Q J Med 2003; 96: 657-662
© 2003 Association of Physicians
Abdominal malignancies in patients with Wilson’s disease
From the 1Department of Neurology, The Middlesex Hospital, London, UK, 2Department of Internal Medicine, University Hospital, Uppsala, Sweden, 3Department of Histopathology, Norfolk and Norwich Hospital, Norwich, UK, and 4Department of Pathology, Sahlgrenska University Hospital, Göteborg, Sweden
Received 20 February 2003 and in revised form 25 June 2003
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: Wilson’s disease is associated with heavy copper overload, primarily in the liver. Copper is a toxic metal, and might be expected to be associated with cancer induction, as iron is in haemochromatosis. However, liver cancer is currently believed to be extremely rare in this disease, and other intra-abdominal malignancies have not been reported.
Aim: To assess the frequency of abdominal malignant disease in patients with Wilson’s disease on long-term follow-up.
Design: Retrospective study in two specialist Wilson’s disease clinics: Cambridge/London and Uppsala.
Methods: We reviewed the case records of 363 patients seen at three centres: Addenbrooke’s Hospital, Cambridge, 1955–1987; the Middlesex Hospital, London, 1987–2000; and the University Hospital, Uppsala, Sweden, 1966–2002. Patients were grouped by length of follow-up: 10–19 years; 20–29 years; 30–39 years; and 40 years or more.
Results: No cancers were seen in patients followed for <10 years. For patients in the 10–19 years group, the frequency was 4.2%; at 20–29 years, it was 5.3%; and at 30–39 years, 15%. No cancers were seen in the 40+ years follow-up group. The cancers consisted of hepatomas, cholangiocarcinomas, and poorly differentiated adenocarcinomas of undetermined primary site.
Discussion: Patients with Wilson’s disease appear to be vulnerable to the formation of aggressive malignant intra-abdominal tumours during long-term follow-up, irrespective of treatment. Ultrasound scanning of the abdomen seems to be a useful screening procedure.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
While iron overload of the liver is associated with a 200-fold increased risk of developing hepatocellular carcinoma,1 the opposite appears to be the case for copper overload in patients with Wilson’s disease and primary biliary cirrhosis.2,3 Interestingly, the Long-Evans Cinnamon (LEC) rat (the animal model for Wilson’s disease, with a deletion of the 3' end of the rat homologue of ATP7B, the gene that is mutated in Wilson’s disease) also has an increased risk of hepatoma formation.4 It has been hypothesized that the presence of copper in the hepatocyte may actually protect against neoplastic change,5 but it was later pointed out that the rarity of hepatic malignancy in patients with Wilson’s disease might be a consequence of early death in untreated patients and removal of the stimulus to malignant change in treated patients,6 although removal of iron does not necessarily protect the liver in haemochromatosis.7 A patient reported by Madden and associates,8 who presented at the age of 61 with cerebellar ataxia, with a story of symptoms for 26 years before the correct diagnosis was made, and who died before treatment could be started, was found at post-mortem examination to have a hepatocellular carcinoma. By 1997, only 15 patients with Wilson’s disease had been reported as developing hepatoma, and one with cholangiocarcinoma;9 as with Madden’s case of hepatoma, this was a patient with long-standing undiagnosed disease. On the other hand, of the four patients with hepatoma reported by Scheinberg and Sternlieb, three had been taking penicillamine for 1, 4 and 11 years, respectively, but it is not stated how long they had been ill before treatment was started.2 In a series of 87 patients with purely hepatic Wilson’s disease, no patient with malignancy was seen.6 When reporting the long-term effects of treatment with trientine (triethylene tetramine 2HCl) Dahlman and colleagues noted that 2/19 patients developed ‘multifocal cancer including the liver’ in fibrotic but not cirrhotic livers.10 They concluded that these were not hepatocellular carcinomas, but made no comment on their pathogenesis. In a communication to the 8th International Conference on Wilson’s disease and Menkes’ disease, we reported nine patients, including Dahlman’s two, with intra-abdominal malignancies,11 and while this manuscript was in progress, we were informed about two patients from the US who were added to the material, making a total of eleven patient described here. However, the two US patients were omitted from the frequency calculations, since neither of them had been seen in any of our clinics. The calculations are therefore based on two patients with a hepatocellular carcinoma (hepatoma), one with ‘liver cancer’, two with cholangiocarcinoma, and five with poorly differentiated adenocarcinoma of unidentified primary site. The clinical data of the American siblings, one with ‘liver cancer’ and one with cholangiocarcinoma are also given. Where available, gene mutations from the patients are also described.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
This was a retrospective study of patients seen between 1955 and 2002, consisting of two series: one at Addenbrooke’s Hospital, Cambridge (1955–1987) and the Middlesex Hospital London (1988–2000) (n = 310) and another at the University Hospital, Uppsala (1966–2002) (n = 53). During this period, a number of patients were lost to follow-up in the first series. Of the patients reported here, patients 1, 3 and 7 died abroad, and only brief details of the final illness and post-mortem findings are available, although tissue for histological and DNA analysis was made available from patient 3. Patients 1 to 7 were seen in Cambridge and London. Patients 8 and 9 were seen in Uppsala. In the Swedish series, all 53 patients were available to follow-up. Details of two additional patients 10 and 11 from the US were received from patient 11. The patients came from the UK (2), Sweden (2), the US (2), and one each from Egypt, Greece, Iran, Italy and Serbia, so it is clear that malignant change is not confined to any one ethnic group. Two other patients were seen with carcinoma, one of the cervix and one of the lung, but as both had risk factors (sexual promiscuity and smoking, respectively), they are not included in this series. Symptoms of Wilson’s disease largely resolved with treatment in all patients, except for patient 8 who had residual tremor to the end.
Mutation analysis
Mutation analysis was done using manifold sequencing.12 DNA was prepared from whole blood when possible using standard methods, or from formalin-fixed and paraffin-embedded biopsies or slides as described by Ren and colleagues.13
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Mutation analysis
Patient 8 was homozygous for the most prevalent mutation in middle-European populations, His1069Gln, and patient 9 was heterozygous for this mutation, the other being an uncommon mutation, Ile967Phe. Slides were available from the livers of two patients, both with only minute amounts of nucleated cells. Two biopsies were available: one that was preserved in formalin, rendering it impossible to obtain any DNA; and one with almost no tissue left, also impossible to amplify. The first two specimens, from patients 3 and 4, yielded enough DNA to amplify by PCR, but only enough to allow sequencing of those exons that previously have been shown to harbour the most frequently occurring mutations in that patient’s ethnic group. No mutations were found there and no material remained for sequencing of the remaining exons.
Frequency of abdominal malignancy and time of follow-up
It takes several years for carcinoma to develop. The patients were followed for variable times, some being lost to follow-up early while others were seen over many years, the longest period being 45 years. The seven patients in the Middlesex Hospital series came from 310 patients seen between 1955 and 2000; the patients with hepatoma (patients 6 and 7) were followed up for 38 and 31 years, respectively, and the diagnoses of hepatoma were made at the ages of 46 and 42 years. The shortest period seen was 11 years. The two Swedish patients came from a series of patients seen between 1966 and the year 2002. Of the 53 patients, three had been liver transplanted and in these, follow-up time was calculated from the onset of the disease to the time of the transplant. Of the 53 patients, 42 had been followed for 10 years or more. Adding these two series together, of a total of 363 patients, we have 159 who were followed for 10 years or more, excluding the two patients from the US. Nine of these patients developed intra-abdominal malignancies (5.7%). We have taken 10 years as a cut-off point, since the shortest duration of follow-up before the development of cancer was 11 years. These figures are broken down further in Table 1, with the patients from the US omitted since neither of them was seen at any of our clinics. Although the numbers are small, there is a worrying trend in the frequency of the development of cancer for duration of follow-up, with a cut off at 39 years, but it would seem unlikely that patients can consider themselves safe after this time.
|
The incidences of hepatic, biliary and pancreatic cancers seen in Sweden in the year 2000, in patients of the relevant age and sex groups, were 58 per 100 000 (patient 8) and 8 per 100 000 (patient 9) (The Swedish Board of Health and Welfare, http://www.sos.se/sos/statisti.htm). It is difficult to give comparable figures for the Cambridge/London patients, who came from such diverse locations as Egypt, Greece, Serbia, Italy and the UK. However, the relevant incidences are unlikely to differ greatly from the Swedish figures, i.e. much below the incidence in our series.
Clinical presentation
There was no uniformity in the clinical presentation of these patients (Table 2). Two were presymptomatic when diagnosed, in each patient after a sibling had been found to be suffering from Wilson’s disease (patients 6 and 11); two had purely neurological illnesses without evidence of liver disease (patients 1 and 4); six had a predominantly neurological illness but gave a history of an earlier attack of ‘jaundice’, which was probably haemolytic in two (patients 2, 7, 8 and 10; patients 3 and 5); one only had a purely hepatic illness (patient 9) although she was diagnosed by an ophthalmologist, upon the finding of Kayser-Fleischer rings and sun-flower cataracts in conjunction with an appointment for contact lenses.
|
The youngest patient at the time of finding the cancer was 28 years and the oldest 85 years (median 46 years). The shortest interval on treatment was 11 years and the longest 38. Special mention should be made of patient 11, who was diagnosed as presymptomatic Wilson’s disease at the age of 63 years, with macronodular cirrhotic liver, splenomegaly and Kayser-Fleischer rings. Her caeruloplasmin was low, and liver histology showed a positive rhodamine stain for copper. She was treated with penicillamine for 22 years, remained symptom-free and lost her Kayser-Fleischer rings. Her tumour was found coincidentally on an ultrasound scan of the renal tract, and was confirmed on biopsy as a cholangiocarcinoma with a moderate iron excess in the hepatocytes. Her brother (patient 10) had predeceased her, dying of ‘liver cancer’, possibly a cholangiocarcinoma. Patient 6, whose hepatoma was also diagnosed by ultrasound scan, underwent a successful liver transplant.
Antibodies to hepatitis A, B and C were sought, but not found, in patients 5 and 6. Alpha-fetoprotein was determined in five patients (patients 5, 6, 7, 8 and 9) but was in the normal range in all but one: patient 7, one of the two hepatoma patients.
Pathology
In only two patients was the cancer proven to be hepatoma (patients 6 and 7), one with pulmonary secondaries (patient 7). In one there was a diagnosis of ‘liver cancer’, three had cholangiocarcinomas (patients 2, 5 and 11), two with moderate and one with heavy iron overload. The remaining patients all had widespread intra-abdominal cancer of undetermined primary site, except patient 3, in whom the pleura was predominantly involved. Patients 1, 3 and 7 died abroad, and full details of the post-mortem examinations are sketchy, but tissue was received from patient 3 for both histological and DNA studies. The remaining patients had widespread intra-abdominal cancer, for which the primary site could not be determined (patients 1, 3, 4, 8 and 9); essentially these are multifocal, disseminated small, poorly differentiated adenocarcinoma.
As mentioned earlier, the hepatitis viruses were not involved in the two patients studied. Cirrhosis is commonly the basis of hepatoma development, and was present in both patients in this series. The incidence of cirrhosis in the Cambridge/London series was 34% (39 patients). The cirrhotic process was confirmed histologically in 26 patients. A further 13 biopsies showed varying degrees of liver damage, from steatosis with vacuolated glycogen nuclei through inflammatory cell infiltration, to fibrosis. The incidence of cirrhosis in the Swedish patients is not precisely known, but is definitely not greater than that for the UK series. Neither of the two Swedish patients showed any residual signs of liver cirrhosis at autopsy.
Pathogenesis
As the pathology of these tumours is not homogeneous, they may well have different causes, but all occurred in patients with Wilson’s disease, so it seems likely that this is at least a facilitating factor. The treatments given to these patients included most of the recognized therapies: penicillamine, trientine (triethylene tetramine 2HCl), tetrathiomolybdate, and patients 10 and 11 also received zinc, although patient 11 could not tolerate it and did not take it for long. Dimercaprol (BAL) was not used in any of these patients. The longest period on any single drug was 38 years for penicillamine (patient 6) and 18 years for trientine (patient 9). The younger sister (by 3 years) of patient 9 demonstrated psychiatric symptoms some 3 years before the detection of Wilson’s disease in her elder sister, but diagnosis was concomitant, as was the start of identical treatment in the two of them. The sister of patient 9 lives without signs of malignancy 31 years following the onset of her disease. Patient 5 had multiple therapies: penicillamine, trientine, molybdate and again trientine over a course of 23 years. During this time, she was also treated for psoriasis with etretinate, which resulted in acute myeloid leukaemia, successfully treated with cytotoxic drugs approximately 10 years before the development of cholangiocarcinoma. The role of copper remains enigmatic, as no figures are available for the copper content of the livers of any of these patients, except for patient 5 in whom it was 36 µg/g (0.57 µmol/g) wet weight (normal < 10 µg or 0.16 µmol/g) and patient 6, in whom it was 19 µg/g (0.30 µmol/g). As all had been on an ‘anticopper’ regimen for many years, it would seem likely that liver copper concentrations were not high, and any copper present was chelated and not metabolically active.
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
That no large series of liver or intra-abdominal cancers has previously been reported in patients with Wilson’s disease may be simply because there are very few available series of this size and with so long a follow-up (up to 45 years). From our experience, all such patients should be considered vulnerable to malignant change in the liver or abdomen in the long run. In this series, alpha-fetoprotein was not of value, but abdominal ultrasound studies gave a diagnosis before overt symptoms, where used, and this may be a test which should be included in the follow-up of patients with Wilson’s disease. Predominantly hepatic illness does not necessarily predispose to this development: patients 2 and 4 had no liver symptoms at any time until ascites developed in the terminal stages of their illnesses. Patients 3, 5, 7 and 8 had predominantly neurological illness, although all gave a history of an earlier, self-limiting hepatic episode before the development of neurological signs, while patients 6 and 11 were treated presymptomatically and only patient 9 had a purely hepatic course to her disease, although she had Kayser-Fleischer rings and sun-flower cataracts at diagnosis. It is thus impossible to say what precautions might avoid the development of malignancy in these patients. However, those followed for long periods do seem to be at risk of developing intra-abdominal malignancies, and should be monitored accordingly. Some of these cancers may be purely coincidental, as the great majority of our patients may be said to have entered the 'cancer age group', although this seems less likely for those with hepatoma or cholangiocarcinoma. Of the three patients with cholangiocarcinoma for whom data were available, all had moderate to severe iron overload in the liver, but 11 other patients were seen with iron overload who have not developed cancer, so the role of this also remains uncertain, but perhaps should suggest increased vigilance.
|
Acknowledgments |
|---|
Our thanks are due to Dr Andrew Taylor, School of Biomedical and Life Sciences, Trace Element Laboratory, University of Surrey, UK, for the estimations of liver copper in patients 5 and 6.
|
Footnotes |
|---|
Address correspondence to Dr J.M. Walshe, 58 High Street, Hemingford Grey, Huntingdon PE28 9BN. e-mail: penicillamine{at}waitrose.com
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Niederau C, Fischer R, Sonnenberg A, Stremmel W, Trampisch HJ, Strohmeyer G. Survival and causes of death in cirrhotic and in noncirrhotic patients with primary hemochromatosis. N Engl J Med 1985; 313:1256–62.[Abstract]
2. Scheinberg IH, Sternlieb I. Major problems in internal medicine XXIII. Wilson’s disease. Philadelphia, Saunders, 1984.
3. Hoogenraad TU. Major problems in neurology 30. Wilson’s disease. Philadelphia, Saunders, 1996.
4. Yoshida MC, Masuda R, Mori M, et al. New mutation causing hereditary hepatitis in the laboratory rat. J Hered 1987; 78:361–5.
5. Wilkinson ML, Portmann B, Williams R. Wilson’s disease and hepatocellular carcinoma: possible protective role of copper. Gut 1983; 24:767–71.
6. Walshe JM. Wilson’s disease presenting with features of hepatic dysfunction: a clinical analysis of eighty-seven patients. Q J Med 1989; 70:253–63.[Web of Science][Medline]
7. Schafer DF, Sorrell MF. Hepatocellular carcinoma. Lancet 1999; 353:1253–7.[CrossRef][Web of Science][Medline]
8. Madden JW, Ironside JW, Triger DR, Bradshaw JP. An unusual case of Wilson’s disease. Q J Med 1985; 55:63–73.[Web of Science][Medline]
9. Kosminkova EN, Generalova S, Ponomarev AB. The development of diffuse cholangiocarcinoma in a female patient with long-term undiagnosed Wilson’s disease. Ter Arkh 1995; 67:85–7.
10. Dahlman T, Hartvig P, Löfholm M, Nordlinder H, Lööf L, Westermark K. Long-term treatment of Wilson’s disease with triethylene tetramine dihydrochloride (trientine). Q J Med 1995; 88:609–16.
11. Walshe JM, Sams V, Westermark K. Malignancy in Wilson’s disease [abstract]. Z Gastroenterol 2001; 39:248 (P44).
12. Waldenström E, Lagerkvist A, Dahlman T, Westermark K, Landegren U. Efficient detection of mutations in Wilson disease by manifold sequencing. Genomics 1996; 37:303–9.[CrossRef][Web of Science][Medline]
13. Ren ZP, Sällström J, Sundström C, Nister M, Olsson Y. Recovering DNA and optimizing PCR conditions from microdissected formalin-fixed and paraffin-embedded materials. Pathobiology 2000; 68:215–17.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. Huster, M. J. Finegold, C. T. Morgan, J. L. Burkhead, R. Nixon, S. M. Vanderwerf, C. T. Gilliam, and S. Lutsenko Consequences of Copper Accumulation in the Livers of the Atp7b-/- (Wilson Disease Gene) Knockout Mice Am. J. Pathol., February 1, 2006; 168(2): 423 - 434. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. M. Tassabehji, J. W. VanLandingham, and C. W. Levenson Copper Alters the Conformation and Transcriptional Activity of the Tumor Suppressor Protein p53 in Human Hep G2 Cells Experimental Biology and Medicine, November 1, 2005; 230(10): 699 - 708. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||