| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Laboratoire d’Explorations Fonctionnelles Endocriniennes, Hôpital Trousseau, 75012 Paris, France
Address all correspondence and requests for reprints to: Dr. Christine Gicquel, Laboratoire d’Explorations Fonctionnelles Endocriniennes, Hôpital Trousseau, 26 Avenue Arnold Netter, 75012 Paris, France.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
TS was evaluated by Southern blot analysis of leukocyte DNA using a multiallelic polymorphic X chromosome marker (88% heterozygosity rate). Y chromosome PCR analysis was carried out if a pattern indicative of TS was obtained. Leukocyte DNA analysis produced an abnormal restriction pattern for 20 of the 375 cases (5.3%). There was a single hybridizing band in 13 cases, an allelic disproportion indicative of mosaicism in 6 cases, and 3 hybridizing bands in 1 case. One patient tested positive in the Y chromosome PCR analysis. Cytogenetic analysis showed 47 XXX trisomy in the patient with a 3-hybridizing-band pattern and confirmed the diagnosis of TS for 17 of the 19 suspected cases: 45 X: n = 7; 45 X/46 Xi(Xq): n = 4; 45 X/46 XX: n = 2; 46 Xi(Xq): n = 1; 45 X/46 Xr(X): n = 1; 45 X/46 XX/47 XXX: n = 1; 45 X/46 XY: n = 1. Cytogenetic analysis was normal (46 XX) for the 2 other patients.
The TS phenotype is variable: dysmorphism is often missing or mild (particularly in cases of mosaicism), but growth is reduced in virtually all patients. Screening of 375 growth-retarded girls identified 18 cases of TS, of which 17 were diagnosed by molecular analysis. This incidence (4.8%) was significantly higher than the expected incidence in this population (0.8–1.6%: P < 0.001).
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Recombinant DNA technology makes possible the analysis of sex chromosome abnormalities at the molecular level. It has helped to identify the parental origin of the abnormality, the existence of cytogenetically hidden mosaicism, and the correlation between genotype and phenotype (5, 6, 7).
Recognition of the disease is important, in view of GH therapy, as is precise diagnosis of the chromosome anomalies responsible for the varying prognoses, in terms of growth and fertility. Growth response would be negatively correlated with age at the start of therapy (8, 9), and so, early diagnosis of TS is important.
Short stature is the most common clinical feature of TS, and the aim of this work was to systematically search for TS in female patients referred on the basis of short stature. Three hundred and seventy-five patients were investigated. We tested for TS by Southern blot analysis of leukocyte DNA with a highly polymorphic marker of the X chromosome (10). Eighteen patients were diagnosed with TS (17 by molecular analysis). Fourteen of theses patients had no clinical features indicative of TS. One of these patients also tested positive for chromosome Y by PCR. The incidence of TS in this population (4.8%) was higher than the expected incidence of TS in a female population of short stature (0.8–1.6%). We could not define a population for molecular diagnosis screening based on clinical, auxological, or hormonal data; but this study suggests that patients with mild growth retardation should be tested for TS.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Clinical data. Growth-retarded patients referred to our center for TS or chronic illness (respectively, 23 and 42 patients for the same period) were not included in this study.
Three hundred seventy-five girls, 1 month to 18 yr old (mean ± SD = 9 7/12 ± 39/12), were referred to our center for investigation of growth retardation. Mean growth retardation was -2.57 SD ± 0.79 (range: -1 to -7). Forty-nine patients had growth retardation less than -2 SD, according to French growth charts (11), but exhibited a decreased height velocity.
Pubertal stage was available for 370 of the patients. Six percent (n = 23) were postpubertal and 25.7% (n = 95) were at puberty at the time of molecular analysis. Two patients exhibited precocious puberty.
Birth length was recorded for 330 of the 375 patients. Two hundred and sixty-four were full-term babies, and their mean birth length was 47.8 ± 2.05 cm. Twenty-five percent of the patients (n = 83/330) had birth lengths below the third percentile (12).
Mean maternal height (n = 363) was 157.5 ± 6.8 cm (138–180 cm), and mean paternal height (n = 360) was 169.7 ± 7.4 cm (150–190 cm). These parental heights were slightly lower than the mean French adult heights, which are 163.2 ± 5.6 cm for women and 174.5 ± 6 cm for men.
Eighty percent (n = 300) had no clinical features of TS. Twenty percent (n = 75) had clinical features consistent with TS. Sixty-three of these patients had one of the following features: mild cubitus valgus, short neck, widely separated nipples, pigmented nevi, or short 4th metacarpal bone. Twelve had a more suggestive Turner’s dysmorphia, with square thorax and two of the other features.
Hormonal data. GH provocative tests were performed in 329 patients (87.7%). The first test for most patients (97%) was an ornithine stimulation test. An impaired GH response to the first provocative test (peak GH level below 10 ng/mL) was found in 70 (21.3%) patients. A second test [insulin induced hypoglycemia or sequential test (arginine-insulin tolerance test or betaxolol-glucagon stimulation test)] was performed in 64 of these 70 patients. This led to the diagnosis of complete GH deficiency (peak GH level below 5 ng/mL) in 5 patients and the diagnosis of partial GH deficiency (peak GH level below 10 ng/mL) in 31 patients.
Thyroid function was investigated in 364 patients (97%) and was normal for all patients, except two with treated congenital hypothyroidism.
Methods
Molecular analysis
Preparation of genomic DNA.
Leukocyte DNA was isolated from 10-mL blood samples collected in
EDTA and extracted as previously described (10).
Southern-blot analysis
DNA (10 µg) was digested with 100 U EcoRI (New
England Biolabs, Beverly, MA), and the digested DNA was analyzed by
Southern blotting, as previously described (10).
Hybridization was carried out using the X-chromosome M27ß probe, which maps to Xcen-p11.22 in a variable number tandem repeat region and recognizes a highly informative (88%) multiallelic polymorphism (13).
Hybridization signals were quantified by densitometric analysis, using a GS700 imaging densitometer and the molecular analyst data system (Bio-rad, Rockland, CA), in comparison with two female reference DNAs (a homozygous 46 XX and a heterozygous 46 XX) and a male reference 46 XY DNA. A single band of reduced intensity (50% of the homozygous 46 XX control DNA) indicates the presence of a single X chromosome. Two bands of different intensities indicate mosaicism. The use of Southern-blot analysis with the M27ß probe has been previously shown to be accurate for the diagnosis of TS, the drawback being its inability to recognize the distal Xp deletion (10).
PCR analysis.
Four different loci of the Y chromosome, the
open reading frame of the SRY gene (Yp11.3), the ZFY gene (Yp11.3), the
DYZ3 sequence (Y centromere), and the repeated sequence DYZ1 (Yq12)
were studied. Sense and antisense primers were, respectively,
5'-GTCGCACTCTCCTTGTTTTTTGAC and 5'- CCGATTGTCCTACAGCTTTGTC for SRY (648
bp), 5'-AATTCATGAGGAGACC-AGAAG and 5'-CACAGAATTTACACTTGTGCAT for ZFY
(400 bp), 5'-ATGATA-GAACGGAAATATG and 5'-AGTAGAATGCAAAGGGCTCC for DYZ1
(170 bp), 5'-TCCACTTTATTCCAGGCCTGTCC and 5'-TTGAATGGAATGGGAACGAATGG for
DYZ1 (154 bp).
Amplified samples were subjected to electrophoresis in a 1.5% agarose gel, stained with ethidium bromide.
Karyotype. Karyotype analysis and fluorescence in situ hybridization (FISH) were done as previously described (14, 15). X chromosome FISH analysis was performed using a DXZ1 probe.
Statistical analysis. If binary variables were examined, between-group comparisons were made using a chiquare test for two independent samples. The Fisher’s exact test was used if necessary.
Comparison of the distribution of a continuous variable was tested using the nonparametric Kolmogorov-Smirnov test.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Normal restriction patterns were found in 355 (94.7%) of the
patients (Table 1
). Eighty-eight percent
of these patients (n = 311) were heterozygous for the X chromosome
marker, with two bands of the same intensity, and twelve percent
(n = 44) were homozygous (Fig. 1).
|
|
).
Abnormal restriction patterns were found in 20 (5.3%) of the patients
(Table 1).
Thirteen had a single hybridization band of 50% reduced intensity,
compared with the homozygous reference 46 XX DNA (Fig. 1). Cytogenetic analyses are presented in
Table 1. One of these patients had a normal chromosomal formula (false
positive).
In 6 cases, there were 2 bands of different intensities, suggesting
that there was mosaicism (Fig. 1). The karyotype analysis of 5 of these
6 patients confirmed the mosaicism TS: 45 X/46 XX, 45 X/46 XX, 45 X/46
Xr(X), 45 X/46 Xi(Xq), and 45 X/46 XX/47 XXX (Table 1), with
(respectively) 15%, 50%, 60%, 50%, and 90% blood cell population
carrying 45 X. The 2-bands pattern of the 45 X/46 Xi(Xq) mosaicism is
caused by a dicentric isochromy of the long arm of X chromosome with
maintenance of a part of the short arm. In the sixth case (false
positive), the restriction pattern was indicative of a mosaicism with
10% 45 X cells, but karyotype analysis of 48 cells was normal.
In 1 case, there was a three-band pattern (Fig. 1). Karyotype analysis
showed there was a 47 XXX trisomy (Table 1).
These data also indicate the good accuracy of molecular diagnosis, with a sensitivity of 94.7% (5.2% false negatives) and a specificity of 97.7% (2.3% false positives).
Y chromosome PCR analysis
PCR analysis was performed for patients with an X-chromosome restriction pattern indicative of TS. One patient tested positive for the Y chromosome, by PCR analysis, with the characteristic bands of the four Y loci studied. There was no clinical or biological hyperandrogenism in this patient. Karyotype analysis with Y chromosome FISH detected a 10% 46 XY cell population.
Correlation of clinical, auxological, hormonal, molecular, and cytogenetic data
Differences in age at diagnosis (Fig. 2A), birth length (Fig. 2B) or severity
of growth retardation (Fig. 2C), were studied in patients diagnosed
with TS and growth-retarded girls without TS. We found that TS patients
who were undiagnosed at birth were referred late to our center (mean
age at diagnosis: 98/12 ± 4 yr) at the same age than
patients without TS (97/12 ± 310/12 yr). Birth
length was normal for 43.75% of TS patients, and there was no
significant difference in mean birth length between the two groups.
However, Fig. 2B shows a bimodal distribution for birth length in TS
girls with 50% (8/16) less than 46 cm. The growth retardation of 33%
of TS patients was within 2 SD of normal, and mean growth
retardation was not significantly different between the two groups.
Finally none of these variables made it possible to select a defined
population for molecular analysis.
|
). Karyotype analysis was performed in
105 of 375 patients. The variable percentage of patients that had
karyotypes in each group (20.3% in group A, 54% in group B, and
83.3% in group C) was determined by a more severe growth retardation
and/or TS-like feature(s). There was a significantly higher rate of TS
(P < 0.001) in patients with TS-like feature(s)
(groups B and C), even if only one feature (group B) was present.
However, 7 TS (including 45 X monosomy) were found in patients without
any clinical features indicative of TS, except growth retardation
(Table 2).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We have previously demonstrated the routine use of Southern-blot analysis for diagnosis of TS and used the same tool to screen for TS in a growth-retarded female population. In the previous study, the sensitivity of the test was estimated to be 92% and its specificity, 100% (10).
The frequency of TS is between 1/2500 and 1/5000 female live births
(1), and growth is reduced in virtually all TS patients. Growth
retardation (
-2 SD) affects 2.5% of the whole female
population; thus, the theorical frequency of TS in growth-retarded
girls was estimated to be between 0.8 and 1.6%. This theorical
frequency is actually overrated for two reasons: 1) 20% of TS cases
are diagnosed at birth; and 2) patients referred to our center for TS
were excluded, and only patients referred for growth retardation were
included in this study. Thus, the expected frequency of TS in
growth-retarded girls is probably below 1%. A Scottish study (17)
of 2- to 4-yr-old children used diagnostic tests (and particularly
conventional chromosome analysis) on girls with heights below -2.5
SD. With these criteria, one TS case was detected in a
group of 208 growth-retarded girls.
In this study, we diagnosed 18 (4.8%) cases of TS in a group of 375 girls, 17 of which were diagnosed by molecular analysis. This incidence is significantly higher than the expected incidence (0.8–1.6%; P < 0.001).
Many authors have suggested that 45 X monosomy is lethal and that most surviving 45 X individuals are actually mosaicisms undetected by cytogenetic tests. The prevalence of micromosaicism recorded for TS patients depends on the methods used and the number of tissues screened (7). 45 X monosomy is usually detected in 50% of patients with TS, and mosaicism and X anomaly are detected in the other half. In this study, mosaicism and X anomaly were detected in 61.1% of the cases of diagnosed TS. These data are consistent with the fact that we selected patients on the basis of growth retardation and that 45 X monosomy is usually diagnosed in the first few years of life because of dysmorphia. However, we detected 45 X monosomy in two patients who had no clinical features consistent with TS (except growth retardation), one of them being at puberty. We also diagnosed five patients with isochromy of the long arm of the X chromosome who had no (n = 1) or only mild (n = 4) Turner’s phenotype. The dissociation between karyotype and phenotype is probably caused by a tissue-confined mosaicism, as previously described in monozygotic triplets (18).
To preselect a group of patients for molecular analysis, we assessed clinical data (such as birth length, age, and severity of growth retardation at the time of molecular diagnosis) and hormonal data in the patients who were diagnosed with TS and in all the other patients with growth retardation. None of these features might allow us to define a group of patients for this analysis. This study showed that 43.75% of patients diagnosed with TS had birth lengths within the normal range. It also showed that 33.3% had moderate growth retardation (within 2 SD of normal). There was impaired GH response to provocative test for 4 of 17 (23.5%) TS patients, and 2 of them were overweighted. Nevertheless, there was a higher frequency of TS (12.7%) for patients with only one of the features of TS (63 patients in this study, group B).
The presence of a Y chromosome or Y-chromosome derivative material in TS is correlated with the risk of gonadoblastoma, and prophylactic gonadectomy could be recommended in these cases (19). Presence of Y-chromosome material accounts for approximately 5% of TS cases (20, 21), but the observed frequency of Y-chromosome material also depends on the methods used (22). The presence of a normal Y chromosome in 10% of blood cells and the lack of clinical virilization and hyperandrogenism in our patient suggest a very low proportion of Y chromosome cells in gonads at the time of gonadal differentiation.
In conclusion, the incidence of TS in this population (4.8%) was higher than the expected incidence of TS in a female population of short stature (0.8–1.6%). It remains to be established whether this higher incidence of TS in patients with growth retardation justifies a systematic screening for TS in growth-retarded girls. In this regard, molecular analysis of the X chromosome is a valuable alternative to cytogenetic analysis in cases where growth retardation is not accompanied by Turner’s phenotype.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received September 17, 1997.
Accepted February 3, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S van Buuren, P van Dommelen, G R J Zandwijken, F K Grote, J M Wit, and P H Verkerk Towards evidence based referral criteria for growth monitoring Arch. Dis. Child., April 1, 2004; 89(4): 336 - 341. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
