College of American Pathologists
Printable Version

  Carrier screening for SMA—why and how it’s done


CAP Today




July 2009
Feature Story

William Check, PhD

It is next to impossible to protect yourself from a threat you don’t even know exists. For instance, it wasn’t long ago that most of us weren’t aware of the danger of identity theft. Then, through bitter personal experience or news reports, knowledge of this form of financial fraud spread. What’s true for financial danger is equally true for danger from genetic diseases. Late last year the American College of Medical Genetics published a practice guideline on spinal muscular atrophy, or SMA, recommending carrier screening for this genetic disorder (Prior TW, for the Professional Practice and Guidelines Committee. Genet Med. 2008; 10: 840–842). As a first step, the guideline said:

“Because SMA is present in all populations, carrier testing should be offered to all couples regardless of race or ethnicity. Ideally, the testing should be offered before conception or early in pregnancy. The primary goal is to allow carriers to make informed reproductive choices.”

“We recommended carrier screening for spinal muscular atrophy for all persons of reproductive age or planning reproduction, just as we do for cystic fibrosis,” says Thomas W. Prior, PhD, the primary author of the ACMG guideline. However, unlike when carrier screening was recommended for cystic fibrosis, the recommendation for SMA screening caught many medical professionals, including many molecular pathologists, by surprise. “The most important thing about carrier screening for SMA is that lots of people have never heard of it,” says Dr. Prior, professor of pathology and neurology and director of molecular pathology at Ohio State University, Columbus. Lack of awareness of SMA is surprising, considering that this condition represents the second most common lethal autosomal recessive disorder after cystic fibrosis (about one in 10,000 live births for SMA versus one in 2,500 for CF) and is the leading cause of death from genetic disease in infancy.

In practical terms, the low level of knowledge about SMA means that professionals and the general population must be educated. And it makes it imperative for the latter to have genetic counseling and effective educational materials. “We cannot do carrier testing without genetic counseling,” Dr. Prior says. “If we don’t have an educational component and genetic counseling, carrier screening programs will fail.” Dr. Prior describes as “beautiful” the educational materials the Claire Altman Heine Foundation has produced.

To Dr. Prior, ignorance of the feasibility of screening for SMA is as surprising as ignorance of the disease itself. “The gene responsible for SMA was not identified yesterday,” he says. The Survival of Motor Neuron, or SMN, gene was discovered in 1995 by Judith Melki, MD, PhD, and colleagues working at Inserm in Paris (Lefebvre S, et al. Cell. 1995;80:155–165). Homozygous deletion of a large segment of the SMN gene underlies SMA. Before this discovery, diagnosis of SMA depended on linkage analysis and muscle biopsy. After the discovery of SMN, Dr. Prior says, “Testing for SMA improved dramatically. We went from a linkage test to a direct deletion test. My lab and others have been offering diagnostic genetic testing since 1995,” he says, “so we have a lot of experience with it.” There has been a CAP Survey for the homozygous deletion assay for at least 10 years, Dr. Prior notes. He started offering carrier testing for SMA in 1997. “Between 95 percent and 98 percent of affected children have a homozygous deletion of SMN,” Dr. Prior says. “So a gene dosage test picks up 95 percent of carriers as having a single copy of the SMN gene.” About one in 40 persons is a carrier, about half the rate for CF. “Unlike CF, SMA has more of a panethnic distribution,” Dr. Prior says.

SMA is a severe neurodegener ative disease in which loss of motor neurons in the anterior horn of the spinal cord causes progressive weakness of the axial and limb muscles, leading to paralysis. “No drug is able to cure it and there is no treatment to date, although physical therapy can improve the patient’s quality of life,” says Dr. Melki, who is now head of the Department of Human Genetics at Hadassah University Hospital, Jerusalem. “There are ongoing clinical trials in the U.S. and Europe with the hope that a drug will improve patient status. We should have some results in one to two years.” Carrier screening is justified, she says, by the severity of the type I form of the disease, the incidence of the disease, and because today there is no therapy.

Type I is the most severe form of SMA and accounts for 60 percent of cases. Onset occurs within the first six months of life; death usually occurs by two years, typically of respiratory failure. Patients never sit up or walk.

Individuals with types II and III SMA survive beyond childhood. Patients with type II can sit up and walk, but many later lose the ability to walk. Type III is the mildest form of the disease. “Patients can walk without any support but can’t run or walk long distances,” Dr. Melki says. “It is difficult for them to go to the market, for instance. They require social integration, but are bright and clever.” Some patients with type III SMA experience progressive motor decline, while others are stable. “We cannot predict who will develop a progressive or stable disease,” Dr. Melki says.

Given the broad clinical heterogeneity of SMA, one of the most striking findings by Dr. Melki and her colleagues was that homozygous absence of a segment of the SMN1 gene causes all types of the disease (Lefebvre S, et al. Cell. 1995;80:155–165). “We found that one gene was responsible for all three clinical forms,” Dr. Melki says. “You can detect a deletion or conversion of the SMN1 gene in almost 100 percent of SMA patients.” More exactly, homozygous absence of exons 7 and 8 of the SMN1 gene affects 95 percent of SMA patients. Intellectually, this was an elegant solution, since it had previously been thought that different genetic mutations would be found in the different types. Such is the power of genetics that deletion or conversion of the SMN1 gene underlies even a rare adultonset form of SMA (Brahe C, et al. Lancet. 1995; 346: 741–742). (Upon discovery of a second gene highly homologous to SMN in the same region of chromosome 5, the original SMN gene was named SMN1 and the copy gene SMN2.)

Genetic homogeneity greatly simplifies diagnosis. “When a patient is suspected of having SMA, the doctor can ask for an SMN1 gene test,” Dr. Melki says. “If there is a lack of SMN1 exon 7, you can say this is SMA. You have confirmed the diagnosis.” Before the SMN1 deletion was discovered, diagnosis relied on muscle biopsy. “Now if there is a high suspicion, you don’t anymore need muscle biopsy,” she says. “You can do a blood test and get the answer very quickly.” The same is true for carrier detection, except that you are looking for the presence of one SMN1 allele, rather than zero.

Richard S. Finkel, MD, is a neurologist who specializes in children with inherited muscle disorders. “Duchenne muscular dystrophy and spinal muscular atrophy are the two biggest conditions we see in our clinic,” says Dr. Finkel, clinical professor of neurology and director of the Neuromuscular Program at The Children’s Hospital of Philadelphia. “There is no cure today for SMA,” he says. “The best we can offer is supportive management.” Without supportive care, type I infants often do not make it to their first birthdays. With loss of respiratory muscles they can die of respiratory infections. They also have severe feeding and malnourishment problems.

“These children really do need multidisciplinary care,” Dr. Finkel says. With him on his team are two other expert physicians, three physical therapists, two nutritionists, a respiratory therapist, occupational therapists, a social worker, a genetic counselor, and nursing support. “We also use pulmonary physicians and surgeons sometimes to place a feeding tube,” he says.

Parents of type I infants face a difficult struggle. “They can choose to be very proactive with breathing and feeding support or let things take more or less the natural course,” Dr. Finkel says. Many parents elect the latter alternative for type I babies, especially after seeing their babies grow progressively weaker. About half of families elect “comfort care” over proactive supportive care.

Type II SMA patients can live stably into their 20s and beyond. Still, Dr. Finkel says, “They require breathing and feeding support and a lot of management.” Type III patients are able to walk, but about half lose that ability by their 10th birthdays.

Even with this broad spectrum of disease from the clinical perspective, nearly all SMA patients are diagnosed correctly by the direct genetic test. “I don’t know of any false-positive cases incorrectly diagnosed with SMA,” Dr. Finkel says. ”It is a very specific test if performed correctly.”

Both parents of an affected child with SMA carrying a homozygous absence of SMN1 are presumed obligate carriers, though new mutations in SMN1 in the gametes (egg and sperm cells) can arise (about two percent of cases) where the risk of having a second affected child is much lower. “At the next pregnancy,” Dr. Melki says, “we can propose prenatal genetic testing if the parents are interested.” Testing of a sample obtained by chorionic villus sampling at 10 to 12 weeks can establish the fetus’ genetic status. “If there is homozygous absence of the SMN1 gene, you can be sure the newborn will develop the disease,” Dr. Melki says. “In contrast, if the fetus has at least one normal copy of SMN1, you can be sure it will not have the disease.” Use of chorionic villus sampling allows parents an early opportunity to terminate an affected fetus.

Another way to avoid a second affected child is through in vitro fertilization and preimplantation genetic diagnosis, or PGD. Genetic testing is done on one cell of the embryo at the six- to eight-cell stage. Only healthy embryos are transferred to the mother. With this alternative approach, “great progress” has been made in preventing SMA, Dr. Melki says (Daniels G, et al. Mol Hum Reprod. 2001;7:995–1000). “PGD is usually accepted by religious people,” she says, and it is appropriate for couples unwilling to terminate an affected fetus. The process takes several months and is not easy for the parents.

In Israel carrier screening for prospective parents is already established, upon the Ministry of Health’s recommendation. “We can reduce the incidence of disease right now by screening at the beginning of pregnancy,” Dr. Melki says. “We are not doing premarital or pre-pregnancy screening.” At the beginning of pregnancy the obstetrician will propose to the mother that she be tested to see if she is a carrier. If the mother agrees, the physician proposes testing the father. Genetic counseling is provided before and after testing. If both parents are carriers, the counselor explains the risk of having a child with SMA and proposes testing the fetus.

“One problem is that there can be false-negatives. We can miss some carriers,” Dr. Melki says. “We demonstrated that the most common mutation [in the SMN1 gene] is deletion or conversion of SMN1 exon 7, leading to reduced number of SMN1 copies. Instead of two alleles you have zero or one.” Detecting zero or two gene doses is easy. “When you look for one it is more technically difficult,” she says. Almost all people have one copy of the SMN1 gene on each copy of chromosome 5. “But a few people have two copies of SMN1 on one chromosome and either zero or one on the other,” Dr. Melki says. When the test shows two SMN1 genes, that person most likely has one copy of the gene on each chromosome and is not a carrier. However, he or she could also have two SMN1 genes on one chromosome and zero on the other, in which case the person would be capable of transferring a copy of chromosome 5 with no SMN1 gene and thus be a carrier. Dr. Melki estimates that the false-negative rate due to having two copies of the SMN1 gene on one chromosome and zero on the other is approximately one in 1,000.

Another drawback: “We cannot predict the level of severity of the disease” based only on homozygous deletion or conversion of SMN1 exon 7, Dr. Melki says. It may be possible to predict severity based on the evidence of a correlation between the SMN2 gene copy number and the clinical expression of the disease. However, this correlation is not absolute. “If there are only one or two copies of the SMN2 gene, disease is usually severe,” she says.

In the American College of Medical Genetics recommendation, Dr. Prior lists the five standard criteria that must be met for carrier screening to be acceptable. It must be for a disease that has a major impact on health and a high frequency carrier rate. There must be a technically feasible assay with which to screen, and screening must be voluntary. And for couples who test positive, there must be reproductive options.

“Those are pretty much the same criteria we set for CF,” Dr. Prior says. “SMA clearly meets those criteria.” Of the diagnostic test, he says, “It’s pretty straightforward. Fewer labs are involved in carrier testing since it is a bit more difficult technically.” (Laboratories that do SMN1 testing can be found at GeneTests)

He identifies two analytical challenges with the gene dosage test. The first is the rare complication Dr. Melki described—a false-negative carrier test in a person with two copies of SMN1 on one chromosome and a deletion on the other chromosome.

A second challenge results from the fact that two percent to five percent of those with SMA are compound heterozygotes for the SMN1 gene deletion. In these people, SMN1 is deleted on one chromosome while the gene on the other chromosome has a point mutation. In almost every instance, one parent of such a child must also have that point mutation in one SMN1 gene. (Among SMA patients, two percent have one de novo mutation, meaning only one parent is a carrier.) This parent would have a false-negative result for carrier status. “If we don’t see a homozygous deletion in an affected child, we sequence the SMN1 gene looking for a point mutation,” Dr. Prior says.

“Gene dosage is not a perfect test,” he adds, “but we recognize its limitations and move ahead.

“Some people probably feel we need more data” before instituting carrier screening, Dr. Prior acknowledges. “But we are losing these kids and things are not getting any better. Certainly prevention is important, just like in CF. I agree we need more educational materials, more genetic counselors, and pilot programs like the ones we had in CF,” he says. “My point is, how much longer do we have to wait? How much longer do kids have to die?”

Dr. Prior advocates even more extensive use of the SMN1 deletion test, saying, “We ought to think about newborn screening, too.” In addition to being able to begin supportive treatment earlier, he points out, “if we could diagnose early, we could put young children into clinical trials while they still have viable motor neurons.” In a feasibility study, Dr. Prior demonstrated 100 percent sensitivity and specificity of newborn screening for SMA using real-time PCR (Pyatt RE, Prior TW. Genet Med. 2006;8:428–437).

Newborn screening can lead to difficult decisions for parents, Dr. Finkel suggests. He raises the same conundrum Dr. Melki noted: A positive SMA genetic test doesn’t say which type of the condition a baby will have. “Therein lies an ethical dilemma,” Dr. Finkel says. “You can diagnose patients pre-symptomatically, because very few patients are symptomatic at birth. Is it ethical to diagnose a progressive disorder for which there is no effective treatment when it is still pre-symptomatic? You could be picking up a child who won’t develop symptoms until adolescence and be a mild type III. The parents could live for 10 years with the notion that their child has a horrible fatal disease. How will that affect how they treat that child?”

On the other hand, early knowledge can mean therapy begins sooner. “You don’t have to have a cure to be able to provide therapy,” Dr. Finkel says. “Therapy is defined as anything that alleviates pain and suffering and improves the condition.” If parents and physicians know that a child will develop SMA, they can monitor for motor problems and intervene more quickly. Benefit in type I is “modest,” he says, “but it is still there.”

He lists another benefit to early diagnosis: If a child is found to have the disease, genetic counseling can be provided to the parents before they have another child. “The main issue to me,” Dr. Finkel says, “as difficult as it is to tell parents that their child has SMA, at least they can make informed reproductive decisions.”

A technical advance may resolve one conundrum of newborn screening—predicting severity in those with a homozygous deletion of SMN1. Accumulating evidence supports the notion that SMN2 gene copy number correlates with SMA type (Feldkotter M, et al. Am J Hum Genet. 2002;70: 358–368; Mailman MD, et al. Genet Med. 2002;4:20–26; Harada Y, et al. J Neurol. 2002; 249: 1211– 1219; Wirth B, et al. Hum Genet. 2006;119: Epub March 1, 2006). Dr. Prior described three asymptomatic individuals with family histories of SMA and homozygous SMN1 deletion who had increased SMN2 copy numbers (Prior TW, et al. Am J Med Genet A. 2004; 130A: 307–310). Summarizing this evidence, Dr. Finkel says, “Individuals with one or two copies of SMN2 will most likely be type I, those with four copies will probably be type III, and those with copy number five will almost certainly be type III or IV (adult onset).” He calls copy number three “a bit murky, with the majority of patients being type II but some being type I or III. Thus, interpretation of the relevance of the SMN2 copy number in an individual patient must be undertaken with caution.” All of which leads him to ask: If a child has a homozygous deletion of SMN1 on newborn screening, should you then measure SMN2 copy number?

“This is an evolving area. It is important that we begin to address these issues,” Dr. Finkel says.

When Dr. Melki and colleagues discovered the SMN gene, Inserm patented it. Subsequently, Inserm sold the rights to Athena Diagnostics, which offers diagnostic, prenatal, and carrier testing. “In the beginning we got samples mostly from parents who had an affected child,” says Sat Dev Batish, PhD, FACMG, D(ABMG), senior director of Athena’s reference laboratory in Worcester, Mass. “More recently we are getting samples for carrier testing from parents who have not yet had a child with SMA.”

Dr. Batish measures SMN1 with a multiple ligation-dependent probe amplification kit from MRC Holland, which, he says, “gives us a good look at both SMN1 and SMN2 genes” (Scarciolla O, et al. Neurogenetics. 2006;7:Epub July 22, 2006; Huang C-H, et al. Genet Med. 2007;9:241–248; MRC Holland: “It’s very important to know the SMN2 copy number for the purpose of diagnosis and prognosis,” Dr. Batish says. In a survey conducted at an ACMG meeting, almost 80 percent of clinical geneticists wanted to know SMN2 copy number and its importance as part of a carrier test report, Dr. Batish says. He recommends sequencing the entire SMN1 gene of patients with a single copy of SMN1 and of the second parent if the first parent was identified to be a carrier of a single copy of SMN1 gene.

Genzyme Genetics licenses from Athena the right to perform SMN1 deletion testing. “Before the discovery of the SMN gene, general population screening was not technically possible and carrier testing was reserved for parents of affected children or their relatives,” says Stirling Puck, MD, FACMG, Genzyme’s vice president for medical affairs. “Only recently has it been possible to do population screening.” Genzyme’s assay uses real-time PCR to amplify exon 7 of the SMN1 gene and an algorithm to calculate the number of SMN1 copies. Its test does not detect point mutations.

Genzyme does the great majority of direct genetic testing for SMA diagnosis and carrier detection in the United States. “We had quite a bit of interest from the beginning,” Dr. Puck says. “I think there had been families waiting for this information and quite a few knowledgeable physicians who started sending samples right away.” Genzyme’s volume hasn’t noticeably increased since the ACMG guideline. “It will take a while for that statement to filter down to practicing physicians,” Dr. Puck says.

Genzyme has genetic counselors in practice locations—typically perinatologists’ offices—around the country. “They see patients under the direction of medical geneticists,” Dr. Puck says, “usually before testing.”

Most testing today takes place during the second trimester, but she wishes there were a stronger push to consider carrier testing during the first trimester. “Everything we are looking for, including SMA, fragile X, and CF, could be done in the first trimester on the same blood draw as the sickle cell test,” she says.

Peter N. Ray, PhD, FCCMG, FACMG, offers an international perspective on SMA carrier testing. “We do molecular testing on children who have a clinical diagnosis of SMA and carrier testing of their parents and close relatives,” says Dr. Ray, head of the Division of Molecular Genetics in the Department of Laboratory Medicine at the Hospital for Sick Children, Toronto. “We also do prenatal diagnosis on high-risk pregnancies.”

Asked about population carrier screening, he replies: “That recommendation seemed to get a lot of promotion. In Canada the health care system is different. We have universal health care funded through the government, which is a little slower in adopting any population screen. In Canada we don’t yet have population screening for CF carriers, and there has been very little discussion about setting up population screening for SMA.”

One reason for supporting population screening could be economic—avoiding some SMA births would reduce expenses for supportive care. “Those studies need to be done,” Dr. Ray says. He notes that SMA is not a common disease, with a prevalence of one in 10,000 and a carrier rate of about one in 50. “To find that one case, you have to screen all parents of those 10,000 fetuses,” Dr. Ray says. “So actually population screening is not a small cost. That doesn’t mean it isn’t worth it because the burden of disease is very high, not just for severe cases but for intermediate wheelchair-bound children as well.” In fact, the costs associated with patients who have intermediate disease may be higher because they live much longer than patients with severe disease.

Perhaps the situation is not so different in the U.S., where test adoption depends on third-party reimbursement. “What I don’t know is whether insurance will pay for SMA carrier screening,” Dr. Finkel says, referring equally to screening of pregnant women, adults planning conception, and newborns. “In general, they don’t pay for broad-based screening of whole populations where there is no pressing need to establish the diagnosis early in order to initiate therapy, as with PKU or congenital hypothyroidism.” In the end, this is likely to determine whether the ACMG recommendation will have the desired effect.

William Check is a medical writer in Wilmette, Ill.