The Effects of Metabolic Bariatric Surgery on Intra-pancreatic Fat Deposition and Total Pancreas Volume: a Systematic Review and Meta-analysis

Abstract

Bariatric procedures are common interventions for weight loss and metabolic improvements; yet, their effects on the pancreas-a key metabolic organ-are largely unknown. This systematic review aimed to summarise data on changes in intra-pancreatic fat deposition (IPFD) and total pancreas volume (TPV) following metabolic bariatric surgery and endoscopic bariatric therapy. The literature search was conducted in MEDLINE and Embase databases. Studies in humans were included if they reported on changes in IPFD and/or TPV from before to after bariatric procedure. Random effects meta-analysis was conducted. Fourteen studies (all involving metabolic bariatric surgery) met the eligibility criteria. Metabolic bariatric surgery resulted in a mean absolute IPFD reduction of 3.9% (p = 0.003) and a mean TPV decrease of 10.7 cm³ (p < 0.001). The mean relative IPFD reduction after metabolic bariatric surgery was 35.9%. Sleeve gastrectomy led to a higher mean relative IPFD reduction than Roux-en-Y gastric bypass (39.9% vs. 20.6%), though this difference was not statistically significant. Analysis of follow-up duration showed a mean relative IPFD reduction of 8.8% in patients with follow-ups of less than 3 months. Metabolic bariatric surgery leads to significant reductions in IPFD and TPV. Follow-up should take place no earlier than 3 months for the full benefits of bariatric procedures on the pancreas to manifest.

Keywords: Endoscopic bariatric therapy; Fatty pancreas disease; Intra-pancreatic fat deposition; Metabolic bariatric surgery; Total pancreas volume.

Fig. 1

Study selection flow chart

Fig. 2

Changes in intra-pancreatic fat deposition (A) and total pancreas volume (B) from before to after metabolic bariatric surgery. Absolute changes are presented. The study by Bai et al. had two non-overlapping cohorts of patients: one cohort was investigated at baseline and three months after metabolic bariatric surgery, whereas the other cohort was investigated at baseline and 12 months or more after metabolic bariatric surgery [17]. The study by Stevens et al. presented data separately for patients with diabetes mellitus at baseline and those without it [25]. CI, confidence interval; IV, inverse variance; SD, standard deviation

Fig. 3

Changes in intra-pancreatic fat deposition according to the type of metabolic bariatric surgery used. Relative changes are presented. Studies were stratified according to the predominant surgery employed in each individual study. IPFD, intra-pancreatic fat deposition; RGYB, Roux-en-Y gastric bypass; SG, sleeve gastrectomy

Fig. 4

Changes in intra-pancreatic fat deposition according to the duration of follow-up. Relative changes are presented. Based on mean relative change in IPFD in each individual study. Data on all available follow-ups in each individual study are presented. IPFD, intra-pancreatic fat deposition

Fig. 5

Correlations of changes in intra-pancreatic fat deposition with baseline IPFD (A) and baseline weight (B) in the included studies. Relative changes (more specifically, reductions) are presented. The solid line denotes the linear regression, whereas the shaded area denotes the 95% confidence interval. The 6-month follow-up data from the study by Hui et al. was used to provide a conservative estimate of change in intra-pancreatic fat deposition [22]. IPFD, intra-pancreatic fat deposition

Fig. 6

Correlations of changes in intra-pancreatic fat deposition with changes in homeostatic model assessment for insulin resistance (A), fasting insulin (B), fasting plasma glucose (C), and high-density lipoprotein cholesterol (D) in the included studies. Absolute changes are presented. The study by Bai et al. had two non-overlapping cohorts of patients: one cohort was investigated at baseline and 3 months after metabolic bariatric surgery, whereas the other cohort was investigated at baseline and 12 months or more after metabolic bariatric surgery [17]. RE, random effects

Fig. 7

Changes in dyslipidaemia (A) and diabetes mellitus (B) status from before to after metabolic bariatric surgery. CI, confidence interval; M-H, Mantel–Haenszel