Laparoscopic Simple Oblique Duodenoduodenostomy in Management of Congenital Duodenal Obstruction in Children

Introduction

Congenital duodenal obstruction (CDO) is a common indication for surgical repair in neonates and rarely in older children. Duodenal atresia, the most common cause of CDO, constitutes nearly half of intestinal atresia events in the newborn. The incidence of duodenal atresia has been reported to be 1:6000–1:10,500 live births. ¹ The standard surgical treatment for CDO is open duodenoduodenostomy (DD) or jejunoduodenostomy or, rarely, membrane excision. ¹ The laparoscopic approach for the treatment of CDO (duodenoduodenostomy) was first reported by Bax et al. ² in 2001. Since then, this approach has been adopted in other centers.^3–11 After initial reports of success, this procedure was abandoned in some centers because of the unacceptably high rate of complications, particularly that of anastomotic leak. ³ With improvement in laparoscopic surgical techniques, good results have been recently reported, but still in small series of patients.^3–5,7–11

For open surgery, two main surgical techniques for making the DD have been described, namely, “simple” and diamond-shaped anastomosis.^1,12,13 The technique of diamond-shaped DD has been reported with a lower risk of complications, ¹² and it is usually preferred to the simple DD, although there has been no randomized trial to compare the two techniques. In the laparoscopic approach, the most commonly used technique in reported series has been again the diamond-shaped DD.^{3–5,7,8,10,11} No large study on laparoscopic simple DD for CDO has been reported to date. In this report, we present our technique of laparoscopic simple oblique DD (LSOD) and its results in the management of CDO in children.

Patients and Methods

The medical records of patients undergoing LSOD for CDO at our center (National Hospital of Pediatrics, Hanoi, Vietnam) between March 2009 and December 2013 were reviewed. Patients' characteristics, clinical presentations, imaging studies, lab tests, intraoperative findings, operative techniques, postoperative course, and length of hospital stay data were collected and analyzed. Our criteria for selection of patients for this report were those with birth weight over 1500 g, stable preoperative cardiorespiratory function, and absence of other gastrointestinal anomalies (intestinal malrotation, imperforate anus, esophageal atresia, etc.) and sepsis, which would preclude the use of our laparoscopic approach. All of the operations were performed by two surgeons with extensive experience with laparoscopic surgery (LS).

For our laparoscopic operations, all newborn patients were placed supine transversally across the table, and older children were placed at the end of the table, with the legs positioned at a lower level than the abdomen. The operating surgeon stood between their legs, facing the screen at the patient's head. The assistant was on the patient's left. One infra- or transumbilical 5-mm port was used for the camera and two 3-mm ports in the lateral abdomen for the operating instruments. CO₂ insufflation was maintained at a pressure of 8–9 mm Hg at a flow of 2–2.5 L/minute.

The round ligament was affixed with 3-0 sutures to the anterior abdominal wall in an anterior–superior direction to maximize the operating field. The duodenum was exposed and mobilized, and the obstructive site was identified. A 5-0 polydioxanone (PDS^®; Ethicon, Somerville, NJ) seromuscular suture was placed on the duodenal wall proximal and distal to the obstruction, then knotted, and tacked to the anterior abdominal wall for traction. The lower duodenum was incised just distal to the traction suture and extended longitudinally downward. The upper duodenum incision was similarly placed away from the traction suture and extended downward obliquely. The length of both incisions had to be at least 15 mm to ensure a large anastomosis. The DD was performed as a “simple” anastomosis, starting from the upper end of the anastomosis (just below the traction suture), using interrupted PDS 6-0 or 5-0 sutures for neonates and running PDS 5-0 sutures for older children, with intracorporal knots. We called this technique the “simple oblique anastomosis” (Fig. 1).

OPEN IN VIEWER

FIG. 1.

The technique of laparoscopic simple oblique duodenoduodenostomy: (A) placement of the traction suture; (B) longitudinal incision on the lower duodenum and oblique incision on the upper duodenum; (C) placement of the first suture on the posterior layer; (D) all sutures on the posterior layer; (E) placement of the first suture on the anterior layer; and (F) finished anastomosis.

No transanastomotic tube was used. A usual nasogastric tube was placed in all patients, and oral feeding started with clearance of gastric aspirates. Patients were discharged when they were on full oral feeding and in good general condition.

Results

Forty-eight patients were identified with a median age at operation of 11 days (range, 1 day–48 months) and median weight of 2650 g (range, 1600 g–10 kg), of which 23 were boys (47.9%) and 42 patients (87.5%) were neonates. Eight (16.7%) had an associated anomaly: 6 patients with Down's syndrome and 3 patients with congenital heart disease (1 patient had both). Thirty-seven patients (77.1%) presented with bilious vomiting, whereas 11 (22.9%) had non-bilious vomiting. All but 2 patients passed meconium after birth. One 4-month-old girl was indicated for LSOD because of persistence of symptoms after laparoscopic excision of a duodenal web during the neonatal period.

Plain abdominal X-rays showed the typical double bubbles in 31 patients (64.6%) but only one gastric bubble in 3 patients (6.2%) and were indeterminate in 14 patients (29.2%). Upper gastrointestinal contrast study was carried out in 26 patients (54.2%), and all showed duodenal obstruction.

Intraoperative findings showed type I atresia (duodenal web or membrane) in 31 patients (64.5%), annular pancreas in 9 patients (18.8%), and type III atresia (with complete separation of the proximal and distal duodenal ends) in 8 patients (16.7%). There were no intraoperative findings of type II atresia (with short fibrous cord connecting two atretic ends). The median operative time was 90 minutes (range, 60–150 minutes). There were no intraoperative complications and no conversion to open surgery. The median time from the day of operation to initial oral feeding was 4.0 days (range, 2–12 days).

There was no anastomotic leakage or stenosis. However, we had 2 cases with a postoperative complication, unrelated to the surgical technique. One neonate had gastrointestinal bleeding at postoperative Day 3 due to neonatal coagulopathy (prothrombin index decreased to 10%), which was treated successfully by transfusion and vitamin K. Another patient (a neonate weighing 1700 g) suffered from severe ventilator-associated pneumonia and died at postoperative Day 17, although feeding via a nasogastric tube was well tolerated from postoperative Day 5. For the remaining patients, the median postoperative hospital stay was 7.0 days (range, 4–17 days). All of these patients were discharged in good health. At a median follow-up of 12 months (range, 2–45 months), all patients were asymptomatic.

Discussion

Standard open surgery for the treatment of CDO has excellent results in more than 90% of cases with minimal mortality, of which most is due to associated anomalies. ¹ The well-known advantages of the laparoscopic approach over open surgery such as minimal invasiveness and better postoperative cosmesis may explain its introduction and development in the surgical management of numerous diseases, including CDO in children. In a study comparing LS and open surgery for CDO, LS patients had faster recovery with a shorter time to oral feeding and shorter hospital stay, ⁸ suggesting that the advantages of LS are not limited to cosmesis but also extend to functional outcome. Functional outcome of laparoscopic repair for CDO has also been reported to be at least comparable to the open surgery in two other reports.^9,10

Although DD for CDO using the open approach is not technically demanding, laparoscopic repair for CDO in children is considered to be one of the most demanding pediatric laparoscopic procedures, especially in the newborn. ³ Limited operating space and difficulties in exposing the duodenum and performing anastomosis with subsequent risks of anastomotic stenosis and leakage may be the main problems limiting widespread application of laparoscopic DD in the practice. In the laparoscopic approach for CDO, a conversion rate to open surgery as high as 35% has been reported. ¹⁰

According to our experience, the transabdominal suture for fixation of the round ligament superiorly and ventrally to lift the liver upward helped create more space for surgical manipulation and facilitated duodenal exposure. The usefulness of similar techniques has also been reported by other authors. ⁹

We modified our anastomotic technique from the standard simple anastomosis or diamond-shaped anastomosis ¹ to facilitate the laparoscopic approach. In the standard simple anastomosis, the incisions on the upper and the lower duodenum are both longitudinal ¹ or parallel in a transversal manner. ¹³ In our opinion, anastomosis created by the longitudinal duodenal incisions allows for a larger size, but with a subsequent tension due to the longer distance between the two ends. On the other hand, anastomosis created by the parallel transversal duodenal incisions allows less tension but increases the risk of stenosis because of the small diameter of the lower duodenum. Our technique of simple oblique anastomosis aims to resolve the aforementioned disadvantages by using the longitudinal incision for the lower duodenum and an oblique incision for the upper duodenum, thus creating a large anastomosis with minimal tension. In CDO the proximal duodenum is always significantly bigger than the distal one, thereby creating a good opportunity to make the mentioned oblique incision.

According to our experience and other reports,^3,5,9 lifting the duodenum greatly facilitates performing of the laparoscopic anastomosis. Other authors have used one or two traction sutures to lift the first part of the duodenum ventrally toward the anterior abdominal wall^3,5 or two separate traction sutures for the proximal and the distal duodenum. ⁹ Our use of a single traction suture to the planned site for the duodenal anastomosis by taking separate bites of the duodenum proximal and distal to the site of obstruction allows a precise cut into the upper and lower duodenum, optimizes the position of the duodenum for the length and oblique extension of the incision, and simplifies the oblique anastomosis during the laparoscopic approach. All of our LSOD operations were carried out successfully using three trocars, and no additional fourth trocar was needed.

In both laparoscopic and open DD, the most common complications are anastomostic stenosis or leak.^1,3,6 For open surgery, the diamond-shape anastomosis has been reported to be better than the simple anastomosis. ¹² We achieved excellent results with our technique of LSOD, without any complication of stenosis or leakage, showing that the simple oblique anastomosis is a good alternative technical option for laparoscopic CDO repair. The use of surgical U-clips for creating DD to reduce leak complications has been reported, but it required four ports plus direct introduction of the clip through the abdominal wall. ⁷ Our study confirmed the findings of other studies^4,8,9 that good surgical outcome could be achieved with standard laparoscopic intracorporeal suturing.

Although the experiences of LS for CDO in children are scant, our results were comparable to other series as shown in Table 1. The functioning outcome of LSOD was very good because our patients' median time to start oral feeding and median time of postoperative hospital stay were, respectively, 4 and 7 days, shorter than those in most other series.^4,8–10 Our operating time was similar to that in other reports.^3,4,8,9

In our opinion, because many neonates with CDO have low birth weight and may have associated anomalies, not all CDO patients are good candidates for LS, and therefore LSOD was performed for selected patients in this study. Patients in a good general condition, with stable cardiorespiratory function, without septicemia, and without other gastrointestinal anomalies (including intestinal malrotation), were selected for our technique. An additional consideration is the patient's body weight. Although successful LS repair of CDO in a 1350-g neonate has been reported, ⁵ we do not routinely use laparoscopic repair for the newborn weighing less than 1500 g because of the severely limited operating space as well as resuscitation issues, especially respiratory function. Similarly, a multicenter study has also selected CDO patients with a weight over 1700 g for LS. ¹⁰ We share the opinion that LS repair for CDO should be restricted to centers with extensive experience in LS in newborns. ³

In conclusion, this study shows that the technique of LSOD is safe and efficacious and can be a viable option in management of selected cases of CDO in children at experienced centers.