A m e r i c a n Journal of Medical Genetics 44:816-82 1 (1992)
Cytogenetic Guidelines for Fragile X Studies Tested in Routine Practice Gordon W. Dewald, Diane D. Buckley, J a c k L. Spurbeck, and Syed
M.Jalal
Cytogenetics Laboratory, Mayo Clinic and Mayo Foundation, Rochester, Minnesota
Several organizations have proposed guidelines for fra(X) studies on peripheral blood lymphocytes. To evaluate these guidelines, we reviewed 1,033 consecutive specimens referred for fra(X) analysis. Each specimen was cultured with medium 199 and RPMI 1640 with 5-fluorodeoxyuridine or excess thymidine. The karyotype and expression of fra(X) were established from 20 GTL- or QFQbanded cells and by screening of up to 130 more banded cells. We found anomalies other than fra(X) in 37 (3.6%) of the patients. We found 4% or more fra(X) cells in 38 (3.7%) cases from 36 unrelated families, including 33 (3.9%) of 850 males and 5 (2.7%) of 183 females. Another 4 females had 1 to 3% fra(X) cells. Six specimens were fra(X)-positive in only one stress system, and 32 were positive in 2 systems. To find the first 2 fra(X)cells in males, we needed to study up to 36 cells in 31 cases, 50 in one case, and 57 in another. To find the first 2 fra(X) cells in females, we needed to study up to 17 cells in 4 cases and 57 in another. A strong family history of fra(X) occurred in 5 patients, and each one was fra(X)-positive. Some manifestations of the fragile X syndrome occurred in 507 cases, 17 (3%)of which were fra(X)-positive. Abnormalities considered unlikely to be the fragile X syndrome occurred in 103 cases, 3 (3%)of which were fra(X)-positive. Use of chromosome breakage and fra(3Mp14)as quality control indicators of the fra(X) stress systems was found to be unreliable. Our findings support most of the proposed guidelines for fra(X) studies but indicate a need for modifications of others. 8 1992 Wiley-Liss, Inc.
KEY WORDS: Fragile X, fra(X) guidelines, Xq27.3
Received for publication June 24, 1991; accepted October 11, 1991. Addressreprint requests to GordonW. Dewald, Ph.D.,Cytogenetics Laboratory, Mayo Clinic, 200 First Street SW, Rochester, MN 55905.
0 1992 Wiley-Liss, Inc.
INTRODUCTION A fragile site at Xq27.3 [Krawczun et al., 19851 has been associated with the fragile X syndrome [Lubs, 1969; Sutherland, 1985; Opitz, 19861. The manifestations of the fragile X syndrome are well known, but they vary considerably among patients [Sutherland and Hecht, 19851. Thus, the use of clinical criteria alone to diagnose the fragile X syndrome can often be difficult. Consequently, many physicians routinely request fra(X) studies on all patients with mental retardation and patients with physical findings resembling those of the fragile X syndrome or a family history of X-linked mental retardation [Turner et al., 1980a,b; Shapiro et al., 1991bl. Several regional and international organizations have proposed guidelines for fra(X) studies (Table I). These guidelines vary considerably and are largely untested in routine practice. To help evaluate and formulate standards for routine practice, we reviewed 1,033 consecutive specimens referred to us for fra(X) studies between 1985 and 1990. We attempted to evaluate the following guidelines for routine practice: (1)the need for constitutional chromosome analysis, (2) the usefulness of multiple stress systems, (3) the optimal number of metaphases necessary for analysis, (4) the labor effort needed to conduct these studies, (5)the clinical reasons for referring specimens, and (6) the efficacy of recommended quality control procedures. METHODS We retrieved data from our laboratory computer system for all peripheral blood specimens referred for fra(X) studies from 1985through 1990. We also reviewed data for other blood specimens referred for routine cytogenetic studies during this same period which proved to be fra(X)-positive. We process all blood specimens for routine chromosome studies with medium 199 and RPMI 1640 to permit identification of fra(X) chromosomes. For each peripheral blood specimenreferred for fra(X) studies, we initiated cultures using 2 different stress systems within 48 hours of arrival of the specimen. The stress systems included medium 199 and either RPMI 1640 with 5-fluorodeoxyuridine(FUdR) or RPMI 1640 with excess thymidine (TdR). For the medium 199 method, 1 mL of heparinized whole blood was cultured
Fragile X Studies Tested in Routine Practice
817
TABLE I. Summary of Reported Guidelines for fra(X) Studies Guidelines
ACT
CPT
No. of cells Males Females Additional cells Standard banded study Induction systems Definition of fra(X) syndrome Quality control 26% cells with breakdgaps 24% cells with fra(3)(p14) Other
...
100
50 100 Yesb Yes Yes' 2 4%
.. ... ,
Yesg
...
... ...
Yes
...
... ... ... ...
IV fra(X) workshoD
... 100 150 Yes' Yes
2-3
CAP ... ...
... ... ...
...
24%
...
... ... ...
... ... Yesh
GLaRGG
... 50 100 Yesd Yes
22 24%'
Yes Yes
...
HCFA ... 50 100 Yesd Yes ...
...
Yes Yes
...
PSRGN
NYS'
74 ...
...
...
Yes' Yes 22
24%'
...
... Yes'
50 100 Yesd Yes 22 24%'
Yes Yes ...
New York State does not include criteria for fra(X)studies in their laboratory checklist, but they do recommend HCFA guidelines in their manual for inspectors. If one or a small number of positive cells is observed, it is recommended that a second blood sample be requested and/or that multiple induction methods be used. If only one positive cell is observed, score an additional 100 cells. If more than one cell but less than 3%of metaphases show the fragile X, the results should be regarded as equivocaland another blood sample requested. It may be necessary to request cessation of dietary folate supplementation, i.e., multivitamins. 'When a frequency of less than 4% is obtained, a repeat specimen should be requested; if no higher frequency of expression is obtained, then the results should be interpreted with caution and on a case-by-case basis. dIf one metaphase is positive for the fragile X chromosome, 50 additional cells should be scored. 'There is no recommendation on the number of induction systems, only that a proven induction system should be used. 'Results with less than 5% fragile X expression should be interpreted with caution. gIf a laboratory reports several consecutive negative studies in patients with a fragile X phenotype and/or positive family history, it is suggested that the culture system be checked by either splitting fragile X samples with a laboratory known to have success with the technique or by obtaining cells from a known fragile X positive individual to use a s a control. I,Appropriate QC systems may include correlation with phenotype and family history, folate-sensitive fragile sites. 'If there is less than 4% expression of known folate-sensitive sites, the results should be interpreted with caution. Abbreviations: ACT, Association of Cytogenetic Technologists; CF'T, Current Procedural Terminology; IV fra(X) Workshop, IV International Workshop on the Fragile X Syndrome and X-Linked Mental Retardation; CAP, College of American Pathologists; GLaRGG, Great Lakes Fkgional Genetics Group; HCFA, Health Care Financing Administration; PSRGN, Pacific Southwest Regional Genetics Group; NYS, New York State.
a
in 10 mL medium 199 with 5% fetal bovine serum for 72 to 96 hours [Sutherland, 19791.For the FUdR procedure, 1mL of heparinized whole blood was cultured in 10 mL of working RPMI 1640 medium and cultured for 48 to 72 hours. FUdR was then added (final concentration, 0.123 kg/mL)for an additional 18 hours [Glover, 19811. For the excess TdR method, 1mL of whole blood was cultured in 10 mL of RPMI 1640 for 48 to 54 hours. TdR was then added (final concentration, 250 kg/mL) for another 18to 24 hours [Sutherland et al., 19851. All cultures were incubated a t 37°C in 5% COO,5%02, and 90% NO.Standard cytogenetic procedures were used to harvest the specimens for chromosome analysis. In each case, we used GTL- or QFQ-banded metaphases for chromosome analysis. We routinely analyzed 20 banded metaphases and prepared 2 to 3 karyotypes to establish the constitutional karyotype. In both males and females, we routinely screened 30 t o 80 more metaphases for the presence of fra(X) or sex chromosome mosaicism. These metaphases were also banded to identify the X chromosomes and to distinguish between the common fragile site at Xq27.2 [Sutherland and Baker, 19901 and fra(XNq27.3). Our cytogenetic definition for the fragile X syndrome required the observation of 4% or more metaphases with a fra(X). We used particular caution to interpret and process cases with l to 3%cells with fra(X) and cases in which there was considerable risk for fragile X syndrome. We attempted to evaluate the usefulness of the guidelines for monitoring breaks and gaps and fragility at 3p14 to assess the potency of the stress systems. This was done by establishing the incidence and normal range for these criteria on 100 consec-
utive specimens and the fra(X1-positivecases. This work was done by a group of technologists who were specifically trained t o study chromosome breakage to ensure consistent and accurate scoring of breaks, gaps, and fra(3)(p14). RESULTS During this 6-year period of this investigation, we processed 1,085 specimens for fra(X1 studies, 1,033 (95%)of which had sufficient numbers of metaphases for a suitable study. Of the successful studies, 850 (82%) specimens were from males and 183 (18%)from females. Ages were provided for 829 males and 181 females. The mean age of the males was 8.8 2 8.7 years (range, newborn to 62 years) and that of the females was 12.2 2 11.5 years (range, newborn to 49 years). Each female withfra(X) was older than 15years. By comparison, only 8 (26%) of 31 males with fra(X) whose age we knew were older than 15 years. We studied an average of 52 ? 10 (range, 20 to 101) metaphases per male patient and 74 2 26 (range, 30 to 150)metaphases per female patient. Only 5 males and 3 females had fewer than 50 suitable metaphases for analysis. For each of these cases we recommended that another study be done. With use of the number of bands on chromosome 10, the band resolution for each case averaged approximately 500 bands per haploid chromosome set. Chromosome anomalies other than fra(X) occurred in 37 (3.6%) of the specimens. These abnormalities included structural anomalies in 29 patients (11balanced and 18unbalanced) and 8 patients with numeric anoma-
Dewald et al.
818 60 50
culture, harvest, and establish the constitutional karyotype based on analysis of 20 metaphases and 47.8 i 16.1 minutes to screen 30 cells for fra(X)and sex chromosome mosaicism. Thus, 154 2 0.56 minutes were needed per metaphase screened. The need for full-time equivalents for fra(X) studies increases significantly with higher numbers of analyzed metaphases. Two of the authors (D.D. Buckley and J.L. Spurbeck) restudied 30 fra(X)-positive cases to establish the incidence of chromosome breakage and fra(3Xp14) expression in each of the culture stress systems. Most chromosome breakage occurred in cultures with excess TdR and the least occurred with medium 199 (Table 11). By combining 2 stress systems, 27 of 30 cases would have had chromosome breakage in 6% or more cells. The expression of fra(3)(p14) was a less reliable quality control indicator. An observable fra(3)(p14) was most common in cultures with excess TdR. Cultures with medium 199 or FUdR were relatively poor inducer systems for fra(3Kp14).Even after combining 2 stress systems, only 9 of 30 specimens had 4% or more cells with an apparent fra(3)(p14). We studied chromosome breakage and fragility at 3p14 for medium 199 and excess TdR in a consecutive series of 100 blood samples (Table 111).The results indicated that the vast majority of specimens cultured in medium 199 would not meet the guidelines recommended by many organizations. More specimens cultured with excess TdR met the minimal criteria than those cultured with medium 199, but many specimens still would need to be judged unsatisfactory. The mean number of metaphases with chromosome breakage and fragility at 3p14 is much greater with excess TdR than with medium 199. However, the lower 95% limit for medium 199 and for excess TdR is 0% for both chromosome breakage and fragility at 3p14. This finding indicates that in routine practice, no threshold would be reliable to identify system problems for individual cases. We attempted to correlate the reason for referral with our chances of observing a fra(X). A strong family history of fra(X) was indicated in 5 specimens, and each one was fra(X)-positive. We also studied 13 unaffected relatives of these patients, and each one was fra(X)-negative. Some manifestations of fra(X) were indicated in 507 specimens; 17 13%) were fra(X)-positive. These anomalies were mainly mental retardation (207 patients), developmental delay (192 patients), autism (103 patients), and other anomalies (5 patients). Traits unlikely to be the fragile X syndrome were indicated in 103
El Cells with fragile X.
% Cells analyzed to find 2 fraaile X
40
30 20
10 n " 1 3
5
7
9 11 13 15 17 19 21 23 25 27 29 31 33 1 3
-1
33 males
5
5 females
Fig. 1. Percentage of fra(X)-positive cells and number of cells needed to find the first 2 cells with fra(X) in 33 males and 5 females.
lies. Among the specimens referred specifically for fra(X) studies, we found more than 4% fra(X)cells in 38 (3.7%)specimens. These patients included 33 males and 5 females from 36 unrelated families. We found 2 additional patients with more than 4% fra(X) cells from specimens referred for cytogenetic studies other than fra(X). Among the cases referred for fra(X) studies, 6 were fra(X)-positive in only one of 2 stress systems and 32 were positive in 2 stress systems. Four males were fra(X)-negative by medium 199 and fra(X)-positive by either FtMPI 1640 with FUdR or excess TdR.One male was fra(X)-positive by medium 199 and negative by excess TdR. One female was fra(X)-negativeby medium 199 and fra(X)-positive by FUdR. For each patient, we established the number of cells needed to find the first 2 fra(X) cells (Fig. 1).In 31 males, 2 fra(X) cells were observed in the first 36 cells. For the 2 other males, 50 and 57 cells were analyzed before 2 fra(X) cells were found. In 4 females, 2 fra(X) cells were found in the first 17 metaphases. In the other female, 57 cells had to be analyzed to observe 2 containing a fra(X). The mean percentage of fra(X)-positive cells was 18.5 ? 5.7% (range, 4 to 50%)among the fra(X) males and 12.1 2 5.8% (range, 6 to 22%)among the females (Fig. 1).In general, the higher the percentage of fra(X) cells in a culture, the fewer the number of cells needed to observe two fra(X)-positive cells. We considered the effort needed to screen cells for fra(X). Work-load recording indicated 432 minutes to
TABLE 11. Incidence of Cells With Chromosome Breakage and fra(3)(p14)in Cases That Are fra(X)-Positive Breakage Method Medium 199 FUdR" Excess TdRb Overall
No. of cases
29 12 16 30 FUdR, 5-fluorodeoxyuridine. bT&, thymidine.
In
6%cells, no. cases 17 9 16 27 2
Mean ? SD (range), % 9 2 7 (0-32) 10 2 6 (0-20) 21 k 12 (8-50) 13 t 7 (4-32)
In
fra(3Mp14) Mean c SD no. cases ___.___ (range), % 2 0.3 2 1 (0-4) 2 1.0 t l(0-4) 12 6.0 '. 5 (0-18) 9 2.0 2 2 (0-81 2
4%cells,
~
Fragile X Studies Tested in Routine Practice
819
TABLE 111. Chromosome Breakage and Fragility at 3p14 in a Consecutive Series of 100 Blood Specimens Medium 199
Metaphases with breakdgaps Mean 2 SD
Excess TdR
(%)
Range 95% normal limits % cases 26%"
Metaphases with fragility at 3p14 (%) Mean 2 SD Range 95% normal limits % cases 24%"
4 + 4 0-17 0-16 21
22
0 5 2
10
0-8 0-7
8
?
14
0-84
0-51 91 5
9
0-56 0-32 72
"Common guidelines for fra(X) studies (Table I).
specimens;3 (3%)were fra(X)-positive.No clinical information was provided in 405 cases; 13 (3%)were fra(X)positive. Four female patients in this series had 1 to 3%fra(X) cells. In one case, we found 3 of 100 metaphases with an apparent fra(X). In each of the other 3 cases, we found only 1 of 100 metaphases with an apparent fra(X). We have not received samples for further study from any of these patients.
DISCUSSION Several organizations have recommended guidelines for fra(X) studies (Table I). These organizations include the Great Lakes Regional Genetics Group (GLaRGG) [Higgins et al., 19901, Pacific Southwest Regional Genetics Network (PSRGN)[19901,Health Care Financing Administration (HCFA)[Department of Health and Human Services], Association of Cytogenetic Technologists (ACT) [Knutsen et al., 1989,19911, and the IV International Workshop on the Fragile X Syndrome and X-Linked Mental Retardation [Jacky et al., 19911. The Physicians' Current Procedural Terminology (CPT)code [American Medical Association, 19901 suggests criteria for fra(X) studies for billing purposes. The College of American Pathologists (CAP) [19911 has not yet dealt with this issue in its entirety. The New York State Department of Health (NYS) [19901 suggests some guidelines for fra(X) in their interpretive manual for inspectors. For each organization, the guidelines were derived from consensus agreement of committees comprised primarily of cytogeneticists. Reasons for Referral Our experience indicates that only about half of the specimens referred for fra(X) studies involve patients with any clinical manifestations (49%)or a family history (1.7%)of the fragile X syndrome. Thus, it may not be surprising that only 38 (3.7%)of the patients in our series had more than 4% cells with fra(X). It may also explain why we found another 37 (3.6%)of the patients who had chromosome abnormalities other than a fra(X). On the basis of a 3-year fra(X) study involving 1,012 patients with developmental delay, Voullaire et al. [1989] observed 31 (3.1%)patients with fra(X) and 53 (5.3%) with other chromosome abnormalities. Their study and ours agree with the recommended guidelines
(TableI) that in practice, in addition to fra(X)analysis, it is important to establish the constitutional karyotype of these patients. Our results suggest that about 93%of specimens referred for fra(X) studies do not have any apparent chromosome abnormality. This finding probably reflects the effort of physicians to rule out the fragile X syndrome as well as other chromosome abnormalities in all mentally retarded or autistic patients [Shapiro et al., 1991bl. In practice, this is probably a good policy because our studies identified many patients with different kinds of chromosome abnormalities. In addition, sporadic patients with Klinefelter syndrome [Filippi et al., 19881 and Down syndrome [Arinami et al., 19871 are also known to have the fragile X syndrome. A molecular test is available that will identify the fra(X) mutant gene [Oberle et al., 1991; Yu et al., 19911, although the molecular test would not detect other chromosome abnormalities and may not be diagnostic for the syndrome in every case [Craig, 19911. Thus, a molecular test may well be effective for screening high-risk individuals. Cytogenetic analyses will most likely continue to be used to provide an important diagnostic test in routine practice for patients with mental retardation of uncertain origin.
Number of Metaphases Analyzed Our work-load recording indicates that, on the average, it takes 432 minutes to culture, harvest, and establish the constitutional karyotype and another 1.54 minutes to find and screen each additional metaphase for fra(X). The analysis of more metaphases than is necessary t o diagnose this disease may be an excessive expense. Thus, it is important to establish the minimal number of metaphases for analysis without sacrificing accuracy in routine practice. Several organizations recommend that analysis of 50 metaphases in males and 100 in females (TableI). Some organizations recommend the analysis of far more metaphases [Jacky et al., 19911, whereas some cytogeneticists think that in routine practice it may not be necessary to analyze so many metaphases. To assess this issue empirically, we determined the number of metaphases needed to find 2 fra(X)-positivecells in each patient with the fragile X syndrome. We would have identified 27 of 33 males and 4 of 5 females by analyzing only 20 meta-
820
Dewald et al.
phases (Fig. 1).We would have found 32 of 33 fra(X) males and 4 of 5 fra(X) females by analysis of 50 metaphases; these cases meet the minimal criterion of 4% metaphases with a fra(X). Only one fra(X) male and one fra(X) female would have required the analysis of more than 50 metaphases. For each male and female in our series, fra(X) would have been diagnosed by analysis of only 57 metaphases. These results suggest that, at least for females, fewer metaphases probably could be analyzed in routine practice than indicated by the existing guidelines (Table I). Furthermore, they suggest that the analysis of at least 100 cells for males and 150 for females, as recommended by the IV International Workshop on Fragile X [Jacky et al., 19911,may be excessive for routine practice. Our data support those guidelines that recommend analysis of 50 cells in males and 74 in females (Table I). This conclusion assumes that if 1to 3% fra(X) cells are encountered in the initial analysis, the investigator will analyze 50 more cells or consider other options. Using statistical logic, some investigators have attempted to establish the number of metaphases necessary to analyze in order to find 4% fra(X) cells [De Arce, 19831.We are concerned that this approach may not be reliable because it is based on the perception that the cellular expression of fra(X) is random. The inability to observe a fra(X) in every metaphase of a patient with fragile X syndrome may be due more to an imperfect cytogeneticprocedure or possible intercellular variation in fra(X) gene amplification [Yu et al., 19911 than to predictable biologic criteria. Tommerup et al. [1988] attempted to determine the minimal number of cells needed for fra(X) investigations based on a retrospective analysis of fra(X) studies on population in Europe and Australia. They estimated that if these studies had been based on analysis of 50 cells per patient rather than 100 cells per patient, about 1.1 to 4.1% of fra(X) males and 2.6 to 5% of fra(X) females would not have been identified. These studies were done using only one stress system (medium 199). The results of our study and others indicate that the sensitivity of fra(X) studies is considerably enhanced by the use of two or more fra(X) stress systems used either separately or together [Howard-Peebles,1991; Jenkins et al., 19911. Our data suggest that it may be more important to use 2 or more fra(X) stress systems than it is to analyze more than 100 metaphases.
Multiple Induction Systems The oldest cytogenetic method used to process specimens for fra(X) utilizes folic acid-deficient medium 199. Many investigators think that culture medium supplemented with either FUdR [Glover, 19811or excess TdR [Sutherland et al., 19851 may be more reliable. Many cytogenetic laboratories use 2 or more stress systems in routine practice, and this method is consistent with most published guidelines for fra(X)studies (Table I). In our series, 6 of 38 specimens were positive in only one of 2 stress systems. The routine use of 2 stress systems for fra(X) studies provides an important quality assurance factor and helps identify certain cases that might be missed with asingle stress system.
Regardless of how carefully the cytogenetic fra(X) studies are done, we think the results will always be equivocal in a few patients. Among the specimens referred for fra(X) studies, we found four females with fewer than 4% fra(X)-positive cells. In such ambiguous cases it may be useful, as recommended (Table I), to analyze more metaphases, study additional blood specimens, investigate whether the patient is using folatesupplement therapy, and examine the patient and review the family history. In these cases, the new direct molecular tests for fra(X) may be particularly valuable. Alternatively, RFLP studies may yield important diagnostic information if there is an established history of the fragile X syndrome.
Quality Assurance Indicators To ensure that the fra(X) stress system is working, some laboratories process a control blood specimen or cells from a lymphoblastoid cell line from patients known to have more than 4% fra(X) cells. This approach requires either a readily available fra(X) donor or the maintenance of lymphoblastoid cell lines. In addition, it does not provide quality assurance for every case. Several organizations recommend that the observation of 6% or more cells with breaks or gaps is evidence that a fra(X) stress system is working. However, our findings suggest that the 6%threshold is not adequate for either medium 199 or excess TdR because many specimens do not achieve this level of chromosome breakage (Table 11). Furthermore, the apparent incidence of chromosome breakage varies among laboratories because of different culture conditions and even among technologists in the same laboratory because of the subjectivity of scoring breaks and gaps. Thus, it may be necessary for each laboratory to establish its own laboratory statistics for chromosome breakage in fra(X) stress systems. Our data suggest that it may not be possible to determine a reliable threshold for chromosome breakage for medium 199 and excess TdR because the lower 95% limit for both of these methods was 0% (Table 111). Thus, monitoring chromosome breakage to identify technical problems may not be reliable for individual cases. It may be better to look for three or four cases in a row with 0% chromosome breakage before becoming concerned about technical problems. Several organizations recommend monitoring the percentage of metaphases with an apparent fra(3Kp14)as a quality control indicator for fra(X) studies. This approach assumes that stress systems work equally well on fra(3Np14) and fra(Xl(q27.3). In our experience with medium 199 and excess TdR, monitoring the frequency of fra(3)(p14)is not as consistent as scoring chromosome breakage (Table III), and in many cases that are fra(X) positive, fewer than 4% of cells express a fra(3)(p14) (Table 11). The incidence of fra(3)(p14)also varies significantly between medium 199 and excess TdR. Thus, we agree with Shapiro et al. [1991a] that it may not be reliable to monitor fra(3)(p14).Monitoring fra(3)(p14) may be more useful with other induction systems. For example, Daniel et al. [19841suggested that monitoring fragility at 3p14 may be a useful quality control indicatorfor FUdR induction systems. -
Fragile X Studies Tested in Routine Practice
SUMMARY We believe there is a need for a uniform set of guidelines for fra(X) studies among the certification agencies. Our findings suggest that many of the existing guidelines summarized in Table I work relatively well in routine practice. Our results suggest that the cytogenetic fra(X) method is most efficient and economical without sacrificing quality when the study involves the following criteria. 1. The constitutional karyotype is routinely established from a t least 20 banded cells and 2 karyotypes. 2. Two or more stress culture systems should be used. 3. Routinely, a total of 50 cells for males and 75 for females should be analyzed. 4. For patients with 1 to 3% cells with fra(X), more metaphases should be analyzed or the study should be repeated. Additional clinical, molecular, and cytogenetic approaches may also be considered (Table I). 5 . The monitoring of chromosome breakage may be an acceptable quality control indicator. However, each laboratory should establish its own statistical values, and it may be necessary to look for 3 or 4 cases in a row with low chromosome breakage to identify technical problems.
REFERENCES American Medical Association (1990): “CPT: Physicians’ Current Procedural Terminology.” 4th ed. Chicago: American Medical Association, p 500. Arinami T, Kondo I, Hamaguchi H, Tamura K, Hirano T (1987): A fragile X female with Down syndrome. Hum Genet 77:92-94. College of American Pathologists (1991):Section IX. Inspection checklist: Cytogenetics. In “Clinical Laboratory Improvement Manual 11: Checklist and Commentaries.” Chicago: College of American Pathologists, pp 1-11. Craig I(1991): Methylation and the fragile X. Nature 349:742-743. Daniel A, Ekblom L, Phillips S (1984):Constitutive fragile sites lp31, 3p14,6q26, and 16q23 and their use as controls for false-negative results with the fragile(X). Am J Med Genet 18:483-491. De Arce MA (1983): Tables for the cytogenetic study of fragile X chromosomes for diagnostic purposes. Clin Genet 24:320-323. Department of Health and Human Services (1991): Health Care Financing Administration. State Operations Manual No. 246. Filippi G, Pecile V, Rinaldi A, Siniscalco M (1988): Fragile-X mutation and Klinefelter syndrome: A reappraisal. Am J Med Genet 30:99-107. Glover TW (1981): FUdR induction of the X chromosome fragile site: Evidence for the mechanism of folic acid and thymidine inhibition. Am J Hum Genet 33:234-242. Higgins JV, Amarose AP, Dewald GW, Palmer CG, Sekhon GS, Stallard R, Bader (1990): “Great Lakes Regional Genetics Group Quality Assurance for Cytogenetics Laboratories.” pp 1-21. Howard-Peebles PN (1991):Fragile X expression: Use ofadouble induction system (letter to the editor). Am J Med Genet 38:445-446. Jacky PB, Ahuja YR, Anyane-Yeboa K, Breg WR, Carpenter NJ, Froster-Iskenius UG, Fryns J-P, Glover TW, Gustavson K-H, Hoegerman SF, Holmgren G, Howard-Peebles PN, Jenkins EC, Krawczun MS, Neri G, Pettigrew A, Schaap T, Schonberg SA, Shapiro LR, Spinner N, Steinbach P, Vianna-Morgante AM, Wat-
821
son MS, Wilmot PL (1991): Guidelines for the preparation and analysis of the fragile X chromosome in lymphocytes. Am J Med Genet 38:400-403. Jenkins EC, Krawczun MS, Stark-Houck SL, Duncan CJ, Kunaporn S, Gu H, Schwartz-Richstein C, Howard-Peebles PN, Gross A, Sherman SL, Brown WT (1991): Improved prenatal detection of Fra(Xl(q27.3):Methods for prevention of false negatives in chorionic villus and amniotic fluid cell cultures. Am J Med Genet 38:447-452. Knutsen T, Bixenman HA, Lawce H, Martin PK (1989):Chromosome analysis guidelines preliminary report. Karyogram 15131-135. Knutsen T, Bixenman HA, Lawce H, Martin PK (1991): Chromosome analysis guidelines preliminary report. Cancer Genet Cytogenet 52:ll-17. Krawczun MS, Jenkins EC, Brown WT (1985):Analysis ofthe fragile-X chromosome: Localization and detection of the fragile site in high resolution preparations. Hum Genet 69209-211. Lubs HA (1969): A marker X chromosome. Am J Hum Genet 21:231-244. New York State Department of Health (1990): Cytogenetics: Laboratory Survey Checklist Review Summary. DOH-1484:ll. Oberle I, Rousseau F, Heitz D, Kretz C, Devys D, Hanauer A, B o d J, Bertheas MF, Mandel J L (1991):Instability of a 550-base pair DNA segment and abnormal methylation in fragile X syndrome. Science 252:1097-1102. Opitz J M (1986): Editorial comment On the gates of hell and a most unusual gene. Am J Med Genet 23:l-10. Pacific Southwest Regional Genetics Network (1990): “Cytogenetics Quality Assurance Guidelines.” Health Resources and Services Administration, Department of Health Services, pp 1-46. Shapiro LR, Eallonardo SJ 111, Wilmot PL (1991a):Internal control for successful induction of the fragile X chromosome: Other folate sensitive fragile site expression eliminated (abstract).Cytogenet Cell Genet 56:221. Shapiro LR, Wilmot PL, Shapiro DA, Pettersen IM, Casamassima AC (1991b): Cytogenetic diagnosis of the fragile X syndrome: E f i ciency, utilization, and trends. Am J Med Genet 38:408-410. Sutherland GR (1979):Heritable fragile sites on human chromosomes. 111. Detection of fra(XMq27)in males with X-linked mental retardation and in their female relatives. Hum Genet 53:23-27. Sutherland GR (1985):The enigma ofthe fragile X chromosome. Trends Genet 1:108-112. Sutherland GR, Baker E (1990): The common fragile site in band q27 of the human X chromosome is not coincident with the fragile X. Clin Genet 37:167-172. Sutherland GR, Baker E, Fratini A (1985):Excess thymidine induces folate sensitive fragile sites. Am J Med Genet 22:433-443. Sutherland GR, Hecht F (1985): “Fragile Sites on Human Chromosomes.” New York Oxford University Press. Tommerup N, Laing S, Christensen IJ, Turner G (1988):Screening for the fragile X: How many cells should we analyse? Am J Med Genet 30417-422. l i m i e r G, Brookwell R, Daniel A, Selikowitz M, Zilibowitz M (1980a): Heterozygous expression of X-linked mental retardation and X-chromosome marker fra(X)(q27).N Engl J Med 303:662-664. Turner G, Daniel A, Frost M (1980b): X-linked mental retardation, macro-orchidism, and the Xq27 fragile site. J Pediatr 96:837-841. Voullaire LE, Webb GC, Leversha M (1989): Fragile X testing in a diagnostic cytogenetics laboratory. J Med Genet 26:439-442. Yu S, Pritchard M, Kremer E, Lynch M, Nancarrow J , Baker E, Holman K, Mulley JC, Warren ST,Schlessinger D, Sutherland GR, Richards RI (1991): Fragile X genotype characterized by a n unstable region of DNA. Science 252:1179-1 181.
Cytogenetic Guidelines for Fragile X Studies Tested in Routine Practice Gordon W. Dewald, Diane D. Buckley, J a c k L. Spurbeck, and Syed
M.Jalal
Cytogenetics Laboratory, Mayo Clinic and Mayo Foundation, Rochester, Minnesota
Several organizations have proposed guidelines for fra(X) studies on peripheral blood lymphocytes. To evaluate these guidelines, we reviewed 1,033 consecutive specimens referred for fra(X) analysis. Each specimen was cultured with medium 199 and RPMI 1640 with 5-fluorodeoxyuridine or excess thymidine. The karyotype and expression of fra(X) were established from 20 GTL- or QFQbanded cells and by screening of up to 130 more banded cells. We found anomalies other than fra(X) in 37 (3.6%) of the patients. We found 4% or more fra(X) cells in 38 (3.7%) cases from 36 unrelated families, including 33 (3.9%) of 850 males and 5 (2.7%) of 183 females. Another 4 females had 1 to 3% fra(X) cells. Six specimens were fra(X)-positive in only one stress system, and 32 were positive in 2 systems. To find the first 2 fra(X)cells in males, we needed to study up to 36 cells in 31 cases, 50 in one case, and 57 in another. To find the first 2 fra(X) cells in females, we needed to study up to 17 cells in 4 cases and 57 in another. A strong family history of fra(X) occurred in 5 patients, and each one was fra(X)-positive. Some manifestations of the fragile X syndrome occurred in 507 cases, 17 (3%)of which were fra(X)-positive. Abnormalities considered unlikely to be the fragile X syndrome occurred in 103 cases, 3 (3%)of which were fra(X)-positive. Use of chromosome breakage and fra(3Mp14)as quality control indicators of the fra(X) stress systems was found to be unreliable. Our findings support most of the proposed guidelines for fra(X) studies but indicate a need for modifications of others. 8 1992 Wiley-Liss, Inc.
KEY WORDS: Fragile X, fra(X) guidelines, Xq27.3
Received for publication June 24, 1991; accepted October 11, 1991. Addressreprint requests to GordonW. Dewald, Ph.D.,Cytogenetics Laboratory, Mayo Clinic, 200 First Street SW, Rochester, MN 55905.
0 1992 Wiley-Liss, Inc.
INTRODUCTION A fragile site at Xq27.3 [Krawczun et al., 19851 has been associated with the fragile X syndrome [Lubs, 1969; Sutherland, 1985; Opitz, 19861. The manifestations of the fragile X syndrome are well known, but they vary considerably among patients [Sutherland and Hecht, 19851. Thus, the use of clinical criteria alone to diagnose the fragile X syndrome can often be difficult. Consequently, many physicians routinely request fra(X) studies on all patients with mental retardation and patients with physical findings resembling those of the fragile X syndrome or a family history of X-linked mental retardation [Turner et al., 1980a,b; Shapiro et al., 1991bl. Several regional and international organizations have proposed guidelines for fra(X) studies (Table I). These guidelines vary considerably and are largely untested in routine practice. To help evaluate and formulate standards for routine practice, we reviewed 1,033 consecutive specimens referred to us for fra(X) studies between 1985 and 1990. We attempted to evaluate the following guidelines for routine practice: (1)the need for constitutional chromosome analysis, (2) the usefulness of multiple stress systems, (3) the optimal number of metaphases necessary for analysis, (4) the labor effort needed to conduct these studies, (5)the clinical reasons for referring specimens, and (6) the efficacy of recommended quality control procedures. METHODS We retrieved data from our laboratory computer system for all peripheral blood specimens referred for fra(X) studies from 1985through 1990. We also reviewed data for other blood specimens referred for routine cytogenetic studies during this same period which proved to be fra(X)-positive. We process all blood specimens for routine chromosome studies with medium 199 and RPMI 1640 to permit identification of fra(X) chromosomes. For each peripheral blood specimenreferred for fra(X) studies, we initiated cultures using 2 different stress systems within 48 hours of arrival of the specimen. The stress systems included medium 199 and either RPMI 1640 with 5-fluorodeoxyuridine(FUdR) or RPMI 1640 with excess thymidine (TdR). For the medium 199 method, 1 mL of heparinized whole blood was cultured
Fragile X Studies Tested in Routine Practice
817
TABLE I. Summary of Reported Guidelines for fra(X) Studies Guidelines
ACT
CPT
No. of cells Males Females Additional cells Standard banded study Induction systems Definition of fra(X) syndrome Quality control 26% cells with breakdgaps 24% cells with fra(3)(p14) Other
...
100
50 100 Yesb Yes Yes' 2 4%
.. ... ,
Yesg
...
... ...
Yes
...
... ... ... ...
IV fra(X) workshoD
... 100 150 Yes' Yes
2-3
CAP ... ...
... ... ...
...
24%
...
... ... ...
... ... Yesh
GLaRGG
... 50 100 Yesd Yes
22 24%'
Yes Yes
...
HCFA ... 50 100 Yesd Yes ...
...
Yes Yes
...
PSRGN
NYS'
74 ...
...
...
Yes' Yes 22
24%'
...
... Yes'
50 100 Yesd Yes 22 24%'
Yes Yes ...
New York State does not include criteria for fra(X)studies in their laboratory checklist, but they do recommend HCFA guidelines in their manual for inspectors. If one or a small number of positive cells is observed, it is recommended that a second blood sample be requested and/or that multiple induction methods be used. If only one positive cell is observed, score an additional 100 cells. If more than one cell but less than 3%of metaphases show the fragile X, the results should be regarded as equivocaland another blood sample requested. It may be necessary to request cessation of dietary folate supplementation, i.e., multivitamins. 'When a frequency of less than 4% is obtained, a repeat specimen should be requested; if no higher frequency of expression is obtained, then the results should be interpreted with caution and on a case-by-case basis. dIf one metaphase is positive for the fragile X chromosome, 50 additional cells should be scored. 'There is no recommendation on the number of induction systems, only that a proven induction system should be used. 'Results with less than 5% fragile X expression should be interpreted with caution. gIf a laboratory reports several consecutive negative studies in patients with a fragile X phenotype and/or positive family history, it is suggested that the culture system be checked by either splitting fragile X samples with a laboratory known to have success with the technique or by obtaining cells from a known fragile X positive individual to use a s a control. I,Appropriate QC systems may include correlation with phenotype and family history, folate-sensitive fragile sites. 'If there is less than 4% expression of known folate-sensitive sites, the results should be interpreted with caution. Abbreviations: ACT, Association of Cytogenetic Technologists; CF'T, Current Procedural Terminology; IV fra(X) Workshop, IV International Workshop on the Fragile X Syndrome and X-Linked Mental Retardation; CAP, College of American Pathologists; GLaRGG, Great Lakes Fkgional Genetics Group; HCFA, Health Care Financing Administration; PSRGN, Pacific Southwest Regional Genetics Group; NYS, New York State.
a
in 10 mL medium 199 with 5% fetal bovine serum for 72 to 96 hours [Sutherland, 19791.For the FUdR procedure, 1mL of heparinized whole blood was cultured in 10 mL of working RPMI 1640 medium and cultured for 48 to 72 hours. FUdR was then added (final concentration, 0.123 kg/mL)for an additional 18 hours [Glover, 19811. For the excess TdR method, 1mL of whole blood was cultured in 10 mL of RPMI 1640 for 48 to 54 hours. TdR was then added (final concentration, 250 kg/mL) for another 18to 24 hours [Sutherland et al., 19851. All cultures were incubated a t 37°C in 5% COO,5%02, and 90% NO.Standard cytogenetic procedures were used to harvest the specimens for chromosome analysis. In each case, we used GTL- or QFQ-banded metaphases for chromosome analysis. We routinely analyzed 20 banded metaphases and prepared 2 to 3 karyotypes to establish the constitutional karyotype. In both males and females, we routinely screened 30 t o 80 more metaphases for the presence of fra(X) or sex chromosome mosaicism. These metaphases were also banded to identify the X chromosomes and to distinguish between the common fragile site at Xq27.2 [Sutherland and Baker, 19901 and fra(XNq27.3). Our cytogenetic definition for the fragile X syndrome required the observation of 4% or more metaphases with a fra(X). We used particular caution to interpret and process cases with l to 3%cells with fra(X) and cases in which there was considerable risk for fragile X syndrome. We attempted to evaluate the usefulness of the guidelines for monitoring breaks and gaps and fragility at 3p14 to assess the potency of the stress systems. This was done by establishing the incidence and normal range for these criteria on 100 consec-
utive specimens and the fra(X1-positivecases. This work was done by a group of technologists who were specifically trained t o study chromosome breakage to ensure consistent and accurate scoring of breaks, gaps, and fra(3)(p14). RESULTS During this 6-year period of this investigation, we processed 1,085 specimens for fra(X1 studies, 1,033 (95%)of which had sufficient numbers of metaphases for a suitable study. Of the successful studies, 850 (82%) specimens were from males and 183 (18%)from females. Ages were provided for 829 males and 181 females. The mean age of the males was 8.8 2 8.7 years (range, newborn to 62 years) and that of the females was 12.2 2 11.5 years (range, newborn to 49 years). Each female withfra(X) was older than 15years. By comparison, only 8 (26%) of 31 males with fra(X) whose age we knew were older than 15 years. We studied an average of 52 ? 10 (range, 20 to 101) metaphases per male patient and 74 2 26 (range, 30 to 150)metaphases per female patient. Only 5 males and 3 females had fewer than 50 suitable metaphases for analysis. For each of these cases we recommended that another study be done. With use of the number of bands on chromosome 10, the band resolution for each case averaged approximately 500 bands per haploid chromosome set. Chromosome anomalies other than fra(X) occurred in 37 (3.6%) of the specimens. These abnormalities included structural anomalies in 29 patients (11balanced and 18unbalanced) and 8 patients with numeric anoma-
Dewald et al.
818 60 50
culture, harvest, and establish the constitutional karyotype based on analysis of 20 metaphases and 47.8 i 16.1 minutes to screen 30 cells for fra(X)and sex chromosome mosaicism. Thus, 154 2 0.56 minutes were needed per metaphase screened. The need for full-time equivalents for fra(X) studies increases significantly with higher numbers of analyzed metaphases. Two of the authors (D.D. Buckley and J.L. Spurbeck) restudied 30 fra(X)-positive cases to establish the incidence of chromosome breakage and fra(3Xp14) expression in each of the culture stress systems. Most chromosome breakage occurred in cultures with excess TdR and the least occurred with medium 199 (Table 11). By combining 2 stress systems, 27 of 30 cases would have had chromosome breakage in 6% or more cells. The expression of fra(3)(p14) was a less reliable quality control indicator. An observable fra(3)(p14) was most common in cultures with excess TdR. Cultures with medium 199 or FUdR were relatively poor inducer systems for fra(3Kp14).Even after combining 2 stress systems, only 9 of 30 specimens had 4% or more cells with an apparent fra(3)(p14). We studied chromosome breakage and fragility at 3p14 for medium 199 and excess TdR in a consecutive series of 100 blood samples (Table 111).The results indicated that the vast majority of specimens cultured in medium 199 would not meet the guidelines recommended by many organizations. More specimens cultured with excess TdR met the minimal criteria than those cultured with medium 199, but many specimens still would need to be judged unsatisfactory. The mean number of metaphases with chromosome breakage and fragility at 3p14 is much greater with excess TdR than with medium 199. However, the lower 95% limit for medium 199 and for excess TdR is 0% for both chromosome breakage and fragility at 3p14. This finding indicates that in routine practice, no threshold would be reliable to identify system problems for individual cases. We attempted to correlate the reason for referral with our chances of observing a fra(X). A strong family history of fra(X) was indicated in 5 specimens, and each one was fra(X)-positive. We also studied 13 unaffected relatives of these patients, and each one was fra(X)-negative. Some manifestations of fra(X) were indicated in 507 specimens; 17 13%) were fra(X)-positive. These anomalies were mainly mental retardation (207 patients), developmental delay (192 patients), autism (103 patients), and other anomalies (5 patients). Traits unlikely to be the fragile X syndrome were indicated in 103
El Cells with fragile X.
% Cells analyzed to find 2 fraaile X
40
30 20
10 n " 1 3
5
7
9 11 13 15 17 19 21 23 25 27 29 31 33 1 3
-1
33 males
5
5 females
Fig. 1. Percentage of fra(X)-positive cells and number of cells needed to find the first 2 cells with fra(X) in 33 males and 5 females.
lies. Among the specimens referred specifically for fra(X) studies, we found more than 4% fra(X)cells in 38 (3.7%)specimens. These patients included 33 males and 5 females from 36 unrelated families. We found 2 additional patients with more than 4% fra(X) cells from specimens referred for cytogenetic studies other than fra(X). Among the cases referred for fra(X) studies, 6 were fra(X)-positive in only one of 2 stress systems and 32 were positive in 2 stress systems. Four males were fra(X)-negative by medium 199 and fra(X)-positive by either FtMPI 1640 with FUdR or excess TdR.One male was fra(X)-positive by medium 199 and negative by excess TdR. One female was fra(X)-negativeby medium 199 and fra(X)-positive by FUdR. For each patient, we established the number of cells needed to find the first 2 fra(X) cells (Fig. 1).In 31 males, 2 fra(X) cells were observed in the first 36 cells. For the 2 other males, 50 and 57 cells were analyzed before 2 fra(X) cells were found. In 4 females, 2 fra(X) cells were found in the first 17 metaphases. In the other female, 57 cells had to be analyzed to observe 2 containing a fra(X). The mean percentage of fra(X)-positive cells was 18.5 ? 5.7% (range, 4 to 50%)among the fra(X) males and 12.1 2 5.8% (range, 6 to 22%)among the females (Fig. 1).In general, the higher the percentage of fra(X) cells in a culture, the fewer the number of cells needed to observe two fra(X)-positive cells. We considered the effort needed to screen cells for fra(X). Work-load recording indicated 432 minutes to
TABLE 11. Incidence of Cells With Chromosome Breakage and fra(3)(p14)in Cases That Are fra(X)-Positive Breakage Method Medium 199 FUdR" Excess TdRb Overall
No. of cases
29 12 16 30 FUdR, 5-fluorodeoxyuridine. bT&, thymidine.
In
6%cells, no. cases 17 9 16 27 2
Mean ? SD (range), % 9 2 7 (0-32) 10 2 6 (0-20) 21 k 12 (8-50) 13 t 7 (4-32)
In
fra(3Mp14) Mean c SD no. cases ___.___ (range), % 2 0.3 2 1 (0-4) 2 1.0 t l(0-4) 12 6.0 '. 5 (0-18) 9 2.0 2 2 (0-81 2
4%cells,
~
Fragile X Studies Tested in Routine Practice
819
TABLE 111. Chromosome Breakage and Fragility at 3p14 in a Consecutive Series of 100 Blood Specimens Medium 199
Metaphases with breakdgaps Mean 2 SD
Excess TdR
(%)
Range 95% normal limits % cases 26%"
Metaphases with fragility at 3p14 (%) Mean 2 SD Range 95% normal limits % cases 24%"
4 + 4 0-17 0-16 21
22
0 5 2
10
0-8 0-7
8
?
14
0-84
0-51 91 5
9
0-56 0-32 72
"Common guidelines for fra(X) studies (Table I).
specimens;3 (3%)were fra(X)-positive.No clinical information was provided in 405 cases; 13 (3%)were fra(X)positive. Four female patients in this series had 1 to 3%fra(X) cells. In one case, we found 3 of 100 metaphases with an apparent fra(X). In each of the other 3 cases, we found only 1 of 100 metaphases with an apparent fra(X). We have not received samples for further study from any of these patients.
DISCUSSION Several organizations have recommended guidelines for fra(X) studies (Table I). These organizations include the Great Lakes Regional Genetics Group (GLaRGG) [Higgins et al., 19901, Pacific Southwest Regional Genetics Network (PSRGN)[19901,Health Care Financing Administration (HCFA)[Department of Health and Human Services], Association of Cytogenetic Technologists (ACT) [Knutsen et al., 1989,19911, and the IV International Workshop on the Fragile X Syndrome and X-Linked Mental Retardation [Jacky et al., 19911. The Physicians' Current Procedural Terminology (CPT)code [American Medical Association, 19901 suggests criteria for fra(X) studies for billing purposes. The College of American Pathologists (CAP) [19911 has not yet dealt with this issue in its entirety. The New York State Department of Health (NYS) [19901 suggests some guidelines for fra(X) in their interpretive manual for inspectors. For each organization, the guidelines were derived from consensus agreement of committees comprised primarily of cytogeneticists. Reasons for Referral Our experience indicates that only about half of the specimens referred for fra(X) studies involve patients with any clinical manifestations (49%)or a family history (1.7%)of the fragile X syndrome. Thus, it may not be surprising that only 38 (3.7%)of the patients in our series had more than 4% cells with fra(X). It may also explain why we found another 37 (3.6%)of the patients who had chromosome abnormalities other than a fra(X). On the basis of a 3-year fra(X) study involving 1,012 patients with developmental delay, Voullaire et al. [1989] observed 31 (3.1%)patients with fra(X) and 53 (5.3%) with other chromosome abnormalities. Their study and ours agree with the recommended guidelines
(TableI) that in practice, in addition to fra(X)analysis, it is important to establish the constitutional karyotype of these patients. Our results suggest that about 93%of specimens referred for fra(X) studies do not have any apparent chromosome abnormality. This finding probably reflects the effort of physicians to rule out the fragile X syndrome as well as other chromosome abnormalities in all mentally retarded or autistic patients [Shapiro et al., 1991bl. In practice, this is probably a good policy because our studies identified many patients with different kinds of chromosome abnormalities. In addition, sporadic patients with Klinefelter syndrome [Filippi et al., 19881 and Down syndrome [Arinami et al., 19871 are also known to have the fragile X syndrome. A molecular test is available that will identify the fra(X) mutant gene [Oberle et al., 1991; Yu et al., 19911, although the molecular test would not detect other chromosome abnormalities and may not be diagnostic for the syndrome in every case [Craig, 19911. Thus, a molecular test may well be effective for screening high-risk individuals. Cytogenetic analyses will most likely continue to be used to provide an important diagnostic test in routine practice for patients with mental retardation of uncertain origin.
Number of Metaphases Analyzed Our work-load recording indicates that, on the average, it takes 432 minutes to culture, harvest, and establish the constitutional karyotype and another 1.54 minutes to find and screen each additional metaphase for fra(X). The analysis of more metaphases than is necessary t o diagnose this disease may be an excessive expense. Thus, it is important to establish the minimal number of metaphases for analysis without sacrificing accuracy in routine practice. Several organizations recommend that analysis of 50 metaphases in males and 100 in females (TableI). Some organizations recommend the analysis of far more metaphases [Jacky et al., 19911, whereas some cytogeneticists think that in routine practice it may not be necessary to analyze so many metaphases. To assess this issue empirically, we determined the number of metaphases needed to find 2 fra(X)-positivecells in each patient with the fragile X syndrome. We would have identified 27 of 33 males and 4 of 5 females by analyzing only 20 meta-
820
Dewald et al.
phases (Fig. 1).We would have found 32 of 33 fra(X) males and 4 of 5 fra(X) females by analysis of 50 metaphases; these cases meet the minimal criterion of 4% metaphases with a fra(X). Only one fra(X) male and one fra(X) female would have required the analysis of more than 50 metaphases. For each male and female in our series, fra(X) would have been diagnosed by analysis of only 57 metaphases. These results suggest that, at least for females, fewer metaphases probably could be analyzed in routine practice than indicated by the existing guidelines (Table I). Furthermore, they suggest that the analysis of at least 100 cells for males and 150 for females, as recommended by the IV International Workshop on Fragile X [Jacky et al., 19911,may be excessive for routine practice. Our data support those guidelines that recommend analysis of 50 cells in males and 74 in females (Table I). This conclusion assumes that if 1to 3% fra(X) cells are encountered in the initial analysis, the investigator will analyze 50 more cells or consider other options. Using statistical logic, some investigators have attempted to establish the number of metaphases necessary to analyze in order to find 4% fra(X) cells [De Arce, 19831.We are concerned that this approach may not be reliable because it is based on the perception that the cellular expression of fra(X) is random. The inability to observe a fra(X) in every metaphase of a patient with fragile X syndrome may be due more to an imperfect cytogeneticprocedure or possible intercellular variation in fra(X) gene amplification [Yu et al., 19911 than to predictable biologic criteria. Tommerup et al. [1988] attempted to determine the minimal number of cells needed for fra(X) investigations based on a retrospective analysis of fra(X) studies on population in Europe and Australia. They estimated that if these studies had been based on analysis of 50 cells per patient rather than 100 cells per patient, about 1.1 to 4.1% of fra(X) males and 2.6 to 5% of fra(X) females would not have been identified. These studies were done using only one stress system (medium 199). The results of our study and others indicate that the sensitivity of fra(X) studies is considerably enhanced by the use of two or more fra(X) stress systems used either separately or together [Howard-Peebles,1991; Jenkins et al., 19911. Our data suggest that it may be more important to use 2 or more fra(X) stress systems than it is to analyze more than 100 metaphases.
Multiple Induction Systems The oldest cytogenetic method used to process specimens for fra(X) utilizes folic acid-deficient medium 199. Many investigators think that culture medium supplemented with either FUdR [Glover, 19811or excess TdR [Sutherland et al., 19851 may be more reliable. Many cytogenetic laboratories use 2 or more stress systems in routine practice, and this method is consistent with most published guidelines for fra(X)studies (Table I). In our series, 6 of 38 specimens were positive in only one of 2 stress systems. The routine use of 2 stress systems for fra(X) studies provides an important quality assurance factor and helps identify certain cases that might be missed with asingle stress system.
Regardless of how carefully the cytogenetic fra(X) studies are done, we think the results will always be equivocal in a few patients. Among the specimens referred for fra(X) studies, we found four females with fewer than 4% fra(X)-positive cells. In such ambiguous cases it may be useful, as recommended (Table I), to analyze more metaphases, study additional blood specimens, investigate whether the patient is using folatesupplement therapy, and examine the patient and review the family history. In these cases, the new direct molecular tests for fra(X) may be particularly valuable. Alternatively, RFLP studies may yield important diagnostic information if there is an established history of the fragile X syndrome.
Quality Assurance Indicators To ensure that the fra(X) stress system is working, some laboratories process a control blood specimen or cells from a lymphoblastoid cell line from patients known to have more than 4% fra(X) cells. This approach requires either a readily available fra(X) donor or the maintenance of lymphoblastoid cell lines. In addition, it does not provide quality assurance for every case. Several organizations recommend that the observation of 6% or more cells with breaks or gaps is evidence that a fra(X) stress system is working. However, our findings suggest that the 6%threshold is not adequate for either medium 199 or excess TdR because many specimens do not achieve this level of chromosome breakage (Table 11). Furthermore, the apparent incidence of chromosome breakage varies among laboratories because of different culture conditions and even among technologists in the same laboratory because of the subjectivity of scoring breaks and gaps. Thus, it may be necessary for each laboratory to establish its own laboratory statistics for chromosome breakage in fra(X) stress systems. Our data suggest that it may not be possible to determine a reliable threshold for chromosome breakage for medium 199 and excess TdR because the lower 95% limit for both of these methods was 0% (Table 111). Thus, monitoring chromosome breakage to identify technical problems may not be reliable for individual cases. It may be better to look for three or four cases in a row with 0% chromosome breakage before becoming concerned about technical problems. Several organizations recommend monitoring the percentage of metaphases with an apparent fra(3Kp14)as a quality control indicator for fra(X) studies. This approach assumes that stress systems work equally well on fra(3Np14) and fra(Xl(q27.3). In our experience with medium 199 and excess TdR, monitoring the frequency of fra(3)(p14)is not as consistent as scoring chromosome breakage (Table III), and in many cases that are fra(X) positive, fewer than 4% of cells express a fra(3)(p14) (Table 11). The incidence of fra(3)(p14)also varies significantly between medium 199 and excess TdR. Thus, we agree with Shapiro et al. [1991a] that it may not be reliable to monitor fra(3)(p14).Monitoring fra(3)(p14) may be more useful with other induction systems. For example, Daniel et al. [19841suggested that monitoring fragility at 3p14 may be a useful quality control indicatorfor FUdR induction systems. -
Fragile X Studies Tested in Routine Practice
SUMMARY We believe there is a need for a uniform set of guidelines for fra(X) studies among the certification agencies. Our findings suggest that many of the existing guidelines summarized in Table I work relatively well in routine practice. Our results suggest that the cytogenetic fra(X) method is most efficient and economical without sacrificing quality when the study involves the following criteria. 1. The constitutional karyotype is routinely established from a t least 20 banded cells and 2 karyotypes. 2. Two or more stress culture systems should be used. 3. Routinely, a total of 50 cells for males and 75 for females should be analyzed. 4. For patients with 1 to 3% cells with fra(X), more metaphases should be analyzed or the study should be repeated. Additional clinical, molecular, and cytogenetic approaches may also be considered (Table I). 5 . The monitoring of chromosome breakage may be an acceptable quality control indicator. However, each laboratory should establish its own statistical values, and it may be necessary to look for 3 or 4 cases in a row with low chromosome breakage to identify technical problems.
REFERENCES American Medical Association (1990): “CPT: Physicians’ Current Procedural Terminology.” 4th ed. Chicago: American Medical Association, p 500. Arinami T, Kondo I, Hamaguchi H, Tamura K, Hirano T (1987): A fragile X female with Down syndrome. Hum Genet 77:92-94. College of American Pathologists (1991):Section IX. Inspection checklist: Cytogenetics. In “Clinical Laboratory Improvement Manual 11: Checklist and Commentaries.” Chicago: College of American Pathologists, pp 1-11. Craig I(1991): Methylation and the fragile X. Nature 349:742-743. Daniel A, Ekblom L, Phillips S (1984):Constitutive fragile sites lp31, 3p14,6q26, and 16q23 and their use as controls for false-negative results with the fragile(X). Am J Med Genet 18:483-491. De Arce MA (1983): Tables for the cytogenetic study of fragile X chromosomes for diagnostic purposes. Clin Genet 24:320-323. Department of Health and Human Services (1991): Health Care Financing Administration. State Operations Manual No. 246. Filippi G, Pecile V, Rinaldi A, Siniscalco M (1988): Fragile-X mutation and Klinefelter syndrome: A reappraisal. Am J Med Genet 30:99-107. Glover TW (1981): FUdR induction of the X chromosome fragile site: Evidence for the mechanism of folic acid and thymidine inhibition. Am J Hum Genet 33:234-242. Higgins JV, Amarose AP, Dewald GW, Palmer CG, Sekhon GS, Stallard R, Bader (1990): “Great Lakes Regional Genetics Group Quality Assurance for Cytogenetics Laboratories.” pp 1-21. Howard-Peebles PN (1991):Fragile X expression: Use ofadouble induction system (letter to the editor). Am J Med Genet 38:445-446. Jacky PB, Ahuja YR, Anyane-Yeboa K, Breg WR, Carpenter NJ, Froster-Iskenius UG, Fryns J-P, Glover TW, Gustavson K-H, Hoegerman SF, Holmgren G, Howard-Peebles PN, Jenkins EC, Krawczun MS, Neri G, Pettigrew A, Schaap T, Schonberg SA, Shapiro LR, Spinner N, Steinbach P, Vianna-Morgante AM, Wat-
821
son MS, Wilmot PL (1991): Guidelines for the preparation and analysis of the fragile X chromosome in lymphocytes. Am J Med Genet 38:400-403. Jenkins EC, Krawczun MS, Stark-Houck SL, Duncan CJ, Kunaporn S, Gu H, Schwartz-Richstein C, Howard-Peebles PN, Gross A, Sherman SL, Brown WT (1991): Improved prenatal detection of Fra(Xl(q27.3):Methods for prevention of false negatives in chorionic villus and amniotic fluid cell cultures. Am J Med Genet 38:447-452. Knutsen T, Bixenman HA, Lawce H, Martin PK (1989):Chromosome analysis guidelines preliminary report. Karyogram 15131-135. Knutsen T, Bixenman HA, Lawce H, Martin PK (1991): Chromosome analysis guidelines preliminary report. Cancer Genet Cytogenet 52:ll-17. Krawczun MS, Jenkins EC, Brown WT (1985):Analysis ofthe fragile-X chromosome: Localization and detection of the fragile site in high resolution preparations. Hum Genet 69209-211. Lubs HA (1969): A marker X chromosome. Am J Hum Genet 21:231-244. New York State Department of Health (1990): Cytogenetics: Laboratory Survey Checklist Review Summary. DOH-1484:ll. Oberle I, Rousseau F, Heitz D, Kretz C, Devys D, Hanauer A, B o d J, Bertheas MF, Mandel J L (1991):Instability of a 550-base pair DNA segment and abnormal methylation in fragile X syndrome. Science 252:1097-1102. Opitz J M (1986): Editorial comment On the gates of hell and a most unusual gene. Am J Med Genet 23:l-10. Pacific Southwest Regional Genetics Network (1990): “Cytogenetics Quality Assurance Guidelines.” Health Resources and Services Administration, Department of Health Services, pp 1-46. Shapiro LR, Eallonardo SJ 111, Wilmot PL (1991a):Internal control for successful induction of the fragile X chromosome: Other folate sensitive fragile site expression eliminated (abstract).Cytogenet Cell Genet 56:221. Shapiro LR, Wilmot PL, Shapiro DA, Pettersen IM, Casamassima AC (1991b): Cytogenetic diagnosis of the fragile X syndrome: E f i ciency, utilization, and trends. Am J Med Genet 38:408-410. Sutherland GR (1979):Heritable fragile sites on human chromosomes. 111. Detection of fra(XMq27)in males with X-linked mental retardation and in their female relatives. Hum Genet 53:23-27. Sutherland GR (1985):The enigma ofthe fragile X chromosome. Trends Genet 1:108-112. Sutherland GR, Baker E (1990): The common fragile site in band q27 of the human X chromosome is not coincident with the fragile X. Clin Genet 37:167-172. Sutherland GR, Baker E, Fratini A (1985):Excess thymidine induces folate sensitive fragile sites. Am J Med Genet 22:433-443. Sutherland GR, Hecht F (1985): “Fragile Sites on Human Chromosomes.” New York Oxford University Press. Tommerup N, Laing S, Christensen IJ, Turner G (1988):Screening for the fragile X: How many cells should we analyse? Am J Med Genet 30417-422. l i m i e r G, Brookwell R, Daniel A, Selikowitz M, Zilibowitz M (1980a): Heterozygous expression of X-linked mental retardation and X-chromosome marker fra(X)(q27).N Engl J Med 303:662-664. Turner G, Daniel A, Frost M (1980b): X-linked mental retardation, macro-orchidism, and the Xq27 fragile site. J Pediatr 96:837-841. Voullaire LE, Webb GC, Leversha M (1989): Fragile X testing in a diagnostic cytogenetics laboratory. J Med Genet 26:439-442. Yu S, Pritchard M, Kremer E, Lynch M, Nancarrow J , Baker E, Holman K, Mulley JC, Warren ST,Schlessinger D, Sutherland GR, Richards RI (1991): Fragile X genotype characterized by a n unstable region of DNA. Science 252:1179-1 181.
