EUS-guided Versus Percutaneous Ultrasound-guided Biopsy for Parenchymal Liver Disease

Aim and Objective - EUS-LB with a 19G Franseen needle is non-inferior to PC-LB with a 18G BioPince Needle, with regards to total specimen length and diagnostic adequacy.

Hypothesis - i. EUS-LB with a 19G FNB needle is non-inferior to PC-LB with a 18G full core biopsy needle, with regards to total specimen length and diagnostic adequacy.

ii. There would be lesser procedure related pain and higher patient satisfaction with EUS-LB compared to PC-LB.

Study population:

• Patients > 18 years of age.
• Eligibility for deep sedation with propofol.
• Undergoing liver biopsy for diagnosis/evaluation/staging.

Study design: All consecutive in-patients and out-patients requiring liver biopsy for evaluation of diffuse parenchymal liver disease will be evaluated for inclusion.

Study period: 1 year

• Sample size with justification: The percentage success rate of procurement of adequate liver biopsies (defined by specimen length ≥ 15 mm and ≥ 8 CPTs) was assumed to be 90% with PC-LB based on two recent retrospective studies comparing PC-LB with EUS-LB. These studies used the bigger 16G Biopince needles. (Literature: Bhogal N. Endosc Int Open 2020, 99%; Facciorusso A. Cancers 2021, 100%). With similar criteria for adequacy, the success rates with EUS-LB were assumed to be 85% for non-inferiority comparison, based on recent studies of EUS-LB using 19G Franseen needles. (Literature: Hashimoto R. Dig Dis Sci 2020, 100%; Aggarwal SN. GIE 2021, 97.2%; Nallapeta N. Hepatology 2021, 55.6%)
• With alpha-value of 10%, power of 80%, and non-inferiority margin of 10%, we need to enroll 44 patients in each of PC-LB and EUS-LB arms, for a total sample size of 88 patients. We plan to enroll 90 patients in this trial to account for technical inadequacies.
• Intervention: Patient after screening for all exclusion criteria will be randomized into either EUS guided liver biopsy or USG guided percutaneous biopsy
• Monitoring and assessment: All patients would undergo vital and baseline parameter screening before randomization. Based on randomization they will undergo the liver biopsy procedure. Post procedure all the patients will be closely observed in the recovery area for at least 4 hours after the procedure, VAS for pain will be noted & 4-point patient satisfaction scale will be provided to them. They will be followed up after 72 hours by a phone call or in their rooms if updated to check for any adverse effects.
• Pain will be assessed with a visual analogue scale (VAS) score (minimum score = 0, maximum score = 10).
• Patient satisfaction will be evaluated by using a four-point scale ('Very satisfied', 'Satisfied', 'Less satisfied' and 'Not satisfied at all').
• Statistical Analysis: Descriptive data following a normal distribution will be reported as means (+ standard deviation [SD]), whereas nonparametric data will be reported as median (range). Comparisons between the two needles will be performed with a t test, Wilcoxon rank sum test, or chi-square test as appropriate.
• A Mantel-Haenszel test of trend will be performed to determine whether a linear association existed between EUS-LB sampling and VCTE and CAP results.
• Adverse effects:

AEs will be defined as any deviation from the anticipated intra-procedure and post-procedural course.

Inclusion criteria

• Patients > 18 years of age.
• Eligibility for deep sedation with propofol.
• Undergoing liver biopsy for diagnosis/evaluation/staging.

Exclusion criteria

• Coagulopathy, defined by platelet count < 70,000/mL and/or INR > 1.7.
• Inability to discontinue anticoagulation or antiplatelet agents for 5 days prior to the procedure.
• Presence of unrelieved biliary obstruction.
• Presence of ascites.
• Suspected or known hepatic malignancy.
• Hypertensive patients with uncontrolled blood pressure. (SBP >150 and/or DBP>100mmHg)
• Medically unfit for sedation.
• Inability to provide informed consent.
• Cannot understand Hindi or English should be excluded since they will not be able to reply objectively to questionnaire.