SAFETY OF ENDOSCOPIC ULTRASOUND-GUIDED TISSUE ACQUISITION ON DIRECT ORAL ANTICOAGULANTS: A RANDOMIZED PRECLINICAL TRIAL

Aims To assess the safety and outcome of endoscopic ultrasound-guided fine needle biopsy (EUS-FNB) without interruption of a direct oral anticoagulant in a porcine model.

Methods Twenty pigs were randomized (1:1) to oral apixaban or placebo and underwent EUS-FNB of the pancreas (22G needle). Apixaban (0.5mg/Kg/12h) was administered 3 days before EUS-FNB and continued for 72 hours to end of survival. Specimens were submitted for rapid on-site evaluation (ROSE) and histology. Hemoglobin levels were checked pre EUS-FNB and 72 hours later. Apixaban concentration was analyzed before EUS-FNB. Study endpoints were 1. Occurrence of clinically significant bleeding (CSB, a composite outcome of drop in hemoglobin level≥2g/dL and evidence of intraprocedural or postprocedural bleeding) 2. Diagnostic specimens defined by ROSE or histological confirmation of pancreas tissue and 3. Quality of specimens graded by the degree of blood contamination. Endosonographers and pathologists were blinded to treatment allocation.

Results CSB occurred in 1 animal in the apixaban group (p=1). Minor bleeding occurred in 10 animals, 6 from the apixaban group (5 intraprocedural bleeding; 10 hematoma at necropsy). Median drop in hemoglobin was 0.85g/dL (IQR: 0.3-1.35), without difference between groups (p=0.78). All specimens were considered diagnostic by ROSE or histology criteria. There was no difference in specimen quality for ROSE (p=0.21). In histological specimens, blood accounted for 74% (IQR: 45.6-96.3) of the core area in the apixaban group and 97.6% (IQR: 92.6-98.6%) in placebo group, p=0.05.

Conclusions EUS-FNB of the pancreas with apixaban did not significantly increase CSB, nor did it limit a diagnostic cytopathological evaluation.

Publication History

Article published online:
14 April 2022