In Wilson disease, as in most metabolic diseases, effective treatment can evolve despite ignorance of cause as long as some aspects of the condition's pathogenesis are understood. In 1948, Cumings [5] demonstrated excess copper in the brain caused by Wilson disease and suggested chelation therapy. After trials showed that dimercaprol (BAL) and EDTA were difficult to administer and unacceptably toxic, Walshe [6] started a therapeutic revolution in 1953 by proving the efficacy of oral penicillamine in reducing total-body copper stores. Today, D-penicillamine remains the mainstay of therapy [7]; it is particularly effective when therapy is started before signs of organ involvement appear [3, 4]. Therefore, diagnosing the disorder at as early an age as possible is crucial in reducing morbidity and mortality. Because Wilson disease occurs in only about 2 to 3 persons per 100 000 and because no reliable biochemical marker exists in newborn infants, screening has been reserved for persons at high risk for the disease, such as sibs of patients.

Confirmatory diagnostic tests have been problematic for both relatives of probands and patients who manifest only one of the features of Wilson disease. Determining copper levels in tissue, for example, requires a biopsy. Levels of serum copper and ceruloplasmin are often low but not invariably, and both levels can vary greatly among affected sibs. Incorporation of radioactive copper into ceruloplasmin is the best test [3], but this procedure is not routinely available. A test based on the underlying biochemical defect or analysis of the gene itself might prove both reasonably specific and sensitive. Until recently, however, no such test was available.

In 1960, Bearn [8] examined the clinical course of patients with Wilson disease from eastern Europe and other regions. He concluded that genetic heterogeneity accounted for the phenotypic differences among some ethnic groups. Cox and colleagues [9] studied disease characteristics, such as age of onset, serum ceruloplasmin level, and urinary copper excretion, and supported genetic heterogeneity as one cause of variability of expression. One group in their study largely comprised persons of eastern European (that is, Slavic) ethnicity who tended to receive a diagnosis at an older age (in the third or fourth decade of life) and presented with neurologic signs.

As was the case with alkaptonuria [10], it has taken much of the 20th century to achieve even a partial understanding of the pleiotropy and clinical variability of Wilson disease by discovering the defective protein and its genetic locus. Both of these inborn errors were found by using positional cloning. First, the gene was localized to a specific chromosomal region (13q21) by linkage analysis in affected families [11, 12]. The gene was then identified through demonstration that mutations were present in patients and carriers but not in unaffected persons [13, 14].

Molecular genetic investigations confirm that Wilson disease shows considerable genetic heterogeneity: More than 50 different mutations in ATP7B, a relatively large gene [13-15], have been found to cause Wilson disease [16-18]. Although only one gene is involved in Wilson disease, many different mutations cause various defects in its product, a copper-transporting adenosine triphosphatase. Much of the variability in Wilson disease among different families is due to the existence of different alleles and combinations of two alleles (intralocus genetic heterogeneity). Correlations of genotype with phenotype suggest that knowledge of the type of mutation may have prognostic importance. For example, mutations that result in the complete absence of the gene product are associated with development of liver disease at a particularly early age [16]. Many of the alleles are rare, but several have been detected in multiple, apparently unrelated patients. One of these mutant alleles, a single nucleotide alteration that results in substitution of glutamine for histidine at position 1069 (H1069Q, called position 1070 in earlier reports), occurs in almost one third of all patients with Wilson disease who are of European ancestry [16]. Persons who have two copies of this mutation (homozygotes) or one copy plus another mutant allele (compound heterozygotes) develop symptoms in their mid-teens, and either neurologic or hepatic disease may be more evident at presentation. Nonetheless, severity of disease varies considerably even among relatives who have the same genotype involving this mutation. Rapid and accurate detection of this mutation by a polymerase chain reaction-based test should facilitate many important clinical applications and further investigations.

As Maier-Dobersberger and colleagues [19] point out in this issue, relatives of patients found to have the 1069 or any other mutation in ATP7B can be screened at an early age. If a relative is found to have the same genotype as the patient, presymptomatic institution of D-penicillamine therapy can be considered. As a corollary, relatives who do not have two mutant alleles can be reassured that they will not develop Wilson disease; if they are carriers, reproductive genetic counseling should be advocated. In addition, the basis of variability in expression among persons with the H1069Q allele can be studied, for example, by correlating phenotype with dietary copper intake, examining patient exposure to hepatotoxins or viruses, or searching for a modifying gene.

The discovery of such a modifying gene could have several benefits. First, prognosis could be even more accurate because it would be based on a person's genotype at both loci. Second, therapy chosen on the basis of the activity of the product of the modifying gene might prove more effective than correcting the defect in ATP7B. Third, studies of natural history and new therapies (such as zinc or better chelating agents [4, 20]) can be stratified so that patients with specific mutations in ATP7B can be compared with each other and, as a group, with patients who have other mutations.

Heterozygotes are invariably not affected by any features of Wilson disease; given the rarity of carriers (about 1 person in 100), there is little need to identify heterozygotes, except in inbred populations. Nonetheless, possessing one mutant allele at the ATP7B locus could potentially render a person more susceptible to liver, brain, or heart damage from other causes. The Wilson disease locus should become a useful candidate for investigations targeted at understanding the multifactorial causes of many common diseases.

Reed Edwin Pyeritz, MD, PhD

MCP-Hahnemann School of Medicine; Allegheny University of the Health Sciences; Pittsburgh, PA 15212