2211 The Ghost of TB: Danger of an Indolent Infection in the Setting of Alcoholic Hepatitis

Abstract

INTRODUCTION: The benefits of corticosteroids in patients with severe alcoholic hepatitis has recently been questioned. Corticosteroids should be used with caution for a febrile patient in the setting of ETOH Hepatitis. We present a case in which steroids lead to a flare-up of an indolent infection and dissemination of tuberculosis. CASE DESCRIPTION/METHODS: A 23-year-old male who migrated from Mexico 8 years earlier with a past medical history of alcohol-related liver disease presented with fever, diarrhea and abdominal pain. On admission, labs were suspicious for alcoholic hepatitis (Table 1), paracentesis showed SAAG of 1.6 (Table 2) and complicated by portal HTN and SBP. He was started on Ceftriaxone and flagyl. He was started on steroids for 2 days that was replaced by pentoxifylline for the diagnosis of SBP then back to steroids after peritoneal fluid culture was negative. He continued to have a persistent fever which was attributed to severe alcoholic hepatitis for which he had a repeat paracentesis (2) that showed a lymphocytic predominance with SAAG 1.7. TB tests were ordered and showed indeterminate Quantiferon, negative AFB smear/TB PCR from paracentesis x2 and HIV was negative. Antibiotics (Zoysn) was restarted due to the persistent fever. A liver biopsy was consistent with alcoholic steatohepatitis with grade 3-4 fibrosis. Lille score was calculated at 0.010 so steroids were continued. The patient developed hemoptysis with EGD showing only grade I nonbleeding esophageal varices. Repeat CT of the lung revealed numerous new pulmonary nodules. Steroids were stopped by a rapid taper given the new pulmonary nodules. He had a diagnostic bronchoscopy and BAL grew mycobacterium TB. His ascitic fluid culture from three weeks prior eventually revealed Mycobacterium species. He was initiated on treatment for disseminated tuberculosis with RIPE therapy. The patient then developed extensive hemoptysis complicated by PEA. CPR achieved ROSC but he was made CMO and passed away. DISCUSSION: Tuberculous peritonitis is rare in the United States but continues to be reported to occur in certain high-risk populations. The diagnosis of this disease requires a high clinical index of suspicion especially in the presence of ascites with a lymphocyte predominance and serum-ascitic albumin gradient of <1.1 mg/dl. Microbiological or pathological confirmation remains the gold standard for diagnosis. Ascitic fluid cultures have a low yield, but peritoneoscopy with biopsy or cultures frequently confirms the diagnosis.