Genetic heterogeneity and clinical disease.
S. Reeders
Western Journal of Medicine
Abstract:Genetic Heterogeneity and Clinical Disease THE CORRECr CLASSIFICATION of a hereditary disease is critical for the study of its pathogenesis. It is particularly important to evaluate genetic diseases clinically because each distinct clinical subtype may correspond to a particular mutation or limited set of mutations in the human genome. If a simple relationship can be shown between a mutation (the genotype) and the resulting clinical pattern of disease (the phenotype), early prediction of prognosis becomes possible based on analysis of deoxyribonucleic acid (DNA) alone. Moreover, different mutations may suggest different mechanisms of disease pathogenesis, which in turn may allow therapies tailored to each disease subtype. Different mutations may be associated with a phenotype that, at first glance, is uniform (genetic heterogeneity). The mutations may be clustered at a single genetic locus or may even affect different genes. Some diseases are essentially homogeneous: sickle-cell disease in Africa is caused by a single mutation. Other disorders are remarkably heterogeneous-more than 40 mutations have been detected in the CFTR gene to produce cystic fibrosis. Although some of these mutations are more commonly associated with the development of pancreatic insufficiency, many CFIR mutations produce indistinguishable phenotypes. In this issue Gardner reviews the evidence for genetic heterogeneity in autosomal dominant polycystic kidney disease, Alport's syndrome, and medullary cystic kidney disease, three common hereditary renal disorders.' Careful evaluation of the clinical patterns of these disorders shows that none ofthem are homogeneous. Evaluation ofthe pattern of inheritance of the three disorders indicates that they are also genetically inhomogeneous. In the case of autosomal dominant polycystic kidney disease, the mutations in the genes of some families are clearly linked to genetic markers on the short arm of chromosome 16,2 although the mutated gene has not been cloned. In a minority of families, perhaps 5%, the mutation is not found on chromosome 16.3 Initial evaluation suggested that the pattern of disease is the same whether the mutation resides on chromosome 16 or elsewhere. More detailed characterization, however, reveals a striking difference in the age of onset of end-stage renal failure in the two groups.4 The chromosome 16-linked mutations seem to produce a much more severe form ofthe disease with earlier onset and a higher incidence ofhypertension than the non-linked form. Although the overlap in age of onset between the two groups is such that the site of the mutation cannot be used to provide an accurate prediction of the age at which renal replacement therapy will be required, a cautious optimism is warranted in persons who have the nonchromosome 16 form, many of whom will live a normal life span without therapy. Of the inherited renal diseases, Alport's syndrome is the only one for which a molecular basis has been found.' 6 In three kindreds from Utah, Barker and co-workers6 found mutations in the a5(IV) isoform of basement membrane collagen. The gene that encodes this collagen, designated COL4A5, is found on the X chromosome, explaining the sexlinked pattern of inheritance frequently observed in Alport's syndrome pedigrees. The availability of a genetic marker for the COL4A5 gene enables analysis ofthe correlation between the genotype and phenotype in this clinically heterogeneous disease. Sensorineural deafness, lenticonus, and platelet abnormalities, all ofwhich accompany nephropathy in a subset of patients, are prime areas for study. Certain correlations between the clinical pattern of Alport's syndrome and the associated mutation are already becoming apparent. Antignac recently showed that the presence of esophageal leiomyomas in Alport's syndrome correlates with deletions of COL4ASthat extend upstream ofthe gene (C. Antignac, MD, INSERM, Paris, oral communication, August 1991). She postulates that these deletions disrupt a second gene or genes upstream of COMA5 to produce susceptibility to leiomyomatosis. A small fraction of patients with Alport's syndrome develop antiglomerular basement membrane antibodies after heterologous renal transplantation. Kashtan has evidence that a high proportion of such patients have extensive deletions of the COL4AS gene (C. E. Kashtan, MD, University of Minnesota, oral communication, August 1991). Presumably absence ofthe a5(IV) collagen isoform in the recipient of the graft accounts for the failure to develop immunologic tolerance to this molecule. In his review Gardner points out that the advent ofmolecular analysis of genetic disease and the predictive testing that it makes possible have far-reaching ethical implications. The ethical issues in genetic testing are particularly complex for several reasons, but perhaps the most important of these is that the ramifications of a genetic diagnosis go beyond the individual member to the extended family: the diagnosis of polycystic kidney disease in a single at-risk individual often affects dozens of family members, including some who have no close links with the proband. Frequently, even close family members have no idea that they are at risk. The duty ofthe physician in these instances has not been resolved adequately.2 Much depends on the nature of the particular disease in question and the benefits (and disadvantages) of diagnosis. The need for research into the use and ethical implications of testing is apparent. The problem is not, however, simply a result of the advent of molecular analysis. After all, in the case of autosomal dominant polycystic kidney disease, for example, presymptomatic diagnosis by excretory urography has been available for decades. Nevertheless, the problem has been brought into focus by the development of tests based on DNA, which are easily performed and, because oftheir novelty and apparent power, are appealing to patients and their physicians. It is inevitable that their use will grow rapidly. A systematic evaluation of the effects of these tests on individuals and on society as a whole is timely.