0021-972X/97/$03.00/0
10
Journal of Clinical Endocrinology and Metabolism
Copyright © 1997 by The Endocrine Society
Menstrual Bleeding in a Female Infant with Congenital
Adrenal Hyperplasia: Altered Maturation of the
Hypothalamic-Pituitary-Ovarian Axis
NAVEEN ULI, DAISY CHIN, RAPHAEL DAVID, NANCY GENEISER,
KEVIN ROCHE, FLAVIA MARINO, ELLEN SHAPIRO, KRIS PRASAD,
AND SHARON OBERFIELD
Departments of Pediatrics and Pediatric Endocrinology (N.U., D.C., R.D., F.M., K.P., S.O.), Pediatric
Radiology (N.G., K.R.), and Urology (E.S.), New York University Medical Center, New York, New York
10016
ABSTRACT
Vaginal bleeding during the neonatal period is commonly related
to the withdrawal of maternal estrogens. Vaginal bleeding has also
been reported in female infants with congenital adrenal hyperplasia
and has been proposed to be due to a treatment-induced activation of
the hypothalamic-pituitary-ovarian axis.
We report a female infant with the salt-losing form of congenital
adrenal hyperplasia due to 21-hydroxylase deficiency, who had the
onset of vaginal bleeding at 3 months of life. Adrenal steroid sup-
pression had been achieved by 2.5 weeks of age. At the time of bleed-
ing, imaging studies revealed an enlarged right ovary with a domi-
nant 3-cm cyst and additional small cysts that had not been seen on
the newborn sonogram. The uterus was enlarged and stimulated.
Three weeks later (1 week after the cessation of bleeding), repeat
G ONADOTROPIN secretion patterns differ in males and
females during early infancy. In females, FSH rises
earlier and LH rises later than in males. The rise in FSH is
considerably greater in females than in males. Normally
thereafter, as the hypothalamic-pituitary unit matures, in-
creasing sensitivity to the negative feedback of gonadal ste-
roids is associated with a decline in gonadotropin levels. This
state is usually maintained until the onset of puberty.
In the first 2 months of life, elevated estradiol values are
also seen in female infants (1). An increase in ovarian fol-
licular maturation, breast tissue enlargement, and even uter-
ine bleeding have been observed during the early neonatal
period, a time when the highest serum FSH and estradiol
levels are seen (2). These findings probably represent the
ovarian response to pituitary gonadotropins (3). It has been
suggested that the more prolonged FSH elevation in females
during infancy represents either a relative ovarian resistance
to gonadotropin stimulation or a delay in the maturation of
the feedback mechanism controlling gonadotropin secretion
(3).
A delayed onset of genital bleeding has previously been
reported in female infants with congenital adrenal hyper-
Received September 20, 1996. Revision received February 24, 1997.
Rerevision received June 6, 1997. Accepted June 11, 1997.
Address all correspondence and requests for reprints to: Sharon E.
Oberfield, M.D., Department of Pediatric Endocrinology, New York
University Medical Center, 550 First Avenue, New York, New York
10016.
3298
Vol. 82, No.
Printed in U.S.A.
Downloaded from https://academic.oup.com/jcem/article-abstract/82/10/3298/2823264 by guest on 23 July 2019
ultrasound demonstrated a marked decrease in the size of the right
ovary, and the dominant cyst was no longer seen. The patient had a
heightened FSH response to GnRH and elevated levels of estradiol for
age. At 5 months of age, no further episodes of sustained vaginal
bleeding were observed. Repeat hormonal levels were prepubertal,
and pelvic sonogram demonstrated no evidence of stimulation.
The findings in our patient suggest that a decline in adrenal an-
drogens after glucocorticoid treatment resulted in an increase in go-
nadotropin levels, which then triggered a transient and augmented
end-organ response (menses). Further, we suggest that our infant’s
hormonal findings may reflect a delay in the timely development of
the negative restraint by sex steroids on gonadotropins that is nor-
mally observed in infancy. (J Clin Endocrinol Metab 82: 3298 –3302,
1997)
plasia (CAH) due to 21-hydroxylase deficiency during the
neonatal period (4). It has been postulated that prolonged
exposure to excessive adrenal androgens during fetal life
followed by a rapid decline of these hormones postnatally
with glucocorticoid treatment results in a more prolonged
activation of the hypothalamic-pituitary-ovarian axis. Addi-
tionally, a greater increase in the responsiveness of internal
genitalia to gonadotropins and sex hormones has been sug-
gested to occur (4).
In this report we present a 3-month-old female infant with
salt-wasting CAH due to 21-hydroxylase deficiency who had
the onset of delayed menstrual bleeding with a large ovarian
cyst. The possible mechanisms of this occurrence are
discussed.
Materials and Methods
Case report (Table 1 and Figs. 1–3)
The infant was a full-term 2860-g 46,XX product of a normal vaginal
delivery. At birth, she was noted to have clitoromegaly, a single perineal
orifice, and hyperpigmentation of the labia majora. A pelvic sonogram
during the newborn period revealed the presence of a uterus without
visualization of adnexal structures. The diagnosis of congenital adrenal
hyperplasia due to 21-hydroxylase deficiency was made based on the
17-hydroxyprogesterone level on day 1 of life (.14,000 ng/dL), with a
serum testosterone level of 530 ng/dL and a dehydroepiandrosterone
sulfate level of 390 mg/dL. Treatment with hydrocortisone was initiated
on day 1, and due to evidence of salt wasting (continued weight loss,
serum Na of 133 mEq/L, serum K of 7.0 mEq/L, and PRA of 195
ng/mL/h), treatment with 9a-fluorohydrocortisone and sodium chlo-
ride was initiated on day 11 of life. Suppression of adrenal androgen
 MENSTRUAL BLEEDING IN AN INFANT WITH CAH 3299

TABLE 1. Sequential serum estradiol, FSH/LH, and FSH/LH response to GnRH stimulation

Day of life: 70 90 95 102 109 151 528

Postbleed 2 Postbleed, Postbleed, Postbleed
Clinical findings: Prebleed Day 3 of bleed Day 8 of bleed
days 1 week 2 months 14 months
Time (min): Random 0/30/60/90 Random Random 0/30/60 Random Random
Estradiol (ng/dL) 1.0 5.6 6.5 8.2 4.9 1.7 ,0.5
FSH (mIU/mL) 6.8/20/26/23 10 10/62/69 17
LH (mIU/mL) 0.68/6.7/5.3/4.4 0.64 0.66/21/14 0.5
DFSH 19.2 59
DLH 6.02 20.3
DFSH/DLH 3.19 2.9

Downloaded from https://academic.oup.com/jcem/article-abstract/82/10/3298/2823264 by guest on 23 July 2019

FIG. 1. Midline longitudinal (A) and
transverse (B) ultrasound images show
prominence of the lower uterine seg-
ment/cervix, with fluid in the vagina.
An endometrial echo is identified on the
longitudinal image (arrow). A trans-
verse image through the vagina dem-
onstrates a large amount of debris and
fluid in the vaginal canal.
3300 ULI ET AL. JCE & M • 1997
Vol 82 • No 10

precursors was achieved on day 17 (17-hydroxyprogesterone, 36 ng/dL;
testosterone, 5 ng/dL; androstenedione, 21 ng/dL; PRA, 1.45 ng/mLzh;
Na, 140 mEq/L; K, 5.5 mEq/L; levels obtained about 6 h after the
morning dose of hydrocortisone) The infant was maintained on hydro-
cortisone (2.5 mg every 8 h), 9a-fluorohydrocortisone (0.15 mg once
daily), and sodium chloride (500 mg every 6 h).
On day 88 of life, the infant had the onset of vaginal bleeding ac-
companied by transient fever. Physical examination revealed ambigu-
ous genitalia as before, with no palpable breast buds and absent pubic
hair. Catheterization of the vagina yielded a minimal amount of blood.
Urine, blood, and cerebrospinal fluid cultures were negative for bacteria.
A pelvic ultrasound revealed the presence of a right ovary with a
dominant 3-cm cyst in the right lower quadrant, a stimulated uterus, and
a fluid/debris-filled vaginal cavity (Fig. 1). A computed tomography
an enhancing rim of tissue, suggesting that it most likely represented an
ovary (Fig. 2). A magnetic resonance imaging (MR) examination per-
formed 3 days later showed a stimulated uterus measuring 4 cm in
length (Fig. 3). In the right lower quadrant, small cysts adjacent to the
dominant cyst confirmed the presence of a stimulated right ovary. To
assess the activity of the hypothalamic-pituitary-ovarian axis, a GnRH
stimulation test was performed. The results are presented in Table 1. At
this time, the 17-hydroxyprogesterone level was 359 ng/dL, and the
androstenedione level was 26 ng/dL.
On the fifth hospitalization day, the infant was noted to have bilateral
breast buds, with glandular tissue measuring 0.75 cm on the right side
and 0.5 cm on the left side. She continued to have scant vaginal bleeding
for a total of 12 days. During a follow-up visit on day 102 of life, she had
smaller breast buds, measuring less than 0.5 cm bilaterally, and the

Downloaded from https://academic.oup.com/jcem/article-abstract/82/10/3298/2823264 by guest on 23 July 2019

(CT) scan performed the following day demonstrated that the cyst had vaginal bleeding had resolved. Serum hormone measurements at that

FIG. 2. CT with oral and iv contrast

demonstrates a 3-cm cystic structure in
the right lower quadrant with an en-
hancing rim of tissue (arrow).

FIG. 3. Sagittal (A) and coronal (B) T2-weighted MR images (TR 6500 ms, TE 130 ms). The sagittal image demonstrates a debris/fluid-filled
vagina (straight arrow) and uterus with a prominent endometrium (curved arrow). On the coronal image, multiple small cysts (straight arrow)
and a large dominant cyst are indicative of an ovary in the right lower quadrant. A prominent lower uterine segment is also noted (curved arrow).
 MENSTRUAL BLEEDING IN AN INFANT WITH CAH
time revealed a persistently elevated estradiol level of 8.2 ng/dL, with
a FSH level of 10.0 mIU/mL and a LH level of 0.64 mIU/mL.
On day 109 of life, the infant still had palpable breast buds, measuring
less than 0.5 cm bilaterally, and was thriving. The GnRH stimulation test
was repeated (see Table 1). Serum 17-hydroxyprogesterone was 62 ng/
dL, and testosterone was less than 3 ng/dL. Pelvic sonogram performed
on the same day revealed the uterus to measure 3.8 cm in length, with
an identifiable endometrial echo. The right ovary was smaller and now
measured 2.2 3 1.4 3 1.5 cm and contained multiple cysts. However, the
dominant cyst was no longer seen. The left ovary was not definitively
identified. Apart from minimal spotting on one diaper on day 114 of life,
she had no more episodes of vaginal bleeding.
Three weeks after cessation of the vaginal bleeding, on day 127 of life,
a follow-up pelvic sonogram was performed, and the uterus was noted
to have decreased in size and now measured 2.4 cm in length, with the
fundus being smaller than the cervix. An endometrial echo was again
visible, but the ovaries were not definitively visualized, consistent with
nonstimulated pelvic organs.
Two months after the initial bleeding, the infant had barely palpable
breast buds. Examination was otherwise unremarkable. Endocrine eval-
uation revealed estradiol of 1.7 ng/dL, FSH of 17 mIU/mL, and LH of
0.5 mIU/mL. At 7 months of age, or 4 months after her bleeding, breast
tissue was no longer present, and her examination was completely
prepubertal. Fourteen months after her initial bleeding episode, she
remains prepubertal on physical examination. A random serum estra-
diol measurement was less than 0.5 ng/dL. She continues to receive
treatment with hydrocortisone (2.5 mg every 8 h), 9a-fluorohydrocor-
tisone (0.15 mg daily), and sodium chloride (2 g daily).
17-Hydroxyprogesterone and testosterone were measured by minor
modification of previously described standard RIA methods after col-
umn extraction with Celite by the Pediatric Endocrine Laboratory of
New York University-Bellevue Hospital Center or Endocrine Sciences
(Calabasas Hills, CA).
Androstenedione, estradiol, and PRA were measured by RIA, and
FSH and LH were determined by immunochemiluminometric assay
(ICMA) at Endocrine Sciences Laboratories.
Sequential pelvic ultrasonography was performed and interpreted by
the same pediatric radiologists (N.G. and K.R.).
The GnRH stimulation test was performed using Factrel (gonadorelin
hydrochloride, Ayerst Laboratories, Philadelphia, PA; 100 mg/m2 given
as an iv bolus over 2 min). FSH and LH levels were determined before
and 30, 60, and on one occasion 90 min after the administration of Factrel.
Results
Clinical and hormonal findings
As stated, at the time of the initial bleeding (day 88 of life),
our patient demonstrated hormone levels consistent with
reasonable control of her 21-hydroxylase deficiency. She did
not have a significant maternal estrogen effect at birth, as
evidenced by the lack of breast buds during the neonatal
period. On day 70 a random estradiol measurement was 1.0
ng/dL. However, on day 90, 2 days after the onset of vaginal
bleeding, her estradiol had risen to 5.6 ng/dL, with further
increases to 6.5 and 8.2 ng/dL on days 95 and 102, respec-
tively, when breast buds were conspicuous. On day 109, 1
week after vaginal bleeding had resolved, estradiol was de-
clining, and by day 151 it reached a level of 1.7 ng/dL, or only
slightly higher than expected for age.
A GnRH stimulation test on day 90 at the onset of bleeding
demonstrated a maximal FSH response to 26 mIU/mL (D
FSH, 19.2) and of LH to 6.7 mIU/mL (D LH, 6.02), with a ratio
of D FSH to D LH of 3.19. A second GnRH stimulation test
on day 109, 1 week after bleeding had stopped, demonstrated
more pronounced FSH and LH responses. The maximal FSH
response was 69 mIU/mL (D FSH, 59), and the maximal LH
response was 21 mIU/mL (D, LH 20.3), but the D FSH to D
LH ratio remained essentially the same at 2.9.
3301
Pelvic imaging studies
Pelvic ultrasound was performed on day 88 (Fig. 1), CT on
day 89 (Fig. 2), and MR on day 92 (Fig. 3). They demonstrated
a stimulated uterus with a prominent endometrial echo, mea-
suring approximately 2.5 mm in the antero-posterior direc-
tion, a fluid- and debris-filled vagina, and a stimulated right
ovary with a dominant 3-cm cyst. (MR was helpful, in that
the small cysts confirming the presence of ovarian tissue
were not appreciated on either the ultrasound or the CT.) A
follow-up ultrasound on day 109 (not shown) demonstrated
the uterus and ovary to be less prominent, and the dominant
Downloaded from https://academic.oup.com/jcem/article-abstract/82/10/3298/2823264 by guest on 23 July 2019
cyst was no longer identified. A follow-up study on day 127
(not shown) failed to visualize either ovary, consistent with
pelvic organs in a nonstimulated state.
Discussion
Technical advances in ultrasonography, including meth-
ods of enhanced anatomical resolution, have revealed that
microcysts of the ovary can be a normal finding in females
during the later part of infancy and early childhood (5, 6).
Maturation of the ovaries begins in early childhood, with the
ovarian volumes correlated with age. The uterus, however,
does not begin to mature until about 7 yr of age, after which
increases in uterine volume and corpus to cervix ratio be-
come apparent (6). Females with premature thelarche, a de-
velopmental variant of normal puberty, have been shown to
have an increased frequency of follicular cysts of the ovary,
measuring up to 10 mm in diameter (5, 7), with ovarian and
uterine volumes being maintained in the normal range (7).
Girls with idiopathic precocious puberty, on the other hand,
have increases in both uterine and ovarian volumes, with the
additional finding of larger and more numerous ovarian
cysts (8). Females with congenital adrenal hyperplasia due to
21-hydroxylase deficiency have been reported to have a
greater frequency of nonhomogeneous ovarian structure.
During early childhood, although the presence of a few small
cysts has been noted, no increases in total ovarian volume or
uterine size have been reported (9). In the only other study
of infants with CAH in whom vaginal bleeding was reported,
sonographic evaluation was not performed (4). The presence
of a dominant ovarian cyst measuring more than 10 mm, as
noted in our patient, has not been reported to date in an
infant with CAH.
As at the time of the initial vaginal bleeding on day 88 of
life, our patient demonstrated a hormonal profile consistent
with good metabolic control of her 21-hydroxylase defi-
ciency, the source of her estrogens was probably not due to
peripheral conversion of adrenal androgens. On the third
day of the bleeding, serum estradiol was elevated, and the
GnRH stimulation test demonstrated a brisk response to both
FSH and LH, while maintaining the early infantile, albeit
exaggerated, pattern of FSH predominance. We acknowl-
edge that no data are currently available for GnRH-induced
FSH and LH levels by ICMA in this age group. However, we
suggest that the response indeed mimicked the pattern
observed in prepubertal infants (Reiter, E., personal
communication).
When related temporally to the initial sonographic find-
ings of multiple small cysts and a single large dominant cyst
3302 ULI ET AL.
in the right ovary, the clinical course indicates an ovarian
response to exaggerated pituitary gonadotropin activity. At
the time that breast development was noted, serum estradiol
was continuing to rise; it reached a maximum level 2 weeks
after the onset of uterine bleeding and declined thereafter. A
reduction in the size of the ovary with nonvisualization of the
dominant cyst and a regression of breast size followed the
decline in estradiol levels. Indeed, 14 months after the bleed-
ing episode, she remains prepubertal with estradiol levels
below 0.5 ng/dL. A repeat GnRH test demonstrated the
preservation of FSH predominance while maintaining brisk
responses to both FSH and LH. This is in keeping with the
previous observation of the maintenance of the sex dichot-
omy of FSH secretion in response to GnRH in patients with
CAH despite exposure of the females to excessive androgens
during the intrauterine and neonatal periods (10).
In contrast to the findings of our report, elevated basal
serum LH levels were recently reported in female infants
with 21-hydroxylase deficiency during the first 3 months of
life after treatment with hydrocortisone; the LH levels were
similar to those observed in control infant males. The re-
searchers suggested that the elevated basal LH levels in their
treated female infants represented a reversal of the normal
sex dichotomy. We are unable to adequately compare these
reported findings (11) to those in our infant inasmuch as the
assays used were different (immunoenzymoassay vs. ICMA),
and the FSH and LH responses to GnRH were not assessed.
Additionally, the roles of FSH and LH in fetal ovarian
maturation are currently under investigation (12, 13). In fetal
pigs, lower levels of FSH and LH receptor concentrations
have been demonstrated in the ovaries compared to those in
the fetal testes (13). It has been suggested that this dichoto-
mous finding is responsible for the relative lack of fetal
ovarian growth compared to fetal testicular growth. Further,
for both sexes an inverse correlation has been observed be-
tween fetal gonadotropin levels and fetal gonadal receptors,
i.e. elevated gonadotropin levels appear to down-regulate
gonadal gonadotropin receptors. We suggest that these go-
nadotropin receptor-mediated events may account for the
previously reported phenomenon of the relative resistance of
infant ovaries to gonadotropin stimulation.
We extrapolate from the animal model that in our patient,
the suppression of pituitary gonadotropin by elevated an-
drogens during fetal life prevented the occurrence of down-
regulation of the ovarian gonadotropin receptors. We sug-
gest further that the glucocorticoid treatment-mediated
JCE & M • 1997
Vol 82 • No 10
androgen suppression led to a marked rise in gonadotropin
levels, resulting in significant stimulation of the existing
ovarian gonadotropin receptors. Growth of the ovarian fol-
licles and estrogen secretion subsequently occurred and
manifested clinically as thelarche and endometrial bleeding.
Further studies of the ontogeny of fetal gonadal development
in humans are needed to allow better understanding of our
patient’s findings.
Acknowledgments
We thank Dr. Natalie Geary for her help with the evaluation of this
Downloaded from https://academic.oup.com/jcem/article-abstract/82/10/3298/2823264 by guest on 23 July 2019
patient. We also thank Ms. Vi Catena for her secretarial help with the
preparation of this manuscript.
References
1. Grumbach MM, Kaplan SL. 1990 The neuroendocrinology of human puberty.
An ontogenetic perspective. In: Grumbach MM, Sizonenko PC, Aubert ML,
eds. Control of the onset of puberty II. Baltimore: Williams and Wilkins; 1– 62.
2. Root AW, Shulman DI. 1986 Isosexual precocity: current concepts and recent
advances. Fertil Steril. 45:749 –766.
3. Winter JD, Hughes IA, Reyes FI, Faiman C. 1976 Pituitary-gonadal relations
in infancy. II. Patterns of serum gonadal steroid concentrations in man from
birth to two years of age. J Clin Endocrinol Metab. 42:679 – 686.
4. Maesaka H, Suwa S, Tachibana K, Katsumata N. 1985 Prolonged activation
of hypothalamo-pituitary-ovarian axis during early infancy in female patients
with salt-losing 21-hydroxylase deficiency. Pediatr Res. 19:1258 –1262.
5. Freedman SM, Kreitzer PM, Elkowitz SS, Soberman N, Leonidas JC. 1993
Ovarian microcysts in girls with isolated premature thelarche. J Pediatr.
122:246 –249.
6. Salardi S, Orsini LF, Cacciari E, Bovicelli L, Tassoni P, Reggiani A. 1985
Pelvic ultrasonography in premenarcheal girls: relation to puberty and sex
hormone concentrations. Arch Dis Child. 60:120 –125.
7. Nakamura M, Okabe I, Shimoizumi M, Yanagisawa M, Taniguchi N, Itoh
K. 1991 Ultrasonography of ovary, uterus and breast in premature thelarche.
Acta Paediatr Jpn. 33:645– 648.
8. Stanhope R, Adams J, Jacobs HS, Brook CGD. 1985 Ovarian ultrasound
assessment in normal children, idiopathic precocious puberty, and during low
dose pulsatile gonadotropin releasing hormone treatment of hypogonadotro-
phic hypogonadism. Arch Dis Child. 60:116 –119.
9. Salardi S, Orsini LF, Cacciari E, et al. 1988 Pelvic ultrasonography in girls with
precocious puberty, congenital adrenal hyperplasia, obesity or hirsutism. J Pe-
diatr. 112:880 – 887.
10. Reiter EO, Grumbach MM, Kaplan SL, Conte FA. 1975 The response of
pituitary gonadotropes to synthetic LRF in children with glucocorticoid-
treated congenital adrenal hyperplasia: lack of effect of intrauterine and neo-
natal androgen excess. J Clin Endocrinol Metab. 40:318 –325.
11. Belgorosky A, Chahin S, Rivarola MA. 1996 Elevation of serum luteinizing
hormone levels during hydrocortisone treatment in infant girls with 21-
hydroxylase deficiency. Acta Paediatr. 85:1172–1175.
12. Brooks AN, McNeilly AS, Thomas GB. 1995 Role of GnRH in the ontogeny
and regulation of the fetal hypothalamo-pituitary-gonadal axis in sheep. J
Reprod Fertil. 49(Suppl):163–175.
13. Goxe B, Prunier A, Remy J, Salesse R. 1993 Ontogeny of gonadal lutein-
izing hormone and follicle stimulation hormone receptors in the fetal pig
and related changes in gonadotropin and testosterone secretion. Biol Re-
prod. 49:609 – 616.