Laparoscopic Fundoplication in Children with Ventriculo-Peritoneal Shunts

Introduction

Laparoscopic fundoplication is well established for the treatment of severe gastroesophageal reflux in infants and children. Many of these patients have other co-morbidities and may have required already previous surgical interventions, including ventriculo-peritoneal (V-P) shunt placement.^1,2

Infection is a significant cause of morbidity and mortality in children with a V-P shunt in situ. ³ The purpose of this report was to assess the outcomes of laparoscopic fundoplication with or without simultaneous gastrostomy placement in this group of children and to determine whether the results are different in patients without a V-P shunt.

Subjects and Methods

Details for all children who underwent laparoscopic fundoplication at our institution between July 1998 and March 2011 were collected prospectively at the time of surgery. A retrospective review of this database of patients with a V-P shunt in situ at the time of fundoplication was conducted. Primary outcomes included intra- and postoperative complications as well as shunt-related problems within a 6-month period after surgery. Further parameters assessed and subsequently compared between the V-P shunt group and the main cohort included duration of surgery, need for postoperative admission to a high-dependency unit or pediatric intensive care unit, opiate requirements, time to full feeds, and length of hospital stay.

All operations—either a total fundoplication according to Nissen or a partial anterior wrap—were performed in a standardized operative way previously described by the authors. ⁴ All patients had a V-P shunt system with a pressure valve in place (the Codman^® Hakim^® [Hakim USA, LLC (Coconut Grove, FL) used under license by Codman & Shurtleff, Inc. (Raynham, MA) system] with the valve adjusted to a low-medium pressure [7–10 cm of H₂O] was used as a standard device). Children with a V-P shunt received a single dose of amoxillin/clavulanic acid (Co-Amoxiclav) at induction. Pneumoperitoneum pressure was chosen relatively low and ranged from 7 to 10 mm Hg of CO₂ according to patient size. Depending on the underlying condition and the weight status, at the end of the procedure a gastrostomy was placed. As a device of choice a percutaneous endoscopic gastrostomy (PEG) (Corflo^® PEG Kit; Merck Serono Ltd., Geneva, Switzerland) was used. This was inserted under laparoscopic guidance in order to avoid damage to the shunt system and to minimize gastric leakage. In small infants ≤7 kg the diameter of the gastroscope and the size of the smallest available PEG device were too big to be placed safely. In these children a balloon gastrostomy (Dilation Gastrostomy Kit; Medicina Ltd., Bolton, United Kingdom) was inserted if required.

Variables were compared using the two-tailed Student's t test, chi-squared test, or Fisher's exact test, where appropriate. Significance was defined as P≤.05. Data were expressed as median (range) or mean (±SEM) values as stated. SPSS software version 15.0 for Windows (SPSS Inc., Chicago, IL) was used for statistical analysis.

Results

During the study period a total of 343 children underwent a laparoscopic fundoplication, of which 11 (6 girls, 5 boys) (3.2%) had a V-P shunt in place at the time of surgery. Mean age (4.3±4.2 years) and sex distribution as well as weight at fundoplication (15.9±12.9 kg) did not differ significantly compared with the main cohort. The median follow-up was 18 months (range, 6–75 months). Nine of the 11 (82%) had a variety of underlying neurological disorders, and 7 (64%) were born premature. The mean duration between the insertion of a V-P shunt and the fundoplication was 42.6±50.7 months. All but 1 patient (91%) had a gastrostomy inserted at the end of the procedure.

At laparoscopy the shunt was identified in all cases. In 6 patients (55%) adhesions were documented in the operating notes, of which in 3 cases they were severe enough to cause intraoperative problems (e.g., difficulties in identifying anatomical landmarks, bleeding, etc.) (Table 1). One boy (9.1%) required conversion to open surgery when freeing of dense adhesions resulted in diffuse bleeding from the omentum. The incidence of conversion was not statistically different compared with five conversions in the main cohort (due to bleeding from the liver, equipment failure, poor visibility associated with adhesions, a large hiatus hernia, and pneumothorax) (P=.179). In a second patient with a V-P shunt the colon was perforated during insertion of the PEG. This was immediately noted and subsequently repaired laparoscopically with no deleterious consequences.

There was no significant difference between patients with a V-P shunt and the main cohort with regard to duration of surgery, need for admission to a high care unit (pediatric intensive care unit or high-dependency unit), requirement for postoperative parenteral opiates, time to full feeds, or the length of hospital stay (Table 1).

Apart from one wound infection that was treated successfully with a course of antibiotics, there were no significant postoperative complications in the V-P shunt patients. None of the children had evidence of shunt dysfunction or infection related to the laparoscopic procedure. There was no documented episode of air embolism into the shunt.

Discussion

Minimally invasive techniques for the treatment of gastroesophageal reflux have generally replaced the open approach in the pediatric population. The reported benefits include less postoperative pain, faster recovery, shorter hospital stay, and better cosmesis. ⁵ Even in children who had undergone a variety of previous open operations, laparoscopic fundoplication has been shown to be feasible and safe. ¹ A special subgroup consists of patients with V-P shunts in situ who subsequently require antireflux surgery. There is concern about intraoperative problems due to adhesions and the possibility of retrograde introduction of carbon dioxide into the ventricles as well as the risk of postoperative infection and associated malfunction of the shunt system in these patients. With a view to the literature there is a paucity of data on the outcome after laparoscopic fundoplication in children with V-P shunts.

In 2011 Allam et al. ⁶ reported on 23 adult patients with V-P shunts who underwent cholecystectomy. In the laparoscopic group 57% were converted to open surgery because of dense adhesions. They observed a significant higher conversion rate in patients with V-P shunts compared with their main group. ⁶ Equally, in this study there were moderate to severe adhesions noted in more than half of the patients. However, in this study, although dissection was sometimes demanding, conversion to open surgery was only required in 1 patient when freeing of severe adhesion resulted in diffuse bleeding from the omentum.

In a study by Liu et al. ² of laparoscopic Nissen fundoplications all 11 children with a V-P shunt were completed laparoscopically.

Overall, our conversion rate (1.7%) is comparable with the literature, ranging from 0% to 4.4%,^2,5,7–9 and, in contrast to Allam et al., ⁶ there was no statistical difference between V-P shunt patients and the main cohort.

In this study the procedures were timed from skin incision to skin closure. The extra time for placing a gastrostomy was documented but not counted in the fundoplication. Similar to other series^2,10 a trend toward a longer operating time in children with V-P shunts was noted, although there was no significant difference compared with the main cohort despite of the presence of adhesions.

There was no significant difference in the need to observe children with V-P shunts more frequently on a pediatric intensive care unit/high-dependency unit. This result may reflect the fact that most patients undergoing fundoplication at our institution suffered from other co-morbidities including neurological problems, and these were equally distributed between both groups. Similarly, there was no difference in the length of hospital stay between the two cohorts. The time in the hospital was comparable with other studies that had a similar patient case load, with a trend toward a longer stay the more patients had associated neurological problems.^5,8

In children with V-P shunts there have been reports of subsequent shunt infections^11,12 associated with surgery on the gastrointestinal tract. In a retrospective review Bui et al. ¹³ reported on 79 children with a variety of cerebrospinal fluid shunts who underwent fundoplication with gastrostomy. Although the difference was not statistically significant, they observed a higher infection rate following open surgery (n=7) compared with laparoscopy (n=3). The overall infection rate (12.7%) was similar to those following an initial shunt placement operation. ¹³ In another study ¹⁴ 91 pediatric patients with a V-P shunt underwent 51 laparoscopic and 48 open abdominal procedures. Subsequently four shunt infections (4.4%) occurred, one after laparoscopic gastrostomy insertion and three after open surgery (P=.56). Postoperatively, a total of 16 patients required shunt revision within a 6-months follow-up period. ¹⁴ In contrast, we did not observe a case of postoperative shunt infection, nor did any patient suffered from a decline in neurological status after the procedure. Each of our patients received a dose of Co-Amoxiclav at induction of anesthetics for prophylaxis.

In 1997 a study reported on two children with a V-P shunt ¹⁵ who underwent laparoscopic bladder augmentation and subsequently developed increased intracranial pressure to a maximum of 25 mm Hg. One year later Baskin et al. ¹⁶ described a case of a 52-year-old man with communicating hydrocephalus in whom laparoscopic insertion of a jejunostomy 5 days after V-P shunt placement resulted in immediate shunt malfunction. At emergency shunt revision the distal catheter was found to be blocked possibly by soft tissue or air caused by the pneumoperitoneum. ¹⁶ Since then concern about adverse effects associated with carbon dioxide insufflation at laparoscopy has been an issue in patients with V-P shunts.

In this study we did not observe any episodes of pneumocephalus. Although all our patients had a shunt system with a valve we have kept the intraperitoneal pressure relatively low—between 7 and 10 mm Hg—not only to ensure continues drainage of cerebrospinal fluid, but also to prevent adverse effects such as temporary increase of intracranial pressure. ¹⁵

Fundoplication is a major surgical intervention, which still carries an appreciable complication rate, particularly when combined with a gastrostomy.^7,17 A well-documented problem (which did not occur in this series) in patients with V-P shunts is leakage around the gastrostomy. ¹⁸ In this study an injury to the colon occurred during PEG placement. The perforation was identified immediately and repaired laparoscopically. In a review from Rotterdam on 467 children following PEG insertion, considerable numbers of patients were noted to have major intra- and postoperative problems. Of the various risk factors only previous V-P shunt placement revealed a significantly higher complication rate. ¹⁹ From the data in this study the authors would agree with Vervloessem et al. ¹⁹ that any PEG procedures with V-P shunts should be performed under laparoscopic guidance.

This study demonstrated that laparoscopic fundoplication and gastrostomy are feasible in children with a V-P shunt in situ and that there is no increased risk of shunt infections and/or malfunction. At laparoscopy a considerable amount of adhesions can be expected in this group of patients. If required, a simultaneous PEG should be inserted under laparoscopic guidance in these patients, in order to minimize complications.