Intestinal bypass of the esophagus.

Esophageal replacement by a segment of isoperistaltic ileum with cecum or by transverse or left colon will allow near-normal swallowing for many years. The authors reviewed the course of 59 children who had bypass of their entire esophagus and of four whose distal esophagus was resected and replaced. The follow-up period ranges from 1 to 37 years; in 36 cases, it exceeds 5 years. Thirty children had caustic strictures and 25 had either isolated esophageal atresia or atresia with fistula. Two children with esophageal injury caused by foreign body ingestion and two with congenital strictures also required complete bypass. Four patients required resection and replacement of the distal esophagus only; two had acquired strictures from gastroesophageal reflux, one had varices, and one had a teratoma involving the esophagus. A retrosternal isoperistaltic ileocolic segment is our preference for complete esophageal replacement. Forty-eight patients underwent esophageal reconstruction with this procedure. The esophagus damaged by caustic ingestion was left in place in all patients, without any subsequent problem. The authors have not used the distal esophagus for anastomosis in patients with atresia, because this segment may be abnormal; and, in any case, an isoperistaltic cologastric anastomosis does not reflux. The right or left colon or jejunum was used in the other cases. Three children lost an interposed intestinal segment from necrosis even though the bowel appeared to be well vascularized at the end of the operation. Each patient had successful reconstruction using another type of interposition. An intrathoracic leak occurred in one infant. A cervical anastomotic leak developed in 11 children, and a stricture in 13. Strictures were more common in patients who had caustic burns. Three patients required surgery for adhesive intestinal obstruction. A redundant colon transplant with ulceration, and the herniation of an ileal segment into the pleural cavity with obstruction prompted reoperation in two other patients. There were two deaths early in the series, one of which was secondary to postoperative respiratory arrest. The other death occurred in a child who had a caustic pharyngeal burn and chronic aspiration. All patients were seen in our office recently, or they or their parents were interviewed by phone. All of them are taking all of their nutrition by mouth. Forty-three of the 61 survivors have had no difficulty with swallowing. One required reoperation to enlarge the thoracic inlet. Seventeen other have mild dysphagia that does not require treatment. The patients with esophageal atresia or atresia and fistula consistently have not grown as well as those who required replacement for an acquired condition or injury.

Long-term outcome in children with opsoclonus-myoclonus and ataxia and coincident neuroblastoma.

We reviewed the neurologic and developmental courses in 10 children with opsoclonus-myoclonus ("dancing eyes syndrome") and neuroblastoma. All patients are alive without evidence of neoplastic disease after 8+ to 111+ months of follow-up. All had localized disease and 50% had extraabdominal tumors. Neuroblastomas of nine children had favorable Shimada histologic characteristics, and all tumors had single copies of the N-myc oncogene. After neuroblastoma resection, all patients had persistent opsoclonus-myoclonus or ataxia that responded to therapy with adrenocorticotropic hormone. Nine children had relapses of neurologic symptoms. Three years after resection, six of seven patients with sufficient follow-up were free of symptoms and had discontinued therapy. However, nine children had chronic neurologic deficits, including cognitive and motor delays, language deficits, and behavioral abnormalities. All six patients in educational programs required special assistance. Five children required physical, occupational, or speech therapy. Long-term developmental and cognitive problems should be anticipated in patients with neuroblastoma who have opsoclonus-myoclonus or ataxia or both, and early intervention should be instituted to try to minimize these deficits.

A 25-year experience with renal tumors of childhood.

A retrospective analysis of the medical records of 234 children with renal tumors managed over a 25-year period at the Children's Memorial Medical Center was undertaken to evaluate long-term morbidity and mortality. There was a significant increase in survival over the years of the study. The 5-year survival for patients treated during the period 1985 to 1989 was 94% versus 68% for the period 1965 to 1969. Thirty-three children have died, 15 with known disease progression. Long-term morbidity included scoliosis (39), cardiorespiratory insufficiency (13), hypertension (7), renal insufficiency (7), small bowel obstruction (10), chest wall deformity (3), amenorrhea (1), leg length discrepancy (1), and 1 patient with an esophageal stricture. One patient with cardiomyopathy secondary to adriamycin has recently undergone cardiac transplantation. Five patients with renal insufficiency have required dialysis. Of these five, one patient has had two renal transplants. The presence of distant metastases and positive hilar or regional lymph nodes were the only findings at operation that were associated with an increased mortality (P = .005). There was a significantly increased mortality in those children operated on by general surgeons or urologists at other hospitals (11/43) versus those operated upon at our hospital (22/191) (P = .033). There was no statistical difference in the staging or histology among these children. We feel that the careful and systematic approach of a radical nephrectomy assures accurate staging of the tumor removing gross and microscopic disease in the abdomen.

Choledochal duct cyst: resection with physiologic reconstruction.

BACKGROUND: The accepted surgical treatment of choledochal duct cyst is complete excision and enteric drainage through an intestinal conduit. Peptic ulceration and fat malabsorption have been reported after Roux-en-Y reconstruction. Such long-term complications may be avoided by a technique that simulates normal physiology. METHODS: Twenty-one patients have undergone resection of a choledochal duct cyst in the past 12 1/2 years. The pathologic duct is resected to the level of normal mucosa. A short segment of jejunum with a intussusception valve (1.5 to 2 cm) is interposed between the common hepatic duct and the duodenum. The medical records and all radiographs of each patient were reviewed. Eighteen children were reexamined or the parents were contacted by phone. RESULTS: Twenty of 21 patients recovered without major perioperative complications. Twelve of them are well and have no symptoms at 3 to 12+ years (mean, 6 years) after operation. Four children are currently well 6 to 19 months after operation. Three children were well when lost to follow-up. Two patients have radiographic evidence of incompetence of the interposition valve. One of these, who initially underwent operation at 9 months of age, was reexplored at 10 months and at 10 years for a stricture at the hepaticojejunal anastomosis. The other, a 7-year-old girl who was admitted with jaundice and pancreatitis, has had episodic abdominal pain for 7 years after operation but is well. CONCLUSIONS: The valved jejunal interposition hepaticoduodenostomy offers superior biliary reconstruction after excision of a choledochal duct cyst. Normal physiology is simulated, with bile draining directly into the duodenum. A short conduit prevents stasis, and biliary reflux is minimized with the addition of an intussusception valve.

Aggressive excision of pulmonary metastases is warranted in the management of childhood hepatic tumors.

Although most children who die of liver malignancies do so as the result of complications of pulmonary metastases, little has been published regarding the efficacy of surgically excising such lesions. To the 12 previously reported cases of children who have undergone excision of pulmonary metastases of hepatic tumors, are added 5, 4 with hepatoblastoma and 1 with hepatocellular carcinoma. Total excision of a primary hepatic tumor leads to survival much more frequently than does incomplete excision. No patient had metastases at diagnosis. The length of time between resection of the primary tumor and the development of pulmonary disease resistant to chemotherapy is available for 9 of the 17 children; it was under 6 months for the 2 who died but over 6 months for the 7 who survived. Postoperative alpha-fetoprotein (AFP) levels accurately predicted the development of metastases in our 5 patients. Resection of metastases benefitted the 4 whose AFP levels had declined to less than 25 ng/mL following initial chemotherapy and who underwent operation before their levels increased above 1,000 ng/mL. They are alive and free of disease 4 to 83 months following excision of their lesions. Resection did not benefit the 1 nonsurvivor whose AFP level fell only to 5,000 ng/mL before beginning to increase, eventually reaching 58,000 ng/mL at the time of operation. Incomplete resection of metastases unresponsive to chemotherapy predictably leads to death. Multiple thoracotomies were successful in achieving the long-term survival of 4 children in this series.(ABSTRACT TRUNCATED AT 250 WORDS)

Transfusion requirements in conservative nonoperative management of blunt splenic and hepatic injuries during childhood.

Nonoperative management of splenic and hepatic injuries in children is safe, and the majority of those with isolated injuries do not require blood transfusion. Thirty-seven children were treated for blunt splenic or hepatic trauma from November 1983 to September 1989. There was one death in a patient with a lethal head injury. No operations were performed on those with isolated splenic or hepatic injuries. Three of those with multiple injuries underwent delayed laparotomy. Two had perirenal and retroperitoneal hematomas without active bleeding, and one had a bowel obstruction secondary to an intramural jejunal hematoma. There were no late complications related to the splenic or hepatic injuries. Eight children (22%) required surgery for other injuries. Twelve children were not transfused, including the majority (8/11) of those with isolated splenic or hepatic injury. The hematocrit of four of these children fell to below 28% and this anemia was well-tolerated. Two children with bleeding disorders (factor VIII [antihemophilic factor] and factor XII [Hageman factor] deficiency) did not require packed red blood cells transfusion. Two clinically distinct groups of children received blood transfusions: (1) eight patients with multiple injuries were transfused during initial resuscitation when unstable or during early operation for other system trauma (mean, 62.0 mL blood/kg body weight); and (2) three hemodynamically stable patients with isolated injuries and 14 stable patients with multiple injuries were transfused empirically after initial resuscitation solely because of decreasing blood counts. They received an average of 16.5 and 21.1 mL blood/kg body weight, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

Preoperative CT and MR imaging of ischiopagus twins.

Each case of conjoint twins is unique. Preoperative imaging is helpful to determine the feasibility of separation. Shared and separate organs can be delineated; and operative technique and problems anticipated. We performed ultrasonography, CT, magnetic resonance imaging, and arteriography preoperatively on ischiopagus conjoint twins. The single most helpful study was CT performed during a bolus intravenous contrast medium injection into one twin. Arteriography was also very helpful. The other studies were complementary but added little additional structural details. Sedation of conjoint twins is complicated yet crucial for optimal imaging studies. A combination of oral and intramuscular sedation was used and worked well for all of the studies.

Cervical esophageal obstruction by ectopic gastric mucosa.

A 5-year-old boy presented with dysphagia, and barium swallow revealed a narrowing in the upper esophagus. Biopsies of the circumferential lesion showed gastric mucosa. The heterotopic gastric mucosa was resected via a cervical approach. After resolution of a minor anastomotic leak, he remains asymptomatic 10 years following resection with no evidence of recurrence by esophagrams and endoscopy.

Bronchoplastic techniques for pediatric lung salvage.

Bronchoplastic reconstruction can salvage normal bronchial and pulmonary tissue distal to diverse obstructive bronchial lesions. Five children underwent proximal bronchial and limited pulmonary resections allowing preservation of normal distal parenchyma. Three patients were infants between 2 1/2 and 12 months of age. All five survived without immediate or late complications. Follow-up included interval bronchoscopy and chest radiography, and ranged from 8 months to 12 years. Cardiopulmonary bypass was unnecessary even in the case of a large carinal lesion. Techniques used to insure adequate ventilation during bronchotomy included positioning the endotracheal tube above the carina with clamp occlusion of the open mainstem bronchus, positioning the tube within the opposite mainstem bronchus, and inserting a sterile tube into an open mainstem bronchus (with connections to sterile ventilator tubing passed across the operative field). Preoperative bronchoscopy and radiographic studies localized the extent and base of the lesion, allowing precise placement of the bronchotomy. Conservative excision was accomplished by frozen section examination of all margins of resection. Ingenious usage of available tissue was essential to a successful reconstruction.

Right middle lobe syndrome in children.

Right middle lobe syndrome (RMLS) is characterized by a spectrum of diseases from recurrent atelectasis and pneumonitis to bronchiectasis of the middle lobe. It has been described among all age groups, although the diagnosis in pediatrics may be delayed or missed because of non-specific symptoms or findings. Twenty-one children with RMLS were evaluated during the past 10 years with particular attention to the history, bronchoscopic and pathologic findings. Most of these patients had asthma or a family history of atopic disorders; 3 patients had a family history of RMLS. Only two of the 21 patients had sufficient obstruction on bronchoscopy to account for their disease. Four had evidence of concomitant laryngeal pathology. The various theories of pathogenesis are discussed. In this series, the non-obstructive (impaired collateral ventilation) theory appeared to be most plausible. Bronchoscopy was performed in all instances to rule out obstruction due to foreign body or tumor. It was therapeutic in two-thirds of the cases. Resolution occurred promptly in one-third, and eventually in another third. Of the remaining patients, 4 required lobectomy and were cured; two have decreased but persistent symptoms. An aggressive medical management following bronchoscopy is warranted in all cases, especially when there is a possibility of asthma.

Emergency lung biopsy: friend or foe of the immunosuppressed child?

An acute pneumonic process in an immunosuppressed child poses a diagnostic and therapeutic challenge. These patients tolerate infection poorly. An open lung biopsy may provide prompt diagnosis. Nevertheless, a beneficial change in therapy that results in survival does not necessarily follow. Fifty-six immunosuppressed children with acute respiratory symptoms and interstitial pulmonary infiltrates underwent lung biopsy from 1974 to 1985. The most common underlying diagnosis was acute lymphocytic leukemia (60%). A specific etiology was determined in 46 (82%). Operative morbidity in 52% included prolonged intubation, recurrent pneumothorax, and hemorrhage. Overall, mortality was 34%. Those patients with solid tumor and those who required postoperative ventilation had a statistically significant higher mortality than all others. We defined biopsy "patient benefit" as follows: (1) the biopsy yielded an etiology for which a change of treatment was required; and (2) the child survived this acute illness. Despite the successful diagnostic results of this procedure, only 13 (23%) of the patients derived clinical benefit. Even though a specific infectious etiology was diagnosed in 39 (69%) patients only ten (18%) of these improved and survived after an appropriate change in therapy. Eight of these had Pneumocystis carinii. One survivor benefited from the treatment of documented radiation pneumonitis. Another was successfully treated for graft v host reaction but this diagnosis also was made by skin biopsy. One half of the biopsies were performed very early in the course of the illness, specifically to exclude Pneumocystis carinii of which we saw a peak incidence in 1978 to 1979.(ABSTRACT TRUNCATED AT 250 WORDS)

The valved conduit prevents ascending cholangitis: a follow-up.

Reflux of bacteria-laden intestinal contents into the biliary tree can be prevented by an intussuscepted valve in an isolated segment of jejunum interposed between the porta hepatis and the duodenum. This method of bile duct reconstruction was adopted in 1979 and since then has been used in 17 children with biliary atresia and 9 with choledochal duct cysts. Those with choledochal duct cysts are well. Follow-up HIDA scans demonstrate normal bile flow, and ultrasound examinations have not revealed dilated bile ducts. Bile flow was established in 13 infants with biliary atresia and was persistent in 8. Two expired because of unrelated problems between 2 to 6 months postoperatively. One of these children developed cholangitis, but at autopsy the nipple valve was incompetent. Six children currently have normal serum bilirubin levels and are clinically well. The average postoperative stay for these patients was 7.5 days. Three were rehospitalized for brief periods for suspected cholangitis which was not proven. Bile flow was never established in 4 babies and was present only briefly in 5. Five of these children died of progressive liver failure and one from unrelated causes. Two have had liver transplants, and a third is awaiting transplantation.

Diagnostic dilemmas of mediastinal cysts.

Children with foregut cysts of the mediastinum can present at any age with nonspecific respiratory symptoms or dysphagia. Chest radiograph and barium esophagram are recommended for initial evaluation, but they fail to identify some lesions. In other cases, an obvious mass may be confused with a solid neoplasm. We have operated upon 34 infants and children with mediastinal bronchogenic cysts and esophageal duplications from 1968 through 1985. This review of their clinical course and radiographic imaging studies emphasizes some of the diagnostic pitfalls that lead to operative delay. Twelve (35%) of these patients were asymptomatic. The correct diagnosis was delayed longer than 3 months from the onset of symptoms in 14 of the 22 symptomatic children. Fifteen of these presented with pneumonia or symptoms or airway obstruction. In 11, the cyst was in a perihilar or subcarinal location, areas in which a lesion can be "hidden" behind the cardiac silhouette. Five children with esophageal duplication had severe neonatal respiratory failure, chest pain, hematemesis or dysphagia. The immediate preop chest radiograph revealed a mass in 29 of 34 cases. However, the lesion was initially missed or never seen in eight of those who were symptomatic. Esophagram, performed in 23, was diagnostic in only six and was normal in four. Prior to 1979 when CT scanning became available at this institution, 11 of 19 children (57%) underwent extensive work-up, but the preop diagnosis was correct in only 50%. Since 1979, only 4 of 15 (26%) have required similar evaluation, and the preop diagnosis has been correct in all.(ABSTRACT TRUNCATED AT 250 WORDS)

Gastroschisis in 106 consecutive newborn infants.

Primary tissue closure of gastroschisis remains controversial. Some surgeons routinely place a silicone rubber sheet silo over the exposed bowel, planning a staged closure. In the past 14 1/2 years, we have cared for 106 newborns with gastroschisis, closing the defect primarily in 80%. The success of this technique depends on enlarging the abdominal cavity and decreasing the volume of bowel that must be replaced in the peritoneal cavity. Thorough preoperative rectal irrigation should evacuate all meconium. After undermining the skin around the abdominal wall defect for only 1 cm, a midline subcutaneous fasciotomy is created from the xiphoid to the pubis. The abdominal wall is then stretched in all quadrants beginning at the flanks. The eviscerated small bowel can often be returned without enlarging the initial skin defect. The skin is closed with subcuticular absorbable sutures reinforced by long skin tapes. The small ventral hernia that results is closed at about 1 year of age. Fascia could be closed primarily in 28% of these patients, and 17% required a prosthetic pouch. The duration of postoperative ileus and length of hospital stay were statistically significantly shorter in the infants who underwent primary closure. Even though more complicated patients were included in the primary closure group, the incidence of mortality and morbidity was not higher than in patients treated with silicone rubber pouches. Deaths were inevitable in five infants with gangrenous bowel, multiple anomalies, and extreme prematurity. Deaths were related to sepsis in three infants and were the result of operative or anesthetic technique in four. Only two preoperative factors were prognostic of morbidity and mortality: gestational age (but not birth weight) and the presence of intestinal ischemia or atresia.

The subcutaneous injection port: a new chest-wall artifact.

A totally implanted central venous catheter with a subcutaneous injection port is now in use in patients in whom long term venous access is required. This injection port creates a new artifact that can mimic a pulmonary lesion on chest radiographs.

The "critical" neonate with diaphragmatic hernia: a 21-year perspective.

We have seen a modest improvement in the survival of a homogeneous group of critically ill newborns with congenital diaphragmatic hernia since 1979. Twenty-seven "critical" infants have been treated who developed respiratory distress shortly after birth, required urgent resuscitation, and could not be stabilized before operation. Two died with other anomalies that appeared incompatible with prolonged survival. Ten of the 27 lived. This survival contrasts with that of only two of 17 similarly affected babies treated from 1962 to 1978. In addition, there has been no operative mortality outside of this "critical" group since 1979; whereas six noncritical babies died between 1967 to 1978. Our current therapeutic plan includes the early establishment of a respiratory alkalosis and vasodilator therapy before or during transport. Postoperatively we have attempted to maintain the baby's arterial pH greater than 7.5, Pco2 less than 25 to 30 and the PO2 approximately 150 torr. The most effective ventilatory parameters have been a rate of 130, PEEP of 5 and an inspiratory:expiratory ratio of 1:1. Peak airway pressures are kept as low as possible. Pharmacologic and ventilator therapy are weaned slowly, and intensive support has been required for at least 48 hours in each baby. Retained secretions and atelectasis of the hypoplastic lung persisted for two to several weeks postoperatively. Two babies that are one year or older still appear to have severely hypoplastic lungs on chest x-ray. M-mode echocardiography has been used to measure ventricular ejection periods. The right ventricular systolic time interval correlates with the degree of pulmonary hypertension.(ABSTRACT TRUNCATED AT 250 WORDS)