Eus-guided liver biopsy provides diagnostically adequate cores and reduces length of hospital stay compared with percutaneous and transjugular techniques: results from a prospective observational study

Introduction

While non-invasive tests are increasingly used to risk stratify fibrosis, liver biopsy (LB) remains the gold standard for diagnosis and staging. Percutaneous LB (P-LB) and transjugular LB (TJLB) are well-established methods for tissue acquisition, while endoscopic ultrasound-guided liver biopsy (EUS-LB) is a novel approach. Our primary aim was to compare diagnostic adequacy of EUS-LB against P-LB and TJLB. Secondary aims were to assess adequacy of fibrosis staging, length of stay post-procedure and 30-day complication rates between different modalities.

Methods

Data was collected prospectively from patients undergoing non-lesional P-LB, TJLB or EUS-LB in a single tertiary hospital from September 2022 to February 2023. The study was approved by the Trust review board. P-LB used 16G Biopince® needle performed by hepatologists or interventional radiologists, TJLB used 18G Argon® needle by interventional radiologists and EUS-LB used a 19G Acquire® FNB needle by endohepatologists. Demographics, biopsy indication, cumulative specimen length, number of complete portal tracts (CPT), length of stay (LoS), 30-day complications and diagnostic yield were recorded. Diagnostic adequacy was assessed by comparing cumulative specimen length and CPTs between techniques, using BSG guideline standards (>11 CPTs; specimen length ≥ 20mm). Histopathology assessment blinded to technique gave a fibrosis score and clinical diagnosis for each sample.

Results

121 patients (51% male) had a liver biopsy. Mean age was 55.2 years, and was similar across groups (p= 0.82). Fifty nine (49%) biopsies were P-LB, 10 (8.2%) were TJLB and 52 (43%) were EUS-LB. Fifty eight samples reported number of CPTs and 100 reported total specimen length. Median cumulative length was 27mm, 26mm and 41.5mm for P-LB, TJLB and EUS-LB respectively (p<0.01). Median number of CPTs was comparable between techniques: 19 for P-LB, 14.5 for TJLB and 22.5 for EUS-LB, (p=0.29). Samples were sufficient for histological diagnosis in 97% in P-LB, 100% in TJLB and 97% of EUS-LB cases. Complications occurred in 2 following P-LB (significant bleed; pain) and 1 following EUS-LB (duodenal wall perforation); all resolved with medical management. Patients undergoing EUS-LB had a significantly lower median LoS (3.22 hours), than those undergoing P-LB (8.01 hours) and TJLB (4.55 hours);p <0.001. Adequacy for fibrosis staging was 96% in P-LB samples, 90% in TJLB and 95% of EUS-LB (p=ns).

Discussion

The diagnostic utility and adequacy of fibrosis staging using EUS-LB compared favourably with other established techniques. Length of post-procedural stay was significantly lower for patients undergoing EUS-LB, which may release cost-savings in Daycase services, but presents challenges in training and service delivery compared to traditional LB pathways. Future comprehensive liver assessment may include single session EUS elastography, variceal screening, biopsy and portal pressure assessment.