INTRODUCTION: Spontaneous fungal peritonitis (SFP) is an uncommon, but serious complication of liver cirrhosis. In one study, the one-month mortality rate was 73.3% as compared to 28% in SBP alone. The mortality remains high despite appropriate antifungal therapy partly due to delayed diagnosis. CASE DESCRIPTION/METHODS: We present a case of a 67 years-old male patient who presented with acute confusion and worsening abdominal distension for unknown duration. He has history of heart failure with reduced ejection fraction secondary to ischemic cardiomyopathy and liver cirrhosis presumed secondary to heart failure. He was found to have acute hepatic encephalopathy, anasarca, acute kidney injury, and septic shock. Abdominal exam showed tender, distended abdomen with positive fluid thrill. Spontaneous bacterial peritonitis (SBP) was suspected and ascitic fluid analysis was sent. It reported nucleated cells 3766/uL, 80% neutrophils, SAAG 1.7, glucose <20 mg/dl, and total protein 3.2 g/dl. Computed tomography of abdomen and pelvis without contrast showed massive ascites, but there was no evidence of bowel perforation (Figure 1). He was treated with Vancomycin and Zosyn for SBP and lactulose for hepatic encephalopathy. On day 3, gram stain of ascitic fluid reported budding yeast, and he was started on anidulafungin. The final culture reported Candida glabrata. Bacterial and fungal blood cultures were negative. Despite treatment, his encephalopathy persisted, and he had rapid accumulating ascitic fluid. Ascitic fluid culture was repeated which reported the same. A repeat CT abdomen showed perihepatic hematoma and free intraperitoneal air (Figure 2). He underwent exploratory laparotomy that revealed duodenal perforation. Despite surgical repair and antifungal, septic shock progressed, and patient died from multi-organ failure. Candida glabrata was sensitive to anidulafungin on retrospective review of sensitivity report. DISCUSSION: Although initial ascitic fluid analysis suggested secondary peritonitis, CT scan was negative for perforation or loculated infection. Our case highlights the high mortality rate of Candida peritonitis despite appropriate treatment. It is likely secondary to delayed antifungal and undiagnosed bowel perforations.

Pembrolizumab, an immune checkpoint inhibitor (ICI) that acts against receptor programmed cell death-1 (PD-1), is currently being used in the treatment of a variety of cancers. As PD-1 is also present on other non-malignant tissues, this results in side effects involving a multitude of organ systems termed immune-related adverse effects (irAEs). Programmed cell death-1 is expressed on the beta cells of islets of the pancreas, and their destruction can result in hyperglycemia and the onset of new diabetes mellitus (DM). Thus, the anti-PD1 action of pembrolizumab can lead to autoimmune-related DM. We present a case of a 62-year-old male who developed new-onset DM after 12 cycles of pembrolizumab with a severe presentation in the form of diabetic ketoacidosis (DKA) and ICU stay. Our case underscores the importance of physician awareness, frequent lab monitoring and patient education about this rare but potentially fatal irAE of ICI. It also strengthens existing data in literature suggesting the association of irAEs with improved efficacy of ICI therapy.

No abstract