LETTERS TO THE EDITOR
Are elevated androgens a cause of discordant pubertal development? Evidence from a case of adrenocortical carcinoma
We present a case of an adolescent with an adrenocortical carcinoma that demonstrates the potential inhibitory effects of androgens on breast and endometrial development.
A previously well 141/2-year-old girl presented with a 2-year history of worsening hirsutism and abdominal pain. She had onset of pubarche at 11 years, but had not yet had breast devel- opment or menarche. In contrast, her mother and sister had menarche at 12 and 11 years of age, respectively. On examina- tion, her height was 150-4 cm (3rd percentile) and weight 57.3 kg (50th percentile). She had a deep voice, muscular build, and hirsutism involving her upper lip, chin, chest and lower back. There were no features suggestive of glucocorticoid excess. She was Tanner stage 1 for breast and 5 for pubic hair. The remainder of the examination was unremarkable.
Investigations revealed significantly elevated androgens, with no evidence of glucocorticoid or mineralocorticoid excess (Table 1). Despite the lack of breast development, the gonadotro- pins and estradiol were in the pubertal range, and bone age was significantly advanced at 18 years. Imaging demonstrated a large left-sided retroperitoneal mass arising from the adrenal gland, consistent with an adrenocortical carcinoma, with no evidence of metastatic spread. On ultrasound, the uterus was peri-pubertal in size, but without a defined endometrial stripe. The patient pro- ceeded to have an open trans-abdominal left adrenalectomy, with pathological confirmation of adrenocortical carcinoma.
Postoperatively she had biochemical resolution of the hyper- androgenism (Table 1) and was monitored for pubertal
progression. She had a menstrual bleed 2 months postopera- tively and subsequently had regular monthly menses. Breast development also occurred postoperatively with progression from Tanner stage 1-5 within 15 months.
Adrenocortical carcinomas are rare tumours, presenting in children most commonly with virilization. Our patient with an adrenocortical carcinoma presented with an unusual pubertal picture: complete lack of breast development and no increase in the endometrium, despite evidence of pubertal activation of the hypothalamic-pituitary-gonadal axis. We hypothesize that excessively high androgen levels interfered with breast and endo- metrial development.
Pubertal mammary gland development is primarily initiated and driven by the presence of increasing oestrogen levels. The effect of androgens on breast tissue has been more controver- sial.1,2 By providing increased precursor supplies of oestrogen through aromatization, elevated androgen levels have been theo- rized to lead to stimulation of breast tissue.2 However, increas- ing evidence from in vitro and animal studies has shown that androgens, by down-regulating oestrogen receptor gene expres- sion,1 can instead reduce proliferation and increase apoptosis in cultured human breast tissue.2
This has led to interest regarding the potential protective role of androgens in the development of breast cancer.1,3 Epidemio- logical studies have shown a reduced risk of breast cancer in women with polycystic ovarian syndrome.3 Postmenopausal women treated with trans-dermal testosterone and hormone replacement therapy demonstrate less breast cell proliferation on needle aspiration biopsy than those on hormone therapy alone.3 In addition, remission rates of breast cancer were doubled in a
| Parameter and (reference range) | Pre-operatively | Day 1 post-op | 3 months post-op | 7 months post-op | 15 months post-op |
|---|---|---|---|---|---|
| Breast Tanner stage | 1 | 1 | 2 | 3 | 5 |
| Menses | Absent | Absent | Regular | Regular | Regular |
| Uterine appearance on ultrasound | |||||
| Length (cm) | 4.4 | 5-8 | |||
| Volume (cm3) | 6.9 | 26-7 | |||
| Endometrial stripe | Absent | Present | |||
| LH (IU/1) (pre-pubertal: ≤ 0-2, pubertal: 0-3-30) | 5.7 | 2.0 | 0-6 | 1-4 | 2.8 |
| FSH (IU/l) | 4.2 | 2.0 | 0-6 | 1-4 | 2-8 |
| Estradiol (pmol/l) | 290 | 31 | 221 | 110 | 100 |
| Testosterone (nmol/l) (females: < 2-1) | 20-2 | <0.7 | <0.7 | 0-8 | <0.7 |
| Androstenedione (nmol/1) (0-5-6-1) | >41-4 | 2.1 | 4.8 | 1.7 | |
| Dehydroepiandrosterone Sulfate (umol/1) (1-12-0) | 58-6 | 8.1 | 3.6 | 5.0 | 5.6 |
| Urinary free cortisol (nmol/day) (<151) | 97 | ||||
| Renin (ng/l/s) (0-1-7) | 0-36 | 0-63 | |||
| Aldosterone (pmol/l) (upright 111-860) | 440 |
phase 3 trial, comparing the use of tamoxifen plus an androgen to tamoxifen alone.3 The inhibitory effect of androgens on breast cells appears to occur in a dose-dependant manner.1 Although confounded by the use of gonadotropin-releasing hormone ana- logue therapy, histological data derived from mastectomies from female-to-male transsexuals exposed to high-dose androgens have demonstrated reduction in glandular tissue in up to 93% of cases.4
Our case provides clinical evidence that pathologically high levels of androgens cannot only reduce proliferation, but may inhibit normal breast development, despite appropriate pubertal levels of oestrogen and an apparently mature hypothalamic- pituitary-gonadal axis. Thus, it appears that the elevated andro- gens were having a direct inhibitory effect on the breast tissue. Moreover, the rapid progression of breast development (15 months to attain Tanner 5 breasts, compared with average progression from Tanner 2-4 over 36 months) provides further evidence that removal of the source of the elevated androgens allowed breast development to proceed.
Although androgens have been associated with both endome- trial atrophy and proliferation,4 our case provides evidence that high-dose androgens can inhibit endometrial maturation. While at initial presentation, there was no endometrial stripe on ultra- sound, postoperatively, with resolution of the hyperandroge- nism, regular menses commenced within 2 months. High androstenedione levels have been shown to be a risk factor for primary amenorrhoea and thinner endometrial lining in girls with polycystic ovarian syndrome,5 supporting the hypothesis that significantly elevated androgens can inhibit the endome- trium.
In conclusion, our patient, presenting with discordant biochemical and physical signs of puberty, provides unique insight into the potential inhibitory effects of significantly ele- vated androgens on breast and endometrial maturation.
Disclosure Statement
The authors have nothing to disclose.
Jennifer Harrington and Mark R. Palmert Division of Endocrinology, Departments of Pediatrics and Physiology, The Hospital for Sick Children, The University of Toronto, Toronto, ON, Canada E-mail:jennifer.harrington@sickkids.ca
doi:10.1111/cen.12021
References
1 Labrie, F. (2006) Dehydroepiandrosterone, androgens and the mammary gland. [Review] [139 refs]. Gynecological Endocrinology 22, 118-130.
2 Eigeliene, N., Elo, T., Linhala, M. et al. (2012) Androgens inhibit the stimulatory action of 17beta-estradiol on normal human breast tissue in explant cultures. Journal of Clinical Endocrinology and Metabolism 24, 24.
3 Shufelt, C.L. & Braunstein, G.D. (2008) Testosterone and the breast. Menopause International 14, 117-122.
4 Grynberg, M., Fanchin, R., Dubost, G. et al. (2010) Histology of genital tract and breast tissue after long-term testosterone administration in a female-to-male transsexual population. Reproductive Biomedicine Online 20, 553-558.
5 Rachmiel, M., Kives, S., Atenafu, E. et al. (2008) Primary amen- orrhea as a manifestation of polycystic ovarian syndrome in ado- lescents: a unique subgroup? Archives of Pediatrics and Adolescent Medicine 162, 521-525.
Is bariatric surgery really inefficient in hypothalamic obesity?
Hypothalamic obesity is a frequent and serious complication following treatment of hypothalamic tumours, especially cranio- pharyngiomas. Hypothalamic obesity is associated with poor quality of life and frequent adverse metabolic complications. Moreover, its pathophysiology is complex and that may explain in part the inefficacy of behavioural or medical treatments.1 Weismann and collaborators recently published in this journal a retrospective analysis of the effectiveness of bariatric surgery in hypothalamic morbid obesity and concluded that no significant loss of body weight was achieved in nine adolescent or young adult patients with craniopharyngioma associated morbid obesity.2
We report here the body weight decrease observed in three of four adults suffering from hypothalamic obesity after they went through bariatric surgery (2 sleeve gastrectomies and 1 gastric bypass).
Four patients are reported here (Table 1). Each of them had been through hypothalamic surgery (in addition to radiotherapy in one case) for a craniopharyngioma (n = 3) or a choristoma (n = 1). At the time of hypothalamic surgery, their respective BMI were 27.0, 31.5, 24.0 and 21 kg/m2. The first two patients started to gain weight during the months before the diagnosis of the hypothalamic tumour as frequently reported.1 Following neurosurgery, all patients developed pan hypopituitarism and obesity, concomitant with increased hunger and a decrease in satiation. Patients 2, 3 and 4 also developed type 2 diabetes with chronic hyperglycemia (mean HbA1C 9-6 ± 1.5%) despite the intensive therapy including insulin for the first female patient. As patients continued to gain body weight despite the multidis- ciplinary lifestyle intervention based on exercise, behaviour and nutrition therapy performed in a university hospital, bariatric surgery was proposed and performed by the same surgeon. For the younger man, sleeve gastrostomy allowed BMI to decrease from 51 to 39.5 kg/m2 at the end of the first year. It then slightly increased to 41 kg/m2 18 months after surgery and remained stable during follow-up. For the second 39-year-old man, BMI decreased from 37.6 to 31.7 kg/m2 in the first year after surgery, eventually reached 34 kg/m2 and then stayed at this level during the 30 months of follow-up. However, HbA1 remained high (7-8%) despite an increase in metformin doses and additional therapy with sitagliptin. Thirty-six months after the bariatric surgery, BMI of the third patient had decreased