Recognizing Exocrine Pancreatic Insufficiency in Patients with Diabetes: A Case Study

Introduction

Exocrine pancreatic insufficiency affects the ability of the pancreas to secrete sufficient pancreatic digestive enzymes to the small intestine, leading to maldigestion and malabsorption. EPI is most commonly caused by diseases that destroy the pancreatic parenchyma, such as chronic pancreatitis, cystic fibrosis, and pancreatic ductal adenocarcinoma, or as a result of pancreatic resection procedures. Non-pancreatic parenchymal etiologies such as untreated celiac disease, inflammatory bowel disease, hypersecretory states such as Zollinger-Ellison syndrome, prior intestinal surgery, and Roux-en-Y bypass surgery have also been shown to be associated with EPI [1,2]. The prevalence of EPI in patients with type 2 diabetes ranges from 16.8% to 49.2% across various systematic reviews, with a median of 29% [1,3]. Despite being considered a “moderate-risk” clinical condition, EPI in patients with diabetes is under-recognized, partially due to significant overlap with other gastrointestinal symptoms caused by complications or treatment of diabetes, low clinical suspicion, and lack of routine testing [4]. Without appropriate treatment, EPI can result in significant nutritional deficiencies, weight loss, osteoporosis, and decreased quality of life. Therefore, clinicians should keep a high index of suspicion for EPI in appropriate clinical settings to allow for earlier diagnosis and treatment to prevent consequences of untreated EPI.

Case Report

A 71-year-old man with a longstanding history of type 2 diabetes mellitus (A1c 7.3% at time of presentation), Roux-en-Y bypass surgery 35 years ago, cholecystectomy 50 years ago, and desmoplastic melanoma of the left neck status post wide local excision 3 years prior, presented with new-onset severe diarrhea resulting in 41-kg weight loss, progressive muscle weakness, and recurrent falls. He reported a sudden onset of 10–15 oily, greasy stools per day and excessive foul-smelling gas approximately 8 months prior to presentation. His diarrhea and weight loss persisted without improvement despite consuming five 1000-calorie meals per day and trialing several over-the-counter anti-diarrheal medications. He had no history of smoking, denied any alcohol consumption following his bariatric surgery, and had no history of heavy drinking, and no acute or chronic pancreatitis. He was diagnosed with type 2 diabetes on screening A1c approximately 15 years prior to presentation and had fluctuating glycemic control, with the highest documented A1c of 12.7% and blood sugars in the 600s, 2 years prior to presentation. He had discontinued metformin due to intolerance several years prior and his diabetes was treated with insulin degludec 8 units every morning at the time of presentation. He was also adherent to 200 mg oral iron supplementation. Physical examination was notable for weight of 84 kg (body mass index 26.8 kg/m2), mild bilateral triceps and iliopsoas weakness, and bilateral peripheral edema up to the knees. Laboratory findings revealed multiple vitamin deficiencies: 25-hydroxy D total 9.4 ng/mL, vitamin A 15.8 mcg/dL, vitamin E 4.7 mcg/dL, total serum calcium 7.9 mg/dL, and albumin 2.8 g/dL. B12 and B1 were unremarkable at 254 ng/L and 221 nmol/L, respectively. A comprehensive malignancy workup, including skull-to-thigh PET-CT scan given his history of desmoplastic melanoma, colonoscopy with random colon biopsies, and celiac disease testing were unremarkable. Given his proximal muscle weakness, EMG was performed and showed length-dependent peripheral neuropathy (consistent with known diabetic neuropathy) and myopathy, with no evidence of neuromuscular transmission defects. Biopsy of the right triceps muscle was performed and demonstrated type 2 fiber atrophy with rimmed vacuoles, consistent with nutritional myopathy. CT enterography demonstrated marked pancreatic atrophy without any sign of pancreatic tumor (Figure 1). A 48-hour fecal fat study demonstrated massive steatorrhea (total fat in 24 hours of 277 g) and fecal elastase less than 40 μg/g. He was started on pancreatic enzyme replacement therapy (PERT) with 2 tablets of lipase-protease-amylase (24 000–76 000–120 000 units) by mouth 3 times a day with meals. His diarrhea improved 2 months after initiation of PERT, allowing him to regain weight to his previous baseline while reducing his caloric intake to 1800 to 2000 calories per day.

Discussion

Gastrointestinal symptoms such as abdominal pain, bloating, nausea, diarrhea, and dyspepsia are common in patients with type 1 and 2 diabetes mellitus, occu rring in up to 75% of patients presenting to diabetes clinics [5]. Given the nonspecific nature of these symptoms, they are often attributed to effects of autonomic neuropathy such as diabetic enteropathy and gastroparesis or adverse medication effects from metformin or GLP-1 agonists. Common comorbid conditions such as small-intestine bacterial overgrowth, celiac disease, gastroesophageal reflux disease, or irritable bowel syndrome can also present similarly.

In patients with diabetes presenting with such nonspecific GI symptoms, particularly steatorrhea with or without diarrhea, bloating, weight loss, or fat-soluble vitamin deficiencies, fecal elastase-1 (FE-1) testing should be considered, as it is an affordable and non-invasive test to evaluate for EPI. FE-1 measures a pancreatic exocrine-specific enzyme that indicates overall pancreatic secretion, with levels <200 μg/g suggestive of moderate EPI, and <100 μg/g indicating severe EPI [6]. The overall sensitivity for moderate and severe EPI is 100%, while sensitivity for mild EPI is lower, at 63% [7]. In a study of 288 consecutive diabetes patients (72% with type 2 diabetes and 28% with type 1 diabetes) presenting for regular diabetes management, fecal elastase concentrations were measured in all patients who endorsed having types 5–7 Bristol stools, steatorrhea, or weight loss. Of those with GI symptoms, 42% of patients had a low fecal elastase consistent with EPI, highlighting the prevalence of this condition in patients with diabetes [8]. While fecal fat testing can be performed to quantify steatorrhea, defined as coefficient of fat absorption (CFA) <93% (CFA=(fat intake-fecal fat excreted/fat intake) ×100), formal testing for steatorrhea requires adherence to diet of known fat content for 5 days with stool collection over the final 3 days, and given the logistical challenges, is not typically performed outside of clinical research [4].

While EPI can cause significant weight loss resulting in a decrease in body mass index (BMI) such as in our patient, normal or elevated BMI should not decrease the suspicion for EPI. One study reported that diabetic patients with a BMI above 25 were more likely to have EPI, although this association was not seen in non-diabetic controls, suggesting a possible synergistic effect [9].

The underlying pathophysiology of EPI in type 1 and 2 diabetes has not yet been fully elucidated but is thought to be multifactorial. Insulin is a trophic factor for pancreatic acinar tissue and its depletion, particularly in type 1 diabetes, can cause pancreatic atrophy. Diabetic microangiopathy and autonomic neuropathy have been proposed to cause ischemia of the exocrine pancreas and impaired enteropathic reflexes, respectively [10]. Hyperglucagonemia is also a well-documented feature of type 2 diabetes due to impaired insulin signaling and has been associated with inhibition of pancreatic enzyme production [1]. One recent meta-analysis demonstrated that in type 2 diabetes, the prevalence of EPI was positively correlated with insulin requirements [11]. Similarly, a study examined rates of EPI in patients with good (A1c <7%) and poor (A1c >7%) glycemic control as defined by the American Diabetes Association and found fecal elastase levels were significantly lower in patients with poor glycemic control, but there was no correlation with duration of diabetes or with BMI [12]. However, Softeland et al demonstrated a significant association between the duration of type 1 and 2 diabetes and EPI, defined by FE-1 <200 μg/g (linear regression, r=−0.27, P=0.009), and suggested a biologic plausibility, as the duration of diabetes may be associated with increased frequency of complications such as autonomic neuropathy, which can contribute to development of EPI [13]. In addition to type 1 and 2 diabetes, type 3c diabetes mellitus, which occurs as a result of pancreatic β- and α-cell damage in the setting of exocrine pancreatic disease, can also cause EPI [14]. Type 3c diabetes was considered in our patient, given the pancreatic atrophy seen on CT imaging; however, the lack of exocrine pancreatic disease such as chronic pancreatitis and onset of diabetes with history of morbid obesity (peak body mass index of 54 kg/m2 prior to bypass surgery) at least 15 years prior to the development of EPI symptoms decreased the clinical suspicion in this case. Other potential causes of EPI in this patient included EPI as a complication of bariatric surgery, given his history of Roux-en-Y gastric bypass, as the procedure bypasses the duodenum, preventing normal stimulation of pancreatic enzyme secretion. However, recent studies report a mean time from bariatric surgery to EPI diagnosis ranging from 12 months to 5 years, making this etiology less likely in our patient, given that his surgery occurred approximately 35 years prior to symptom onset [2,15].

PERT is the cornerstone of EPI treatment and is effective in management of the condition, with improvement in gastrointestinal symptoms such as steatorrhea, diarrhea, stool frequency, and bloating, as well as improved fat and protein absorption [16,17]. Current guidelines for PERT in patients with diabetes and EPI demonstrate variability in dosing recommendations. The most commonly recommended dose is 40 000–50 000 units of lipase per meal, with increases of up to 2–3 times this amount before pursuing additive therapies, which mirrors recommendations for PERT dosing in other co-conditions [16]. A recent retrospective study also demonstrated the safety of PERT usage in diabetes, with no significant differences in severe hypoglycemia rates after PERT initiation. In fact, hypoglycemia symptoms improved in 80% of patients with pancreatogenic diabetes [18]. A retrospective, subgroup analysis of a double-blind, randomized, placebo-controlled trial study of pancrelipase in patients with EPI with and without diabetes found no effect on glucose levels or insulin dosing, but did find a reduction in mild and moderate hypoglycemia in those receiving PERT [17]. Our patient remained on the same dose of insulin, degludec 8 units daily, while on PERT, and did not experience any new hypoglycemia or hyperglycemia.

Conclusions

The patient presented in this case report was diagnosed with EPI based on his clinical symptoms, low fecal elastase, and improvement of symptoms with initiation of PERT. While EPI is not uncommon in patients with type 2 diabetes mellitus, the diagnosis can easily be delayed or overlooked given the high rate of overlapping symptoms with other gastrointestinal disease comorbid with diabetes, and clinicians should maintain a high index of suspicion for EPI in patients presenting with steatorrhea with or without diarrhea, abdominal pain, and bloating.