Long-term testosterone replacement therapy with slow-release silastic implants

E. M. Coutinho, C. Eiden Teixeira, A. Maltez and H. Maia, Jr


In healthy women the level of ovarian secretion of estradiol drops from 40-80 µg/day during reproductive years to less than 20 lig/day during  menopause.  Ovarian  secretion  of estrone drops from 20-50 pg/day to less than 10 pg/day¹. Unlike estrogen, androgen levels do not drop as markedly at the menopause. However, their decline with increasing age is greater than that of estrogen². Total circulatory levels in women in their forties are approximately 50% of those of women in their twenties. In addition, reduction of ovarian secretion rates of androgens also occurs and, although not as marked as that of the estrogens, it is also significant. Testosterone drops lioni 50-70 pg/day to 40-50 fig/day, whereas androstenedione declines from 1.0-1.5 mg/clay to 0.3-0.6 mg/day3. The physiologic decrease in the estrogen—androgen ratio is the cause of the increased facial hair growth that occurs in some women postmenopausally.  Replacement the rapy with estrogen corrects the estrogen deliciency but increases the estrogen—androgen ratio in favor of estrogen, creating a relative androgen deficiency in which loss of libido and sexual satisfaction becomes one of the most disturbing consequences.  In addition, lowered  testosterone  levels  contribute  to increased risk of invertebral crush fractures¹.


For the past twenty years we have been providing testosterone supplementation as replacement therapy to women with low testosterone blood levels. We have also offered hormone replacement therapy (HRT) with estrogen and testosterone using silastic implants to both oophorectomized and menopausal women who were unhappy with their estrogen—alone or estrogen—progestogen HRT. The silastic implants were made with segments of silastic tubing (Technical Products Inc., Decatur, Georgia), measuring 4 cm in length and filled with 50 mg of testosterone-17ß-ol or 50mg 17ß-estradiol. The implants were sterilized with either steam or ethylene oxide.  Insertion was made in the gluteal region following local anesthesia with procaine 2%. The standard dose consisted of 6-8 capsules of estradiol, totalling 300-400 mg, and 2-4  capsules of testosterone,  totalling 100-200 mg. The set of estradiol implants provided blood levels of estradiol of 35-50 pg/ml of estradiol whereas the two implants of testosterone provided 50-80ng/dl of testosterone. The release rate of estradiol was in the range of 0.8-1.09 mg/day, whereas that of testosterone was 0.2-0.8 mg/day over one year⁵. Implants containing levonorgestrel (Norplant®) or other progestins such as nomegestrol acetate (Uniplant®) were available for those women whose endometrial thickness surpassed 10 min for a period of two months or longer. The women included in these studies were: (1) women with intact ovaries but low testosterone blood levels; (2) young, oophorectomized wonien; and (3) menopausal women. The third group included both those with intact ovaries and oophorectomized women. Most women had been treated previously with estrogen, alone or in combination with progestin, and considered the previous treatment  unsatisfactory because they felt depressed,  lacked  energy  and  enthusiasm, and suffered loss of libido. Sonic of them complained of memory loss.


Testosterone treatment in intact, cycling women

Eighty-eight intact, cycling women complaining solely of loss of libido and presenting total testosterone levels below 80 ng/d1 (normal levels 50-180 ng/dI) were inserted with two capsules of testosterone. Blood levels of the hormone measured during implant use were higher than 80 ng/dl but never higher than the physiological limit of 180 ng/dl. Following insertion of the implants, these women were evaluated for effects on libido. In this group of 88 women the effect of the hormone on libido on a scale of 0-10 was rated 8 by 36 women (41%). Twenty-eight women (32%) rated the response between 5 and 7, and the remaining 24 (27%) rated it below 5. Half of these latter women (12) had no change in libido at all (Table 1).

No virilizing effects, such as hoarseness, hirsutisin and acne, were observed at the 100 mg (two implants) close level. In women who reported no effect or a response well below that which was desired, an additional capsule or set of two capsules was offered with the warning that  the higher blood levels resulting from this additional dosage could result in side-effects. All 12 women who had no response and ten of those who had rated the response below 5 opted for a supplement. Of the 22 women receiving a supplement, 20 received one more capsule and only two received two additional capsules. After three months these women were re-evaluated. Of the 20 women who were now bearing three capsules, eight women rated  the response superior to that attained with two capsules, i.e. above 5 but below 8. Three women rated the response above 8. One of these women considered  the effect exaggerated and requested removal of the extra capsule. This subject reported the development of an unwelcome desire for other women. The remaining nine women reported no improvement with the extra implant. One of the two women who received two additional capsules rated the response between 5 and 7 but the other remained unresponsive. Both developed hoarseness, seborrhea and acne. In over 50% of all women receiving these implants a line of very soft, downy hair grew over the skin covering the implant. In some women (15%), a slight depression at the site of the implant developed, indicating loss of fat.

Testosterone replacement therapy in young oophorectomized women

In women who had been oophorectomized for various  reasons, which  included endometriosis, abscesses, benign cysts and malignant disease, testosterone replacement therapy with implants was carried out in conjunction with estrogen replacement therapy. The combined set consisted of 6-8 implants containing 50 mg each (total of 300-400 mg) and 2-4 implants of testosterone containing 50 mg each (total of 100-200 mg). The sets provided enough hormones to maintain blood levels of estrogen in the range of 40-80 pg/ml and testosterone in the range of 50-80 ng/dl. Forty-eight oophorectomized women, aged 15-45, were inserted with either 6 or 8 implants of estradiol and 2-4 implants of testosterone, depending on body weight, and were followed up for 1-5 years. (Women with ovarian agenesis were not included in this group). Twenty-eight of these women who had an intact uterus (subgroup A) were offered two alternative regimens of progesterone therapy: (1) oral administration of medroxyprogesterone 5 mg for two weeks followed by  two weeks off; or (2)  an  implant of nomegestrol acetate containing 50 mg of the pmgestin inserted from the outset. The second alternative was the choice of 19 women (68%). Nine women (32%) opted for cyclic withdrawal bleeding. Testosterone replacement therapy provided the necessary support for normal sexual receptivity and libido according to the subject’s own evaluation at six months, one year and every successive year in over 50% of the women in both subgroups. Twenty-one out of 28 women who had intact uteri (subgroup A) reported a satisfactory response to masturbation. Over half of this group had no sexual partner, which made evaluation of sexual receptivity,  sex  drive and orgasm during coitus impossible. In the 20 women who had no uterus and who received no progestin supplementation (subgroup B), the response to sexual stimulation was reported as satisfactory in 17/20 subjects (85%). Only two subjects reported no orgasm when masturbating.

Testosterone replacement therapy in menopausal women

The use of subdermal implants with testosterone in menopausal women has the advantage of providing natural testosterone in physiological dosage over a long period of time. In clinics at both the Federal University of Bahia’s Climerio de Oliveira Maternity Hospital and at CEPARH (Centro de Pesquisa e Assistência em Reprodução Humana) in Salvador, Bahia, Brazil, we have been providing our menopausal patients with hormone replacement  therapy,  including testosterone implants, for over twenty years. Although several thousand women over the last ten years have received a small supplement of subdermal testosterone to maintain a physiological level of the hormone, we are reporting here on only 153 women who opted for that form of therapy and who have been followed up for 1-5 years (mean 3.5 years). The women were aged 42-72 years. The standard procedure was to offer all the women a set of implants containing 4-6 capsules of estradiol and 1-3 capsules  of testosterone.  In  intact women whose  blood levels of both estrogen and testosterone, albeit low, remained within the physiological range, the combination set contained four capsules of estradiol and one to two capsules of testosterone. Whenever the blood levels of these hormones were below the physiological range, the higher dose was prescribed. Of the 153 women, 108 were intact women who required monitoring of the uterus by ultrasound at 6-monthly intervals for evaluation of endometrial thickness. The other 45 women had been  hysterectomized prior to admission  to this study  for reasons which included  leiomyomata, endornetrial hyperplasia and/or excessive bleeding. In addition to sonographic monitoring, all the women were evaluated on a yearly basis with a mammography and breast ultrasound, as well as blood analysis which included cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL), triglycerides, testosterone, estradiol, thyroid-stimulating hormone (TSH), follicle-stimulating hormone (FSH), luteinizing hormone (LH) and prolactin. Bone densitometry was also carried out on a yearly basis. A total of 634 woman-years of use of the estradiol-testosterone implant combination was analyzed. Except in the case of smokers, the symptomatology associated with the menopause,  which  included  hot  flushes, sweating and tachycardia, was maintained under control during implant use with the standard close for a full year following implant insertion. The control of depression and headaches was likewise complete in over 72% of those who presented with these symptoms. A total of 426 woman-years of use were completed by women with an intact uterus and the remaining 208 woman-years of use were from women hysterectomized before initiation of therapy.

At  the end  of the  first week following  implant insertion, most women (84%) who had complained of hot flushes, tachycardia and sweating at admission reported complete relief from these symptoms. By the end of the second month of use, 72% of implant users reported relief from depression and a moderate (32%) to marked (56%) improvement in their sexuality. Smokers in both intact and oophorectotnized groups were those who felt that the standard dose was insufficient to control their symptoms. In these women, blood levels of both estrogen and testosterone were lower than in non-smokers. Moreover, the steroid supply from the implant needed replacement much earlier (3-6 months) in the smokers, never reaching the one year they lasted in non-smokers. Weight changes at one year were moderate (mean 1.5 kg). Thirty-eight per cent of women presented no change or a reduction in body weight. Thirty-three per cent presented an increase of 0.5-3.0 kg. The remaining women (29%) presented increases greater than 3 kg. Two women, both with hypothyroidism, had increases in body weight above 8 kg and one subject, who had type II diabetes, gained 12 kg during the first year of use but lost 8 kg during the second year with proper dieting and medication. There were no complaints of hoarseness or other voice changes. Thirty-two women (21.3%) complained of body and facial hair growth but none of them found the effect serious enough to discontinue use of the implant. Only one woman complained of clitoral hypertrophy. Discontinuation at one year was only 7.8% (12/153 women). The reasons for discontinuation were known in only three cases, where subjects felt the implantation procedure to be painful or unpleasant (2) or felt that the treatment was below expectations (1). The other nine women dropped out because they moved away from Salvador or because  they had been advised  to  discontinue HRT. Of the 153 women admitted to the study, 141 wanted to be reinserted for the second year (92.2%). At the end of the second year, 14 women dropped out (9.2%). One hundred and twenty-seven remained for a third year. Reasons for dropout. at the end of the second year were similar to those given at the end of the first year: moving away from the area or lack of motivation for continuation. At the end of the third year, 16 women dropped out (10.5%). Some gave as a reason being too old (over 60) to carry on, others the loss of their husband. In no case were side-effects presented as a reason for discontinuation. One hundred and eleven women opted for continuation for a fourth year. At the end of the fourth year, only nine women (8.1%) dropped out, leaving 102 from the original group of 153, who felt they should continue with the therapy. Continuation rate at five years is therefore 2/3 (6 (i%) of the original group (table 2). Twenty-six out of 108 women who had an intact uterus developed endometrial thickness above the 10 mm limit, which did not regress following shedding after a course of levonorgestrel (Nortrel®) or medroxyprogesterone acetate (Provera®) daily for ten clays.  These women were submitted to hysteroscopy and/or endometrial ablation. Five of these opted for hysterectomy. Endometrial changes detected by hysteroscopy were diagnosed as hyperplasia of the glandular epithelium or endometrial polyp. In none of these cases was malignant change detected. No adverse changes were detected which could be correlated with the therapy in blood levels of cholesterol, HDL, LDL, triglycerides or total lipids. Elevated blood pressure developing during treatment in seven women, one of them diabetic, did not seem to be associated with the therapy according to their cardiologists’ comments. No abnormal bone loss occurred during treatment as documented by bone densitometry carried out at yearly intervals in all patients.


Study conclusions

These observations of the use of suardermal silastic implants, which release physiological doses of testosterone, show that the hormone should be included in HRT whenever possible because of its remarkable benefits on mood, memory and sexuality of women with low blood levels of the hormone. That the users feel better when on testosterone than when without the hormone is indicated by the high continuation rates, which show two-thirds of the subjects remaining faithful to the therapy at 5 years. It is possible that other benefits of the implant method, which include liberating the woman from the need to apply or take medication for a full year, make this method more attractive. The relative lack of side-effects associated with the use of progestins, such as lowered libido and depression, is another attractive feature of this method. There is a strong association between the menopause and a marked decrease in coital frequency and sexual interest, which is independent of age and seems to be more associated with testosterone than estrogen decrease⁶. The response to testosterone included an increase in sexual desire, sex drive, receptivity, skin sensitivity, particularly at the nipples and clitoris, better lubrication and enhanced response at orgasm. In addition to the benefits  to psychosexual function, testosterone plays an important role in the development and maintenance of bone mass, reducing bone loss and the risk of fractures. This is supported in our studies by normal bone densitometry showing normal densities over the duration of the study. The studies also show that the association of testosterone and estrogen at physiological doses reduces the need to use progesterone in order to protect the  endometrium.  It has been shown recently that androgens, particularly androstenedione,  can inhibit human endometrial epithelial cell growth and secretory activity in vitro. No significant effects of the two most potent androgens, testosterone and dihydrotestosterone, were shown in these studies, probably because the binding of the two steroids to proteins of the calfserum in the in-vitro system used reduced bioavailability⁷. Studies in our own laboratory confirm the existence, previously reported, of androgen receptors in the endometrium. The effects of testosterone on the enclometrium are complex. The androgen receptor in the endometrium is located mainly in the stroma with very few detected in the glandular epithelium⁸. This predominantly stromal distribution of receptors is observed in various benign endometrial pathologies such as polyps and non-atypical hyperplasia. This indicates that most of the effects of testosterone on the normal endometrium are exerted through the stroma, where it stimulates, for instance, prolactin production⁹. This effect of testosterone on prolactin is similar to the one observed with progestins. In fact, several studies have shown that high and prolonged doses of androgens can induce endometrial atrophy. This has been observed to occur in women with gender dysphoria, who had used androgens for at least a year¹⁰. Androgens, when used to treat endometriosis, are also effective in causing endometrial atrophy despite the presence of normal circulating levels of estrogens¹¹. Estrogens can induce the synthesis of androgen receptors in the endometrium in the same way as they do for progesterone receptors¹². Testosterone, like progesterone, is also known to affect predominantly cellular differentiation and not proliferation in the endometrium¹³. These findings suggest that testosterone may exert an antiproliferative effect on the normal endometrium when a proper ratio with estrogen can be achieved. In fact, in postmenopausal patients using estradiol and testosterone implants, the endometrium remains atrophic throughout the treatment with very few signs of endometrial proliferation.

When these patients develop an abnormal thickening of the endometrium, this is due in most cases to the  growth of endometrial polyps¹⁴. The growth of polyps in the uterine cavity seems to he unabated by testosterone. Similarly, endometrial polyps also seem to be insensitive to the antiproliferative effects of progesterone, growing in response to estrogen and displaying various grades of hyperplasia in their glandular epithelium¹⁵. The reasons for the insensitivity of endometrial polyps to both testosterone and progestins are still far from  being understood. We have recently found that approximately 50% of endometrial  polyps show amplification for the c-erbB2 oncoprotein and this is related to higher proliferation rates¹⁶. However, the association between c-erbB2 amplification and unresponsiveness  to  testosterone  or progestins has not yet been established. Recent studies, on the other hand, have suggested that epidermal growth factor receptor could activate the estrogen receptor in the absence of estrogens by means of a chemical signal originating at the plasma membrane. This chemical signal would probably involve the phosphorylation of estrogen receptors and their subsequent   activation¹⁷. Because c-erbB2 protein is a membrane glycoprotein growth factor receptor, which shares molecular homology with the EGF receptor, its overexpression in endometrial polyps may therefore permanently activate the estrogen receptor, markedly increasing the proliferative esponse to estrogens and possibly blocking the  differentiating effect of testosterone. Although an association between c-erbB2 overexpression and testosterone resistance in endometrial polyps is purely speculative at present, such an association might explain the development of endometrial polyps in postmenopausal patients using implants of testosterone and estradiol and the low incidence of diffuse endometrial hyperplasia as the  cause of endometrial thickness and abnormal uterine bleeding in these patients.


Testosterone supplementation with slow-releasing subdermal silastic implants was provided to women with low testosterone blood levels. The women included in these studies were women with intact ovaries but testosterone blood levels below 80 ng/dl, young oophorectomized women and menopausal women. In all three groups of women, testosterone replacement therapy provided the necessary support for normal sexual receptivity and libido, according to the subjects’ own evaluation at 6 months, one year and every successive year in over 50% of the women. The use of subdermal implants with testosterone and estrogen in menopausal women has the advantage of providing natural hormones in physiological dosages over a long period of time.


1. Judd HL. Hormonal dynamics associated with the  menopause. Clin. Obstet Gynecol 1976;19: 775-88

2. Judd HL, Yen SSC. Serum androstenedione and testosterone levels during the menstrual cycle. J Clip Endocrinol Metab 1973;36:481

3. Muschayandebvu T, Castracane DV, Gimpel T, et al. Evidence for diminished midcycle ovarian androgen  production  in  older  reproductive aged women. Fertil Sind 1996;65:721-3

4. Longscope C, Baker RS, Hui SI„ et al. Androgen and estrogen dynamics in women with vertebral  crush  fractures.  Maturitas 1984;6: 309-18

5. Coutinho EM. Clinical experience with implant contraception. Contraception 1978;18:411-27

6. Montgomery  M,  Coutinho  EM.  Androgens, sexuality and menopause.  In Coutinho EM, Spinola P, eds. Current Knowledge in Reproductive Medicine 2000. Excerpta Medica International Congress  Series 1206.  Amsterdam. Elsevier Science BV, 2000;409-13

7. Tuckerman EM, Okon MA, Tin-Chiu Li Ana, Laird SM. Do androgens have a direct effect on endometrial fUnction? An in vitro study. Fertil 13. Steril 2000;74:771-9

8. Maia H Jr, Maltez A, Fahel P, et al. Detection of testosterone and estrogen receptors in post-menopausal  endometrium.  Maturities 2000; in press 

9. Narukawas S, Kanzaki H, Inoue T Androgens induce  prolactin production by human endometrial stromal cells in vitro. J Glin Endocrinol Metab 1994;78:165-8

10. Miller N, Bedard VC, Cooter NB, et al. Histo-logical changes in the genital tract in transsexual women following androgen therapy. Histopathology 1986;10:661-9

11. Rose GL, Dowsett. M, Mudge JE, et al. The inhibitory effects of danazol, danazol metaboites,  gestrinone and testosterone on  the growth of human endometrial cells in vitro. Fertil Steril 1998;49:224-8

12. Adesanya-Famuyiwa 00,  Zhou  J,  Wu Gu,. et al. Localization and sex steroid regulation

of androgen receptor gene expression in rhesus monkey uterus. Obstet Gynecol 1999;93;


13. Neulen J, Wagner B, Runge M, et al. Effect of progestins,  androgens,  estrogens  and  anti-estrogens on 3H-thymidine uptake by human endometrial and endosalpinx cells in vitro. Arch Gynecol Obstet 1987;240:225-32

14. Maia H Jr; Maltez A, Fahel Y, et al. Hysteroscopic findings in postmenopausal patients with a thick endometrium after using implants of estradiol and testosterone. Gynaecol Endoscopy 2000;9:259-65

15.Maia  H Jr,  Maltez A,  Calmon  LC, et al. Histopathology  and  steroid  receptors  in endometrial polyps of postmenopausal patients under hormone replacement therapy. Gynaecol Endoscopy 1998;7:267-72

16.Maia H Jr, Maltez A, Fahel P, et al. Histochemical detection of c-erbB2 over-expression in

endometrial polyps removed by hysteroscopy.  Gynaecol Endoscopy 2000;9:253-8

17.Sperolt L. The estrogen receptor: changing concepts. Clinical lessons from molecular biology. In Coutinho EM Spinola P, eds. Reproductive Medicine: A Millennium Review. Carnforth, UK: Parthenon Publishing, 1099;155-61


Cadastre-se e receba os informes ELMECO