The co-occurrence of intestinal and pancreatic diseases and the common involvement of both organs in systemic diseases have been barely investigated. Postulating a similar mechanism in the pathophysiology of this association, the role of impaired gut homeostasis in influencing systemic inflammation might be crucial ^[1].

2. Pancreatic Disorders in Inflammatory Bowel Diseases

IBDs, including Crohn’s disease (CD) and ulcerative colitis (UC), are highly heterogeneous, chronic immune-mediated diseases affecting the gastrointestinal tract, with unknown but multifactorial etiology involving a complex interaction between the genetic, environmental, and/or microbial factors and the immune responses ^[2]. A wide spectrum of pancreatic abnormalities has been reported among patients with IBD, ranging from acute or chronic inflammation of the pancreas to asymptomatic imaging alterations or elevation of the pancreatic enzymes without any clinical relevance ^[3]. The occurrence of these conditions is often challenging for clinicians, due to the difficulty in finding an etiology, the decision to withdraw important drugs, and the management of the recurrences or the long-standing conditions. Acute pancreatitis (AP) is the most frequently observed pancreatic disorder in IBD patients; however, data about the diagnostic workup, clinical management, and outcome of these patients are lacking. The diagnosis of chronic pancreatitis (CP) with exocrine pancreatic insufficiency (EPI) might overlap with the symptoms of IBD and make it difficult to manage the patient. Chronic pancreatic enzyme elevation, moreover, is a benign but little-known condition. Its recognition and exclusion of pancreatic disease are essential to avoid clinical decisions that have an impact on the patient’s management.

2.1. Acute Pancreatitis

Acute pancreatitis (AP) consists of a sudden inflammation of the pancreas caused by the inappropriate activation of local digestive enzymes, with various grades of severity, diagnosed when at least two out of the following three criteria are present: (i) a typical pancreatic pain (high-intensity epigastric pain that can radiate to the upper quadrants and the back, with or without nausea and vomiting), (ii) a threefold or more elevation of the serum pancreatic amylase and/or lipase, and (iii) typical imaging findings ^[4]. The crude pooled incidence of AP is reported between 33 and 74 cases (95% CI 23–33; 48–81) per 100,000 person-years ^[5]. AP represents the most common pancreatic disorder in IBD patients with a twofold and fourfold risk in UC and CD, respectively, compared to the general population ^[6]. The most common etiologies of AP in IBD patients include gallstones and medications, while alcohol-induced AP seems less common in IBD compared to the general population. Less frequent etiologies include metabolic causes, autoimmune pancreatic involvement, and duodenal obstruction or papillary inflammation ^[3]. When AP occurs in IBD patients, both pancreatic and intestinal disease outcomes appear unaffected by one another [7,8]^[7][8]. Gallstones are frequently described in IBD patients, mainly in CD, involving up to 34% of cases ^[9]. Due to inflammation or previous ileal resection, impaired enterohepatic circulation leads to the hepatic excretion of cholesterol-supersaturated bile ^[10] and increased conjugated and unconjugated bilirubin excretion, with an unbalancing of the Admirand–Small triangle ^[11] and subsequent risk of biliary stones ^[12]. Additionally, prolonged fasting, parenteral nutrition use, use of non-steroidal anti-inflammatory drugs, and duration and activity of bowel disease could be contributing factors ^[13]. According to universally recognized criteria, in the care of AP, examinations necessary for diagnosing gallstone-induced acute pancreatitis should include blood tests and ultrasonography (US). In the presence of increasing ALT levels over 150 U/L, the sensitivity ranges between 48 and 93% and the specificity ranges between 34 and 96%. Even more so, when an alteration by blood tests is detected in more than three of the items including bilirubin, ALP, GGT, ALT, and the ALT/AST ratio, 85% for sensitivity and 69% for specificity are reported. Finally, the combination of US and blood tests yields a sensitivity of 95–98% and a specificity of 100%. To date, several studies about AP in IBD patients only consider the presence of gallbladder stones to formulate a biliary-related AP diagnosis, and often no data about the imaging study, and even more so, laboratory tests, are available [18,19]^[14][15]. Therefore, a prompt search for the etiology should be conducted when an AP episode occurs, and the gallstones should have priority both for their frequency in IBD patients, and for the need to choose the correct treatment policy, i.e., the cholecystectomy need, followed by the possible role of medications. Due to hypersensitivity or less frequent direct toxicity [14]^[16], drug-induced pancreatitis (DIP) requires caution and the presence of the following diagnostic criteria: (i) temporal sequence between medication introduction and the onset of AP; (ii) symptom cessation after drug discontinuation; (iii) AP recurrence after drug re-exposure [15]^[17]. According to the classification by Badalov et al. about the potential to induce AP based on the literature data, Azathioprine (AZA) and 6-mercaptopurine (6-MP), mesalamine, sulphasalazine, and metronidazole are classified as definite, while cyclosporine, prednisone, and prednisolone, are considered a possible cause of AP, albeit with less evidence [15]^[17]. In the last decade, AZA has been increasingly studied, while the role of mesalamine in inducing AP has been called into question [16]^[18]. In a Scandinavian nationwide register-based pediatric cohort, AZA was found to be associated with an increased risk of AP (incidence rate ratio 5.82 [95% CI 2.47–13.72]; events per 100 patients) during the 90-day risk period [20]^[19]. In adult patients, the incidence of AZA-induced pancreatitis seems to be higher, up to 7.3% in 510 patients treated with AZA, from a multicenter prospective registry. Several studies have shown the occurrence of DIP after a few weeks from AZA introduction, unrelated to dosage or other AZA-related adverse effects and typically affecting CD patients. A moderate clinical course, without hospitalization or a brief hospital stay, and a rapid recovery after drug withdrawal are described in most cases [14,21,22]^[16][20][21]. In a recent study, smoking, concurrent therapy with budesonide, and single-dose AZA were found as predictors of AZA-induced pancreatitis [23]^[22]. Interestingly, the AP occurrence within the first month by the initiation of AZA therapy is not similarly observed when thiopurines are used to manage other diseases. This association remains unclear but is probably related to a genetic predisposition. Polymorphisms in the gene encoding thiopurine methyltransferase enzyme are associated with dose-dependent adverse effects, including myelosuppression and hepatotoxicity, but are unrelated to AZA-AP risk. Recent studies identified the association of the HLA-DQA1*02:01-HLA-DRB1*07:01 haplotype with a 17% risk of developing pancreatitis in patients homozygous for the at-risk allele. Despite the paucity of data about the translation of these observations in clinical practice and the cost-effectiveness rate, a possible role of an immune-mediated mechanism in this association has been suggested [24]^[23]. In a relevant proportion of IBD patients, the etiology of AP remains unknown, and idiopathic AP (IAP) is eventually diagnosed, historically referred to as true extraintestinal manifestations, influenced by IBD activity with a rapid response when bowel disease is treated ^[3]. These data are not confirmed in a recent study on this topic, where 38 IAP (20.6%) among 185 IBD patients with an AP episode were retrospectively selected and studied [15]^[17]. No relationship was found between the extent/activity of disease or other extraintestinal manifestations, and a mild course was observed in all cases. Interestingly, additional diagnostic workup (magnetic resonance cholangiopancreatography, MRCP, and/or endoscopic ultrasound, EUS) after the first episode of IAP was performed in less than half of the patients, and an autoimmune pancreatitis (AIP) diagnosis was made in 13% of the cases of first IAP patients, due to a recurrence during the follow-up period. Thus, it is probable that a correct diagnostic route at the onset, with a more extensive workup including MRCP and/or EUS, could clarify the etiology of AP in IBD patients, reducing the percentage of IAP and preventing new AP episodes or pancreatic changes with an evolution toward chronic pancreatitis (CP) [25]^[24]. Moreover, in about 20% of IAP, the pancreatic inflammation preceded IBD diagnosis, mainly in the UC and pediatric population, suggesting that AP could be an early event in an undiagnosed IBD, particularly UC in young patients [15]^[17].

Pathogenesis of AP in IBD Patients

The innate immune activation and the acinar cell inflammatory signaling play a pivotal role in the pathogenesis of AP, with the need to balance pro-inflammatory cytokines/chemokines (Tumor necrosis factor, TNF-α; interleukin-1, IL-1) and anti-inflammatory or regulatory molecules (IL-10). Intestinal dysfunction and secondary inflammatory issues aggravate AP retroactively and are prodromes of systemic complications [26]^[25]. So, changes in the microcirculation, gut permeability/motility, bacterial translocation, and activation of the gut-associated lymphoid tissue, variously present in IBD patients, were postulated as factors promoting pancreatic inflammation. In vitro studies aimed to prevent intestinal barrier disruption, with modulation of the immune cell activation, inflammatory cytokine production, and pancreatic inflammatory signaling, have been conducted with promising results, but to date, with scanty reproducibility in the clinical setting. An essential role in maintaining intestinal barrier integrity seems to be attributable to microbiota, which has been increasingly studied in pancreatic disease. Several studies support the existence of a gut microbiota–pancreas axis, where pancreatic juice and pancreatic diseases may alter the gut microenvironment and composition [27]^[26] and conversely, the gut microbiota may influence the occurrence of pancreatic diseases [28]^[27]. Nevertheless, little is known about translating these observations into clinical practice to prevent and treat pancreatic diseases. The administration of probiotics in severe AP was only attempted in Gou et al.’s metanalysis, but the significant heterogeneity among trials does not allow uresearchers to conclude their efficacy in real-life experience [29]^[28]. The oxidative stress and the lysosomal damage/dysfunction, generally linked to neurodegeneration and IBD in humans, may represent another important AP pathogenetic pathway. Using antioxidant substances [26,30]^[25][29] and targeting lysosomes for cell death induced by lysosomal membrane permeabilization are considered additional potential strategies in recent studies. Interestingly, Sudhakar et al. showed that loss of a carbohydrate-binding intracellular protein, galectin 9 (Gal-9), in mice causes unstable lysosomes in autophagy-active selectively in intestinal and pancreatic cells at the steady state, with compromising autophagy and consequent increased susceptibility to disease pathogenesis. Gal-9 preferentially targets intestinal Paneth and pancreatic acinar cells, promoting autophagic degradation to prevent cell death. So, the lysosome dysfunction, due to loss of Gal-9, could serve as a shared cell-intrinsic defect in the intestine and pancreas, supporting the pathophysiologic link between IBD and extraintestinal manifestations (EIMs), especially pancreatitis [31]^[30].

2.2. Autoimmune Pancreatitis

The AIP is a peculiar form of pancreatitis, characterized by an immune-mediated inflammation of the pancreatic gland, with a rapid response to steroid therapy. It is histologically classified into two types, type 1 and type 2, with different clinical courses, biochemical markers, histological findings, and clinical outcomes regarding relapse rate. Even in the absence of a histological diagnosis, the International Consensus Diagnostic Criteria (ICDC) may be applied to reach a clinical diagnosis. Moreover, when no specific type 1 or 2 criteria are present, the not-otherwise specified AIP (type NOS) could be defined [29]^[28]. Rare and underdiagnosed, the AIP prevalence in the general population is still not confirmed but estimated as 0.82:100,000 [30]^[29]. The association between IBD and AIP has been described in 5% to 40% of patients [32,33,34]^[31][32][33]. Elevated serum IgG4 levels characterize type 1 AIP, possibly associated with extrapancreatic manifestations [41]^[34] and often clinically characterized by jaundice and/or pancreatic mass. Abdominal pain, weight loss, and diabetes are other possible presentations. Elderly male subjects are usually affected, and a tendency to relapse is observed after treatment discontinuation in a relevant percentage of cases (up to about 60%) [34]^[33]. A combination of parenchymal and ductal imaging, extrapancreatic disease, serum IgG4 concentrations (twofold upper normal values), pancreatic histology, and steroid responsiveness are needed for diagnosis. Imaging findings include diffuse or multi-focal enlargement of the pancreatic gland (sausage-like shape) and thin peripancreatic edematous rim, with altered signal intensity or enhancement pattern after contrast injection. However, a focal presentation, mimicking malignancy, could be observed, and histology becomes essential to avoid useless surgery. Type 2 AIP usually affects young patients without a predilection for sex. No serological biomarkers are available, and IgG4 levels are typically normal. On histology, granulocytic epithelial lesions (GEL) are typically characterized by intraluminal and intraepithelial neutrophils, often leading to the destruction and obliteration of the duct lumen. Both diffuse and focal involvement of the pancreatic gland can be present, and no other extrapancreatic disease location is described, except for IBD and mainly UC, that occurs in a relevant percentage of type 2 AIP patients (up to 80% in ourthe experience in a cohort of AIP-selected patient) ^[35]. The clinical presentations include an incidental finding of focal lesion mimicking cancer, AP, not specific abdominal pain/weight loss, and increased pancreatic enzymes. The IBD may be previously or subsequently diagnosed, compared to AIP onset, but usually, a concomitant diagnosis is made within 2 years [35,37]^[35][36]. The clinical relevance of these data, reported by studies with a high number of patients despite the rarity of AIP, is that awareness is needed when an AIP occurs to reach an early IBD diagnosis with subsequent strict management. Conversely, when an idiopathic AP occurs, AIP should be suspected in IBD patients. Conflicting data about the relationship between bowel disease activity, pancreatic inflammation, and IBD outcomes in AIP patients are available. While Lorenzo et al. suggests a possible association of UC and AIP as a consequence of systemic inflammation and a higher colectomy rate in UC-AIP patients over the follow-up period, compared to only UC patients [37]^[36] in other studies, these data are not confirmed, suggesting a different pathogenetic pathway and the need for further prospective studies to understand the real impact of AIP on UC outcome [8,35,38,39]^{[8][35][37][38]}. Finally, a relevant proportion of patients may be initially classified in NOS AIP type and reclassified during the follow-up period. A study published on this topic reported a 17.4% reclassification rate of NOS AIP to type 2 for developing IBD during the follow-up period [42]^[39].

Pathogenesis of AIP in IBD Patients

The pathogenetic correlation between IBD and AIP is still unknown, but the immune-mediated mechanism may be postulated. Type 1 AIP has been defined as the prototype of IgG4-related disease (IgG4-RD), where the pivotal role of IgG4 could also involve the bowel. However, the gastrointestinal location of IgG4-RD is rare and histologically proven CD and UC without IgG4 infiltration are sporadically associated with type 1 AIP [43]^[40]. The role of chronic inflammation and lymphocyte recruitment by inflamed tissue, known as “lymphocyte homing”, described in the last decade, could be the underlying pathogenetic link. In most chronic inflammatory disorders, including IBD and AIP, the formation of tertiary lymphoid tissues properly occurs when the lymphocytes are recruited [44]^[41] with the subsequent morphological and functional change of postcapillary venules into endothelial venule-like (HEV-like) vessels. In the first phase of this process, a mucosal addressin cell adhesion molecule 1 (MAdCAM-1), which under normal conditions is expressed only in gut-associated lymphoid tissue (GALT), has been studied. In inflamed sites of IBD, HEV-like vessels expressing MAdCAM-1 have been extensively found [45]^[42]. In a study by Kobayashi et al., the UC-inflamed site MAdCAM-1 protein can be post-translationally glycosylated with MECA-79+ 6-sulfo sialyl Lewis X-capped carbohydrates, on immunoblotting, especially in the active phase [46]^[43]. The distribution of MECA-79+ HEV vessels and the less spread positive MAdCAM1 was also found in pancreatic cells of AIP patients and salivary glands in sclerosing sialoadenitis, suggesting a common immune-mediated mechanism [45]^[42]. Differently, in type 2 AIP, the frequency of IBD is higher and histological resemblances between the diseases have been described. The same neutrophil infiltration involving colonic crypt epithelium (cryptitis) and lumen (crypt abscess) in UC has been observed in the epithelium and lumen of pancreatic acini and small and medium-sized ducts in type 2 AIP [47]^[44]. The presence of shared antigenic molecules between both organs has been postulated, suggesting type 2 AIP as an extraintestinal manifestation of IBD, mainly UC.

2.3. Chronic Pancreatitis

Chronic pancreatitis (CP) is a multifactorial, fibroinflammatory syndrome characterized by repetitive episodes of pancreatic inflammation with subsequent infiltration of the pancreas by fibrosis, resulting in chronic pain, exocrine and endocrine pancreatic insufficiency, and reduced quality of life. Recurrent upper abdominal and back pain often appear as initial symptoms of CP, with attenuation as the disease stage advances and both exocrine and exocrine functions of the pancreas gradually decrease. Multiple complications, including chronic pain, characterize the end stage of CP, exocrine pancreatic insufficiency (EPI) and endocrine impairment, referred to as pancreatogenesis or type 3c diabetes mellitus, metabolic bone disease, and pancreatic ductal adenocarcinoma (PDAC) [48]^[45]. However, a painless pattern is described in about 10% of CP patients [49]^[46], even more than in IBD patients, where the disease is clinically unapparent in most cases. Occasionally, the diagnosis is suspected because of the onset of exocrine dysfunction [50,51]^[47][48]. The incidence of CP in the general population ranges from 5 to 12 per 100,000 with a prevalence of 50/100,000 person-years, and the risk seems to be increased in immune-mediated diseases, such as IBD ^[5]. The association of IBD with CP was postulated in 1950 when Ball et al. [52]^[49] described macroscopic and microscopic pancreatic alterations in autoptic specimens in up to 53% of UC patients. In a more recent series, a higher incidence for IBD patients has been described (5.75 vs. 0.56/10,000 person-years, respectively) compared to the general population [9,53]^[9][50]. In addition, a nationwide population-based cohort study revealed a 10.3-fold higher risk of IBD in patients with CP compared to a control group [53]^[50].

Pathogenesis of CP in IBD Patients

Few studies explored the association of CP in IBD patients, mainly before the 2000s, thus including cases of mass-forming CP, with a possibility to include the known AIP. Thus, although the presence of pancreatic duct changes in IBD patients, both on ERCP and EUS imaging studies [51]^[48], the pathogenesis and the possible outcome remain elusive. Genetic, immunologic, and obstructive causes could be involved, but no human studies are available to date. In animal models of CP, when the inflammatory cytokine IL-1 is overexpressed under the control of the elastase promoter in the pancreas of mice, prominent histologic features of CP in terms of fibrosis and an inflammatory response dominated by T cells has been observed. Interestingly, despite the fibrotic replacement, the IL-1 transgenic mice developed neither pancreatic exocrine nor endocrine insufficiency after 8–10 months of life, in a comparable pattern of pancreatic alterations in IBD patients [54]^[51]. The role of IL-1 has also been recognized in animal models of bowel inflammation, as well as in cultural cells of IBD patients [55]^[52]. It is conceivable that in genetically predisposed IBD patients, the bowel disease activity with systemic cytokine- and, specifically, IL-1-mediated inflammation may involve other organs, as a pancreatic gland, without necessarily determine clinically evident CP. Another cytokine of the IL-1 family, IL-33, has been proposed as the missing link between IBD and CP. The IL-33 produced by myofibroblasts and epithelial cells has been shown to increase the Th1 responses associated with CD, especially when the intestinal epithelium is damaged, as well as to mimic CU in experimental models with circulating levels correlating with bowel disease activity [56]^[53]. Moreover, it can promote the activation of key players in fibrogenesis occurring in intestinal and pancreatic inflammation. The enhanced concentration of IL-33 in acinar cells and pancreatic stellate cells and the IL-33-mediated production of profibrogenic molecules suggest its role in CP pathogenesis [57]^[54]. Further studies are needed to estimate the actual prevalence of CP in IBD patients and to evaluate the correlation of CP with IBD, mainly with bowel activity flares over the years. Interestingly, a significant gut microbiota dysbiosis has been reported in CP patients, with increased opportunistic pathogens and depleted reduced taxa with a potentially beneficial role in intestinal barrier function. However, it is reasonable that these changes are primarily the result, rather than the cause, of the CP, probably due to the altered composition of the intestinal content, with resulting changes in the availability of substrates for microbial metabolism [27]^[26]. Moreover, no data about the effects of gut microbiota composition on the progression and severity of CP are available. So, the relationship between IBD and CP, mediated by microbiota alterations, cannot be proved. Finally, considering the occurrence of pain in the setting of CP as well as in the natural history of IBD, the role of the enteric nervous system (ENS) and neuronal plasticity might be an interesting field for investigation [58]^[55]. The enteric neurons and intestinal immune cells share common regulatory mechanisms, responding to environmental factors, and the relay of neurotransmitters and neuropeptides is indispensable for effective immunity and tissue homeostasis. Emerging studies [59,60,61,62]^{[56][57][58][59]} have started to elucidate the role of specific immune mediators, including mast cells and the related cytokine release, macrophages and T cells, with or without the interaction with microbiota, in mediating pain sensitivity in the IBD context. The translation of these observations in the clinical setting have led to the possibility to stimulate the vagus nerve for the pain control and outcome in CD patients with promising results [63]^[60].

2.4. Exocrine Pancreatic Insufficiency

EPI consists of a deficiency of pancreatic enzymes, resulting in the inability to digest food properly. It is usually associated with pancreatic diseases, mainly CP, due to the progressive depletion of pancreatic exocrine function. However, the cause may be the alteration in the digestive chain involving the pancreas, including pancreatic stimulation, pancreatic juice production or outflow, and synchronization with gastrointestinal secretions for mechanical and functional reasons [65]^[61]. Thus, pancreatic and extrapancreatic diseases such as IBD may be associated with EPI. Symptoms of EPI include steatorrhea, abdominal distension, flatulence, and/or weight loss, whose intensity may range from scarce symptoms to the impairment of quality of life. Moreover, the resulting malnutrition may be responsible for osteopenia, sarcopenia, reduced immunocompetence, and a nutritional deficiency [65]^[61]. The prevalence of EPI in the general population is unknown. It is mostly a late-stage manifestation of CP, occurring when 90% of the pancreatic gland is lost, but it could be influenced and anticipated when recurrent pancreatic disease flares occur [66]^[62]. On the other hand, the frequency of EPI in IBD patients varies considerably, ranging between 18 and 80% of cases, depending on patient selection and the diagnostic tests used (PABA test, secretin-caerulein test, qualitative or quantitative fecal fat test, and more recently fecal elastase-1, FE-1). In a prospective sectional study by Maconi et al., 14% of patients with CD and 22% of patients with UC suffered from EPI, using the FE-1 levels as a diagnostic tool. Compared to control subjects, the odds ratios (OR) for EPI were 8.34 for patients with CD and 12.95 for patients with UC. Moreover, in CD, EPI seems to be related to the extension and activity of bowel disease, mainly in the ileal location [67]^[63].

Pathogenesis of EPI in IBD

Some authors suggest that EPI in IBD patients may result from idiopathic CP, secondary to ductal obstructive changes, or previously undiagnosed AIP. However, as mentioned above, EPI in CP is a late-stage symptom, and the entity of ductal changes is insufficient to explain the EPI onset in these patients. Another hypothesis of IBD and EPI association is the immune-mediated mechanism, suggested by the presence of autoantibodies against exocrine pancreas (PABs), found in up to 39% of CD patients and up to 23% of UC patients. The PABs belong to the IgG and IgA isotypes and show two distinct response patterns, suggesting at least two different pancreatic autoantigens as targets of the autoimmune responses. The identity of PAB-specific autoantigens has been recently elucidated as a glycoprotein predominantly expressed in the pancreas and known as GP2. It was previously believed that GP2 was exclusively expressed by pancreatic acinar cells, but recent studies showed it also to be in the epithelium of Peyer’s patches. However, the role of exocrine pancreatic autoantibodies in inducing EPI was postulated in other autoimmune diseases, such as by Sjogren [58]^[55]; no data about the association of PABs with EPI or with morphologic changes in the pancreatic duct in IBD patients are available. In a study by Barthet et al. assessing the frequency of radiological and biological alterations in IBD patients with a previous AP episode by comparing data with IBD patients without a history of AP, no differences were found in serum levels of PABs (p = 0.17), although the higher rate of reduced FE-1 in AP experienced patients [68]^[64]. To date, the role of PABs has been studied as a possible disease marker in association with ANCA and ASCA autoantibodies rather than as a cause of EPI. Only one study [69]^[65] documented the most frequent presence of extraintestinal autoimmune manifestations in CD patients with PAB compared to seronegative cases. Further prospective studies are needed to explore the clinical meaning of these observations.