The Value of the Double-Track Sign as a Differentiating Factor between Pylorospasm and Hypertrophic Pyloric Stenosis in Infants

Abstract

The development of vomiting in the infant after the immediate neonatal period frequently requires radiographic investigation of the underlying cause. The classical case of hypertrophic pyloric stenosis presents little problem, but an infant with a severe and prolonged pylorospasm may be difficult to differentiate from one with hypertrophic pyloric stenosis. In this communication, a group of proved cases of hypertrophic pyloric stenosis is compared with a similar group of cases of pylorospasm to determine identifying features of each. The main roentgenographic signs of hypertrophic pyloric stenosis are the string (Fig. 1), the shoulder (Figs. 1 and 2), the beak (Figs. 1 and 3), and the tit (Figs. 1 and 4). With the string sign an area of constant narrowing appears to be rigid and free of peristaltic activity in the prepyloric region and pylorus. This is due to the restriction of distensibility by the hyper-trophied muscle of the pylorus. The narrowed channel presents a curve upward and slightly to the left. The remaining signs are due to draping of the antrum against the thickened pyloric wall. The cases of hypertrophic pyloric stenosis proved at surgery between Nov. 1, 1963, and Oct. 15, 1965, were reviewed. Of this group, preoperative upper gastrointestinal studies had been obtained in 14 instances, 10 per cent of the entire series. The x-ray findings revealed the string sign in 11 cases, the shoulder sign in 5, the beak sign in 11, and the tit sign in 6. During this review, an additional sign was observed in all patients. It consisted of two parallel linear streaks of barium with an interposed radiolucent band extending from the prepyloric region to the base of the cap and corresponding in length to the elongated pylorus (Figs. 3 and 5). This we refer to as the double-track sign of hypertrophic pyloric stenosis. The cases of pylorospasm over the same period of time were reviewed and totaled 21 in number. Nineteen of these were successfully treated medically. One required gastrostomy; at that time a pyloro-myotomy was performed, but no pyloric tumor was present. This patient subsequently died of sepsis, and autopsy confirmed the absence of hypertrophic pyloric stenosis. A second patient died two daysafter examination, with small bowel infarction. This latter ease presented with severe spasm of the pylorus which appeared rigid and nondistensible. After an interval of twenty minutes, a slight widening of the pyloric channel occurred, but complete dilatation was never observed. Postmortem study revealed a normal stomach, pylorus, and duodenum. The double-track sign of pyloric stenosis could not be identified in any of the 21 infants with pylorospasm. That the beak sign could be demonstrated in some cases of pylorospasm indicated that this sign is not specific for pyloric stenosis.