Biochemical Markers for the Diagnosis and Monitoring of Wilson Disease

Abstract

Wilson disease (WD) is an autosomal recessively-inherited disorder of copper metabolism and characterised by a pathological accumulation of copper. The ATP7B gene encodes for a transmembrane copper transporter essential for biliary copper excretion. Depending on time of diagnosis, severity of disease can vary widely. Almost all patients show evidence of progressive liver disease. Neurological impairments or psychiatric symptoms are common in WD patients not diagnosed during adolescence. WD is a treatable disorder, and early treatment can prevent the development of symptoms in patients diagnosed while still asymptomatic. This is why the early diagnosis of WD is crucial. The diagnosis is based on clinical symptoms, abnormal measures of copper metabolism and DNA analysis. Available treatment includes chelators and zinc salts which increase copper excretion and reduce copper uptake. In severe cases, liver transplantation is indicated and accomplishes a phenotypic correction of the hepatic gene defect. Recently, clinical development of the new copper modulating agent tetrathiomolybdate has started and direct genetic therapies are being tested in animal models. The following review focuses especially on biochemical markers and how they can be utilised in diagnosis and drug monitoring.

Introduction and History

Wilson disease (WD) is a rare autosomal recessively-inherited disorder of copper metabolism caused by mutations in the ATP7B gene which codes for a transmembrane copper transporting ATPase. The genetics of WD is complex with more than 450 disease-causing mutations identified. Compound homozygous mutations (the presence of two several mutant alleles) in the ATP7B gene are common. Under physiological conditions the role of ATP7B is two-fold, mediating the excretion of copper into the bile and copper transport into the trans-Golgi network (TGN) for copper loading of cupro-enzymes. In patients, the impaired copper excretion leads to toxic levels of copper accumulation primarily in the liver and consecutively in other organs, particularly the brain. The consequences are a wide range of symptoms, e.g. progressive liver disease and liver failure, neurological extrapyramidal disease as well as psychiatric symptoms.¹

Indeed, it was the autosomal recessive inheritance of liver cirrhosis in combination with severe progressive neurological disease that Samuel Alexander Kinnier Wilson observed in his first description in 1912 as ‘hepatolenticular degeneration’ in a total of 12 patients. The first description of pathognomonic corneal rings – copper deposition in the corneal membrane – was made in 1902 by Kayser and Fleischer. The discovery of copper deposition as the underlying aetiology followed decades later. In 1956, D-penicillamine was introduced as the first oral copper chelator by J Walshe, a milestone in WD treatment. Alternative anti-copper drugs followed with zinc salts in 1961 and trientine 1982. The first ever liver transplantation was performed by Thomas E Starzl in 1963 in Denver, US.^2,3 He was also the surgeon of the first liver transplantation in WD.⁴ The discovery that mutations in the ATP7B gene on chromosome 13q were the underlying genetic defect followed in 1993 (Figure 1).^5,6

Figure 1.

Milestones in the history of Wilson disease.

Epidemiology and Genetics

WD is a rare disorder with an estimated prevalence of symptomatic disease of 1:~30,000 and a heterozygous ATP7B mutation carrier frequency of 1:90 (almost 1% of a population). These numbers were partially based on assumptions and have occasionally been questioned. Mass screening studies in East Asia suggested an even higher prevalence (1:1500–1:3000) based on caeruloplasmin (Cp) level measurements.^7,8 A genetic prevalence study of WD in the UK revealed in 1000 healthy-born neonates, a number of 1:40 for heterozygous ATP7B mutation carriers (including mutations of unclear significance) and thus a worst-case scenario prevalence of 1:7000 for WD in the UK population.⁹ Prevalence numbers in WD also differ between certain ethnic groups and geographic regions with particularly high prevalence in isolated populations like Crete, Sardinia and Costa Rica due to genetic founder effects.¹⁰ The incidence of WD in Costa Rica is the highest in the world with 4.9–6:100,000 inhabitants.^11,12 More recent studies on prevalence of WD marked a discrepancy between estimated prevalence data in WD and new findings. This might be due a reduced penetrance of some ATP7B mutations. Nevertheless, there is also concern that WD is still an underdiagnosed disease. Late diagnosis is still the most common cause of death and severe consequences in WD.¹³ With increasing knowledge of WD and increasing availability of genetic testing, numbers of patients with WD appear to be rising.

WD is an autosomal recessive disease caused by mutations affecting the ATP7B gene locus on the short arm of chromosome 13. It contains 20 introns and 21 exons. The spectra of mutations in WD include missense or nonsense mutations, small deletions/insertions in the coding region or splice junction mutations. Whole exon deletions, promotor region mutations or monogenic disomy are less common.^1,9,14 More than 700 pathogenic variants of the ATP7B gene have been identified and compound heterozygosity in affected patients is frequent.¹⁵ The majority of variants affect the transmembrane region of the protein. In pre-symptomatic patients or patients with mainly hepatic manifestations, mutations are often found to be located in the M- and N-domains of the ATPase protein or transporter.¹⁶ A model of ATP7B is shown in Figure 2.

Figure 2.

A schematic representation of human ATP7B protein. ATP7B belongs to class 1B of the highly conserved P-type ATPase superfamily responsible for the transport of copper and other heavy metals across cellular membranes (M). The protein contains 1465 amino acids including eight transmembrane fractions (Tm). The N-terminal metal-binding domain (NH2) is composed of six copper binding sites (Cu), which play a central role in accepting copper from copper chaperone ATOX1 through protein-protein interactions. Unique amino acid motifs are present at the core structure of each domain, such as TGDN and SEHPL. The SEHPL domain functions as a phosphorylation domain, the TGDN as an ATP binding site. (Adapted from Fanni D et al. Eur J Histochem 2005;49:371–8.)

NH2, N (nucleotide)-terminal metal-binding domain; Cu, copper; Tm, transmembrane domain; M, phospholipidic bilayer of the membrane; PD, Phosphorylation domain.

The most common cause of WD in central, eastern and northern Europe is a histidine to glutamate substitution at position 1069 (p.H1069Q) in a histidine-containing sequence of the N-domain of ATP7B protein. About 50–80% of WD patients carry at least one allele with the H1069Q mutation.¹⁷ Animal models demonstrated a reduced ATP-mediated phosphorylation but no significant misfolding of the protein. However, this suggests that phosphorylation plays an important role for localisation and further catalytic effects of ATP7B in copper metabolism.¹⁸ More recent clinical studies suggested that homozygous mutation for H1069Q leads to later onset but to more frequent neurological manifestation in comparison to H1069Q compound heterozygotes.^19–21 However, this was not confirmed by larger more recent studies.²²

Other common mutations in ATP7B include p.E1064A, p.R778L, p.M769V and p.G943S. The p.R778L mutation can be found more often in Southeast Asia. Several studies investigated the challenging task of correlation between genotype and phenotype, yet without establishing genotype-specific phenotype features.²³

In vitro experiments showed that copper transporting activity as well as function depend on the different mutant types.²⁴ Missense mutations, for example, led to higher levels of Cp than frameshift or nonsense mutations.²⁵ Truncating mutations of ATP7B have been associated with early onset of disease and occurrence of acute liver failure.^26,27 Genotype and phenotype correlation still lack evidence and larger studies to be conclusive in WD. This is also due to significant inter-individual differences regarding cellular susceptibility of copper toxicity and endogenous copper chelation capacity.²⁸

Copper Homeostasis and Physiology

Copper is an essential trace mineral in human physiology. It acts as a cofactor of cupro-enzymes that are involved in radical defense (superoxide dismutases (SOD), SOD 1 and 3, catalase), respiration (cytochrome c oxidase), pigmentation (tyrosinase), catecholamine synthesis and clearance (dopamine-β-monooxygenase), activation of neuroendocrine peptides (peptidyl-α-monooxygenase), collagen synthesis (lysyloxidase) and many other cellular processes.¹ In the blood, copper is bound to Cp, the major copper-binding protein which can carry six copper atoms per protein (holocaeruloplasmin (holoCp)). Copper loading of apocaeruloplasmin (apoCp) is ATP7B protein dependent. Albumin and other proteins containing SH-groups also function as copper-binding proteins.²⁹ Normal dietary consumption and absorption of copper is not usually controlled by food uptake but regulated by biliary excretion of copper. Copper is present in shellfish, organ meat, beef, nuts and seeds, chocolate (especially dark chocolate), legumes and potato.^1,30 Dietary copper restriction has long been considered an important aspect of treatment WD. However, evidence supporting this approach is limited. WD patients are generally advised to avoid foods high in copper during the initial treatment phase.³¹

Copper is absorbed in the intestine by the enterocytes and transported protein-bound by albumin or transcupreine to the sinusoidal plasma membrane in liver (Figure 3). Copper atoms are transported into the hepatocyte via copper transporter 1 (CTR1) into the cytosol.^32–35 Copper chaperones such as copper chaperones for superoxide dismutase (CCS) and antioxidant protein 1 (Atox1; donor for ATP7B) shuttle copper to specific intracellular targets.³⁴ These targets are e.g. SOD1, ATP7A and ATP7B, respectively. Glutathione mediates other intracellular transfer, including incorporation into metallothionein.

Figure 3.

Copper homeostasis in the hepatocyte. Role of Wilson ATPase (ATP7B) in the hepatocellular disposition of copper: a hepatocyte is shown, with one side connected with the bile canaliculus, the other connected with the sinusoidal membrane. Starting at the left side of the diagram, copper (small green dots) is taken up by CTR1 (grey square), picked up and carried by ATOX1 to the Wilson ATPase (ATP7B; red square) in the trans-Golgi network (TGN). The Wilson ATPase either directs copper to production of caeruloplasmin (Cp; blue round shaped) or to excretion into bile. When intracellular copper concentrations are low or normal, the Wilson ATPase participates in the production of holocaeruloplasmin (HoloCp-Copper) in the Golgi apparatus; holocaeruloplasmin is then secreted into the blood. When intracellular copper concentrations are elevated, the Wilson ATPase expedites biliary excretion of copper.

In the normal hepatocyte under steady state conditions, ATP7B is predominantly located in the TGN, into which copper is transported for incorporation into apoCp to assemble enzymatically-active holoCp. The metal-binding domain in the nucleotide-binding domain of the Cp protein has six copper-binding sites, from which copper is accepted from Atox1 via protein-protein interaction. If levels of intracellular copper are high, ATP7B translocates its cellular compartment to biliary canalicular-associated structures and facilitates the process of copper excretion into the biliary tract.³⁶

Copper Toxicity

Copper toxicity has been the subject of intensive research over several decades and evidence supports the involvement of excess copper-induced neurotoxicity in WD and other neurodegenerative disorders (especially Alzheimer’s disease and Parkinson’s disease).³⁷ ATP7B is primarily expressed in the liver and its mutations cause a progressive hepatic copper burden that results in excess copper accumulation. Once the endogenous copper storage capacity of the liver is exhausted, toxic effects of reactive non-protein-bound copper are evident. Recent studies identified mitochondria as a potential early target of copper toxicity.³⁸ Copper increasingly affects the mitochondrial structure and mitochondrial proteins that are essentially involved in bioenergetic functions. This leads to a concomitant decrease in their ATP production capacity, and copper accumulates in mitochondrial destruction in hepatocytes.^39–42 Copper-induced mitochondrial dysfunction and destruction provoke hepatic failure distress and apoptosis, which could be demonstrated in ATP7B −/− knockout rats.^43,44 However, in contrast to the liver phenotype of WD, copper toxicity in the brain is much less-well understood.

Besides mitochondrial damage, formation of reactive oxygen species or direct interaction of copper with lipid biosynthesis can additionally result in impaired energy metabolism for the cell and downregulation of genes involved in cholesterol biosynthesis. These mechanisms contribute to hepatic steatosis, followed by chronic liver injury, inflammation, fibrosis and apoptosis, triggered by release of cytochrome c from damaged mitochondria.^45,46

In comparison to the clinically significant effects on the liver and brain, abnormalities in other organs are rare.^47–50 The metal accumulation in the brain is associated with neurological impairments as WD patients present with tremor, Parkinsonism, gait disturbances, dysarthria or seizures.⁴⁷ Our current understanding of this pathomechanism supports the theory that brain effects in WD are secondary to liver effects, caused by copper overload circulating under conditions of exhausted endogenous hepatic copper storage capacity or released from destroyed hepatocytes. Such copper spillover into the bloodstream then secondarily might affect the brain.^51,52

The toxic effect of ‘free’ copper in brain is considered to be first buffered by astrocytes, which increase in number (astrogliosis) and undergo cellular swelling. Metallothionein synthesis is upregulated as a response of brain storage for copper.^53,54 Different brain regions show distinct patterns regarding their sensitivity for copper-induced cell damage. Long-term copper exposure leads finally to damaged astrocytes, destroys blood-brain barrier and afflicts other neurons (e.g. oligodendrocytes) and other brain regions. Consistently, some studies have demonstrated an improvement of neurological symptoms after liver transplantation, indicative of restoration of a physiological copper excretion via the new liver transplant which also reduces copper concentration in the blood stream and the brain.^55–57

Clinical Symptoms and Diagnosis

WD may become symptomatic at any age. The majority of patients are diagnosed at 5–35 y, but late onset of the disease is reported.^58,59 In concordance with the pathogenesis of WD, copper accumulates in different organs and therefore causes a wide range of clinical symptoms. Figure 4 gives a short overview of potential symptoms in different organs. In the vast majority of cases, only hepatic and neurological/neuropsychiatric manifestations are clinically relevant. A diagnostic score based on all available tests was proposed by the Working Party at the 8th International Meeting on Wilson disease in Leipzig, 2001 (Table 1), bringing together biochemical parameters and clinical symptoms.⁶⁰

Figure 4.

Clinical manifestations in the pathogenesis of Wilson disease.

Table 1.

Diagnostic scoring system developed at the 8^th International Meeting on Wilson disease, Leipzig, 2001 (ref. ¹⁵⁵).

Typical clinical symptoms and signs		Other tests
Kayser-Fleischer rings		Liver copper (in the absence of cholestasis)
Present	2	>250 μg (>4 μmol)/g dry weight	2
Absent	0	50–249 μg (0.8–4 μmol)/g	1
Neurologic symptoms*		Normal: <50 μg (<0.8 μmol)/g	−1
Severe	2	Rhodanine-pos. granules†	1
Mild	1	Urinary copper (in the absence of acute hepatitis)
Absent	0	Normal	0
Serum caeruloplasmin		1–2 × ULN	1
Normal (>0.2 g/L)	0	>2 × ULN	2
0.1–0.2 g/L	1	Normal but >5 × ULN after D-penicillamine	2
<0.1 g/L	2	Mutation analysis
Coombs-negative haemolytic anaemia		On both chromosomes detected	4
Present	1	On 1 chromosome detected	1
Absent	0	No mutations detected	0
TOTAL SCORE	Evaluation
4 or more	Diagnosis established
3	Diagnosis possible, more tests needed
2 or less	Diagnosis very unlikely

^*Or typical abnormalities at brain magnetic resonance imaging;

^†If no quantitative liver copper available; ULN, upper limit of normal.

Hepatic Manifestations

WD may present as a cholestatic liver disease or, rarely, as haemolytic anaemia.^1,61 Other forms of liver disease have been reported but they are uncommon. Presenting symptoms of liver disease are independent of the presence of neurologic symptoms and can be highly variable, ranging from asymptomatic or mild (with only biochemical abnormalities) to overt cirrhosis with all its complications. In rare cases (~5%), WD may also present initially as acute hepatic failure sometimes associated with Coombs-negative haemolytic anaemia and acute renal failure, affecting more females than males.⁶² Such fulminant WD accounts for 6–12% of all patients with acute liver failure who are referred for emergency transplantation.^63,64 A clinical score is available to guide the decision on the necessity of an emergency liver transplantation.^64,65 A score >11 points is always fatal without liver transplantation (Table 2). Patients presenting with neurologic symptoms fare better with respect to life expectancy, especially if liver disease is limited. In patients undergoing orthotopic liver transplantation, survival may be slightly reduced early on, but appears normal compared to the general transplant population thereafter.⁶⁶ For chronic liver disease, presentation may be indistinguishable from other forms of chronic active hepatitis, with symptoms including jaundice, malaise and vague abdominal complaints. It is still unknown why some patients present with liver disease (acute or chronic) whereas others evolve neurological manifestations.

Table 2.

Wilson disease prognostic index. New Wilson Index modified from the Nazer scoring system (ref. 196) by Dhawan et al. (ref. ¹⁶⁸). A score of ≥11 points is associated with high probability of death without liver transplantation and is an indication for liver transplantation.

	0 points	1 point	2 points	3 points	4 points
Serum bilirubin (μmol/L)	0–100	101–150	151–200	201–300	>300
Aspartate aminotransferase (U/L)	0–100	101–150	151–300	301–400	>400
International normalised ratio	0–1.29	1.3–1.6	1.7–1.9	2.0–2.4	>2.4
White blood cell count (109/L)	0–6.7	6.8–8.3	8.4–10.3	10.4–15.3	>15.3
Albumin (g/L)	>45	34–44	25–33	21–24	<21

In untreated patients, WD presents as a chronic liver disease with the known complications: portal hypertension and its consequences (bleeding of oesophageal varices, hepatosplenomegaly, etc.), coagulopathy, hepatic encephalopathy, jaundice and ascites. Laboratory results reveal low serum albumin, low cholinesterase and low platelets.

Like in all other hepatic diseases in which cirrhosis is a common complication, the Model for End-Stage Liver Disease and Child-Pugh-Score is used to reflect the severity of cirrhosis and the potential need for transplantation.^67,68 Imaging methods such as ultrasound can also easily reflect status of cirrhosis. If needed, either CT or MRI can be performed. Common findings are a fatty infiltration, contour irregularities and atrophy of the right liver lobe. Measurements for liver stiffness (e.g. Fibroscan™ elastrography) are still poorly examined in small cohorts and its importance remains unclear.⁶⁹ Other prognostic scores such as APRI (AST to Platelet Ratio Index) and the PGA (prothrombin time, GGT, apolipoprotein A1) score have also been used to predict significant cirrhosis.^70–72 However, there is still need for accuracy and validation in the WD population.

Due to development of cirrhosis, patients with WD are at increased risk for hepatobiliary malignancies, but this is generally lower than in patients with nonalcoholic steatohepatitis (NASH) or viral hepatitis. Hepatocellular carcinoma or intrahepatic cholangiocellular carcinoma occur in about 1% of patients.^73,74 Still, the best strategy for screening is not clear and a screening recommendation is not yet established. Histological analysis through surgical resection or biopsy should be mandatory when a suspect liver lesion is detected.⁷⁴ The influence of copper depletion from WD-specific medical treatment on tumour activity remains to be elucidated.

Neurological Manifestations

Generally occurring later than the first hepatic biochemical abnormalities, neurological symptoms are often the first clinical symptoms leading to the diagnosis of WD. First neurological symptoms occur in 18–68% of patients with mean age at onset of symptoms 20–30 y with a wide range (6–72 y with presentation of new neurological symptoms).^75,76 The symptoms include especially movement disorders with a spectrum of involuntary movements.^1,76 Common neurological features of WD are the following: predominance of tremor, dystonia, Parkinsonism, dysarthria, gait and posture disturbances, drooling and dysphagia.^44,77

Tremor is very characteristic of WD, experienced by up to 55% WD patients at diagnosis (tremor differentiation: resting, postural or kinetic). It might start initially as unilateral or bilateral tremor and often with distal upper extremities involvement. Asterixis or flapping tremor is a negative myoclonus affecting the hands. This can be observed in WD patients with liver failure as a symptom of hepatic encephalopathy and needs to be distinguished from the usual Wilson tremor.^44,77

Dystonia is also often reported as an initial symptom in 11–65% of WD patients. Characteristic are abnormal facial expression or risus sardonicus, which presents as a fixed smile of the risorius muscle. Parkinsonism occurs frequently in WD patients (19–62%), presenting as rigidity, hypomimia, symmetric bradykinesia, gait and posture disturbances as well as dysarthria or dysphagia.⁷⁸ Ataxia as a symptom of cerebellar dysfunction occurs in 30% of WD patients, usually not as a solitary symptom but in combination with other movement disorders.⁷⁸ Chorea, characterised by rapid, irregular involuntary movements of the face, head, trunk or extremities, occurs rarely in neurologic patients with WD (6–16%). Dysarthria appears to be the most frequent neurological symptom, being reported in up to 97% of WD neurologic patients. In some cases, it is so severe and persistent that verbal communication is impossible. There is no specific WD dysarthria, but manifestations can be divided based on the predominance of clinical symptoms: cerebellar, extrapyramidal (dystonic, Parkinsonian) and mixed (unclassified due to symptoms overlapping).^44,77–79 Dysphagia is reported in up to 50% of neurological WD patients.^80–82 Dysphagia varies from mild to severe including bronchoaspiration, pneumonia, malnutrition and weight loss. Other neurological syndromes like epilepsy may occur.^83,84 Seizures are usually generalised. Due to the wide spectrum of neurological symptoms in WD, clinical scales like the Unified Wilson’s Disease Rating Scale (UWDRS) or Global Assessment Scale for WD assessing neurological deficits and functional impairment were established.^76,79,85

Brain MRI is currently the most important neuroradiological imaging for diagnosis and may be helpful for treatment monitoring, especially in patients with known brain copper deposits. Abnormalities include symmetric hyperintense changes visualised in T2-weighted images located in basal ganglia (mainly putamen and caudate nuclei), thalami, midbrain and pons.^86–89 The most spectacular WD changes are described as the ‘face of the giant panda’ in the midbrain.

Psychiatric symptoms occur abundantly in WD, secondary to the somatic and brain pathology. Co-morbidity of psychiatric illness are, for example, major depressive disorder and bipolar disorder; 20–60% of WD patients develop depression,⁹⁰ often presenting with a high rate of suicidal attempts (4–16%).^78,91,92 Bipolar disorder has been reported in 14–18% of WD patients.⁹³ Psychosis occurs no more frequently in WD than in the general population. The most common behavioural and personality disorders are irritability, aggression and antisocial behaviour.⁹³ Frequently, patients with initial psychiatric manifestation suffer from a delayed diagnosis of WD.

Biochemical Parameters

An overview of the diagnostic tests for WD is given in Table 3.

Table 3.

Routine tests for the diagnosis of Wilson disease.

Test	Typical finding	Diagnosis	Monitoring
Serum caeruloplasmin	Decreased	Part of the Leipzig score 2001 (ref.155)	Not helpful
24 h urinary copper	Adults: >1.6 μmol (100 μg)/24 h Children: >0.64 μmol (40 μg)/24 h	Part of the Leipzig score 2001 (ref.155)	Increased when measured on taking chelating agents
Non-caeruloplasmin-bound copper (NCC)	>3.9 μmol/L (250 μg/L) (ref. 51)	Not recommended for diagnosis of WD among guidelines (normal population: <150 μg/L; <2.4 μmol/L)	Additive marker of copper control in combination with 24 h urine copper
Exchangeable copper (CuEXC)	Threshold for extra-hepatic diagnosis >2.08 μmol/L (ref. 117)	Experimental	Experimental
Hepatic copper	4 μmol (250 μg)/g dry weight	Part of the Leipzig score 2001 (ref.155)	Not helpful as expected; changes under treatment are not established
Kayser-Fleischer rings by slit-lamp examination	Present	Part of the Leipzig score 2001 (ref.155)	Could become negative under chelating therapy

Biomarkers in Established Clinical Practice

Caeruloplasmin

Cp is the main copper-carrying protein in blood, binding 70–95% of circulating copper. It is synthesised in hepatocytes as inactive, non-copper bound apoCp.⁹⁴ The active variant of Cp binds six to eight copper atoms and activates the protein, then named holoCp. Copper loading of apoCp is ATP7B dependent. Non-copper loaded apoCp is rapidly degraded. Thus, low holoCp (and consequently low copper) is the most prominent biochemical finding in serum in WD. Besides its involvement in copper transport, holoCp acts as a ferro-oxidase and is functional in iron metabolism, antioxidant defense, coagulation and angiogenesis.^94–96 Cp levels can be measured by radioimmunoassays (antibody-dependent), immunodiffusion, nephelometry (quantitative immunochemical) or by enzymatic methods. Immunological methods do not distinguish between holoCp and apoCp. Therefore the accurate concentration of active Cp (holoCp) tends to be overestimated by the non-enzymatic measurements.^97–99 The normal concentration of Cp measured by enzymatic assay varies between laboratories (with a lower limit between 0.15 and 0.2 g/L).⁷⁴

Physiological levels are low in early infancy (<6 months-of-age) and reach a peak in early childhood (~0.3–0.5 g/L).^51,100 In adulthood, they again settle to a lower range in which women tend to have higher reference intervals than men because Cp is increased by oestrogen, contraception and pregnancy.^101,102 Cp is an acute-phase protein which responds in cases of infection or inflammation with increased plasma levels, and also depends on seasonal changes (lowest concentrations during winter).^63,103–105 A measured concentration below 0.2 g/L is consistent with the diagnosis of WD but does not establish the diagnosis. A threshold of <0.14 g/L is even more strongly associated with WD.^60,106,107 However, differential diagnosis of WD may not be based on low Cp alone as 20% of healthy heterozygous subjects have Cp levels below the lower reference limit, and also in WD patients, normal or near normal Cp concentrations were reported.^108–110 Normal or even elevated concentrations of Cp may occur in WD patients with major hepatic involvement and corresponding hepatic inflammation.^76,111,112 Low concentrations of Cp may occur in patients with severe malabsorption or malnutrition, fulminant hepatitis or in cases of acaeruloplasminaemia, which is very rare.^109,112,113 All these limitations render Cp measurement unreliable as an isolated biomarker for copper excess in WD.

Serum Copper and Non-Caeruloplasmin Bound Copper

Reference intervals for total copper in serum, including copper bound to Cp, are about 14–21 μmol/L. Copper determinations can be made by either colorimetric or atomic absorption spectrometry.¹¹⁴ Counterintuitive to copper overload in WD, serum copper is decreased in WD with typical levels of serum copper <14 μmol/L.⁹⁸ Several clinical conditions can influence copper levels in blood, e.g. damage to hepatocytes during acute liver failure or severe liver injury, independent of Cp levels. In these cases, serum copper is markedly elevated due to the sudden release of copper from liver tissue.

In general, copper bound to the copper transporter Cp is considered non-reactive and thus non-toxic. So, in a hypothetical scenario, copper in serum can be assumed to be present in two different pools: Cp-bound copper or non-caeruloplasmin bound copper (NCC). In WD, serum NCC has been proposed as a tool to assess the de-coppering state of patients under treatment.^115,116 It has been proposed as an estimate of the free (toxic) portion of total serum copper. NCC is elevated above 250 μg/L in most untreated patients (reference value 150 μg/L).^94,115,116

The amount of copper associated with Cp is approximately 3.15 μg of copper/mg of Cp. Thus, for mass units, the NCC is the difference between the serum copper concentration in μg/dL and three times the serum caeruloplasmin concentration in mg/dL. For SI units, both serum copper and Cp should be expressed as ‘per litre’; the conversion factor (3) is unchanged, and the reference value is 2.4 μmol/L).^51,76 The following formula can be used to calculate NCC in μmol/L:

$$\text{NCC}=\text{total\hspace{0.17em}serum\hspace{0.17em}copper\hspace{0.17em}}(\mathrm{\mu }\text{mol}/\text{L})-0.049\times \text{Cp\hspace{0.17em}}(\text{mg}/\text{L})$$

The usefulness of this calculation in WD has been the subject of intensive discussions in the field. The value of Cp determined by immunological methods has been questioned due to the falsely high values, as these tests do not discriminate between holoCp and apoCp.^115,117 When using immunological tests, negative values for NCC might be found in up to 10–25% of WD patients.^118–120 Thus, current European Association for the Study of the Liver (EASL) guidelines do not recommend NCC for diagnosis of WD.⁷⁶

Other methods, such as direct measurements of Cp via Cp-oxidase activity, appear to be more promising. Nevertheless, they are still rarely studied and not part of common routine testing.¹²¹

Urinary Excretion of Copper

In untreated WD patients, the 24 h urinary excretion of copper is elevated and corresponds to the NCC fraction in serum.¹²¹ Thus urinary copper has an established role in diagnosis. Correct collection for a comprehensive result is an accurate 24 h collection (exact volume and total creatinine excretion/24 h) as well as an avoidance of copper contamination of the device for urine collection. Typical ranges for a 24 h urinary collection for copper in WD patients depend on the laboratories and appear in literature to be about >1.6 μmol/24 h (>100 μg/24 h).^{76,94,122,123}

Asymptomatic children with WD or heterozygotes might have a basal copper excretion below the threshold of 1.6 μmol/24 h.^124–126 Some study groups suggest even a lower threshold for native urinary copper excretion (without D-penicillamine) to 0.64 μmol/24 h (40 μg/24 h). This would eliminate the need for D-penicillamine stimulation and increases sensitivity of the normal 24 h urinary copper excretion.^65,127

Other liver diseases can cause false positive 24 h urine copper levels. In autoimmune hepatitis, cholestasis as well as chronic or acute liver failure, urinary copper excretion can be increased without WD.^123,128,129 Increased urinary copper excretion can be understood as an attempt to compensate for the impaired biliary excretion of copper in these cases. In contrast to serum NCC, the urinary copper excretion in treated patients is likewise dependent on chelating agents (increased urinary copper excretion under medication with chelating agents). Therefore its role in treatment monitoring is limited.¹³⁰

Chelating agents, especially D-penicillamine, enhance urinary copper excretion. The amount of excreted copper shows a correlation to the D-penicillamine dose and the available amount of NCC that is elevated in WD.

In children, a D-penicillamine challenge test has been standardised for measuring the 24 h urinary copper excretion. During the test, 500 mg D-penicillamine is administered at the beginning and then after 12 h of collecting urine. This can increase the amount of excreted copper in paediatric WD patients and can distinguish WD from other chronic liver diseases. Nevertheless, sensitivity in comparison with other liver diseases is only 12.5%. Reference values for paediatric WD patients undergoing the D-penicillamine challenge test are >25 μmol/24 h (>1600 μg/24 h). The same challenge test has been used in adults, but dosing as well as timing for administration remains unclear and has not been uniformly applied.^123,131,132

Liver Biopsy with Quantitative Copper Determination

If determination of Cp, serum copper, NCC or 24 h urinary copper excretion and clinical findings do not allow the final diagnosis, a liver biopsy can add to the diagnostic yield.¹³³ The liver biopsy procedure is an invasive diagnostic test that is not required for the diagnosis of WD in all cases.

Regarding conventional histopathology, early stages of WD resemble non-alcoholic fatty liver disease (NAFLD) or NASH, including mild steatosis (micro- and macronodular) and focal hepatocellular necrosis.¹³⁴ Histopathological features of autoimmune hepatitis can also be found in WD. In progressive disease, biopsy reveals parenchymal damage with fibrosis and subsequently a development of cirrhosis.¹³⁵

Copper detection methods in hepatocytes are highly variable depending on the stage of disease. In early stages, copper is mainly in the cytoplasm bound to methallothionein and therefore hardly detectable.¹³⁶ The amount of copper varies from pre-cirrhotic stages to cirrhosis and even within the nodules in cirrhosis. Lysosomal copper is stained by methods such as rhodamine or orcein stain. These methods reveal focal copper stores in <10% of cases. More detailed analysis of hepatocytes presents mitochondrial abnormalities in cases of WD.¹³⁷

In the light of these, in nearly all cases, non-specific reports of histopathology and copper staining, the standard assessment of liver specimens in WD is quantitative copper determination of ideally 10–20 mg dry needle biopsy.¹³³ In healthy subjects, normal copper concentrations in the liver are 15–55 μg/g;^138,139 in WD, copper levels ≥250 μg/g (≥4 μmol/g) are common.⁵¹

In 2005, Ferenci et al. examined 114 liver biopsies in WD and compared them to 219 patients with cholestatic liver disease. By lowering the threshold from 250 to 75 μg/g, sensitivity increased (83.3% vs 96.5%) but specificity decreased (98.6% vs 95.4%).¹⁴⁰ A more recent study by Yang et al. from 2015 revealed the best cut-off of 209 μg/g copper in dry weight liver, using an entire score biopsy sample.¹⁴¹ A liver biopsy is warranted when other clinical and biochemical markers fail to establish the final diagnosis of WD. In these cases the biopsy can contribute to the diagnostic Leipzig score.¹⁴²

Biomarkers in Experimental Practice

Exchangeable Copper and Relative Exchangeable Copper

Besides Cp-bound copper, NCC and the copper used in variant enzymes, another labile copper fraction is exchangeable copper (CuEXC). This concept was discussed in 2009 by El Balkhi et al. and primarily refers to the copper loosely complexed to albumin in the circulation.¹⁴³

Based on the idea that copper in the circulation can only be reactive if it is easily exchangeable, a two-step method (ultrafiltration-determination) was carried out to determine these fractions using EDTA as a chelator of high-copper-affinity (incubation for 1 h). The ultrafiltration procedure and the instrumental determination showed good repeatability and a very low limit of detection was obtained (0.7 nmol/L). In vitro stability of both ultrafiltrable copper (CuUF) and CuEXC was studied. Plasma was ultrafiltered in 44 presumably healthy subjects to determine CuUF and CuEXC and to set reference intervals. The method was applied to a few patients, showing good correlation between both parameters and the clinical and biological features of the patients.¹⁴³ The main advantage of CuEXC is that it is not dependent on Cp. Relative exchangeable copper (REC) is a calculated ratio of CuEXC/total serum copper, reflecting the toxic free fraction of copper in the circulation.

Limited clinical evidence suggests that the severity of WD might be associated with toxic free copper levels.^144,145 A recent study by Poujois et al. evaluated the role of CuEXC in 48 newly-diagnosed WD patients for monitoring purposes.¹¹⁷ These patients were prospectively evaluated using hepatic, neurological, ophthalmological and brain MRI scores. Three phenotypic presentations were distinguished: pre-symptomatic, hepatic and extra-hepatic. CuEXC was determined in addition to standard copper assays before de-coppering therapy. Correlations between biological parameters and the different scores were determined and compared in the hepatic and extra-hepatic groups. Extra-hepatic patients had significantly higher CuEXC values than those with the hepatic form (P <0.0001). The overall ability of CuEXC to separate the two forms was satisfactory (AUC 0.883) with an optimal threshold for extra-hepatic diagnosis of 2.08 μmol/L (sensitivity 85.7%; specificity 94.1%). In extra-hepatic patients, CuEXC was the only biological marker to be positively correlated with the neurologic disease burden (assessed by Unified Wilson Disease Rating Score). CuEXC determination is, in consequence, useful when diagnosing WD with a value >2.08 μmol/L, and is indicative of the severity of the extra-hepatic involvement. However, CuEXC did not indicate the severity of liver damage. CuEXC is an interesting experimental biomarker, but needs to be interpreted yet with caution, especially in WD patients with hepatic manifestation.

Treatment

Once the diagnosis of WD is made, lifelong treatment is necessary. Non-compliance and underdosage of anti-copper drugs are the main risk factors for an unfavourable clinical course.¹⁴⁶ To prevent non-compliance and improve adherence to the lifelong therapy, treatment schemes should be as simple as possible.¹⁴⁴ Studies suggest that up to 45% of patients have a problematic long-term adherence.¹⁴⁷ Therefore, it is essential to monitor adherence carefully and evaluate non-compliance in any case of disease progression.^51,76

Because of the low disease frequency and the lack of randomised controlled trials, the available medical treatment options are still not standardised. The efficacy of the commonly used drugs is satisfactory for hepatic disease but disappointing in regard to the treatment response of neurological symptoms. This, in part, might be explained by the irreversibility of structural neurologic damage, but includes the risk of neurological deterioration after the initiation of chelation therapy. An approach to overcome this problem is the careful and systematic assessment of biochemical response patterns and the quantitative monitoring of symptoms using validated rating scales and biomarkers. Available drugs for WD are listed in Table 4.

Table 4.

Drugs used in the treatment of Wilson disease.

Drug	Mechanism	Frequency of adverse events under drug therapy	Effectiveness of treatment (Drug monitoring by 24 h urine copper and NCC)
D-penicillamine (DPA)	Exaggerates urinary excretion of copper/chelating agent	20–30% during treatment (refs. 168,170 ) -- hypersensitivity -- arthralgia -- leucopaenia -- proteinuria -- paradoxical neurological worsening -- renal impairment -- gastric symptoms	-- at the beginning of the treatment copper urinary excretion >1000 μg/24 h) -- under therapy with DPA, copper urinary excretion about 200–500 μg/24 h -- serum NCC 50–150 μg/L
Trientine (TN)	Exaggerates urinary excretion of copper/chelating agent	7.1% (less than with DPA; ref 170) - sideroblastic anaemia (commonly indicative of overtreatment)	-- at the beginning of the treatment copper urinary excretion >1000 μg/24 h) -- under therapy with DPA, copper urinary excretion about 200–500 μg/24 h -- serum NCC* 50–150 μg/L
Zinc salts (ZS)	Inhibits intestinal absorption of copper	3–7% (ref. 168) -- gastritis -- biochemical pancreatitis	-- copper urinary excretion <75 μg/24 h -- serum NCC 50–150 μg/L; (>12 months of treatment)
Tetrathiomolybdate	Copper modulator	-- anaemia; neutropaenia -- reversible elevation of transaminases	Experimental, in clinical development

NCC, non-caeruloplasmin-bound copper. NCC is not routinely used.

^*NCC <50 μg/L indicates over treatment and modification of therapy might be needed. Decrease of NCC during treatment is currently used as an efficacy measure (ref. ⁶).

A sequential treatment concept differentiates between the initial (acute) de-coppering therapy and the subsequent maintenance therapy.¹⁴⁴ This concept takes into account that after the initial, more aggressive treatment phase addressing the high copper load, a reduced dose of the chelating drug might be sufficient to maintain copper homeostasis. Standardised dosage strategies that address changes in copper pools might improve adherence and reduce side effects. The ‘medical need’ in WD is defined by such phenotypes and identifies three subgroups of patients with increased mortality and morbidity: the incompliant patient, the patient with fulminant hepatic failure and the symptomatic neurologic patient.

In asymptomatic patients or patients with only mild hepatic symptoms, all available treatments have proven effective.^{122,146,148–153} The choice of the drug – chelators (D-penicillamine, trientine) or zinc salts – should primarily be driven by safety considerations. This favours trientine over D-penicillamine.^154,155

In patients with hepatic manifestation, the degree of liver damage determines further treatment. Validated scores can guide the decision whether the timeframe for medical treatment is wide enough or whether an urgent liver transplantation is needed.^65,156 Chelating agents are preferred to zinc salts due to the faster establishment of a negative copper balance.^153,157,158 Zinc monotherapy can be insufficient to control hepatic disease in a subpopulation of these patients.^144,154 A more recent study reports hepatic improvement with D-penicillamine or trientine in more than 90% of symptomatic hepatic cases.¹⁵⁵ However in the same study, the clinical response of symptomatic neurologic patients was less encouraging (62%). Response rate to chelation therapy was in the same range as reported for zinc monotherapy in the neurologic subgroup.^148,159 An explanation might be irreversible cell damage in the central nervous system which is already established at the time of diagnosis.¹⁶⁰ A worsening of the clinical and especially neurological symptomology is associated with chelating agents and reported in up to 20% of WD patients after initiation therapy.^{148,161–163}

Long-Term Monitoring of Wilson Disease

The overall goals of long-term treatment in WD are a balanced copper homeostasis and to confirm a clinical and biochemical improvement or stability. Monitoring should, therefore, take into account screening for patient compliance as well as signs of re-coppering or over-de-coppering. Frequency monitoring depends on the impairment of patients, but it is recommended to perform a check-up once or twice a year.⁵¹

During the initial phase of treatment or in case of worsening of symptoms, more frequent monitoring is necessary including: physical examinations; ophthalmologic examination for Kayser Fleischer rings (when patient incompliance is suggested); and laboratory testing, including liver enzymes, hepatic synthetic function values and indices of copper metabolism (serum copper and caeruloplasmin).⁹⁴ In neurologic symptomatic cases, a thorough neurologic examination is likewise recommended, preferably using (semi-)quantitative rating scales.⁸⁵

An analysis of 24 h urinary copper excretion (on medication) reflects overall CuEXC and is helpful for monitoring as well as for checking compliance. According to current guidelines, 24 h urinary copper excretion values for patients taking D-penicillamine or trientine are 3–8 μmol/day (200–500 μg/day).⁵¹ For therapy with zinc salts, it should be not more than 1.2 μmol/ day (75 μg/day). Unexpected fluctuating values for urine copper might be suggestive of non-adherence to treatment, which should be discussed with the patient. Low values for urine copper excretion under chelation therapy could also indicate overtreatment. In the latter condition, NCC values are typically low.^51,94

NCC is elevated above 250 μg/L in most untreated patients (reference interval <150 μg/L). NCC concentrations <50 μg/L are suggestive of systemic copper depletion that can occur in patients with prolonged treatment. Long-term patients taking sub-therapeutic dosages may also present with low values for 24 h urine copper excretion while having elevated NCC values. Compliance in patients taking zinc salts can be checked by measuring serum zinc or 24 h urinary zinc excretion (reference threshold: 2 mg/day). Urinalysis as well as total blood count should be performed regularly to assure safety under chelation therapy to detect adverse proteinuria or anaemia.^51,94

Emerging Therapies and Management of Extra-Hepatic Copper

An unmet medical need in WD is the clinical non-response or even drug-induced worsening of neurologic disease. A suspected correlation with inadequate biochemical response pattern of copper metabolism (serum copper, NCC and copper excretion) is discussed.¹²¹

Following this concept, the mobile ‘free’ copper pool (assessed by the surrogate NCC) would essentially contribute to the neurologic copper toxicity. Evidence for different copper pools (hepatic vs NCC) and such shift under therapy were noted early.¹⁶⁴ Only a few studies assessed copper concentrations in the cerebral fluid under treatment in humans or the copper levels in the brain in WD animal models.^145,165 However, these studies are consistent with the association of increased free copper and a neurologic deterioration. In Brewer’s study from 2009, worsening neurological condition was associated with significant spikes in serum free copper levels at the time of deterioration.¹⁴⁵ Patients who did not worsen neurologically generally did not show significant spikes in free copper. This raises the hypothesis that effective medical therapy in the neurologic WD patient should aim for a rapid and substantial control of the free copper, which can be verified by free copper calculation.¹⁴⁵ In summary, a high free copper – either calculated as NCC or directly measured as CuEXC – while undergoing treatment and an increase in urinary copper excretion can be suggestive of an upcoming difficult treatment situation.

The future concept to develop dosing strategies based on changes in copper pools (NCC or CuEXC) is highly promising. The question of whether copper peaks provoke clinical deterioration is still unsolved and under debate for chelating drugs. In general, chelator-induced neurologic or clinical deterioration seems to occur more frequently in patients on a high starting dose.¹⁴⁴ Therefore an up-titration of the dosage should be used instead of a high initiation dosage.

Currently, the most promising candidate for a biochemical guided dose and treatment regimen is bis-choline-tetrathiomolybdate (WTX101), an oral first-in-class copper-protein-binding agent.^{145,161,166–168} In a proof-of-concept phase II trial, once-daily WTX101 over 24 weeks rapidly lowered NCC levels and this was accompanied by improved neurological status without apparent initial drug-induced paradoxical worsening, reduced disability, stable liver function, with a favourable safety profile.¹⁶⁸ WTX101 directly removes excess copper from intracellular hepatic copper stores and also forms an inert tripartite complex with copper and albumin in the circulation and promotes biliary copper excretion. These mechanisms may explain the rapid biochemical and clinical improvements observed. A phase III trial of WTX101 is ongoing and results are eagerly awaited.

Another promising approach for restoration of hepatic copper metabolism in WD is gene therapy.¹⁶⁹ By successful ATP7B gene transfer, hepatic or neurological damage could be prevented before the clinical manifestation of the disease. For cell or gene therapy in WD, the rationale for targeting the liver first and foremost is based on the fact that ATP7B expression is physiologically restricted to the liver. Even early gestational gene transfer might be considered to adjust impaired copper metabolism during foetal development. One of the foremost mandatory requirements for successful gene therapy is a stable and sustainable gene transfer, ensuring a stable long-term expression of the transferred gene. In this regard, HIV-derived lentiviral vectors (LV) show favourable capacity, whereas LV are able to efficiently integrate a stable gene replacement into the genome of non-dividing cells. Animal studies gave proof-of-principle that transduction of ATP7B LV leads to hepatic expression of the transgene in hepatocytes, resulting in lower liver copper levels and improvement of fibrosis compared to untreated animals.¹⁷⁰ In another murine model of WD, even prenatal gene transfer by injection of LV containing the human ATP7B resulted in decreased hepatic copper content. This study applied for the first time an in utero gene therapy in WD.¹⁷¹ Another current approach of gene therapy used a small single-stranded adeno-associated virus (AAV), to transfer ATP7B with an hepatocyte-specific promoter into murine models for WD.¹⁷² Gene transfer yielded a sustainable expression of ATP7B with a subsequently sufficient restoration of copper metabolism six months after only a single injection. The application of this liver-directed gene therapy seems to be one of the most promising techniques in WD for the future.

Conclusion and Outlook

Improved diagnostic techniques and more systematic approaches in diagnosis offer an easier path to confirmation or exclusion of a diagnosis of WD and therefore an earlier diagnosis. Currently-used zinc salts and chelators are mostly effective but have some limitations. Especially chelating agents are associated with a deterioration of neurological symptoms. We need better information about dosing, monitoring and studies of the comparative effectiveness of available medical therapies and standardisation of definitions for treatment success and failure. A recent clinical trial with WTX101 that leads to the formation of an inert complex of WTX101 with copper and albumin in the circulation and increases biliary copper excretion seems very promising, especially with respect to low potential for neurological worsening with initiation of therapy.¹⁶⁸ Independent of which parameter the decision is based on, the quest for the optimal drug should include the optimal dosage regimen. This requires the collection of as many pharmacodynamic data as feasible, including individual biochemical patterns and neurologic response scores. Rapid control of ‘free’ copper could be the dosage or treatment goal of the future for all chelators. However, broad clinical evidence is lacking to allow general recommendations for dosage adjustments and biochemical target corridors for NCC, urinary copper excretion as well as for CuEXC. A new approach for restoration of hepatic copper metabolism in WD is gene therapy, which has shown promising results in animal studies and, in future, needs to be transferred to the human model. There are many opportunities for early disease diagnosis, successful medical therapies and rescue therapies such as liver transplantation. New opportunities, especially in dosing strategies of WD and biochemical response patterns, will emerge in future with perhaps even better outcomes.