Case Presentation: A 41 year old female from Haiti with a history of uterine fibroids and menorrhagia presents to the emergency room with subacute abdominal distension. A CT shows moderate ascites with peritoneal enhancement and infiltration of mesenteric fat by soft tissue densities suggestive of omental caking, thus favoring peritoneal carcinomatosis versus peritonitis. Serum CA-125 is elevated to 205 U/mL and thus a malignant process is suspected. TVUS does not demonstrate an ovarian mass. She undergoes endometrial biopsy only revealing weakly proliferative endometrium. A diagnostic paracentesis yields a serum-albumin-associated gradient (SAAG) of 0.6 g/dL and WBC 1846 (11% PMNs and 57% lymphocytes), thus ruling out spontaneous bacterial peritonitis (SBP). Ascites cytology is negative for malignant cells. MRI of the pelvis demonstrates myomatous uterus, ascites, and unremarkable adnexa.Given the negative malignancy work-up, serum QuantiFERON-gold is checked and results positively. Though she had initially denied risk factors for mycobacterium tuberculosis (MTB), the patient later discloses she may have had a prior positive TB screening test but never sought confirmation or treatment. Repeat paracentesis is ultimately negative for acid fast bacilli (AFB) smear/culture and MTB-PCR, but shows elevated adenosine deaminase (ADA) to 72 U/L (ref range 0-30 U/L). The patient desires discharge and is started on empiric Rifampin, Isoniazid, Pyrazinamide, and Ethambutol (RIPE). At her follow-up appointment she reports improved symptoms since starting RIPE. She is continued on treatment and undergoes diagnostic laparoscopy for confirmation of TB. The operative report notes copious adhesions with omental caking and peritoneal deposits which are biopsied. Though AFB culture is negative, pathology demonstrates non-caseating granulomas.
Discussion: New onset ascites is a presenting symptom of many disease entities encountered by hospitalists. Utilizing a structured approach to undifferentiated ascites, hospitalists can accurately diagnose the underlying etiology. SAAG predicts portal hypertension with 97% accuracy. If SAAG is < 1.1g/dL, the differential diagnosis is narrowed. The differential for low SAAG ascites includes carcinomatosis, peritoneal TB, pancreatitis, serositis, and nephrotic syndrome. The gold-standard for diagnosis of TB is isolation of MTB from bodily fluid or tissue. In suspected abdominal TB, ascites should be tested for AFB and MTB culture, though each is associated with low positivity rates, thus peritoneal biopsy may be required. ADA >39 U/L can support diagnosis of abdominal TB but is not diagnostic in isolation. On biopsy, the presence of thickened peritoneum, yellow-white tubercles, or dense adhesions is characteristic of TB. Though biopsy has higher sensitivity for TB detection compared to ascites studies, TB remains a clinical diagnosis in 15-20% of cases.
Conclusions: A systematic approach to new-onset ascites efficiently narrows the differential. Paracentesis should be performed in all patients with new-onset ascites. Abdominal TB should be suspected in the appropriate clinical context of low SAAG with lymphocytic predominance. The gold-standard for diagnosis is isolation of MTB from bodily fluid or tissue. Ascites studies are often low-yield for TB, requiring sampling via biopsy. Abdominal TB remains a clinical diagnosis in many cases, and empiric treatment should be initiated in the appropriate context even in the absence of definitive microbiologic data.