Inflammatory reaction in endometriotic tissue: an immunohistochemical study

European Journal of Obstetrics & Gynecology and Reproductive Biology, 48 (1993) 43-50

43

0 1993 Elsevier Scientific Publishers Ireland Ltd. All rights reserved. 0028-2243/93/%06.00

EUROBS 01470

Inflammatory reaction in endometriotic tissue: an immunohistochemical study U. Cirkel”, H. Ochs”, B. Muesb, G. Zwadlob, C. Sorgb and H.P.G. Schneider” ‘Department of Obstetrics and Gynecology and bhtstitute of Experimental Dermatology. University of Miktsrer. Germany

Accepted for publication 8 October 1992

Summary A panel of monoclonal antibodies specific for macrophage subtypes appearing in early (27ElO), downregulatory (RM3/1) and late (25F9) stages of inflammation had been applied to 20 endometriotic implants of 14 women. Of those patients 9 were in the follicular phase of the cycle, two on danazol,.one on LHRHanalogue (buserelin) and another two on oral contraceptives. Beside the macrophage subsets, antibodies against T4, T8 lymphocytes as well as proliferating cells (EN7/44 and Ki67) were examined. In all specimens immunologically competent cells could be detected to a varying degree and within the same implant different stages of inflammation were discernible. Endometriosis presented with signs of early inflammation indicated by 27ElO+ macrophages and CD4+ lymphocytes (15 specimens) and with downregulatory, late inflammatory reactions as shown by RM3/1+, 25F9+ macrophages and CD8+ lymphocytes (19 biopsies). Additionally, in 14 specimens cells of the EN7/44+ and Ki67+ type were detected. These preliminary results showed no significant correlation to either extension of endometriotic implants or adhesions or concomitant therapy and clearly indicate, that there is an immunological dynamic process within the lesion itself in addition to that one of the peritoneal fluid. Endometriosis; Immunity; Macrophages; Lymphocytes

Introduction Essential questions regarding etiology and remain pathophysiology of endometriosis unanswered until now. Since Sampson [l] had postulated that uterine endometrium implants via retrograde menstruation and gives origin to endometriotic foci in the peritoneal cavity, it is still unknown, why some women develop the disease Correspondence IO: Dr. U. Cirkel, Universitiits-Frauenklinik Miinster, Albert Schweitzer Str 33, 4400 Miinster, FRG.

while others don’t although nearly every woman experiences the retrograde blood flpw [2,3]. Koninckx et al. developed the concept, that implantation of flushed endometrium is caused by an altered local hormonal milieu of the peritoneal fluid (PF) resulting from luteinized unruptured follicles [4]. In recent years the hypothesis was favoured of an immunological deficiency in endometriosis patients. Support to this idea derives from the demonstration of a decrease in cellmediated immune response in rhesus monkeys [5] and an increase in number and activity of macro-

44

phages and lymphocytes in peritoneal fluid in endometriosis patients [6-91. In contrast to these studies dealing with the peritoneal environment we concentrated on the endometriotic implant itself. By means of immunohistochemical methods we classified leucocyte and lymphocyte subpopulations in the endometriotic lesions and tried to delineate any evidence for a local disturbance of immunity in endometriosis patients. Patients and Methods Clinical observations Endometriosis biopsies were obtained during laparoscopy or laparotomy in 14 women aged 28.8 years (range: 19-34 years). The operation was performed as part of an infertility workup and/or for pelvic pain in the late follicular phase of the menstrual cycle in 9 patients. Three women underwent a second look operation immediately after cessation of danazol therapy (600 mg/day orally for 6 months) or LHRH-analogue (900 &day intranasally of buserelin for 3 months) to assess the effect of treatment. Two others had been examined who used oral contraceptives (ethinylestradiol/ levonorgestrel monophasic preparations, used for more than 3 months). Staging of the disease was performed according to the AFS criteria of 1985 [lo]. Two patients were classified as AFS I, 7 as AFS II and four as AFS III. In one woman only the vagina was affected, while no other deposits could be detected. The total of 20 specimens originated from the following locations: 7 from the pouch of Douglas, 7 from the ovaries, 3 from the vesicouterine pouch, 2 from the sigmoidorectal region and 1 from the vagina. In four women multiple biopsies were taken: in two from two different and in another 2 from 3 different locations. According to history, routine laboratory findings, and gynecological examination none of the patients neither presented general or a localized pelvic inflammation. Histological verification of endometriosis was done by preparing hematoxilin and eosin-stained sections adjoining the specimen for immunohistochemistry and by careful examination of the immunohistochemically-stained slice itself.

Immunological methods For immunohistochemical examination biopsies were snap-frozen in liquid nitrogen immediately after excision and stored at -8O’C. Frozen serial cryostat sections, 5 pm thin, were cut with a microtome (SLEE, Mainz, FRG), fixed in acetone at room-temperature and air-dried. Thereafter the sections were washed with phosphate buffered saline (PBS) and incubated in 0.4% sodium-azide with 0.1% H202 in PBS for 7 min in order to inactivate endogenous peroxidases. After washing twice in PBS the indirect immunoperoxidase staining was performed. Preincubation was carried out with 1% bovine serum albumin (BSA) for 30 min. Thereafter each biopsy was incubated with 100 ~1 of the specific monoclonal antibody (mAb). The antibodies used to identify lymphocyte and macrophage subpopulations as well as proliferating cells are listed in Table I. The mAb 910D7, serving as positive control, reacts with an formaldehyderesistant epitope on the non-polymorphic part of the bimolecular complex of HLA-DR antigens [ 151. Mouse IgG and IgM and for the biotinylated lC5 biotin were included as negative control. After washing three times with PBS the sections were incubated for 30 min with the corresponding peroxidase conjugate at a dilution of l/100 in PBS, supplemented with 1% BSA. Incubation steps were done at 37°C. After three additional washes bound peroxidase was developed in the dark with 3-amino-4-ethylcarbazole (AEC, Sigma, Miinchen) at pH 4.9 in 0.1 m acetate buffer plus 0.015% HZ02 for 7 min. Counterstaining was performed with Mayer’s Hemalaun (Merck, Darmstadt, FRG). Semiquantitative evaluation of positive staining was done at a 200 x magnification: 0:

(+): +: ++: +++:

no positive cell per vision field l-5 positive cells 6- 10 positive cells 1l-20 positive cells >20 positive cells

Results

Distribution of the monoclonal antibodies In all of the samples tested a quite heterogenous

45 TABLE I Monoclonal antibodies used in the study mAb

Speciticity

Reference

Subclass

27 E 10 RM 3/l 25 F9

Zwadlo 1986 [II] Zwadlo 1987 [12] Zwadlo 1985 (131 Michels 1985 [14] Sorg 1985 [15]

IgG 1 IgG I IgG I

EN 1144 Ki 67 IOT-4 CD 4

Macropbages in early types of inflammatory processes Macropbages in the healing phase of inflammation Mature macrophages in the late stage of inflammation Intermediate type of macrophages producing MIF Epitope on the nonpolymorphic part of the bimolecular complex of HLA-DR antigens Proliferating endothelial cells Proliferating cells Helper/inducer lymphocytes

IgM IgG 1

IOT-8 CD 8

Cytotoxkuppressor

Hagemeier 1986 116) Gerdes 1983 [17] Immunotech Lab.’ Marseille, France Immunotech Lab.a Marseille, France

1 cs 910 D 7

lymphocytes

IgG 2 a

Wnpublisbed data.

appearance of most of the identified macrophage and/or lymphocyte subpopulations was present. There had been specimens with only few and rare positive reactions with the monoclonal antibodies, whereas on the other hand some of those were highly positive for nearly all of the markers used (Table II). Reside the description of the quantity of the immunological response there had been qualitative differences in the distribution of the individual antibodies throughout all of the examined samples. Patient Ja (number 6) presented with about 20 mononuclear cells of the 27ElO+ type, which is standing for the early phase of an inflammatory reaction. For three specimens either RM3/1 or 25F9 was the dominating antibody indicating the healing or late stage of inflammation with signs of chronicity. In two of the cases with pronounced 25F9+ cells T8 lymphocytes showed a moderate reaction as well (1 I-20 cells/vision field), which is additionally stressing a late form of immune response. Another four specimens were positive for nearly all antibodies investigated in this study. Early, intermediate and late macrophage subsets were seen beside cells with proliferating capabilities (Ki67, EN7/44) indicating a great variety of immunoregulatory mechanisms within the same tissue.

Correlation to clinical parameters

For further evaluation of the immunohistochemical data association of these findings to clinical parameters had been investigated. Table III lists implant and adhesion score - adding up to the total AFS score of 1985 - together with the hormonal treatment of all of the patients. The severity of the disease as judged by the total AFS score showed no relation to immunohistochemical findings. Samples with only few positive reactions with the monoclonal antibodies were present in endometriosis stage I, II and III. For further elucidation of these findings implant and adhesion score had been looked at separately. Pronounced adhesive disease (1 7 AFS points) was related to a high number of positive cells and to hardly any immunohistochemical staining as well. The same holds true for extensive endometriotic implants (> 10 AFS points). Danazol or LHRH-analogue treatment did not provoke uniform reactions in residual endometriosis as found during second look operation immediately after cessation of therapy. The two cases after 6 months of 600 mg danazol demonstrated only few positive cells, whereas the samples of the patient following three months buserelin were strongly positive for nearly all of the markers

vo II We I We II We III Wi I Wi II Wi III

Pi II Sc Ti vo I

30 a

-

LHRH-analogue

-

30a

28 a

-

Neogynon 21@b

29a 34a

31a

Neostediril@ a -

Pouch of Douglas Pouch of Douglas Ovary Ovary Pouch of Douglas Ovary Ovary

Ovary Ovary Pouch of Douglas Vesicouterine pouch

Vagina Pouch of Douglas Pouch of Douglas Vesicouterine pouch Sigmoidorectal region Vesicouterine pouch Pouch of Douglas Ovary Sigmoidorectal region

Danazol Danazol -

34a 30a 29a 34a 22a 19a 28a 25a

Do Eh EP He Ho Ja La MU Pi I

aFifty micrograms ethinylestradiol + 0.125 mg levonorgestrel. bFifty micrograms ethinylestradiol + 0.250 mg levonorgestrel.

(14)

(13)

(10) (11) (12)

(3) (4) (5) (6) (7) (8) (9)

(1) (2)

Localization of biopsy

Therapy

(age)

Patient

Distribution of antibodies within the endometriotic implant

TABLE II

t+,

(+) + 0 ++ +++

25 F 9

0 + 0 + (+I +++ +++ ii+ 0

0 + ++

(o+, +++

0

++ 0

;=, (+)

(+) +

:+I

lC5

0 0 ++ ++

F+)

0 +

0 0

0 0 (+I + ++ 00

ICi 67

(+) + ++ ++

+ + 0 cl

0

F+,

0

0 0 (+) + ++

EN II44

0 0 0

I:;

0 0 0 +

0

+ +

0 0 0 + ii 0 0 0

0 + +++ +++ 0 +++ +

0 + + ++

0 ++ ++ 0 + 0 +

IOT-8 CD8

IOT-4 CD4

; + + h-i +++ +++ ii+ +++ +++ +++ ii + 0 +++ ++ +++ + 0

0

0 ++ ++ 0

0

(+) +++ 0 0

+ :+I ii +

(+) 0 + + ii 0

RM 3/I

0 +++

t+, (+)

0

(+) 0 0 + ++ ++ 0 0 ++

21 E 10

TABLE III Operative findings (r AFS 1985) Total AFS score

Therapy

(1) (2)

Do Eh

Danazol Danazol

(3) (4) (5) (6) (7) (8) (9)

EP He Ho Ja La MU Pi Pi II SC Ti vo I vo II We I We II We III Wi I Wi II Wi III

-

I

-_

3 7 3 3 7 7

8 0 0 6 3 3 0 10

14 9 2

6 0 3

III II 1

12

7

III

18

4

III

(10) (11) (12) (13)

(14)

Neostediril@ Neogynon 21@

LHRH-analogue

Implants

Adhesions

Patient

0

tested - 25F9+ being dominant in the first specimen. The number of investigated patients however is yet too small to draw any definite conclusions. The same is true for those women on oral contraceptives. Further on there was no clear relation to the location, where biopsies had been taken. Even multiple specimens from the same patient were not necessarily identical in immunohistochemical findings. Discussion Endometriosis has, been regarded as a chronic intraperitoneal inflammatory process associated with infertility even when normal ovulatory function and anatomically patent oviducts are present [6]. Therefore many efforts have been made to detect cellular and noncellular components attributing to the disease. Until now most of the investigations relate to the peritoneal fluid. Cellular composition of this compartment consists of 90% macrophages, 8% lymphocytes and 2% poly-

0 (vagina ~0s.) II II I II II II II III

morphonuclear leukocytes [6]. In endometriosis patients the number of macrophages apparently is increased [6,8,18,19] and their functional state has obviously changed [7,20]. Significantly elevated levels of interleukin 1 (IL-l) and the tumor necrosis factor (TNF) - both products of activated macrophages - could be demonstrated in the peritoneal fluid by Fakih [21] and Eisermann [22], resp. The altered peritoneal environment of endometriosis is in accordance with the intensification of sperm phagocytosis by macrophages in vitro [23], the negative effect on sperm/ova interaction in a mouse fertilization assay 1241, the reduced capability of sperms to penetrate zonafree hamster eggs incubated with macrophages of endometriosis PF [25] and the increased ovum capture time of timbria in the presence of endometriotic implants [26]. In contrast to these studies dealing with the altered composition of the peritoneal fluid our study concentrated upon the implant itself. We are able to show, that macrophages and lymphocytes are present in the endometriotic foci themselves, even though there is

48

obviously a great variety in the maturation and distribution of these cells among different locations in the pelvis and between all of the patients examined so far. The origin of these immunocompetent cells could be the peritoneal cavity, from where they might migrate into the surrounding tissue like the ovary and the peritoneum itself. On the other hand blood monocytes perfuse the pelvic capillaries, could diffuse into the perivascular region and differentiate to macrophages induced by several local factors. According to the studies of Zwadlo et al. [ 1 l- 131 macrophage-subsets can be determined by several monoclonal antibodies. 27ElO detects a surface antigen (mol. wt. 17 000 f lOOO), which is expressed on macrophages in an acute type of inflammatory reaction [ll]. In our study 13 specimens, 6 of them with more than 10 positive cells in the endometriotic implant, showed this antigen indicating the acute beginning of an immune response. On the other hand 80% of samples were positive for 25F9 detecting a surface-antigen with a mol. wt. of 86 000. The presence of this subset is characteristic for the mature tissue fixed macrophage standing for a state of chronicity [ 131. These alterations are paralleled by the appearance of CD4+ and CDS+ cells. While there were only 6 out of 20 specimens positive for the helper/inducer type of lymphocytes, the majority of cases (75%) was indicative for the suppressor/cytotoxic T-8 cells, a conformation of the late stage of inflammation. The investigations had been completed by measurements of RM3/1+ cells - being present in the down regulatory phase of an inflammatory process [12] and those intermediate macrophages producing the migration inhibiting factor (MIF), which are indicated by a positive reaction with the human mAb lC5 [27]. First reports on MIF date back to 1966 and meanwhile it is proven, that under in vitro conditions macrophages in the presence of MIF are able to produce IL-l and to induce T-4 lymphocytes [ 141. RM3/1 and lC5 positive staining is quite comparable in all of the samples investigated with 6 out of 20 specimens showing more than 10 cells/vision field for both markers. These results are indicative for the down regulation of an inflammatory process, which might lead to a state of

chronicity. Additionally the presence of lC5+ cells is supporting those studies reporting about elevated levels of IL-1 in the presence of endometriosis. Another important aspect of the study is the demonstration of Ki67 and EN7144 in endometriotic implants. Angiogenesis as proven by the antibody EN7144 (IgM-class) is present during embryonic development, ovulation, wound-healing, inflammatory reaction and tumor-development [ 161. Induced by angiogenetic signals endothelial cells of capillaries will start to migrate into the surrounding tissue, this way destroying the basal membrane of their original vessels. Thereafter growth of new capillaries will start. Ki67 as well belongs to the group of those monoclonal antibodies detecting proliferating cells [ 171. In our study EN7144 and Ki67 are positive in some endometriotic foci, even though their appearance within the specimens investigated is quite heterogeneous. On the other hand these data present another proof for the proliferative process of the disease. Even though in our investigation the number of patients after specific therapy for endometriosis (danazol, LHRH-analogue buserelin) is yet too small to draw any definite conclusions, some positive effect on the immune response following danazol appears likely. This is in accordance with results of El-Roeiy and coworkers [28], who reported about gradually decreased autoantibodies in concentration and in number/patient after danazol, not so however after treatment with LHRH-analogues. Another study found danazol immunosuppressive by inhibiting lymphocyte proliferation in cultures [29]. Further efforts with greater numbers of patients and at comparable intervals of therapy in the different treatment protocols are mandatory in order to prove the effect of those substances on mononuclear cell subsets in the endometriotic implants themselves. Although the results of our study are heterogenous concerning patients and different locations of endometriotic foci, we can clearly demonstrate the presence of macrophage and lymphocyte subtypes as well as proliferating antigens in the investigated implants. The inflammatory sequence of chemoattraction, extravasation,

49

differentiation of monocytes to macrophages obviously is at work, even though there is a great variety of differentiation between specimens and patients. Additionally these data might give further support to the hypothesis of endometriosis to be an autoimmune disease in a way that this condition may create an antigen from the host’s individual endometrial proteins to be recognized by the host’s immune mechanisms as foreign [30]. Ectopic endometrial tissue originated by retrograde menstruation, coelomic metaplasia, direct or canalicular (haematogenicfiymphogenic) invasion or by a composition of all these mechanisms - might produce several local factors, e.g. proteins, which will gradually attract monocytes. These mononuclear cells could proliferate, differentiate and be activated according to the local environment. After being activated macrophages will produce substances like IL-l and TNF, which are responsible for the development of fibroblast proliferation, collagen deposition, T-cell and enzymatic activation. All these alterations are building up a dynamic process, which might be affected at several steps leading to a great variety in the immune response between specimens and patients. Further studies should be directed to characterize those substances being created by endometriotic lesions and to point out those mechanisms by which the development of the disease might be impeded and possibly be stopped. In particular the role of immunosuppressive substances in the treatment of endometriosis should be clarified more extensively. Just recently Simon and co-workers [31] found a significant decreased embryotoxicity of serum from infertile patients with minimal to moderate endometriosis in the in vitro development of two-cell mouse embryos after 3 days of treatment with glucocorticoids. These steroids are known to inhibit the production of IL-l, IL-6 and TNF in vivo. According to these mechanisms endometriosis treatment seems to benefit from direct inhibitory effects on the local immune response. References 1 Sampson JA. Peritoneal endometriosis due to menstrual dissemination of endometrial tissue into peritoneal cavity. Am J Obstet Gynecol 1927;14:422-469.

2

3

13

14

15

16

Bhunenkrantz MJ, Gallagher N, Bashore RA, Tenckhoff H. Retrograde menstruation in women undergoing Obstet chronic peritoneal dialysis. Gynecol 1981;57:667-670. Halme J, Hammond MG, Hulka JF, Raj SG, Talbert LM. Retrograde menstruation in healthy women and in pawith endometriosis. Obstet tients Gynecol 1984;64:151-154. Koninckx PR, Ide P, Vandenbroucke W, Brosens IA. New aspects of the pathophysiology of endometriosis and associated infertility. J Reprod Med 1980;24:257-260. Dmowski WP, Steele RW, Baker GF. Deficient cellular immunity in endometriosis. Am J Obstet Gynecol 1981;141:377-383, Haney AF, Muscat0 JJ, Weinberg JB. Peritoneal fluid cell populations in infertility patients. Fertil Steril 1981;35:696-698. Halme J, Becker S, Hammond MG, Raj MHG, Raj S. Increased activation of pelvic macrophages in infertile women with mild endometriosis. Am J Obstet Gynecol 1983;145:333-337, Hill JA, Faris HMP, Schiff I, Anderson DJ. Characterization of leukocyte subpopulations in the peritoneal fluid of women with endometriosis. Fertil Steril 1988;50:216-222. Badaway SZA, Cuenca V, Kaufmann L, Stitzel A, Thompson M. The regulation of immunoglobulin production by B cells in patients with endometriosis. Fertil Steril 1989;51:770-173. Revised American Fertility Society Classification of Endometriosis: 1985. Fertil Steril 1985;43:351-352. Zwadlo G, Schlegel R, Sorg C. A monoclonal antibody to a subset of human monocytes found only in the peripheral and J lmmunol blood inflammatory tissues. 1986;137:512-518. Zwadlo G, Voegeli R, Schulze Osthoff K, Sorg C. A monoclonal antibody to a novel differentiation antigen on human macrophages associated with the down-regulatory phase of the inflammatory process. Exp Cell Biol 1987;55:295-304. Zwadlo G, Brijcker E-B, v Bassewitz D-B, Feige U, Sorg C. A monoclonal antibody to a differentiation antigen present on nature human macrophages and absent from monocytes. J Immunol 1985;134:1487- 1492. Michels E, Burmeister G, Malomy U, Sorg C. Phenotypic alterations induced in macrophages by migration inhibitory factors. In: Sorg C, Schimpl A, eds. Cellular and molecular biology of lymphokines, Orlando: Academic Press, Inc., 1985;321-327. Sorg C, Briicker E-B, Zwadlo G, Redmann K, Feige U, Ax W, Feller AC. A monoclonal antibody to a formaldehyde-resistant epitope on the nonpolymorphic constant part of the HLA-DR antigens. Transplantation 1985;39:90-93. Hagemeier H-H, Vollmer E, Goerdt S, Schulze-Osthoff K, Sorg C. A monoclonal antibody reacting with endothelial cells of budding vessels in tumors and inflammatory tissues and non-reactive with normal adult tissues. Int J Cancer 19#6;38:481-488.

50

17 Gerdes J, Schwab U, Lemke H, Stein H. Production of a mouse monoclonal antibody reactive with a human nuclear antigen associated with cell proliferation. Int J Cancer 1983;31:13-20. 18 Hahne J, Becker S, Hammond MC, Raj S. Pelvic macrophages in normal and infertile women: The role of patent tubes. Am J Obstet Gynecol 1982;142:890-895. 19 Chacho KJ, Stronkowski Chacho M, Andresen PJ, Scommegna A. Peritoneal fluid in patients with and without endometriosis: Prostanoids and macrophages and their effect on the spermatozoa penetration assay. Am J Obstet Gynecol 1986;154:1290-1299. 20 Hoffman M, Haney AF, Weinberg JB. Reduced trypsinbinding capacity of a*-macroglobulin in the peritoneal fluid of women with endometriosis: possible relevance to alterations in macrophage function. Fertil Steril 1988;50:39-47. 21 Fakih H, Baggett B, Holtz G, Tsang K-Y, Lee JC, Willimson HO. Interleukin-1: a possible role in the infertility associated with endometriosis. Fertil Steril 1987;47: 213-217. 22 Eisermann J, Gast M, Pineda J, Odem RR, Collins JL. Tumor necrosis factor in peritoneal fluid of women Steril Fertil undergoing laparoscopic surgery. 1988;50:573-579. 23 Muscat0 JJ, Haney AF, Weinberg JB. Sperm phagocytosis by human peritoneal macrophages: A possible cause of infertility in endometriosis. Am J Obstet Gynecol 1982;144:503-510.

24

25

26

27

28

29 30

31

Sueldo CE, Lambert H, Steinleitner A, Rathwick G, Swanson J. The effect of peritoneal fluid from patients with endometriosis on murine sperm-oocyte interaction. Fertil Steril 1987;48:697-699. Chacho KJ, Andresen PJ, Scommegna A. The effect of peritoneal macrophage incubates on the spermatozoa assay. Fertil Steril 1987;48:694-696. Suginami H, Yano K, Watanabe K, Matsuura S. A factor inhibiting ovum capture by the oviductal Iimbriae present in endometriosis peritoneal fluid. Fertil Steril 1986;46:1140-1146. Burmeister G, Sorg C, Tarcsay L. Production and characterization of a monoclonal antibody to human MIFs. In: Sorg C, Schimpl A, eds. Cellular and molecular biology of lymphokines. Orlando: Academic Press, Inc., 1985;315-320. El-R&y A, Dmowski WP, Gleicher N, Radwanska E, Harlow L, Binor Z, T-on I, Rawlins RG. Danazol but not gonadotropin-releasing hormone agonists suppresses autoantibodies in endometriosis. Fertil Steril 1988;50:864-871. Hill JA, Barbieri RL, Anderson DJ. Immunosuppressive effects of danazol in vitro. Fertil Steril 1987;48:414-418. Weed JC, Arquembourg PC. Endometriosis: can it produce an autoimmune response resulting in infertility. Clin Obstet Gynecol 1980;23:885-893. Simon C, Gomez E, Mir A, De 10s Santos MJ, Pellicer A. Glucocorticoid treatment decreases sera embryotoxicity in endometriosis patients. Fertil Steril 1992;58:284-289.