Major Surgical
Intervention During Extracorporeal Membrane Oxygenation
By James B. Atkinson,
Hiroaki Kitagawa,
and Bridget Humphries
Los Angeles, California 0 Of 135 patients treated with extracorporeal membrane oxygenation (ECMO) between January 1987 and December 1989, 19 (14.0%) patients underwent surgical procedures while on ECMO. Thirteen (68%) patients had operations related to hemorrhage, including cannula site (6), mediastinal (l), hemoperitoneum (3). and hemothorax (3). Six of 13 patients required repetitive operations for bleeding; 4 of 6 died. Six (35%) patients had operations for congenital pathology including patent ductus (PDA) ligation (2). repair of transposition of the great vessels (2). repair of coarctation (1). and repair of congenital diaphragmatic hernia (3). One patient had multiple simultaneous procedures performed. Of these 6 patients, 4 were decannulated immediately and 2 were decannulated within 28 hours following surgery without any bleeding complications. Fifteen of 19 patients were operated on in the neonatal intensive care unit. The 4 remaining patients required transport on ECMO to the surgical suite. Thirteen of the 19 patients requiring surgical intervention on ECMO survived. In the 13 survivors, the mean time to decannulation postoperation was 45 hours, and in those that died it was 90 hours. Our experience suggests that surgical intervention while on ECMO is technically feasible with the best results achieved when rapid discontinuation of ECMO can be accomplished postoperatively. Due to this fact major surgical intervention should be postponed if possible until near the conclusion of the ECMO therapy. Copyright 8 1992 by W.8. Saunders Company INDEX
WORDS:
Extracorporeal
membrane
oxygenation
(ECMO).
E
XTRACORPOREAL membrane oxygenation (ECMO) is now widely available for the support of neonates with severe respiratory distress.ll’ Major surgical intervention may be required during ECMO due to complications or to correct congenital pathology.” This report summarizes the operative experience resulting from the sequential care of 135 infants treated consecutively for respiratory distress while being supported with ECMO.
MATERIALS
AND METHODS
A retrospective review of all charts for infants treated with ECMO from 1987 to 1989 was performed. Data from each chart were recorded in those infants who required any surgical procedure other than cannulation and decannulation. The infants requiring surgical intervention were analyzed for details of diagnosis, ECMO indications, surgical indications, type of surgery, complications, and outcome. Prospectively, all infants prior to surgical intervention were prepared by lowering the activated clotting time (ACT) to 175 to 200 seconds, transfusing platelets to 100,000/mm3, and maintaining blood pressure less than 90 mm Hg. These preparatory measures were also maintained intraoperatively and postoperatively to the conclusion of the ECMO run. The wounds JournalofPediatric Surgery, Vol27, No 9 (September), 1992: pp 1197-l 198
Table 1. The Etiology of Hemorrhagic Complications Operative
Indication
Precipitating
Factors
Cannula site bleeding
6
Cannula repositioning
Mediastinal bleeding
1
Sternotomy for cannulation
1
Hemoperitoneum
3
CDH repair
2
2
Paracentesis with omental laceration (Pre-ECMO) Hemothorax
were sealed with a fibrin
3
1
Thoracostomy
2
Lobectomy (CCAM)
1
glue consisting
of cryoprecipitate
and
thrombin.
RESULTS A total of 135 patients were treated with ECMO for severe respiratory insufficiency between January 1987 and November 1989. A total of 19 patients (14%) were identified who underwent surgical procedures in addition to cannulation and decannulation. The patients were grouped according to the presence of bleeding or congenital pathology as the major surgical indication. In group I (13119) surgery was to control primary hemorrhagic complications. The sites of bleeding and precipitating factors are listed in Table 1. Bleeding as an indication for surgery was spontaneous in only 4 of 13 patients. With the exception of cannula related procedures, the precipitating events took place prior to heparinization and initiation of ECMO. The indications for surgery due to congenital pathology are also listed in Fig 1. Overall survival for surgical intervention during ECMO was 68% (13/19 patients). The survival of infants was similar in the two groups (69% v 67%) whether the indication was for bleeding or correction of congenital pathology. Multiple operations were required only in group I in 2 of 6 patients in whom the surgical indication was bleeding and multiple procedures had a very poor prognosis (survival = 33%). Surgical procedures were performed in the neonatal intensive care unit in 15 of 19 patients. Correction of
From the Division of Pediatric Surgery, Childrens Hospital of Los Angeles, Universiry of Southern California, Los Angeles, CA. Date accepted: April 23, 1991. Address reprint requests to James B. Atkinson, MD, Division of Pediatric Surgery, Childrens Hospital of Los Angeles, 4650 Sunset Blvd, Los Angeles, CA 90027. Copyright 0 1992 by W.B. Saunders Company 0022-3468/9212709-0007$03.00l0 1197
1198
During 19/135
ATKINSON, KITAGAWA, AND HUMPHRIES
ECMO (14%)
L
Hemothorax
3
r
PDA Ligation
2
Transposition .oll(l panent Ilad multIpIe procedures
Fig 1. There were 19 patients identified who underwent secondary surgical procedures. The patients are grouped by the presence of hemorrhage or congenital pathology as the primary surgical indication.
congenital heart disease (3) and a single case of liver laceration complicating CDH were the reasons for transport to the operating suite. Decannulation and reversal of heparinization was possible at a mean of 45 hours postsurgery in survivors and 90 hours in those that eventually died. DISCUSSION
Major surgical intervention in a heparinized neonate on prolonged extracorporeal support can be both a challenging and rewarding effort. The infants surgical indications can be divided into those with congenital pathology that must be corrected to allow weaning from bypass and those with bleeding requiring intervention for hemostasis4 In spite of an anticipated high mortality for operations in a critically ill neonate who is fully anticoagulated, successful operation was performed in 68% (13/ 19) of these patients. Success rates were similar whether the initial indication was hemorrhage or congenital pathology. Morbid-
Fig 2. The plan of management outlined is based on the indication for intervention and urgency based on the rate of bleeding.
ity was primarily due postoperatively to bleeding and only 33% of infants bleeding significantly enough to require repetitive operations survived. In the majority of infants with hemorrhage as the primary indication for surgery, the cause of bleeding could be traced to a pre-ECMO manipulation. These infants had the highest probability of persistent bleeding and an unsuccessful outcome. An algorithm for the management of patients bleeding while on ECMO is presented in Fig 2. The results reported support the use of this protocol to safely perform major surgery in this group of infants. It is concluded that surgical intervention while on ECMO is technically feasible. The best results can be achieved when surgical intervention can be postponed until respiratory recovery has occurred allowing decannulation soon after surgery. Because the requirement for repetitive surgery carries a high mortality (67%), postponing surgery to as late as possible in the ECMO run allows treatment by reversal of anticoagulation. This recommendation must be weighed against the severity of blood loss that is occurring as suggested in the algorithm.
REFERENCES 1. Bartlett RH, Gazzaniga AB, Jefferies R, et al: Extracorporeal membrane oxygenation (ECMO) cardiopulmonary support in infancy. Trans Am Sot Artif Intern Organs 20:80-881976 2. Bartlett RH, Toomasian J, Roloff D, et al: Extracorporeal membrane oxygenation (ECMO) in neonatal respiratory failure. Ann Surg 2041236-2451986 3. Langham MR, Krnmmel TM, Bartlett RH, et al: Mortality
with extracorporeal membrane oxygenation following repair of congenital diaphragmatic hernia in 93 infants. J Pediatr Surg 22:1150-1154, 1987 4. Sell LL, Whittlesey GC, Yedlin St, et al: Hemorrhagic complications during extracorporeal membrane oxygenation: Prevention and treatment. J Pediatr Surg 21:1087-1091, 1986
Intervention During Extracorporeal Membrane Oxygenation
By James B. Atkinson,
Hiroaki Kitagawa,
and Bridget Humphries
Los Angeles, California 0 Of 135 patients treated with extracorporeal membrane oxygenation (ECMO) between January 1987 and December 1989, 19 (14.0%) patients underwent surgical procedures while on ECMO. Thirteen (68%) patients had operations related to hemorrhage, including cannula site (6), mediastinal (l), hemoperitoneum (3). and hemothorax (3). Six of 13 patients required repetitive operations for bleeding; 4 of 6 died. Six (35%) patients had operations for congenital pathology including patent ductus (PDA) ligation (2). repair of transposition of the great vessels (2). repair of coarctation (1). and repair of congenital diaphragmatic hernia (3). One patient had multiple simultaneous procedures performed. Of these 6 patients, 4 were decannulated immediately and 2 were decannulated within 28 hours following surgery without any bleeding complications. Fifteen of 19 patients were operated on in the neonatal intensive care unit. The 4 remaining patients required transport on ECMO to the surgical suite. Thirteen of the 19 patients requiring surgical intervention on ECMO survived. In the 13 survivors, the mean time to decannulation postoperation was 45 hours, and in those that died it was 90 hours. Our experience suggests that surgical intervention while on ECMO is technically feasible with the best results achieved when rapid discontinuation of ECMO can be accomplished postoperatively. Due to this fact major surgical intervention should be postponed if possible until near the conclusion of the ECMO therapy. Copyright 8 1992 by W.8. Saunders Company INDEX
WORDS:
Extracorporeal
membrane
oxygenation
(ECMO).
E
XTRACORPOREAL membrane oxygenation (ECMO) is now widely available for the support of neonates with severe respiratory distress.ll’ Major surgical intervention may be required during ECMO due to complications or to correct congenital pathology.” This report summarizes the operative experience resulting from the sequential care of 135 infants treated consecutively for respiratory distress while being supported with ECMO.
MATERIALS
AND METHODS
A retrospective review of all charts for infants treated with ECMO from 1987 to 1989 was performed. Data from each chart were recorded in those infants who required any surgical procedure other than cannulation and decannulation. The infants requiring surgical intervention were analyzed for details of diagnosis, ECMO indications, surgical indications, type of surgery, complications, and outcome. Prospectively, all infants prior to surgical intervention were prepared by lowering the activated clotting time (ACT) to 175 to 200 seconds, transfusing platelets to 100,000/mm3, and maintaining blood pressure less than 90 mm Hg. These preparatory measures were also maintained intraoperatively and postoperatively to the conclusion of the ECMO run. The wounds JournalofPediatric Surgery, Vol27, No 9 (September), 1992: pp 1197-l 198
Table 1. The Etiology of Hemorrhagic Complications Operative
Indication
Precipitating
Factors
Cannula site bleeding
6
Cannula repositioning
Mediastinal bleeding
1
Sternotomy for cannulation
1
Hemoperitoneum
3
CDH repair
2
2
Paracentesis with omental laceration (Pre-ECMO) Hemothorax
were sealed with a fibrin
3
1
Thoracostomy
2
Lobectomy (CCAM)
1
glue consisting
of cryoprecipitate
and
thrombin.
RESULTS A total of 135 patients were treated with ECMO for severe respiratory insufficiency between January 1987 and November 1989. A total of 19 patients (14%) were identified who underwent surgical procedures in addition to cannulation and decannulation. The patients were grouped according to the presence of bleeding or congenital pathology as the major surgical indication. In group I (13119) surgery was to control primary hemorrhagic complications. The sites of bleeding and precipitating factors are listed in Table 1. Bleeding as an indication for surgery was spontaneous in only 4 of 13 patients. With the exception of cannula related procedures, the precipitating events took place prior to heparinization and initiation of ECMO. The indications for surgery due to congenital pathology are also listed in Fig 1. Overall survival for surgical intervention during ECMO was 68% (13/19 patients). The survival of infants was similar in the two groups (69% v 67%) whether the indication was for bleeding or correction of congenital pathology. Multiple operations were required only in group I in 2 of 6 patients in whom the surgical indication was bleeding and multiple procedures had a very poor prognosis (survival = 33%). Surgical procedures were performed in the neonatal intensive care unit in 15 of 19 patients. Correction of
From the Division of Pediatric Surgery, Childrens Hospital of Los Angeles, Universiry of Southern California, Los Angeles, CA. Date accepted: April 23, 1991. Address reprint requests to James B. Atkinson, MD, Division of Pediatric Surgery, Childrens Hospital of Los Angeles, 4650 Sunset Blvd, Los Angeles, CA 90027. Copyright 0 1992 by W.B. Saunders Company 0022-3468/9212709-0007$03.00l0 1197
1198
During 19/135
ATKINSON, KITAGAWA, AND HUMPHRIES
ECMO (14%)
L
Hemothorax
3
r
PDA Ligation
2
Transposition .oll(l panent Ilad multIpIe procedures
Fig 1. There were 19 patients identified who underwent secondary surgical procedures. The patients are grouped by the presence of hemorrhage or congenital pathology as the primary surgical indication.
congenital heart disease (3) and a single case of liver laceration complicating CDH were the reasons for transport to the operating suite. Decannulation and reversal of heparinization was possible at a mean of 45 hours postsurgery in survivors and 90 hours in those that eventually died. DISCUSSION
Major surgical intervention in a heparinized neonate on prolonged extracorporeal support can be both a challenging and rewarding effort. The infants surgical indications can be divided into those with congenital pathology that must be corrected to allow weaning from bypass and those with bleeding requiring intervention for hemostasis4 In spite of an anticipated high mortality for operations in a critically ill neonate who is fully anticoagulated, successful operation was performed in 68% (13/ 19) of these patients. Success rates were similar whether the initial indication was hemorrhage or congenital pathology. Morbid-
Fig 2. The plan of management outlined is based on the indication for intervention and urgency based on the rate of bleeding.
ity was primarily due postoperatively to bleeding and only 33% of infants bleeding significantly enough to require repetitive operations survived. In the majority of infants with hemorrhage as the primary indication for surgery, the cause of bleeding could be traced to a pre-ECMO manipulation. These infants had the highest probability of persistent bleeding and an unsuccessful outcome. An algorithm for the management of patients bleeding while on ECMO is presented in Fig 2. The results reported support the use of this protocol to safely perform major surgery in this group of infants. It is concluded that surgical intervention while on ECMO is technically feasible. The best results can be achieved when surgical intervention can be postponed until respiratory recovery has occurred allowing decannulation soon after surgery. Because the requirement for repetitive surgery carries a high mortality (67%), postponing surgery to as late as possible in the ECMO run allows treatment by reversal of anticoagulation. This recommendation must be weighed against the severity of blood loss that is occurring as suggested in the algorithm.
REFERENCES 1. Bartlett RH, Gazzaniga AB, Jefferies R, et al: Extracorporeal membrane oxygenation (ECMO) cardiopulmonary support in infancy. Trans Am Sot Artif Intern Organs 20:80-881976 2. Bartlett RH, Toomasian J, Roloff D, et al: Extracorporeal membrane oxygenation (ECMO) in neonatal respiratory failure. Ann Surg 2041236-2451986 3. Langham MR, Krnmmel TM, Bartlett RH, et al: Mortality
with extracorporeal membrane oxygenation following repair of congenital diaphragmatic hernia in 93 infants. J Pediatr Surg 22:1150-1154, 1987 4. Sell LL, Whittlesey GC, Yedlin St, et al: Hemorrhagic complications during extracorporeal membrane oxygenation: Prevention and treatment. J Pediatr Surg 21:1087-1091, 1986
