93 Variations of Diagnosis and Management of Immune Checkpoint Inhibitor Pancreatic Injury (ICIPI) and Immune Checkpoint Inhibitor Pancreatitis: A Single Institution Experience

INTRODUCTION: Immune checkpoint inhibitors (ICI) have revolutionized the treatment of malignancies, but are limited by immune-related adverse events (iRAE). Pancreatic iRAEs, broadly defined by asymptomatic lipase elevation to complicated pancreatitis, lack clear management. We seek to report rates of varying definitions, evaluations, and management of pancreatic iRAEs, classified as immune-checkpoint inhibitor pancreatic injury (ICIPI) and ICI pancreatitis. METHODS: From a database of 1672 patients treated with ICI, data was collected on those with possible pancreatic injury. They were classified into four groups: ICIPI, as defined by either lipase elevation or imaging of pancreatitis due to ICI; ICI pancreatitis, as defined by fulfilling Atlanta criteria; non-ICI asymptomatic lipase elevation; non-ICI pancreatitis. Data on rates of diagnosed pancreatitis, lipase elevations, imaging evidence, symptomatology, treatment, management of ICI, and patient outcome after cessation of ICI were collected. RESULTS: Of 31/1672 (2%) patients with pancreatic injury, 7 had ICI pancreatitis, 14 had ICIPI, 6 had non-ICI pancreatitis, and 4 had non-ICI asymptomatic lipase elevations. In the 7 ICI pancreatitis patients meeting Atlanta criteria, 5/7 (71%) had follow up evaluation to rule out other etiologies of pancreatitis. In contrast, only 1/13 (8%) ICIPI patients had a follow up evaluation to meet Atlanta criteria or rule out other etiologies. Of the ICIPI patients, 6/14 (43%) were diagnosed as pancreatitis, 8/14 (57%) were asymptomatic, 9/14 (64%) had lipase >3x upper limit of normal (ULN), and only 1/14 (7%) ICIPI was due to imaging alone (Table 1). Most ICI pancreatitis, 5/7 (71%), and ICIPI, 10/14 (71%) patients received corticosteroids (CS). While 4/7 (57%) ICI pancreatitis patients received intravenous fluids (IVF), only 2/14 (14%) of ICIPI patients did. Of 6/7 (86%) of ICI pancreatitis patients that stopped ICI, 3/6 (50%) reinitiated them. In contrast, 7/14 (50%) of ICIPI patients stopped ICI and 5/7 (71%) reinitiated them (Table 2). CONCLUSION: Pancreatic injury is known to be associated with ICI. Efforts to identify and treat those with pancreatic iRAE are critical to avoid delay in ICI therapy. In our study, most patients with ICIPI and ICI pancreatitis received CS. Almost all ICI pancreatitis patients stopped ICI with half reinitiating them, while only half of ICIPI patients stopped ICI with most reinitiating them. Further characterizing outcomes to develop a diagnostic algorithm will enhance care.