Introduction An adequate liver biopsy should contain ≥10 portal tracts. We asked: what length of fresh tissue would guarantee enough microscopic tissue for an adequate liver biopsy after allowing for tissue shrinkage during fixation and processing?

Methods Data for liver biopsies taken between 2004 and 2008 were analysed for the following: macroscopic length – length of tissue before processing, microscopic length on pathology slides, number of portal tracts in each biopsy and diagnosis. We also conducted a prospective pilot study to correlate length of fresh tissue at the time of biopsy with the length of formalin-preserved tissue at the time of arrival in the lab.

Results There were 412 ultrasound-guided liver biopsies taken with 18 Fr needles in 195 females (F) and 217 males (M). Portal tracts could not be counted in the following: 98 cases of malignancy, 42 cases of cirrhosis and 14 insufficient biopsies. Of the remaining 258 biopsies, statistical analysis was performed on 192 (75%) samples where all of macroscopic length, microscopic length and number of portal tracts were recorded. The diagnoses were: 119 viral hepatitis (F=45, M=74 mean 37 years) 35 fatty liver disease (F=14, M=21 mean 49 years) 10 haemochromatosis (F=2, M=8 mean 59 years) 7 autoimmune hepatitis (F=6, M=1 mean 50 years) 6 normal liver (F=3, M=3, mean 43 years) 4 cholestatic liver disease (F=2, M=2 mean 54 years) 4 drug induced liver injury (F=3, M=1 mean 39 years) 7 others (F=5, M=2 mean 51 years). Tissue shrank consistently (r=0.837, p<0.01) from a mean length of 23.8 mm to 21.0 mm after processing. Using linear regression analysis, macroscopic length was found to be highly predictive of microscopic length (variance 95.1%) and the number of portal tracts (variance 85.2%). χ² analysis showed that age and sex were non-significant parameters for an adequate liver biopsy. We calculated that 17.55 mm of macroscopic tissue equating to 15.89 mm of microscopic tissue (p<0.01) would guarantee an adequate liver biopsy with at least 10 portal tracts in any diagnosis excluding cirrhosis and malignancy. Similarly, 30 mm of fresh tissue would guarantee ≥15 portal tracts in >99% of cases. Our pilot study of 10 consecutive biopsies showed no tissue shrinkage during transport in formalin to the lab.

Conclusion The length of fresh tissue from a liver biopsy obtained using an 18 Fr needle directly correlates to the number of portal tracts in all diagnoses excluding cirrhosis and malignancy. To ensure that patients have an adequate diagnosis and avoid a repeat biopsy we suggest that operators ensure at least 18 mm of fresh liver tissue at a cut-off of 10 portal tracts and 30 mm of tissue at a cut-off of ≥15 portal tracts. We advise operators to actively measure the length of tissue at the time of the procedure.