Article Text
Abstract
Introduction Liver biopsy (LB) can be associated with adverse outcomes including technical failures, complications and inadequate tissue for histological analysis. The indication for a chosen route of LB is variable between physicians and the impact of the chosen route of LB is often unclear.
Method Retrospective review of all LB performed at a UK transplant centre over a 12 month period focussing on indication and route of LB, complications, histological adequacy and variability in operator specialty and experience.
Results Over the 12 month period 570 LB were performed, of which 330 were transjugular liver biopsies (TJLB) and 240 were percutaneous liver biopsies (PCLB). A LB was categorised as being histologically adequate if a histological diagnosis could be made, or if the sample met the histological standard in the literature (a 15 mm sample, with 6 or more complete portal tracts). Overall, 94% of LB were adequate based on the histological standard, and 98% allowed histological diagnosis. There were no significant differences in adequacy when comparing TJLB vs. PCLB route; however, a consultant radiologist was more likely to provide an adequate sample than a radiology registrar (p = 0.02). The overall complication rate for a liver biopsy was 5.6%. This included 8 major complications (1.4%) and 24 minor complications (4.2%). Neither the route of LB (TJLB vs. PCLB) or operator (consultant radiologist vs. radiology registrar vs. hepatologist) was significantly associated with the development of a complication. The technical failure rate of a TJLB was 4%; these included failure to puncture internal jugular vein, failure to cannulate the hepatic vein, or an unstable sheath position. There were no technical failures in the PCLB group. Screening time for a TJLB was significantly longer if the procedure was done by a radiology registrar when compared to both a hepatologist (p = 0.01) or consultant radiologist (p = 0.003). Apart from the standard indications for a TJLB we found a patient was significantly more likely to undergo a TJLB if they were post-transplant compared to pre-transplant (p = 0.001) and if the biopsy was deemed an emergency compared to an elective procedure (p = 0.001). In our cohort we identified 130 patients who had a TJLB with no indication for this procedure.
Conclusion We have demonstarted that TJLB and PCLB have no significant differences in complication rates or adequacy of histological sample. Technical failure is more common in TJLB compared to PCLB. A TJLB is our preferred method in emergency cases and post-transplant patients. Further information regarding cost and patient experience is needed to identify if a change in practice is indicated.
Disclosure of interest None Declared.