Time to cast a wider net in fragile X testing?

April 2007
Feature Story

William Check, PhD

Shakespeare famously defined seven ages of human life, from infancy to dotage. It now appears that the mutation responsible for fragile X syndrome, or FXS, has serious clinical manifestations across nearly that whole spectrum, for men and women, a fact with important implications for medical genetics laboratories.

Currently, testing is recommended for diagnosis in symptomatic individuals, carrier detection in those who have an affected family member, and prenatal testing when the mother is a known carrier. However, the broader clinical impact of various forms of the fragile X mutation, which include a tremor/ataxia syndrome in older men with the premutation and premature ovarian failure in middle-aged women, along with an upward revision in the frequency of female carriers from one in 250 to one in 130, has raised calls for increased recognition of fragile X-related clinical syndromes, which could augur more frequent testing. Overall, mutations in the fragile X gene may affect more than 1 million men, women, and children in the United States.

In addition, some people advocate consideration of general prenatal screening and even of universal newborn screening, neither of which is now recommended. Donald Bailey, PhD, distinguished fellow at RTI International, favors universal newborn screening “with some caveats.”

“There are many benefits to newborn screening and earlier identification,” Dr. Bailey says. “These are much broader than the typical benefits that have been used as criteria for newborn screening, such as a medical treatment that would change the course of disease.”

How broader testing would play out in practice is not clear, not least because of the demanding nature of the test for fragile X mutations, which Elaine Lyon, PhD, medical director of molecular genetics in the Department of Pathology at ARUP Laboratories, calls “technically challenging.”

“The challenge,” she says, “is to have one simple test that would check for the full mutation, ascertain the size of premutations, and detect methylation. We are not there yet. Most of us use several methods to do comprehensive testing.” Jean Amos Wilson, PhD, scientific director of human genetics at Focus Diagnostics, says, “Large alleles are difficult to size and accurate sizing has profound implications for genetic testing. In this case, size matters.” Dr. Amos Wilson and 28 other members of the Association for Molecular Pathology and the CDC formed the Fragile Xperts to study this problem. This group is verifying reference materials of known allele size that are publicly available to clinical laboratories. “Reference materials will not solve the problem by themselves,” Dr. Amos Wilson notes. But they are a helpful first step.

Dr. Amos Wilson calls testing for fragile X “very intense.”

“The stakes are high. This is something that labs can miss,” she says. “Fragile X should be done only by very experienced molecular genetics labs. Doing cystic fibrosis testing is not an adequate qualification.” Says Benjamin Roa, PhD, vice president of technology development, Myriad Genetic Laboratories, “Fragile X syndrome testing is likely to remain in the high-complexity expert lab setting” for the near future. Even so, availability of the test is not likely to play a major role in constricting access. The online directory GeneTests currently lists 72 laboratories in the United States that offer the test, with another nine in Canada and 30 internationally. Dr. Roa lists several possible reasons for the fairly large number of labs doing fragile X testing, including the speculation that it is “to position the lab for potential [fragile X] population carrier screening.”

While these considerations will remain true for now, they may be dramatically overturned in the near term by an innovative new assay that was announced in late March that brings fragile X carrier screening one step closer to practice. (Related article: New assay may ‘change face’ of fragile X testing.)

It was back in 1943 that the classic facial features of this syndrome were described in individuals with mental retardation—long face, prominent forehead, and large and slightly dysmorphic ears, but it was not recognized as FXS. In men, enlarged testicles are also part of the syndrome. Later, a folate-sensitive fragile site on the X chromosome was identified, which is now called FRAXA and represents the most common mutated site. It is well re8212;ognized that FXS should be considered in young children—both boys and girls&#with unexplained mental retardation, developmental delay, or autism. Many such children are diagnosed by three years of age. “Fragile X is important because it is the most common inherited cause of mental retardation,” says Randi Hagerman, MD, medical director of the MIND (Medical Investigation of Neurodevelopmental Disorders) Institute and endowed chair in fragile X research in the Department of Pediatrics at the University of California at Davis. About one in 4,000 males and one in 8,000 females in the general population are affected with FXS. There is a very strong association between FXS and autism: 30 percent of males with the full mutation have autism. When autism is present in FXS, the IQ tends to be lower than when autism is absent.

Females typically have a milder version of FXS—only 25 percent to 30 percent of girls with the full fragile X mutation have mental retardation, compared with 85 percent of boys. The other 15 percent of boys with the full mutation are so-called high functioning—that is, they have an IQ above 70. “Children who are not strictly in the retarded range—high functioning males—may be missed by pediatricians if they have learning disabilities but their IQ is in the normal range,” says Annette K. Taylor, MS, PhD, president, CEO, and laboratory director at Kimball Genetics Inc. Such children usually have behavioral features of fragile X but may not have typical facial features. “Pediatricians should lower the bar for testing for fragile X and have a higher level of suspicion,” Dr. Taylor urges.

Since fragile X is one of the more common causes of mental retardation, it should be differentiated from other potential causes, Dr. Lyon says. She points out that, in one study, FXS was the fourth most common diagnosis for patients with development delay or mental retardation (Rauch A, et al. Am J Med Genet A. 2006;140: 2063–2074). “One of the reasons for identifying fragile X is that, if we can do earlier diagnosis, we can identify carriers and hopefully help with family planning and early intervention,” she says. “Unfortunately, early intervention is more behavioral modification. There is no real treatment for the disease itself.” A trial of a drug that modulates neurotransmission through AMPA-type glutamate receptors found no benefit (Berry-Kravis E, et al. J Child Adolesc Psychopharmacol. 2006;16:525–540). Another class of compounds is under investigation (Di Prospero NA, Fischbeck KH. Nat Rev Genet. 2005;6:756–765).

The American College of Obstetricians and Gynecologists and the American College of Medical Genetics recommend testing for the fragile X (FMR1) mutation for diagnosis, carrier detection, and in the prenatal setting (Obstet Gynecol. 2006;107:1483–1485; Sherman S, et al. Genet Med. 2005;7:584–587). Indications include children with unexplained mental retardation or developmental delay, family members of those found to have a fragile X mutation, individuals seeking reproductive counseling who have a family history of fragile X syndrome or a family history of undiagnosed mental retardation, and fetuses of known carrier mothers.

Newer investigations have extended the impact of FXS into school-age children. “Probably 60 percent of girls with the full fragile X mutation have learning disorders or emotional problems without mental retardation,” Dr. Hagerman estimates. Her group and others have identified an increased rate of shyness, social anxiety, social phobia, mood instability, math difficulties in school, and language delays in both girls and boys with the full mutation. She urges pediatricians to be aware of these behaviors and to test for FXS as a possible explanation.

Even more significant, over the past five years discoveries that Dr. Taylor calls “astonishing” have substantially expanded our appreciation of the clinical impact of a class of mutations in the fragile X gene that are not extensive enough to cause frank FXS. These discoveries were based on basic knowledge of mutations in the fragile X mental retardation 1 (FMR1) gene. In 1991 it was discovered that FXS is due to expansion of an unstable CGG repeat sequence in the 5’ untranslated region of the FMR1 gene. More than 98 percent of cases of FXS are due to this triplet expansion. (Point mutations and deletions occur rarely.) Further work uncovered that a complete spectrum of repeat numbers can be found in the population. The normal number of CGG repeats in the FMR1 gene is between five and 44. From 45 to 54 repeats is called an intermediate or grey zone. From 55 to 200 or 230 repeats is called a premutation. FXS is due to full mutations, with more than 200 to 230 and typically several hundred to several thousand repeats.

Full mutations arise from spontaneous expansion of CGG repeats in the fetus of a woman who is a premutation carrier. (How and why a premutation expands is one of the open questions in the fragile X story. Also unexplained is that males with either premutations or full mutations are not at risk for having daughters with full mutations and fragile X syndrome.) The risk of expansion of a premutation to a full mutation during transmission from a female carrier depends on the size of the woman’s repeat. The smallest repeat size to expand to a full mutation in one generation is 59. Probability of expansion to a full mutation ranges from five percent or less for repeat numbers under 70 to 100 percent for greater than 100 repeats. (Strom CM, et al. Genet Med. 2007;9:46–51). The risk for a woman who is a fragile X carrier to pass on the mutant fragile X gene is 50 percent for each child.

For many years the premutation allele was considered to be important only as conferring risk for expansion to the full mutation, not as causing clinical symptomatology itself. Over the past few years, however, evidence has accumulated to suggest that some males with premutation alleles demonstrate cognitive and behavioral problems, including social deficits, autism spectrum disorders, and attention problems, which seem to be mild variants of those seen in FXS. These findings are still under investigation.

Much more solid and significant is the evidence linking premutation status in older males with a neurological syndrome called fragile X-associated tremor/ataxia syndrome, or FXTAS, which Dr. Hagerman and her colleagues recognized in about 2001. “As part of our research, we evaluated families who have a child with fragile X syndrome,” Dr. Hagerman says. Routinely they asked about other family members. “Mothers who were talking to me about their children with FXS also described problems in their [own] fathers, who were obligate carriers.” When Dr. Hagerman saw these grandfathers, they all had the same neurological complex—progressive intention tremor and ataxia. Dr. Hagerman’s group found an age-related penetrance of this syndrome, with between 30 percent and 40 percent of males older than 50 who carry the premutation having FXTAS, a number that increases from the sixth decade into the seventh decade and beyond (Jacquemont S, et al. JAMA. 2004;291:460– 469). These authors concluded, “Since male premutation carriers are relatively common in the general population, older men with ataxia and intention tremor should be screened for the FMR1 mutation, especially if these signs are accompanied by parkinsonism, autonomic dysfunction, or cognitive decline, regardless of family history.” An estimated one per 813 men in the population carries a premutation allele.

“It is important for medical specialists to screen for premutation when they see symptoms of FXTAS,” Dr. Hagerman says. “It has implications for other family members.” All of the daughters of a grandfather who develops FXTAS will be premutation carriers and are at 50 percent risk of passing on the fragile X mutation to their offspring, either as a premutation or expanded to a full mutation.

The ACMG recommends considering testing for the fragile X premutation in “men and women who are experiencing late onset intention tremor and cerebellar ataxia of unknown origin, especially if they have (a) a family history of movement disorders, (b) a family history of fragile X syndrome, or (c) male or female relatives with undiagnosed mental retardation or autism.”

FXTAS develops only rarely in female premutation carriers. Middle-aged women with a premutation allele develop a different syndrome—premature ovarian failure, or POF, defined as a cessation of menstrual periods before age 40. A fragile X premutation allele is the leading cause of POF, according to Dr. Hagerman. She estimates that four percent to 14 percent of women who present with POF will be fragile X premutation carriers. Conversely, about 21 percent of women who are premutation carriers will have POF, 21-fold higher than in the general population.

“We think an awareness of testing is really important for OB-GYN physicians,” Dr. Hagerman says. It’s not only that they treat women with obvious POF. “They also see women who have infertility problems and who might have POF. The couple might spend much money to get pregnant and have a child at high risk of fragile X.” ACOG recommends testing for the fragile X premutation in women with “ovarian failure or an elevated follicle-stimulating hormone level before 40 years of age without a known cause.” The ACMG’s recommendation is similar; it includes “women who are experiencing reproductive or fertility problems associated with elevated follicle stimulating hormone.”

With wider testing for fragile X and a deeper knowledge of its implications, the ACMG’s document mentions ethical concerns. “[P]remutation carrier status associated with late onset disorders may be inadvertently uncovered in an individual who is tested as part of a family study,” its statement says. “For late onset disorders such as these, ethical issues arise as to whether or not a given individual wishes to know his or her carrier status.”

Mutations in the fragile X gene exert a strong presence during the early, middle, and later ages of life; it seems the only stage not greatly affected is the very end of life. Dr. Hagerman notes that a person with FXS can die early of seizures. And a rare person has died suddenly from an arrhythmia related to mitral valve prolapse. “But other than these issues,” she says, “they usually have a normal life span.”

Insights into the pathological mechanisms in FXS and premutation syndromes have arisen from a collaboration between Dr. Hagerman’s group and neuropathologist Claudia Greco, MD, associate professor of pathology at the University of California at Davis and neuropathologist at the MIND Institute. Dr. Greco discovered unique ubiquitin-positive intranuclear inclusions throughout the brain that contained FMR1 mRNA and more than 20 other proteins in the brains of 11 men who died with FXTAS (Greco, CM, et al. Brain. 2006;129:243–255; Iwahashi CK, et al. Brain. 2006;129:256–271). More recently, Dr. Greco observed similar inclusions in the testes and pituitary glands of two men who died with FXTAS and who had early-onset impotence with no known cause (Greco, CM, et al. J Urol. 2007;177:1434–1437). She is now studying the ovaries of women with POF. “General pathologists need to be aware of these [brain] inclusions,” she says. “They are most numerous in the hippocampus, a section taken routinely in the examination of the post mortem brain.” She suggests referring autopsy brains from individuals with atypical parkinsonism to a neuropathologist.

Evidence is accumulating that the pathology in premutation syndromes is caused by an excess of FMR1 mRNA that leads to RNA toxicity and induces formation of intranuclear inclusions (Hessl D, et al. Am J Med Genet B Neuropsychiatr Genet. 2005;139:115–121). In FXS the situation appears to be quite different: Pathology results from a lack of transcription due to the fully methylated status of the FMR1 gene, though it is not understood why methylation occurs when the repeat number exceeds about 200.

Further information on the relation among FMR1 mRNA, FMR1 protein (FMRP), and clinical syndromes comes from work by Dr. Taylor, whose laboratory is one of the few clinical facilities that measures FMRP. She, Dr. Hagerman, and Flora Tassone, PhD, now of UCD and the MIND Institute, studied 16 carrier males, 26 mosaic males (men with some cells containing a full mutation allele and some cells containing a premutation allele), and 92 females with a full mutation. They also examined 31 males with a full mutation but in whom not all FMR1 genes were fully methylated—another finding awaiting explanation. They found that FMRP expression in more than 50 percent of lymphocytes correlated with high functioning status, stressing the crucial role of FMRP in cognitive development (Tassone F, et al. Am J Med Genet. 1999; 84:250–251; Tassone F, et al. Am J Hum Genet. 2000; 66:6–15).

Consistent with the RNA toxicity hypothesis, in premutation carriers RNA increased with the number of CGG repeats while protein levels decreased. The drop in FMRP started with repeat numbers in the grey zone and was accentuated in the upper premutation range. Results in mosaic males were also consistent with that idea. “These discoveries by Dr. Tassone in 2000 were very exciting,” Dr. Taylor says.

As indications for fragile X testing expand, with a probable increase in the number of tests ordered, more laboratories may consider initiating this test. Testing for fragile X is particularly challenging. “When you do clinical diagnosis for fragile X, you are looking at repeat numbers over a very broad range,” says Dr. Amos Wilson, “from as small as five repeats up to thousands. It can be quite difficult to determine the exact length of the repeat in the FMR1 allele.” This is due to the high G+C content. “It is difficult to amplify through the full mutation,” Dr. Lyon explains. “Not because of methylation, but because the high number of CGG repeats makes it harder to separate the DNA strands. This is also true for the premutation range, but it is not as much of a problem.” Current laboratory-developed assays reliably amplify up to 150 to 200 repeats. In addition to problems in assessing larger allele size, Myriad’s Dr. Roa notes that using only PCR can cause confusion with individuals who are mosaic for mutations in the FMR1 gene.

To overcome these limitations, most laboratories also run samples by Southern analysis, which has lower resolution for size but shows the entire length spectrum. “Southern analysis has greater dynamic range than PCR but is not as precise,” says Dr. Amos Wilson. “With current lab-developed assays, you can detect full mutations and most premutations using Southern analysis, but sizing of these alleles is imprecise.” Using a methylation-sensitive enzyme, Southern analysis also detects methylation status, which is important for establishing the presence of a full mutation.

All laboratories use some version of both PCR and Southern analysis for fragile X testing, either simultaneously or sequentially. Says Dr. Lyon, “I’m not comfortable dropping the Southern yet. We do them simultaneously.” Using only one of the assays might work for population screening, she says. “With population screens you will accept some false-negatives. But you don’t want to accept false results with a diagnostic test.”

Testing for fragile X is also complicated by the presence of sex aneuploidy—Klinefelter and Turner’s syndromes—which are about as common as fragile X itself. These chromosomal anomalies pop up among samples submitted for fragile X testing because the conditions have some similar features, according to Dr. Lyon. To avoid confusion in these cases, she uses gender markers. “The Abbott ASR retains gender markers,” she says, “which helps us to be sure of gender. If we get a male sample and see a female pattern—one methylated allele and one unmethylated—and confirm a Y chromosome marker, we suspect Klinefelter syndrome right off instead of suspecting a sample switch.” In a laboratory report on such a case, Dr. Lyon would recommend cytogenetic testing if it hasn’t already been done.

Dr. Taylor says that, in the clinical setting, FMR1 protein analysis would be done only in males and is most helpful when the patient appears to have a deletion in all or part of the FMR1 gene based on Southern analysis. FMRP analysis will reveal the effect of the deletion on expression of the fragile X protein. FMRP analysis can also be instructive if Southern analysis indicates a full mutation that is unmethylated in a significant proportion of cells or indicates mosaicism (full mutation plus premutation) with the premutation in a significant proportion of cells. In these cases, appreciable FMRP expression may be present.

To investigate the accuracy of fragile X testing in experienced laboratories, the Fragile Xperts group conducted a consensus survey on 16 cell lines with various CGG repeat numbers ranging from normal to full mutation. To set the scope of the problem, Dr. Amos Wilson cites a patient sample in a 2005 CAP Survey that had a CGG repeat number around 90. “Participating laboratories reported from 67 to 128 repeats,” she says. “Only half reported numbers between 88 and 92. Think about the risk to expand to a full mutation for the extreme values: The risk for a repeat length of 67 is about five percent, while the risk for a repeat length of 128 to expand is 100 percent.” In the Fragile Xpert survey, the laboratories achieved consensus when alleles were less than 100 repeats, but not for repeat numbers much above 100. “That demonstrates how hard it is to do, because our labs are among the most experienced,” Dr. Amos Wilson says.

This consensus survey grew from recommendations made in a study commissioned as part of a contract to DynCorp eight years ago by the Laboratory Practice Evaluation and Genomics Branch of the Centers for Disease Control and Prevention to study the needs of genetic testing. “One of the main findings from the DynCorp study was that there is a serious lack of materials that can be used for quality control for these tests,” says Lisa Kalman, PhD, who heads the Genetic Testing Reference Material Coordination Program (GeT-RM; www.phppo.cdc.gov/dls/genetics/qcmaterials/default.aspx), which grew out of that contract. GeT-RM has coordinated the development of reference materials for about a dozen genetic tests. (For a list of available materials, some characterized by the GeT-RM program and some not, see www.phppo.cdc.gov/dls/genetics/qcmaterials/materialsavailability.aspx.) Fragile X is one of the most challenging of these tests. “In CAP proficiency surveys, fragile X has been a very difficult test,” Dr. Kalman says. “Often labs have difficulty accurately sizing samples.”

Cell lines that were studied and characterized in the Fragile Xpert survey have been deposited at the Coriell Cell Repositories and, once Coriell finishes the acceptance process, they will be publicly available for QC and proficiency testing. In fact, it was the DNA from these cell lines that was characterized, Dr. Kalman notes. “Very soon, you will be able to buy this same DNA from Coriell,” she says. “We hope that laboratories will purchase the characterized DNA rather than the cell lines to use as reference materials. The characterization of these FX reference materials was a completely voluntary effort,” she says. “We had amazing support and enthusiasm from the genetics testing community.”

In the Fragile Xpert survey, each laboratory used its own laboratory-developed assay plus a “common” assay: Celera provided analyte-specific reagents and general-purpose reagents that each of the laboratories assembled into assays. Among the 16 experimental samples, statistical consensus was achieved for both platform types in 11 and the common platform only for one sample. Celera’s reagents do not constitute a kit, emphasizes Mike Zoccoli, PhD, vice president of development, instrument systems and software at Celera. Rather, it is a proprietary set of reagents with which a laboratory can develop an assay. “The advantage [of our reagents] is that they can be used to make the most powerful PCR system for amplifying GC-rich targets up to 900 or 1,000 repeats, three times the highest published figure,” Dr. Zoccoli says. Celera’s materials achieve this by three innovations. First, what Dr. Zoccoli calls “pretty intriguing” cycling conditions, which have been largely disclosed. Not much has been published so far about the other two factors—a particular enzyme reagent and a special PCR buffer. “We had a pre-IDE meeting with FDA on what they will require for clearance of an IVD kit,” Dr. Zoccoli says. “We intend to submit data for both FDA and CE clearance of an IVD kit based on our reagents and methods.”

Broader indications for testing, a challenging assay, and ethical considerations about privacy and genetic counseling—a complex mix. One commercial reference laboratory, Genzyme Genetics, has already ventured cautiously into this minefield. “What we do is to ask physicians whose patients we see for genetic counseling if they would like us to offer fragile X carrier testing to their patients,” says Stirling Puck, MD, national medical director at Genzyme. Dr. Puck acknowledges that fragile X testing in women with no family history is not recommended now by either ACOG or the ACMG. About half of physicians agree to have Genzyme’s counselors mention fragile X testing to their low–risk patients. In one publication from Genzyme, less than 10 percent of patients agreed to have fragile X testing; in this group the premutation allele frequency was one in 382 (Cronister A, et al. Genet Med. 2005;7:246–250). This raises the question—what number justifies a test? “We do not make that determination,” Dr. Puck says. “Are we in the brave new world where patients want to be tested for everything? There are patients who show up and say, ‘I want to be tested for everything that’s available.’ Where do you draw the line? It’s up to individual physicians and patients.” Yet, one could argue that, simply by offering the test, Genzyme has already determined that it is justified.

In the in vitro fertilization arena, some clinics test for the fragile X mutation in potential mothers and in oocyte donors, since the carrier rate is quite high. “We have noticed that this trend is growing,” Dr. Taylor says. Another broad application would be universal newborn screening. Dr. Bailey lists potential benefits as access to early intervention programs and family information about reproductive risk. He has expressed his views in several commentaries (Bailey DB Jr., et al. Ment Retard Dev Disabil Res Rev. 2006; 12:270–279; Bailey DB Jr. Ment Retard Dev Disabil Res Rev. 2004;10:3–10; Bailey DB Jr., et al. Am J Public Health. 2005; 95:1889–1893). “In fragile X and other conditions, research supports that parents want to know as early as possible, even if there is no medical treatment,” he says. “Most parents have a definition of benefits that is broader than the medical community’s.” But Dr. Bailey’s surveys were done among parents with a child affected with FXS. Genzyme’s experience suggests that families who are not at risk have a different view.

Dr. Bailey recognizes several objections to his position. At the recent ACMG meeting, he listed them: Would newborn testing have to be done with informed consent? Would it overwhelm the ability of the available genetic counselors to provide information and support? Also, he notes, “Detecting carrier status in infants is something people are concerned about. Some conditions are late onset in carriers and could heighten anxiety. I still support the concept,” he says, “but we need to make sure we know how to do it right and provide support to families before we jump into it.”

Dr. Bailey also recognizes the technical realities. “First you need an accurate and inexpensive test,” he says. “That is kind of a dealbreaker. If you don’t have that, things will not move forward.” He is awaiting validation now of a novel test by a collaborating laboratory. “We have pilot studies funded and ready to go and are waiting for final validation of the test,” he says. “We will study how people react to finding both full and premutations and what kind of support they need.”

Dr. Bailey is an optimist about technology and finds obstacles in social issues. “Several university researchers and companies are working on tests,” he points out. “We are confident that something will be ready soon. We will be able to detect so much,” he marvels. “Technology has gotten way ahead of social science. We have to do research on how people react to a variety of types of genetic information and what we have to have in place to make sure that disclosure of genetic information results in benefits and minimizes harms.”

Between geneticists who are meticulous about method and social scientists who are meticulous about psychological concerns, perhaps the future of genetic testing is not so threatening.

Related Links

• New assay may ‘change face’ of fragile X testing