0013-7227/90/1266-3263$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society
Vol. 126, No. 6 Printed in U.S.A.
Disparate Effect of Tamoxifen in Rats with Experimentally Induced Endometriosis* R. KADABAf AND C. W. SIMPSON Reproductive Biology Research Unit, Department of Obstetrics and Gynecology, University of Saskatchewan, Saskatoon, Saskatchewan, S7N OWO Canada
ABSTRACT. The effect of tamoxifen on the growth of endometrial implants was examined in rats with experimentally induced endometriosis. Administration of tamoxifen (1 mg/kgday) for 60 days completely regressed the endometrial implants. When tamoxifen treatment was withdrawn, recurrence of endometrial implants was observed. Ectopic endometrial implants regressed completely 2 months after ovariectomy in rats with experimentally induced endometriosis. Tamoxifen administration (1 mg/kg day) for 60 days to ovariectomized rats with regressed implants led to complete recurrence of ectopic endometrial implants. In ovariectomized rats administration of tamoxifen antagonized estradiol-enhanced thymidine incorporation into uterine DNA, but tamoxifen per se stimulated thymi-
dine incorporation. This indicates the agonist/antagonist nature of tamoxifen. Treatment of rats with tamoxifen blocked the normal estrous cycle. However, serum progesterone levels were higher during tamoxifen treatment in rats with intact ovaries than in ovariectomized rats. When ovariectomized rats with regressed implants were administered tamoxifen (1 mg/kg-day) and progesterone (10 mg/kg day) for 60 days, recurrence of ectopic endometrial implants was not observed. These findings suggest that tamoxifen is effective in regressing endometrial implants in rats with experimentally induced endometriosis, and ovarian progesterone may have a facilatory effect. (Endocrinology 126: 3263-3267, 1990)
E
opausal patient treated for benign breast disease (6) clouds the potential use of this drug in endometriosis. In the present study rats with experimentally induced endometriosis were used as an animal model to examine the hypothesis that tomoxifen treatment is effective in inducing regression of ectopic endometrial implant. Our results suggest that tamoxifen is effective in regressing ectopic endometrial implants, and ovarian progesterone participates in the regressive effect.
NDOMETRIOSIS is a disease characterized by the presence and proliferation of endometrial tissue outside the uterus. Approximately one third of women who complain of pelvic pain and/or infertility have endometriosis (1). Radical surgery may be recommended in extreme cases. A more conservative approach to the treatment of endometriosis is hormone therapy. Growth of endometriosis is stimulated by cyclic ovarian secretion of estrogen and progesterone (2, 3). In the absence of these hormones, i.e. after oophorectomy, endometriosis typically undergoes regression (3). Medical treatment of endometriosis depends on suppression of ovarian function and induction of endometrial atrophy. Several drug regimens have been used for this purpose. Danazol is currently the most widely used drug for the medical treatment of endometriosis (4). Tamoxifen, an antiestrogen, has been recently demonstrated to completely regress endometriotic implants in two patients (5). Tamoxifen has been suggested as an alternative modality in the treatment of endometriosis, especially for women desiring to conceive, as it does not suppress ovarian function. However, a report indicating development of endometriosis during tamoxifen treatment in a 54-yr-old premen-
Materials and Methods Experimental induction of endometriosis in rats
Received January 8, 1990. * This work was supported by the Saskatchewan Health Research Board. t To whom requests for reprints should be addressed.
Mature cycling female rats (Sprague-Dawley), 150-200 g, were used. All surgical procedures were performed under aseptic conditions in a sterile flow hood. Animals were administered 0.05-0.1 mg ketamine hydrochloride solution (200 mg/ml), followed by anesthesia using halothane. During laparotomy, the blood vessels supplying the right uterine horn were ligated, and approximately 2 cm of uterus was resected. End to end anastomosis of the uterus after resection was performed with 6-0 nylon sutures (Ethicon, Peterborough, Ontario, Canada). The resected horn was opened in isotonic sterile saline, and the myometrium was peeled from the endometrium under a dissecting microscope. Two pieces of endometrium measuring 4 X 4 mm (approximate) were cut and then sutured ip to the inside of the peritoneal membrane, with the sutures (6-0 nylon, Ethicon) passing through the membrane and anchoring into the abdominal musculature on all four sides of the implant. The
3263
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 April 2016. at 15:17 For personal use only. No other uses without permission. . All rights reserved.
3264
TAMOXIFEN AND ENDOMETRIOSIS
criterion used to assess the viability of the graft was the formation of a cyst within 1 month after transplantation. Effect of tamoxifen on experimentally induced endometriosis in rats One month after induction of endometriosis, the endometrial implant size was measured during laparotomy. The animals were then subdivided into two subgroups. One subgroup of animals was treated with tamoxifen (1 mg/kg-day) sc every day for 2 months. The other subgroup served as a control and was injected ever day with vehicle. One and 2 months after treatment with tamoxifen, the animals from both control and treated groups underwent laparotomy, and endometrial implant surface area (length x width) was measured using Vernier callipers (Manostat, Switzerland). Serum samples were collected at 2 months for progesterone estimation. Tamoxifen treatment was withdrawn at the end of 2 months. The animals underwent laparotomy 2 months after withdrawal of tamoxifen treatment, and implant size was measured. Effect of tamoxifen on experimentally induced endometriosis in ovariectomized rats One month after induction of endometriosis, the rats underwent laparotomy. During laparotomy, the endometrial implant size was measured, and the ovaries were removed. Two months after ovariectomy, the implants were examined under laparotomy. The animals were then divided into three subgroups. One group was treated with tamoxifen (1 mg/kg-day) sc every day for 2 months. The second group was treated with the same doses of tamoxifen and 10 mg/kg progesterone every day for 2 months. The third group remained untreated. Two months later endometrial implants were examined at laparotomy. Serum samples were collected at the end of 2 months for progesterone estimation. To measure the progesterone levels in serum samples, serum was extracted twice with ethyl acetate, and progesterone levels were determined by a RIA technique (7). The recovery of progesterone with ethyl acetate ranged between 85-90%. Effect of tamoxifen on I3 H] thymidine incorporation into DNA in rat uterus Mature female rats (Sprague-Dawley) were ovariectomized. Four weeks after ovariectomy, the animals were divided into four subgroups. One group of animals received a sc injection of 10 ng 17/3-estradiol. Tamoxifen (1 mg/kg) was administered to the second group. The third group of animals was given a combination of estradiol and tamoxifen. The fourth group remained as a control. Rats were killed 24 h later by cervical dislocation. The uterus was rapidly removed, placed on ice, and trimmed free of fat. The uterine horns were then slit lengthwise and incubated at 37 C in Minimum Eagle's Medium containing 20 juCi [3H]thymidine for 2 h with continuous shaking. At the end of incubation thymidine incorporation was stopped by removing the horns and placing them in 5 ml ice-cold saline. The uterine horns were then washed twice, blotted dry, and frozen on dry ice. Thymidine incorporation into DNA was determined as described by Stewart and Webster (8).
Endo • 1990 Vol 126 • No 6
Statistical analysis The results of this study were subjected to nested analysis of variance (9). When a significant F value was present, Fisher's least significant ditference test was used for individual comparison of means.
Results Effect of tamoxifen on experimentally induced endometriosis in rats The ectopic endometrial implants took well to the abdominal environment and produced histologically identifiable glands and stroma. All of the transplanted endometrial sections exhibited accumulation of fluid at laparotomy, conducted 4 weeks after operation (Fig. 1). The implants were well vascularized and covered by a sheath of connective tissue. When animals were treated with tamoxifen (1 mg/kg-day, sc), four of seven implants completely regressed within 1 month of treatment. The remaining three implants were partially regressed. Oneway analysis of variance indicated a significant change in the implant size during tamoxifen treatment. The mean decrease in the implant size after 1 month of tamoxifen treatment was statistically significant (Table 1; P < 0.05). When treatment with tamoxifen continued for 2 months, all implants were completely regressed (P < 0.01; Table 2). In the control group implant size did not significantly change during the 2-month period (data not shown). In the tamoxifen-treated group, when the treatment was withdrawn three of seven implants grew within the following 2 months. However, the remaining four implants remained regressed. The mean implant size 2 months after withdrawal was significantly larger (P < 0.05; Table 1) than those at 1 and 2 months of tamoxifen treatment. Rats with experimentally induced endometriosis from the control group exhibited a normal estrous cycle. Cornified epithelial cells were observed in vaginal smears
f
FIG. 1. A healthy endometrial implant on the peritoneal surface of the abdominal wall of a rat 1 month after surgical implantation.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 April 2016. at 15:17 For personal use only. No other uses without permission. . All rights reserved.
TAMOXIFEN AND ENDOMETRIOSIS TABLE 1. Effect of tamoxifen on regression of endometrial implants in rats with experimentally induced endometriosis Parameter Before treatment 1 month after tamoxifen treatment (1 mg/kg-day, sc) 2 months after tamoxifen treatment (1 mg/kg-day, sc) 2 months after withdrawal of tamoxifen treatment
4.07 0.71°
7 7
0.00 + 0.00*
7
± 6.58C
7
20.59 1.63
14.54
± ±
All values are Mean ± SEM. n, Number of animals. " P < 0.05 compared with implant size before treatment. b P < 0.01 compared with implant size before treatment. c P < 0.05 compared with implant size 1 and 2 months after tamoxifen treatment. TABLE 2. Effect of ovariectomy on regression of endometrial implants and recurrence with tamoxifen administration Parameter 1 month after induction 2 months after ovariectomy Tamoxifen (1 mg/kg-day) for 60 days 2 months after ovariectomy Tamoxifen (1 mg/kg-day) and progesterone (10 mg/kg-day) for 60 days 2 months after ovariectomy
TABLE 3. Serum progesterone levels in intact and ovariectomized rats Parameter
Implant size (mm2)
3265
Intact rats (estrous) Intact rats (diestrous) Intact rats + tamoxifen (1 mg/kg-day, sc) for 2 months 2 months after ovariectomy Ovariectomy + tamoxifen (1 mg/kg-day, sc) for 2 months Ovariectomy + tamoxifen (1 mg/kg-day, sc) for 2 months + progesterone (10 mg/ kg-day, sc) for 2 months
Progesterone (ng/ml) 20.50 66.25 30.42
± 2.27 ± 9.07° ± 3.17
8 5 6
12.61 15.80
± 0.62 ± 0.79*
6 6
52.15
±
6
1.25C
All values are the mean ± SEM. n, Number of animals. "P < 0.01 compared with progesterone levels in rats in estrus and intact rats treated with tamoxifen. b P < 0.01 compared with progesterone levels in intact rats treated with tamoxifen. c P > 0.01 compared with progesterone levels in ovariectomized rats treated with tamoxifen.
Implant size (mm2) 19.90 ± 2.52 0.00 ± 0.00° 16.89 ± 4.34"
9 9 9
0.00 ± 0.00
8
All values are the mean ± SEM. n, Number of animals. " P < 0.01 compared with implant size before ovariectomy. 6 P < 0.01 compared with implant size 2 months after ovariectomy.
during estrous. Progesterone levels in serum samples collected during estrus and diestrus were 20.50 ± 2.27 and 66.25 ± 9.07 ng/ml, respectively. Animals treated with tamoxifen did not show the characteristic changes in vaginal smears associated with the estrous cycle. The vaginal smears contained a preponderance of leucocytes, with a variable admixture of nucleated and cornified cells. Serum progesterone levels in random blood samples obtained at the end of tamoxifen treatment were significantly lower than progesterone levels in control rats during diestrus (P < 0.01; Table 3). Effect of tamoxifen on experimentally endometriosis in ovariectomized rats
induced
When rats underwent ovariectomy, after induction of endometriosis, the implants were completely regressed within 2 months of ovariectomy in all animals (Fig. 2). When these rats were then treated with tamoxifen (1 mg/kg-day) for 2 months, eight of nine implants grew back. One implant remained completely regressed at the end of treatment. The mean implant size was significantly larger than that of implants before treatment (Table 2; P < 0.01). In the control group implants remained completely regressed compared to those in the
FIG. 2. The total morphological regression of an endometrial implant after ovariectomy.
tamoxifen-treated group. In a different group of ovariectomized rats with induced endometriosis, none of the implants grew after treatment with tamoxifen (1 mg/kgday) and progesterone (10 mg/kg-day) every day for 2 months. Ovariectomy significantly reduced progesterone levels compared to those in intact rats during estrus (P < 0.05) and diestrus (P < 0.01). In ovariectomized rats receiving tamoxifen, progesterone levels were significantly lower than those in intact rats during tamoxifen treatment (P < 0.01). When ovariectomized rats were treated with tamoxifen and progesterone, serum progesterone levels were higher than those in rats treated with tamoxifen (P < 0.01). Effect of tamoxifen on pHJthymidine incorporation into DNA in rat uterus The effect of tamoxifen alone or in combination with estrogen on thymidine incorporation into DNA by rat
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 April 2016. at 15:17 For personal use only. No other uses without permission. . All rights reserved.
TAMOXIFEN AND ENDOMETRIOSIS
3266
uterus is depicted in Fig. 3. Four weeks after ovariectomy, rat uteri were capable of incorporating thymidine into DNA. Administration of estrogen increased thymidine incorporation into DNA by 5-fold (P < 0.01; Fig. 3). When tamoxifen was administered with estrogen, thymidine incorporation into DNA was significantly reduced compared to that after estradiol treatment alone (P < 0.01; Fig. 3). Thymidine incorporation into DNA was significantly enhanced with tamoxifen treatment alone (P < 0.05) compared to that in the untreated controls. Discussion The present study demonstrates that induction of experimental endometriosis in the female rat is feasible. We have earlier reported the histological characteristics of the endometrial implant identifying the presence of stroma and gland (10). Endometrial implants in rats are ovarian steroid dependent (10). In the present study administration of tamoxifen to rats with experimentally induced endometriosis led to complete regression of ectopic endometrial implants. This corroborates the recent preliminary clinical report of Haber and Behelak (5), who demonstrated regression of ectopic endometrial implants in two patients with endometriosis after tamoxifen therapy. Our study further showed that recurrence of endometriosis is possible after withdrawal of tamoxifen therapy in the rat model. Further studies are required to determine whether longer periods of tamoxifen treatment reduce the chances of recurrence after withdrawal. The nonsteroidal antiestrogen tamoxifen exhibits a paradoxical species-specific pharmacology. The drug is a full estrogen in the mouse (11), a partial estrogen/antiestrogen in humans (12), and an antiestrogen in the chick oviduct (13). In the present study when tamoxifen was administered to ovariectomized rats at the dose level that induced regression of endometrial implants, it enhanced thymidine incorporation into DNA of rat uterus. 60r 50 40 30 20 10
CONTROL
TAM
Endo • 1990 Voll26-No6
However, when administered with estradiol, tamoxifen partially antagonized estradiol-stimulated thymidine incorporation into DNA of rat uterus. This indicates the estrogen/antiestrogen activity of tamoxifen in rats, as in humans. In rats with experimentally induced endometriosis, ectopic endometrial implants completely regressed 2 months after ovariectomy, which confirms our previous finding (10). In our earlier study although we were unable to detect viable endometrial tissue by histology after complete morphological regression, recurrence of endometriosis was observed with estrogen administration. This suggests the presence of estrogen-responsive cells at the implant site after complete morphological regression (10). In the present study administration of tamoxifen to ovariectomized rats with completely regressed implants led to recurrence of endometriosis. We suggest that recurrence of endometriosis during administration of tamoxifen may be due to the estrogen-like activity of the drug. Recently, Ford et al. (6) reported the development of endometriosis in a premenopausal woman treated with tamoxifen for recurrent benign breast disease. The results of this study suggest that development of endometriosis reported in a clinical study during tamoxifen therapy may be related to its partial estrogenic activity. The disparate effects of tamoxifen, that is regression of endometrial implants in rats with intact ovaries but recurrence of regressed endometrial implants in ovariectomized rats, suggest a possible ovarian modulatory effect on the tamoxifen treatment of endometriosis. In rats 2 months after ovariectomy serum progesterone levels were decreased compared to those in the control group. However, significant serum progesterone levels persisted in spite of complete ovariectomy. These may be of adrenal origin, as it has been demonstrated that in rats the adrenal is capable of secreting large amounts of progesterone (14). In rats with induced endometriosis treated with tamoxifen, a normal estrous cycle pattern was not present, based on vaginal cytology. It has been reported that tamoxifen administration blocks ovulation in rats by preventing the ovulatory surge of LH from the pituitary (15). However, when serum progesterone levels were measured during tamoxifen therapy in ovariectomized rats and rats with intact ovaries, the rats with intact ovaries had higher levels of progesterone. When ovariectomized rats with regressed implants were treated with tamoxifen and progesterone, the recurrence of endometrial implants was not observed. This supports the facilitatory role of progesterone in tamoxifen treatment of endometriosis in rats.
TAM4€
FlG. 3. Effect of estrogen (E) and tamoxifen (TAM) treatment on thymidine incorporation into uterine DNA in ovariectomized rats. All values are the mean ± SEM (n = 5).
In clinical studies, the reason for the various effects of tamoxifen in treatment of endometriosis is not clear. Development of endometriosis was reported in a 54-yr-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 April 2016. at 15:17 For personal use only. No other uses without permission. . All rights reserved.
TAMOXIFEN AND ENDOMETRIOSIS
old premenopausal woman (6), while complete regression of endometrial implants was reported during tamoxifen therapy (10 mg twice a day) in patients 24 and 29 yr of age (5). As the present study in rats indicates a modulatory effect of ovarian progesterone on tamoxifen treatment of endometriosis, clinically reported disparate effect of tamoxifen in endometriosis may be related to serum progesterone levels. The rat endometrium, unlike the human endometrium, does not slough off during the estrous cycle. However, there are similarities between the human and rat endometrium in regulation of endometrial growth by ovarian steroids. Recent studies demonstrate that estrogen enhances DNA polymerase activity and DNA content, which are antagonized by progesterone in both the rat and human endometrium (16, 17). This further validates the use of rats with experimentally induced endometriosis as an animal model to investigate the clinically reported disparate effect of tamoxifen. Successful regression of endometriotic implants in rats receiving tamoxifen treatment suggests the potential value of this drug therapy in endometriosis. Tamoxifen offers some important advantages over GnRH agonists (GnRHa), which have been introduced into the management of endometriosis. The loss of trabecular lumbar bone during GnRHa-induced ovarian suppression has been a recent concern (18-20). Tamoxifen, on the other hand, has been demonstrated to inhibit bone resorption in ovarian hormone-deficient rats (21, 22), which has been implicated in the estrogen-like effect of tamoxifen. However, the teratogenic effect of tamoxifen (23, 24) may limit its advantages over GnRHa. Recently, one of the 7-alkyl amide analogs of estradiol (ICI 164,384) has been demonstrated to be a potent antiestrogen devoid of estrogenic activity (25). A comparative study of tamoxifen and ICI 164,384 indicates that pure antiestrogen (ICI 164,384) is free of some of the toxicological effects of the partial agonist (25). The successful regression of endometrial implants with tamoxifen treatment observed in this study clearly indicates that partial selective suppression of estrogen activity can regress ectopic endometrial implants. Further studies with pure antagonist to determine whether partial suppression of endogenous estrogen activity will resolve ectopic endometrial tissue without inducing bone resorption are important. References 1. Kistner RW 1979 Endometriosis and infertility. Clin Obstet Gynecol 22:101
3267
2. Gerbie AB, Merril JA 1988 Pathology of endometriosis. Clin Obstet Gynecol 31:779 3. diZerega GS, Barber DL, Hodgen GD 1980 Endometriosis; role of ovarian steroids in initiation, maintenance, and suppression. Fertil Steril 33:649 4. Dmowski WP 1988 Danazol induced pseudomenopause in the management of endometriosis. Clin Obstet Gynecol 31:829 5. Harber GM, Behelak YF 1987 Preliminary report on the use of tamoxifen in treatment of endometriosis. Am J Obstet Gynecol 156:582 6. Ford MRW, Turner MJ, Wood C, Soutter WP 1988 Endometriosis developing during tamoxifen therapy. Am J Obstet Gynecol 158:1119 7. Rajkumar K, Malinek J, Murphy BD 1985 Effect of lipoproteins and luteotrophins on progesterone accumulation by luteal cells from the pregnant pig. Steroids 45:119 8. Stewart PJ, Webster RA 1983 Intrauterine injection of cholera toxin induces estrogen-like uterine growth. Biol Reprod 29:671 9. Sokal RR, Rohlf FJ 1981 Nested analysis of variance. In: Sokal RR, Rohlf FJ (eds) Biometry. Freeman, New York, p 271 10. Rajkumar K, Schott PW, Simpson CW, The rat as an animal model for endometriosis to examine recurrence of ectopic endometrial tissue following regression. Fertil Steril, in press 11. Harper MJK, Walpole AL 1966 Contrasting endocrine activation of cis and trans isomers in a series of substituted triphenylethylenes. Nature (Lond) 212:87 12. Furr BJA, Jordan VC 1984 Pharmacology and clinical use of tamoxifen. Pharmacol Ther 25:127 13. Sutherland RL, Mester J, Baulieu EE 1977 Tamoxifen is a potent "pure" antiestrogen in the chick oviduct. Nature (Lond) 267:434 14. Ogle TF, Kitay JI 1977 Ovarian and adrenal steroids during pregnancy and the oestrous cycle in the rat. J Endocrinol 74:89 15. Labsetwar AP 1970 Role of estrogens in ovulation: a study using the estrogen antagonist ICI 46,474. Endocrinology 87:542 16. Cummings AM 1989 The role of steroid hormones and decidual induction in the regulation of adenosine diphosphoribosyl transferase activity in rat endometrium. Endocrinology 124:1408 17. Usuki S, Kubota S, Shioda M 1988 Activity of deoxyribonucleic acid polymerase alpha stimulated by estrogen in the endometrium of the human uterus during the menstrual cycle. Gynecol Endocrinol 2:283 18. Lewis V, Ramos J, Dawood MY, Changes in bone mineral content in endometriosis patients treated with GnRH agonist. 34th Annual Meeting of the Society for Gynecologic Investgation, Atlanta GA, 1987, p 181 (Abstract) 19. Cann CE, Henzl M, Burry K, Anreko J, Hanson F, Adamson D, Trobough G, Reversible bone loss is induced by GnRH agonists. 68th Annual Meeting of The Endocrine Society, Anaheim CA, 1986, p 37 (Abstract) 20. Dawood YM, Lewis V, Ramos J 1989 Cortical and trabecular bone mineral content in women with endometriosis: effect of gonadotropin-releasing hormone agonist and danazol. Fertil Steril 52:21 21. Turner RT, Wakley GK, Hannon KS, Bell NH 1988 Tamoxifen inhibits osteoclast mediated resorption of trabecular bone in ovarian hormone deficient rats. Endocrinology 122:1146 22. Jordon VC, Phelps E, Lindgren JU 1987 Effects of antiestrogen on bone in castrated and intact female rats. Breast Cancer Res Treat 10:31 23. Chamness G, Bannayan G, Landry L, Sheridan P, McGuire W 1979 Abnormal reproductive development in rats after neonatally administered antiestrogen/tamoxifen. Biol Reprod 21:1087 24. Clark J, Guthrie S, McCormack S 1981 Neonatal stimulation of the uterus by clomiphene, tamoxifen and nafoxidine: relationship to the development of reproductive tract abnormalities. Adv Exp Med Biol 138:87 25. Wakeling AE, Bowler J 1988 Novel antiestrogens without partial agonist activity. J Steroid Biochem 31:645
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 April 2016. at 15:17 For personal use only. No other uses without permission. . All rights reserved.
Vol. 126, No. 6 Printed in U.S.A.
Disparate Effect of Tamoxifen in Rats with Experimentally Induced Endometriosis* R. KADABAf AND C. W. SIMPSON Reproductive Biology Research Unit, Department of Obstetrics and Gynecology, University of Saskatchewan, Saskatoon, Saskatchewan, S7N OWO Canada
ABSTRACT. The effect of tamoxifen on the growth of endometrial implants was examined in rats with experimentally induced endometriosis. Administration of tamoxifen (1 mg/kgday) for 60 days completely regressed the endometrial implants. When tamoxifen treatment was withdrawn, recurrence of endometrial implants was observed. Ectopic endometrial implants regressed completely 2 months after ovariectomy in rats with experimentally induced endometriosis. Tamoxifen administration (1 mg/kg day) for 60 days to ovariectomized rats with regressed implants led to complete recurrence of ectopic endometrial implants. In ovariectomized rats administration of tamoxifen antagonized estradiol-enhanced thymidine incorporation into uterine DNA, but tamoxifen per se stimulated thymi-
dine incorporation. This indicates the agonist/antagonist nature of tamoxifen. Treatment of rats with tamoxifen blocked the normal estrous cycle. However, serum progesterone levels were higher during tamoxifen treatment in rats with intact ovaries than in ovariectomized rats. When ovariectomized rats with regressed implants were administered tamoxifen (1 mg/kg-day) and progesterone (10 mg/kg day) for 60 days, recurrence of ectopic endometrial implants was not observed. These findings suggest that tamoxifen is effective in regressing endometrial implants in rats with experimentally induced endometriosis, and ovarian progesterone may have a facilatory effect. (Endocrinology 126: 3263-3267, 1990)
E
opausal patient treated for benign breast disease (6) clouds the potential use of this drug in endometriosis. In the present study rats with experimentally induced endometriosis were used as an animal model to examine the hypothesis that tomoxifen treatment is effective in inducing regression of ectopic endometrial implant. Our results suggest that tamoxifen is effective in regressing ectopic endometrial implants, and ovarian progesterone participates in the regressive effect.
NDOMETRIOSIS is a disease characterized by the presence and proliferation of endometrial tissue outside the uterus. Approximately one third of women who complain of pelvic pain and/or infertility have endometriosis (1). Radical surgery may be recommended in extreme cases. A more conservative approach to the treatment of endometriosis is hormone therapy. Growth of endometriosis is stimulated by cyclic ovarian secretion of estrogen and progesterone (2, 3). In the absence of these hormones, i.e. after oophorectomy, endometriosis typically undergoes regression (3). Medical treatment of endometriosis depends on suppression of ovarian function and induction of endometrial atrophy. Several drug regimens have been used for this purpose. Danazol is currently the most widely used drug for the medical treatment of endometriosis (4). Tamoxifen, an antiestrogen, has been recently demonstrated to completely regress endometriotic implants in two patients (5). Tamoxifen has been suggested as an alternative modality in the treatment of endometriosis, especially for women desiring to conceive, as it does not suppress ovarian function. However, a report indicating development of endometriosis during tamoxifen treatment in a 54-yr-old premen-
Materials and Methods Experimental induction of endometriosis in rats
Received January 8, 1990. * This work was supported by the Saskatchewan Health Research Board. t To whom requests for reprints should be addressed.
Mature cycling female rats (Sprague-Dawley), 150-200 g, were used. All surgical procedures were performed under aseptic conditions in a sterile flow hood. Animals were administered 0.05-0.1 mg ketamine hydrochloride solution (200 mg/ml), followed by anesthesia using halothane. During laparotomy, the blood vessels supplying the right uterine horn were ligated, and approximately 2 cm of uterus was resected. End to end anastomosis of the uterus after resection was performed with 6-0 nylon sutures (Ethicon, Peterborough, Ontario, Canada). The resected horn was opened in isotonic sterile saline, and the myometrium was peeled from the endometrium under a dissecting microscope. Two pieces of endometrium measuring 4 X 4 mm (approximate) were cut and then sutured ip to the inside of the peritoneal membrane, with the sutures (6-0 nylon, Ethicon) passing through the membrane and anchoring into the abdominal musculature on all four sides of the implant. The
3263
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 April 2016. at 15:17 For personal use only. No other uses without permission. . All rights reserved.
3264
TAMOXIFEN AND ENDOMETRIOSIS
criterion used to assess the viability of the graft was the formation of a cyst within 1 month after transplantation. Effect of tamoxifen on experimentally induced endometriosis in rats One month after induction of endometriosis, the endometrial implant size was measured during laparotomy. The animals were then subdivided into two subgroups. One subgroup of animals was treated with tamoxifen (1 mg/kg-day) sc every day for 2 months. The other subgroup served as a control and was injected ever day with vehicle. One and 2 months after treatment with tamoxifen, the animals from both control and treated groups underwent laparotomy, and endometrial implant surface area (length x width) was measured using Vernier callipers (Manostat, Switzerland). Serum samples were collected at 2 months for progesterone estimation. Tamoxifen treatment was withdrawn at the end of 2 months. The animals underwent laparotomy 2 months after withdrawal of tamoxifen treatment, and implant size was measured. Effect of tamoxifen on experimentally induced endometriosis in ovariectomized rats One month after induction of endometriosis, the rats underwent laparotomy. During laparotomy, the endometrial implant size was measured, and the ovaries were removed. Two months after ovariectomy, the implants were examined under laparotomy. The animals were then divided into three subgroups. One group was treated with tamoxifen (1 mg/kg-day) sc every day for 2 months. The second group was treated with the same doses of tamoxifen and 10 mg/kg progesterone every day for 2 months. The third group remained untreated. Two months later endometrial implants were examined at laparotomy. Serum samples were collected at the end of 2 months for progesterone estimation. To measure the progesterone levels in serum samples, serum was extracted twice with ethyl acetate, and progesterone levels were determined by a RIA technique (7). The recovery of progesterone with ethyl acetate ranged between 85-90%. Effect of tamoxifen on I3 H] thymidine incorporation into DNA in rat uterus Mature female rats (Sprague-Dawley) were ovariectomized. Four weeks after ovariectomy, the animals were divided into four subgroups. One group of animals received a sc injection of 10 ng 17/3-estradiol. Tamoxifen (1 mg/kg) was administered to the second group. The third group of animals was given a combination of estradiol and tamoxifen. The fourth group remained as a control. Rats were killed 24 h later by cervical dislocation. The uterus was rapidly removed, placed on ice, and trimmed free of fat. The uterine horns were then slit lengthwise and incubated at 37 C in Minimum Eagle's Medium containing 20 juCi [3H]thymidine for 2 h with continuous shaking. At the end of incubation thymidine incorporation was stopped by removing the horns and placing them in 5 ml ice-cold saline. The uterine horns were then washed twice, blotted dry, and frozen on dry ice. Thymidine incorporation into DNA was determined as described by Stewart and Webster (8).
Endo • 1990 Vol 126 • No 6
Statistical analysis The results of this study were subjected to nested analysis of variance (9). When a significant F value was present, Fisher's least significant ditference test was used for individual comparison of means.
Results Effect of tamoxifen on experimentally induced endometriosis in rats The ectopic endometrial implants took well to the abdominal environment and produced histologically identifiable glands and stroma. All of the transplanted endometrial sections exhibited accumulation of fluid at laparotomy, conducted 4 weeks after operation (Fig. 1). The implants were well vascularized and covered by a sheath of connective tissue. When animals were treated with tamoxifen (1 mg/kg-day, sc), four of seven implants completely regressed within 1 month of treatment. The remaining three implants were partially regressed. Oneway analysis of variance indicated a significant change in the implant size during tamoxifen treatment. The mean decrease in the implant size after 1 month of tamoxifen treatment was statistically significant (Table 1; P < 0.05). When treatment with tamoxifen continued for 2 months, all implants were completely regressed (P < 0.01; Table 2). In the control group implant size did not significantly change during the 2-month period (data not shown). In the tamoxifen-treated group, when the treatment was withdrawn three of seven implants grew within the following 2 months. However, the remaining four implants remained regressed. The mean implant size 2 months after withdrawal was significantly larger (P < 0.05; Table 1) than those at 1 and 2 months of tamoxifen treatment. Rats with experimentally induced endometriosis from the control group exhibited a normal estrous cycle. Cornified epithelial cells were observed in vaginal smears
f
FIG. 1. A healthy endometrial implant on the peritoneal surface of the abdominal wall of a rat 1 month after surgical implantation.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 April 2016. at 15:17 For personal use only. No other uses without permission. . All rights reserved.
TAMOXIFEN AND ENDOMETRIOSIS TABLE 1. Effect of tamoxifen on regression of endometrial implants in rats with experimentally induced endometriosis Parameter Before treatment 1 month after tamoxifen treatment (1 mg/kg-day, sc) 2 months after tamoxifen treatment (1 mg/kg-day, sc) 2 months after withdrawal of tamoxifen treatment
4.07 0.71°
7 7
0.00 + 0.00*
7
± 6.58C
7
20.59 1.63
14.54
± ±
All values are Mean ± SEM. n, Number of animals. " P < 0.05 compared with implant size before treatment. b P < 0.01 compared with implant size before treatment. c P < 0.05 compared with implant size 1 and 2 months after tamoxifen treatment. TABLE 2. Effect of ovariectomy on regression of endometrial implants and recurrence with tamoxifen administration Parameter 1 month after induction 2 months after ovariectomy Tamoxifen (1 mg/kg-day) for 60 days 2 months after ovariectomy Tamoxifen (1 mg/kg-day) and progesterone (10 mg/kg-day) for 60 days 2 months after ovariectomy
TABLE 3. Serum progesterone levels in intact and ovariectomized rats Parameter
Implant size (mm2)
3265
Intact rats (estrous) Intact rats (diestrous) Intact rats + tamoxifen (1 mg/kg-day, sc) for 2 months 2 months after ovariectomy Ovariectomy + tamoxifen (1 mg/kg-day, sc) for 2 months Ovariectomy + tamoxifen (1 mg/kg-day, sc) for 2 months + progesterone (10 mg/ kg-day, sc) for 2 months
Progesterone (ng/ml) 20.50 66.25 30.42
± 2.27 ± 9.07° ± 3.17
8 5 6
12.61 15.80
± 0.62 ± 0.79*
6 6
52.15
±
6
1.25C
All values are the mean ± SEM. n, Number of animals. "P < 0.01 compared with progesterone levels in rats in estrus and intact rats treated with tamoxifen. b P < 0.01 compared with progesterone levels in intact rats treated with tamoxifen. c P > 0.01 compared with progesterone levels in ovariectomized rats treated with tamoxifen.
Implant size (mm2) 19.90 ± 2.52 0.00 ± 0.00° 16.89 ± 4.34"
9 9 9
0.00 ± 0.00
8
All values are the mean ± SEM. n, Number of animals. " P < 0.01 compared with implant size before ovariectomy. 6 P < 0.01 compared with implant size 2 months after ovariectomy.
during estrous. Progesterone levels in serum samples collected during estrus and diestrus were 20.50 ± 2.27 and 66.25 ± 9.07 ng/ml, respectively. Animals treated with tamoxifen did not show the characteristic changes in vaginal smears associated with the estrous cycle. The vaginal smears contained a preponderance of leucocytes, with a variable admixture of nucleated and cornified cells. Serum progesterone levels in random blood samples obtained at the end of tamoxifen treatment were significantly lower than progesterone levels in control rats during diestrus (P < 0.01; Table 3). Effect of tamoxifen on experimentally endometriosis in ovariectomized rats
induced
When rats underwent ovariectomy, after induction of endometriosis, the implants were completely regressed within 2 months of ovariectomy in all animals (Fig. 2). When these rats were then treated with tamoxifen (1 mg/kg-day) for 2 months, eight of nine implants grew back. One implant remained completely regressed at the end of treatment. The mean implant size was significantly larger than that of implants before treatment (Table 2; P < 0.01). In the control group implants remained completely regressed compared to those in the
FIG. 2. The total morphological regression of an endometrial implant after ovariectomy.
tamoxifen-treated group. In a different group of ovariectomized rats with induced endometriosis, none of the implants grew after treatment with tamoxifen (1 mg/kgday) and progesterone (10 mg/kg-day) every day for 2 months. Ovariectomy significantly reduced progesterone levels compared to those in intact rats during estrus (P < 0.05) and diestrus (P < 0.01). In ovariectomized rats receiving tamoxifen, progesterone levels were significantly lower than those in intact rats during tamoxifen treatment (P < 0.01). When ovariectomized rats were treated with tamoxifen and progesterone, serum progesterone levels were higher than those in rats treated with tamoxifen (P < 0.01). Effect of tamoxifen on pHJthymidine incorporation into DNA in rat uterus The effect of tamoxifen alone or in combination with estrogen on thymidine incorporation into DNA by rat
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 April 2016. at 15:17 For personal use only. No other uses without permission. . All rights reserved.
TAMOXIFEN AND ENDOMETRIOSIS
3266
uterus is depicted in Fig. 3. Four weeks after ovariectomy, rat uteri were capable of incorporating thymidine into DNA. Administration of estrogen increased thymidine incorporation into DNA by 5-fold (P < 0.01; Fig. 3). When tamoxifen was administered with estrogen, thymidine incorporation into DNA was significantly reduced compared to that after estradiol treatment alone (P < 0.01; Fig. 3). Thymidine incorporation into DNA was significantly enhanced with tamoxifen treatment alone (P < 0.05) compared to that in the untreated controls. Discussion The present study demonstrates that induction of experimental endometriosis in the female rat is feasible. We have earlier reported the histological characteristics of the endometrial implant identifying the presence of stroma and gland (10). Endometrial implants in rats are ovarian steroid dependent (10). In the present study administration of tamoxifen to rats with experimentally induced endometriosis led to complete regression of ectopic endometrial implants. This corroborates the recent preliminary clinical report of Haber and Behelak (5), who demonstrated regression of ectopic endometrial implants in two patients with endometriosis after tamoxifen therapy. Our study further showed that recurrence of endometriosis is possible after withdrawal of tamoxifen therapy in the rat model. Further studies are required to determine whether longer periods of tamoxifen treatment reduce the chances of recurrence after withdrawal. The nonsteroidal antiestrogen tamoxifen exhibits a paradoxical species-specific pharmacology. The drug is a full estrogen in the mouse (11), a partial estrogen/antiestrogen in humans (12), and an antiestrogen in the chick oviduct (13). In the present study when tamoxifen was administered to ovariectomized rats at the dose level that induced regression of endometrial implants, it enhanced thymidine incorporation into DNA of rat uterus. 60r 50 40 30 20 10
CONTROL
TAM
Endo • 1990 Voll26-No6
However, when administered with estradiol, tamoxifen partially antagonized estradiol-stimulated thymidine incorporation into DNA of rat uterus. This indicates the estrogen/antiestrogen activity of tamoxifen in rats, as in humans. In rats with experimentally induced endometriosis, ectopic endometrial implants completely regressed 2 months after ovariectomy, which confirms our previous finding (10). In our earlier study although we were unable to detect viable endometrial tissue by histology after complete morphological regression, recurrence of endometriosis was observed with estrogen administration. This suggests the presence of estrogen-responsive cells at the implant site after complete morphological regression (10). In the present study administration of tamoxifen to ovariectomized rats with completely regressed implants led to recurrence of endometriosis. We suggest that recurrence of endometriosis during administration of tamoxifen may be due to the estrogen-like activity of the drug. Recently, Ford et al. (6) reported the development of endometriosis in a premenopausal woman treated with tamoxifen for recurrent benign breast disease. The results of this study suggest that development of endometriosis reported in a clinical study during tamoxifen therapy may be related to its partial estrogenic activity. The disparate effects of tamoxifen, that is regression of endometrial implants in rats with intact ovaries but recurrence of regressed endometrial implants in ovariectomized rats, suggest a possible ovarian modulatory effect on the tamoxifen treatment of endometriosis. In rats 2 months after ovariectomy serum progesterone levels were decreased compared to those in the control group. However, significant serum progesterone levels persisted in spite of complete ovariectomy. These may be of adrenal origin, as it has been demonstrated that in rats the adrenal is capable of secreting large amounts of progesterone (14). In rats with induced endometriosis treated with tamoxifen, a normal estrous cycle pattern was not present, based on vaginal cytology. It has been reported that tamoxifen administration blocks ovulation in rats by preventing the ovulatory surge of LH from the pituitary (15). However, when serum progesterone levels were measured during tamoxifen therapy in ovariectomized rats and rats with intact ovaries, the rats with intact ovaries had higher levels of progesterone. When ovariectomized rats with regressed implants were treated with tamoxifen and progesterone, the recurrence of endometrial implants was not observed. This supports the facilitatory role of progesterone in tamoxifen treatment of endometriosis in rats.
TAM4€
FlG. 3. Effect of estrogen (E) and tamoxifen (TAM) treatment on thymidine incorporation into uterine DNA in ovariectomized rats. All values are the mean ± SEM (n = 5).
In clinical studies, the reason for the various effects of tamoxifen in treatment of endometriosis is not clear. Development of endometriosis was reported in a 54-yr-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 April 2016. at 15:17 For personal use only. No other uses without permission. . All rights reserved.
TAMOXIFEN AND ENDOMETRIOSIS
old premenopausal woman (6), while complete regression of endometrial implants was reported during tamoxifen therapy (10 mg twice a day) in patients 24 and 29 yr of age (5). As the present study in rats indicates a modulatory effect of ovarian progesterone on tamoxifen treatment of endometriosis, clinically reported disparate effect of tamoxifen in endometriosis may be related to serum progesterone levels. The rat endometrium, unlike the human endometrium, does not slough off during the estrous cycle. However, there are similarities between the human and rat endometrium in regulation of endometrial growth by ovarian steroids. Recent studies demonstrate that estrogen enhances DNA polymerase activity and DNA content, which are antagonized by progesterone in both the rat and human endometrium (16, 17). This further validates the use of rats with experimentally induced endometriosis as an animal model to investigate the clinically reported disparate effect of tamoxifen. Successful regression of endometriotic implants in rats receiving tamoxifen treatment suggests the potential value of this drug therapy in endometriosis. Tamoxifen offers some important advantages over GnRH agonists (GnRHa), which have been introduced into the management of endometriosis. The loss of trabecular lumbar bone during GnRHa-induced ovarian suppression has been a recent concern (18-20). Tamoxifen, on the other hand, has been demonstrated to inhibit bone resorption in ovarian hormone-deficient rats (21, 22), which has been implicated in the estrogen-like effect of tamoxifen. However, the teratogenic effect of tamoxifen (23, 24) may limit its advantages over GnRHa. Recently, one of the 7-alkyl amide analogs of estradiol (ICI 164,384) has been demonstrated to be a potent antiestrogen devoid of estrogenic activity (25). A comparative study of tamoxifen and ICI 164,384 indicates that pure antiestrogen (ICI 164,384) is free of some of the toxicological effects of the partial agonist (25). The successful regression of endometrial implants with tamoxifen treatment observed in this study clearly indicates that partial selective suppression of estrogen activity can regress ectopic endometrial implants. Further studies with pure antagonist to determine whether partial suppression of endogenous estrogen activity will resolve ectopic endometrial tissue without inducing bone resorption are important. References 1. Kistner RW 1979 Endometriosis and infertility. Clin Obstet Gynecol 22:101
3267
2. Gerbie AB, Merril JA 1988 Pathology of endometriosis. Clin Obstet Gynecol 31:779 3. diZerega GS, Barber DL, Hodgen GD 1980 Endometriosis; role of ovarian steroids in initiation, maintenance, and suppression. Fertil Steril 33:649 4. Dmowski WP 1988 Danazol induced pseudomenopause in the management of endometriosis. Clin Obstet Gynecol 31:829 5. Harber GM, Behelak YF 1987 Preliminary report on the use of tamoxifen in treatment of endometriosis. Am J Obstet Gynecol 156:582 6. Ford MRW, Turner MJ, Wood C, Soutter WP 1988 Endometriosis developing during tamoxifen therapy. Am J Obstet Gynecol 158:1119 7. Rajkumar K, Malinek J, Murphy BD 1985 Effect of lipoproteins and luteotrophins on progesterone accumulation by luteal cells from the pregnant pig. Steroids 45:119 8. Stewart PJ, Webster RA 1983 Intrauterine injection of cholera toxin induces estrogen-like uterine growth. Biol Reprod 29:671 9. Sokal RR, Rohlf FJ 1981 Nested analysis of variance. In: Sokal RR, Rohlf FJ (eds) Biometry. Freeman, New York, p 271 10. Rajkumar K, Schott PW, Simpson CW, The rat as an animal model for endometriosis to examine recurrence of ectopic endometrial tissue following regression. Fertil Steril, in press 11. Harper MJK, Walpole AL 1966 Contrasting endocrine activation of cis and trans isomers in a series of substituted triphenylethylenes. Nature (Lond) 212:87 12. Furr BJA, Jordan VC 1984 Pharmacology and clinical use of tamoxifen. Pharmacol Ther 25:127 13. Sutherland RL, Mester J, Baulieu EE 1977 Tamoxifen is a potent "pure" antiestrogen in the chick oviduct. Nature (Lond) 267:434 14. Ogle TF, Kitay JI 1977 Ovarian and adrenal steroids during pregnancy and the oestrous cycle in the rat. J Endocrinol 74:89 15. Labsetwar AP 1970 Role of estrogens in ovulation: a study using the estrogen antagonist ICI 46,474. Endocrinology 87:542 16. Cummings AM 1989 The role of steroid hormones and decidual induction in the regulation of adenosine diphosphoribosyl transferase activity in rat endometrium. Endocrinology 124:1408 17. Usuki S, Kubota S, Shioda M 1988 Activity of deoxyribonucleic acid polymerase alpha stimulated by estrogen in the endometrium of the human uterus during the menstrual cycle. Gynecol Endocrinol 2:283 18. Lewis V, Ramos J, Dawood MY, Changes in bone mineral content in endometriosis patients treated with GnRH agonist. 34th Annual Meeting of the Society for Gynecologic Investgation, Atlanta GA, 1987, p 181 (Abstract) 19. Cann CE, Henzl M, Burry K, Anreko J, Hanson F, Adamson D, Trobough G, Reversible bone loss is induced by GnRH agonists. 68th Annual Meeting of The Endocrine Society, Anaheim CA, 1986, p 37 (Abstract) 20. Dawood YM, Lewis V, Ramos J 1989 Cortical and trabecular bone mineral content in women with endometriosis: effect of gonadotropin-releasing hormone agonist and danazol. Fertil Steril 52:21 21. Turner RT, Wakley GK, Hannon KS, Bell NH 1988 Tamoxifen inhibits osteoclast mediated resorption of trabecular bone in ovarian hormone deficient rats. Endocrinology 122:1146 22. Jordon VC, Phelps E, Lindgren JU 1987 Effects of antiestrogen on bone in castrated and intact female rats. Breast Cancer Res Treat 10:31 23. Chamness G, Bannayan G, Landry L, Sheridan P, McGuire W 1979 Abnormal reproductive development in rats after neonatally administered antiestrogen/tamoxifen. Biol Reprod 21:1087 24. Clark J, Guthrie S, McCormack S 1981 Neonatal stimulation of the uterus by clomiphene, tamoxifen and nafoxidine: relationship to the development of reproductive tract abnormalities. Adv Exp Med Biol 138:87 25. Wakeling AE, Bowler J 1988 Novel antiestrogens without partial agonist activity. J Steroid Biochem 31:645
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 05 April 2016. at 15:17 For personal use only. No other uses without permission. . All rights reserved.
