Wilson's disease (WD) is an autosomal recessive disorder of copper metabolism leading to copper accumulation in various organs. The most common clinical presentation of WD is hepatic or neuropsychiatric. The spectrum of liver disease encountered in patients with WD is variable.1 Most patients present with chronic hepatitis, which progresses to cirrhosis if untreated. In 20–30% of patients with WD encountered in tertiary centers, the first disease manifestation is fulminant hepatic failure (FHF).2
Fulminant WD is almost invariably fatal, and liver transplantation is the only life-saving treatment, whereas nonfulminant decompensated chronic WD mostly responds well to chelation therapy.3, 4 To facilitate a prompt diagnosis of WD in the setting of FHF, 2 screening indexes, alkaline phosphatase (ALP)/bilirubin <2.0 and aspartate aminotransferase (AST)/alanine aminotransferase (ALT) >4.0, have been proposed.5, 6 Unfortunately, their specificity for WD compared to the urinary copper excretion test was shown to be unsatisfactory in a later study by Sallie et al.6
In parallel, 2 prognostic indexes that predict the risk of a fatal outcome without orthotopic liver transplantation (OLT) in patients with fulminant or decompensated chronic WD have been published. First, the WD prognostic index (WPI) was proposed in 1986 by Nazer et al.7 The severity of hepatic dysfunction on admission, evidenced by the prolongation in prothrombin time, AST activity, and serum bilirubin, was graded, and an index based on the sum of these scores was calculated. Seventeen patients with a score of 6 or less survived on chelation therapy, whereas 18 patients with a score of 7 or more died. When prospectively evaluated in a further 9 cases, the index correctly predicted the outcome. However, these prognostic criteria have not been validated in more studies. Recently, Dhawan et al.8 reviewed the WPI in 32 patients with decompensated chronic WD and 25 patients with fulminant WD. Five children with a score ≥7 survived, and 4 with a score <7 died, compromising the performance of the test to predict the outcome. In view of the limitations of the WPI, the revised WD prognostic index (RWPI), calculated from a combination of AST, albumin, bilirubin, international normalized ratio, and white cell count, has been proposed at King's College London, UK.8 The RWPI ≥11 had a sensitivity, specificity, positive and negative predictive values above 90% for prediction of death if liver transplantation was not performed. All children who received a liver transplant had a RWPI ≥11, and none of the patients who improved on chelation therapy exceeded this value. The high discriminative power of the RWPI with a cutoff value of 11 was confirmed in a subsequent prospective evaluation of pediatric cases.
Following the introduction of the model for end-stage liver disease (MELD) allocation system9 in 2004, it has been shown that the MELD score may also serve as a suitable predictor of the outcome in patients with FHF.10, 11 A MELD score ≥30 is accepted as indication for urgent listing for liver transplantation in patients with fulminant liver failure. However, the MELD scoring system has not been applied to a representative group of patients with chronic decompensated WD.
In our study, we validated the WPI, RWPI, and MELD scoring systems in adult patients with chronic decompensated WD treated in our institute between 1995 and 2005. In addition, we reviewed the versatility of the 2 biochemical indexes designed for rapid diagnosis of fulminant WD.
WD, Wilson's disease; FHF, fulminant hepatic failure; ALP, alkaline phosphatase; AST, aspartate aminotransferase; ALT, alanine aminotransferase; OLT, orthotopic liver transplantation; WPI, WD prognostic index; RWPI, revised WD prognostic index (Revised King's College score); MELD, model for end-stage liver disease.
PATIENTS AND METHODS
Twenty-one patients with fulminant WD (group 1), 14 patients with chronic decompensated WD (group 2), and 19 patients with other causes of FHF (group 3), admitted from March 1995 to December 2005, were analyzed in a retrospective manner.
Fulminant WD (group 1) was defined as FHF at the first presentation of WD, severe coagulopathy (international normalized ratio >2), and either encephalopathy or hemolytic anemia. Decompensated chronic WD (group 2) was defined as known chronic liver disease with sudden onset of jaundice with liver synthetic dysfunction with neither encephalopathy nor hemolytic anemia. The reasons for decompensation were withdrawal of D-penicillamine treatment or unknown etiology of liver disease before admission. The diagnosis of WD was based on the presence of liver disease and the following laboratory criteria: serum ceruloplasmin <0.2 g/L (normal range, 0.2–0.6 g/L) and/or urinary copper excretion >100 μg/24 hours together with either increased hepatic copper >250 μg/g dry liver or presence of pathogenic mutation on both alleles of the WD gene. Mutations found in our patients are presented in Table 1. Details of mutation analysis protocol have recently been reported in a study by Vrabelova et al.12 In patients with WD (groups 1 and 2), other causes of liver disease, namely autoimmune and cholestatic liver diseases, viral hepatitis, α1-antitrypsin deficiency, and other metabolic liver disorders were excluded by appropriate investigations. FHF of other etiology than WD (group 3) was diagnosed according to the King's College criteria.13
Table 1. Mutations in ATP7B Gene Found in Patients With Fulminant WD (Group 1) and Chronic Decompensated WD (Group 2)
Genotype (mutation 1/mutation 2)
Group 1 (n = 21)
Group 2 (n = 14)
NOTE: Three patients who died shortly after admission and 1 patient with fulminant WD were not investigated. — indicates that no patient with the particular genotype was found in the respective group.
To assess the differences of continuous variables between groups (age, laboratory values, and copper metabolism parameters), the 1-way ANOVA or the Kruskall-Wallis test, when appropriate, was employed. To compare the accuracy of the 3 prognostic indexes as predictors of survival on chelation therapy for patients with decompensated chronic WD (group 2), the concordance c statistics (area under the receiver operating characteristic curve) were calculated. The c statistics may vary from 0 to 1, with 1 indicating perfect discrimination and 0.5 indicating what is expected by chance alone. For this analysis, fulminant WD patients (group 1) known to have 100% mortality on medical treatment were used as a reference group.
The c statistics method was also employed for comparison of the ALP/bilirubin and AST/ALT indexes for diagnosis of fulminant WD (group 1). Patients with FHF of other etiology (group 3) served as a reference group. Sensitivity, specificity, and positive and negative predictive values of the prognostic and diagnostic indexes were calculated. For correlation analysis of the prognostic indexes, Pearson correlation coefficients were computed. The SPSS statistical software system version 14.0 (SPSS, Inc., Chicago, IL) was used to carry out the calculations. A P value < 0.05 was considered to indicate statistical significance.
Clinical Presentation and Laboratory Data
Patients with fulminant WD (group 1)
The median age of 21 patients with fulminant WD was 18 (range, 15–48) years, and only 1 patient was male. The mean time interval between the first manifestation and admission was 20.6 (range, 2–58) days. On admission, jaundice was present in all patients, ascites in 13 (61.9%) patients, and encephalopathy in 11 (52.3%) patients. Renal insufficiency was present in 7 (33.3%) patients. Clinical and laboratory data on admission are shown in Table 2. Two (9.5%) of the 21 patients had Kayser-Fleischer ring, 16 (76.2%) had Coombs' negative hemolytic anemia, 5 (23.8%) had decreased ceruloplasmin concentration, and 7 (33.3%) had abnormal values of serum copper. Elevated urinary copper excretion was detected in 20 investigated patients; in the other patient, the estimation of urinary copper excretion prior to OLT could not be performed due to anuria. Liver copper content >250 μg/g dry weight was determined in 13 available patients from liver explant or autopsy specimen. Data on histological examination of the liver were available for 13 patients. Cirrhosis was present in 11 patients and severe fibrosis was present in 2 patients.
Table 2. Clinical and Laboratory Data of Patients With Fulminant WD (Group 1), Chronic Decompensated WD (Group 2), and FHF of Other Etiology (Group 3)
The median age of 14 patients that presented with decompensated chronic WD was 33.5 (range, 20–56) years (Table 2). Eight (57.2%) patients were male. All patients had liver cirrhosis, and 4 of them (28.5%) had Kayser-Fleischer ring. Seven out of 8 patients investigated for serum copper and all 9 patients with documented serum ceruloplasmin concentration had abnormal values. Data on urinary copper excretion and liver copper content were available for 9 and 5 patients, respectively. All 5 patients with elevated liver copper content had increased urinary copper excretion. The mean urinary copper excretion was significantly lower than in patients with fulminant WD (P < 0.05). The mean concentration of serum ceruloplasmin was higher in group 1 (P < 0.05, Table 2). Treatment with D-penicillamine (300 mg, 3 times a day) was initiated in all 14 patients.
Patients with FHF of other etiology (group 3)
In the same period, 19 patients (10 females) were admitted with FHF caused by acute hepatitis B (10 patients), autoimmune hepatitis (5 patients), acute hepatitis C (2 patients), fulminant Budd-Chiari syndrome (1 patient), and rapidly progressing cholangiocarcinoma (1 patient). After admission, all patients except for the subject with cholangiocarcinoma were listed for OLT.
Of the 21 patients with fulminant WD, 4 died a median of 1.5 (range, 0–15) days after being listed for OLT. Seventeen patients proceeded to OLT after a median of 5 (range, 0–27) days. Five transplanted patients died in the early posttransplant period (median, 16 days; range, 1–29 days). The 1-year survival in the subgroup of transplanted patients was 70.6%. In contrast, no patient with decompensated chronic WD on chelation therapy required urgent liver transplantation. Six patients on D-penicillamine therapy progressed to end-stage liver disease and had to be electively transplanted 3.7 (range, 1-8) years after treatment initiation, with no reported death. The remaining 8 patients are clinically well and have not progressed to end-stage cirrhosis during a mean follow-up time of 5.8 (range, 1–10) years.
Validation of Prognostic Scores
In the original publication of Nazer et al.,7 a WPI cutoff value of 7 or higher indicated a high risk of death in patients with fulminant WD provided that liver transplantation was not performed. In our series of 14 adult patients with decompensated chronic WD, 13 patients had WPI <7 and 1 patient had WPI equal to 7. Nineteen patients with fulminant WD reached or exceeded the cutoff; however, 2 patients had WPI <7 (Fig. 1A). The discrimination power between both groups was characterized by sensitivity of 86.7%, specificity of 95.0%, positive predictive value of 92.8%, and negative predictive value of 90.4%.
When evaluated according to the system proposed by Dhawan et al.,8 13 patients with chronic decompensated WD had RWPI <11, and 1 patient had RWPI equal to the proposed cutoff limit of 11. In contrast, all but 1 patient with fulminant WD reached or exceeded this threshold at presentation (Fig. 1B). The sensitivity of RWPI <11 for prediction of response to chelation therapy in patients with decompensated chronic WD (group 2) was 92.8%. The calculated specificity was 95.2%, positive predictive value was 92.8%, and negative predictive value was 90.4%.
None of our 14 patients with decompensated chronic WD exceeded a MELD score cutoff point of 30. However, only 13 patients with fulminant WD had a score ≥30, whereas 8 patients remained below the MELD limit for urgent liver transplantation (Fig. 1C), rendering the score 63.6% sensitive and 100% specific. The observed positive and negative predictive values were 100% and 61.9%, respectively.
Following calculations with receiver operating characteristic curves (Fig. 1D, Table 3), the RWPI showed the highest overall performance, followed by WPI and MELD. The WPI showed a strong correlation with RWPI, contrary to a weaker correlation with MELD (Fig. 2 and Table 3).
Table 3. Accuracy of Prognostic Scoring Systems for Prediction of Response to Chelation Therapy in Patients With Decompensated Chronic WD and Their Correlation
95% CI for AUC
Correlation with RWPI (r)
AUC, area under the curve; CI, confidence interval; r, Pearson correlation coefficient.
Validation of Diagnostic Indexes
In all patients with fulminant WD (group 1), the ALT activity was only moderately raised. The same was true for ALP activity. Both values significantly differed from patients with FHF of other etiology (group 3, P < 0.05). A significant difference between the mean values of ALP/bilirubin and AST/ALT ratios was observed between group 1 and group 3 (Table 2). The sensitivity and specificity for the cutoff ALP/bilirubin <2 were 42.9% and 100%, respectively. The values of the same parameters for the cutoff AST/ALT >4 were 38.1% and 94.7%, respectively. The accuracy of the ALP/bilirubin and AST/ALT indexes reached 0.877 ± 0.054 and 0.862 ± 0.067, respectively.
Adult patients with fulminant WD tend to be younger and of female gender. All but 1 of the patients with fulminant WD included in our study were female. This observation is in line with other human studies conducted in adult patients with WD3, 14 and with animal studies.15 The difference between the high prevalence of females after the onset of puberty and equal prevalence of both genders in childhood reported by Dhawan et al.8 supports the role of sex hormones in triggering fulminant WD.
WD may be clinically silent for years, even in the presence of substantial organ damage, and most patients acutely presenting with WD have underlying chronic liver disease— either cirrhosis or chronic active hepatitis with fibrosis.4 Our finding that 11 of the 13 patients with fulminant WD in whom the liver histology was available had liver cirrhosis and the remaining 2 patients had liver fibrosis, together with the results of Durand et al.,4 indicate that acute liver failure without preexisting chronic liver damage is exceptional.
Liver transplantation is irreplaceable in the management of both FHF and end-stage liver disease of any etiology. The indication for OLT in patients presenting with fulminant WD is straightforward, since chelation therapy proved no benefit in fulminant setting. In contrast, the decision when to list a patient with decompensated chronic WD without encephalopathy for transplant is difficult because chelation therapy can rescue most of them, avoiding the complications of urgent transplantation and saving resources. To predict the outcome of fulminant and decompensated chronic WD patients, Nazer et al.7 and, later, Dhawan et al.8 developed prognostic indexes to be used upon admission. The former index was evaluated in a mixed population of young adults and children (age range, 6–33 years), and the latter was validated only in children (age limit, 17.9 years). Similarly, the MELD score11 has been shown to have a prognostic value in patients with FHF regardless of etiology. Neither of the scoring systems has been validated in a group of adult patients with decompensated chronic WD. All of our patients with decompensated chronic WD had a MELD score <30, WPI ≤7, and RWPI ≤11, none of them required urgent liver transplantation, and all improved on chelation therapy. This indicates that all 3 scoring systems reliably predict favorable outcome, provided that the cutoff limits of 30, 7, and 11 for MELD score, WPI and RWPI, respectively, are not exceeded. Our data based exclusively on patients in group 2 did not provide enough information to address the setting of the cutoff values. In fact, clinical signs and laboratory findings of fulminant WD and chronic decompensated WD may not differ upon admission. The scoring systems should assist predicting the clinical course on medical treatment as reliably as possible. This concept encouraged us to compare the MELD score, WPI, and RWPI between groups of fulminant WD and chronic decompensated WD. The MELD score demonstrated the worst performance because of high false negativity for more than one-third of our patients with fulminant WD (Fig. 1C). On the contrary, there was almost no overlap in RWPI values between patients with fulminant WD and with chronic decompensated WD (Fig 1B). Our data confirm that RWPI is the most reliable predictor of response to medical therapy, and its use can be extended for adult patients with chronic decompensated WD.
Establishing the diagnosis of WD in the setting of FHF is critical, because there is no role for conventional medical therapy, and prompt liver transplantation is the only life-saving therapeutic option. In line with the published results,16 1-year survival in patients urgently transplanted for fulminant WD was 70.6%, contrary to zero survival in nontransplanted patients. A straightforward algorithm for a prompt diagnosis of WD is based on the presence of Kayser-Fleischer ring and decreased serum ceruloplasmin.2, 17 Unfortunately, Kayser-Fleischer ring may be absent in patients with hepatic WD, and ceruloplasmin as an acute phase reactant paradoxically may be elevated.18 The only useful marker indicative of fulminant WD is the urinary copper that, in accordance with our data, invariably exceeds 100 μg/24 hours.19 Although some reports18, 20 state that elevated urinary copper is not useful in distinguishing fulminant WD from other causes of hepatic failure, we believe that urinary copper excretion is the most reliable diagnostic marker, since liver biopsy with the measurement of tissue copper is frequently precluded by the bleeding propensity and typically is not available in timely fashion. The turnaround time of this examination is usually several days. Therefore, surrogate serologic markers that would enable a rapid diagnosis of fulminant WD were proposed. The ALP/bilirubin index was based on the observation that in the course of fulminant WD, the ALP activity declines and concentration of bilirubin rises.21 The AST/ALT index reflects the concept of significant mitochondrial injury in fulminant WD.22 The ratio ALP/bilirubin <2.0 and AST/ALT >4.0 was shown to discriminate WD from other causes of FHF in adults.5 In our patients, both diagnostic indexes had a high false negativity resulting in an unacceptably low sensitivity. Therefore, we confirm the finding of Sallie et al.6 that the diagnostic indexes are useless for clinical practice.
In conclusion, the results of our study indicate that adult patients with chronic decompensated WD with the Revised King's College Score for (RWPI), up to 11 are likely to improve on medical treatment with D-penicillamine. In contrast, simple diagnostic indexes proved unacceptably low accuracy and should not be used for rapid diagnosis of WD in the setting of FHF.