Upsala J Med Sci 79: 65-7 I , 1974 Ultrastructure of Human Uterine Epithelium at the Time of Imp I an ta t i on after Post ovu I at ory Administration of Norethindrone OVE NILSSON AND KARL-GOSTA NYGREN From the Department of Anutomy, Biomedical C e n t e r , S-75123 U p p s a l a , und Department of Obstetrics and G y n e c o l o g y , University H o s p i t a l , S-750 14 Uppsulu, S w e d e n ABSTRACT The present investigation was performed to elucidate the mechanism of action of the previously demonstrated contraceptive effect in women of postovulatory ad- ministration of a synthetic gestagen, norethindrone (NET). Seven women participated during 3 control cycles and during 8 treatment cycles, in which NET was given orally after ovulation. Daily peripheral plasma levels of progesterone, estradiol and NET were assayed. An endometrial biopsy was taken in all cycles a t about the expected time of implantation. Light microscopy revealed no consistent differences between non-treat- ment and treatment cycles hut electron microscopy indicated that, after NET treatment. the mitochondria treatment was found to have a certain contra- ceptive effect although the efficacy was not suffi- ciently high t o merit its use in clinical practice. One possible reason for the relatively low con- traceptive efficacy of the NET treatment, in spite of the drastic effect upon the progesterone levels, could be that the drug itself substitutes for the decreased progesterone levels (23). From previous investigations ( I , 2, 7, 8, 10, 17, 26) it is known that NET influences the endometrium, but in these studies the administration of N E T did not follow the present schedule. Therefore, the pres- had grown larger and that nucleolar channel system ent investigation was designed t o study the impact had appeared. These changes suggest an i n c ~ a s e d of postovulatory N E T treatment upon the human progesterone-like influence upon the epithelium, despite uterine epithelium of normal at the time of the decreased progesterone plasma levels, caused by implantation, as visualized by light- and electron- NET. It is assumed that these structural changes, caused by the postovulatory NET treatment, might microscopy, in relation to the peripheral plasma change the functional properties of the endometrium levels of the ovarian steroids and of NET. and thereby impair the possibilities for normal im- plantation. MATERIAL AND METHODS INTRODUCTION Postovulatory administration of synthetic gestagens to women has been shown to markedly reduce peripheral plasma levels of progesterone ( 14). Estradiol levels are also decreased but this de- crease is not so pronounced (23). It was pro- posed that the “luteolytic” effect of such treat- ment might be utilized for postovulatory contra- ception (14). The concept was that the hormonal effects of the treatment would induce changes in the endometrium significant enough to impair the possibility of normal implantation of the blastocyst When tested in clinical trials, in which a syn- thetic gestagen, norethindrone (NET), was given in various doses and dose schedules (23, 24) the (23). Volunteers Seven healthy regularly menstruating women, in the age range 25-30 years, participated in this investigation. Five of the women had previously been pregnant and the remaining 2 women have both become pregnant since the present study. N o pathological changes were found upon pelvic examination (including a vaginal smear for cytology) before the start of the investigation. General design of the investigation A total of I I menstrual cycles were investigated. Daily peripheral venous blood samples were collected, com- mencing before the expected day of ovulation until after the day when implantation is expected to occur in a normal pregnancy. The concentrations of pro- gesterone and estradiol were assayed in each blood sample. From the results of these assays the probable day of ovulation was calculated: After the mid-cyclic peak the first significant decrease in the estradiol level and the second day with a progesterone level above 5-742855 Upsala J Med Sci 79 66 0. Nilsson mnd K . - G . N y w n I ng/ml plasma was called the day of ovulation (14). Progesterone levels of 5 nglml plasma or more were re- garded as postovulatory levels, i.e. a s an indication of ovulation (15). Three of the eleven cycles served a s controls. In eight of the eleven cycles oral N E T treatment was given postovulatory and the N E T plasma levels were assayed. The treatment was aimed to start on the third postovulatory day, when the corpus luteum is functioning well and when, therefore, peripheral blood levels of progesterone have become high enough to permit the conclusion that ovulation has taken place (15). However, for practical reasons t h e results of both the progesterone and the estradiol assays were some- times not available until a couple of days after the day when the blood samples were taken. Therefore, preliminary calculation of the day of ovulation, which was i n some cycles based upon only one or two in- creased progesterone plasma levels, sometimes re- quired correction when all the results of the hormonal assays were available. In three of the cycles, treat- ment had actually been started on the fourth and in one cycle on the fifth postovulatory day, as indicated in the Table. In order to study the influence of varius schedules of treatment the doses of N E T ranged in total from 150 to 300 mg and the period of treatment from 2 to 4 days. This is also indicated in the Table. The women were instructed to take their tablets in the evening and to give their blood sample in the morning. However, one woman (woman D, indicated in the Table) took her N E T dose in the morning about one hour before the blood sample was taken. During each control and treatment cycle an endo- metrial biopsy was obts,ned for light- and electron- microscopy. The timing of biopsy removal was aimed to coincide with the probable time when implantation occurs in a normal pregnancy, i.e. about I week after the time of ovulation (9). For practical reasons the period of time elapsing from the calculated day of ovulation until the biopsy was taken varied between 6 and 8 days. All women were instructed to use mechanical and/or chemical contraceptives during the investigation. The biopsies were immediately fixed by immersion in a 2.5% solution of glutaraldehyde in Soerensen phos- phate buffer (pH 7.4) and were kept in the fixative for periods of one day up to several months. At embedding, small specimens were cut from the biopsies, rinsed in the buffer, and post-fixed for 3 hours in a solution of 1 % osmium tetroxide in Soerensen phosphate buffer, pH 7.4. Dehydration was performed in ethanol, and the specimens were embedded in Epon 812. For light-microscopy, the sections were stained in a basic solution of Toluidine Blue. After microscopical examination appropriate areas of glandular or luminal epithelium were selected and trimmed for sectioning for electron-microscopy. For transmission electron- microscopy, the sections were stained with uranyl acetate followed by lead citrate. For scanning electron- microscopy (21), specimens with an endometrial sur- face area of about I x 1 mm were cut, carefully rinsed in distilled water, and taken to 10% ethanol in water for freeze-drying. The specimens were frozen in iso- pentane cooled by liquid nitrogen and dried in Torr for 3 days at -79°C (solid carbon dioxide). The dry specimens were mounted for scanning electron- microscopy and coated with carbon followed by gold. A Jeol JSM-U3 scanning microscope was used. R E S U L T S Hormonal assays All cycles in this investigation were found t o b e ovulatory, as judged by the levels of progesterone in plasma. The results of the assays of pro- gesterone, estradiol and N E T a r e shown in the Table, which summarizes the daily plasma levels of these hormones from the d a y after t h e cal- culated day of ovulation until the day when the endometrial biopsy was taken. This gives a n estimation of the total amount of the hormones in plasma between the time of ovulation and t h e date of t h e biopsy. Woman A had, initially, high progesterone levels and low estradiol levels compared with those found Hormonal assays in most of the other treatment cycles, but respond- Progesterone in p l a s m a was estimated by a com- petitive protein binding assay, as described by Johans- ed as expected to the N E T treatment. son (13). had comparatively low estradiol levels in both Estradioi in p l a s m a was determined by a radio- cvcles investisated. while in her second treatment Y immunoassay as described by Hotchkiss et ( 1 I ) , modified a s described by Edqvist & Johansson (4). Norethindrone ( N E T ) in p l a s m a was assayed by a cycle progesterone levels were exceptionally low. T h e remaining 5 women, C through G, showed a radioimmunoassay as described by Nygren et al. ( 2 5 ) . more consistent Pattern. Comparing control cycles and treatment cycles from t h e same women ( E , F and G) it was found Endometrial morphology that the summarized progesterone levels were The endometrial biopsies were obtained with a Genell lower i n the treatment cycles. Estradiol levels curette without anesthesia and without dilation of the cervix. The biopsies were taken from the middle part were not changed but it be of the anterior wall of the corpus uteri. noted that in 2 of the women (F and G) the Upsala J Med Sci 79 Endometrium after norethinclrone treutment 67 Table. Peripherul plusmu concentrutions of progesterone, estradiol, und N E T , summurized jrom the day after ovulution until the d u y when the endometriul biopsy M ' U S tuken. The day oj'ovulution is cxilled duy 0 Control cycles Treatment cycles Proges- Proges- terone Estradiol terone Estradiol N E T Biopsy (nglml (pglml N E T treatment Biopsy (ng/ml (pg/ml (nglml Woman on day plasma) plasma) mg on day on day plasma) plasma) plasma) A 100 4-6 7 B 100 5-7 8 l o o 3-4 6 C loo 3-4 7 D 100 4-5 7 E 7 87.1 1475 50 3-6 7 F 7 66.9 I226 50 4-6 8 G 6 73.2 I032 50 3-5 7 138 52.5 18.8 49.5 43.8 41.7 49.2 41.4 445 310 645 I285 I305 I445 I328 995 456 696 332 not assayed 1413" 349 219 I14 0 Woman D took her N E T tablets about one hour before the blood sample was taken. hormonal levels were summarized from o n e more day in the treatment cycle as compared with the corresponding control cycle. O n t h e day of the biopsy the progesterone plasma level in woman E decreased from 14.5 ng/ml in the control cycle to 0.6 ng/ml in the treatment cycle, in woman F from 9. I ngfml t o 0.9 ng/ml and in women G from 15.5 ng/ml t o 5.5 ng/ml. Corresponding estradiol plasma levels were decreased from 245 t o I10 pg/ ml in woman E , from 210 t o 60 pg/ml in woman F and from 250 t o 135 pglml i n woman C. To illustrate the hormonal patterns, the daily plasma levels of progesterone, estradiol and N E T in the control cycle and the treatment cycle of woman E a r e shown in Fig. I . Endometrial morphology Light microscopy. T h e biopsies were examined with regard t o t h e amount of glycogen, stromal oedema, and decidual changes (9, 22, 27). Micro- scopy demonstrated slight differences among the biopsies, but n o difference was sufficiently con- sistent to s e r v e a s a criterion of NET-induced change. However, the glycogen content of t h e uterine epithelium appeared t o increase somewhat after the N E T treatment. Electron microscopy. The ultrastructure of t h e luminal and glandular epithelium from biopsies obtained on the 6th-8th postovulatory d a y of t h e normal untreated women corresponds rather well to what has been previously reported (3, 5 , 6, 12, 18, 20, 21, 28, 29, 31): Apical protrusions were ~ present although they contained only a few glyco- 35. 30. 25. 20. EF ?, Control cycle : *, Days from ovulailon F i g . 1. Woman E. Daily peripheral plasma levels of progesterone, estradiol and N E T during one control cycle and one N E T treatment cycle. In both cycles an endometrial biopsy was taken 7 days after the calculated day of ovulation (=day 0). Upsala J Med Sci 79 68 0. Nilsson and K.-G. Nygren Fig. 2. Luminal surface of normal endometrium, d a y 7 postovulatory. T h e microvilli of the secretory cells are noticed. Several cells also possess apical pro- trusions. x 10000. gen granules (Fig. 2). T h e endoplasmic reticulum was well developed and the Golgi apparatus was rather large. Most mitochondria were in t h e normal size range and only a few giant mito- chondria were observed. Lysosome-like bodies occurred sparsely. The amount of glycogen gran- ules varied but never reached the high levels present a few days earlier in the cycle when glycogen content reaches its peak (28). The ultrastructure of the cells of the luminal and glandular NET-influenced uterine epithelium differed in some aspects from that of the control biopsies. Apical protrusions were present but they generally contained more vesicles and glycogen granules than were observed in biopsies from untreated women (Fig. 3). T h e mitochondria were generally larger (Fig. 4), and giant mitochondria were more often present basally. T h e glycogen granules seemed t o be more numerous, often occurring in groups containing glycogen vacuoles. Nucleolar channel systems were rather frequently observed (Fig. 5 ) . No marked differences related t o the various dose schedules were found. Obviously, this does not exclude t h e existence of functional differences although the technique used in the present in- Upsala J Med Sci 79 vestigation did not reveal a n y structural differences. T h e biopsies from woman A showed a glandular epithelium containing more dense granules than were usually observed. DISCUSS ION T h e duration between the ingestion of N E T tablets and the blood sampling for the determination of plasma N E T levels was not entirely consistent, f o r practical reasons. After oral ingestion. N E T has been found t o have a plasma half-life, after its initial peak, of about 9 hours only (25). Thus. the plasma levels of N E T recorded in the present investigation cannot be used for strictly quantitative comparison but merely as an indicator that the women took their tablets as intended. A decrease in progesterone and estradiol plasma levels was found during treatment cycles at the d a y of biopsy, as anticipated from previous ex- perience with postovulatory N E T treatment (14, 23). O n e woman ( A ) had extremely high progesterone levels a s compared with the o t h e r women in this study. On examining the electronmicrographs, the uterine epithelium of this woman was found t o contain more dense granules than did the epithe- lium from t h e other NET-treated women. However, since this i s a n observation from o n e single wo- man, no conclusion can a t present b e drawn from t h e finding. Further, it must b e remembered that o u r knowledge of t h e inter-relationship between a hormonal plasma concentration and its effect upon the target tissues is meager. F o r instance, we d o not yet know the time lag between a change in t h e plasma level of a hormone and the structural response it may induce in the uterine epithelium. Therefore, in the present communication only general trends in hormonal and structural changes have been considered. Light microscopical findings indicating that the glycogen content of the gland cells increased after N E T treatment a r e not particularly conclusive, con- sidering the inconsistency of t h e findings and t h e relatively few biopsies available. Although electron microscopical observations imply ,various diffi- culties (19), changes were found which seem to be sufficiently consistent t o warrant the con- clusion that N E T did affect the uterine epithelium. T h e changes comprised t h e appearance of large mitochondria and nucleolar channel systems. Both EtidomcJtriurn ufter norethindrone treutmcnt 69 F i g . 3. Luminal part of uterine epithelium, day 7 trusion is observed. I t contains vesicles a n d glycogen postovulatory after N E T treatment. A n apical pro- granules. X20000. these features a r e regarded, by most authors, as an effect of progesterone ( I . 7, 17, 30). Since the serum levels of both progesterone and estradiol were lower in the treatment cycles at the time of biopsy, the progesterone-like action on the epithe- lium is probably d u e to t h e synthetic gestagen. Results from animal studies (16) suggest that the relation between the action of progesterone and N E T upon uterine hormone receptors might b e of great importance in this respect. Several explanations for a contraceptive effect of the N E T treatment a r e possible. T h e tubal function might be affected. The proportion of tubal pregnancies, found in previous clinical trials with postovulatory N E T treatment was unexpected- ly high (24), even though t h e actual number was too low t o provide conclusive statistical evidence of a n increased incidence of tubal pregnancies. Furthermore, the luteotrophic effect of human chorionic gonadotropin ( H C G ) from the blastocyst upon the corpus luteum after successful implantation, might b e affected by the presence of N E T . However, both experiments with exogenous HCG administered to women treated with N E T after ovulation (14) and the normal hormonal and histological development of pregnancies occurring after such treatment (24) suggest that this is not a probable explanation. The most likely explana- Upsala J Med Sci 79 70 0. Nilsson and K . - G . N y g r e n F i g . 4 . Basal part of glandular uterine epithelium, day 7 postovulatory after N E T treatment. The mito- chondria generally are larger than at a similar stage of a control cycle. ~ 3 0 0 0 0 . tion is perhaps that the NET-induced changes observed in the uterine epithelium might result in a secretion which is not optimal for t h e survival of the blastocyst or in a n extra-cellular coat on the luminal surface, counteracting the attachment of the blastocyst. It is possible that the use of a drug with a more potent progesterone-like action might in- fluence t h e uterine epithelium more drastically, thereby giving a better contraceptive efficacy than postovulatory N E T treatment. On the other hand a drug having the same depressive effect a s N E T upon the function of the corpus luteum but which F i g . 5 . Nucleolar channel system of glandular uterine epithelium, day 8 postovulatory after N E T treatment. x25 000. does not exhibit progesterone-like actions o n the uterine epithelium might also induce a more pronounced disturbance in the development of the endometrium. A C K N O W L E D G E M E N T This investigation was supported by grants from the International Committee for Contraception Research of the Population Council (No. M71, 138), the Ford Foundation (grant to Prof. Gemzell, no. 66-405-A) and from the Swedish Medical Research Council (B74- 17X-3495-03B). The authors are indebted to their technical staffs for skilled assistance. The assay of progesterone and estradiol by Dr Elof D. B. Johansson is gratefully acknowledged. R E F E R E N C E S I . Ancla, M., de Brux, J . , Belaisch, J . & Musset, R.: Influence de I’equilibre oestrogene-progesterone sur les ultrastructures de I’endometre humain. I . Morphologie et evolution de corpuscules intra- nucleaires presents dans les cellules glandulaires de I’endometre sous I’effet de la norethisterone et dans les sterilites essentielles. Gynec Obstet (Paris) 63: 239, 1964. Upsala J Med Sci 79 Endometrium ufter norethindrone treatment 7 I 17. Kohorn. E. I . . Rice. S. I.. Hemperly, S . & Gordon, M.: The relation of the structure of progestational steroids to nucleolar differentiation in human endo- metrium. J Clin Endocr34: 257. 1972. 18. Morgenroth. K. & Verhagen. A.: Sublichtniikro- skopische Obertlachenstruktur des Endometrium. Arch Gynaek 212: 30, 1972. 19. Nilsson, 0.: Electronmicroscopy of the glandular epithelium in the human uterus. I . Follicular phase. J . Ultrastruct Res6: 413, 1962. 20. - Electron microscopy of the glandular epithelium in the human endometrium. 11. Early and late luteal phase. J. Ultrastruct Res 6: 422. 1962. 2 1 . Nilsson. 0. & Nygren, K. G.: Scanning electron microscopy of human endometrium. Upsala J Med Sci 77: 3, 1972. 22. Noyes, R. W . , Hertig. A . T. & Rock. J . : Dating the endometrial biopsy. Fertil Steril 1: 3, 1950. 23. Nygren, K. G . . Johansson, E. D. B. & Wide, L . : Postovulatory contraception in women with large doses of norethindrone. Contraception 5 : 445, 1972. 24. - The contraceptive efficacy of 10 mg or 25 mg of norethindrone given orally on day 15-21 of the human menstrual cycle. Contraception. in press. 2 5 . Nygren, K . G . , Lindberg. B . , Martinsson, K . , B o w , W. T. K. & Johansson. E. D. B.: Radioimmuno- assay of norethindrone: peripheral plasma levels after oral administration to humans and rhesus monkeys. Contraception, i n press. 26. Rice-Wray, E., Schulz Contreras, M . & Rosell, A . A , : Endometrial changes with the administration of norethindrone. Ginec Obstet Mex 17(98): 333, 1962. 27. Rock, J . & Barlet, M. K.: Biopsy studies of human endometrium. J Amer Med Ass 108: 2022, 1937. 28. Sakuma, S.: Glykogengenhalt und- verteilung in der Uterusschleimhaut der Frau wahrend des nor- malen Zyklus im elektronenmikroskopischen Bild. Beitr path Anat 140: 454, 1970. 29. Sengel, A . & Stoebner, P . : Ultrastructure de I’endometre humain normal. 11. Les glandes. Z Zellforsch 109: 260, 1970. 30. Terzakis, J . A.: The nucleolar channel system of human endometrium. J Cell Biol27: 293, 1965. 31. Wynn, R . M. & Woolley, R. S.: Ultrastructural cyclic changes in the human endometrium. 11. Normal postovulatory phase. Fertil Steril 18: 721, 1967. 2. Boschann, H. W. & Kur, S.: The effect of 17- aethinyl-19-nortestosterone-enanthate. A new long- acting gestagen on the endometrium and the atrophic vaginal epithelium. Geburtsh Frauenheilk 17: 928, 1957. 3. Cavazos, F., Green, J. A , , Half, D. G . & Lucas, F. V.: Ultrastructure of the human endometrial glandular cell during the menstrual cycle. Amer J Obstet Gynec 99: 833, 1967. 4. Edqvist, L. E. & Johansson. E. D. B.: Radio- immunoassay of estrone and estradiol in human and bovine peripheral plasma. Acta Endocr (Kbh) 71: 716. 1972. 5 . Ferenczy, A,, Richart, R. M., Agate, Jr, F. J . , Purkerson, Mable L. & Dempsey, E. W.: Scanning electron microscopy of the human endometrial sur- face epithelium. Fertil Steril23: 515, 1972. 6. Ferenczy. A. & Richart, R . M . : Scanning and transmission electron microscopy of the human endometrial surface epithelium. J Clin Endocr Metabol36: 999, 1973. 7. Ferin, J. & Delforge. J. P . : Endometrial assessment of relative potency of contraceptive progestational compounds. Mode of action of luteal supplementa- tion. Histochemical study. Proc 3rd I n t Congr Hor- monal Steroids. Hamburg Sept. 1970. Int Congr Ser 219. Excerpta Medica Foundation (Amsterdam) 878, 1971. 8. Henzl. M.. Jirasek, J., Horsky, J . & Presl, J.: Proliferative effect of 17-alpha-ethinyl-19-nortesto- sterone. Arch Gynaek 199: 335, 1964. 9 . Hertig, A. T.: Gestational hyperplasia of endo- metrium. A morphologic correlation of ova, endo- metrium and corpora lutea during early pregnancy. Lab Invest 13: 1153, 1964. 10. Hertz, R., Waite, J. H. & Thomas, L. B . : Pro- gestational effectiveness of 19-nor-ethinyl-testo- sterone by oral route in women. Proc Soc Exp Biol Med 91: 418, 1956. 1 1 . Hotchkiss, J., Atkinson, L . E. & Knobil, E.: Time course of serum estrogen and luteinizing hormone (LH) concentrations during the menstrual cycle of the rhesus monkey. Endocrinology 89: 177, 1971. 12. Johannisson, E & Nilsson, L.: Scanning electron microscopic study of t h e human endometrium. Fertil Steril23: 613, 1972. 13. Johansson, E. D. B.: A simplified procedure for the assay of progesterone. Acta Endocr (Kbh), Suppl. 147: 188, 1970. 14. -Depression of the progesterone levels in women treated with synthetic gestagens after ovulation. Acta Endocr (Kbh) 68: 779, 1971. 15. Johansson, E. D. B., Wide, L. & Gemzell, C.: Luteinizing hormone (LH) and progesterone in plasma and LH and oestrogens in urine during 42 normal menstrual cycles. Acta Endocr (Kbh) 68: 502, 1971. 16. Janne, O., Kontula, K., Rajakoski, E . , Tanhuan- paa, E. & Vikho, R.: Specificity of progesterone- binding proteins. Acta Endocr (Kbh), Suppl. 177: 249. 1973. Received December 23, 1973 A d d r e s s f o r reprints: K.-G. Nygren, M. D. Department of Obstetrics & Gynecology University Hospital S-750 14 Uppsala Sweden Upsala J Med Sci 79