Phenotypic and Genetic Characterization of a Cohort of Pediatric Wilson Disease Patients

Wilson's disease (WD) is a rare autosomal recessive disorder of copper metabolism, with a prevalence of about 1 in 30 000 people. Its frequency increases in populations where consanguinity is more common [1]. It is characterized by decreased biliary copper excretion and defective incorporation of copper into ceruloplasmin, leading to copper accumulation in different tissues mainly the liver, brain and kidneys [2].

WD typically begins with a pre-symptomatic period, during which copper accumulation in the liver causes subclinical hepatitis and progresses to liver cirrhosis and development of neuropsychiatric symptoms. The type of hepatic and neurological symptoms can be highly variable. It may also present as fulminant hepatic failure with an associated Coomb's-negative hemolytic anemia and acute renal failure [3]. If untreated WD causes progressive fatal liver and brain damage [4] and therefore, early recognition and treatment is required.

In 1993, the WD gene, ATP7B, was cloned. This gene codes for a membrane-bound copper -transporting ATPase expressed primarily in the liver [5, 6]. More than 400 mutations within the ATP7B have been identified so far (personal data). Although molecular genetic diagnostics are increasingly available for clinical use, in practice their use is very limited.

The diagnosis of WD is based on the results of several clinical and biochemical tests, each has its limitations, and only the combination of clinical, biochemical and genetic tests provides a powerful and reliable tool for the diagnosis [3].

Zinc sulfate and D-penicillamine (D-PCA) have both proved to be clinically effective in the treatment of WD, but zinc sulfate offers the advantage of having a very low toxicity, especially in the case of neurological involvement. For asymptomatic or pre-symptomatic patients, zinc salts or maintenance dosages of chelating agents may be used [7].

In Egypt, where viral hepatitis (HCV, HBV) is by far the major player in the field of liver disease, WD seems to be under diagnosed and clinical data on large cohorts are limited owing to its low frequency, being a rare disease. The aim of this study was to highlight the clinical, laboratory and genetic characteristics of WD in our pediatric population as well as to report our experience with both treatment options and outcome aiming at increasing awareness about diagnosis and management. Based on the mutational spectrum rapid screening approaches may be applied for Egypt in the future.

The study included 77 WD patients (75 children and 2 adult cousins) (43 males and 34 females), 62 of whom (34 males and 28 females) were followed up for a mean period of 58.9 ± 6.4 months. Patients were from 50 unrelated families presenting to Yassin Abdelghaffar Charity Center for liver disease and research and the Pediatric Hepatology clinic, Ain Shams University (both are tertiary referral centers) from 1992 to 2009. One or more authors cared for all patients. Data were collected retrospectively by record analysis and patient interviews.

The study was approved by the ethical committee of both Yassin Abdelghaffar Charity Center for liver disease and research and Ain Shams University and informed consent for inclusion in the study was obtained from parents of children.

Initial diagnosis of WD was made if the patient had hepatic and/or neurologic disease in addition to at least two of the following six criteria (adapted and modified from Dhawan et al., 2005)[8]:

According to the type of presentation, symptomatic patients were classified into three groups:

Brothers and sisters of all patients diagnosed with WD were screened for the disease by being subjected to full history taking, physical examination, serum ceruloplasmin, liver function tests, slit lamp examination for KF ring, and molecular testing. Those who had no symptoms at presentation and were discovered to have WD during screening of family members of the index case comprised Group 4 (G4): Asymptomatic cases (n = 27).

Patients with liver symptoms (both G1 and G 3) were then re-classified according to type of hepatic presentation into:

Liver function tests and other routine laboratory data (including viral markers and autoantibodies) were performed using standard methods to evaluate the degree of liver affection as well as to identify/exclude any associated hepatic condition. If not contraindicated due to severe coagulopathy or advanced refractory ascites, liver biopsy was obtained (n = 39) before initiation of treatment. MRI brain was performed in ten patients with neurological symptoms

Diagnosis was confirmed by mutational analysis (sequencing of ATP7B gene) in 64 patients (25 asymptomatic and 39 symptomatic) (48 of whom were previously reported in Abdelghaffar et al., 2008 study) [9].

The standard computer program SPSS for Windows, release 13.0 (SPSS Inc, USA) was used for data entry and analysis. All numeric variables were expressed as mean ± standard deviation (SD). Comparison of different variables in various groups was done using student's t test and Mann Whitney test for normal and nonparametric variables, respectively. Multiple regression analysis was also performed to determine effect of various factors on a dependent variable. Chi-square (χ²) test was used to compare frequency of qualitative variables among the different groups. Spearman's correlation test was used for correlating non-parametric variables. For all tests a probability (p), less than 0.05 was considered significant. Survival statistics was done using the Kaplan Meire curve [11].

Sequencing of the ATP7B gene was performed in 65 patients. Mutations within the gene were identified in 64 of them and in only one patient no mutations were detected (he was excluded from the study). Eleven mutations were novel (table 5). Fifty four patients (85.7%) had homozygous mutations while only 9 (14.3%) were compound heterozygous. For better genotype-phenotype analysis, only homozygous mutations were considered for further analysis:

In this cohort of children with WD who are mostly Egyptians, there are a number of diagnostic and clinical features, which differ from those previously reported in the Western literature. The KF ring, which in absence of chronic cholestasis is the best diagnostic sign of WD, has been found in 96.4% of our children with purely hepatic manifestations. All patients with acute hepatitis as well as all patients with ALF had a KF ring. Furthermore, it was found in 17.4% of asymptomatic siblings. These figures are substantially higher than those reported in children with hepatic WD. Merle et al., 2007 reported the presence of a KF ring in 66.3% of their pediatric cohort. In most other studies, a KF ring was present in less than 50% of the patients [3, 8, 13]. The youngest child in this cohort positive for KF ring was only 6 years old. Although this high frequency of KF ring could indicate early onset of neurological disease, the possibility of environmental exposure to excessive amounts of copper could not be ruled out.

Only 6.5% of our patients had normal ceruloplasmin. Dhawan et al., 2005 found a normal ceruloplasmin in 20% of their pediatric cohort [8]. It seems thus that within this cohort, it is very uncommon to find symptomatic patients negative for KF ring and having a normal ceruloplasmin level, suggesting that the value given to different items in the Leipzig score might have to be modified for different ethnic groups.

Coomb's negative hemolytic anemia occurred at presentation in only 5.4% of patients, a percentage that is much less than that reported by Walshe 1987 (11%) [14], and slightly more than in Japanese patients (1.2%) [4].

Although in this study urinary copper excretion before and after D-PCA challenge yielded results diagnostic of WD in 84.2% of patients, it was performed in only 25 patients in this cohort, thus revealing the difficulty in obtaining accurate 24 hour urine collection for performing this test in children and confirming previous reports [8].

Other features characteristic of this pediatric cohort was the low age of onset of the disease. Median age of onset in our children was 10 years. Previous studies reported an onset of 12.2 years in Japan, 11 ± 7 years in Iran and 7.2 years in India [15‐17]. This finding may suggest that Egyptians may have an early age of onset of WD. Whether this finding is related to the effect of environmental factors or to polymorphisms in the MURR1 gene needs further studies.

While cirrhosis is frequently found in most patients with WD by the second decade of life [18], in our study almost two thirds of cirrhotic patients were ≤ 10 years old (13/21). In addition, 45% of asymptomatic children were already cirrhotic at the time of diagnosis (as confirmed by liver biopsy) with their median age being only 8 years. The youngest patient with cirrhosis in this cohort was only 7 years old. Ascites was present at presentation in more than half of the patients with liver symptoms. Again, a figure higher than that reported in other studies [4, 19]. This implies that the hepatic disease runs a more aggressive course in the Egyptian children a finding that might be related to genetic factors, undetermined environmental factors or gene modifiers.

Co-existence of WD and viral infections may lead to early onset, severe presentation and rapidly progressive liver disease [20]. True co-morbidity of WD at the time of diagnosis is rarely reported, no more frequently than expected by chance [21]. It is to be noted that in this study the three patients who had schistomiasis ran a detrimental course (one died, one developed neurological symptoms and the last had deterioration of his liver functions). To the best of our knowledge, the effect of schistomiasis on the course of WD has not been reported before.

Due to the high incidence of consanguineous marriage (78.9%) and the extended family size in our community, asymptomatic siblings constituted 35% of this cohort, thus allowing for studying the disease in its presymptomatic phase. Most of those children (81.5%) had hepatomegaly, one third had splenomegaly (37.5%) and the only disturbed LFTs were ALT/AST. In spite of being asymptomatic, yet histologically evident cirrhosis was present in 45% and a KF ring was detected in 17.4%. The youngest affected sibling was only one year old. This emphasizes the need for screening of all sibs of affected children even at a younger age than previously recommended.

Nearly one third of the patients (29.5%) with hepatic manifestations presented with ALF. A slightly higher percentage was reported by Dhawan et al. (33%) [8], yet, lower figures were reported in other studies [22]. Again, the age of onset of patients with ALF in this cohort (median: 9 years) is less than generally reported. In our cohort as well as that of Dhawn et al., there was a male predominance. This is in contrast to previous reports where females were more commonly represented among WD patients with ALF [21]. It has been suggested that the female predominance was due to the effect of sex hormone as shown in the animal model of WD [23]. Dhawan et al. attributed these contrasting results to the fact that most of their patients were prepubertal, which holds true for our group of patients as well [8].

In this study, as in others [8, 24‐27], D-PCA has been shown to be an effective copper chelator in both symptomatic and asymptomatic WD patients who are compliant to treatment.. Side effects were observed in 30.76% and were severe enough to discontinue the drug in 11.5%. Such figures are slightly less than those reported by others for both adults [3] and children [8, 28]. In Egypt, the only available decoppering agent is D-PCA, so we tried hard to limit its side effects by starting therapy with a very small dose that was very gradually increased to reach the maximum desired dose meanwhile giving zinc sulphate starting with a maximal dose and gradually reducing it as D-PCA was increased.

Of the children born to WD mothers one suffered from Fallot's tetralogy and another was diagnosed as Budd-Chiari syndrome at 6 months of age. In both cases, the mother did not receive D-PCA during pregnancy and they both were totally non-compliant on zinc therapy. This may implicate intrauterine exposure to excess copper, a teratogenic substance, as the cause in both children.

Zinc monotherapy has been shown to be an effective non-toxic alternative to D-PCA in asymptomatic siblings [29]. It was approved in 1997 by the US FDA as maintenance therapy. Yet some reports indicate its efficacy also as first line treatment of symptomatic patients [7]. Three of the asymptomatic children in this study have been effectively treated with zinc monotherapy. It was also used in the six patients for whom D-PCA had to be discontinued and in all it was effective. The mechanism of action of Zinc is different from that of D-PCA as it inhibits copper absorption by inducing enterocyte metallothionin. It may also induce hepatocyte metallothionin [30] and can generate a negative balance for copper thereby removing stored copper [31]. We used the combination of Zinc sulphate (a single midday dose) and D-PCA for our patients, spacing them at least 6 hours apart. Although the use of a single midday zinc sulfate dose decreases the chance of non compliance, its effectiveness in the production of a negative copper balance is uncertain.

Six patients in this series died within two weeks of presentation. They all had a score >11 in the new Wilson predictive index proposed by Dhawan et al. 2005. None of the patients with score <11 experienced early death, yet two patients with a score >11 survived on medical management without transplantation. In their original description of the score, Dhawan et al. 2005 indicated that a score >11 was predictive for death with a sensitivity of 93%, specificity of 98% and positive predictive value of 88% [8].

During follow up ten patients died, in six of whom death resulted from hepatic decompensation. All six were non-adherent to therapy. The adverse effect of non- compliance to therapy on survival has been clearly demonstrated in the Kaplan Meier survival curve. Nearly 24% of our patients were non adherent to therapy more than 50% of the time. Brewer et al. (1994) found that about 10% of the patients had serious non adherence problems and another 20% had episodic non adherence when followed up for 5-10 years [29]. Arnon et al. (2007) reported that poor adherence to therapy might be responsible for the persistence of elevated ALT [10]. The patients in this study who developed liver decompensation could not be rescued since liver transplantation became available in Egypt only since late 2001.

Homozygous mutations were present in 54 patients (85.7%). This high percentage may be explained by the high percentage of consanguinity in our study (78.9%) compared to the consanguinity in the Egyptian population (32-35%) [32]. In spite of the high percentage of homozygous mutations, it was difficult to find genotype-phenotype correlations due to the large number of mutations detected. Heterogeneity of the Egyptian population with respect to ethnicity may justify the presence of such big number of different mutations that may also reflect a high carrier rate for WD in our population.

Protein truncating mutations profoundly affect gene functions [27, 33]. Patients with protein-truncating mutations in this study had slightly earlier age of onset, died earlier from ALF and neurological symptoms were more common among them.

Egyptian children with WD present with early KF rings, and early onset of liver and neurological disease. The mutation spectrum identified differs from that observed in other countries. This study offers screening strategies for WD in Egypt, which may not depend on genetic testing.