Society: AGA

Diabetes mellitus (DM) and prediabetes are common complications of acute pancreatitis (AP) and have aroused general concern in recent years. The underlying mechanisms need to be addressed. Dedifferentiation of β cells has been proposed as an important factor accounting for type 2 diabetes, but has not been studied in AP yet. Notch signaling is a relatively conservative pathway which regulates the differentiation of pancreatic lineages during embryogenesis, and studies have shown a detrimental role of persistent high Notch activity in β cell identity and function in adult pancreas. In our previous study, we found an unregulated Notch activity in the exocrine and endocrine pancreas of AP mice. We hypothesize that β cell dedifferentiation due to excessive Notch activity in injured pancreas may account for prediabetes and DM following AP.
Objective: The aim of this project was to 1) Explore the existence of β cell dedifferentiation in pancreas of acute necrotizing pancreatitis (ANP) patients and AP mice; 2) Investigate the role of Notch activity in β cell dedifferentiation and endocrine function following AP.
Materials and methods: The pancreatic tissues from ANP patients were obtained for analyses. The murine AP model was induced by six hourly intraperitoneal injection of cerulein (100 μg/kg) dissolved in 0.9% saline for 4 days. Control mice received saline instead. In Notch activation group, Hes1-encoding adeno-associated viruses (AAV) of serotype 8 (AAV8-Hes1) were used to transfect the mice 3 weeks before AP model establishment. In Notch inhibition group, DAPT (50 mg/kg) was administered after last cerulein injection for consecutive 7 days. Intraperitoneal glucose tolerance test were performed at 7, 15 and 30 days post AP model establishment. Mice pancreases were harvested at indicated time points for subsequent analyses.
Results: Pancreatic specimens from ANP patients showed increased β cell dedifferentiation (Aldh1a3+ cells) compared with control pancreas. Upregulated Notch activity in both exocrine pancreas and islets were noticed as indicated by increased Hes1 expression. AP mice showed impaired glucose tolerance compared to control mice. The pancreases of AP mice showed increased Notch activity in both exocrine pancreas and islets compared with control mice, together with increased expression of β cell dedifferentiation markers, Ngn3 and Aldh1a3, and decreased β cell markers, Mafa, Nkx6.1, and Nkx2.2. The Hes1 overexpression in AP mice resulted in even worse glucose tolerance and increased β cell dedifferentiation, while Notch inhibition in AP mice with DAPT led to improved glucose tolerance and decreased β cell dedifferentiation.
Conclusion: β cell dedifferentiation accounts for impaired glucose tolerance following AP. Inhibition of Notch activity reduces β cell dedifferentiation and mitigates glucose intolerance after AP.

β cell loss and dedifferentiation in ANP versus control pancreas

β cell loss and dedifferentiation in ANP versus control pancreas

Background: Type 3 autoimmune pancreatitis (AIP) is a chronic inflammatory disorder of the pancreas caused by heightened immune activity triggered by immune checkpoint inhibitor therapy (ICI-Rx) employed for advanced malignancies. We describe the clinico-radiological spectrum and response to steroids in type 3 AIP.
Methods: We retrospectively performed a detailed clinical and radiological review of the 248/11,165 (2.2%) adult patients receiving ICI-Rx who developed type 3 AIP (>3-fold serum lipase elevation + pain, absent other etiologies). 110/248 had a normal computerized tomography scan (CT) of the pancreas at the time of first lipase elevation. We selected 48 of these patients that had serial pancreas volumetry before and after onset of type 3 AIP. We examined the impact of steroids on pain and the disease course.
Results: A majority of patients with type 3 AIP are Caucasian women in the seventh decade of life with exposure to PD-1/L1 blockade. Genitourinary cancer is the most common cancer type. At onset, type 3 AIP was asymptomatic in 154 (62%) and painful in 94 (38%). (Table 1) CT showed a mild, acute interstitial pancreatitis in a minority 27/110 (24.5%), less so in asymptomatic patients (8% vs 24%, p=0.01). 1 yr after pancreatitis both groups showed pancreatic volume loss, even when CT at the time of initial injury was normal (figure1). Steroids were given for pain (n=44). In ~80%, the pain lasted a short duration (<14 days) and was mild (visual analog score <3 in 62%) regardless of steroid use. (Table 1) Biochemically, a minority achieved complete and persistent biochemical resolution, more often when ICI-Rx was held (32/89) vs no intervention (13/84, p.003) or vs steroid treatment alone (2/23, p=.01); relapses occurred equally frequently regardless of intervention. Steroids did not impact pancreas volume loss (figure1) or occurrence of diabetes 1-year post-pancreatitis. None of these patients developed pancreatic calcifications.
Conclusions: Type 3 AIP is a unique form of pancreatitis that causes a rapid pancreatic volume loss despite the lack of pain or pancreatitis on CT at onset. Immunosuppression with systemic steroids do not alter the disease progression of this form of chronic pancreatic injury.

Introduction: Diabetes associated with exocrine pancreatic disease secondary to chronic pancreatitis (DEP, T3cDM) is complicated by insulin secretory defects and metabolic dysregulation. We investigated metabolomic signatures in chronic pancreatitis (CP) patients to identify the dysregulated metabolic pathways in the progression of T3cDM.
Methods: In this prospective study, we recruited 203 participants; CP patients (n=107), type2 diabetes(T2DM) patients (n=70) and non-diabetic healthy controls (HC,n=26).OGTT was performed to categorize CP patients into normal glucose tolerance, prediabetes, diabetes per ADA criteria. Fasting plasma samples were analyzed using liquid chromatography/ high-resolution mass spectrometry. Fatty acids and conjugates identified from the full scan in discovery phase of CP patients were confirmed by tandem mass spectrometry (LC-ESI-Q-orbitrap-MS/MS). Metabolites with highest peak intensities were compared across glycemic status vs controls/T2DM. Multivariate regression analysis was performed employing PLS-DA (metaboanalyst). Pathway analysis was carried out using KEGG and Reactome database as reference and confirmed by mummichog pathway analysis. ROC was used to estimate sensitivity, specificity, AUC of metabolites (MedCalc).
Results: Of the 107 CP patients, 60% were normoglycemic (ND-CP), 19% had prediabetes (PD-CP) and 21% had diabetes(T3cDM). A total of 4325 in ND-CP, 3666 in PD-CP,5493 in T3cDM and 4104 metabolites in T2DM were identified in comparison to HC. Untargeted metabolome profiling identified increased fatty acids/conjugates in PD-CP (60%) in comparison to HC (42%), ND-CP (28%), and T3cDM (13%).Among these, 57 metabolites had high peak intensities. Multivariate analysis identified glycerolipid pathway [PE (34:1,34:2,36:3), PC (37:6)], glycerophospholipid pathway [PG(32:0)] and sphingolipid metabolic pathways [Cer(d18:1/18:0,d18:1/24:1,d18:2/24:1)] enrichment in ND-CP in comparison to HC. As glycemic status progressed to prediabetes, inositol signaling pathway was enriched [PI (36:0), LPI (20:4)]in addition to glycerophospholipid metabolic pathway [PG (32:0), PG (34:1)],Fig1. By the time CP patients develop T3cDM,glycerolipid pathway metabolite lysophosphatidic acid LPA(16:0) (P=0.01) was enriched, Fig2. T2DM patients in contrast, had enrichment in DAGs (33:2,33:4) (P<0.001) and PS (28:2) (P=0.02) in glycerolipid pathway. Among these metabolites, PG (32:0) [90% sensitivity 53.8% specificity,AUC 65%(CI 0.54-0.75),p=0.009] and PI(36:0) [80% sensitivity,52.3% specificity,AUC 68% (CI 0.57-0.77) p=0.004] showed higher sensitivity and AUC.
Conclusion: Inositol signaling and glycerophospholipid metabolic pathways can be considered as potential therapeutic targets to arrest or delay progression of ND-CP to T3cDM. These pathways can also be evaluated for their prognostic and diagnostic value in progression of CP.

Figure 1

Figure 2

Background: Adrulipase alfa (formerly MS1819) is a recombinant yeast-derived lipase in clinical development by First Wave BioPharma Inc., for the treatment of exocrine pancreatic insufficiency (EPI) due to cystic fibrosis (CF), chronic pancreatitis (CP) and other indications. CF patients with EPI are currently prescribed porcine pancreatic enzymes (PPE) to facilitate food digestion and adequate nutrition. However, up to 25% of patients on PPE are not adequately controlled, exhibiting refractory steatorrhea, symptoms of malabsorption and loss of weight. Reasons for this may be reaching maximal recommended doses of PPE, or in some cases lack of tolerance of increasing PPE doses. A clinical trial was conducted in several clinical centers in Hungary and Turkey to evaluate the effects of the addition of adrulipase to the enzyme regimen of CF patients. The primary objectives of the clinical study were to determine the safety and the efficacy of adrulipase on top of a stable dose of PPEs on triglyceride digestion assessed by coefficient of fat absorption (CFA).

Methods: For study schematic see Figure 1. Male or female subjects ≥12 years of age at Screening with a clinical diagnosis of CF under stable dose of PPE ≥1 month, were enrolled. Subjects must have had a nutritional status defined by body mass index (BMI) ≤ 22.0 kg/m2 for females, ≤ 23.0 kg/m2 for males or ≤ 50th percentile for those 12 to < 18 years of age. Subjects must have had fecal elastase <100 μg/g of stool, baseline CFA < 80% with a maximum daily PPE dose of 10,000 lipase units/kg/day. The CFA measurements were determined after a quantitative fecal fat test administered to patients on standardized high-fat meals for 72h.

Results: In the full analysis set (N=20) the mean %CFA at baseline was 67.3 and mean increase in %CFA at Visit 4, visit 5 and visit 6 were 5.7, 6.6 and 5.0 respectively. The average improvement in CFA over baseline was 5.7%. These results were supported by secondary endpoints including improvements in stool weight, stool consistency, stools per day, body weight and BMI (Table 1). A mean increase in body weight of 1.7kg (~4 lbs) over 45 days was observed with the addition of adrulipase.

Conclusions: Escalating doses of adrulipase (700 mg/day – 2240 mg/day) on top of a stable dose of PPEs resulted in a clinically significant increase of the mean CFA (Brady et al, 2006), in addition to improvements in signs, symptoms and nutritional status. Adrulipase was safe and well tolerated when administered in escalating doses up to 2240 mg/day. The results of this study support further studies of this product in patients currently not well controlled on PPE alone. We are currently developing an improved formulation of adrulipase with better protection from stomach acid, and potentially more lipase activity in the duodenum.

Objectives
The human pancreas is a complex organ and plays an essential role to perform exocrine and endocrine functions. Acinar cells produce digestive enzymes that are transported through the duct into the duodenum. Pancreatic islets produce hormones and secrete them directly into the bloodstream to maintain glucose homeostasis in the body. The pancreas is an important organ, but there is lack of studies on culturing the three major pancreatic cell types: acinar cells, pancreatic ductal epithelial cells (PDECs) and islets. PDECs are located in close proximity to islets and we demonstrated that inhibiting an ion channel function of PDECs (e.g., CFTR) contributes to impaired insulin secretion in islets (Nat. Commun; PMID: 31311920). In this study, we will develop a novel in vitro co-culture model, triple-channel pancreas-on-a-chip, that will help to investigate their exocrine and endocrine functions simultaneously.
Methods
Acinar cells, PDECs, and islets were isolated from patient’s donor tissue who underwent total pancreatectomy with islet autotransplantation or brain death. Pancreas-on-a-chip, which is composed of three cell culture chambers and a thin layer of porous membrane, was fabricated through photolithography technique. Acinar cells and PDECs were cultured in the top chamber separated by cylinders having a narrow gap (<10 mm) between cylinders and islets were cultured in the bottom chamber separated by the porous membrane (thickness <5 mm). The pancreas-on-a-chip was connected to a peristaltic pump to apply fluid flow and control flow rate and direction, mimicking in vivo physiology. To monitor exocrine and endocrine functions, we collected supernatant and quantified secreted amylase and insulin using ELISA.
Results
We successfully isolated acinar cells, PDECs, and islets from donor pancreatic tissues and cultured them using Matrigel and were able to monitor amylase and insulin secretion in vitro. However, the amount of secretion from the cells was diminished over time, and amylase was not detected after two weeks from culturing cells. Given that acinar cells in culture can be damaged by powerful digestive enzymes, to overcome this limitation, we developed a microfluidic-based pancreas-on-a-chip to co-culture pancreatic cells and monitor endocrine and exocrine functions simultaneously.
Conclusion
The novel triple-channel pancreas-on-a-chip allows us not only to co-culture three major pancreatic cells in the same chip, but also to investigate exocrine and endocrine functions simultaneously, using a small number of cells. Importantly, the microfluidic-based pancreas-on-a-chip will allow us to apply fluid flow to the cell culture chamber that will help digestive enzymes to flow through the PDECs, preventing in this way the digestion of cells by the enzymes, which is not the case in standard 2D culture models.

Background:
Diabetes mellitus (DM) following an episode of acute pancreatitis (AP) is an increasingly described complication. However, the pathogenesis remains poorly understood and prospective studies assessing the cumulative incidence are lacking. The aim of this study is to prospectively assess the short-term incidence and risk factors of diabetes following an episode of AP.

Methods:
Adults (≥18 years) hospitalized with AP were consecutively invited to participate in this multicenter, prospective cohort study from 06/2017-10/2021 (Post-Acute Pancreatitis Exocrine Pancreatic Insufficiency, PAPPEI). Patients with pancreatic cancer, chronic pancreatitis, exocrine pancreatic insufficiency, or other malabsorption disorders at baseline were excluded. This is a secondary analysis of the PAPPEI study that was restricted to participants who completed the 12-months questionnaire data and who were diabetes-free at time of enrollment. Diabetes status was ascertained at 3- and 12- months using a combination of physician diagnosis/self-report and/or protocol-directed glycated hemoglobin (HbA1c) values. Baseline characteristics were compared between those who did and did not develop DM.

Results:
126 subjects completed the 12-months follow-up questionnaire. Approximately one-third (40/126, 32%) of participants had pre-existing DM at enrollment and were not included in additional analyses. In the remaining 86 at-risk subjects, the cumulative incidence of DM (based on either physician diagnosis or protocol-directed testing) was 5.8% (5/86) at 3 months and 7% (6/86) at 12 months. There were no differences in demographic or pancreatitis-related characteristics between those who did versus did not develop DM (Table). The proportion of participants with new hyperglycemia (i.e., either pre-DM or DM) was 12.7% at 3 months and 11.1% at 12 months. In a subgroup analysis limited to at risk participants with HbA1c values at baseline, a paired HbA1c value was available for 20 and 12 participants at 3 and 12 months, respectively. In this subgroup, the rate of hyperglycemia was 25% (n=5) at 3 months and 17% (n=2) at 12 months.

Conclusions:
The cumulative incidence of new onset DM within 1 year following AP is 7% in this prospectively ascertained multicenter cohort. There were no clinical risk factors identified with our small sample size. Self-reporting of DM underestimates the true disease burden, so additional, larger prospective studies are needed to investigate the incidence, risk factors, and mechanisms of DM following AP.

Vikesh Singh

Johns Hopkins University School of Medicine

Appearances