Persistent chylothorax in the pediatric patient, regardless

Transcription

1 Treatment of Refractory Chylothorax With Externalized Pleuroperitoneal Shunts in Children Andrew B. Wolff, BA, Mark L. Silen, MD, Evan R. Kokoska, MD, and Bradley M. Rodgers, MD Department of Surgery, St. Louis University Health Sciences Center, St. Louis, Missouri and Departments of Surgery and Pediatrics, University of Virginia Health System, Charlottesville, Virginia Background. Traditional therapy for refractory chylothorax in the pediatric population has included pleurodesis and thoracic duct ligation. These s are associated with high morbidity and questionable success rates. Methods. We retrospectively reviewed our experience with 15 patients who underwent treatment for chylous effusions using pleuroperitoneal shunts with exteriorized pump chambers. Mean patient age at time of shunt placement was 2.1 (0.1 to 11.5) years and the most common indication (7 of 15) was refractory chylothorax following surgical correction of congenital heart disease. Mean chylothorax duration before shunt placement was 76 (5 to 810) days and shunts were in place for an average of 104 (12 to 365) days. A total of 19 chylous effusions (pleural or pericardial) were treated with shunts. Results. Nine of 11 right-sided chylothoraces, 5 of 6 left-sided chylothoraces, and 2 of 2 chylopericardia resolved with shunt therapy (84% total). Pleuroperitoneal shunting failed to clear the effusion in 3 children. There were six episodes of shunt malfunction that were repaired and two episodes of infection. Inguinal or umbilical hernia developed in 4 patients. Conclusions. Externalized pleuroperitoneal shunting is a safe, effective, and minimally invasive treatment for children with refractory chylous effusions. (Ann Thorac Surg 1999;68:1053 7) 1999 by The Society of Thoracic Surgeons Persistent chylothorax in the pediatric patient, regardless of its etiology, has always presented a difficult clinical challenge. Prolonged severe chylothorax has been shown to result in compromise of the immune system, nutritional depletion, metabolic instability, electrolyte imbalance, impaired cardiac and respiratory functions, and death [1 4]. The morbidity and mortality resulting from chylothorax is a result of local, metabolic, and immunologic effects. Locally, chyle leakage into the pleural cavity causes compression of the ipsilateral lung. If the accumulation is sufficiently severe, mediastinal shift can occur, compromising the contralateral lung and impairing cardiac function. The metabolic and immunologic effects of persistent chylothorax are caused by the loss of chyle that is comprised mostly of T lymphocytes, chylomicrons from long-chain fatty acids, and an electrolytic composition similar to plasma. Upon recognition of chylothorax, conservative therapy is indicated, consisting of restriction of oral fat intake to medium-chain triglycerides or use of total parenteral nutrition with medium-chain triglyceride supplementation. Drainage is accomplished by tube thoracostomy or serial thoracentesis. Conservative therapy often fails and Presented at the Forty-fifth Annual Meeting of the Southern Thoracic Surgical Association, Orlando, FL, Nov 12 14, Address reprint requests to Dr Rodgers, Department of Surgery, University of Virginia Health System, PO Box 10005, Charlottesville, VA 22906; bmr@virginia.edu. is no longer indicated if the patient s condition worsens or fails to improve after 2 weeks [2, 5 7]. Traditional therapy for chylothorax refractory to conservative management has included chemical or talc pleurodesis, pleurectomy, and thoracic duct ligation. Pleurodesis is a painful and the long-term results are unreliable. Ligation of the thoracic duct using either video-assisted thoracoscopy (VATS) or a thoracotomy may fail as a result of anatomic variations of the thoracic duct or failure to identify the sites of chylous leakage [2, 8]. In 1982, Weese and Schouten [9] reported the successful use of pleuroperitoneal shunts in adults to palliate respiratory symptoms secondary to malignant pleural effusions. In 1983, Azizkhan and colleagues [10] described the treatment of refractory chylothorax with a pleuroperitoneal shunt in neonates. The original version of the shunt was completely subcutaneous with a 1.5-mL pump chamber implanted along the costal margin. However, experience demonstrated that the small volume of the pump chamber necessitated frequent pumping, and that many children became irritable due to pain during chamber compression, discouraging use of this version of the shunt. The current version of the shunt (Denver Biomaterials, Inc, Golden, CO) consists of a flexible 2.5-mL pump chamber containing two one-way valves with 15.5F fenestrated and cuffed pleural and peritoneal catheters on either end (Fig 1). The size of these catheters has proved suitable for use in patients with a wide range of body weights, from 1999 by The Society of Thoracic Surgeons /99/$20.00 Published by Elsevier Science Inc PII S (99)

2 1054 WOLFF ET AL Ann Thorac Surg TREATMENT OF CHYLOTHORAX WITH PLEUROPERITONEAL SHUNTS IN CHILDREN 1999;68: Table 1. Characteristics of Patients Treated With Pleuroperitoneal and/or Pericardialperitoneal Shunts Patient Number (Sex) Primary Diagnosis Age at Shunt Placement Shunt(s) Duration of Chylothorax Prior to Shunt Placement (days) Duration of Shunt Therapy (days) Complications Outcome 1 (M) AV septal defect, s/p AV canal repair 2 (M) Tricuspid and pulmonary atresia, s/p Fontan 3 (F) Tricuspid atresia, s/p Fontan with closure of ASD 4 (M) TGA, s/p Jatene 5 (M) Tricuspid atresia, s/p Fontan 8 months Left None Effusion cleared 5 years Right Broken valve; right empyema; right inguinal hernia 4 years Right Shunt obstruction (2 episodes); shunt replacement required 2.5 months Right Shunt infection; shunt removal 4 years Left Shunt obstruction (2 episodes); bilateral inguinal hernia; umbilical hernia Failure Failure; required decortication Effusion cleared Failure 6 (F) Meconium 1.5 months Right Shunt obstruction Effusion cleared aspiration syndrome, SVC and innominate vein obstruction 7 (F) SVC obstruction 2.5 months Right None Effusion cleared 8 (F) Congenital 27 months Right Fluid loculation Effusion cleared chylothorax 9 (M) SVC obstruction 4 months Right None Effusion cleared Left None Failure; talc pleurodesis 10 (F) VSD, PDA, s/p repair 2 months Left 5 60 Bilateral inguinal hernia Effusion cleared 11 (M) Down s, TEF 23 months Right 7 90 None Effusion cleared 12 (M) TGA, s/p Jatene 11.5 years Right None Effusion cleared Left None Effusion cleared Pericardial Umbilical hernia Effusion cleared 13 (M) Left CDH, s/p 3 months Right Ascitic leak Effusion cleared ECMO 14 (F) Lymphangioma 1.5 months Right None Effusion cleared Pericardial Shunt infection; shunt Effusion cleared removal 15 (F) Left CDH, s/p ECMO 1.5 months Left None Effusion cleared ASD atrial septal defect; AV atrioventricular; CDH congenital diaphragmatic hernia; ECMO extracorporeal membrane oxygenation; PDA patent ductus arteriosus; s/p status post; SVC superior vena cava; TEF tracheoesophageal fistula; TGA transposition of the great arteries; VSD ventricular septal defect. infancy to adolescence. The shunt is implanted with a minimally invasive surgical and allows for chyle to be pumped into the peritoneal cavity where it is reabsorbed without the loss of nutrients, lymphocytes, and electrolytes associated with tube thoracostomy. We report here on the successful use of shunt therapy in two medical centers in pediatric patients with chylothoraces of various etiologies. Material and Methods We reviewed the records of 15 patients treated with pleuro- or pericardialperitoneal shunts (Denver Biomaterials) at The Children s Medical Center at the University of Virginia and at Cardinal Glennon Children s Hospital at Saint Louis University Health Sciences Center between 1988 and Upon recognition of chylo-

3 Ann Thorac Surg WOLFF ET AL 1999;68: TREATMENT OF CHYLOTHORAX WITH PLEUROPERITONEAL SHUNTS IN CHILDREN 1055 thorax or chylopericardium, conservative therapy was undertaken for variable periods of time. Therapy consisted of drainage of the pleural space by thoracentesis or tube thoracostomy, and restriction of oral fat intake to medium-chain triglycerides or the use of total parenteral nutrition. If a patient s condition failed to improve or began to deteriorate, a pleuroperitoneal or pericardialperitoneal shunt was placed. Primary diagnosis, age at shunt placement, location of shunt, duration of chylothorax or chylopericardium prior to and following shunt placement, complications, and overall outcome of shunt treatment are shown in Table 1. Before shunt placement, pleural effusions were allowed to accumulate, most often by clamping the chest tube for 12 to 24 hours preoperatively, to facilitate pleural catheter placement. General anesthesia and endotracheal intubation were used routinely. Correct positioning of the patient is important to facilitate placement of the shunt. The patient was placed in slightly less than a lateral decubitus position allowing access to the affected hemithorax and the abdomen. The entire shunt was immersed in bacitracin solution to confirm shunt patency. The shunt was placed on the skin of the abdomen and thorax and the ultimate positions of the Dacron (C.R. Bard, Haverhill, PA) cuffs, subcutaneous tunnels, and sites for entry into the pleural and peritoneal cavities were carefully designed and drawn on the skin. It is critical that sharp angles in any portion of the catheter be avoided so that mechanical obstruction is eliminated. A short transverse incision was made overlying the seventh or eighth intercostal space in the midaxillary line. From this incision a 5- to 6-cm tunnel was constructed with a clamp and the pleural end of the shunt was delivered, placing the cuff within 1 to 2 cm of the thoracic incision. The pleural catheter was then cut to an appropriate length for the patient and the catheter was tunneled across the intercostal muscle into the posterior hemithorax. The pump was then compressed to confirm free flow of fluid from the pleural space. A short transverse incision was made overlying the rectus muscle in the supraumbilical region. From this incision a 5- to 6-cm subcutaneous tunnel was created with a clamp and the Fig 1. Externalized pleuroperitoneal shunt. Fig 2. Two externalized shunts implanted in an 11-year-old patient (Patient #12) with bilateral chylothoraces and a chylopericardium. The left pleural shunt system is not visible in the photograph. These shunts were successfully removed after 1 year of drainage. peritoneal end of the shunt was drawn through this tunnel. The anterior rectus sheath was divided transversely and the rectus muscle split to expose the posterior fascia and peritoneum. Two concentric purse string sutures of nonabsorbable suture material were placed in the posterior fascia. The peritoneal end of the catheter was trimmed to appropriate length and passed into the peritoneal cavity through the purse string sutures. The purse strings were secured around the catheter and tied to the Dacron cuff of the catheter. All surgical incisions were closed in multiple layers to avoid potential exposure and infection of the shunt. The catheters were secured to the skin at the exit site with nonabsorbable sutures. Every effort should be made to accomplish a watertight seal of peritoneum around the abdominal end of the shunt system. Circumferential ligatures of nonabsorbable suture material are effective and do not risk laceration of the peritoneum accompanying suture ligation. Immediately after placement, the shunt was pumped repeatedly until the chamber no longer filled, indicating that the hemithorax was cleared of fluid. The patients were placed on a daily schedule of pumping based on the

4 1056 WOLFF ET AL Ann Thorac Surg TREATMENT OF CHYLOTHORAX WITH PLEUROPERITONEAL SHUNTS IN CHILDREN 1999;68: Fig 3. Treatment algorithm for pediatric patients diagnosed with chylothorax. volume of fluid produced per day and the parents were instructed on how to use the pump. Usually this involved pumping the chamber every 2 to 4 hours until the chest was empty. Pumping frequency was reduced as the amount of chylous effusion declined, although the parents were instructed to pump the chamber at least twice daily to ensure shunt patency. Shunt therapy was considered successful and elective removal of the shunt was undertaken when serial radiographs showed resolution of the effusion over a 1-month period of time, during which the pump was not compressed. Results Fifteen patients were treated for a total of 19 chylous effusions (11 in the right pleural cavity, 6 in the left pleural cavity, and 2 in the pericardium). Patients ranged in age from 1 month to 11.5 years (mean: 2.1 years; median: 3 months). The most common indication (7 of 15) was refractory chylothorax following surgical correction of congenital heart disease. The duration of the chylothorax prior to shunt placement varied from 5 to 810 days with a mean duration of 76 days and a median duration of 14 days. Fourteen of the 17 pleural effusions (83%) resolved secondary to shunt therapy in an average of 95 days (range: 12 to 365 days). Both of the pericardialperitoneal shunts were associated with successful outcomes. In all cases in which the chylous effusions cleared, the shunts were removed without further complications (Fig 2). Of the three chylothoraces that failed to resolve with shunt therapy, one was left-sided and two were rightsided. One of the failures occurred in a 1-month-old infant with bilateral chylothoraces secondary to vena caval thrombosis. This patient was treated with bilateral pleuroperitoneal shunts. The right shunt cleared the right pleural effusion, whereas the left shunt failed to clear the effusion in the left pleural cavity and the patient eventually underwent successful talc pleurodesis on the left side. The other two failures occurred in children ages 4 and 5 years following a Fontan. One of these patients subsequently underwent thoracoscopic thoracic duct ligation, whereas the other eventually required heart transplantation in part for right heart failure. During the course of the shunt therapy, there were a total of six shunt malfunctions that required intervention. Five of the six malfunctions involved shunt or valve obstruction and the other involved a broken pump valve. Shunt obstruction was managed by dividing the afferent limb and flushing the system with sterile saline. Some episodes of shunt obstruction due to omentum were managed by removing the peritoneal end of the shunt, removing the offending omentum from the shunt lumen, and replacing the shunt within the peritoneal cavity. Shunt obstruction resulted in system failure in 2 of 3 patients. Additionally, there were two shunt-associated infections that precipitated shunt removal. No infections resulted in failure of shunt therapy. Four male patients developed inguinal or umbilical hernia during therapy requiring repair. Comment Externalized pleuroperitoneal and pericardial-peritoneal shunts are safe and effective therapies for children with refractory chylous effusions. Pleuroperitoneal shunts alleviate the respiratory symptoms of chylothorax without the nutritional and immunological depletion or metabolic instability associated with long-term tube thoracostomy. This method of treatment allows pleural effusions to resolve without resorting to pleurodesis or thoracic duct ligation. Externalized shunt therapy also empowers parents who often feel helpless when their child is sick by involving them directly in their child s healing process. The pleuroperitoneal shunt with the exteriorized pump chamber has significant advantages over the original subcutaneous version of the shunt. Not only is placement of these shunts technically simpler, but exteriorization of the pump chamber allows for a large volume pump chamber, even in small infants. Because the pump chamber is larger and exteriorized, the pump needs to be compressed less frequently and the compressions are less painful and disturbing to the young patient. There are theoretical disadvantages to the externalized system, including infection and the need for frequent dressing changes. All of the chylous effusions that failed to resolve during the course of shunt therapy were in patients with conditions associated with elevated vena cava or right atrial pressures. Although some researchers have reported success with pleuroperitoneal shunting after Fontan pro-

5 Ann Thorac Surg WOLFF ET AL 1999;68: TREATMENT OF CHYLOTHORAX WITH PLEUROPERITONEAL SHUNTS IN CHILDREN 1057 cedures [11, 12], there is substantial evidence to suggest that chylous effusions in patients with elevated right atrial or vena caval pressures, such as those that often exist after a Fontan or with vena cava thrombosis, are more difficult to manage [10, 11]. In this subset of patients, the efficacy of pleuroperitoneal shunting requires further investigation. We recommend externalized pleuroperitoneal shunts for the therapy of chylous effusions to avoid the problems associated with long-term chyle loss and respiratory compromise. This approach avoids a major operative or pleurodesis, although these treatments may be necessary for patients in whom shunt therapy fails. We propose a general treatment algorithm (Fig 3) that advocates pleuroperitoneal shunting of some refractory chylothoraces after 2 weeks of conservative therapy. Patients in whom a localized injury to the thoracic duct has occurred (eg, penetrating trauma) are likely to benefit from early thoracic duct ligation, rather than from shunt therapy. Shunt therapy is most appropriate for patients in whom the chylous leak is more diffuse (eg, postcardiotomy). References 1. Van Straaten HL, Gerards LJ, Krediet TG. Chylothorax in the neonatal period. Eur J Pediatr 1993;152: Merrigan BA, Winter DC, O Sullivan GC. Chylothorax. Br J Surg 1997;84: Machleder HI, Paulus H. Clinical and immunological alterations observed in patients undergoing long-term thoracic duct drainage. Surgery 1978;84: McWilliams BC, Fan LL, Murphy SA. Transient T-cell depression in postoperative chylothorax. J Pediatr 1981;99: Robinson CL. The management of chylothorax. Ann Thorac Surg 1985;39: Milsom JW, Kron IL, Rheuban KS, Rodgers BM. Chylothorax: an assessment of current surgical management. J Thorac Cardiovasc Surg 1985;89: Marts BC, Naunheim KS, Fiore AC, Pennington DG. Conservative versus surgical management of chylothorax. Am J Surg 1992;164: Ferguson MK, Little AG, Skinner DB. Current concepts in the management of postoperative chylothorax. Ann Thorac Surg 1985;40: Weese JL, Schouten JT. Pleural peritoneal shunts for the treatment of malignant pleural effusions. Surg Gynecol Obstet 1982;154: Azizkhan RG, Canfield J, Alford BA, Rodgers BM. Pleuroperitoneal shunts in the management of neonatal chylothorax. J Pediatr Surg 1983;18: Rheuban KS, Kron IL, Carpenter MA, Gutgesell HP, Rodgers BM. Pleuroperitoneal shunts for refractory chylothorax after operation for congenital heart disease. Ann Thorac Surg 1992;53: Sade RM, Wiles HB. Pleuroperitoneal shunt for persistent pleural drainage after Fontan. J Thorac Cardiovasc Surg 1990;100: DISCUSSION DR G. HOSSEIN ALMASSI (Milwaukee, WI): I enjoyed your presentation and the data regarding this challenging group of children. The question I have is about the 1-month old patient who underwent talc pleurodesis, presumably for effusions that were nonmalignant. I was wondering if you could comment on the advisability of using talc in such young patients and the long-term outlook for these children. DR KOKOSKA: We try to avoid using talc, especially in young patients, because the long-term effects of talc pleurodesis are unknown. Although the results we have seen with talc do not suggest any long-term harmful effects with regard to lung trapping, data are not yet definitively known. DR ROSS M. UNGERLEIDER (Durham, NC): I wonder if you could tell us, maybe I did not quite understand I understand the indications for placement of the shunt quite clearly how do you decide when to remove these shunts? DR KOKOSKA: Whenever the effusions are clear radiographically, we ask the parents to stop pumping or depressing the chamber and then we follow the chest films usually weekly over a 1-month period. Once it has been demonstrated to be clear over a month s time, then we remove the shunt. DR UNGERLEIDER: So some of the shunts can be removed after a couple of weeks; there is flexibility in how long you wait, I take it? DR KOKOSKA: Yes, there is.