Failed Nissen fundoplication in children: causes and management.

INTRODUCTION: Nissen fundoplication (NF) is the gold standard procedure for the treatment of gastroesophageal reflux (GER) in children. However, it fails in a significant proportion of patients. The purpose of this study was to identify preoperative predictors of failure and to examine the results of reoperations.

PATIENTS AND METHODS: The charts of patients who underwent NF at our institution between 1992 and 2011 were retrospectively reviewed. Surgery was indicated in patients with symptomatic GER in whom medical treatment failed, particularly, in cases of esophageal atresia (EA), congenital diaphragmatic hernia (CDH), and neurologic impairment (NI). Chi-square comparisons and logistic regression were used to test comorbidities, previous abdominal surgery, surgical technique, gastrostomy, pyloromyotomy or pyloroplasty, age, weight, and surgical complications as possible predictors of NF failure.

RESULTS: A total of 360 children (217 male and 143 female) underwent NF. Comorbidities were NI (n = 100, 27.8%), EA (n = 50, 13.9%), CDH (n = 22, 6.1%), and abdominal wall defects (AWD) (n = 6, 1.7%). A total of 35 patients (9.7%) had esophageal stenosis. Age at surgery was 3.06 years (0.04-20.7 years) and weight was 12 kg (2-77 kg). NF was open in 196 patients (54.4%) and laparoscopic in 164 patients (45.6%) (with 9 conversions). Follow-up was 6.7 years (0.01-18.7 years). A total of 42 patients (11, 7%) had postoperative complications (10 wound infection, 9 dumping syndrome, 8 gastrostomy related complications, 7 intestinal obstruction, 5 evisceration, 2 chylothorax, and 1 pneumothorax). Reflux recurred in 42 patients (11.7%) and 35 patients (9.7%) underwent redo NF 1.01 years (0.02-8.4 years) after the initial surgery. A total of nine patients (2.5%) required further interventions (five another redo NF, three esophageal replacements, and one esophago-gastric disconnection). A total of 29 patients (8.1%) died during the follow-up (25 because of their baseline disease, 3 in the postoperative period, and 1 because of pulmonary aspiration 3 years after surgery). EA (31.6% failure) and CDH (46.7% failure) were the only comorbidities predictive of NF failure (p < 0.05).

CONCLUSIONS: Failure of NF is particularly frequent in patients previously operated upon for EA or CDH and can be predicted preoperatively. However, the benefits of the operation may outweigh this risk. Redo NF is indicated if symptoms of GER recur, but the proportion of failure is even higher. In subsequent failures, other options like esophageal replacement or esophagogastric dissociation should also be considered.