'Spontaneous' neonatal gastric perforation: is it really spontaneous?

BACKGROUND/PURPOSE: Most historical reports have described gastric perforation in the neonatal population as "spontaneous." More recently, several variables, including prematurity and nasal ventilation, have been implicated as contributing factors. The authors sought to analyze the etiology, course, and outcome of newborns with spontaneous gastric perforation from one institution over a 16-year period. METHODS: The authors reviewed retrospectively the charts of all infants who underwent operation or had perforation of the stomach diagnosed in the newborn period. RESULTS: Among more than 84,000 live births, 7 newborns were identified with perforation of the stomach. Four had coexisting gastrointestinal lesions (2 necrotizing enterocolitis, 1 undiagnosed tracheoesophageal fistula, 1 meconium plug), and 1 received nasal continuous positive airway pressure (CPAP). In only 2 cases were no other gastrointestinal lesions or other presumed contributing factors (nasal CPAP) present, and thus, only 2 cases could be classified as "spontaneous." Mortality rate was 57%. Three of the patients were premature, all of whom died. CONCLUSIONS: Whereas in the older literature, most cases of gastric perforation were considered spontaneous, and were full term, the authors' review of 7 cases over a 16-year period leads us to question the cause as spontaneous. The authors found that prematurity and concomitant gastrointestinal lesions were associated with gastric perforation in the neonate and that few cases truly are spontaneous. The authors suggest that when gastric perforation occurs in neonates, a contributing cause should be sought.

Abdominal pain and appendicitis: is there a difference in referrals between HMO pediatricians and private pediatricians?

BACKGROUND: In the last few years, there have been many reports of the rising incidence of late appendicitis and perforated appendicitis. The rise of managed care medicine has been blamed for this, because the health maintenance organizations (HMO) and gatekeepers allegedly want to keep the child away from the surgeon and hospital to save costs. METHODS: The authors were in a unique position a number of years ago because they had only a single HMO in their area of practice (New Brunswick, NJ) employing 14 pediatricians, and 86 pediatricians were in private practice or on the medical school staff. The HMO had a yearly contract with the pediatric surgeons, and all visits and surgeries were covered by a yearly agreement (capitated agreement). The authors compared the number of children with appendicitis, perforated appendicitis, or just abdominal pain (not requiring surgery), between both groups over a 5-year period (1991 to 1996). RESULTS: A total of 492 patients were referred with abdominal pain over a 5-year period. Two hundred eight (42%) had appendicitis, and 284 (58%) had just abdominal pain and eventually were sent home. The HMO pediatricians diagnosed appendicitis accurately in 46% of their patients (55 of 118), and 54% were abdominal pain that was not appendicitis (63 of 118). For the private pediatricians, the percentages were: 40% of their patients had appendicitis (153 of 374), and 60% had abdominal pain without appendicitis (221 of 374). The incidence of perforated appendicitis was similar (28% HMO v 30% private), and the incidence of negative explorations were similar (8% HMO v 9% private). CONCLUSIONS: The results are very similar in both groups, with a similar percentage of abdominal pain not being appendicitis (60% v 54%). HMO patients did not pay for consultations, and the HMO was not charged for any consultations. However, it does not appear that they overused the consultations for abdominal pain that were not appendicitis. Their accuracy for correct diagnosis of appendicitis was similar to the private group (46% v 40%), and the incidence of perforated appendicitis for both groups was 28% to 30%. The authors cannot draw any conclusions from their experience that HMOs in general do refer patients to surgeons in a timely fashion, and do not delay the referrals to avoid the cost of consultation, because the authors had a capitated agreement with their HMO, and consultations were free. Both groups of pediatricians, the HMO and the private pediatricians, did not realize financial gain or loss by sending children to the pediatric surgeon and were not penalized by sending patients with abdominal pain to the surgeon. On the contrary, their referral habits and judgments were similar and did not appear to change by being part of an HMO.

Is it necrotizing enterocolitis, microcolon of prematurity, or delayed meconium plug? A dilemma in the tiny premature infant.

Failure of a small premature newborn to adequately evacuate meconium for days or weeks has been attributed to "probable necrotizing enterocolitis (NEC)" or "microcolon of prematurity." The authors present an unusual type of "meconium plug syndrome" with the same clinical picture, seen in tiny premature babies (500 to 1,500 g), which required a contrast enema or Gastrografin upper gastrointestinal (UGI) series to evacuate the plugs. The obstruction resolved. Twenty babies (480 to 1,500 g) presented with the same clinical picture without any x-ray suggestion of NEC; contrast enemas were performed because of the suspicion of meconium plug syndrome. All 20 had extensive meconium plugs that were evacuated by the enema or by a Gastrografin UGI series. Most of them improved after the plugs were passed. These infants differ from typical full-term babies with meconium plug syndrome in a number of ways: (1) many of the mothers were on magnesium sulfate (MgSO4) or had eclampsia; (2) the plugs were diagnosed late rather than shortly after birth; and (3) the plugs were significant, extending to the right colon. The authors believe that when a tiny premature baby has findings consistent with meconium plug syndrome, the baby should be transported to radiology, for a Gastrografin enema, despite the difficulties involved. Delay postpones the start of feedings, and increases the number of radiographic studies.

Is early fascial closure necessary for omphalocele and gastroschisis?

Before the introduction of the "silo" for gastroschisis, the main goal of surgery was to cover the defect with skin. Since the silo has been used, the goals have been (1) to cover the defect with SILASTIC sheets and return the extraabdominal contents to the abdominal cavity by progressive plication of the silo and (2) to eventually close the defect by fascia-to-fascia approximation, before 1 month of age. In many series, early definitive abdominal wall closure resulted in mortality rates of 10% to 30%, usually because of bowel necrosis and resulting sepsis. At the author's institution, 20 newborns with large omphaloceles or gastroschisis have been treated, and fascial closure was obtained by the second week in 10 infants. In ten babies it was impossible to obtain early fascial closure without tension, and these children were managed differently. A nonaggressive two-stage approach was used, in which the goals were (1) early return of contents to the abdominal cavity and (2) only skin and granulation coverage of the defect (without aiming for early fascial closure or partial fascial closure) with a small central SILASTIC patch. Stage 1 is reduction of abdominal contents to the abdomen, through plication of the silo, over a 9 to 14 day period. Stage 2 is removal of the silo and closure of the ventral abdominal wall defect using a SILASTIC patch to close most of the defect, after approximating fascia in the superior and inferior portions. If the skin cannot be closed, the patch usually separates in 14 to 21 days, the pellicle remaining becomes completely epithelialized in 1 to 2 months, and further surgery has not been necessary.(ABSTRACT TRUNCATED AT 250 WORDS)

Allopurinol protects the bowel from necrosis caused by indomethacin and temporary intestinal ischemia in mice.

The present study was undertaken to evaluate if allopurinol administration protects mice from bowel necrosis caused by temporary intestinal ischemia followed by indomethacin (INDO). We have previously reported that ischemia (15-minute occlusion of superior mesenteric vessels) followed by intravenous (i.v.) INDO caused significant bowel necrosis in CD-1 mice. Ischemia or INDO alone did not cause necrosis. To investigate protective measures against necrosis, we used CD-1 mice, 25 to 30 g. Forty-four animals were gavage fed 1 mL of water for 7 days and 32 animals were gavage fed 10 mg/kg allopurinol for 7 days. On the seventh day all animals were anesthetized and the superior mesenteric vessels occluded for 15 to 20 minutes, followed by i.v. INDO (0.5 mg/kg) once daily for 3 days. Animals who died were examined for bowel necrosis and all animals were killed 7 days after surgery and necropsied. Of the 44 saline-fed animals, 12 developed bowel necrosis (27%). Of the 32 allopurinol-fed animals, 1 developed necrosis (3%). The result of Fisher's exact two-tailed test was P = .006. Pretreatment with oral allopurinol significantly protects the mice from developing bowel necrosis when the mesenteric vessels are temporarily occluded and INDO is administered. Allopurinol may prevent reperfusion injury by inhibiting formation of xanthine oxidase generated, oxygen-derived free radicals and may be valuable in pretreating premature infants with patent ductus arteriosus who have had an ischemic episode in whom INDO use is contemplated.

Risk of malignancy in bilateral streak gonads: the role of the Y chromosome.

Although girls with Turner's syndrome (45,X) are not at risk for malignancy, patients with feminizing testicular syndrome with XY chromosome composition and patients with "mixed gonadal dysgenesis" are at risk for malignancy, and bilateral gonadectomy is performed. We have treated seven girls with "Turner-like" syndrome, who we believe are also "at risk" for development of malignancy and in whom gonadectomy should be performed. We present seven cases of phenotypically typical females, without sexual ambiguity, who presented with primary amenorrhea and short stature (5) and/or minor dysmorphic features (2). Chromosome analysis showed 45,X karyotype plus a fragment that we could not rule out as being part of a Y chromosome (in one patient a complete Y chromosome). In two patients, the fragment was subsequently positively identified as a Y, using a DNA probe. In view of the known high incidence of development of gonadoblastoma in the dysgenetic gonads of phenotypic females with a Y chromosome, bilateral gonadectomy was performed in these girls. Bilateral genital streaks with normal uterus and Fallopian tubes were found in all patients. In two patients unsuspected gonadoblastoma, without metastases, was found. In five cases, Leydig cells and tubular structures resembling rete testes were found, cells that are associated with Y-chromosomal tissue. We stress the need for complete chromosomal evaluation of phenotypically female patients with primary amenorrhea or features of Turner's syndrome. If a Y chromosome or the possibility of a Y chromosome cannot be excluded, gonadectomy should be performed because of the risk of malignancy.

Unilateral leg edema caused by abdominoscrotal hydrocele: elegant diagnosis by MRI.

A 5-month-old boy presented with bilateral hydroceles since birth and right leg edema. An ultrasound of the pelvis showed a 4-cm cystic mass that was diagnosed as a teratoma or cystic hygroma. Magnetic resonance imaging was performed, which showed a dumbbell shaped contiguous, fluid filled mass extending intraabdominally through the inguinal canal from the scrotum. The cystic portion in the right iliac fossa was lying on the right iliac vessels, which were patent. A bilateral hydrocelectomy was performed, and the intraperitoneal sac was completely excised through the inguinal incision. The edema of the right leg disappeared a few days after surgery.

Indomethacin administration after temporary ischemia causes bowel necrosis in mice.

Several reports have suggested indomethacin administration causes necrotizing enterocolitis (NEC) in infants. Few experimental studies have addressed this relationship. We studied the effects of indomethacin after temporary intestinal ischemia in CD-1 mice, using a previously reported method of simulating NEC in mice. This involved occlusion of both superior mesenteric vessels for 15 minutes. Three groups were studied. In group 1, 12 mice had superior mesenteric vessels occluded for 15 minutes, followed by intravenous (IV) administration of saline for three doses over 3 days. In group 2, 12 mice had sham laparotomy, without occlusion of vessels, followed by IV administration of indomethacin for three doses over 3 days. In group 3, 36 mice had mesenteric vessel occlusion for 15 minutes, followed by IV administration of indomethacin for three doses over 3 days. The results were as follows: group 1, bowel necrosis developed in 1 of 12 animals (8%); group 2, all 12 animals survived without bowel damage (0%); and group 3, 22 of 36 animals developed bowel necrosis (61%) (Fisher's Exact Test: occlusion alone v occlusion and indomethacin, P = .002; indomethacin alone v occlusion and indomethacin, P = .00015.) We conclude that whereas occlusion alone or indomethacin alone does not cause bowel necrosis, temporary intestinal occlusion followed by indomethacin causes bowel necrosis in over 60% of animals studied. In the shocked preterm infant who may have suffered temporary intestinal ischemia, administration of indomethacin may be the second step to development of NEC.

Life-threatening fluid extravasation of central venous catheters.

The majority of life-threatening injuries secondary to the placement of central venous catheters, such as bleeding and pneumothorax, occur at the time of initial insertion. When a catheter extravasates in the neck, edema of the neck wall or chest is usually seen, and the pump indicates occlusion. We present four cases in which an uneventful, successful placement of four central lines (three superior vena cava, one inferior vena cava) were followed at greater than 48 hours by either hydrothorax or hydroperitoneum, which resulted in either cardiorespiratory collapse or intraabdominal sepsis. In reviewing these cases, all showed both a change in catheter location on a subsequent x-ray and poor or no blood return on aspiration; paradoxically, the infusion pump in each case did not sense a catheter malposition or occlusion. We conclude that, although the success of central line placement may be documented on insertion, a continual reappraisal of both the function and location of the line is necessary.

Intussusception in childhood.

We present our experiences with 41 consecutive cases of intussusception. Most cases had a preliminary barium enema performed that successfully reduced the intussusception in 12 cases. Twenty-nine patients were explored surgically; bowel resection was carried out in 6 patients, and surgical reduction was carried out in 23 cases.

Esophageal perforation in the neonate: an emerging problem in the newborn nursery.

We report 11 cases of esophageal perforation in the neonate, in whom no surgery was performed for repair of the perforation, nor was any cervical or mediastinal drainage carried out. The perforation was in the cervical esophagus in all cases where an esophagram was performed. Nine were in premature babies (580 to 1,350 g), and two were full-term babies. There were two deaths in small prematures (580 and 935 g), from extreme prematurity and intraventricular hemorrhage, with no morbidity or mortality related to the esophageal perforation. The babies presented as esophageal atresia, or pneumothorax with the feeding tube in the right chest, or an abnormal right upper extrapleural air collection with infiltrate. Barium esophagram showed a classic "double esophagus" configuration. Two babies were mistakenly operated on, one with a diagnosis of esophageal duplication, and one had a gastrostomy for a diagnosis of esophageal atresia. Esophageal perforation in the neonate is an iatrogenic disease that may mimic esophageal atresia, and may be managed without surgical intervention.

Acute trans-rectus traumatic hernia.

Abdominal wall herniae following blunt trauma are rare. Two cases of acute traumatic rectus abdominis muscle and anterior rectus sheath hernia, an apparently unreported entity, are described. Early recognition of the injury by physical examination and sometimes ultrasonography or CT is important. Repair should be early except in selected cases.

A mouse model for the study of necrotizing enterocolitis.

A mouse model for the study of necrotizing enterocolitis is presented. It is a model of temporary intestinal ischemia and consists of occluding both superior mesenteric vessels with a bulldog clamp for varying periods of time. The resultant lesions resemble the intestinal lesions seen in necrotizing enterocolitis in respect to the gradual development of the necrotizing lesions and their patchy distribution. We also studied the effect of intravenous saline and low molecular weight dextran in preventing the development of these ischemic lesions. In moderate ischemia, saline and dextran show a similar protective effect, and in severe ischemia, both show a protective effect, with dextran being more effective than saline.

Failure of autotransplantation of the spleen in dogs: an anatomic, radionuclide imaging, and pathologic study.

Although splenic autotransplantation is successful in mice and rats, with regeneration occurring in any location, no extensive study had been performed on dogs. We transplanted the spleen into subcutaneous and intraperitoneal locations on 40 dogs. Four to six months later, splenic scanning and reexploration was carried out. Spleens were removed, weighed, and microscopic sections were made. Months later, no uptake was present on scanning, weight was less than 6% of original weight, and little identifiable splenic tissue was present on microscopic sections. When 15 small sections were transplanted to the omentum in a "necklace" fashion, good uptake and preservation were seen after six months. The small amount of spleen transplanted by this method, however, makes this an unsatisfactory option. We conclude, therefore that our large-animal experiments do not provide a basis on which to recommend autotransplantation of the spleen in humans. Preservation of splenic tissue by splenorrhaphy is still the treatment of choice.

The importance of surgical evaluation of patients with "Turner-like" sex chromosomal abnormalities.

Children with "pure" gonadal dysgenesis do not require abdominal explorations to remove their gonads, because the risk of development of malignancy is minimal. Children with "mixed" gonadal dysgenesis, who usually present with sexual ambiguity, need to have gonadectomy because of the high incidence of malignancy in these gonads. We present four children (three adolescents and one two year old) who presented with minimal somatic anomalies, who did not have sexual ambiguity. The prominent features were primary amenorrhea and short stature in the adolescents, and minor dysmorphic features in the small child. On the basis of these physical features, complete genetic evaluation was carried out, and a Y chromosome or an unidentifiable fragment was identified in all four cases. At exploration, bilateral genital streaks were found, and were removed. In addition to the usual microscopic features of the streaks, Leydig cells and tubular structures resembling rete testis were seen in all the gonads. In one gonad, a gonadoblastoma was found. Since this type of gonad is "at risk" for the development of gonadoblastoma, the need for complete genetic evaluation of all females with "Turner-like" chromosomal abnormalities, although phenotypically female, is emphasized. If a Y chromosome or any unusual fragment is identified, exploration should be carried out, and any gonad that is not a normal ovary should be removed.