Division of Reproductive Medicine, Department of Obstetrics and Gynecology (J.S.E.L., B.C.J.M.F.), Erasmus University Medical Center Rotterdam, 3015 GD Rotterdam, The Netherlands; Department of Hormonology (S.L.), University Hospital Montpellier, 34295 Montpellier, France; Department of Hormonology and Unit of Pediatric Endocrinology and Gynecology, Department of Pediatrics (C.S.), University Hospital Montpellier, 34295 Montpellier, France
Address correspondence and requests for reprints to: Joop S. E. Laven, M.D., Ph.D., Division of Reproductive Medicine, Department of Obstetrics and Gynecology, Erasmus University Medical Center Rotterdam, Dr. Molewaterplein 40, 3015 GD Rotterdam, The Netherlands. E-mail: laven{at}gyna.azr.nl
McCune-Albright syndrome is a sporadic disease
characterized by polyostotic fibrous dysplasia,
café-aux-lait lesions, and a variety of endocrine
disorders (1, 2). The molecular basis of this syndrome has
recently been elucidated. Missense point mutations in the GNAS1 gene
located on the long arm of chromosome 20 and encoding for the
subunit of Gs (the G protein that stimulates
cAMP) of transmembrane glycoprotein receptors have been identified
(3, 4). Mutations at codon 201 substituting Arg with
either Cys or His give rise to abnormal Gs
proteins that reduce the
intrinsic guanosine triphosphatase activity, thereby constitutively
activating the Gs protein. The mutation is found in variable abundance
in different endocrine and nonendocrine tissues, consistent with the
mosaic distribution of abnormal cells generated by a somatic cell
mutation early in embryogenesis. Severe disease may be associated with
an earlier mutational event leading to more widespread distribution of
mutated cells (5).
The most commonly encountered endocrine dysfunction in McCune-Albright syndrome is gonadal hyperfunction. Precocious puberty represents the usual initial manifestation of McCune-Albright syndrome in girls. Ovarian cysts may be found on pelvic ultrasound (6, 7, 8). Other endocrine abnormalities include hyperfunction of the thyroid and adrenal cortex, as well as excessive GH secretion. The majority of patients have abnormally elevated sex steroids with low or undetectable gonadotropin levels (5). Whereas pregnancies have been described later in life (9, 10), polymenorrhea and amenorrhea due to continued gonadotropin-independent estrogen production have also been reported (11). However, clinical information regarding ovarian dysfunction in McCune-Albright patients during adolescent and adult life is scant.
Case Report
A 22-yr-old patient previously diagnosed as McCune- Albright attended our outpatient clinic for fertility counseling. She exhibited the classical clinical triad of polyostotic fibrous dyplasia along with large café-aux-lait spots in the lumbosacral region and a history of precocious puberty and irregular menstrual bleeding.
Computed tomography scans showed fibrous dysplastic bone in the left humerus as well as in the sphenoid and maxillary sinus. The field of vision of the left eye was restricted due to facial bone involvement. Furthermore, she complained of recurrent maxillary sinusitis on the left side. At the age of 20 she underwent surgery in which dysplastic bone was removed from her maxillary and ethmoïd sinus on the left side. At age 21 she fractured her left clavicle following an accident with a horse. Healing was markedly delayed. Several typical café-aux-lait spots were located in the lumbosacral region, having a triangular shape, predominantly located at the left side and extending from L4 until S3, as well as in the neck and at the flexor surface of the lower left leg.
Menarche occurred at age 5 along with left-sided breast enlargement and development of pubic hair. At that time she exhibited low serum FSH and LH levels, whereas estradiol (E2) was in the normal adult range. Bone age and height were normal at the onset of menstruation. Symptoms did not progress during 5 yr of treatment with cyproterone acetate (12). Sexual maturation started after cessation of cyproterone acetate at 10 yr of age and was completed at 15 yr of age. Thereafter, several combined steroid contraceptive pills were prescribed for irregular menstrual bleeding without success. As far as she could remember, her menstrual cycle had been irregular (bleeding interval, 12 weeks) throughout life. During several periods without hormonal contraception she had unprotected intercourse with different male partners without conceiving a pregnancy. Before consultation she also suffered from intermittent pelvic pain predominantly on the right side. On physical examination, her height was 175 cm and body weight 58 kg. Pubic hair and breast development, as well as the appearance of her external genitals, were according to Tanner stage V. On transvaginal pelvic ultrasound several cysts were observed only in the right ovary, together with a thickened endometrial lining of the uterus and engorged uterine veins. At the time of referral, increased serum E2 (805 pmol/L), normal FSH (2.6 IU/L), LH (3.0 IU/L), PRL (3.6 µg/L), TSH (1.6 mU/L), androstenedione (9.4 nmol/L), and dehydroepiandosteronsulphate (2.4 µmol/L) levels were found.
Materials and Methods
Ultrasonography
Transvaginal ultrasound was carried out on initial screening and at 2-day intervals during two months. Ovarian volume, number of follicles and cysts, features of ovulation, and endometrial thickness (anterior and posterior layers measured in the longitudinal axis) were recorded. For sonographic imaging we used a 6.5-MHz vaginal transducer (model EUB-415; Hitachi Medical Corporation, Tokyo, Japan), as described previously (13, 14).
Source of tissue and preparation of cells
In the beginning of a bleeding period (day 3) laparoscopy and dilatation and curettage were performed to collect ovarian and endometrial biopsies. Before the laparoscopy, several ovarian cysts were punctured separately using transvaginal ultrasound guidance. Ovarian biopsies were taken from both ovaries for genetic analysis using monopolar scissors. Endometrial biopsies were taken using a Pipell microcurette (Laboratoire CCD, Paris, France). Samples were taken from the anterior and posterior endometrial wall. All tissue samples as well as the aspirates from cysts were placed in in vitro fertilization (IVF) medium immediately after collection. Aspirates were centrifuged at 3000 x g for 10 min, and the pellets of cells were reemerged into IVF medium. The remaining fluid was analyzed for its hormonal content.
Light microscopy
Sections of both ovaries and endometrium obtained by biopsy were prepared in neutral buffered formalin and embedded in paraffin. Thereafter, 10-µm thin sections were cut and subsequently stained with hematoxilin and eosin (15).
Hormone assays
Blood samples were obtained by venepuncture during the initial visit and at 2-day intervals during two monitored months. They were processed within 2 h after withdrawal. Serum and aspirates from the cysts was stored at -20 C. Serum was assayed for FSH, LH, E2, progesterone (P), testosterone, androstenedione, sex hormone-binding globulin, inhibin A, and inhibin B as described previously (16, 17). Normal serum values were obtained from previous longitudinal studies in 42 normo-ovulatory volunteers (17, 18). Follicular fluid was assayed for E2 and P. Normal follicular fluid values were obtained from normo-ovulatory volunteers (19).
DNA analysis
DNA was extracted from blood lymphocytes, endometrium, left and right ovarian tissues, and fluid obtained from ovarian follicles and cysts using commercial kits (QIAGEN, Courtaboeuf, France). PCR was performed on extracted DNAs. With the exception of minor modifications, we have used a method described previously (20) for selective enrichment of mosaic Arg 201 mutations.
Informed consent
The local Institutional Review Board was informed of the investigations being carried out. Because only one fully informed patient was involved, a research potocol approval was not required.
Results
Light microscopy
Microscopic analysis of ovarian biopsies showed primordial,
primary, and secondary follicles along with Graaffian structures most
pronounced in the right ovary. Although all stages of follicular
development were present, larger follicles were luteinized. Secretory
as well as proliferative elements were present side by side in the
endometrium (Fig. 1, A and B).
|
Direct DNA sequencing showed the presence of a guanine to adenine transversion leading to an Arg-His substitution at position 201 in the anterior endometrial lining and right ovary. The mutation was neither detectable in the left ovary nor in fluid obtained from ovarian cysts, in the posterior endometrial lining, or in blood lymphocytes.
Both canonical and mutant sequences were present. The latter with a weaker intensity, indicating a somatic mutation and, thus, a mosaicism of normal and abnormal cells as observed in McCune-Albright syndrome.
Cycle 1
In the beginning of this cycle the maximum FSH level observed was
4.0 IU/L (median during early follicular phase of normo-ovulatory
controls, 5.2 IU/L; range, 2.511.2), whereas LH levels were increased
at 15.5 IU/L (median during early follicular phase of normo-ovulatory
controls, 3.2 IU/L; range, 1.910.1) (17, 18). The
initial ultrasound investigation on day 5 showed four dominant
follicles in the enlarged (58 mL) right ovary with a diameter ranging
between 14 and 23 mm. Beyond day 9 these follicles ceased growing at a
diameter beyond 25 mm. In contrast, single dominant follicle selection
and normal development could be observed in the normal sized (18 mL)
left ovary. At a follicle diameter beyond 20 mm signs of ovulation (a
sudden decrease in follicle diameter and the appearance of free fluid
in the pouch of Douglas) could be observed. This occurred once between
the 10th and 15th day and once on day 23. E2
levels were 971 pmol/L on cycle day 7 (median mid follicular phase in
controls, 220 pmol/L; range, 91462) and decreased to 467 pmol/L on
day 16. Thereafter, an increase was observed up to 930 pmol/L before
menses (median late luteal phase in controls, 169 pmol/L; range,
144228). P levels were as high as 9.2 nmol/L in the beginning of the
cycle (median early follicular phase in controls, 1.0 nmol/L; range,
0.29.0) and rose to a peak value of 30.2 nmol/L on day 12 (median mid
luteal phase in controls, 49.5 nmol/L; range, 16.760.8).
Subsequently, P levels decreased to a value of 4.5 nmol/L on cycle day
23. At day 23, LH and FSH levels were 4.3 IU/L and 3.4 IU/L,
respectively. The total endometrial thickness was 10 mm on cycle day 10
and decreased to 4.5 mm on day 20. These results are depicted in Fig. 2A.
|
During the second cycle, three distinctive rises in FSH were
observed on cycle day 12 (6.8 IU/L), cycle day 24 (5.9 IU/L), and cycle
day 39 (12.3 IU/L). Similarly, rises in LH were observed on days 12,
24, and 39, reaching values of 12.1 IU/L, 11.7 IU/L, and 50.0 IU/L,
respectively. Inhibin B serum concentrations were only determined
during the second cycle because the sample volume during the first
cycle was insufficient. Inhibin B levels increased from early
follicular phase levels of 76 ng/L (median early follicular phase in
controls, 40 ng/L; range, 3260) to a maximum of 220 ng/L on cycle
days 15 and 17 (median late follicular phase in controls, 170 ng/L;
range, 98223). Thereafter, inhibin B fell to a nadir level of 45 ng/L
on cycle day 30 with a subsequent rise to 96 ng/L on cycle day 37.
Inhibin A levels were 28 ng/L on cycle day 3 (median early follicular
phase in controls, 7 ng/L; range, 1.110.2) and gradually increased to
a maximum value of 49 ng/L on cycle day 32 (median mid to late luteal
phase in controls, 35 ng/L; range, 1643). Thereafter, inhibin A
decreased to a level of 4 ng/L on cycle day 39. In the right ovary,
multiple dominant follicle development was observed and again growth
was arrested at a diameter around 25 mm. In the left ovary, single
dominant follicle growth was noted later during the cycle with a
maximum diameter of 22 mm on day 32. E2 levels
were 521 pmol/L on cycle day 3 and remained fairly constant until cycle
day 24. Thereafter, E2 levels fell gradually to a
nadir of 69 pmol/L on cycle day 37. P levels were 4.2 nmol/L on cycle
day 3 and remained fairly constant during the follicular phase. On
cycle day 24 a rise in P levels could be noted up to a peak value
of 44.1 nmol/L reached on cycle day 32. A sharp decline was noted
between days 32 and 39 in this cycle to a level of 5.8 nmol/L.
Endometrial thickness was 3.7 mm at initial screening and increased
gradually to 8.4 mm. On ultrasound examination a triple line was
observed throughout the cycle. During menses the endometrium remained
8.4 mm thick. These results are depicted in Fig. 2B.
Follicular fluid
In the right ovary five follicles measuring 22, 13, 12, 10, and 10
mm were punctured whereas in the left ovary three follicles with a
diameter of 5, 5, and 4 mm were aspirated separately.
E2 and P levels were compared with values
obtained from similar follicles from regularly cycling women
(19). Results are summarized in Fig. 3. Intrafollicular
E2 and P concentrations were increased in small
follicles (<10 mm) but diminished in large follicles.
|
Our findings represent the first longitudinal assessment of ovarian dysfunction in an adult patient suffering from McCune-Albright syndrome. The anticipated phenotype is the development of multiple dominant follicles as a result of increased FSH signaling. Indeed, FSH levels are increased in mothers of dizygotic twins resulting from multiple ovulations (21). Next to the development of multiple preovulatory follicles, premature luteinization and follicle maturation arrest may also be anticipated in this patient due to increased LH receptor signal transduction. This latter phenomenon may be comparable with a premature rise in serum LH observed during initial protocols for ovarian hyperstimulation for IVF without GnRH agonist cotreatment (22).
High FSH levels occurring during the luteo-follicular transition give rise to continued growth of a limited number (cohort) of follicles (23). Subsequent development of this cohort during the follicular phase is dependent on continued stimulation by gonadotropins. FSH levels decrease during the follicular phase due to negative feedback by ovarian inhibin B and E2 synthesis. Except for the dominant follicle remaining, follicles enter atresia due to insufficient support by reduced FSH levels (24). During the late follicular phase, aromatase enzyme activity of granulosa cells from the dominant follicle is also stimulated by LH (25). Under normal conditions, a good correlation between dominant follicle diameter and follicle fluid or serum E2 levels is observed (18, 19).
Increased FSH receptor signaling induced multiple dominant follicle development in the right ovary of the current patient. Consequently, E2 levels were increased at the beginning of the cycle. In McCune-Albright patients, cyst-like structures produce E2 in vitro comparable with normal preovulatory follicles. In contrast, small follicles synthesize substantially more E2 compared with these cysts (26). In the current study, follicular fluid E2 levels were increased in small follicles and decreased in preovulatory follicles compared with normal control subjects. Because McCune- Albright patients respond well to treatment with aromatase inhibitors (12), it might be speculated that aromatase is overexpressed in granulosa cells of McCune-Albright patients due to constitutive FSH signaling resulting in supraphysiological intrafollicular E2 concentrations.
Follicular fluid P levels were also increased in small follicles whereas in larger preovulatory follicles P concentrations were normal. Due to continuous LH receptor activation, P is synthetized prematurely by small follicles. Increasing P production is accompanied by decremental E2 synthesis due to luteinization of granulosa cells. Consequently, growth is arrested, atresia occurs, and these follicles become cysts due to premature P exposure.
In the left ovary (without the mutation), normal single dominant follicle growth and normal intrafollicular steroid levels were found. The follicle reached a normal preovulatory diameter and subsequently showed signs of ovulation. In contrast to the luteo-follicular rise in FSH during the normal cycle, a dominant follicle emerged in our patient after a distinct rise in serum FSH during the midfollicular phase of the cycle. This rise was accompanied by a transient increase in inhibin B like in the normal cycle. While the dominant follicle is growing, E2 output increases coinciding with the rise in inhibin A (27). Following ovulation, serum P levels increased along with an increase in inhibin A, suggesting normal corpus luteum function. Subsequently, P, inhibin A, and E2 levels decrease, constituting a pattern comparable with luteolysis in the normal cycle. The overall hormonal pattern was virtually the same in both cycles monitored.
It seems that some negative feedback activity of inhibin B on endogenous FSH levels remains because FSH serum levels decreased after a rise in inhibin B levels during the midfollicular phase. LH levels were very low throughout the cycle. There is, however, no consistent relationship between LH concentrations and either E2 levels or P levels indicating appropriate feedback. Collectively, these data suggest that the feedback mechanism of the pituitary gonadal axis for FSH is functioning properly in McCune-Albright patients.
Patients presenting with a Gs "gain of function" mutation
represent a "human model" to study increased signal transduction of
gonadotropin receptors. The only activating FSH receptor mutation was
identified in a hypophysectomized male who remained fertile despite
undetectable gonadotropin levels (28). Activating LH
receptor mutations have been described in male-limited precocious
puberty, whereas they do not seem to have any particular phenotype in
females (29). Hence, in females with McCune-Albright
syndrome symptoms might be due mainly to increased FSH signal
transduction. Although the effects of activating FSH receptor mutations
in the context of normal pituitary function in the female are not
known, they might resemble the phenotype of McCune-Albright patients.
Indeed, enlarged ovaries with multiple cysts have been described in
women with FSH-producing pituitary tumors (30). Finally,
FSH receptor polymorphism seems to be associated with the amount of
exogenous FSH required for adequate ovarian stimulation for IVF
(31).
This patient showed a typical unilateral involvement of tissue. The
mutation was only found in the right polycystic ovary. This might be
due to the absence or a smaller number of mutated cells present in the
left ovary not being detected by DNA analysis. This difference in
expression of the mutation is compatible with "normal"
monofollicular growth in the left ovary. This ovary might still be
dysfunctional as a result of the abnormal endocrine environment induced
by the right ovary. Removal of the right ovary might restore normal
function of the remaining left ovary and should be considered.
Similarly, differences were observed in distribution of mutated cells
throughout the endometrium. Consequently, the endometrium was out of
phase presumably due to elevated P levels throughout the cycle. This
implies that endometrial receptivity is disturbed and natural fertility
is compromised. This may represent an additional cause of infertility,
even in case normal function of at least one ovary could be restored.
This condition might also render future IVF procedures unsuccessful.
Some women with McCune-Albright syndrome achieve normal menses and
fertility, as well as pregnancies (9, 10, 11, 32, 33, 34). They
might constitute a subgroup of patients in which the extent of the
Gs-mutated cells is limited. On the contrary,
other patients have persistence of autonomous gonadal function
resulting in irregular cycles, metrorhagia, and other gynecological
problems (8, 11, 34). These abnormalities might be
underestimated because the vast majority of papers address clinical
findings in younger patients.
In conclusion, the present report provides evidence for persistent autonomous unilateral ovarian dysfunction during early adulthood in McCune-Albright syndrome not compatible with normal fertility. Increased FSH and LH signaling gives rise to development of multiple dominant follicles, premature luteinization, anovulation, and cyst formation. Single dominant follicle development and normal ovulation and subsequent corpus luteum function could be observed on the contralateral unaffected ovary. Endometrial morphology is abnormal. The gynecological implications of these findings may include cycle disturbances and untreatable infertility. Extended suppression of endogenous FSH or unilateral ovariectomy should be considered when pregnancy is desired.
Acknowledgments
We kindly acknowledge Dr. S. Chadha for morphology investigations, Prof. F. H. de Jong for hormone assays, and Dr. N. S. Macklon and Prof. P. Bouchard for critically reviewing the manuscript.
Footnotes
1 Presented in abstract form during the annual meeting of the Society
of Gynecological Investigation, Chicago, Illinois, 2000.
Received August 28, 2000.
Revised November 3, 2000.
Revised December 8, 2000.
Accepted December 13, 2000.
References