Long-term outcomes following failure of Nissen fundoplication.

ABSTRACT

PURPOSE: To determine the management and outcomes of patients with gastro-oesophageal reflux (GOR) that requires further intervention following failure of Nissen fundoplication (NF). METHODS: After institutional audit department approval, a retrospective review of paediatric patients who had further intervention following failure of primary NF between January 2006 and December 2015 for GOR at our centre was performed. Data are presented as median (range). RESULTS: Of 820 patients who underwent NF, 190 (23%) received further procedures for GOR management at a median of 21 months of age (6-186); 90/190 (47%) had gastro-jejunal feeding (GJ). Of these, 67 (74%) remained on GJ feeds up to a median of 48 months and 23/90 (26%) had a second NF after GJ feeding. 97/190 (51%) had a redo fundoplication without having had a GJ; thus, 120/190 (63%) of patients having a further procedure went on to have a second NF after a median period of 15 months (1-70 months). Three patients (2%) had early emergency wrap revision 4 days after first fundoplication (we classed this as an 'early complication'). Of the seven patients who failed a 3rd NF, 4 continued GJ feeding, 2 of had oesophagogastric dissociation; 2 had 4th NF of which 1 was successful and 1 patient had gastric pacemaker and is successfully feeding orally. Patients who were finally successfully managed with GJ underwent 2 (2-5) tube changes/year. We found patients who had a previous GJ were more likely to have failure of the redo fundoplication than those who had not to have the GJ (16/24 vs. 30/90, p = 0.005). CONCLUSION: The chance of success decreases with every further attempt at fundoplication. The only factor significantly associated with failure of redo fundoplication was whether the patient previously had a GJ tube. In patients with failed fundoplications, when symptom free on jejunal feedings, further anti-reflux surgical intervention should be avoided. A randomized prospective study is needed for patient selection.