| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Pediatric Endocrine Unit (M.M.) and Pediatric Surgical Research Laboratory (D.T.M., P.K.D.), MassGeneral Hospital for Children and Harvard Medical School, Boston, Massachusetts 02114; and Pediatric Endocrine Division (M.M.L.), Duke University Medical Center, Durham, North Carolina 27710
Address all correspondence and requests for reprints to: Mary M. Lee, M.D., Pediatric Endocrine Division, Box 3080, 308 Bell Building, Duke University Medical Center, Durham, North Carolina 27710. E-mail: Lee00140{at}mc.duke.edu.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
Although the role of MIS in the regression of Müllerian ducts during male sexual differentiation is well understood, its role in females is less clear. MIS induces follicular growth and differentiation (4, 5, 7) and inhibits recruitment of primordial follicles (25) in animal studies. In women, MIS is measurable in ovarian follicular fluid (26) and its concentration in serum fluctuates across the menstrual cycle, peaking in the late follicular phase (13, 14). Women with polycystic ovarian syndrome have higher serum MIS values, which correlate inversely with serum estradiol, indicating that MIS may play a role in regulating estradiol synthesis as well as in the disordered folliculogenesis of polycystic ovarian syndrome (26).
Because MIS is a hormone restricted in expression to the gonads and expressed at much higher levels in testicular than ovarian tissue, we reasoned that MIS determination would help differentiate between gonadal and nongonadal causes of mild virilization. MIS values would be low in female infants with a 46,XX karyotype and normal ovarian tissue who are virilized from in utero exposure to maternal androgens or elevated adrenal androgens as with congenital adrenal hyperplasia (CAH). On the other hand, serum MIS would be elevated above the normal female range if the virilization is caused by testicular androgens in a child with a 46,XY karyotype and male pseudohermaphroditism. With acquired virilization, MIS determination might help distinguish between gonadal androgens and those from other sources. For example, androgen-secreting ovarian tumors such as Sertoli-Leydig cell tumors and selected virilizing granulosa-theca cell tumors (27, 28) might express high levels of MIS (29, 30). Therefore, we examined serum MIS concentrations in relation to gonadal histology and cytogenetic findings in phenotypic girls with mild virilization (nonpalpable gonads, clitoromegaly, and partial labial fusion).
| Patients and Methods |
|---|
|
|
|---|
We identified 65 phenotypic females who presented with virilization between 1 d and 17 yr of age and had MIS measurements from December 1995 to April 2000. These patients were evaluated initially in the Pediatric Endocrine Unit or the Pediatric Surgical Unit of the MassGeneral Hospital for Children (n = 20), or at other institutions (n = 45). At Massachusetts General Hospital, the parents gave verbal consent for MIS determination, in accordance with approved institutional guidelines. However, because MIS determination is available as a standard diagnostic test, this was not uniformly obtained at other institutions. All patients underwent surgical exploration, gonadal biopsy, or diagnostic imaging. Clinical findings, cytogenetic analysis, operative details, and pathologic details of the gonadal biopsies were obtained from chart review and communications with the patients physicians.
Serum MIS assay
Serum MIS was measured by an ELISA as described in earlier studies (12, 16, 22). This assay does not cross-react with other members of this gene family, and is specific for primate MIS. The intraassay coefficient of variation for the assay is 9%, and the interassay coefficient of variation is 15%. The limit of sensitivity is 0.3 ng/ml. MIS values that fell below this (<0.3 ng/ml) were considered undetectable. Values of MIS that fell within the range for age-matched females were designated as normal, and MIS values above the female range were designated as elevated. We previously reported mean serum MIS values of 0.7 ng/ml (range, 0.21.9 ng/ml) in female infants, ages 212 months (12). By 18 months, MIS rises minimally to prepubertal values of 0.9 ng/ml (range, 0.23.9 ng/ml), then increases further to pubertal and adult values of 2.9 ng/ml (range, 0.28.9 ng/ml) (12). All MIS measurements were made on specimens that were freshly collected or stored for less than 2 yr.
Statistical analysis
The Students t test was used to compare means between groups. Sensitivity, specificity, and positive and negative predictive values with 95% confidence intervals were calculated using two-way contingency table analysis of Javastat, based on methods described by Fleiss (31), and the program Matlab (binofit.m) produced by The Mathworks Inc. (Newton, MA), which uses methods described by Johnson et al. (32).
| Results |
|---|
|
|
|---|
The subjects included 65 phenotypic females with mild degrees of virilization (Prader stages 1 and 2). All subjects were considered to have clitoromegaly by the examining physician. Some also had posterior labial fusion. The diagnoses, gonadal histology, karyotype, and MIS values are elaborated in Table 1
. Mean age was 2.4 ± 0.5 yr (range of 1 d to 17 yr), and most were prepubertal (61/65).
|
|
All subjects with ovarian tissue had individual MIS values that fell within the normal female range. The lowest mean MIS value was found in this gonadal category (Fig. 1
). The highest MIS values were found in 46,XX females with MISsecreting ovarian tumors, followed by patients with bilateral testes. The individual MIS values in both of these categories were clearly above the female range but within the normal male range (Fig. 2
). Subjects with bilateral or unilateral testes (n = 7) had higher MIS values than those with dysgenetic gonads or ovotestes (n = 22) (23.6 ± 11.1 ng/ml vs. 7.3 ± 1.6 ng/ml, P < 0.05, Fig. 1
). In this latter group with abnormal gonads, a number of individual MIS values fell within the normal female range, although the mean value was significantly higher than that of girls with ovaries (7.3 ± 1.6 ng/ml vs. 0.7 ± 0.1 ng/ml, P < 0.0001). Among those with an ovotestis, the presence or absence of Y chromosomal material did not alter the mean MIS value (7.5 ± 1.4 ng/ml vs. 8.1 ± 3.2 ng/ml).
|
|
All 28 subjects with elevated serum MIS had gonads containing testicular tissue or MIS-secreting ovarian tumors (Table 2
). Five of the seven subjects with testes were subsequently diagnosed with partial androgen insensitivity. The other two had 46,XX true hermaphroditism with unilateral testes. Eleven subjects had ovotestes, defined by the presence of both ovarian follicles and seminiferous tubules containing germ cells. Seven of these children had a 46,XX karyotype. The karyotype was unknown in one. The other three had the following karyotypes: 46,XY, 46,XX/46,XY, and 45,X/46,XY. The presence of ovarian follicles was unexpected in the patient with a 45,X/46,XY karyotype, raising the possibility of a gonadal 46,XX cell line. Seven subjects had dysgenetic testes, defined by the presence of dense fibrosis between tubules and absence of germ cells. Of these, the karyotype was 46,XY in three, 45,X/46,XY in three, and unknown in one. All three subjects with virilizing ovarian tumors had 46,XX karyotypes.
Among the 37 virilized subjects with MIS values in the normal female range, 19 had measurable values, whereas 18 had undetectable serum MIS (Table 2
). Among those with measurable MIS values, one 8 yr old with a 46,XX karyotype and an MIS value of 0.7 ng/ml had an ovotestis, and a 7 month old with a 45,X/46,XY karyotype and low MIS of 0.3 ng/ml had dysgenetic testicular tissue. The rest had presumably normal ovaries. None of the subjects in the group with undetectable MIS values had testicular tissue or dysgenetic gonads with testicular elements. Sixteen had ovaries, whereas two patients aged 2.5 and 4 yr with 45,X/46,XX karyotypes and clitoromegaly had ovarian dysgenesis without any testicular elements identified.
All 33 subjects with presumptive ovarian tissue (no testicular elements or dysgenetic gonads detected by imaging studies) had 46,XX karyotypes. Thirteen of these subjects were confirmed to have 46,XX CAH, whereas the remaining 20 girls had mild idiopathic virilization of unidentified etiology and low neonatal testosterone values. Further work-up was advised in these 20 girls, all of whom were classified as Prader 1 with clitoromegaly ± minimal posterior labial fusion.
Comparison of MIS and karyotype determination
Testicular tissue was found in all subjects in this series with Y chromosomal material (nine with a 46,XY karyotype and six with sex chromosome mosaicism). As anticipated from previous clinical observations, the presence of an Y chromosome was associated with the presence of testicular tissue (33, 34). The absence of Y chromosomal material in the peripheral karyotype, however, was not predictive of absent testicular tissue. In this series, 10 subjects with a 46,XX karyotype had testes or ovotestes. Therefore, MIS determination was compared with cytogenetic analysis for discerning the presence of testicular tissue (Table 3
). This analysis indicated that an elevated MIS value and Y chromosomal material were both highly specific and predictive for testicular tissue (Table 3
). However, the absence of Y chromosomal material was less sensitive and predictive for the absence of testicular tissue than MIS determination (negative predictive value of 72.9% vs. 94.6% (Table 3
).
|
| Discussion |
|---|
|
|
|---|
Across species, the expression of MIS is limited to the gonads, making it a highly specific marker of gonadal tissue. During early childhood, MIS expression is much higher in the testis than the ovary, with no overlap in mean serum concentrations (12). These characteristics of MIS expression, i.e. its sexual dimorphism and specificity for gonadal tissue, have facilitated the use of the MIS assay for evaluating children with diverse gonadal disorders (19, 20, 21, 22, 23, 24, 35). In boys with cryptorchidism, MIS determination reliably distinguishes between anorchia and undescended testes (22, 24), with the rare exception of a patient with the persistent Müllerian duct syndrome caused by an MIS gene defect (36, 37, 38). In children with intersex conditions, MIS determination helps ascertain the presence of testicular tissue and aids differential diagnosis (19, 21, 23, 35). This study extends these findings by focusing on a subgroup of patients with lesser degrees of genital ambiguity who were deemed to be phenotypic females with virilization.
The evaluation of virilized girls can be challenging, particularly in the prepubertal years when gonadotropins and sex steroids are nondiagnostic (39, 40). Cytogenetic analysis, although not indicative of a specific condition, can help direct further evaluation. If Y chromosomal material is detected, quantitative assessment of MIS secretion may help distinguish between conditions primarily affecting androgen synthesis or action (such as androgen insensitivity) from those due to abnormal formation and structure of the gonad such as a dysgenetic testes or anorchia (19, 21, 23). A normal or high normal MIS value is consistent with a structurally normal testis, whereas a low value may indicate an abnormality of testicular formation. One caveat must be kept in mind, however. In children with dysgenetic testes, serum MIS can fall either within the normal female range or above the female norms to the low normal male range (19, 22). Therefore, an MIS value within the normal female range in a virilized female does not exclude dysgenetic testicular tissue or ovotestis and cannot be used as confirmation of normal ovaries. In this situation, more extensive evaluation including assessment of basal or stimulated testosterone levels or imaging studies may help establish the diagnosis.
In virilized girls, it is not uncommon to find a 46,XX karyotype or a mosaic karyotype without Y chromosomal material. Only 15 of our subjects had Y material detected, whereas 48 had a 46,XX karyotype and 2 had mosaicism (45,X/46,XX). In this study, 13 of the 48 virilized patients with a 46,XX karyotype had a diagnosis of CAH, the predominant cause of virilization in 46,XX individuals. For this subset of patients, MIS determination would not provide additional useful information. In the remaining subjects, cytogenetic analysis is not discriminatory for gonadal pathology; for example, a 46,XX karyotype can be associated with normal ovaries, true hermaphroditism (41) or even 46,XX sex reversal (males with intact testes) (42). In a virilized patient with no Y sequences detected peripherally, however, MIS determination would help identify testicular tissue or ovarian tumors. Of the 12 patients with a 46,XX karyotype and elevated MIS in this study, 9 had either testes or ovotestes (all with diagnoses of true hermaphroditism) and 3 had ovarian tumors. In these subjects, the cytogenetic analysis was nondiagnostic, whereas MIS determination reliably predicted testicular tissue or a virilizing tumor.
Our data demonstrate that MIS has a high sensitivity, specificity, and positive and negative predictive values for detecting testicular tissue or an MIS-secreting virilizing ovarian tumor in phenotypic females with virilization. MIS determination is especially valuable during late infancy and the prepubertal years when gonadotropins and testosterone are nondiagnostic (42 , 43). MIS values that are elevated to the male range are consistent with the presence of testicular tissue or MIS secreting tumors, mandating additional evaluation and surgical exploration to verify the gonadal histology and resect testicular or tumor tissue. MIS can then be measured postoperatively to document complete removal of testicular tissue or tumor. Our study confirms the value of MIS determination in conjunction with cytogenetic, steroid, and radiologic studies in the evaluation of phenotypic females with virilization.
| Acknowledgments |
|---|
| Footnotes |
|---|
Abbreviations: AMH, Anti-Müllerian hormone; CAH, congenital adrenal hyperplasia; MIS, Müllerian inhibiting substance.
Received June 7, 2002.
Accepted November 12, 2002.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. A. Rey, E. Codner, G. Iniguez, P. Bedecarras, R. Trigo, C. Okuma, S. Gottlieb, I. Bergada, S. M. Campo, and F. G. Cassorla Low Risk of Impaired Testicular Sertoli and Leydig Cell Functions in Boys with Isolated Hypospadias J. Clin. Endocrinol. Metab., November 1, 2005; 90(11): 6035 - 6040. [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 |