Kidney Week

Abstract: FR-PO1003

The AngioJet "Blood Leak": A Diagnostic Pitfall in Hemodialysis

Session Information

• Hemodialysis: Clinical Challenges, Patient-Centered Outcomes, and Quality of Life
  October 23, 2026 | Location: Exhibit Hall A, Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Dialysis

• 801 Dialysis: Hemodialysis and Frequent Dialysis

Authors

• Kapyur, Amitha, Massachusetts General Hospital, Boston, Massachusetts, United States

Amitha Kapyur, MD, MS, MSc

• Apodaca, Javier, Massachusetts General Hospital, Boston, Massachusetts, United States

Javier Apodaca, MD

• McFadden, Kathleen, Massachusetts General Hospital, Boston, Massachusetts, United States

Kathleen McFadden, MD

• Seethapathy, Harish Shanthanu, Massachusetts General Hospital, Boston, Massachusetts, United States

Harish Shanthanu Seethapathy, MD, MBBS

• Steele, David J. R., Massachusetts General Hospital, Boston, Massachusetts, United States

David J. R. Steele, MD

• Tolkoff-Rubin, Nina E., Massachusetts General Hospital, Boston, Massachusetts, United States

Nina E. Tolkoff-Rubin, MD

• Lundquist, Andrew L., Massachusetts General Hospital, Boston, Massachusetts, United States

Andrew L. Lundquist, MD, PhD

Introduction

The blood leak detector in hemodialysis is a critical safety mechanism designed to identify dialyzer membrane rupture. While normally triggered by whole blood, these sensors can also detect free hemoglobin. This results in a "false-positive" blood leak alarm that may serve as a critical early marker for profound intravascular hemolysis rather than mechanical circuit failure.

Case Description

A 73-year-old woman with end-stage kidney disease presented with arteriovenous fistula thrombosis. Following a fistulogram, clot removal with Fogarty balloon and Angiojet thrombectomy of the inflow and outflow tracts was accomplished. Soon after the procedure, dialysis was initiated but a blood leak alarm activated, prompting termination of the session without return of blood. Concerns arose regarding a “faulty membrane", so another machine was set up. However, a blood leak was detected again, and testing of the dialysate revealed the presence of blood. Pre-dialysis labs obtained were reported as “hemolyzed.” Dialysis was not attempted again and instead labs and electrolytes were closely monitored. Hemolysis studies revealed haptoglobin 17 mg/dL, LDH 846 U/L, indirect bilirubin 1.2 mg/dL and retic 6.4%. The patient subsequently developed shock and was found to have a prominent leukocytosis (60.89 K/uL), which led to admission to the ICU. She was started on vasopressor support and broad-spectrum antibiotics. Unfortunately, her clinical condition deteriorated due to acute limb ischemia, persistent shock and respiratory failure, leading to transition to comfort measures.

Discussion

The blood leak alarm in hemodialysis is classically interpreted as a sign of dialyzer membrane rupture, necessitating immediate termination of therapy to prevent blood loss and dialysate contamination. Our case illustrates a critical diagnostic pitfall where recurrent alarms across multiple machines were actually triggered by free hemoglobin (32 kDa). Unlike whole red blood cells, free hemoglobin is small enough to pass through an intact high-flux semipermeable membrane and reach the optical sensors in the dialysate compartment. In this clinical scenario, the use of an AngioJet thrombectomy device for clot removal likely caused significant mechanical hemolysis. Clinicians should recognize that a "false-positive" blood leak can serve as an early marker for profound intravascular hemolysis, notably in patients who have recently undergone mechanical thrombectomy.