Biochemical testing for the diagnosis of Wilson's disease: A systematic review

Abstract Background Wilson's disease (WD) is a rare inherited disorder that leads to copper accumulation in the liver, brain, and other organs. WD is prevalent worldwide, with an occurrence of 1 per 30,000 live births. Currently, there is no gold standard diagnostic test for WD. The objective of this systematic review is to determine the diagnostic accuracy for WD of three biochemical tests, namely hepatic copper, 24‐hour urinary copper, and ceruloplasmin using the Leipzig criteria. Methods Adhering to PRISMA guidelines, databases including PubMed/MEDLINE, CINAHL Plus, Web of Science, and Cochrane were searched. Studies that comprised of confirmed or suspected WD along with normal populations were included with adult and pediatric group. The sensitivity, specificity, negative predictive value and positive predictive value were computed using RevMan 5.4. Results Nine studies were included. The best practice evidence for 24‐hour urinary copper test ranged from a cutoff value of 0.64–1.6 μmol/24 h (N = 268; sensitivity = 75.6%, specificity = 98.3%). Hepatic copper test was optimally cutoff based on the ROC curve analysis at 1.2 μmol/g yielding a power of 96.4% sensitivity and 95.4% specificity (N = 1,150); however, the tried and tested 4 μmol/g cutoff, with 99.4% sensitivity and 96.1% specificity, is more widely accepted. The ceruloplasmin test cutoff value was found to be ranging from 0.14 to 0.2 g/L (N = 4,281; sensitivity = 77.1%–99%, specificity = 55.9%–82.8%). Conclusion This paper provides a large‐scale analysis of current evidence pertaining to the biochemical diagnosis of WD employing the Leipzig criteria. The laboratory values are typically based on specific subgroups based on age, ethnicity, and clinical subgroups. The findings of this systematic review must be used with caution, given the over‐ or under‐estimation of the index tests.


| INTRODUC TI ON
Wilson's disease (WD) was first described in 1912 by Samuel Wilson as an autosomal recessive metabolic disorder occurring due to mutations of the ATP7B gene. 1 It is a rare inherited disorder that leads to copper accumulation in the liver, brain, and other vital organs. 2 WD is found worldwide, with an estimated prevalence of 1 per 30,000 live births across populations, 3 although the data obtained by molecular sequencing from the United Kingdom suggest higher prevalence of 1 per 7,021 live births. 4 A large proportion of patients are diagnosed between the ages of 5 and 35 years, but it may affect the older population as well. 1,2 At present, there is no gold standard diagnostic test for Wilson's disease, where only the measurements of liver copper content are being used to improve the diagnostic accuracy. 5 Owing to the nonspecific clinical features of Wilson's disease, the battery of laboratory and clinical tests for diagnosis is oftentimes delayed. 6 Ultimately, this may affect the clinical outcomes and has implications for members in the family tree when considering late or missed diagnosis earlier during the disease course. 7,8 With the Leipzig criteria, the shortcoming of no gold standard diagnostic test may be overcome by promoting the standardization of diagnosis and treatment of disease. 9 The objective of this systematic review is to determine the diagnostic accuracy, that is, sensitivity and specificity of biochemical tests, including hepatic copper, 24-hour urinary copper content, and ceruloplasmin. These laboratory markers are ideally tested in suspected patient and control groups. With inconsistent gold standard testing for the diagnosis of Wilson's disease, the Leipzig criteria are used as the standard for this investigation. The key purpose is to scale the benefits of these tests for patients under an index of suspicion or whether the test should be used at large.

| MATERIAL S AND ME THODS
In accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement 2020, 10 observational studies (retrospective/prospective cohorts and case controls) that tested the diagnostic accuracy of any or all three index tests in the context of diagnosing WD were included. Databases including the following were searched: PubMed/MEDLINE, CINAHL Plus, Web of Science, and Cochrane. The date of the last database search was October 20, 2021. There were no language restrictions. A combination of the following keywords was used: Wilson's disease, Wilson's disease, ceruloplasmin, urinary copper, hepatic copper, liver copper, and hepatolenticular degeneration.
The studies included participants with confirmed or suspected WD and also comprised of normal population and heterozygotes in cases where genetic testing was used. In case the study evaluated the index test in the normal population with the Leipzig criteria but did not include a WD comparator group, it was omitted.
The target population was the pediatric and adult population with suspected WD as assessed by the Leipzig criteria, which was essential for inclusion. Studies that did not use the Leipzig criteria and did not define WD were excluded. The three index tests were ceruloplasmin, liver copper content, and urinary copper that were evaluated for diagnosing WD. The cutoff thresholds for each of the index test are provided in Figure 1. The clinical reference standard to the diagnosis of WD is outlined in the Leipzig criteria below.
Two pairs of authors (AS-ZS and HMS-JS) independently extracted data onto a google spreadsheet, and any discrepancies were resolved by active discussion. The data were entered as au-

| RE SULTS
The PRISMA flowchart is depicted in Figure 2. In total, 13,783 studies were identified from the enlisted databases. On removing 6,436 duplicates, 7,347 studies were screened. Overall, 21 studies were retrieved and thereby assessed for eligibility. Finally, 9 studies were included in this analysis.

| Characteristics of included studies
The characteristics of the included studies are listed in Table 4.
Only 5 of the 9 studies provided data for ceruloplasmin. Overall, while 6 of the 9 studies were case controls, with one each being a cross-sectional, a prospective cohort, and a genetic cohort study, they were well-designed and had robust inclusion criteria along with a battery of index testing. Individualistic differences of index tests, patient populations, and evaluations are listed (Table 4). A proportion of the sample size presented mild hepatic WD, whereas neurological cases and asymptomatic WD presentations were also included.  Table 1). In these five studies, the optimal cutoff for the ceruloplasmin index test was determined to be between 0.14 and 0.2 g/L. The studies were well-designed, clearly defined WD, and the details of the laboratory cutoffs were conducted using receiver operating characteristic (ROC) curve analysis.

| Ceruloplasmin index test
It may be noted that ceruloplasmin levels are lower in the neonatal age group; however, the levels rise in women who are currently pregnant or taking oral contraception and those undergoing acute  Therefore, the cutoff values identified as 0.14-0.2 g/L serve as an essential method that is paramount when the same sample is run using the same analyte with a different method, but will possibly lead to different results as seen in the Merle study. 14 presents three studies that enlist the 24-hour urinary copper cutoffs of children with WD-associated CLD (Lu et al. 16 Table 2). Lu and colleagues utilized 24-hour urinary copper was measured by ICP mass spectrometry using a Chinese population, 16 whereas Nicastro/Sezer and et al. 11,12 used an atomic absorption spectroscopy in an Italian and Turkish population, respectively (Table 2). Both these studies used a clear criterion for diagnosing WD, and there were gender-and age-matched controls. Wherever differences in cutoffs were present, they may have occurred due to slight differences in methods and ethnicity.

| Hepatic copper
On considering the cutoffs for copper, four studies in total were eligible, and differences in age groups, methods, and index test analysis were presented (Table 3). Ferenci and colleagues clearly defined the WD criteria and employed a pediatric and adult population, and adequate laboratory methods were utilized to evaluate hepatic copper by using atomic absorption spectroscopy. 17 The optimal cutoff based on the ROC curve analysis was 1. ing a sensitivity of 65% and specificity of 77%. 11 On the contrary, Sezer noted that when the cutoff was decreased to a threshold of 1.5 μmol/g, the sensitivity increased to 91.4%, but the specificity decreased to 65.8%. 12 Moreover, while ceruloplasmin levels are helpful in diagnosing WD, the feasibility and cost-effectiveness of utilizing serum ceruloplasmin in presymptomatic WD may be low. 19 With these limitations in mind, we believe that the methodological weaknesses of the included studies may have resulted in overestimation of the accuracy of the index tests, despite still offering pertinent information to medical communities worldwide.

ACK N OWLED G EM ENTS
None.

CO N FLI C T O F I NTE R E S T
None.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data used and acquired for this study are available online.