Robotic-Assisted Laparoscopic Nissen Fundoplication with Gastrostomy Preservation in Neurologically Impaired Children

Background and Significance

Laparoscopic fundoplication has become the gold standard for the surgical management of gastroesophageal reflux disease in children. It has been shown to have improved outcomes, shorter hospital stay, less postoperative pain, and earlier return of bowel function, compared with open fundoplication.^1–3 The Nissen, Thal, and Toupet procedures have all been employed, with no clear benefit of one over the other. ⁴ The robotic Nissen procedure has equivalent hospital stay, alimentation, and safety profiles to laparoscopic fundoplication, but is not indicated for routine repair due to its higher cost, increased system utilization, decreased availability, and longer procedure time.^5–9 Robotic surgery does offer key advantages over standard laparoscopy because it employs internally articulating arms, which allow 360-degree mobility for maneuvers and access to areas not possible with standard laparoscopy. ¹⁰ Other advantages of robotic surgery include a stable camera platform, three-dimensional (3D) imaging, and ergonomic operating position for the surgeon.

Children with neurologic impairment (NI) frequently have feeding difficulties that lead to malnutrition and failure to thrive. Gastrostomy placement is standard care in these children to support nutritional efforts. Gastrostomy tube feeding has been shown to increase weight, improve overall health, and decrease feeding time in these patients. ¹¹ The quality of life of caregivers is also increased. ¹² In addition to feeding difficulties, children with neurologic impairment often have factors that contribute to gastroesophageal reflux, such as impaired gastrointestinal motility, spasticity, seizure disorders, scoliosis, chronic cough, and prolonged supine positioning. ¹³ It has been shown that placement of a gastrostomy tube may produce or worsen reflux disease in these children.^14,15 For these reasons, many NI children undergo antireflux surgery at the time of gastrostomy placement. However, several studies have shown a higher complication rate when a simultaneous antireflux procedure is added to gastrostomy, compared with gastrostomy alone.^16–18 Many children with NI will require primary or redo fundoplication after having undergone the placement of a feeding gastrostomy.

Children with NI have a higher morbidity and mortality for fundoplication, compared with neurologically normal children.^18–20 One of the most common, vexing causes of failure is crural disruption with wrap herniation. ¹⁹ Recently, St. Peter et al. described a minimally invasive technique for fundoplication with minimal esophageal dissection and four-quadrant suturing of the esophagus to the crura. ²¹ There was a statistically significant reduction in wrap herniation when using this technique, compared with previous mobilization and closure techniques. This technique of minimal paraesophageal dissection has been shown to have clinically significant lower esophageal sphincter pressure increase, decreased acid reflux, and good durability for up to 2 years. ²²

Although laparoscopic fundoplication after gastrostomy has been described, it is technically difficult and often requires either revision or repair of the gastrostomy. Liu et al. showed that laparoscopic fundoplication was possible in 25 of 26 patients with previous abdominal surgery. ²³ However, 4 of 17 patients with previous gastrostomies required a takedown of the gastrostomy in order to complete the procedure. In addition, there was 1 child who developed a pneumothorax, which was attributed to the technical difficulty of dissection in a small space due to the gastrostomy. Jesch et al. performed laparoscopic Nissen fundoplication on 22 children with prior percutaneous gastrostomies. ²⁴ Two of these patients required open conversion, one due to intra-abdominal adhesions. Two patients had detachment of the gastrostomy during fundoplication. The improved internal dexterity of the surgical robot allows a higher quality of repair in children with previous abdominal surgery without disruption of the feeding gastrostomy. Although the children in these published series did not have complications with the revised gastrostomies, redo gastrostomy tube placement is associated with leaking and the formation of new hypergranulation tissue. In addition, parents and caregivers have an easier time and less anxiety managing a previously functioning gastrostomy. In this article, we present a retrospective review of 15 children with neurologic impairment and previous gastrostomy who underwent Nissen fundoplication by using the da Vinci^® surgical robot (Intuitive Surgical, Inc., Sunnyvale, CA) between January 2006 and June 2008.

Materials and Methods

All patients were referred for fundoplication by gastroenterology after failing medical management, having positive endoscopic findings for hiatal hernia and/or esophagitis, having a positive 24-hour pH-probe study, or having radiographic evidence of wrap failure. The operations were performed at three Richmond-area hospitals by two pediatric surgeons who have expertise in minimally invasive surgery. All patients underwent a floppy Nissen fundoplication with minimal esophageal dissection, crural closure, and placement of four anchoring stitches to the crura, as described by St Peter et al. ²¹ A Nathenson liver retractor was placed in the epigastrium, while the camera port was placed adjacent to the gastrostomy (Figure 1). While a 5-mm robotic scope was initially used, as in this patient, we now routinely use the 8-mm robotic scope, which provides 3D visualization. The working ports were all 5 mm and were placed laterally in the abdomen with an accessory port at the umbilicus.

OPEN IN VIEWER

FIG. 1.

Port placement for gastrostomy preservation. The patient's gastrostomy tube is temporarily replaced by a Foley catheter for the operation.

Results

For the 15 patients in this study, ages ranged from 1 to 16 years (average, 8), with their follow-up ranging from 24 to 46 months (average, 32). Operative times ranged from 2 hours 39 minutes to 7 hours 50 minutes (average, 4 hours 23 minutes). Longer operative times were attributed to increased complexity of the cases included in this study, as well as simultaneous procedures undertaken. Six patients (40%) had redo Nissen fundoplication. Four children (26.7%) underwent extensive lysis of adhesions for greater than 1 hour. Five patients (33.3%) had replaced left or accessory hepatic arteries, all of which were preserved. Seven patients underwent the repair of a hiatal hernia (46.7%), 2 of which had a biologic mesh placed. One child had a revision of the gastrostomy site, because the prior percutaneous endoscopic gastrostomy (PEG) had been placed through the transverse mesocolon. One child had a simultaneous cholecystectomy and another simultaneous tracheostomy and central line placement. There were no conversions to open or intraoperative complications. Six of the children (40%) were discharged home on postoperative day 2 and 3 more (20%) on postoperative day 3. Even though some of the children presented on continuous feeds preoperatively, usually through a gastrojejunostomy tube, all children were discharged home tolerating bolus gastrostomy tube feeds.

Postoperative complications included 1 child with a small pneumothorax, which resolved on oxygen therapy alone without the placement of a chest tube. This child was discharged home on postoperative day 3 in good condition. One child had postoperative bleeding, which required a transfusion of blood, but no reoperation. This child was discharged home on postoperative day 14. Four children had a postoperative ileus and were discharged on postoperative days 5, 6, 6, and 18. The last of these children had a very complicated underlying disease state, including cerebral palsy, laryngomalacia, intestinal malrotation, hiatal hernia, seizure disorder, and previous aspiration pneumonia. He underwent a tracheostomy and central line placement at the time of fundoplication. His ileus required the administration of total parenteral nutrition for the short term, but resolved without further intervention. His respiratory status contributed, as well, to the prolonged hospital course. One child was discharged home on postoperative day 7 due to an underlying uncontrolled seizure disorder. To date, there are no children with evidence of wrap failure. Most children received 2 doses of Toradol (Hoffman-La Roche, Nutley, NJ) with, possibly, a few doses of supplemental morphine and were easily transitioned to Tylenol (Ortho-McNeil Pharmaceuticals, Raritan, NJ) with codeine prior to discharge.

Discussion

Although robotic-assisted laparoscopic fundoplication is not indicated for routine fundoplication, it has benefits in children with previous gastrostomy, adhesions from previous abdominal surgery, or failed primary fundoplication. Prior gastrostomy contributes to the tethering of the stomach away from the diaphragm, which is accentuated in the setting of the pneumoperitoneum. This can limit the mobilization of the fundus to create the 360-degree Nissen wrap and cause tension on the gastrostomy site, which can lead to dislodgement. In our series, the robotic approach enabled adequate wrap formation without gastrostomy dislodgement in all patients. One patient was noted to have the previous PEG placed through the colonic mesentery and was thus revised. It is our opinion that the robotic approach provides improved visualization and internal mobilization of tissues over standard laparoscopy. These factors allow high-quality suture placement for crural repair, esophageal tacking, and fundic wrap, which should contribute to the long-term durability of the repair. This, combined with limited dissection of the esophagus with preservation of the deep phrenoesophageal ligament, may help prevent wrap herniation. This is especially important in children with NI who have a higher incidence of wrap failure and other postfundoplication complications.^18–20

Immediate postoperative complications of fundoplication in NI-impaired children can be multiple and severe. Smith et al., in a series of 39 open fundoplications in 35 profoundly neurologically impaired children, had 9 major complications in the first month after fundoplication. ²⁰ Complications included pneumothorax, hepatic vein laceration, gastric outlet obstruction, jejunal perforation, sepsis, fluid and electrolyte imbalances, aspiration, and small-bowel obstruction. This led to 3 mortalities and 3 immediate reoperations. The robotic approach, in our patients, enabled rapid recovery, toleration of bolus feeds, and discharge within the first 3 days, in a majority of our patients. There were no perioperative deaths and no immediate reoperations. Although 1 child had bleeding postoperatively, this was managed with judicious transfusion. Hospital stay was prolonged in the setting of postoperative bleeding, postoperative ileus, respiratory difficulties, and seizures. Although several patients were delayed in hospital due to initial feeding intolerance, all patients were able to be discharged home on bolus feeds. A number of patients preoperatively had only tolerated continuous feeds via a gastrojejunostomy tube.

Although the rate of an aberrant left hepatic artery has been reported to be 15%, we saw a much higher incidence (33.3%). ²⁵ While the number of patients in our study was relatively small, the increased incidence may also have been due to some of these vessels being accessory hepatic arteries. However, we chose to preserve all of these arteries in order to limit the risk of hepatic ischemia. This technique can lead to increased difficulty with dissection and wrap creation. However, preservation was feasible with the robotic technique in all patients in our series.

Conclusions

In summary, we believe that robotic-assisted laparoscopic Nissen fundoplication is a safe, effective operation for children with existing gastrostomy, particularly those that are neurologically impaired. The argument for the utilization of this technique is further strengthened in the setting of a child, with multiple previous abdominal operations, that require adhesiolysis. In addition, we place a piece of biologic mesh in those children that are NI and need a redo fundoplication, which is facilitated by the surgical robot. Short-term complications are low, in comparison with other techniques, in this population. Robotic fundoplication in these children also enabled the rapid recovery of bowel function, tolerance of bolus feeding, and a relatively rapid discharge from the hospital.