Wilson disease: the diagnostic challenge and treatment outcomes in a series of 262 cases

Abstract Background  Wilson disease (WD) is an autosomal recessive disorder that leads to organ toxicity due to copper overload. Early diagnosis is complicated by the rarity and diversity of manifestations. Objective  To describe the diagnostic features and response to treatment in our cohort of WD patients. Methods  This was a retrospective analysis of 262 WD patients stratified by clinical presentation, complementary exams, ATP7B genotyping, and response to treatment. Results  Symptoms occurred at an average age of 17.4 (7–49) years, and patients were followed up for an average of 9.6 (0–45) years. Patients presented mainly with hepatic (36.3%), neurologic (34.7%), and neuropsychiatric (8.3%) forms. Other presentations were hematologic, renal, or musculoskeletal, and 16.8% of the patients were asymptomatic. Kayser-Fleischer rings occurred in 78.3% of the patients, hypoceruloplasminemia in 98.3%, and elevated cupruria/24h in 73.0%, with an increase after D-penicillamine in 54.0%. Mutations of the ATP7B gene were detected in 84.4% of alleles. Brain magnetic resonance imaging showed abnormalities in the basal ganglia in 77.7% of patients. D-penicillamine was the first choice in 93.6% of the 245 patients, and 21.1% of these patients were switched due to adverse effects. The second-line therapies were zinc and trientine. The therapeutic response did not differ significantly between the drugs ( p  = 0.2). Nine patients underwent liver transplantation and 82 died. Conclusion  Wilson disease is diagnosed at a late stage, and therapeutic options are limited. In people under 40 years of age with compatible manifestations, WD could be considered earlier in the differential diagnosis. There is a need to include ATP7B genotyping and therapeutic alternatives in clinical practice.


INTRODUCTION
Wilson disease (WD) is an autosomal recessive disorder caused by mutations in the ATP7B gene that leads to an accumulation of copper (Cu), particularly in the liver, brain, and cornea.][3][4] In the 1940s, the WD diagnosis was based on typical neurologic and hepatic findings, the presence of Kayser-Fleischer (KF) rings detected on the biomicroscopic examination with the slit-lamp, family history, and postmortem liver histology.In the 1950s, WD diagnosis was confirmed by hypoceruloplasminemia, increased 24-hour urinary Cu excretion, and increased Cu content in biopsy-derived liver tissue. 5However, in many cases, these parameters are not sufficient to establish a diagnosis.][6] Medications to be taken orally include Cu chelating agents and metallothionein inducers.Liver transplantation is appropriate in patients with decompensated liver disease who do not respond to oral medications and do not have severe neurological impairment, or in patients with severe acute liver failure. 5,80][11][12][13][14][15][16][17] Given the complexity of WD and the difficulty in diagnosing it, the aim of this study was to describe the clinical presentation, results of diagnostic tests, patterns of response to treatment, and outcomes of WD patients treated at our center.

METHODS
A cohort of WD patients (n ¼ 262) from 198 distinct families was analyzed.Data were obtained from a database (n ¼ 280) of patients treated between 1946 and 2020 in the departments of gastroenterology, neurology, or pediatrics of Hospital das Clínicas of the School of Medicine of Universidade de São Paulo.After 18 patients were excluded due to insufficient records, the data were collected retrospectively by reviewing the medical charts.
The diagnosis of WD was based on typical clinical manifestations or familial screening confirmed by complementary exams.Diagnostic parameters, including neuroimaging (computed tomography [CT] or magnetic resonance imaging [MRI] of the brain), and 24 h urinary Cu stimulated by D-penicillamine (DPA), have evolved over the years, as standardized in considered earlier in the differential diagnosis.There is a need to include ATP7B genotyping and therapeutic alternatives in clinical practice.
Kayser-Fleischer rings were seen on biomicroscopic examination with the slit-lamp.Response to treatment response at ! 12 to 18 months was determined by history and laboratory tests.Response was complete if there was a complete improvement in symptoms, laboratory changes, disappearance of KF rings, and stabilization of urinary Cu excretion at 200 to 500 µg/24h.It was incomplete when some of these parameters remained abnormal, and it was inefficient when these parameters worsened.Adherence was good if the medication was not interrupted for > 7 days, irregular if the interruption time was between 8 and 30 days, and poor if it was interrupted frequently or !30 days.The χ 2 test, Fisher test, and unpaired t test were used for the statistical analyses (significance at p < 0.05).

RESULTS
We included 262 individuals from 198 families diagnosed with WD, 154 men (18.7; 5-53 years) and 108 women (8.6; 4-51 years).There was no overall gender difference (p > 0.05), but males predominated in neurological manifestations (p ¼ 0.02).Symptoms appeared at 17.6 AE 6.5 (6-49) years, while the diagnosis was made at 18.7 AE 7.5 (4-53) years.The psychiatric form was invariably accompanied by neurological symptoms; therefore, it is referred to as the 'neuropsychiatric' form.►Figure 1 illustrates the age at onset of symptoms and at diagnosis.►Table 1 illustrates the time between symptom onset and diagnosis over the decades.

Clinical presentation
►Table 2 illustrates the distribution of clinical presentations.Dysarthria (89.0%) was the most common neurologic manifestation, as previously reported. 21During follow-up, motor impairment disorders were noted in 155 patients, classified as mild in 20.6%, moderate in 45.8% or severe in 33.6% (►Figure 2).
Decompensated chronic liver disease was found in 57.9% of patients with the hepatic form.Acute liver failure occurred in 5 patients (4 women, 2 of whom had Coombs-negative hemolytic anemia, mean age 13.2 (9-18) years.
Correlations of clinical presentation with diagnostic parameters are shown in ►Table 3. Mutations of the ATP7B gene were detected in 78.3% of alleles (191/244).Kayser-Fleischer rings were present in all hematologic patients and remarkably also in 5 asymptomatic patients, while they were absent in only 2 neurologic/neuropsychiatric patients (p < 0.001).
In patients with acute liver failure (n ¼ 5), ALT levels were twice the reference value, with a median of 29 (22-67) IU/L and a mean alkaline phosphatase (AP) level of 85 (9-883) IU/L.These patients had AST values 5 times higher than the upper reference, with a median of 135 (58-195) IU/L and hyperbilirubinemia, with a median of 28.6 (5.3-46.3)mg/dL.In 3 patients, the AST/ALT ratio was > 2.2, and in 2 patients, the AP/total bilirubin ratio was < 4. Hypoceruloplasminemia was found in 98.7% of patients (207/257) and was mostly < 10 mg/dL, with a median of 4.2 (0-26) mg/dL.The neuropsychiatric patients had the lowest values (►Table 3).

Radiological imaging of the brain
Magnetic resonance imaging of the brain showed abnormalities in the basal ganglia in 42/54 (77.7%) patients.In five neurological cases, no imaging signs were detected.One patient with a neuropsychiatric form had pontine myelinolysis.Changes in signal intensity in the cerebellum were observed in seven patients.Magnetic resonance imaging was normal in four neurological cases.Cranial CT showed cortical atrophy in 11/32 (34.4%), basal ganglionic hypodensity in 9/32 (28.1%), and ventricular dilatation in 7/32 (21.9%).

Liver parenchymal findings
Liver biopsies were performed in 100 cases.Seven samples were obtained from liver explants and 18 came from necropsies.Cirrhosis was already present in 64% of the samples.Steatosis was found in 33% and glycogenotic nuclei in 20% of the samples.Rhodanine staining was performed in 33 specimens, 18 of which were positive (54.5%), and in 3 cases changes suggestive of autoimmune hepatitis were detected.Copper levels were quantified in 10 patients, 2 of them in a postmortem sample.Four individuals had hepatic Cu concentrations < 250 µg/g dry weight.

Analysis of the ATP7B gene
In the sequencing of the ATP7B gene, 37 distinct mutations were identified, and a detection rate of 89.8% (230/256 alleles) was achieved.Mutations were found in both alleles in a total of 108 individuals (84.4%), while no mutations were detected in 3.1% patients.There was no significant difference in clinical presentation or number of alleles affected (►Table 3).The most frequent mutation was p.A1135Qfs (32.8%), followed by p.L708P (14.8%), while p.H1069Q had an allelic frequency of 5.1%.

Treatment
The patients were followed up for 9.6 years (range 0-45 years).Unfortunately, 55 patients were lost to follow-up, and 82 patients (39.6%) died.
The first case of WD occurred in the neurological clinic in 1946.The first 3 patients in this study died within 5 years of symptom onset, prior to the introduction of specific WD therapy.During this time, treatment options were limited to palliative measures or dimercaprol, which proved ineffective.Nine patients receive no treatment as they died during the investigative period.D-penicillamine (DPA) was the mainstay of therapy in 229 of 245 (93.6%) drug-treated patients.After an initial dose of 250 mg/day, it was gradually increased to 750 to 1,500 mg/day.In pediatric patients, the starting dose is usually 20 mg/kg/day.Pyridoxine was consistently administered in combination with DPA at a dose of 25 to 50 mg/day.Trientine was the treatment of first choice in four patients and was prescribed at a similar dose to DPA.Elemental zinc at a dose of 150 mg/day was given to 10 patients.One patient was given dimercaprol at a daily dose of 200 mg via the intramuscular route.All patients were advised to avoid foods with a high Cu content for at least the 1 st year.
Data on the results of pharmacotherapy were available for 183 people, of whom 29 (15.8%)responded completely and 78 (42.6%) incompletely to the therapy.It was ineffective in 41.5%.Adherence to therapy was good in 112 (61.2%) patients, irregular in 37 (20.2%), and poor in 34 (18.6%).Clinical deterioration was observed in 29 (85.3%)patients with poor adherence, while 35 (31.3%) had good adherence (p < 0.001).There was no significant difference between the therapeutic outcomes of the 3 main drugs (p ¼ 0.2).All patients in the asymptomatic group who showed good adherence were healthy.There were significant differences between the asymptomatic patients treated with DPA and those treated with neurological (p ¼ 0.012) or hepatic symptoms (p < 0.001).
Two out of 5 patients with acute liver failure died before undergoing liver transplantation.Nine patients were successfully treated by transplantation (3 with acute liver failure and 6 with decompensated chronic liver disease).One patient died 3 years after transplantation due to bronchoaspiration.

DISCUSSION
The current series provides a comprehensive overview of our experience with WD, from diagnosis to treatment and outcomes over the past 7 decades.Previous publications have focused on specific aspects of subgroups of this cohort, [20][21][22][23][24] such as genotype-phenotype correlation, neurologic manifestations, and rare presentations.However, long-term follow-up data could provide food for thought for better management of WD.
Patients with a hepatic presentation were younger than patients with neurological symptoms. 12,25,26One possible explanation could be the natural course of the disease.However, the oldest (53 years old) patient in this cohort who started with extrapyramidal manifestations had no signs of clinical or histologic liver disease.Another patient from a familial screening at age 51 had a ceruloplasmin level of 3 mg/dL with no other signs of WD; therefore, she was taken off medication and monitored every 6 months.It was not until the 6 th year of follow-up that a slight increase in AST/ALT ratio and changes in liver imaging were noted.Liver histology revealed steatohepatitis that did not react to rhodanine staining.ATP7B genotyping revealed p.A1135-Qfs/p.M645R mutations.The patient's brother had been diagnosed with decompensated cirrhosis at the age of 34.It is known that WD can occur at any age and tends to be less severe when symptoms begin at an older age.This indeterminate pattern may be due to various environmental factors (such as dietary Cu levels), which are associated with genetic factors.Compound heterozygosity, including the p.M645R mutation, has been found in patients with hypoceruloplasminemia, but the findings are unusual. 27Unfortunately, in WD patients diagnosed by familial screening, with the exception of asymptomatic patients, the time between the onset of symptoms and the start of treatment was not shortened.
Direct sequencing of the first 60 patients in this series showed a mutation detection rate of 90% (108/120 alleles), and mutation in both alleles were detected in 81.7% of patients, as previously reported. 20In the present study, 68 subjects were added to this cohort using the same techniques; the detection rate was 89.8% (230/256), and mutations in both alleles were detected in 108 patients (84.4%).We emphasize that the 3 patients from 2 different families in whom no ATP7B mutation was detected were diagnosed as definite WD cases, with typical clinical, laboratory, and imaging features and adequate response to treatment.9][30][31][32] The improvement of primer design and amplification techniques could contribute to a higher yield.In addition, there are mutations in promoters and intronic regions that are not accessible by direct sequencing.Recently, next-generation sequencing (NGS) has emerged as a technique that can elucidate many of these cases. 33pigenetic factors causing genomic imprinting and other genetic disorders that mimic WD may also be considered.The large number of different mutations in patients with very different clinical features does not yet allow to establish clear genotype-phenotype associations.Although more than 900 ATP7B mutations have been described in the literature over the last 2 decades, the mechanisms involved in the development of WD are not yet fully understood. 17,34nterestingly, the group from Austria claims that age and gender actually influence the WD phenotype more than the ATP7B genotype does. 32ayser-Fleischer rings were detected in 78.3% of patients but were absent in 2 (1.1%) patients with the neurologic form.As expected, 86.8% of the asymptomatic patients did not have these symptoms.Hypoceruloplasminemia was detected in 98.7% of patients. 13The penicillamine challenge test was positive in 18/33 adults (54.5%).Based on the European scoring system 19 criteria, five patients were definitively diagnosed with WD after the penicillamine challenge test.Although this is helpful, we emphasize that most healthy parents of WD patients also show a pronounced reaction to penicillamine, sometimes more than 15 times the reference value. 35mong the patients with neurological symptoms, most were male (68.1%).One hypothesis is that men have higher concentrations of ferritin-bound iron in the brain, and that this metal is associated with neurodegenerative diseases. 36owever, only 1/6 of patients with hepatic siderosis had neurological symptoms.
Both Cu excess and deficiency correlate with fatty liver disease. 37In this cohort, 28 patients with liver biopsy had steatosis and inflammation.Liver biopsy was performed on admission (n ¼ 21), during Cu chelator therapy (n ¼ 7), or after removal of excess Cu (n ¼ 5).
Interestingly, 2 patients had smooth muscle antibody levels > 1/320 (female) and > 1/160 (male).Both patients responded well to Cu chelator without immunosuppressive drugs; thus, the antibodies are rare epiphenomena. 38Penicillamin was the first choice as trientine is not yet easily accessible for the vast majority of patients.Zinc salts were used in the early 1980s as a second option for the treatment of WD in patients who could not tolerate DPA.Since the 1990s, zinc salts have been used as first-line treatment for severe neurological conditions and asymptomatic patients.Although 77.3% improved liver function after starting treatment, 20% deteriorated.Liver biochemical tests improved in 39.7%, but worsened in 17%.There was a clear distinction between clinical and laboratory improvement; however, when patients deteriorated clinically, laboratory parameters reflected this trend.Iorio et al. 39 also observed persistent elevation of alanine transaminase (ALT) in patients who remained clinically stable.Therefore, lowering AST/ALT levels may not be the goal of successful therapy.
In the subgroup of decompensated chronic liver disease treated with DPA, 11 patients died within 6 months of starting therapy, 7 died between 7 months and 4 years, and 1 patient did not respond satisfactorily to therapy.Thus, some patients may not respond to copper chelation and liver transplantation is a plausible alternative for nonresponders when liver function deteriorates. 40f the 33 patients who continued with KF rings, 21 (63.6%)experienced clinical deterioration, while the 28/29 (96.6%) in whom the KF rings disappeared also showed a regression of neurological symptoms.Kayser-Fleischer rings can, therefore, indicate treatment compliance and response.
Three asymptomatic patients had poor adherence to Cu chelators and manifested liver disease in the follow-up.All asymptomatic patients with satisfactory adherence had a good outcome.
While therapeutic response did not differ between the three drugs, 53% of patients treated with DPA experienced adverse reactions, 21.1% experienced significant reactions requiring drug substitution, and one patient died from nephrotoxicity.One patient reported elsewhere developed myeloneuropathy after 14 years of zinc therapy. 22This could be due to Cu deficiency or the toxic effects of zinc, as both excess and deficiency are harmful.
In conclusion, WD still mainly affects young patients, who also find it difficult to adhere to treatment.ATP7B genotyping can help with early diagnosis but cannot predict severity or presentation.A high suspicion of WD is crucial for early diagnosis.Penicillamine is the most commonly prescribed treatment, although it does not work satisfactorily.Another challenge in the treatment of WD patients is the identification of better tolerated medications and the development of strategies to improve adherence.formal analysis; GP: conceptualization, data curation, formal analysis, and investigation visualization; ERB: conceptualization, data curation, formal analysis, investigation, methodology, supervision, and visualization; ELRC: conceptualization, data curation, formal analysis, investigation, methodology, project administration, resources, supervision, visualization, and writingreview & editing.Support Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) 06/00499-1.

Figure 2
Figure 2 Neurological presentation of WD.A severely disabled young man presents with rigid dystonic posture, dysarthria, dysphagia requiring feeding through a gastrostomy tube.

Table 1
Wilson disease onset and time to diagnosis in a cohort of 262 patients Abbreviation: SD, standard deviation.

Table 2
Clinical presentation at disease onset in Wilson disease patients (n ¼ 262)

Table 3
Clinical presentation form and diagnostic parameters in Wilson disease (n