The repair of the straightforward duodenal atresia is relatively simple and in many centers is routinely being performed laparoscopically. However, not every case is clear-cut, and there are many potential pitfalls: what to do with a duodenal web (resection or duodenoduodenostomy); recognizing and managing the windsock deformity; ruling out a distal atresia; performing a proper Ladd procedure when malrotation is found; whether to taper a dilated duodenum; and when and how to do a proper duodenojejunostomy. Annular pancreas is not included here, because the treatment is the same as for a simple duodenal atresia. The pediatric surgeon needs to be prepared for every variation and even unique anatomic variants that have yet to be clearly defined.
There is some disagreement over the proper management of the duodenal web, namely the question as to whether one should resect the web or simply bypass it. This is because of the risk of injury to the ampulla, which always enters the medial portion (mesenteric side) of the web. Though either approach is acceptable, I prefer to excise the web, being careful to preserve the medial aspect, but only after identifying the ampulla. The ampulla is identified by compressing the gallbladder and looking for the flow of bile.
The windsock deformity is a web that is stretched out distally for a variable distance. The circumferential origin of the web is still way back at the level of the ampulla while the central portion can be several centimeters downstream, where the obstruction appears to be. These can be extremely difficult to recognize and are easily missed. It is important to remember that duodenal obstruction almost always arises at the level of the ampulla (at the junction of the first and second portion of the duodenum) so that an obstruction that appears to be involving the third portion of the duodenum should be investigated further to rule out a windsock. Treatment involves excision or bypass once the anatomy has been properly sorted out.
It is important to rule out the presence of a second more distal obstruction, typically by passing a red rubber catheter distally and injecting saline to ensure flow through the lumen. This currently poses a problem for laparoscopists, but I suspect that this will be resolved creatively in the near future.
A significant number of infants with duodenal atresia also have malrotation and this needs to be addressed at the time of the duodenoduodenostomy by performing a proper Ladd procedure. The anatomy can be confusing and the unused small bowel is often very delicate and of small caliber, demanding patience and gentle technique.
Even when the bowel is normally rotated, the duodenum can be quite dilated, which creates the potential for dysmotility and stasis. It is usually best to taper the duodenum, as there is very little risk involved and it appears to be effective. The technique is well-described elsewhere, but it is a good idea to use a 20 Fr red rubber catheter within the lumen as a guide, a series of Babcock clamps on the antimesenteric border, and a gastrointestinal stapling device fired sequentially around the circumference of the dilated duodenum.
If a duodenoduodenostomy cannot be performed with minimal tension and in such a way that the proper flow of intestinal contents can be ensured (long-gap between duodenal ends, bulky annular pancreas), there should be a low threshold to perform a duodenojejunostomy, which is generally straightforward and well-tolerated. Given the small caliber of the jejunum, it is important to make the anastomoses long and with precise technique to ensure patency.
Finally, the postoperative care of these infants can be long and difficult, mostly due to duodenal stasis and prolonged ileus. All infants should be given parenteral nutrition through a PICC line and most with trisomy 21 or significant associated anomalies benefit from having a gastrostomy tube placed at the initial operation. The usual indication of normal bowel function, namely a transition to nonbilious gastric aspirates, might never occur in these infants even when their ileus has resolved, presumably due to the fact that their pylorus is incompetent and they tend to reflux bile into their stomachs for a very long time. If the patient is extubated and stable on the sixth or seventh day, I will empirically clamp the nasogastric tube and start a slowly advancing trial of water or glucose-electrolyte solution. Most infants are ready to feed at this point but some will need another 3 or 4 days before starting another trial. Infants who cannot feed beyond 2–3 weeks should have a contrast study to rule out a stricture or other form of obstruction.