Mining for a diagnosis of Wilson's disease in children: Genetics, score, and ore


  • Richard A. Rosencrantz M.D.,

    1. Department of Pediatrics; Section of Gastroenterology Hepatology, Yale University School of Medicine, New Haven, CT
    2. Transplantation and Immunology, Yale University School of Medicine, New Haven, CT
    Search for more papers by this author
  • Michael L. Schilsky M.D.

    Corresponding author
    1. Transplantation and Immunology, Yale University School of Medicine, New Haven, CT
    2. Digestive Diseases, Yale New Haven Transplantation Center, Yale University School of Medicine, New Haven, CT
    • Yale New Haven Transplantation Center, Yale University School of Medicine, 333 Cedar Street, LMP 1080, New Haven, CT 06520
    Search for more papers by this author
    • fax: 203-785-6645

  • See Article on Page 1948.

  • Potential conflict of interest: Nothing to report.

Establishing a diagnosis of Wilson's disease (WD) is often challenging in young, asymptomatic patients. The consensus on diagnostic criteria using clinical, biochemical, and genetic studies has previously been reviewed, and diagnostic algorithms have been proposed.1 In addition, a WD scoring system for the evaluation of patients was previously set forth and adopted at an international conference on WD.2 This scoring system has been subsequently validated in adult populations but not in pediatric ones.3-5


ATP7B, ATPase, Cu++ transporting, beta polypeptide; CDG, congenital disorder of glycosylation; KF, Kayser-Fleischer; PCT, penicillamine challenge test; WD, Wilson's disease.

In this issue of Hepatology, Nicastro et al.6 evaluate the conventional diagnostic criteria for WD in a pediatric population. They compare a cohort of patients with known WD (n = 40) diagnosed by liver copper concentration or by the ATP7B genotype who were clinically asymptomatic except for elevated aminotransferases (34 of 40 patients) against a control population of patients with liver disease other than WD (n = 58). The evaluated diagnostic parameters include the presence of Kayser-Fleischer (KF) rings, serum copper, ceruloplasmin, 24-hour basal urinary copper excretion, 24-hour urinary copper excretion after a penicillamine challenge test (PCT), hepatic copper content, liver histology, ATP7B genotype, and WD scores2 calculated with two urinary copper measurements with a diagnostic cutoff of >40 μg/24 hours or >100 μg/24 hours.

Let us examine these potential diagnostic variables independently and then together as a WD score. It has previously been shown that in a pediatric age group less than 10 years old, KF rings are more prevalent in symptomatic patients versus asymptomatic patients (75% and 12.5%, respectively).3 In agreement with other pediatric studies,7 in this study, KF rings were present in only 5% (2 of 40 patients), with the youngest with KF rings being 16 years old. From these observations, we can conclude that a slit lamp examination is likely not to be useful in most asymptomatic patients before puberty. However, because of the high specificity of this finding and the noninvasive nature of the testing, it should still be performed when possible.

Lowering the diagnostic cutoff for basal urinary copper from 100 to 40 or 63.5 (1 μmol) μg/24 hours has been shown to be useful in pediatric patients.1, 5, 8 In this study, the cutoff of 40 μg/24 hours was predictive of a diagnosis of WD in 82%. However, 7 of the 38 patients with urinary copper excretion had values less than 40 μg/24 hours, and they had an average age of 3.4 years; the rest were above this cutoff and had an average age of 7.7 years. For another 5 of the 38 patients, the values were between 41 and 99 μg/24 hours. This means that urinary copper is unreliable as a sole screening method in younger patients, and the lower cutoff for basal urine copper excretion for diagnostic screening for WD should be adopted.

Some have touted the usefulness of a 24-hour urinary copper measurement after a PCT.9 However, in this study, the PCT urinary copper measurement was diagnostic in only 3 of 24 patients and did not increase diagnostic accuracy over a basal 24-hour urinary copper measurement alone. In the subgroup of 4 of 38 patients who underwent a PCT with a basal urinary copper level less than the cutoff for diagnosis (<40 μg/24 hours), no patients met their set diagnostic criteria (>1575 μg/24 hours). Indeed, when we look at the controls, we find that the PCT resulted in false-positive results. This was demonstrated in one patient with nodular regenerative hyperplasia whose basal urine copper level was 24 μg/dL and who had a value of 1666 μg/24 hours after PCT. Furthermore, lowering the diagnostic criterion for PCT to 500 or 200 μg/24 hours would capture only another one or two patients, respectively. Therefore, along with the data collected by Dhawan and his colleagues,5 Nicastro et al.'s data6 imply that we should lay the PCT to rest and rely on basal urine copper excretion with lower cutoff values, as noted previously.

How did serum ceruloplasmin fare? Only 2 of 40 WD patients had a serum ceruloplasmin level that was in the normal range, and in all 8 patients under 4 years of age, a diagnostic serum ceruloplasmin level <20 mg/dL was seen. This test, therefore, remains useful. As for the controls, 10 of 58 (17%) had a low serum ceruloplasmin level and required more in-depth workup. Four of these 10 controls were diagnosed with a congenital disorder of glycosylation (CDG) and met other diagnostic features of WD (see comments later). It is known that 1 in 200 individuals may be heterozygous for an ATP7B mutation and also that up to 20% of these ATP7B heterozygotes may have reduced levels of ceruloplasmin10; however, the control group was not likely large enough to account for these individuals.

Serum copper values are low in patients with WD without acute liver failure because ceruloplasmin accounts for the majority of circulating copper.11 In this study, 10 of 13 patients with a decreased serum ceruloplasmin level also had a low serum copper level. The remaining three patients had serum copper levels on the low side within the normal range12; however, these patients also had elevated non–ceruloplasmin-bound copper concentrations with estimates ranging from 23 to 62 μg/dL; these values are often found in untreated WD patients. Among controls, only one patient with CDG was described as having a low ceruloplasmin level and a normal serum copper level with an estimated non–ceruloplasmin-bound copper concentration of 20 μg/dL. Therefore, a low-normal value of the serum copper concentration in the context of a low serum ceruloplasmin level could also provide more evidence for a diagnosis of WD.

All three of the young patients (4 years old and younger) who had hepatic copper quantification met the established diagnostic cutoff of >250 μg/g of dry weight. Interestingly, one patient had no ATP7B genotype mutation, and in another, an unknown heterozygote mutation was identified. The normal liver copper contents reported in 2 of the 30 patients who were biopsied likely represented aberrant results. Although only one of these patients had KF rings, both met other WD diagnostic criteria leading to a high suspicion for WD, such as a low ceruloplasmin level, a diagnostic basal urinary copper level, and at least one mutant ATP7B WD allele. As for the controls, only 2 of 24 patients had a hepatic copper level greater than 250 μg/g, and both had a diagnosis of CDG. Similarly, elevated hepatic copper concentrations in the pediatric age group with conditions other than WD may also be found in term infants and in patients with certain pathological conditions such as biliary atresia, primary sclerosing cholangitis, Alagille syndrome, familial cholestatic syndromes, extrahepatic biliary construction, and cirrhosis.13, 14 Furthermore, by biopsy, the liver can be accurately staged for the presence and degree of fibrosis; this is important because 90% of these asymptomatic patients had biopsy evidence of hepatic fibrosis. Although this subcohort is small, it reinforces the validity and utility of liver biopsy and copper quantification in establishing a diagnosis of WD in younger patients.

The ATP7B genotype testing found a mutation in 34 of the 36 tested patients. Two WD patients (siblings) had no known mutations, but their diagnoses were confirmed by hepatic copper quantification and their clinical response to penicillamine.

As for the WD scoring system, 28 of 30 asymptomatic WD patients were scored as “highly likely.” Only two had a diagnosis of “probable WD”; for one, this was presumably due to the aberrant liver copper quantification described previously, and for the other, the clinical data were not known. In comparison with the control group, 22 of 24 would have required more investigation, and none had a score greater than 4. Therefore, this scoring system displayed reasonable diagnostic accuracy in this young population.

It is worth mentioning that the patients of the CDG cohort had many of the diagnostic features of WD. Specifically, all four had low serum ceruloplasmin values; two had elevated hepatic copper levels and had WD scores totaling 4 points. Intriguingly, all lacked the classic CDG phenotypic spectrum of neurological and multiorgan involvement. In this case, other diagnostic measures such as liver biopsy for electron microscopy, 24-hour urine copper quantification, and ATP7B molecular studies should be employed to ensure the correct diagnosis. A lack of response to treatment for WD would be expected for CDG patients as well.

In summary, Nicastro et al.'s data6 demonstrate that the approach of the current guidelines of the American Association for the Study of Liver Diseases to the diagnosis of WD, which calls for obtaining a slit lamp examination, a serum ceruloplasmin level, and a 24-hour urine copper level (and then liver biopsy in some) is useful even in young, clinically asymptomatic children. The WD scoring system makes use of these data and helps clinicians to gauge the degree of certainty of the diagnosis. In addition, WD scores greater than 4 appear to be validated. Molecular testing for diagnosis has come of age and is perhaps the new standard for family screening; at present, it is still expensive and not always obtainable for all patients, although it is commercially available.

In the pediatric population with liver ailments, WD should be considered in patients with undefined disease, in those rare young patients with concurrent neurological problems, and in those patients with a suboptimal response to therapy directed against another presumed liver disease (especially autoimmune hepatitis). As shown by this study, we can now “mine for copper” and for mutant genes in the very young and use the WD score to be more confident in establishing a correct diagnosis of WD. However, keeping our suspicion high and considering a diagnosis of WD before the “ore” or, more specifically, the copper creates irreversible cellular damage are critical to achieving better outcomes for patients with this treatable disorder.