Ingrid Schwartz, MD1, Tara Keihanian, MD, MPH1, Prateek Sharma, MD, FACG2, Jinendra Satiya, MD3, Harjit Singh4, Rtika Abraham, MD1, Mohit Girotra, MD1; 1University of Miami Health System, Miami, FL; 2University of Miami, Miami, FL; 3New York Medical College Metropolitan Hospital, New York, NY; 4University of Miami Miller School of Medicine, Coral Gables, FL
Introduction: Endoscopic Ultrasound (EUS) can detect subtle pancreatic parenchymal and ductal abnormalities before they are evident on radiological imaging. Patients with Diabetes Mellitus (DM) are at increased risk of acute pancreatitis and vice-versa. However, the prevalence and clinical significance of subtle parenchymal or ductal abnormalities in pancreas of DM patients are unclear of-date and remains a field of immense interest. Methods: This is a retrospective chart review of 103 DM patients, between 2010-2018, who underwent EUS for unexplained abdominal pain (n=34) or other indications including abnormality found in endoscopy (n=13), biliary tree abnormality (n=27), abnormal laboratory values (n=9), abnormal imaging (n=9), staging for known non-pancreatic malignancy (n=3) and recurrent pancreatitis (n=8). Pancreatic parenchymal and ductal abnormalities were recorded and classified using the Rosemont criteria. Results: 103 DM patients (93.2% type-2; 6.8% type-1) had mean age and BMI of 64.9±14 years, and 31.65±9.2 respectively, of which 61.2% (63/103) were females. Average duration of DM in this cohort was 9.6±6.2 years, with mean HbA1c of 7.7±1.9%. The odds of finding any minor Rosemont minor criteria on EUS in DM patients presenting with abdominal pain were 3.2 times higher in male gender (95%CI: 1.4-7.5) and 3.3 times higher with history of alcohol use (95%CI: 1.33-8.42). All other variables tested on univariate analysis (age ≥ 65, race, ethnicity, diabetes duration >10 years, HbA1c≥6.5, presence of diabetes complications, tobacco use, reason for EUS being exclusively unexplained abdominal pain) had no statistically significant association with the finding of any minor Rosemont criteria on EUS. In binary regression model, male gender (OR=4.1; 95%CI: 1.5-11.3; p=0.005), history of alcohol use (OR=4.3; 95%CI: 1.4-13.9; p=0.013) and BMI > 30 kg/m2(OR=4; 95%CI: 1.3-12; p=0.013) were most predictive of the finding any minor Rosemont criteria on EUS in patients with DM irrespective of procedure indication. Discussion: Pancreatic parenchymal and ductal abnormalities not meeting Rosemont criteria diagnostic of CP appear to be present in patients with DM. In our cohort, one patient met required Rosemont criteria for CP and 35.9% (37/103) had one or more minor Rosemont feature in their EUS. Alcohol use and male gender are independently associated with the presence of any of the Rosemont criteria.
TABLE 1: Results
Disclosures: Ingrid Schwartz indicated no relevant financial relationships. Tara Keihanian indicated no relevant financial relationships. Prateek Sharma indicated no relevant financial relationships. Jinendra Satiya indicated no relevant financial relationships. Harjit Singh indicated no relevant financial relationships. Rtika Abraham indicated no relevant financial relationships. Mohit Girotra indicated no relevant financial relationships.