Stenosis of the third portion of the duodenum due to bleeding from the anterior pancreaticoduodenal artery: assessment by multiphase contrast-enhanced examination with multislice CT.

BACKGROUND: This study was undertaken to analyze the CT findings for the rare pathological process that stenosis of the third portion of the duodenum was presumed to be caused by bleeding from the anterior pancreaticoduodenal artery. METHODS: Four consecutive patients presenting with frequent vomiting, who did not have well-known underlying disorders causing duodenal stenosis, were retrospectively recruited. Multiphase contrast-enhanced CT examinations were performed with 0.5- or 1-mm collimation. Two radiologists evaluated 2-mm axial and multiplanar reformatted images. RESULTS: In all patients, endoscopy demonstrated severe edematous stenosis of the third portion of the duodenum not associated with ulcer, bleeding, or neoplasm. The following CT findings were observed in all patients: homogenous swelling of the third portion of the duodenum associated with luminal stenosis in un-enhanced images, a band-like area of lower contrast-enhancement surrounding the walls of the third portion of the duodenum in pancreatic-phase images, and stenosis of the celiac axis. In three patients, aneurysms of the anterior pancreaticoduodenal artery in arterial-phase images were depicted. In the remaining patient, the diameter of the artery was irregular. CONCLUSIONS: Multiphase contrast-enhanced CT examination using a multislice CT scanner helps to establish the diagnosis of this pathological process.

Radiological assessment of duodenal calibre in congenital duodenal obstruction.

A review of barium meal studies in 32 cases of congenital intrinsic duodenal obstruction (CDO) from the Adelaide Children's Hospital was performed to assess and measure pre and post operative duodenal dilatation. Comparisons were made with a group of 153 normal barium studies from the same Hospital with a view to quantifying normal and abnormal proximal duodenal calibre. Duodenal size was expressed as a ratio D1/L1, (diameter of 1st part of duodenum/height of first lumbar vertebra). In this study a normal range of duodenal diameter was found to be 0.90 to 1.62 (+/- 2 SD from mean), with a mean of 1.26. The CDO patients recorded a range of postoperative ratios on first follow-up barium studies of between 1.2 and 4.5 with a mean of 2.28. Where serial studies had been performed, duodenal diameter was found to show some decrease in size postoperatively. In patients who underwent surgery for membrane obstruction there was a trend towards normalisation of duodenal diameter over time, compared with the atresia group. However, this difference did not reach statistical significance.