Introduction

Inflammatory bowel disease (IBD), including both Crohn’s disease (CD) and ulcerative colitis (UC), is an idiopathic, chronic, remitting/relapsing inflammation of the gastrointestinal tract [1]. The etiology and pathogenesis remain poorly understood but is considered a multifactorial inflammatory process involving environmental factors that result in alteration of intestinal immunity and dysbiosis of gut microbiota in genetically susceptible individuals [1, 2]. Extra-intestinal manifestations (EIM) involve up to 24% of patients with IBD [3, 4].

Patients with IBD are at an increased risk of acute pancreatitis (AP) when compared to the general population. Several studies have shown a 3-4-fold increase in risk of developing AP in the setting of CD and a 2-fold risk in the setting of UC [3,4,5]. While most cases of AP have an identifiable etiology, a subset of cases will be defined as acute idiopathic pancreatitis [2]. Evidence indicates that acute idiopathic pancreatitis is associated with IBD [6,7,8,9]. However, no study has explored the prognostic implications of acute idiopathic pancreatitis in the setting of IBD. To address these gaps in knowledge, this study aimed to assess whether acute idiopathic pancreatitis is associated with a more severe disease course in both CD and UC.

Materials and methods

Patients

Using the electronic medical record, a retrospective review was conducted across a tertiary care referral center. Inclusion criteria included patients 18 years or older with established diagnoses of IBD who then developed an episode of AP between March 2011 and June 2020. Patients with recurrent AP were noted; however, only the first episode of AP was included in the analyses. Patients with chronic pancreatitis were excluded from the study. AP was diagnosed by the presence of at least two of the following: (i) acute epigastric pain; (ii) elevated serum lipase ≥ 3 times the upper normal limit; (iii) characteristic CT findings [10]. Medication-induced AP was diagnosed when there was a temporal association between the introduction of the drug and the development of AP, and when symptoms resolved after discontinuation of the offending medication [11]. Acute idiopathic pancreatitis was defined as AP after excluding other possible etiologies (gallstone, alcohol, medication, hypertriglyceridemia, iatrogenic, autoimmune, or malignancy). Recurrent AP was defined if the patient presented with a new episode of AP, as defined above, after complete resolution of symptoms. All variables were collected for the first episode of AP. EIM was defined by the presence of at least one of the following: peripheral arthritis, axial arthritis, erythema nodosum, pyoderma gangrenosum, oral ulcers, ophthalmologic complications (including episcleritis, scleritis or uveitis), and primary sclerosing cholangitis. An aggressive IBD course was defined, utilizing similar parameters described previously in the literature, as the presence of at least one of the following: (i) change in biologic therapy within 1 year of AP; (ii) biologic dose escalation within 1 year of AP; (iii) IBD-related hospitalizations or surgeries within 1 year of AP [12].

Statistical analysis

Statistical analyses were performed using JMP Pro 14.1 (SAS Institute, Cary, NC) and R 4.0.4 (R Foundation for Statistical Computing, Vienna, Austria). Categorical data are expressed as numbers (%) and continuous variables are expressed as median (interquartile range), unless otherwise stated. Categorical and continuous variables were compared using Chi-square test and non-parametric Mann-Whitney U test, respectively. Statistical significance was determined by a P-value < 0.05. Logistic regression modeling was used to identify association between different covariates and an aggressive IBD course. Results were expressed as Odds Ratio (OR) and 95% Confidence Interval (CI). Since there were no more than 20 events in either the CD or the UC group, we could not include more than one variable in the logistic regression model [13,14,15].

Results

After exclusion of two patients with UC due to a chronic pancreatitis diagnosis, fifty-three patients were identified with IBD and a diagnosis of AP (n CD = 29, n UC = 24). There was a total of 60 episodes of AP in those 53 patients, in a total of 7231 IBD patients identified in this tertiary care institution (cumulative incidence of AP in IBD, 0.73%). We ultimately included a total of 20 patients with idiopathic pancreatitis (11 in the CD and 9 in the UC groups) and a total of 33 patients in the comparison group (non-idiopathic pancreatitis) (Fig. 1).

Fig. 1
figure 1

Flowchart of patient selection

Full size image

Baseline characteristics of patients with CD and UC are shown in Table 1. On presentation, 98.11% (52/53) of patients complained of abdominal pain, 50.94% (27/53) complained of nausea or vomiting, and 92.45% (49/53) had elevated lipase on admission. Patients were followed for a median of 2.03 (1.00-4.72) years after the diagnosis of AP. In the CD cohort, two patients had a complicated course of AP (including both idiopathic and non-idiopathic AP) – one patient had an episode of acute idiopathic pancreatitis complicated by pseudocyst and necrotizing pancreatitis, and one patient had an episode of endoscopic retrograde cholangiopancreatography (ERCP)-induced pancreatitis with pseudocyst formation. Similarly, one patient with UC with mesalamine-induced pancreatitis was complicated by pseudocyst. On follow-up, 26.42% (14/53) had recurrent AP (Fig. 2). The median duration of recurrence from the index AP was 8.50 (1.82-36) months. The median duration of recurrence of acute idiopathic pancreatitis was 3.50 (1.19–15.75) months. All patients who had recurrence of AP (including both idiopathic and non-idiopathic AP) had the same etiology as the initial episode of AP except for one patient with CD who had an episode of gallstone pancreatitis followed by an episode of ERCP-induced pancreatitis 6 years later (Footer of Table 1; Fig. 2). No patients died as a direct result of pancreatitis.

Table 1 Baseline characteristics of cohort
Full size table
Fig. 2
figure 2

Natural course of recurrent acute pancreatitis

Full size image

Twenty out of the 53 patients were diagnosed with acute idiopathic pancreatitis in this cohort (n CD = 11, n UC = 9). In patients with CD, there were no statistically significant different characteristics between patients with acute idiopathic pancreatitis as compared to other causes of pancreatitis. Duration of IBD, age at the diagnosis of AP and biologic treatment were not associated with the development of different types of AP. Similarly, in patients with UC there was no statistically significant differences between patients with acute idiopathic pancreatitis as compared to other causes of AP. Patients with UC and acute idiopathic pancreatitis had a significantly higher rate of recurrent pancreatitis compared to UC patients diagnosed with other types of pancreatitis. Six patients (66.67%) with UC developed recurrent acute idiopathic pancreatitis and one patient (6.67%) developed recurrent alcoholic pancreatitis (P-value < 0.01) (Table 2). No risk factors could be identified for developing acute idiopathic pancreatitis when compared to other causes of pancreatitis (Table 3).

Table 2 Baseline characteristics according to the type of pancreatitis
Full size table
Table 3 Univariable analysis of variables associated with development of acute idiopathic pancreatitis
Full size table

A more aggressive IBD course was identified in 63.64% (7/11) of patients with CD and acute idiopathic pancreatitis, as compared to only 11.11% (2/18) of patients with CD and other causes of AP (OR 6.67, 95% CI 1.12–56.45; P-value 0.04). None of the other variables were associated with an aggressive disease course in CD. A more aggressive IBD course was identified in 66.7% (6/9) of patients with UC and acute idiopathic pancreatitis, as compared to only 46.67% (7/15) of patients with UC and other causes of AP (OR 2.29, 95% CI 0.43–14.31; P-value 0.35) (Table 4).

Table 4 Univariable analysis of variables associated with aggressive disease course
Full size table

Discussion

While there are a few studies exploring the association between IBD and AP, this cohort reflects the first focused on the association of IBD and acute idiopathic pancreatitis. These results suggest that acute idiopathic pancreatitis is associated with a more aggressive disease course in CD, but no similar association was identified with UC. Based on these results, the presence of acute idiopathic pancreatitis may represent an EIM and a prognostic factor in identifying a more aggressive disease course in CD.

Acute idiopathic pancreatitis accounts for ~ 10–35% of AP cases in the general population [16, 17]. Interestingly, acute idiopathic pancreatitis accounted for a higher percentage of AP cases in our cohort for both CD and UC (37.93% and 37.50% respectively) indicating that IBD may be a risk factor for developing acute idiopathic pancreatitis. The explanation for the association between IBD and acute idiopathic pancreatitis is unclear and may indicate that acute idiopathic pancreatitis is an EIM of IBD. Authors have speculated that pancreatitis might be an EIM of IBD [2]. AP only accounted for 0.73% of patients with IBD in our institution, however, similarly low incidence rates are found in other EIMs of IBD [18, 19]. EIM such as peripheral arthritis, erythema nodosum, and episcleritis have been associated with disease activity; however, other EIM such as uveitis or ankylosing spondylitis do not correlate with disease activity [20]. Our results demonstrate that acute idiopathic pancreatitis was associated with a more severe disease course in CD, but not in UC. However, it is not unusual for CD and UC to have different prognostic factors e.g., different studies have shown that smoking, antibiotic exposure and physical activity have varying effects on CD, compared to UC [21,22,23].

A possible association between acute idiopathic pancreatitis and IBD might involve a common genetic susceptibility locus between IBD and AP such as the myosin IXB (MYO9B), membrane-associated guanylate kinase inverted 2 (MAGI2) genes, and the gene encoding the partition-defective 3 protein (PARD3 gene) [24]. These genes are tight junction genes coding for scaffolding proteins that affect intestinal permeability. Alterations in the gut microbiota has been shown to be associated with IBD [25]. Such microbiota alteration may be associated with the development of pancreatitis, as noted by various studies demonstrating abnormally low levels of Bifidobacterium and Lactobacillus, as well as abnormally high levels of Enterobacteriaceae in patients with pancreatitis [26, 27].

Understanding the difference in genetic associations, microbiota dysbiosis and immune dysregulation in CD and UC is a possible key to understanding the differences noted between both diseases and might help explain the opposing association between acute idiopathic pancreatitis and the disease severity in CD and UC. An example of a possible factor that may explain the association between CD and acute idiopathic pancreatitis includes the microbiota changes of Bifidobacterium depletion in CD, that has similarly been observed in AP [23, 25, 28]. On the other hand, the previously mentioned MYO9B and PARD3 genes are more associated with UC [29, 30].

There are several limitations of this study. Firstly, it was a retrospective study conducted at a tertiary referral center. Secondly, given the rarity of the addressed disease, the sample size is small limiting its power and the generalizability of the results. This also limited the ability to conduct a multivariable analysis, as the number of variables to be included in a logistic model is dependent on the number of events [13,14,15]. As such, we urge the readers to interpret the results with caution. However, the cumulative incidence of AP in this cohort was similar to previously established data in the literature and thus can be considered an accurate representation of patients with AP and IBD [6, 31, 32]. In addition, the logistic regression model did not identify any other statistically significant variables, thus making it unlikely for the multi-variable analysis to have changed the results. Finally, more studies are needed to study the significance of AP, and in turn acute idiopathic pancreatitis, in an IBD cohort that consists of patients with both AP and without AP to eliminate concerns of inclusion bias.

In conclusion, this study presents data to support an association between acute idiopathic pancreatitis and IBD. The prevalence of acute idiopathic pancreatitis in the setting of IBD was noted to be higher than that in the general population. Furthermore, there is a statistically significant association between acute idiopathic pancreatitis and a more aggressive disease course in CD. Despite the lack of association between acute idiopathic pancreatitis and an aggressive disease course in UC, patients are at an increased risk of acute idiopathic pancreatitis recurrence which may contribute to increased morbidity. Thus, acute idiopathic pancreatitis might best be considered an EIM that prognosticates IBD severity. The high prevalence, association with aggressive IBD course and a possible higher recurrence rate of pancreatitis highlight a potential significant disease burden that acute idiopathic pancreatitis plays in patients with IBD.