Sixty-four cases of necrotizing enterocolitis are reviewed. The diagnosis was based on tissue examination in 57 and on the clinical syndrome, including pneumatosis, in 7. Three factors are important in the development of the disease: injury to the intestinal mucosa, bacteria, and feedings. The indications for surgical intervention are pneumoperitoneum, signs of peritonitis, and intestinal obstruction. The importance of stress in the etiology of the disease is confirmed by the high incidence of perinatal complications, particularly hypoxia. The mortality was high, but results are improving with the institution of early aggressive treatment.

In newborns with respiratory distress being treated with continuous lung distending pressures, the development of pneumoperitoneum could be a manifestation of an air-block phenomenon but it can also be due to a real perforation of the bowel. When this clinical situation arises upright films or cross-table lateral films will be helpful to demonstrate air fluid levels and some consideration should be given to inserting aqueous contrast material into the stomach and take sequential films in order to demonstrate the presence or absence of a perforation. Six illustrative cases are reported.

A case of aseptic necrosis of the femoral head is reported in a 10 and 1/2 year old white male with hemoglobin S-D disease. The patient also had functional asplenia and a history of hand-foot syndrome in infancy. The radiographic, clinical, and hematologic similarities, as well as the differences between S-D and S-S hemoglobinopathy are discussed.

Abstract Midgut malrotation is a surgical emergency usually seen in the first weeks of life. It presents with duodenal obstruction and bile emesis; the signs may be intermittent or minimal. The advantages and limitations of plain films, barium enema studies, and gastrointestinal studies were reviewed in 77 patients with malrotation; volvulus was seen in 65. Additional experience was gained from a questionnaire among members of the Society for Pediatric Radiology. Barium enema studies were preferred by 39 respondents. The upper gastrointestinal series was useful in defining extrinsic duodenal obstruction, but was difficult to interpret if the patient was asymptomatic when studied.

Midgut malrotation is a surgical emergency usually seen in the first weeks of life. It presents with duodenal obstruction and bile emesis; the signs may be intermittent or minimal. The advantages and limitations of plain films, barium enema studies, and gastrointestinal studies were reviewed in 77 patients with malrotation; volvulus was seen in 65. Additional experience was gained from a questionnaire among members of the Society for Pediatric Radiology. Barium enema studies were preferred by 39 respondents. The upper gastrointestinal series was useful in defining extrinsic duodenal obstruction, but was difficult to interpret if the patient was asymptomatic when studied.

ADVANTAGES OF PRONE POSITIONING IN GASTROINTESTINAL AND GENITOURINARY ROENTGENOLOGIC STUDIES IN INFANTS AND CHILDRENWALTER E. BERDON, M.D., DAVID H. BAKER, M.D. and JOHN LEONIDAS, M.D.Audio Available | Share

1. Pulmonary hypoplasia is the major cause of death in infants with diaphragmatic hernia anti eventration. The more severe the hypoplasia, the more confusing the clinical presentation and the greater the similarity to medical respiratory distress syndrome. 2. The hypoplasia affects the ipsilateral lung maximally but the contralateral lung is often involved. The lower lobes are more involved than the remainder of the lung. 3. Serial roentgenographic studies in a case of severe unilateral pulmonary hypoplasia associated with diaphragmatic hernia revealed the development of lower lobe emphysema. This is correlated with pathologic studies suggesting that the hypoplasia is basically an arrest in bronchial (and bronchiolar) growth at the time of maximal fetal bronchial growth. The alveoli present overdistend to compensate for the reduction in bronchial segments and emphysema results.

In 1930 Wangensteen and Rice (9) introduced the still widely used technic of obtaining upside-down abdominal roentgenograms in newborn infants with “imperforate anus.” They made no measurement of any type, but merely employed the apparent distance between the rectal gas and anal dimple in planning surgical therapy. Subsequently, various measurements were devised including a distance of 1.5–2 cm between the rectal gas and anal marker; if this distance was exceeded, the perineal approach for correction was not used. Bony reference points were also considered (5), including the pubococcygeal line and ischial point (8); if the gas terminated above the pubococcygeal line, the perineal approach again was not advised. All these methods obviously were aimed at guiding the surgeon by placing patients into two major groups: those amenable to repair by the perineal route (“low”anomalies) and those not approachable solely from the perineum (“high” anomalies) (4, 6, 8). A basic assumption was implicit in all previous roentgen measurements: that the position of the rectal termination could be determined by these landmarks. In our experience the roentgen division of the anomalies has not been consistent or reliable. Indeed, the roentgen measurements have often been confusing and erroneous in assessing the length of the rectal pouch from the skin. We will illustrate that such roentgen conclusions are largely derived from anatomic and physiologic mechanisms independent of the internal anatomy of the malformation. Our conclusions are based on knowledge of the clinical picture of this complex group of anomalies. The roentgen studies to be presented include both conventional radiologic and cinefluorographic examinations of the neonatal gas-filled rectum in patients with “imperforate anus.” In addition, radiopaque contrast material has been injected into the rectal pouch through the perineum with cinefluorographic studies of the ascent and descent of the rectum in response to changing physiologic states. Basic Approach for Clinician and Radiologist The radiologist, confronted with films of an infant with “imperforate anus” cannot give any meaningful interpretation without knowing the clinical picture. This includes the sex of the patient, the presence or absence of a visible perineal fistula, and the presence or absence of meconium either in the vagina or in the male infant's urine. Unfortunately, even today most radiologic requisitions on such patients merely say “imperforate anus.” Also, the films are obtained on a struggling crying infant hung by the heels with a penny placed, hopefully, on the anal dimple. Measurements are made of the distance from the rectal gas to the marker and a report submitted.

MECONIUM peritonitis, representing in utero perforation of the gastrointestinal tract, is usually associated with neonatal intestinal obstruction, secondary either to atresia, volvulus, or adhesions from the peritonitis. Only rarely1 2 3 4 5 6 may the infants be without any signs of gastrointestinal abnormality. Such male infants may have "hydroceles" at birth, only to return at four weeks of age with hard scrotal "tumor masses." Such masses represent the rapid calcification and tissue reaction of meconium peritonitis; the fetal scrotum communicates with the peritoneal cavity through the patent processus vaginalis through the last months of pregnancy until delivery. Because the correct diagnosis is not suspected . . .

Inferior vena cavography has been suggested for initial roentgenographic evaluation of abdominal masses in infants and children (2, 3, 6, 7). The procedure has been proposed as a means of determining both malignancy and operability (2). The two common tumors encountered, Wilms's tumor and neuroblastoma, are readily diagnosed in the majority of patients by ordinary intravenous urography. We wish to report serious problems, seemingly inherent in the interpretation of inferior vena cavograms, that critically limit their usefulness. There is a high incidence of faint to nonvisualization of the inferior vena cava accompanied by massive shunting through pelvic veins to the rich retroperitoneal ascending lumbar collateral channels leading to the azygos system. These pathways were repeatedly seen in patients in whom the inferior vena cava was either normal at operation or merely displaced and partially compressed. The factors producing such spurious caval obstruction include a large mass adjacent to the inferior vena cava, the lesser effects of crying and straining (Valsalva maneuver), and the supine versus prone positions. Technic of Inferior Vena Cavography To be practical, the technic of inferior vena cavography must be applicable to the small infant with a mass who is referred for urography. Cavography will yield the pyelogram but must be accomplished through a leg vein puncture. The femoral vein is avoided in the Babies Hospital, New York, for fear of both possible hip joint infection and arterial spasm with gangrene. The catheter technic is not accomplished in such small patients with enough facility to make it practical. After preliminary frontal and lateral abdominal films have been examined for osseous metastases (which would largely exclude Wilms's tumor) and for calcifications (which usually mean neuroblastoma), a saphenous vein is used as the site of injection. Dependent on the age of the patient, scalp vein needles ranging from No. 19 to as small as No. 21 or No. 23 can be used. Tourniquets are of value (3) when employed on the opposite leg, to decrease dilution of the bolus. On the side of injection they force the contrast medium into deep veins and, when released from this side, allow it to travel as a bolus to the abdomen. Their major disadvantages are the resultant crying, straining, and struggling which increase the shunting to retroperitoneal collateral channels (1). The first study, made with the baby supine, consists of rapid injection of 2.5 cc/kg of 50 per cent sodium diatrizoate (or 60 per cent meglumine iothalamate), with a cross-table lateral film of the abdomen taken as the last cubic centimeter or so is injected. One to two minutes later, a second injection, equal in amount, is made for the supine anteroposterior projection. The patient has not been moved for these exposures, which can be followed by routine five-minute supine and ten-minute prone abdominal films.

Calcified portal vein thromboemboli are not, as a rule, considered in the differential diagnosis of abdominal calcifications in the hepatic area in infants. A review of 13 autopsy cases with a retrospective examination of available chest and abdominal films showed that this entity could have been recognized in some instances (1). More recently, the roent-genographic diagnosis of calcified portal vein thrombi antemortem was suggested in 2 patients and verified at autopsy. This article presents the main clinical and radiological data from a series of 21 cases, 19 collected since 1954, representing 0.6 per cent of all perinatal and pediatric autopsies in this period. More detailed clinicopathological correlations, pathologic observations, and pathogenetic mechanisms will appear in a subsequent paper. Gross and Microscopic Anatomy The lesions may be single or multiple. They are usually subcapsular, appearing as nodules or as zig-zag or branching lines of a chalky-yellow color, and average less than 1 mm in width and 2 mm to 1 or 2 cm in length. They frequently run parallel to the liver capsule at the bottom of shallow depressions. Deeper thrombi commonly accompany the more superficial ones, and the affected area may show extensive fibrosis secondary to focal infarction (Figs. 1 and 2). All thrombi obliterated completely the involved portal veins, sometimes with evidence of recanalization. Most were organized and calcified en masse or as more or less confluent tiny psammous bodies; a few recent thrombi could be found at the periphery of older lesions, and some purely fibrous obstructions without calcification were identified (Figs. 3 and 4). The pattern of a linear or branching structure was fairly consistent and reflected in the roentgen appearance of peripheral calcifications. In some very extensive lesions, usually associated with infarction, thrombi reached from one large branch toward the periphery in a distinctive “cauliflower” pattern, also identifiable on films. Location and Distribution (Table I) The majority of the lesions were found in the left lobe of the liver. Only two single ones and one group of thrombi were located in the right lobe. The left lobe, usually at its most peripheral zone, twice showed multiple and six times single phleboliths. The calcifications were multiple and distributed in both lobes in nine cases. In addition, one “staghorn” phleb-olith was identified in the ductus venosus extending into the portal sinus. Roentgenographic Appearance and Differential Diagnosis As the patients generally died soon after birth, the only clinical studies were chest films. In sick premature infants, however, portable films are the rule, and the upper abdomen is well included, allowing visualization of the hepatic area in frontal and lateral views.

GASTROINTESTINAL FINDINGS IN CYSTIC FIBROSISHERMAN GROSSMAN, M.D., WALTER E. BERDON, M.D. and DAVID H. BAKER, M.D.Audio Available | Share

The ductus arteriosus may appear as a left mediastinal mass, the "ductus bump," on chest roentgenograms of the newborn infant. "Airblock" (pneumomediastinum) accentuates the findings. The current knowledge of the first day of life hemodynamic adjustments is reviewed, including catheterization and cine-angiographic studies. These lead to the conclusion that the ductus bump is a "normal" mass, transient and physiologic, presenting on chest roentgenogram of infants in the first day of life.

In 1929 Léri and Weill (16) described aI case of dwarfism in a young woman with very short forearms and Madelung-like wrist deformities. They termed the group of abnormalities dyschondrostéose ; under this title a small number of isolated reports have appeared (2–4,8, 11–13, 15). Lamy and Maroteaux (14) gave an excellent review and added additional cases (including one in a child) in their discussion of chondrodystrophies. The rarity of the disorder, its interesting roentgen findings, and recent reports of chromosomal studies (3, 8) make the following case history worthy of record. Case Report G. V., a 6-year-old Cuban-born boy, was seen for evaluation of unusual wrist deformities first noted at three years of age. The parents were not seeking any advice for the patient's short stature. The father, fifty years of age, was normal in height and wrist development; the 36-year-old mother was 5 feet tall, and her wrists were normal. A 9-year-old brother was normal in height and wrist development. Physical Examination: The patient was a bright, sturdy little boy; his height of 94 cm was more than two standard deviations below the normal average of 105 cm. The short stature was due to short lower legs rather than general underdevelopment of the extremities. The forearm hypoplasia was striking (Fig. 1); his fingertip-to-fingertip span was 94 cm (normal 105 cm). The cranium-fingertip to symphysis pubis-heel ratio was 80 per cent (normal 114 per cent at this age), reflecting the extremely short forearms. A bayonet-like wrist deformity was apparent in the lateral view (Fig. 1) due to the dorsal subluxation of the distal ulna from the radiocarpal bones. Wrist motion was normal, but attempts to reposition the distal ulna were both painful and transient. The fourth knuckle bulge was absent when a fist was made, reflecting the short metacarpals demonstrated on the roentgenograms. Laboratory Studies: Calcium phosphorus and alkaline phosphatase levels were normal, as were the hemogram and VDRL test. Roentgen Findings: A skeletal survey revealed abnormalities limited to the mid and peripheral seg­ments; the skull, spine, and proximal segments (femur, humerus) were normal. Radius and ulna: Both the radius and ulna evidenced marked shortening without bowing. The distal ulna, although dorsally subluxated, showed a normal growth plate. By contrast, the distal radius was grossly abnormal with a bulbous and sloping growth plate; the deformity was most marked on the ulnar medial half (Fig. 2). The shortness of the radius created with the ulna a “V”-shaped space in which the carpal bones rested. Lateral views confirmed the dorsal subluxation of the ulna from the radiocarpal bones with a bayonet-like stepoff (Fig. 3). Hand: Radiographs of the hand showed bilateral marked shortening of the fourth metacarpals with normally shaped ossification centers. The digits were normal.

Reversible changes in the gastrointestinal tract of adults due to bleeding and edema have been described clinically (11, 14) and produced experimentally (11). It is not unexpected that examples should be found in the pediatric age group of diseases of diverse etiology with common roentgen manifestations, relating to hemorrhage, ischemia, edema, and necrosis with return to normal. Schönlein-Henoch purpura, giant hemangioma with platelet trapping, classic hemophilia, and postoperative ischemia of a colonic transplant for reconstructing the esophagus in esophageal atresia have all presented with signs of gastrointestinal bleeding and examples of “thumbprinting” and “pseudotumor” defects on barium studies. Schönlein-Henoch purpura (nonthrombocytopenic purpura) In children with Schönlein-Henoch nonthrombocytopenia or anaphylactoid purpura, swollen painful joints, cutaneous purpura, and abdominal pain with associated bleeding from the intestinal tract are characteristic. The abdominal pain, often colicky, may precede the cutaneous and joint manifestations. Surgical exploration, when an acute abdomen has been suspected, has revealed swollen edematous hemorrhagic loops of small bowel. The intestine may grossly resemble an acute regional enteritis. Gastrointestinal studies with nonflocculating barium (Baroloid) have revealed a diffuse involvement of the small intestine with dilated and separated loops, and marginal polypoid defects (“pseudo-tumor” or “thumbprinting”). When the patients have been re-studied in three to four weeks the intestinal pattern has returned to normal (5). Case I: An 8-year-old white male was admitted to Babies Hospital with a four-day history of generalized abdominal pain, rectal bleeding, purpura, and swollen joints. No recent toxin exposure, infection, allergic history, or drug ingestion was noted. Coagulation studies were normal; stool guaiacs were 2+ to 4+. Bone-marrow showed normal megakaryocytes. Small-intestinal barium examination (Fig. 1) disclosed the entire midgut from the ligament of Treitz to the ileocecal valve to be involved; dilated, separated loops had loss of normal mucosal detail with polypoid marginal defects. Steroids were not administered; the patient became asymptomatic and one month later (Fig. 2) he was normal, and the abnormal roentgen signs had disappeared entirely. It had been noted previously that the less severe the clinical signs of abdominal pain and blood per rectum, the less marked were the manifestations of bleeding and edema in the small-bowel roentgenograms (5). The Schönlein-Henoch syndrome is a vasculitis; as such it produces hemorrhage, edema, and even necrosis. Intussusception may occur, though in the majority of patients studied roentgenographically there has been no such involvement.

THE DUCTUS BUMP A TRANSIENT PHYSIOLOGIC MASS iN CHEST ROENTGENOGRAMS OF NEWBORN INFANTSWALTER E. BERDON, M.D., DAVID H. BAKER, M.D. and L. STANLEY JAMES, M.D.Audio Available | Share