EUS-guided paracentesis for the diagnosis of malignant ascites.

Background: EUS and EUS-guided fine-needle aspiration (EUS-FNA) have well-defined roles in the diagnosis and staging of GI and pancreaticobiliary malignancy. Malignant ascites usually represents peritoneal carcinomatosis, increases disease stage, and portends a poor prognosis. There are limited data regarding the yield of EUS-guided paracentesis (EUS-P) for the diagnosis of malignant ascites.

Objective: To determine the usefulness of EUS-P for the diagnosis of malignant ascites.

Design: Prospective case series.

Setting: Tertiary referral academic center.

Patients: Those presenting for EUS examination for suspected or proven malignancy over a 16-month period were evaluated prospectively for the presence of ascites.

Interventions: EUS-P was performed via a transgastric or transduodenal approach if ascites was detected.

Main Outcome Measurements: Sensitivity, specificity, positive predictive value, and negative predictive value of EUS-P for diagnosing malignant ascites.

Results: Six hundred twenty-nine patients were studied. Twenty-five patients with ascites who met inclusion criteria comprised the study cohort. The mean volume of ascites aspirated was 6.8 mL (range, 1-20 mL). Sixty-four percent (16 of 25) of EUS-P samples revealed malignant cytology. Of the group with negative ascitic cytology, 67% (6 of 9) had a proven malignancy. There was one false-negative cytology result. The sensitivity, specificity, positive predictive value, and negative predictive value of EUS-P for diagnosing malignant ascites was 94%, 100%, 100%, and 89%, respectively. The complication rate was 4%; 1 patient developed bacterial peritonitis after EUS-P.

Limitations: The study did not address cost savings in patient care based on the diagnosis of malignant ascites.

Conclusions: EUS-P is highly sensitive and specific for diagnosing malignant ascites. The finding of malignant ascites significantly alters patient management, so an active search for ascites and use of EUS-P should be incorporated into the diagnosis and staging of upper GI and pancreaticobiliary malignancy.