Periampullary Diverticula and ERCP Outcomes: A Systematic Review and Meta-Analysis

Abstract

Introduction

Periampullary diverticulum (PAD) is most often asymptomatically found in elderly population. ERCP in the presence of PAD is technically challenging since the location and orientation of the ampulla could be altered. Various studies have reported differing results on the technical success and safety outcomes of ERCP in the presence of PAD. We aimed at a meta-analysis of such studies to assess the technical success and the occurrence of complications during ERCP in patients with PAD.

Methods

We conducted a comprehensive search of several databases and conference proceedings including PubMed, EMBASE, and Web of Science databases (earliest inception to October 2017). The search was done in accordance with PRISMA guidelines to identify studies. Studies that reported on the ERCP outcomes based on the presence of PAD were included. Both prospective and retrospective studies, manuscripts and abstracts were included. Only articles in English literature were included. The primary analysis focused on the overall technical success of ERCP in the presence of PAD, and the secondary analysis was to estimate the risk of occurrence of complications.

Results

Our search resulted in 16 studies that were included for final analysis. These 16 studies reported on 2794 patients, who had PAD, and the control group included 13,032 patients, who did not have a PAD during ERCP. Our meta-analysis of this data showed an Odd’s ratio estimate of having a successful ERCP procedure in patients with PAD to be 0.51 [95% C.I. (0.35–0.72)] when compared to patients without it. This was statistically significant, with a p value 0.00. Considerable heterogeneity was noted among the studies. The heterogeneity proportion was quantified at 74.6% based on I² statistic. The secondary outcomes measured were complications. We analyzed the pooled Post-ERCP Pancreatitis (PEP), cholangitis, perforation, and bleeding. Only those studies that had the data for these complications in both the study and the control groups were selected. PEP: The pooled Odd’s estimate of having PEP was 1.28, [95% C.I (0.88–1.87)] from 12 studies reporting on 1863 patients with PAD in comparison with 7803 patients without it. The risk of PEP occurrence tended to be more in the group without PAD, though it was not statistically significant, with a p value 0.20. There was some heterogeneity observed between the studies, with the quantification I² statistic being 28.6%. Our analysis shows that having PAD does not put a patient at increased risk for PEP. Bleeding: The pooled Odds estimate was 1.69, 95% C.I. 0.88–3.25 from nine studies reporting on 1816 patients with PAD in comparison with 5327 patients without it. This was not statistically significant, p value 0.11. Considerable heterogeneity was noted, with I² being 55.7%. The risk of having a bleed was noted to be more in control group, and having PAD did not put patients at increased risk for bleeding during an ERCP procedure. Perforation: Patients with PAD undergoing ERCP were not at increased risk for perforation. Seven studies reported on this complication. This was noted in seven patients out of 1245 in study group, and 19 patients out of 4912 in control group. The pooled Odd’s estimate was 1.24, 95% C.I. 0.54–2.87. There was no statistical significance, p value 0.61. No heterogeneity was noted among the studies included in this analysis. Cholangitis: Only four studies reported on this complication. In a total of 778 patients in study group, four had cholangitis and eight had this complication out of 3886 patients in the control group. The pooled Odd’s was 2.12, 95% C.I. 0.61–7.33. There was no statistical significance, p value 0.24. No heterogeneity was noted.

Conclusion

ERCP is technically feasible and increasingly successful when performed by experts in the presence of PAD. The risk of complications such as PEP, bleeding, perforation and cholangitis does not differ between ERCP done in patients with and without PAD.