Carrier Screening for Gaucher Disease: Title and subTitle BreakMore Harm Than Good?

It was a warm summer day more than 30 years ago when my laboratory assistant and collaborator, Wanda Kuhl, showed me some unusual results. Our laboratory had devised a facile method for measuring white blood cell glucocerebrosidase activity and was performing family studies to determine the efficiency of this technique for the detection of heterozygotes for Gaucher disease. But here was a sample from an obligate heterozygote, the mother of a patient with Gaucher disease, in which the range of enzyme activity was as low as in the patient. Bone marrow examination in this 72-year-old asymptomatic parent confirmed that she also had Gaucher disease.¹ Discovery of Gaucher disease in an older, asymptomatic individual was by no means unique—the disease had previously been detected in such persons, including one person who was aged 86 years.² However, many persons homozygous for the N370S mutation associated with Gaucher disease are never diagnosed; based on the discrepancy between the heterozygote frequency and the number of patients diagnosed, an estimated 60% of such homozygous patients may go undiagnosed.³

Herein lies the dilemma faced by Zuckerman and colleagues⁴ in their evaluation of prenatal screening for Gaucher disease reported in this issue of JAMA. Among an estimated 28 893 individuals screened at Israeli genetic centers, 82 couples were at risk for offspring with type 1 Gaucher disease, including 70 (85%) at risk for asymptomatic or mildly affected offspring and 12 (15%) at risk for moderately affected offspring. Following prenatal diagnosis in 68 pregnancies, 16 fetuses with Gaucher disease were detected. Of these pregnancies, 2 of 13 (15%) involving N370S homozygote fetuses and 2 of 3 (67%) involving compound heterozygote fetuses were terminated. Had the mother of my patient with Gaucher disease been conceived in Israel in this century, rather than in the early 1900s, she might never have been born. Her parents would have been informed that the fetus carried by the mother had Gaucher disease, and they might have elected to terminate the pregnancy.

The clinical expression of Gaucher disease, like that of virtually all single-gene diseases, varies enormously.⁵ Physicians remember their patients with severe clinical manifestations—the patients who seek care. Often forgotten are those patients never diagnosed because they have no symptoms of the disorder. In the majority of single-gene genetic diseases, clinicians and genetics researchers have no clue as to the cause of the phenotypic variability, and therefore have no means by which to inform prospective parents whether their child will be severely affected by the disease or will have an asymptomatic life.

Screening of newborns began more than 45 years ago with the introduction of the microbial inhibition test for phenylalanine⁶ for detecting possible phenyketonuria. Providing children with a phenylalanine-free diet in a timely fashion greatly ameliorated the mental retardation that is the clinical hallmark of phenylketonuria. This was soon followed by screening of newborns for galactosemia,⁷ allowing early treatment and prevention of the high incidence of neonatal mortality of infants with galactosemia. Before implementation of screening, infants who survived because they were treated were usually treated too late to avert the development of cataracts, cirrhosis, and mental retardation. Subsequently another approach to screening was introduced, the examination of parents for the carrier state, such as for Tay-Sachs disease, giving them the option after prenatal diagnosis to terminate a pregnancy.⁸

Although there is phenotypic variation in these severe genetic diseases, it is from bad to worse, rather than from bad to normal as with Gaucher disease. For example, galactosemia⁹ and phenylketonuria in adults who have never been treated are rare and are associated with mental retardation.¹⁰ Tay-Sachs disease, once known by the grim name of “amaurotic family idiocy,”¹¹ is currently untreatable and leads to death after the first few years of life. Even these relatively straightforward screening programs have not been without some difficulties. One problem came to light with the discovery that some normal adults lacked serum hexosaminidase activity.^{12 - 13} Although this phenotype was rare, had these individuals been detected during fetal life as a consequence of a screening program, they most likely would have been aborted. Some children who have hyperphenylalanemia caused by mutations of genes other than phenylalanine hydroxylase¹⁴ may not require therapy.

In general, however, cost-benefit analyses for these diseases were quite straightforward. The cost of screening and counseling could be estimated, and the expense associated with each case that was discovered could be determined. Based on these estimates of the potential savings to society, it was generally concluded that screening for these disorders was useful and cost-effective. The reason these programs have been so effective is that the clinical penetrance of the diseases for which screening was performed was very high.

The success of these screening programs has spawned numerous others. In some cases the target disease phenotype has been uniformly severe, but in others, such as Gaucher disease, hereditary hemochromatosis, and glucose-6-phosphate dehydrogenase (G6PD) deficiency, many or most patients do not develop the clinical disease associated with the mutation. There seem to be 2 reasons these programs have proliferated: one economic and the other one altruistic. There is little doubt that the hope or reality that newborn screening will generate profits, either for the medical institution in which the screening is performed or for the vendor who supplies the materials, is one driving force. In some cases companies that profit from the treatment of diseases that are detected also have a motive to encourage screening. The other motivation seems to be based on the philosophy that informing patients about their health and that of their offspring is bound to be a “good thing.”

However, screening for low-penetrance mutations may not necessarily be a “good thing” and is associated with important challenges. Even the most knowledgeable genetic counselor or specialist physician faces a terrible dilemma when trying to explain the probable outcome to a woman carrying a fetus found to have the relatively benign homozygous N370S Gaucher disease genotype. Some of these children may develop splenomegaly, pancytopenias, bone disease, or all 3. Others may live to be 80 years old without encountering any difficulties. Although Zuckerman et al⁴ correctly divide genotypes into asymptomatic or mild and moderate forms, they are quite aware of the fact that there is tremendous overlap in the actual disease phenotypes observed in these groups. For example, we have studied^{15 - 16} a 65-year-old man with the moderate N370S/84GG genotype with a very mild disease severity score of 4, and also have encountered patients who were similarly mildly affected with N370S/L444P genotypes. There are many patients with the asymptomatic or mild homozygous N370S genotype that have much more severe disease at earlier ages.

Attempting to analyze the costs and benefits of a screening program for Gaucher disease is much more complex than doing so for severe, high-penetrance disorders. What is the price for the anguish of parents of a fetus diagnosed to be homozygous for the N370S genotype? To what extent will the psychological development of a child carrying the stigma of this genotype be impaired? Will his parents protect him unnecessarily and not allow him to participate in contact sports when all his friends are on the team? How many millions of dollars will be spent on unnecessary enzyme replacement therapy for children who are fated never to develop clinical disease? Zuckerman et al⁴ point out that the Israeli Medical Geneticists' Association has recommended against Gaucher disease screening, presumably for these reasons. It is apparent that this recommendation has not had much effect in Israel. The countervailing forces enumerated above may be too strong.

Today vast sums of money are being expended to conduct whole genome searches for mutations that modify the risk for disease. Although these efforts will produce some answers and will help in the discovery of modifier genes, the yield will most likely be meager. Although there are probably modifier genes to be discovered, the important answers will probably come from environmental or genetic interactions and, most importantly, from epigenetic differences between genetically identical individuals. This is known to be true not only because of the failure to find important modifying genes in the majority of cases but because identical twins, individuals with identical genomes, show very different clinical courses for Gaucher disease¹⁷ and for other single-gene diseases. Not until clinicians and researchers better understand the factors that determine whether a patient homozygous for the N370S mutation will develop severe disease or none at all will screening for Gaucher disease become useful. Until then, screening for Gaucher disease will likely do more harm than good.