Liver Copper Estimation: Does Liver Biopsy Size Really Matter?

Abstract

Potential conflict of interest: Nothing to report. To the Editor: We read with interest the article by Yang et al. published in hepatology,1 a well‐designed prospective study about the accuracy of hepatic copper measurement for Wilson disease (WD) diagnosis. The authors concluded that the size of the liver sample may reliably reflect hepatic copper content. They recommended using the entire core of the liver biopsy specimen at >1 mg dry weight to obtain the valid hepatic copper content as the coefficient of variation decreases with larger specimens. Unfortunately, the number of liver samples in each subgroup classified by dry weight liver was quite low to use the statistical mean and standard deviation to calculate the coefficient of variation mean centered. Indeed, there was no P value demonstrated in this core content. We retrospectively conducted a similar study in our pediatric liver unit at King's College Hospital between 2005 and 2015. Children with genetically confirmed diagnosis of WD (n = 28, median age 10.61 [range 1.64‐17.26] years) and two liver diseases, autoimmune hepatitis (n = 60, median age 12.35 [range 2.04‐17.36] years) and nonalcoholic fatty liver disease (n = 59, median age 13.00 [range 7.90‐17.98] years) were included. Liver biopsy procedures were performed using Hepafix Luer Lock (Melsungen, Germany), and specimens were divided using a new scalpel for histologic study and copper measurement. Liver specimens were placed onto moistened filter paper with deionized water and then put into a sterile bottle. Samples were stored frozen (‐20°C) or sent immediately to the Trace Elememt Unit for copper measurement by flame atomic absorption spectrophotometry. The hepatic copper content was significantly higher in WD children than in the control group (Table 1). Table 1 - Liver Dry Weight and Liver Copper Content in Three Liver Diseases Disease Dry Weight Liver (mg) Patients (n) Median Dry Weight Liver (mg) Liver Copper Values Copper Content (µg/g Dry Weight) Coefficient of Variation Median Centered (%) P WD All 28 1.12 (0.15‐3.19) 495.5 (7.0‐4225.0) 186.4 0.326 <1 12 0.77 (0.15‐1.00) 621.5 (73.0‐4225.0) 184.1 1‐2 10 1.38 (1.08‐1.89) 350.23 (38.0‐2420.0) 216.3 >2 6 2.51 (2.01‐3.19) 232.50 (7.0‐1117.0) 205.7 Autoimmune liver disease All 63 0.60 (0.01‐1.88) 51.5 (1.0‐877.0)a 244.1 0.973 <1 43 0.44 (0.01‐0.99) 53.0 (1.0‐877.0) 274.0 1‐2 20 1.27 (1.02‐1.88) 35.5 (13.0‐247.0) 182.6 Nonalcoholic fatty liver disease All 59 1.09 (0.26‐5.36) 14.0 (2.0‐123.0)a 163.4 0.584 <1 27 0.52 (0.26‐0.99) 15.0 (2.0‐123.0) 157.2 1‐2 22 1.32 (1.02‐1.90) 15.0 (6.0‐117.0) 177.5 >2 10 3.23 (2.13‐5.36) 9.5 (6.0‐21.0) 51.6 Laboratory parameters are expressed as median (range).aP < 0.001 versus WD. We assessed the receiver operating characteristic curve to evaluate the diagnostic values of liver copper content. The area under the receiver operating characteristic curve was 0.919 (95% confidence interval 0.847‐0.990). According to the cutoff levels used in the scoring system for WD diagnosis,2 liver content >50 and >250 μg/g dry weight had sensitivity and specificity of 92.9% and 72.3%, 64.3% and 98.4%, respectively. The most useful single cutoff value in our study was 120 μg/g dry weight, with sensitivity and specificity of 82.1% and 91.0%, respectively. Liver concentration was <250 μg/g dry weight in eight WD children. Two of eight WD children had liver copper content <50 μg/g dry weight. Both children, aged 12 years and 4.5 years, were investigated for liver copper content because of positive family history. Livers weighing 3.19 and 1.4 g were used to measure the liver concentration, with 7 and 38 μg/g dry weight as a result, respectively. Our data show that estimation of liver copper was reliable in children with WD even when the liver biopsy sample size was <1 mg. Our method is much simpler and could avoid having to do more than one liver biopsy pass, thus making it arguably safer.