Limitations of non‐ceruloplasmin‐bound copper in routine clinical practice

We read with interest the paper on Wilson's disease by Merle et al (this issue, p 115). Some aspects of the investigation profile are omitted from the paper. Although data on the prevalence of reduced ceruloplasmin levels and raised non‐ceruloplasmin‐bound copper levels are provided outside cut‐offs, data on mean (and median) levels, SD (interquartile ranges) and any skew or kurtosis are not. This would be useful as the pattern of result distributions in patients with Wilson's disease has substantial implications for the validity of diagnostic algorithms. In addition, this cohort survey clearly showed that some patients with Wilson's disease have ceruloplasmin concentrations within the reference interval (11.8%)—an important message when devising a diagnostic algorithm for Wilson's disease.

According to Roberts and Schilsky,¹ the upper end of the reference interval for non‐ceruloplasmin‐bound serum copper (NCC) is 15 μg/dl (2.4 μmol/l) and, in most patients with untreated Wilson's disease, the concentration is >25 μg/dl (3.9 μmol/l). They correctly went on to state that interpretation is difficult as the NCC depends on satisfactory copper and ceruloplasmin assays. Since then, we have shown that the upper reference interval in our population was 40 μg/dl (6.3 μmol/l)² but that the lower reference interval was −18.4 μg/dl (−2.9 μmol/l), which is clearly not possible and thus supports the assertion by Roberts and Schilsky about the need for satisfactory copper and ceruloplasmin assays. Furthermore, 40 μg/dl is considerably higher than the cut‐off of 25 μg/dl used by Merle et al to detect 86.6% of patients with Wilson's disease; accordingly fewer patients would be detected using a more appropriate NCC cut‐off. This, with the negative values found in 20% of normal patients, means that the NCC is a poor diagnostic test for the detection of Wilson's disease.