia: ECMO.
C
ONGENITAL diaphragmatic hernia (CDH) presenting with severe respiratory failure in the first few hours of life continues to be associated with significant mortality. Most series report a 50% survival rate, which has not improved despite innovations in ventilator and pulmonary vasodilator therapy.’ Currently, centers using extracorporeal membrane oxygenation (ECMO) in selected high-risk newborns with CDH have improved the survival after surgical repair of the diaphragm to greater than 70%.2-6 Experimental and therapeutic efforts are targeted at further reducing the mortality from refractory pulmonary hypertension and critical pulmonary hypoplasia. One new treatment strategy uses delayed operation in Journal of Pediatric Surgery,
Vol 25, No
10
(October),
1990: pp 1043-1047
certain babies. It is anticipated that a period of stabilization and resuscitation will make them better operative candidates, and reduce the incidence of postoperative deterioration.’ Despite these stabilizing maneuvers, a group of patients still die from complications of pulmonary hypertension. Another group of potentially salvageable infants die before or during surgical correction of a diaphragmatic hernia. Preoperative ECMO may have a significant impact on the treatment of these groups of infants with critical CDH. This report describes our experience with six very high-risk patients placed on ECMO prior to surgery who subsequently underwent repair of their diaphragmatic hernias while on ECMO. The first two of these patients presented in extremis. They appeared unlikely to survive emergency operative repair and were placed immediately on ECMO. Four additional patients deteriorated during preoperative treatment and were placed on ECMO prior to operation. MATERIALS AND METHODS
Thirty-seven newborns with CDHs symptomatic in the first few hours of life were treated at Cardinal Glennon Children’s Hospital, St Louis University Medical Center between January 1984 and May 1989. Twenty-three presented initially to this hospital, and I4 were transferred from another hospital for failure of conventional therapy following repair of their hernia. Twenty-seven of the babies were treated with ECMO. This group consisted of 14 infants transferred to us following repair and 13 babies treated initially by our group. ECMO was initiated prior to surgery in six of these patients. The techniques for newborn venoarterial ECMO using carotid-jugular cannulation have been previously reported.’ During the period from May 1988 to May 1989, selected newborns with CDH were treated with medical management for variable periods preoperatively in an attempt to improve outcome. Following stabilization and clinical improvement, elective operative repair was undertaken. If the baby had severe deterioration during preoperative treatment and met none of the standard exclusion criteria, ECMO was initiated. Exclusion criteria for ECMO treatment were weight less than 1,800 g, gestational age less than 35 weeks, intracranial hemorrhage, or the presence of other lethal anomalies.
From the Division of Pediatric Surgery. Department of Surgery, Cardinal Glennon Children’s Hospital. St Louis University Medical Center, St Louis. MO. Presented at the 38th Annual Meeting of the Surgkal Section of the American Academy of Pediatrics, Chicago, Illinois. October 21-23.1989. Address reprint requests to Robert H. Connors, IUD, Division of Pediatric Surgery, Cardinal Glennon Children’s Hospital, 1465 S Grand Blvd. St Louis. MO 63104. 0 1990 by W.B. Saunders Company. 0022-3468/90/2510-0007$03.00/O
1043
1044
CONNORS ET AL
Table 1. Patient Data Apgars
ABG With Best PostductalPO,
PatientNo.
sex
Weight (kg)
GestationalAge (wk)
HerniaSide
l-minute
5-minute
PCO,
PO,
1
M
3.1
40
R
0
2
6.63
156
68
2
M
2.2
34
R
1
2
7.11
58
33
3
F
3.0
40
R
2
5
6.73
95
64
4
M
4.1
42
L
6
5
7.34
45
69
5
F
3.0
38
L
3
4
6.94
71
51
6
F
1.8
37
L
2
4
7.11
52
44
PH
Abbreviations: ABG, arterial blood gas; M. male: F, female; R. right: L, left
Repair of the CDH on ECMO was undertaken when the clinical course stabilized within an 1% to 30-hour period. Electrocautery was used extensively in these heparinized patients. Patch repairs of the diaphragm, using l-mm polytetrafluoroethylene was preferred to the use of tissue flaps or using chest wall for the closure of large hernias. A chest tube was left on the side of the repair and no abdominal drains were placed. One repair was performed in the intensive care unit and five newborns were moved to the operating room for repair while on bypass. For this report, the clinical courses of all patients with CDH requiring ECMO treatment were reviewed. Data examined for each patient included gestational age, birth weight, Apgar scores, arterial blood gases (ABGs), ventilation parameters, age ECMO was initiated, age of diaphragm repair, the use of prosthetic material for diaphragm repair, total time on the ventilator, and time in the hospital. Details of ECMO treatment included duration of ECMO, blood replacement, and any complications. The Student’s t test was performed to compare this selected group of six patients repaired on ECMO with the remaining 21 patients with CDH requiring ECMO following hernia repair. RESULTS
Initial presenting data are charted in Table 1 and analyzed in Table 2. Three boys and three girls presented with respiratory distress immediately at birth and were diagnosed as having a CDH. Five of the babies were very difficult to ventilate and all had poor oxygenation. None attained a postductal PO, greater than 69 mm Hg despite receiving vigorous mechanical ventilation with 100% oxygen. Significant differences are apparent when these six patients are compared with the remaining group of 21 patients treated with ECMO following operative hernia repair. Apgar scores are lower in the preoperative ECMO group, with the Sminute score being statistically decreased. The ABG recording the best PO, prior to ECMO treatment is listed for each of the six patients. They are significantly worse than those of the rest of the group.
The condition of the patients at the time that ECMO was initiated is noted in Table 3. Despite attempted delay for stabilization in the last four patients, two of these four deteriorated rapidly and were placed on ECMO early in their hospitalization. Although the ABGs immediately prior to ECMO are somewhat worse in the study group, these differences did not reach statistical significance. Details of the ECMO treatment of these patients is summarized in Table 4. Every effort was made to closely control activated clotting time (ACT) and to maintain a platelet count of 100,000. There was a small increase in blood replacement following surgery in five of the patients. The average rate of blood replacement in this group was slightly higher, but not statistically different from the blood replacement in the patients treated with ECMO following CDH repair (Table 2). Four of the six patients survived. One died of a massive intracranial hemorrhage and another of unremitting pulmonary hypertension. Survivors often required prolonged mechanical ventilation and long hospitalizations. Ventilator time averaged 463 hours and ranged from 150 to 768 hours. Hospitalization for these infants ranged from 46 to 92 days with a mean of 68.5 days. Three of the four survivors underwent patch closure of the diaphragm, and each of these three required surgical treatment for severe gastroesophageal reflux prior to discharge. All four surviving patients are doing well. The oldest, who is over 3 years old, initially ate poorly and was sustained for many months by gastrostomy feedings. He is now thriving and his only problem is vision difficulty in the left eye, requiring glasses. The other
Table 2. Analysis of Data ABG With Best PostductalPO,
Apears l-minute
5-minute
Group 1 Group 2
2.3 * .94
3.7 f .56
6.97
k .ll
79.5
PCO,
3.9 f .51
5.3 f .43
7.40
* .04
35.1
P value
NS
.05
PH
Diaphragmatic
Hernia Repair on ECMO
By Robert H. Connors, Tom Tracy, Jr, Patrick V. Bailey, Barbara Kountzman, and Thomas R. Weber St Louis, Missouri 0 Congenital diaphragmatic hernia (CDH) with severe respiratory failure in the first few hours of life continues to be associated with significant mortality. Extracorporeal membrane oxygenation (ECMO) has been successfully used postoperatively to reverse the effects of severe pulmonary hypertension. Since 1964, ECMO has been required in 27 of the patients we treated with CDH. This report describes our experience with six very high-risk patients placed on ECMO prior to the operation who subsequently underwent repair of their diaphragmatic hernias while on ECMO. Two patients presented in extremis, unlikely to survive initial operative repair, and were placed on ECMO prior to the operation. All six patients had immediate respiratory distress after birth with mean Apgars of 2.3 and 3.7. The best pre-ECMO arterial blood gas (postductal) showed mean + SEM values of 6.97 + 0.1; PO, = 54.6 c 5.9; PCO, = 79.5 f 16.9. Immediately prior to ECMO. the mean f SEM ventilatory index (VI = rate x mean airway pressure) was 1,233 * 44, with a mean pH of 7.17 f 0.05; PO, = 32 + 2.9; PCO, = 59 * 5.3 and a mean AaDO, of 622 + 4.6. The timing of the operative repair averaged 25 hours following initiation of ECMO. Three right-sided and three left-sided hernias were treated. Four were repaired through an abdominal approach, and two via thoracotomy; four required a Gortex patch closure. Postoperative bleeding was not a major problem in these heparinized patients. Four of these six patients survived, and follow-up of 2 months to 3 years shows no significant respiratory compromise. Based on this initial experience, we feel that repair of CDH can be done safely while the patient is on ECMO. This approach avoids perioperative barotrauma and protects the patient from the pulmonary function deterioration and worsening pulmonary hypertension that often is associated with operative repair. This option should be considered for high-risk patients with CDH who deteriorate during prolonged preoperative treatment and for those who present in extremis. @ 1990 by W.B. Saunders Company. INDEX WORDS: Congenital diaphragmatic hernia: ECMO.
C
ONGENITAL diaphragmatic hernia (CDH) presenting with severe respiratory failure in the first few hours of life continues to be associated with significant mortality. Most series report a 50% survival rate, which has not improved despite innovations in ventilator and pulmonary vasodilator therapy.’ Currently, centers using extracorporeal membrane oxygenation (ECMO) in selected high-risk newborns with CDH have improved the survival after surgical repair of the diaphragm to greater than 70%.2-6 Experimental and therapeutic efforts are targeted at further reducing the mortality from refractory pulmonary hypertension and critical pulmonary hypoplasia. One new treatment strategy uses delayed operation in Journal of Pediatric Surgery,
Vol 25, No
10
(October),
1990: pp 1043-1047
certain babies. It is anticipated that a period of stabilization and resuscitation will make them better operative candidates, and reduce the incidence of postoperative deterioration.’ Despite these stabilizing maneuvers, a group of patients still die from complications of pulmonary hypertension. Another group of potentially salvageable infants die before or during surgical correction of a diaphragmatic hernia. Preoperative ECMO may have a significant impact on the treatment of these groups of infants with critical CDH. This report describes our experience with six very high-risk patients placed on ECMO prior to surgery who subsequently underwent repair of their diaphragmatic hernias while on ECMO. The first two of these patients presented in extremis. They appeared unlikely to survive emergency operative repair and were placed immediately on ECMO. Four additional patients deteriorated during preoperative treatment and were placed on ECMO prior to operation. MATERIALS AND METHODS
Thirty-seven newborns with CDHs symptomatic in the first few hours of life were treated at Cardinal Glennon Children’s Hospital, St Louis University Medical Center between January 1984 and May 1989. Twenty-three presented initially to this hospital, and I4 were transferred from another hospital for failure of conventional therapy following repair of their hernia. Twenty-seven of the babies were treated with ECMO. This group consisted of 14 infants transferred to us following repair and 13 babies treated initially by our group. ECMO was initiated prior to surgery in six of these patients. The techniques for newborn venoarterial ECMO using carotid-jugular cannulation have been previously reported.’ During the period from May 1988 to May 1989, selected newborns with CDH were treated with medical management for variable periods preoperatively in an attempt to improve outcome. Following stabilization and clinical improvement, elective operative repair was undertaken. If the baby had severe deterioration during preoperative treatment and met none of the standard exclusion criteria, ECMO was initiated. Exclusion criteria for ECMO treatment were weight less than 1,800 g, gestational age less than 35 weeks, intracranial hemorrhage, or the presence of other lethal anomalies.
From the Division of Pediatric Surgery. Department of Surgery, Cardinal Glennon Children’s Hospital. St Louis University Medical Center, St Louis. MO. Presented at the 38th Annual Meeting of the Surgkal Section of the American Academy of Pediatrics, Chicago, Illinois. October 21-23.1989. Address reprint requests to Robert H. Connors, IUD, Division of Pediatric Surgery, Cardinal Glennon Children’s Hospital, 1465 S Grand Blvd. St Louis. MO 63104. 0 1990 by W.B. Saunders Company. 0022-3468/90/2510-0007$03.00/O
1043
1044
CONNORS ET AL
Table 1. Patient Data Apgars
ABG With Best PostductalPO,
PatientNo.
sex
Weight (kg)
GestationalAge (wk)
HerniaSide
l-minute
5-minute
PCO,
PO,
1
M
3.1
40
R
0
2
6.63
156
68
2
M
2.2
34
R
1
2
7.11
58
33
3
F
3.0
40
R
2
5
6.73
95
64
4
M
4.1
42
L
6
5
7.34
45
69
5
F
3.0
38
L
3
4
6.94
71
51
6
F
1.8
37
L
2
4
7.11
52
44
PH
Abbreviations: ABG, arterial blood gas; M. male: F, female; R. right: L, left
Repair of the CDH on ECMO was undertaken when the clinical course stabilized within an 1% to 30-hour period. Electrocautery was used extensively in these heparinized patients. Patch repairs of the diaphragm, using l-mm polytetrafluoroethylene was preferred to the use of tissue flaps or using chest wall for the closure of large hernias. A chest tube was left on the side of the repair and no abdominal drains were placed. One repair was performed in the intensive care unit and five newborns were moved to the operating room for repair while on bypass. For this report, the clinical courses of all patients with CDH requiring ECMO treatment were reviewed. Data examined for each patient included gestational age, birth weight, Apgar scores, arterial blood gases (ABGs), ventilation parameters, age ECMO was initiated, age of diaphragm repair, the use of prosthetic material for diaphragm repair, total time on the ventilator, and time in the hospital. Details of ECMO treatment included duration of ECMO, blood replacement, and any complications. The Student’s t test was performed to compare this selected group of six patients repaired on ECMO with the remaining 21 patients with CDH requiring ECMO following hernia repair. RESULTS
Initial presenting data are charted in Table 1 and analyzed in Table 2. Three boys and three girls presented with respiratory distress immediately at birth and were diagnosed as having a CDH. Five of the babies were very difficult to ventilate and all had poor oxygenation. None attained a postductal PO, greater than 69 mm Hg despite receiving vigorous mechanical ventilation with 100% oxygen. Significant differences are apparent when these six patients are compared with the remaining group of 21 patients treated with ECMO following operative hernia repair. Apgar scores are lower in the preoperative ECMO group, with the Sminute score being statistically decreased. The ABG recording the best PO, prior to ECMO treatment is listed for each of the six patients. They are significantly worse than those of the rest of the group.
The condition of the patients at the time that ECMO was initiated is noted in Table 3. Despite attempted delay for stabilization in the last four patients, two of these four deteriorated rapidly and were placed on ECMO early in their hospitalization. Although the ABGs immediately prior to ECMO are somewhat worse in the study group, these differences did not reach statistical significance. Details of the ECMO treatment of these patients is summarized in Table 4. Every effort was made to closely control activated clotting time (ACT) and to maintain a platelet count of 100,000. There was a small increase in blood replacement following surgery in five of the patients. The average rate of blood replacement in this group was slightly higher, but not statistically different from the blood replacement in the patients treated with ECMO following CDH repair (Table 2). Four of the six patients survived. One died of a massive intracranial hemorrhage and another of unremitting pulmonary hypertension. Survivors often required prolonged mechanical ventilation and long hospitalizations. Ventilator time averaged 463 hours and ranged from 150 to 768 hours. Hospitalization for these infants ranged from 46 to 92 days with a mean of 68.5 days. Three of the four survivors underwent patch closure of the diaphragm, and each of these three required surgical treatment for severe gastroesophageal reflux prior to discharge. All four surviving patients are doing well. The oldest, who is over 3 years old, initially ate poorly and was sustained for many months by gastrostomy feedings. He is now thriving and his only problem is vision difficulty in the left eye, requiring glasses. The other
Table 2. Analysis of Data ABG With Best PostductalPO,
Apears l-minute
5-minute
Group 1 Group 2
2.3 * .94
3.7 f .56
6.97
k .ll
79.5
PCO,
3.9 f .51
5.3 f .43
7.40
* .04
35.1
P value
NS
.05
PH
