Background
Wilson’s disease (WD) is an autosomal recessive disorder caused by mutations in the
ATP7B gene leading to excessive copper overload, predominantly in the liver and the brain [
1,
2]. The severity of the disease varies considerably between patients, and it remains unclear why some patients have hepatic symptoms while others develop neurological, psychiatric, or combined symptomatology [
3‐
5]. Hepatic features range from asymptomatic patients with slightly elevated liver enzymes to liver cirrhosis or acute hepatic failure [
6]. Neurological symptoms are usually characterized by multiple motor impairments characteristic for dysfunctions of the basal ganglia and the cerebellum. This results in various neurological symptoms including rigidity, tremor, dyskinesia, dystonia, ataxia, chorea, dysarthria, dysphagia, or excessive salivation [
6‐
10]. Psychiatric symptoms are diverse and may include concentration difficulties, attention disorders, behavioral abnormalities with alterations of personality, depression, and psychosis [
11‐
14]. The severity of the symptoms clearly determines the extent to which a patient is restricted in their activities of daily living (ADL).
To prevent progression of the disease and lethal outcomes, lifelong therapy pursuing a negative copper balance is imperative; hence, underlining the importance of an early diagnosis and an individually adapted therapy [
15]. However, neurological worsening in WD patients following initiation of penicillamine standard therapy is a major concern [
16]. Given the multifaceted symptomatology in WD, a comprehensive, standardized, and practical clinical rating scale is indispensable to monitor individual therapeutic effects and side effects in daily practice and to use as a valid end point in clinical trials testing therapies.
Currently, there are two clinical rating scales for WD: the Unified Wilson’s Disease Rating Scale (UWDRS) and the Global Assessment Scale for Wilson’s Disease (GAS for WD). The UWDRS has neurological, hepatic, and psychiatric subscales. The neurological subscale of the UWDRS was developed in collaboration with the European Network, EuroWilson, and the German Network of Hereditary Movement Disorders (GeNeMove). The first WD rating scale reflecting the extent of neurological impairment was reported by Czlonkowska et al. in 2007 [
17]. It was extended by hepatic and psychiatric subscales that were evaluated in total by Leinweber et al. in 2008 [
18]. The GAS for WD, considerably shorter than the UWDRS, is comprised of two tiers scoring global disability (Tier 1) and neurological dysfunction (Tier 2) and was proposed by Aggarwal et al. in 2009 [
19]. Potential difficulties that have arisen from both rating scales in daily practice are the relatively long duration of the complete clinical assessment and the formal requirement of two to three different medical specialists (i.e., neurology, gastroenterology, and psychiatry).
So far the UWDRS and the GAS for WD have not been directly compared, and it remains unclear which rating scale may be superior. Hence, the goal of the present study is to evaluate the UWDRS and the GAS for WD in routine clinical practice and to propose a less time consuming patient reported prescreening tool for use outside scientific trials, the “minimal UWDRS”.
Discussion
In the clinical management of WD, thorough clinical monitoring including early detection of potential medical side effects and therapy failure is crucial to improve the patient’s motivation and adherence [
23]. Therefore, liver function tests and the calculation of MELD and Child-Pugh scores on an every three months basis are recommended [
6,
21,
24]. During the initial phase of treatment, the modified Nazer score is a reliable tool to identify patients at risk for hepatic treatment failure requiring liver transplantation [
25,
26].
Additionally, ultrasound of the liver visualizes the status of cirrhotic reorganization. However, an adequate assessment of neurological impairment has long been neglected due to the lack of a suitable tool for evaluation. Standardized neurological assessment allows for both intra- and inter-individual comparison of symptoms, which improves follow up and adds the ability of quantifying the severity of neurological symptoms. Especially for clinical trials on effects and side effects of medical treatments in WD, neurological rating scales are invaluable [
27,
28]. This study presents the first systematic comparison of two clinical rating scales previously developed for WD, the UWDRS and the GAS for WD, and evaluates their feasibility in routine clinical practice. Additionally, we introduce the patient reported part of the UWDRS neurological subscale, the “minimal UWDRS”, as a handy and time saving prescreening tool for the assessment of WD patients in routine clinical practice outside scientific trials.
The significant differences in the UWDRS total score and the UWDRS neurological subscore observed in our study show that the results differ depending on the initial mode of manifestation (Table
2). The strong correlation of the UWDRS total score with the UWDRS neurological subscore (Fig.
1a) demonstrates that the UWDRS neurological subscale represents the UWDRS total scale very well. As the UWDRS neurological subscore can total 208 points at maximum, or 65% of the total score, one reason for the strong correlation is that the UWDRS neurological subscore largely represents the UWDRS total score. In contrast, the correlation of the UWDRS total score with the UWDRS hepatic subscore is weak (Fig.
1b). Hence, the UWDRS hepatic scale does not reflect chronic compensated liver disease, which is common in WD [
6]. The moderate correlation of the UWDRS total score with the UWDRS psychiatric score (Fig.
1c) reveals that patients with severe neurological or hepatic symptoms do not necessarily have pronounced concomitant psychiatric symptoms [
14,
29]. More than 20% of the patients in the study were found to have failing memory (item 40), concentration difficulties (item 41), and reduced social contact (item 42), all being psychiatric disabilities frequently seen in patients with WD [
8,
30,
31]. A brief global psychiatric assessment as provided by the UWDRS psychiatric subscale is doubtlessly inappropriate to reflect such disorders. Nevertheless, it can be supportive in identifying patients with WD who would benefit from referral to a psychiatrist for further exploration.
The UWDRS neurological subscore correlates strongly with the GAS for WD Tier 2 score (Fig.
1d). Since the UWDRS scale includes a larger number of neurology and psychiatry related items than the GAS for WD scale, it documents these disabilities in greater detail (with the caveat that GAS for WD includes Kayser-Fleischer rings as a feature, while the UWDRS does not). However, this comes at some cost. In routine clinical practice, the larger number of items in the neurological and psychiatric subscales of the UWDRS requires significantly more time to assess than the Tier 2 score of the GAS for WD. However, this may not be a concern in clinical trials, particularly in those that assess therapeutic effects when detailed documentation of a change in a single neurological symptom is inevitable [
32]. More significantly, many of the detailed neurological signs in the UWDRS scale are difficult to reliably discern in practice, as is reflected in the low interrater reliability (i.e., small values for intraclass correlation coefficients [ICC] for some of the individual items of the scale) [
18]. In contrast, all the individual items of the GAS for WD have ICC values in the high range (i.e., can be assessed reliably) [
19]. Whether a larger number of detailed items or a smaller number of reliably assessable items are of greater benefit will depend on the goal and the design of the individual clinical trial, as well as the consistency in raters. For instance, a trial interested in a specific neurological feature (e.g., autonomic disturbances) that is included only in the UWDRS is probably preferable. Likewise, the GAS for WD may be more suitable for a longitudinal trial with the goal of assessing a change in item scores in individual patients over a long period of time involving many different raters, as it reduces rater induced variance.
The weak correlation of the UWDRS hepatic subscore with the MELD score (Fig.
2a) and the Child-Pugh score (Fig.
2b) demonstrates that the UWDRS hepatic subscore does not mirror the current liver status very well. The correlation of the GAS for WD Tier 1 “liver” domain with the MELD score (Fig.
2c) and the Child-Pugh score (Fig.
2d) is also weak (i.e., this item does not represent the liver status very well either). Furthermore, we could not reproduce the correlation coefficient of
r = 0.65 with respect to the Tier 1 “liver” domain and the Child-Pugh score published by Aggarwal et al. in 2009 (in our study:
r = 0.12). However as the Child-Pugh score, as well as the MELD score to a lower extent, might have a limitation in detecting significant changes of the hepatic function in clinically stable or non-cirrhotic liver patients, the most meaningful mode of evaluation for the hepatic system in a WD specific rating scale has yet to be established.
The “minimal UWDRS” score correlates strongly with the UWDRS total score (Fig.
3a), the UWDRS neurological subscore (Fig.
3b), and the GAS for WD Tier 2 score (Fig.
3c). Outside clinical trials, the “minimal UWDRS” is a convenient and time saving prescreening tool to document neurological impairments in WD patients. The neurological questionnaire can be handed to the patient in the waiting area before the medical appointment without the requirement of a neurologist. Consequently, the questionnaire will help to better structure the consultation as the patient would have already deliberated his actual complaints and would be more prepared for the physician’s questions. Besides, a short questionnaire, likely to be applied more frequently than an extensive neurological assessment, will support self reflection on disease symptoms and thus improve coping strategies as well as adherence to the treatment. Furthermore, the use of a questionnaire is more economical (after an initial detailed assessment), since it can be applied to figure out which impairments should be focused on and reevaluated at future follow up examinations. As seven out of nine items in the “minimal UWDRS” score the level of independence in the ADL, the physician will learn rapidly whether the neurological symptoms limit the patient’s ADL. Constrictively, the questionnaire does not interrogate the exact reasons underlying potential impairments of the ADL. A further limitation is the “minimal UWDRS” may not adequately capture the full breadth of neurological disability in patients with WD. For instance, the “minimal UWDRS” may miss mild limb dystonia or new onset Kayser-Fleischer rings, which can be early signs of noncompliance or inadequate treatment. Thus, the time saving will come at some cost to patient care.
Altogether, our study collective suffered from relatively mild to moderate WD symptoms (Table
2, Additional file
1: Figure S1). Hence, we can not use a single item total score correlation to select single items for a minimal neurological scale. Otherwise, a minimal neurological scale will be inappropriate to be applied for patients with severe WD symptoms. Internal consistencies of the UWDRS total score and its subscales, as well as the internal consistency of the GAS for WD neurological assessment, are all very similar to those reported by Leinweber et al. 2008 and Aggarwal et al. 2009 [
18,
19]. In our study, Cronbach’s alpha for the UWDRS total score is 0.95 compared with 0.92 reported by Leinweber et al., the UWDRS neurological subscale is 0.95 compared with 0.94, the UWDRS hepatic subscale is 0.65 compared with 0.59, and the UWDRS psychiatric subscale is 0.83 compared with 0.76 [
18]. Cronbach’s alpha for the GAS for WD Tier 2 in our study (0.82) and in the first report (0.89) is also similar [
19]. In summary, we largely confirm the internal consistencies for both the UWDRS and the GAS for WD as reported by the previous two articles.
The high prevalence of the hepatic initial mode of manifestation (70.8% in this study vs. an average of 45% in the literature) observed in our study collective may be due to selection bias caused by the fact that all patients were recruited in a WD outpatient clinic in a Department of Gastroenterology. The low median MELD (7.3) and Child-Pugh (5) scores show that most patients suffered from compensated liver cirrhosis. In general, our study collective is only mildly to moderately affected, which is probably due to the median duration of treatment at study inclusion being 15 years (Table
1). Despite the relatively long median time of pretreatment, patients still were clinically impaired. This finding underlines the importance of future studies focusing on improvement of the medication utilized to treat WD.