Embed Size (px)
Citation preview
UMEÅ UNIVERSITY MEDICAL DISSERTATIONS
New series No. 854 ISSN 0346-6612 ISBN 91-7305-522-0
From the Department of Clinical Sciences, Obstetrics and Gynecology, Department of Radiation
Sciences, Oncology, and Department of Medical Biosciences, Pathology,
Umeå University, Sweden
Apoptosis, proliferation, and sex steroid receptors
in endometrium and endometrial carcinoma
Marju Dahmoun
Umeå 2003
Front cover picture: Apoptosis is seen in benign endometrium on the first day of menstruation. The
apoptotic bodies and cells are stained brownish with TUNEL method. Magnification x430.
Microphotograph by Stefan Cajander.
Copyright © 2003 by Marju Dahmoun
ISBN 91-7305-522-0
Printed in Sweden by
Kaltes Grafiska AB
Sundsvall 2003
To Johan
5
ABSTRACTApoptosis, proliferation, and sex steroid receptors in endometrium and endometrial carcinomaMarju Dahmoun
The cyclic changes in the female genital tract require remodeling of the endometrial tissuerelated to hormonal variations during the menstrual cycle. Apoptosis and proliferation functiontogether in many hormone-dependent organs and during embryogenesis, when rapid growth andregression are needed for tissue modulation. This thesis focuses on the involvement ofapoptosis and proliferation in the mechanisms of menstruation and hormonal replacementtherapy, HRT, as well as in the mechanisms of progesterone therapy in endometrial carcinoma.
Under the assumption that apoptosis is involved in menstruation, the aim of the first study wasto investigate endometrium for 4 days before and for 2 days during menstruation. Endometriumwas examined separately in endometrial glands and stroma during declines in levels of serum17ß-estradiol and progesterone. Different reactions were observed in epithelial and stromaltissues. In the epithelium, decreasing expression of estrogen receptor a (ER) and progesteronereceptor (PR), minimal proliferation, and rapid increase in the apoptotic index were observedprior to menstruation. In the stroma, an increase in the expression of ER and PR andproliferation was seen before the final decrease during menstruation. Stromal apoptosis wasclearly observed, but later than in the epithelium. Thus, apoptosis is involved in the remodelingof the endometrium during menstruation.
Apoptosis and proliferation, as well as high ER and PR expression, were also observed inpostmenopausal endometrium. During substitution therapy, which consisted of 2 differentregimens of HRT, the epithelial glands showed unaffected homeostasis with apoptotic indexand Ki-67 index as proliferation markers. ER expression was decreased both in the epitheliumand stroma, while PR showed different sensitivity, with some increase in receptor expression.The unchanged homeostasis during combined continuous HRT contributes to endometrialsafety, while an increase in proliferation was seen in stroma along with a maintained level ofapoptosis. This increase in proliferation has not been reported before and its importance shouldbe further evaluated. It could have some effect on breakthrough bleedings, as the stromalsupport may be important to the vascular stability in endometrium.
Unchanged apoptosis and increasing proliferation were observed with increasing tumor grade in29 patients with endometrioid endometrial carcinoma, which may contribute to greateraggression as tumor grade increases. The effects of medroxy-progesterone at 20 mg per daywere monitored after 14 days of therapy, and decreased proliferation was observed particularlyin the foci of maximal proliferation in G1 and G2 tumors, while G3 tumors were unaffected bythe progesterone therapy. The expression of ER was unchanged, while PR was decreased in thefoci of maximal expression for PR in G1 and G2 tumors. Since high proliferation and PRexpression also coexisted in the same foci, evaluated in G1 and G2 tumors, the effect ofprogesterone could be facilitated in these tumor groups. High expression of sex steroidreceptors was also a predicting factor for good response to progesterone (= decrease inproliferation), while the amount of stroma could not predict that effect.
6
CONTENTSABSTRACT.................................................................................................................... 5CONTENTS.................................................................................................................... 6ABBREVIATIONS ........................................................................................................ 8PAPERS.......................................................................................................................... 9GENERAL INTRODUCTION..................................................................................... 11BACKGROUND .......................................................................................................... 13
1. Apoptosis ............................................................................................................... 132. Proliferation ........................................................................................................... 14
2.1. Regulation of cell cycle by estrogen and progesterone.................................. 152.2. Monitoring proliferation and apoptosis.......................................................... 15
2.2.1. Proliferation ............................................................................................. 152.2.2. Apoptosis ................................................................................................. 15
2.3. Other regulators of apoptosis and proliferation.............................................. 162.3.1. Bcl-2......................................................................................................... 162.3.2. P53 ........................................................................................................... 17
3. Ovarian hormones.................................................................................................. 173.1. Estrogens ........................................................................................................ 173.2. Progesterone ................................................................................................... 19
4. Receptors ER and PR............................................................................................. 194.1. ER ................................................................................................................... 194.2. PR ................................................................................................................... 204.3. ER and PR in cyclic endometrium ................................................................. 204.4. ER and PR in postmenopausal endometrium................................................. 204.5. ER and PR in endometrial carcinoma ............................................................ 214.6. Heterogeneity of receptor expression............................................................. 21
5. The homeostasis of benign endometrium .............................................................. 215.1. Proliferation in endometrium ......................................................................... 225.2. Apoptosis in endometrium during the menstrual cycle.................................. 225.3. Paracrine mechanisms in the regulation of proliferation and apoptosis inendometrium during the menstrual cycle .............................................................. 225.4. Menstruation mechanisms .............................................................................. 235.5. Apoptosis and proliferation in postmenopausal endometrium: effects of HRT24
6. Endometrial carcinoma .......................................................................................... 256.1. Epidemiological aspects ................................................................................. 256.2. Classification and prognostic factors ............................................................. 266.3. Carcinogenesis................................................................................................ 266.4. Homeostasis in endometrial carcinoma.......................................................... 276.5. Growth factors and endometrial carcinoma ................................................... 276.6. Estrogen metabolism in endometrial carcinoma ............................................ 286.7. Progesterone therapy of endometrial carcinoma ............................................ 28
AIMS OF THE STUDY ............................................................................................... 29METHODS ................................................................................................................... 30
1. Ethical considerations ............................................................................................ 30
7
2. Subjects .................................................................................................................. 302.1. Paper I............................................................................................................. 302.2. Paper II ........................................................................................................... 302.3. Papers III and IV ............................................................................................ 31
3. Blood samples........................................................................................................ 31Paper I.................................................................................................................... 31
4. Tissue processing and immunohistochemistry ...................................................... 314.1. Processing for apoptosis ................................................................................. 314.2. Processing for ER, PR, Ki-67, bcl-2, and p53................................................ 324.3. Evaluation of steroid receptor immunoreactivity........................................... 324.4. Evaluation of the Ki-67 index ........................................................................ 334.5. Bcl-2 and p53 ................................................................................................. 334.6. Evaluation of apoptosis .................................................................................. 344.7. The amount of stroma in carcinoma............................................................... 34
5. Statistical methods ................................................................................................. 34RESULTS AND DISCUSSION................................................................................... 36
1. Results of the methodological evaluations ............................................................ 361.1. Apoptotic index (Ai), morphological and ISEL methods .............................. 361.2. ER and PR in endometrial carcinoma, 3 different methods........................... 36
2. Sex steroid hormones............................................................................................. 373. Tissue sensitivity to sex steroid hormones ............................................................ 38
3.1. ER and PR in benign endometrium................................................................ 383.1.1. ER and PR in cyclic endometrium........................................................... 383.1.2. ER and PR in postmenopausal endometrium .......................................... 38
3.2. ER and PR in endometrial carcinoma ............................................................ 393.2.1. ER and PR in tumors of different grade .................................................. 393.2.2. Comparison between benign and malignant endometrium ..................... 443.2.3. Random areas/specific areas.................................................................... 45
4. Homeostasis, indicated as proliferation and apoptosis .......................................... 474.1. Benign endometrium ...................................................................................... 47
4.1.1. Cyclic endometrium................................................................................. 474.1.2. Postmenopausal endometrium................................................................. 48
4.2. Endometrial carcinoma before, during, and after progesterone therapy........ 494.3. Predictive factors of progesterone therapy..................................................... 53
4.3.1. Epithelial factors ...................................................................................... 534.3.2. Stromal factors......................................................................................... 53
5. Bcl-2 and P53......................................................................................................... 555.1. Bcl-2 ............................................................................................................... 555.2. P53 .................................................................................................................. 55
SUMMARY.................................................................................................................. 58CONCLUSIONS........................................................................................................... 59ACKNOWLEDGEMENTS.......................................................................................... 60REFERENCES ............................................................................................................. 63PAPER I – IV................................................................................................................ 84
ABBREVIATIONS
17β-OH-HSD 17beta hydroxy steroid dehydrogenase
ABC avidin-biotin complex
Ai apoptotic index
APAAP alkaline phosphatase-antialkaline phosphatase
complex
CDK cyclin dependent kinase
E1 estriol
E2 17β-estradiol
EGF epidermal growth factor
ER estrogen receptor
ET endotelin
FIGO Federation Internationale de Gynecologie et
Obstetrique
HNPCC hereditary non polypoid colon carcinoma
hpf high power field
HRT hormone replacement therapy
IGF insulin-like growth factor
IGFBP insulin-like growth factor binding protein
MMP matrix metalloproteinase
PCO polycystic ovary (syndrome)
PG prostaglandin
SERM selektive estrogene receptor modulator
SHBG sex hormone binding globulin
TGF transforming growth factor
TNF tumor necrosis factor
TUNEL terminal uridine deoxynucleotidyl nick end labeling
VEGF vascular endothelial growth factor
WHO World Health Organization
8
9
PAPERSThis thesis is based on the following papers, which will be referred to in the text bytheir roman numerals:
IDahmoun M, Boman K, Cajander S, Westin P, and Bäckström T. Apoptosis,proliferation and sex steroid receptors in superficial parts of human endometrium atthe end of the secretory phase. J Clin Endocrinol Metab 84:1737-1743, 1999.
II Dahmoun M, Ödmark I-S, Risberg B, T. Pavlenko, and Bäckström T. Apoptosis,proliferation and sex steroid receptors in postmenopausal endometrium before andduring HRT.Manuscript.
IIIDahmoun M, Bäckström T, Boman K, and Cajander S. Apoptosis, proliferation andsex hormone receptors in untreated endometrial adenocarcinoma: results depending onmethods of analysis. Int J Oncol 22:115-122, 2003.
IVDahmoun M, Boman K, Cajander S, and Bäckström T. Intratumoral effects ofmedroxy-progesterone on apoptosis, proliferation and sex steroid receptors inendometrioid endometrial adenocarcinoma.Submitted for publication.
Reprints of papers where made with the kind permission of the publishers.
10
11
GENERAL INTRODUCTIONApoptosis and proliferation regulate homeostasis in benign as well as malignant tissue.Proliferation of human endometrium is stimulated mainly by estrogen via nuclearreceptors, a process that can be inhibited by progesterone [1]. In its receptor complex,17ß-estradiol participates in cell cycle stimulation. It also prepares the feedbackmechanism by stimulating PR synthesis. Progesterone stops the estrogen-stimulatedproliferation via nuclear effect, enzyme activation that catalyzes the bioactive 17ß-estradiol (E2) to inactive estrone (E1) [2-5]. Hormonal effects are further modulatedlocally by growth factors both in normal endometrium [6, 7] and in hormone-responsive endometrial carcinoma cells [8].
Apoptosis, programmed cell death, is an active physiological process used byorganisms for disposal of unnecessary or potentially harmful cells [9]. Apoptosis istriggered in many reproductive and hormone-dependent organs after ablation of thehormones [10-14]. In normal endometrium, autodigestion of basophilic granules hadbeen observed before apoptosis was described 1972 [15]. Apoptosis was later observedthrough the luteal phase, menstruation, and the early follicular phase [16-21], and thereare indications of low apoptotic activity in postmenopausal endometrium in somestudies [3, 18, 22, 23].
Changes in the balance between apoptosis and proliferation may be physiological,such as the changes during the menstrual cycle, when the phases of proliferation,specialization, shedding, and renewal all show specific balance between apoptosis andproliferation both in the epithelium and in the stroma. Unopposed estrogen therapymay lead to hyperplasia and carcinoma [24-26]. However, another pathway ofcarcinogenesis also exists, one not related to estrogen [27, 28]. Maintaining thebalance between apoptosis and proliferation in postmenopausal endometrium duringhormonal replacement therapy (HRT) is important for endometrial safety, and it mayalso be essential for good bleeding control [3]. The mechanisms of cytostatica andradiation are often induction of apoptosis in the tumor cells [9, 13, 29, 30], butprogesterone therapy of endometrial carcinoma has been used without knowing theexact mechanisms of action.
12
BACKGROUND
1. Apoptosis
Apoptosis is an energy-consuming active process that multicellular organisms havedeveloped to dispose of unnecessary or potentially harmful cells. The apoptotic stimulilead to a cascade of events that are shared by many cell types: the changes in themitochondrial membrane stop production of ATP and lead to the irreversible pathwayof apoptosis with activation of caspase reactions. Finally, the Ca2+/Mg2+ dependentendonuclease is activated and breaks the DNA into 180 base-pair nucleosomalfragments or to fragments in multiples of that size [9, 31-33].
Affected cells first show coarse aggregation of the chromatin that abuts inside thenuclear membrane. The cytoplasm shrinks, and cells diminish in volume. By that time,the apoptotic cells are seen in halos and the epithelial cells have lost the specificmicrovilli and desmosomal cell-to-cell junctions. The cell continues to shrink, thenucleus is condensed-later fragmented-and blunt protuberances of cytoplasm maydevelop and later separate from the rest of the cell body by enclosure of thefunctionally active cell membrane. Using that mechanism, the whole cell breaks upinto smaller particles called apoptotic bodies, which include relatively intact cellorganelles and parts of the nucleus. Apoptotic bodies are either engulfed byneighboring cells or extruded into the glandular lumen [9, 30, 31] (Fig 1).The lysosomal enzymes of the engulfing cells are involved in the digestion of theapoptotic bodies, but the apoptotic bodies dispersed into lumen or fluid may escapedigestion and undergo a spontaneous degeneration process such as necrosis [34].
13
Figure 1
Apoptosis and necrosis
The two pathways of cell death differ from
each other at most points: Apoptosis is
genetically programmed active process that
starts with DNA fragmentation and increasing
density (decreased volume) fragmentation of
the cell with active cell membrane action and
followed by phagocytosis by neighboring cells
or leukocytes without inflammation process.
Necrosis starts with membrane damage with
subsequent swelling of the cell and lysis of the
cellular organelles followed by inflammation
process.
14
2. Proliferation
Cyclic proliferation is characteristic of the female reproductive organs during fertileyears, and FSH and estradiol are the most important hormones stimulatingproliferation, especially in the ovary and endometrium [35]. The ability to measureproliferation and to know the factors regulating hormonal effects is of great value,since proliferation is associated with carcinogenesis of the endometrium [25, 28].High proliferation rate in endometrial carcinoma is also associated with aggressivebehavior of the tumor [36-39].
The events taking place in proliferating cells are best illustrated in the cell cycle(Fig. 3):� G1-phase represents the period from mitosis to S-phase. Most of the specialized
functions of the cells are carried out during this phase and G0, and the DNAcontent of benign diploid cells in this phase is equivalent with a double set ofchromosomes. The cells are under the influence of several growth factors and otherstimuli, also oncogenes, that may push the cells to enter the next proliferative phaseor to enter the resting phase G0. Before entering the S-phase there is an importantrestriction point, under control of a tumor suppressor gene, wild type p53, that isable to stop proliferation of genetically defective cells and lead them to G0 to berepaired, or to undergo apoptosis (see Section 2.2.2).
� S-phase is the period during which the DNA is duplicated.� In G2-phase the cells show duplicated DNA contents and prepare for mitosis.� Mitosis (M) features cell division, during which the DNA chromatin is divided and
condensed to chromosomes distributed equally to each daughter cell.
New daughter cell
Apoptosis
Synthesis (Doubling of DNA)
Restriction point (point of no return)
Figure 2
Cell cycle
The specialized action of the cells takes
place in G1-phase and the length of this
phase is shortened by estrogens, which
push the cell to the next phase and towards
mitosis, thus shortening the cell cycle. D-
type cinases and Cytocin E are the main
regulators of both estrogen and
progesterone action, but in a different way
as progesterone prolongs the G1-phase, at
least in breast tissue, thus promoting the
specialized function of the cells.
15
Measurements of the ploidy level can be used to describe the genetic contents of thecell population. The results can reflect the number of cells with aberrant chromosomalcontents and the proportions of cells at different points in the cell cycle [40-42].
2.1. Regulation of cell cycle by estrogen and progesterone
In animal studies, 17ß-estradiol reduces the cell-generation time by selectivelyshortening the time that cells stay in G1-phase and by promoting the G1/S transition ofuterine epithelial cells in vivo [43]. D-type cyclins D1 and D3 are cell-cycle-promoting factors induced by estrogen in G1-phase [44]. Cyclin E is also a G1/Sregulatory protein shown in endometrial endometrioid carcinoma in a tumor-grade-dependent manner but not in endometrial hyperplasia or in normal endometrium [45].Unfortunately, little is known about the role of progesterone directly in the cell cycleof endometrial cells.
In breast carcinoma cells, both estrogen and progesterone act mainly via D1 and c-mycas targets. Estrogen stimulates the formation of these proteins as well as the formationof highly specific activity forms of the cyclin E-Cdk2 enzyme complex lacking thecyclin dependent kinase CDK3 inhibitor p21. The delayed growth inhibition ofprogestins involves decreases in cyclin D1 and E gene expression and recruitment ofCDK inhibitors into cyclin D1-Cdk4 and cyclin E-Cdk2 complexes. Progestinspromote the cell differentiation in the prolonged G1-phase [46, 47].
2.2. Monitoring proliferation and apoptosis
2.2.1. Proliferation
Proliferation can be studied by morphological identification and counting of themitotic cells. Later methods have been developed to evaluate the number of cells inother active cell phases as in S-phase (flow cytometric methods) [48, 49], or in phasesG1-G2, i.e., all other active phases except G0-phase (immunohistochemical methodfor staining of Ki-67) [50]. Ki-67 is a nuclear protein (antigen) present only inproliferating cells. This protein can be demonstrated with anti-Ki-67 antibody (MIB-1)during the cell cycle except in G0-phase and the early part of G1-phase [50, 51].Ki-67 is present in endometrial glands during the proliferative phase and during thefirst half of the secretory phase but fades off during the second part of the secretoryphase [52, 53].
2.2.2. Apoptosis
Apoptosis is first detected and described according to morphologically characteristicsigns [9] that make it possible to detect apoptotic cells in routine staining, such ashematoxylin and eosin(H&E) staining in light microscope. In some tissues andconditions, such as inflammation with leukocyte infiltration, the detection of apoptoticcells is difficult. In situ hybridization techniques have been developed to assessidentification of the cells under programmed cell death [54, 55]. Generally, these
16
methods are based on labeling the free 3´hydroxyl ends for marking the fractions ofDNA. When terminal deoxynucleotidyl transferase (TdT) is used to incorporatebiotinylated deoxyuridine at sites of DNA breaks and the reaction then amplified withavidin-peroxidase for conventional histochemical identification by light microscope,the method is usually called the TUNEL-method (TdT-mediated dUTP-biotin nick endlabeling method) [54, 55]. The TUNEL-method has several variations [56, 57] thatenable identification of apoptotic cells even at the beginning of the apoptotic process,when the morphological signs are not observed. Using the combination of bothmorphological and staining criteria may give false low rates of apoptosis, but cellsshowing unspecific staining of non-apoptotic cells can be excluded [56].A flow cytometric TUNEL-method has also been developed and is able to givesemiquantitative images of apoptosis frequency in pure cell solutions [58].
Radiolabeled nucleotides can also be incorporated to the free 3´hydroxyl ends of the180-base-pair DNA fragments or their multimers. In agarose gel electrophoresis, atypical ladder pattern is seen [59, 60]. The thickness of the ladders gives only a coarseimage of the number of apoptotic cells among the cells studied. This method is suitedfor evaluation of pure cell cultures or cells easily separated from the tissues.
Quantitative image analyses to detect apoptosis in situ have been developed only inexperimental studies [61]. Time-consuming ocular methods are still needed forcomparison of apoptosis in different cell populations in the same organ or tumor aswell as for observations of staining heterogeneity [62, 63].
2.3. Other regulators of apoptosis and proliferation
2.3.1. Bcl-2
The proto-oncogene bcl-2 (B-cell lymphoma/leukemia 2) [64] is implicated incontrolling the cell cycle together with other members of the bcl family, such as bax,bcl-X-long (bcl-XL), and bcl-X-short (bcl-XS). Bcl-2 functions to prolong survival ofhealthy and pathological cells by blocking apoptosis [65] and is opposed by bax [14,66, 67]. The other pair of bcl-family-forming heterodimers is bcl-XL / bcl-XS. In thispair, bcl-XL promotes prolonged cell survival, while bcl-XS promotes apoptosis [14].Endometrial epithelium is immunoreactive for bcl-2 in the follicular phase, and itsstrictly cyclic appearance in studies using immunohistochemical methods argues thatbcl-2 is under hormonal control [52, 68, 69].
Bcl-2 may act as an oncogene [67, 70, 71], but the action of bcl-2 in humanendometrium and endometrial carcinoma is complicated because of the coaction withother members of the bcl family, such as bax, bcl-XL, and bcl-XS [52, 67, 71-74].
17
2.3.2. P53
The tumor suppressor gene wild type p53 is a powerful regulator of cell proliferationand apoptosis [75, 76]. It encodes a sequence-specific DNA-binding phosphoproteinthat is able to block stressed or DNA damaged cells in G1 (G0). The blocking processis even dependent on other gene expressions, such as WAF1 [77], and on growthfactors [78]. After successful repair the cell is allowed to enter S-phase and replicate;otherwise, the pathway of apoptosis is chosen (Fig. 2). Wild type p53 has the capacityto protect organisms against carcinogenesis by commanding the defective cells toapoptotic pathway (if not repaired), while mutated p53 is the single genetic mutationmost commonly observed in many different types of tumors [41, 76, 79-81].
Wild type p53 is a short-lived protein and has been difficult to show inimmunohistochemical staining, maybe also because of its great polymorphism [82].The mutated p53 has a longer half-life and has been widely studied in carcinomasusing immunohistochemical methods. (See Section 6.2. Carcinogenesis)
3. Ovarian hormones
Estrogens and progesterone, which are ovarian hormones, control normal cyclicendometrium [83]. Like other members of the steroid hormone family, such asandrogens, they elicit their genomic effects via nuclear receptors. Progesterone andandrogens are bound mainly to the sex-steroid-binding globulin SHBG, and only asmall free fraction of these hormones is responsible for their hormonal effects.Thus the hormonal effects can also be regulated by an excess or shortage of SHBG.The effects may be locally modulated by growth factors [84-87], but the rapidhormonal effects (via neurotransmitters) on the cellular functions of endometrium areless known [88, 89].
3.1. Estrogens
The ovarian steroidogenesis of 17β-estradiol (E2) takes place in follicular granulosacells and depends on follicle-stimulating hormone (FSH) [35]. Estrogen has a mitoticeffect on the endometrium, and it also upregulates both estrogen and progesteronereceptors in the follicular phase of the normal menstrual cycle and during hormonalreplacement therapy (HRT) after menopause [88, 90] [91-94]. Increasing levels of E2are seen in the follicular phase of the menstrual cycle, and high levels of E2 togetherwith increasing values of progesterone are observed during the luteal phase withmidluteal peak 1 week from the ovulation. Decreasing values of both hormones arenoted during the last 6 days of the luteal phase until the start of menstruation.
17β-estradiol (E2) is a biologically active estrogen while estrone (E1), an aromataseproduct, is a low-potency estrogen. The interconversion of E2 and low potency E1 by17β-hydroxysteroid dehydrogenase (17β-HSD) isoenzymes takes place in a tissue-specific way, dependent on which type of the isoenzyme is dominant in each tissue.
18
17β-HSD type 2 (17 β-HSD 2) catalyzes the oxidation of E2 to E1,and 17β-HSD type 1 (17β-HSD 1) catalyzes the reduction of E1 to E2 [95] (Figure 2).In normal endometrium, progesterone acts for cell differentiation and for production of17β-HSD 2, and thus prevents the proliferative effects of E2 [2, 3, 96-99] (Fig.3).17β-HSD 2 is also present in endometrial hyperplasia, even if the tissue showsproliferation and the proliferative normal endometrium lacks 17β-HSD 2.Less than one half of the cases with endometrioid endometrial carcinoma showpresence of 17β-HSD 2 [95, 100, 101], and these carcinomas may still have someprotection against unopposed-estrogen effects. In benign and malignant breast tissue,the 17β-HSD 1 is dominant, and this difference between endometrium and breast isessential when safety aspects of hormonal therapy are discussed [5, 96, 101].
In menopause, the production of estradiol in ovaries ceases, even if some follicles canstill be observed in peri- and postmenopausal ovaries, while the production oftestosterone from the stroma continues during some postmenopausal years [102, 103].Androstenedione of adrenal origin becomes the main source of estrogen products ofpostmenopausal women. Estriol is peripherally aromatized from androstenedione, andsome organs such as the breast can further convert estrone to estradiol in the presenceof 17β-HSD 1 as mentioned above [101].
Figure 3
Gonal steroid synthesis and metabolism
19
3.2. Progesterone
Serum progesterone levels are low in the follicular phase, but there is a rapid increaseafter ovulation, since progesterone synthesis occurs mainly in granulosa cells of corpusluteum during the luteal phase of the menstrual cycle. During pregnancy, the placentais the main source of progesterone synthesis. Progesterone stops estrogen-inducedproliferation [2], down-regulates both receptors [90], and enhances secretorydifferentiation of the epithelial cells as well as decidualization of the stroma.Progesterone is also connected with proliferation-called the second wave ofproliferation-in the stroma at the end of the luteal phase[2].
Natural progesterone cannot be administered orally, and for endometrial safetysynthetic progestagens are used in HRT regimens. Long term use of estrogen-onlyregimens, even using low-potency estrone, has been connected with clearly elevatedrisk of endometrial carcinoma [26, 92, 104-106]. Different doses of progesterone forcontinuous therapy and different lengths of progesterone therapy in sequential therapyhave been tested to find the lowest possible total dose of progesterone that is able toprevent endometrial hyperplasia and cancer. Most therapies used today are reasonablysafe for the endometrium when the end point of the studies has beenhistopathologically evaluated absence of endometrial hyperplasia and cancer [107-115].
4. Receptors ER and PR
Estrogen and progesterone receptors (ER and PR) are members of the steroid receptorfamily, which shares structural similarities with thyroid hormone receptors.Each receptor is loosely bound to the nuclear membrane and able to bind therespective hormone and transport it to the nucleus. This hormone-receptor complex isbound to the specific DNA sites and activates the polymerase transcription [116]. Thenuclear product, messenger RNA, is produced and transported to cytoplasm for furtherprotein production [35].
4.1. ER
Three types of ER (estrogen receptor alpha, ERα; estrogen receptor beta, ERβ; andestrogen receptor gamma, ERγ) exist in several isoforms, and the proportion of each asmediator of estrogen effects differs in organs and tissues [117, 118]. ERα, which for along time was assumed to be the only ER, dominates normal endometrium throughoutthe menstrual cycle [119] and it also dominates in endometrial carcinoma [120] andendometriotic tissue [121], even if ovarian endometrioma may show more ERßexpression [122] and ERß also exists in the endometrium. Further ERß has beendemonstrated in the oviduct, ovary, kidney, brain, and heart as well as male organssuch as the prostate and testis [117]. The most recently discovered estrogen receptor,ERγ [123-125], may have some prognostic value in breast cancer [118].
20
The variation of the estrogen receptor types in endometrium and brain as well as inskeletal and vascular systems has also been a possibility for novel therapies withselective estrogen receptor modulators (SERM), e.g., against osteoporosis [126].The most-used hormonal therapy for breast cancer, namely, tamoxifen, is functionallyalso a SERM, giving different effects for ERα and ERß: it is an agonist for ERα and anantagonist for ERß [127, 128]. Unfortunately, tamoxifen in long-term use has acarcinogenic effect on endometrium via ERα activation [129-132], and this risk haspromoted the use of aromatase inhibitors in the therapy of breast carcinoma.
Thus, in studies of many organs such as ovaries, the evaluation of both ERα and ERßis necessary, while ERα alone is able to illuminate the estrogen sensitivity of theendometrium.
4.2. PR
For PR, two isoforms, A and B, are known. Both homo- and heterodimers (AA, BB,and AB) are activated by the natural ligand progesterone. As well, the function of PRA and PR B differs between organs, e.g., uterus and breast: studies in knock-out micehave shown PR A as necessary for action of progesterone in the genital tract, includingthe uterus, while PR B is required for the normal proliferative effect of progesterone inthe breast [133].
4.3. ER and PR in cyclic endometrium
Estrogen receptor (ER) increases in the endometrial epithelium and stroma during thefollicular phase and decreases after ovulation to reach a low level under the late lutealphase [90, 134]. Progesterone receptor (PR) also increases during the follicular phaseand decreases in the epithelium in the luteal phase, but it stays at a higher level in thestroma until menstruation [90, 135]. ERα is dominant in human endometrium [136],even if ERβ may also modulate estrogen's action, especially in the epithelial cells.In the endometrium, both PR A and PR B are known [137] with PR A as thequantitatively dominant isoform.
4.4. ER and PR in postmenopausal endometrium
There are only a few studies of sex-steroid receptors in postmenopausal endometriumusing an in situ method [132, 138, 139]. One study with quantitative estimation ofreceptors found 92% expression of ER and 54% expression of PR [132]. Up-regulationof the receptors by estrogen is a rapid process, and in biochemical experimental studiesa 3-fold increase in receptor concentration of both ER and PR was observed within 1day and maximal increase in 3 days [140]. Progesterone in high doses is able to down-regulate the receptors in 1 to 2 days [141], and even shorter times have been observed[142].
21
4.5. ER and PR in endometrial carcinoma
A loss of sex steroid receptors is an early event in endometrial carcinogenesis, andendometrial carcinoma generally has a lower level of steroid receptors than doesnormal endometrium or endometrial hyperplasia [80, 143, 144]. There is also a greatvariation in receptor content, especially in PR content, in tumors of high or lowdifferentiation grade with low progesterone content in poorly differentiated tumors[145-148], and in tumors of the subtypes with worse prognoses [149, 150].Low PR content or absence of PR is an unfavorable prognostic factor [151], and somestudies also show prognostic significance of ER [152, 153]. High receptor content isalso associated with other positive prognostic factors as diploid DNA content and lowS-phase fraction (SPF) [154].
4.6. Heterogeneity of receptor expression
In benign endometrial tissue, the expression of sex-steroid hormone receptors maydiffer between epithelium and stroma, between surface epithelium and epithelialglands, and even between superficial glands and deep glands.Meanwhile, homogeneous expression of receptors is seen mostly inside the specifictissue in specific layers of endometrium.
The density of hormone receptors may differ inside the malignant tumor in variousways: between epithelial parts and the stroma [155, 156], between different parts ofthe tumor, and even inside the same fraction [80, 156, 157]. Primary tumor andmetastases may also have different hormone-receptor expression as the aggressivereceptor-negative subpopulations may give rise to metastases more often than therelatively benign receptor-positive subpopulations [158-160]. Very little is knownabout the importance of the heterogeneity of sex hormone receptors in endometrialcarcinoma, but theoretically hormonal therapy should give a different response inreceptor-dense parts compared with parts with low receptor expression.More broadly, very little is known about the implications of receptor heterogeneity incancer therapy in general.
5. The homeostasis of benign endometrium
The endometrium is under sex-steroid control, and its homeostasis is regulated byhormones during the menstrual cycle, which culminates in the implantation process.In a non-fertile cycle, the endometrium rapidly undergoes a remodeling process bymenstruation, proliferation, and specialization of the different endometrial cells, inorder to be ready for the next implantation. Every phase of endometrium has a specificbalance between proliferation on one side and apoptosis and necrosis on the other side[83, 161].
22
5.1. Proliferation in endometrium
Proliferation of the endometrial epithelium is maximal in the proliferative andfollicular phases and is stopped after ovulation. Studied with proliferation markerKi-67, 37% to 38% of the epithelial cells were in active-cell phase in the proliferationphase [53, 162], and the corresponding mitotic index was 2.3% [162].
Stromal proliferation follows mainly that of the epithelial glands during theproliferative phase [162-166]. However, there are exceptions from the rule: stromalcells may show proliferation during the luteal phase [2] as shown, e.g., in implantationprocess [87, 167-169]. The basal layer of endometrium shows lower but constantproliferation throughout the menstrual cycle, and the cyclic changes are most markedin the superficial parts of the endometrium [170].
5.2. Apoptosis in endometrium during the menstrual cycle
Apoptosis is rare during the proliferation phase of endometrium even if it has beenreported in some studies at the beginning of the phase [18, 171]. Increasing frequencyof apoptosis has been reported during the luteal phase and during menstruation [13,18-21, 23, 69, 162, 166, 172-174], and locally in the implantation site [167, 175-177].
Cyclic apoptosis in the endometrium provides evidence for hormonal regulation ofapoptosis in endometrium [19], and this has also been demonstrated in experimentalstudies [178-180]. Different study animals have been used, and results may differ forthat reason: in hamster epithelium, estrogen withdrawal induces apoptosis [179], whilerabbit endometrium is dependent on progesterone, and, consequently, withdrawal ofprogesterone induces apoptosis here [178, 180]. Further, the cycle-specific apoptoticactivity may provide evidence for its importance in inducing menstruation and inremodeling of the endometrium.
5.3. Paracrine mechanisms in the regulation of proliferation and
apoptosis in endometrium during the menstrual cycle
In a complex interaction, growth factors, cytokines, and enzymes, as well as receptors,modulate hormonal effects in endometrium during the menstrual cycle. Epidermalgrowth factor (EGF) and transforming growth factor alpha (TGF-α) are seen duringthe late follicular and secretory phases [6] and insulin-like growth factor 1 (IGF-1) inthe late secretory phase [7]. There is some evidence for the hypothesis that TGF-α isan important molecule in the pathway of estrogen-mediated cellular proliferation [27],and also, besides epidermal growth factor (EGF), important in regulation of stromadecidualization [181]. In a fertile cycle, rapid proliferation and apoptosis occur in theimplantation site when levels of both estrogen and progesterone are high [167].The paracrine signaling system is therefore important for the local regulation of tissueremodeling. As well, IGF-1 and IGF-2 have been connected with mitogenic effectsand differentiation of the epithelial cells [182], and together with the interplay with thebinding protein IGFBP-I they locally regulate the decidua during implantation process
23
and early pregnancy. Another growth factor associated with implantation istransforming growth factor beta (TGF-β) [177, 183]. Disturbances in the regulation ofimplantation may contribute to pregnancy loss and to pre eclampsia [182].
TGF-β has been shown to participate in apoptosis in experimental studies [184], andtumor necrosis factor (TNF-α) in the late luteal phase and during menstruation inhuman endometrium may have a similar role [19, 161, 185].
Changes in vascular endothelial growth factor (VEGF) and its receptor (KDR) as wellas TNF-α may activate matrix metalloproteinases (MMP) in the late secretory phase ofendometrium and may contribute to the menstruation process [87, 161, 186]. VEGF isan important factor in neovascularization and is present in the endometrium in theregeneration process during menstruation and in the early follicular phase [86, 168,186], in decidualization, and in the implantation process [84, 87, 167], and in ectopicimplantation of the endometrium [183], as well as in endometrial carcinoma [187-189]. The disturbances in the regulation of VEGF may contribute to intermenstrualbleedings [85]. Tissue hypoxy has been shown to be able to regulate changes inproteins coded by the VEGT family, proteins that together with angiopoietins regulategrowth and death of endothelial cells [168].
5.4. Menstruation mechanisms
The vascular changes associated with menstrual bleeding were observed by Markee inthe 1940s [190]. He used autotransplants of endometrium in the chamber of the eye ofrhesus monkeys and observed the spiral artery spasm and subsequent necrosis of thefunctional endometrium during menstruation. The existence of a pressor agent oragents was proposed to be responsible for vascular stasis and also for protectionagainst excessive blood loss. His theory came to be established as the dominant modelfor menstrual mechanisms. Later research connected prostaglandins (PGs) andendothelins (ETs) to the vascular changes in menstruation [191, 192]. Prostaglandin(PG) synthesis is stimulated in secretory endometrium by estrogen via cyclo-oxygenase [193], and prostaglandins may also have regulatory effects on angiopoietinand VEGF action. Markee also described earlier changes as thinning of theendometrium, dilatation of the arteries, and leukocytic infiltration. Only part of thefindings could be interpreted by Reynolds, who was critical of Markee's studies [194]using another type of monkey (New World monkey that lacks the spiral arteries andyet menstruates), and discussion focused on the amount of endometrium shed duringmenstruation. Bartelmez argued that only a very small amount of endometrium wasshed [15, 195], the discussion later continued by others [21, 112, 196-198].These authors contributed greatly to the literature through their studies withelectronmicroscopy, e.g., on the loss of cell-to-cell adhesion before the start ofmenstruation [197]. A detailed study by Christiaens also revealed the existence ofdefects in the vascular endothelium in superficial parts of the endometrium duringmenstrual spotting, hours before the menstruation starts [199]. The basophilic granuleswere described in endometrium and the lysosomal activity was shown.The autodigestion, together with the fact that not all primates have blood-yielding
24
menstruation, contributed to what was called the lysosomal concept of menstrualbleeding [200]. In any event, necrosis, vascular bleedings, and thrombosis in menstrualendometrium were also observed [112, 198, 201].
Hopwood and Lewison showed that the basophilic granules described by Bartelmez in1933 [15] were apoptotic bodies. These and apoptotic cells could be found in normal-cycling endometrium in the proliferative phase but mainly at the end of the secretoryphase in cd 24-28 [18]. Even in the ultrastructural study by Verma, apoptosis wasdated to the end of the menstrual cycle [21], and Otsuki names the same results in astudy that actually focuses on bcl-2 [69]. Tabibzadeh showed apoptosis in theglandular epithelium of human endometrium throughout the secretory phase,increasing toward the end of the phase [19], and Kokawa interpreted these findings butdescribed apoptosis even in the early proliferative phase [171].
During menstruation, a part of the functional layer of the endometrium is shed, and there-epithelization of the surface occurs from the stumps of the glands. [99, 161].
At the end of the luteal phase, infiltrating leukocytes may release many regulatoryproteins such as cytokines and proteinases. Molecules such as TNF-α, interleukin(IL)-1, relaxin, and TGF-ß are also locally produced in different cells of endometriumat the end of the luteal phase and during menstruation, and they may regulate MMPproduction [202, 203]. There is strong evidence that MMPs play a critical role in thetissue breakdown in menstruation as also described in Section 5.3. In any case, little isknown about activation of leukocytes and about their interaction with endometrialcells.
The modern understanding of the menstruation process includes both apoptosis,vascular constriction, and collapse with prostaglandin action as well as MMP actionwith an inflammation-like process [193]. Still, there are many questions about themenstrual mechanism and about the remodeling of the remaining endometrium duringmenstruation.
5.5. Apoptosis and proliferation in postmenopausal endometrium:
effects of HRT
There are few studies of postmenopausal endometrium, but some information can beobtained from the baseline studies of HRT. However, in most studies, onlyhistopathological evaluation (and not immunohistology) is used, and baselineendometrial biopsy is not included routinely in all studies. Johannisson et al. reportedatrophic endometrium in 76% to 90%, proliferative endometrium in 8% to 18% andoccasional patients (< 2.3%) with progestational endometrium in the study groupsrecruited [113]. The studies using Ki-67 as an indicator of proliferative status in postmenopausalendometrium are few. Morsi showed Ki-67 positivity in 18% of the cases [22];Mourits, in 2% [132]. An earlier morphological study indicates both low proliferationand apoptosis [18]
25
Histopathological evaluation during clinical studies of HRT is generally favorable tomodern HRT with regard to endometrial safety [107-115]. In most epidemiologicalstudies, therapies with monthly progesterone for > 10 days and daily progesterone incontinuous combined HRT regimens also show low risk for developing endometrialcarcinoma [204-206], even though there are contradictory results [207]. In clinicalstudies, histopathological evaluation is used, and the results in epidemiological studiesare based on the registered diagnosis. However, in some studies, proliferation duringHRT is investigated by proliferation markers (e.g., Ki-67) [208, 209], but thefrequency of apoptosis remains unknown.High proliferation in the endometrium may provide evidence of a risk that hyperplasiawill develop, but together with apoptosis it may simply be a sign of high cell turnoverin a tissue.
Optimal balance between the proliferative effects of estrogen and the antiproliferativeeffects of progesterone is also needed for good bleeding control during HRT. Irregularbleeding during continuous HRT regimens is one of the major problems with HRT[3, 112, 210, 211] and is the most usual reason for discontinuation of the therapy.Progesterone down-regulates both estrogen and progesterone receptors and causesatrophy in the endometrium. Apoptosis appear to occur in endometrial stroma afterprolonged continuous progesterone therapy and may contribute to breakthroughbleedings [3]. Decreased cell-to-cell adhesion as an effect of matrix metalloproteinases(MMP) could be involved in this mechanism of vascular and stromal breakdown,which is seen during progesterone therapy and prior to menstruation, as also discussedin Section 5.3. [212-217]. These changes occurring during the menstrual cycle orunder hormonal manipulation could be expected to somehow be affected by sexsteroids and mediated by receptors, even if other local regulators are of greatimportance. The exact mechanisms leading to vascular fragility and breakthroughbleedings remain unknown.
6. Endometrial carcinoma
6.1. Epidemiological aspects
Endometrial carcinoma is the third most common malignancy among women inSweden, [218] and approximately 1,100 new cases are diagnosed every year.An unopposed estrogen therapy may cause pathologic changes in the endometrium,and in some cases this may lead to malignancy. Other factors related to increased riskof endometrial cancer are obesity, nulliparity, early menarche, and late menopause.Most patients are postmenopausal, and only 5% of patients are younger than 50 yearsof age.
26
6.2. Classification and prognostic factors
The majority of endometrial cancers are diagnosed at an early stage, as the mainsymptom of the disease is postmenopausal bleeding, which usually leads the patient toseek medical advice. In addition, doctors tend to react quickly to potential cases ofendometrial cancer, as the guidelines of diagnostic procedures (ultrasonography and/orendometrial biopsy) of postmenopausal bleeding are clear.
The prognosis of endometrial carcinoma is generally good: overall 5-year survival isapproximately 80%. However, endometrial carcinoma is a heterogeneous entity.Endometrioid carcinoma has a generally better prognosis, while cancers ofseropapillary and clear cell types are associated with less favorable prognoses even inearly stages [219, 220]. Other prognostic factors are as follows: tumor grade, stadium,S-phase fraction, ploidy, sex-steroid-receptor content, age, and invasivity of blood andlymph vessels [37, 145, 147, 151, 221, 222]. In order to find better markers of tumoraggressiveness among early stage tumors, new genetic markers such as p53 [79, 223-225] and bcl-2 [74, 226], and markers reflecting tumor growth, such as Ki-67, [37, 39,224, 225, 227] have been tested (see Section 2).
6.3. Carcinogenesis
Many risk factors of endometrial carcinoma are associated with increased estrogenexposure, but endometrial cancer might also develop without endogenous orexogenous hormonal exposure [27]. Many tumors pass the endometrial hyperplasia[25] while others do not [28]. The heterogeneity of the molecular biology ofendometrial cancer makes it difficult to find a rational stepwise model forcarcinogenesis of the endometrium. At the risk of oversimplifying, endometrialcarcinoma is often divided in 2 types according to the carcinogenesis: Type I ischaracterized as estrogen-related with a carcinogenetic pathway including hyperplasiaand atypical hyperplasia [25], and type II as endometrial carcinoma independent ofestrogen [26, 93]. Type II is also called atrophy-associated carcinogenesis[228].Type I endometrial carcinoma is typically a well-differentiated carcinoma withglandular pattern, and it expresses estrogen and progesterone receptors. Patients areoften near menopause, younger than the patients with type II cancer, which isassociated with advanced stadium in the time of diagnosis and worse prognosis.Histopathologically, type II disease often exhibits more aggressive subtypes ofendometrial carcinoma with higher incidence of oncogenes.
The histopathological subtype has great prognostic significance and has also been usedto divide endometrial tumors into 2 groups: endometrioid adenocarcinoma as 1 groupand together the papillary serous, clear-cell, undifferentiated, and squamous-cellsubtypes as another group with less favorable prognosis [219, 229].
Additionally 1 type of endometrial cancer is associated with hereditary non-polypoidcolon carcinoma (HNPCC) [230].
27
The 2-step model of carcinogenesis is generally well accepted and requires a series ofgenetic events with both oncogenes and tumor suppressor genes, as described forHNPCC [231]. Several oncogenes have been associated with endometrialcarcinogenesis. The RAS family encoding p21 proteins, ERRB-2 (=HER-2 or neu) andC-MYC, are the oncogenes most studied besides the tumor-suppressor gene p53[28, 151, 162, 229, 232-238]. Satayaswaroop has earlier presented a model wheremalignant transformation can occur in any differentiation level of endometrial cells[239].
Wild type p53 protein is known as a tumor-suppressor gene that leads cells to arrestwith the possibility for DNA to be repaired [76]. Deranged p53 protein is the alterationmost frequently documented in human tumors and is often associated with poorprognosis and advanced stadium [41, 76, 79-81]. In endometrial carcinoma, mutationof p53 is observed as a late phenomenon in carcinogenesis [234, 235, 238, 240] and ithas been associated with more aggressive subtypes of carcinoma [229], beingrelatively rare in endometrial carcinoma of the endometrioid type [237, 241-243] andabsent in premalignant endometrial hyperplasia [244]. These results may indicate thatp53 mutation is a part of the pathway independent of estrogen action [27].
Microsatellite instability has been identified in sporadic endometrial cancer in 17% to23% of cases, but is more usual in endometrial cancer associated with HNPCCsyndrome [245, 246].
6.4. Homeostasis in endometrial carcinoma
Proliferation rate in endometrial carcinoma is correlated with tumor grade and withmore malignant histopathological types, but not to the stadium of the tumor [36, 39,53, 149, 224, 225, 227, 236, 247, 248].
There are only few in vivo studies of apoptosis in human endometrial carcinoma.In 2 studies there is some indication of correlation between tumor grade and apoptosis[166, 249]; others are indirect studies showing variation in the gene products of the bclfamily in tumors with different prognoses [74, 250-252].
6.5. Growth factors and endometrial carcinoma
Because several growth factors such as EGF, TGF-α, and IGF-1 have been connectedwith the pathway of receptor mediated estrogen action in normal endometrium, theymay also play a role in estrogen-dependent endometrial carcinoma [8, 27].
IGF-1 acts in regulation of several vascular endothelial growth factors (VEGF)[188, 189] that are associated with angiogenesis and metastasis of endometrialcarcinoma [187].
28
6.6. Estrogen metabolism in endometrial carcinoma
The progesterone-induced activation of the protecting enzyme 17β-HSD 2 is altered inendometrial carcinoma, compared with the activation in the normal luteal-phaseendometrium of premenopausal women [100]. In endometrioid adenocarcinoma,17β-HSD 2 enzyme has been shown in 37% of the tumors, and it has been shown moreoften in tumors of younger patients [95]. In any event, the oxidation of E2 to E1 is stilldominant in endometrial carcinoma compared with the opposite direction of theconversion (there is no 17β-HSD 1), at least in younger patients, and endometrialcarcinoma still has some capacity to defend itself against unopposed-estrogen effects.On the other hand, more than half of the adenocarcinomas have no 17β-HSD 2.Thus, the moderate progesterone effects even in receptor-positive endometrialcarcinoma may be understood as in part a result of the altered metabolism of estrogensin cancer tissue.A worse situation is seen in breast-tissue disorders such as hyperplasia and ductalcarcinoma [101], as the existence of 17β-HSD 1 in these tissues is able to stimulate thearomatization product E1 to be bioactivated to E2.
6.7. Progesterone therapy of endometrial carcinoma
Progesterone therapy has been used mostly in recurrent metastatic endometrialcarcinoma and as primary treatment when surgery and radiation have beencontraindicated [253]. Response rates of about 30% have been reported but varywidely according to the inclusion criteria and the tumor grade [4, 254-256].The empirical effect of progesterone on endometrial carcinoma may consist of thereceptor-mediated inhibition of estrogen-induced proliferation [257] and also of effectson growth factors [258].
Progesterone therapy has been used in treatment of metastatic endometrial carcinomaempirically since clinical studies have shown response to progesterone [253].Different response rates of 10% to 30% have been reported, with lower rates tendingto be found in later studies [4, 158, 254-256, 259]. Generally, higher response rates areobserved in patient groups with PR-positive tumors [260, 261], but the response ratesare not directly correlated with receptors, and variable response can be found in groupsof patients with both receptor-positive and -negative tumors. One reason for thevariable response rates could be the heterogeneous pattern of hormone-receptorexpression [155, 156, 159]. In experimental studies, preceding estrogen therapy hasbeen able to facilitate progesterone's effects [257], which may be mediated via thestromal cells of the tumor [5].
Even if hysterectomy with salpingo-oophorectomy is the first-choice therapy forendometrial carcinoma, progesterone therapy has been shown to be successful in somecases of young patients when an unopposed-estrogen etiology, such as PCO syndrome,has been suspected [262-264]. Although there is some conflicting evidence on adjuvant progesterone therapy [265],on balance the evidence shows that this approach has not been successful [266-268].
29
AIMS OF THE STUDYThis study focuses on the involvement of apoptosis andproliferation in tissue modulation, and on the importance ofhormonal sensitivity for these processes, in benignendometrium and in endometrial carcinoma under differenthormonal circumstances. The specific aims were as follows:
� To investigate endometrial hormone sensitivity (ERand PR), proliferation (Ki-67), and apoptotic index(Ai), as well as an antiapoptotic factor (bcl-2),during major hormonal withdrawal before andduring menstruation, under the hypothesis thatapoptosis is involved in the mechanisms ofmenstruation.
� To elucidate ER, PR, Ki-67, and Ai separately in thestroma and epithelial endometrium before and duringsubstitution with continuous combined HRT, underthe hypothesis that the ratio of proliferation toapoptosis is not increased during HRT.
� To evaluate hormonal sensitivity, apoptosis, andproliferation as well as bcl-2 and the incidence oftumor suppressor gene p53 in endometrioidendometrial carcinoma before, during, and afterhormonal manipulation with progesterone, under thehypothesis that progesterone withdrawal can induceapoptosis.
30
METHODS
1. Ethical considerations
The Ethics Committee of Umeå University approved the 3 studies in papers I, III, andIV included in this work. The Ethics Committee of each center involved in themulticenter study, represented partly in paper II, approved the study, and the EthicsCommittee of Umeå University further approved this specific study. Informed consentwas obtained from all women.
2. Subjects
2.1. Paper I
Endometrial micro-biopsies were taken with a Pipelle® or Endorette® instrument from35 regularly menstruating healthy women who were not receiving any hormonaltherapy during 37 menstrual cycles from 4 days prior to the onset of menstruation(Day -4) until the second menstrual day (Day 1). The biopsies may represent any partof the superficial corpus endometrium. Altogether 75 biopsies were taken, representing6 consecutive days and the number of biopsies varied from 10 to 15 each day (Table 1,paper I). One biopsy per day and 1 to 3 biopsies per cycle were taken from individualpatients. In 2 cases, only a single biopsy was taken; in 29 cases, paired biopsies weretaken; and in 5 cases, 3 biopsies were taken. The length of the menstrual cycle variedindividually but the data in this study were centered on the onset of bleeding.
2.2. Paper II
The patients were recruited in a prospective multicenter study carried out in 14 centersin Sweden [210]. Out of 92 women who had not used HRT during the past 2 months,43 women had biopsy material allowing histological evaluation in both biopsies, i.e.,the biopsy obtained before HRT and the biopsy during HRT (after 1 year of HRT).The therapy consisted of either conjugated estrogen (CE) 0.625 mg + 5 mg medroxy-progesterone acetate (MPA) (= CE/MPA) or 17β-estradiol (E2) 2 mg + 1 mgnorethisterone acetate (NETA) (= E2/NETA). The women included in the study wererequired to be in good health, with an intact uterus, = 52 years and = 2 yearspostmenopausal. The exclusion criteria were: adenomatous hyperplasia with orwithout atypia, undiagnosed vaginal bleeding, history of cancer, cardiovascular orthromboembolic disease, uncontrolled hypertension, diabetes, and long-termmedication with barbiturates, psychotropics, or antiepileptic drugs. No use of steroidhormones besides the study medication was allowed during the study period.
31
2.3. Papers III and IV
This study included a homogenous group of 29 patients with endometrioid endometrialadenocarcinoma: 4 patients with stage IA, 20 with stage IB, and 3 with stage ICaccording to the FIGO criteria for surgical staging [269]. In 2 cases, no surgery wasperformed due to contraindications. Three samples were obtained from each patient:biopsy 1 at the diagnostic endometrial biopsy with Pipelle® (Prodimed, Neuilly-en-Thelie, France) instrument or traditional dilatation and curettage (D&C), biopsy 2 after14 days of treatment with 20 mg medroxy-progesterone acetate per day, and in 20cases biopsy 3 after 2 days of withdrawal of progesterone treatment. In 9 cases, biopsy3 consisted of material obtained from the hysterectomy specimen after 6 days ofwithdrawal of progesterone treatment.
3. Blood samples
Paper I
Blood samples were collected immediately before or after each endometrial biopsy.Samples were centrifuged, and serum was split into small portions that were frozenand stored at-20(C until analysed in the same run by a TRFIA method (time-resolvedfluoro-immunoassay, Wallac).Duplicate analyses were performed, and the resulting mean value was used. Allsamples were assayed in the same run. Intra-assay coefficient of variation (CV) was2.0% for serum progesterone (S-P) and 7.5% for serum estradiol (S-E) analysis.Blood samples of the women in the cancer study (papers III and IV) were collected atthe time of biopsy 2.
4. Tissue processing and immunohistochemistry
4.1. Processing for apoptosis
In paper I, the biopsy material from corpus endometrium was fixed in 4%formaldehyde for 4 to 6 hours and embedded in paraffin according to routineprocedures. The same fixation time was applied also for biopsies 2 and 3 in the cancermaterial (papers III and IV) to enable the use of the in situ end labeling (ISEL)technique used and described in paper I. This technique was used in cancer materialonly for control and not reported because the first biopsy (diagnostic) was often fixedfor longer times and the ISEL method was therefore not used. The fixation time variedfor the biopsies in the multicenter study (paper II), and another TUNEL method(ApopTag®, Intergen Company, Oxford, UK), was applied for identification ofapoptotic cells. For immunohistochemistry, the fixation time was not critical.
32
4.2. Processing for ER, PR, Ki-67, bcl-2, and p53
Immunostaining for materials used in papers I, III, and IV was carried out in theDepartment of Pathology in Umeå University Hospital and for paper II in theDepartment of Pathology in Örebro Regional Hospital as separately described in therespective papers. Monoclonal antibodies directed against ER, PR, Ki-67, bcl-2, andp53 were used. Localization of antigen-antibody complexes was performed with theavidin-biotin-peroxidase complex (ABC) technique, except localization of bcl-2, forwhich the APAAP method was applied. Brown nuclear color was seen in positivestaining of cells, except in bcl-2 staining that indicates the cytoplasmic staining withpurple color.
4.3. Evaluation of steroid receptor immunoreactivity
In normal endometrium (paper I), a semiquantitative scoring method was used forevaluation of ER and PR expression: ER and PR staining in the surface and glandularepithelium as well as in the stroma were evaluated in each section. Positive cellsshowed a brown reaction product in the nucleus, while the cytoplasm remainedunstained. Staining was scored in 4 categories as follows: 0 = cells were negative;1 = most cells were weakly stained or occasional cells were strongly stained;2 = most cells were moderately stained; and 3 = most cells were strongly stained. In postmenopausal endometrium (paper II), the stained cells and the total amount ofcells inside the grid were again counted in minimum 10 high power fields (hpf), and intotal at least 1,000 cells were counted. The whole section was evaluated if there were< 1,000 cells in the specimen, because many postmenopausal biopsy materials werescanty.
A methodological study of ER and PR evaluation was made in cancer material beforeprogesterone treatment (paper III). Estrogen receptor and PR staining were evaluatedwith a standard light microscope using 3 different methods:
(1) The counting method: A grid was mounted in 1 eyepiece of the microscope and100 cells/field were counted in at least 10 representative fields (overall counting) ofthe epithelial fraction of the tumor. Positive cells stained brown in the nucleus, whilethe cytoplasm remained unstained. If the sections showed heterogeneous staining inareas of at least half a high-power field (hpf), the percentage of stained cells wascounted in a similar way in at least 5 fields of maximal staining and in at least 5 fieldsof minimal staining. The percentage of cells that were stained is referred to as thestaining index. Hereafter, the staining indexes for ER or PR in overall staining areabbreviated as ER and PR, respectively, the staining index in the areas of maximalstaining as ER-max and PR-max, and the staining index in the areas of minimalstaining as ER-min and PR-min.
33
(2) The mixed method: The results of the counting method were used and the intensityof nuclear staining for ER and PR was ranked separately from 0 to 3 in overallcounting, in the areas of maximal staining and in the areas of minimal staining.The score was given as follows: 0 was given for absent staining; 1, for weak staining;2, for moderate staining; and 3, for strong staining. The results were given as an index,with the percentage of stained cells multiplied by the staining intensity score (0-3).
(3) The scoring method: The whole section was evaluated according to an integratedanalysis of the visual appearance of intensity and amount of stained cells. Staining wasranked on a 4-point scale, as follows: 0 was allocated if there were no stained cells;1, if most of the cells were weakly stained or if some cells were strongly stained;2, if most cells were moderately stained; and 3, if most cells were strongly stained.
There were good correlations between the results with all 3 methods (paper III).The counting method was used in evaluation of receptor expressions in biopsy 2(during the progesterone therapy) and in biopsy 3 (after withdrawal) and in evaluationof the changes between biopsies 1 and 3 (paper IV).
The heterogeneity (i.e., presence of a difference between the staining indexes in theareas of maximal and minimal staining) in the staining for ER and PR was assessed ineach tumor, further referred to as ER-het and PR-het. In addition, a visual estimation-based scoring of heterogeneity was performed in paper III, in which the score of 1meant no difference; 2 meant there was a difference of 1% to 25%; 3 meant adifference of 26% to 75%; and 4 meant a difference of 76% to 100%.
4.4. Evaluation of the Ki-67 index
The Ki-67 index was evaluated as described for ER and PR in the counting method(Section 4.3) (cells stained/100 cells), and this quantitative method was used toestablish the Ki-67 index in all papers. In paper I, 2 different methods were applied inmicroscopic evaluation of the Ki-67 index, and the results were in good accordancewith each other (P = 0.001).Because the cancer tissue showed marked heterogeneity of proliferation, the countingwas interpreted separately in the areas of maximal and minimal staining for Ki-67expression, and the heterogeneity was counted out as described for the receptors(Section 4.3) and referred to as Ki-het.
4.5. Bcl-2 and p53
The cytoplasm of bcl-2 positive cells stained purple, and in a rapid visual evaluationthe intensity of bcl-2 staining was scored on a 4-point scale as absent (0), weak (1),moderate (2) or strong (3). Infiltrating granulocytes were often seen in premenstrualand menstrual endometrium as well as in cancer tissue and showed strong staining,which served as an internal positive control. The percentage of p53 positive cells in theepithelial fraction of the endometrial cancer was counted in all biopsies obtained(papers III and IV).
34
4.6. Evaluation of apoptosis
With TUNEL methods the nuclear materials of apoptotic cells stain brownish, andcounterstaining with a blue nuclear marker makes it easier to identify apoptotic cellsamong normal cells. In normal benign material (papers I and II), the TUNEL methodwas necessary, because the identification of apoptotic cells among stromal cells withmorphological criteria alone had been difficult. The stroma was often infiltrated withgranulocytes, and it could be troublesome to distinguish between apoptotic cells andgranulocytes. Identification of apoptosis in epithelial cells is easier, and the evaluationof apoptosis with only morphological criteria in H&E staining correlated well withevaluation of the specimen stained with TUNEL method, in the epithelial glands ofbenign endometrium (Spearman's correlation analysis, r = 97; P < 0.001) (paper I).
A grid was used in 1 eyepiece of a standard light microscope. Cells inside the grid areain 20 to 40 random fields in tissue sections were evaluated, and the apoptotic index(Ai) (i.e., number of apoptotic cells per 1,000 cells) was determined after counting2,000 to 10,000 cells in total (in the postmenopausal material, a minimum of 1,000cells was required). The morphological criteria used for apoptotic cells were asfollows: single rounded cells or fragments of cells with densely aggregated chromatinand condensed cytoplasm, often lying in a halo of extracellular space [9].
In carcinoma (papers III and IV), only the epithelial part of the tumor was evaluatedfor apoptosis, and only morphological criteria were used in counting the Ai intraditional hematoxylin and eosin staining. [9].
4.7. The amount of stroma in carcinoma
Evaluation of the amount of stromal tissue was made in sections of biopsy 1 (paper III)stained for ER or PR. Both the stromal and epithelial areas covering the area of thegrid mounted in1 eyepiece of the microscope were counted in 10 non-fragmented partsof the specimen according to stereological principles [62]. The areas weresummarized, and the percentage of the stromal area out of the total area of the tumor inthese 10 sections was used in statistical analysis. The percentage of ER- and PR-positive cells in the stroma was also counted in these 10 sections.
5. Statistical methods
Standard non-parametric methods were used to test for significant differences betweenthe 2 groups: Wilcoxon and Mann-Whitney tests were applied for paired andindependent observations, respectively. Multiple group comparisons of means wereperformed using1-way and 2-way analysis of variance (ANOVA). Fisher's leastsignificant difference (LSD) procedure was applied as a post-hoc test for conductingpairwise comparisons. In order to evaluate sex-steroid receptor expressions in differentfoci of high and low proliferation in the same tumor, a repeated-measures design wasproposed. Repeated measures were then analyzed using a general linear model (GLM)procedure. To study the relationship between feature variables, to show the agreement
35
between different methods of evaluation of receptors, and to compare differentmethods for detection of apoptosis, correlation analysis was applied. Spearman's andPearson's correlation coefficients were used, and regression lines were fitted for theobservations. Regression analysis was also used to identify possible predictors of theeffect of hormonal therapy in endometrial carcinoma. In all statistical tests used,P < 0.05 was considered to be statistically significant.
Plots and calculations were performed using the Statistical Package of Social Science(SPSS), version 10.1 (Scandinavia AB, Stockholm, Sweden), and by MATLAB,version 6.1, (The Math Works, Inc. Natick, Massachusets, USA).
36
RESULTS AND DISCUSSION
1. Results of the methodological evaluations
1.1. Apoptotic index (Ai), morphological and ISEL methods
Comparisons were made between the apoptotic indexes(Ai) in H&E- and ISEL-stainedsections (epithelium) in 23 sections. In the ISEL method, both morphological andstaining criteria were used, and in the H&E-stained sections, the evaluation was basedon the morphological criteria only [9]. There was a strong correlation (r = 0.97;P < 0.001) in apoptotic indexes between the ISEL and H&E methods.
1.2. ER and PR in endometrial carcinoma, 3 different methods
Estrogen receptor (ER) and PR staining was evaluated with 3 different methods:in method 1 the average percentage of stained cells was calculated, in method 2 thestaining intensity was added to the staining percentage and method 3 consisted of asoring including both frequency and intensity of staining All 3 methods correlated wellwith each other (Table 1).
The counting method (method 1) and the mixed method (method 2) were both used inseparate evaluation of the ER staining index in the areas of maximal staining density(ER-max) and in the areas of minimal staining density (ER-min), and the resultsshowed good correlation (r = 0.76 and r = 0.91, respectively; P < 0.001). The scoringmethod (method 3) was not used for separate evaluation of the areas of maximal andminimal staining, because the areas might have been too small for correct visualanalysis.
In evaluation of PR staining, there was high correlation between methods 1 and 2,methods 1 and 3, and methods 2 and 3 (Table 1). Results of methods 1 and 2 alsocorrelated well (r = 0.86 and r = 0.93, respectively; P < 0.001) when PR-max and PR-min were evaluated.
The 3 different methods of evaluating ER and PR status can be used with equal effectin the evaluation of hormone-receptor expression. The visual scoring method is quick,but the more time-consuming counting of the percentage of stained cells is necessary ifthe heterogeneity of receptor status is studied. Considering the staining intensity scorealongside the percentage of stained cells did not provide any further information.The counting method is used in papers II - IV, as it permits evaluation of heterogeneitywithout subjective scores.
Table 1
Different methods of evaluating sex steroid receptor expression in the epithelial part ofendometrial carcinoma. There was a high correlation of the results established with the3 different methods (Pearson’s correlation coefficient, r, and P-values shown).
2. S
Bothmate(ca. 4minimdecrefertilto shethe p
WompostmE2 an
Table
Level
(pape
Receptor Methods tested r P
Scoring /Counting 0.75 P < 0.001
Scoring/Mixed 0.78 P < 0.001
ER
Counting/Mixed 0.88 P < 0.001
Scoring /Counting 0.85 P < 0.001
Scoring/Mixed 0.85 P < 0.001
PR
Counting/Mixed 0.91 P < 0.001
ex steroid hormones
serum 17ß-estradiol (E2) and serum progesterone (paper I) were monitored in therial of 35 healthy menstruating women from the normal luteal-phase values00 pmol/l E2, and ca. 30 nmol/l progesterone) 4 days prior to menstruation, toal values during the 2 menstrual days (P < 0.001) (Fig. 3; paper I). The rapid
ase in availability of estrogen and progesterone as a sign of an unsuccessfule cycle is the inducing factor for a series of changes in the endometrium, leadingdding of a superficial layer and part of a functional layer of the endometrium in
rocess of menstruation.
en in the cancer study (papers III and IV) were postmenopausal (range 1-44enopausal years), except 1 woman, who was perimenopausal. Consequently, thed progesterone levels in serum were low, as shown in table 2.
2
s of sex steroid hormones in 28 out of 29 patients with endometrial carcinoma
rs III and IV).
Steroid hormone Mean (SEM) Min Max
E2 89.0 (7.6) pmol/l 72 pmol/l 265 pmol/l
Progesterone 1.6 (0.19) nmol/l 0.6 nmol/l 5.5 nmol/l
37
38
3. Tissue sensitivity to sex steroid hormones
ER and PR expression, as indicators of tissue sensitivity to sex steroid hormones, wereevaluated at the end of the menstrual cycle (papers I) and in postmenopausalendometrium before and during combined continuous HRT (paper II). In malignantendometrial tumors, receptor expression was analyzed before, during, and aftermedroxy-progesterone therapy (papers III and IV).
3.1. ER and PR in benign endometrium
3.1.1. ER and PR in cyclic endometrium
The ER immunoreactive score decreased in the glandular epithelium of menstruatingwomen during the period from 4 days prior to menstrual start to 2 days prior tomenstrual start, at which time it began to increase again. The ER score in the stromashowed just the opposite pattern, with increasing staining during the first 2 study daysand decreasing staining from the day before menstrual start to the second menstrualday (Figure 3, paper I). Throughout the study period, the surface epithelium showedlittle staining of ER and PR, without any obvious pattern of change.The progesterone receptor score in glands decreased from 1.5, measured 4 days priorto menstruation, to 1.0, measured 1 day later, and it stayed at a low level until thesecond menstruation day. In the stroma, PR scores were on a higher level 4 daysbefore menstrual start and still increased during the first day of the study period(Figure 2, paper I), but then decreased until the second day of menstruation (Figure 3,paper I).An apparent difference was seen in the expression of sex-steroid receptors in epithelialglands and in the stroma: In the epithelium a rapid decrease was observed in receptorexpression, while the stroma was still sensitive at least to the effect of progesterone.Earlier decrease of PR in the epithelium compared with endometrial stroma has beenobserved in a previous study, and our results concur with those results [135].
3.1.2. ER and PR in postmenopausal endometrium
A high and relatively homogenous pattern of sex-steroid receptors was observed inpostmenopausal endometrium, with mean ER expression 96 (cells/100 showingstaining for ER) and mean PR expression 83 before HRT. The levels of ER and PRexpressions in our study were on the same level as reported earlier by 2 other groups[132, 139]. Both ER and PR expression were higher in the epithelium than in thestroma (Wilcoxon signed rank test; P < 0.001), and that difference remained in biopsyobtained during continuous combined HRT with either conjugated estrogen (CE)0.625 mg + 5 mg medroxy-progesterone acetate (MPA) (= CE/MPA) or 17β-estradiol(E2) 2 mg + 1 mg norethisterone acetate (NETA) (= E2/NETA) (Table 3, paper II).The expression of PR was not affected by HRT, while ER expression was decreased,and this decrease was seen in the group with HRT regime E2/NETA in the glandular
39
epithelium but not in the CE/MPA group. A decrease in ER expression was observedalso in stroma (Tables 1 and 2, paper II).
High expression of sex steroid receptors was observed both before HRT, when serumlevels of E2 and progesterone were probably low, and during combined continuousHRT, when serum levels of E2 and gestagen were higher and stable. Differentreactions of ER and PR were observed during HRT, with ER decreasing significantlywhile PR showed a non-significant tendency to increase (Tables 1 and 2, paper II).The decrease of ER expression could be an effect of gestagen and especially an effectof NETA, since the decrease was seen only in the group treated with the E2/NETAregimen, and androgens are known to be powerful down-regulators of ER [270].Progesterone receptors may be more sensitive to the effect of estrogen when estrogenand progesterone are combined, and no difference was seen between the 2 regimens.
3.2. ER and PR in endometrial carcinoma
3.2.1. ER and PR in tumors of different grade
Great heterogeneity of both ER and PR expression was observed in the epithelialfraction of the tumors in all 3 biopsies obtained during the study period: in biopsy 1before progesterone therapy, in biopsy 2 during the therapy, and in biopsy 3, whichwas obtained 2 or 6 days after progesterone withdrawal. For that reason, sex steroidexpressions were separately counted in the foci of maximal and minimal expressionsof receptor staining, and the overall expression (mean value from at least 10 randomareas) was evaluated as well (see Methods). Results from the evaluation of theuntreated endometrial carcinoma are shown in Tables 1 and 2 in paper III and theresults of the composites of the mean values in all 3 biopsies are shown in Table 2.The results clearly show that ER and PR expression is higher in Grade 1 (G1) andGrade 2 (G2) tumors compared with Grade 3 (G3) tumors (Tables 1 and 2, paper III).Progesterone receptor heterogeneity varies with tumor grade, with greaterheterogeneity in G1 and G2 tumors compared with G3 tumors (Table 2, paper III),while the heterogeneity of ER expression was in the same range between the tumors ofdifferent grade in biopsy 1 (Table 3, paper III) and throughout the study period.
Table 3
The composites of the mean values of ER, PR, and Ki-67 expression in endometrial
carcinoma of different grades. Statistics by ANOVA (ad hoc LSD).
b Grade N PR PR-max PR-min PR-het
Mean (SE) Mean (SE) Mean (SE) Mean (SE)
1 13 41.90 (4.4) 75.00 (4.1) 12.82 (2.9) 62.18 (5.2)
2 10 40.30 (6.7) 67.77 (6.3) 19.33 (5.5) 48.43 (5.9)
3 6 21.50 (9.9) 40.44 (13.9) 10.22 (5.6) 30.22 (7.7)
Sign. n.s. P <0.05 n.s. P <0.001
c Grade N Ki Ki-max Ki-min Ki-het
Mean (SE) Mean (SE) Mean (SE) Mean (SE)
1 13 15.18 (2.1) 37.05 (4.2) 4.67 (1.3) 32.39 (3.7)
2 10 28.10 (3.4) 56.63 (4.6) 7.40 (1.8) 49.23 (4.3)
3 6 43.83 (4.1) 71.11 (3.0) 19.28 (4.2) 51.83 (5.5)
Sign. P <0.001 P <0.001 P <0.001 P <0.01
a Grade N ER ER-max ER-min ER-het
Mean (SE) Mean (SE) Mean (SE) Mean (SE)
1 13 76.38 (3.0) 92.15 (2.8) 49.51 (4.1) 41.34 (5.7)
2 10 68.27 (6.6) 88.00 4.0) 40.23 (6.8) 36.35 (5.2)
3 6 40.61 (14.2) 59.89 (11.6) 33.33 (14.4) 39.27 (6.6)
Sign. P<0.01 P <0.01 n.s. n.s.
40
41
The heterogeneity of ER expression stayed in the same range, while the heterogeneityof PR expression decreased during the progesterone treatment from biopsy 1 to biopsy2, and was obviously an effect of the marked decrease in PR expression in the foci ofmaximal expression for PR (PR-max) during the treatment.
There was a significant difference in ER expression of tumors of different grade.Higher mean values of ER were observed in G1 and G2 tumors compared with G3tumors, which showed significantly lower ER density throughout the study period(Table 3 and Fig. 4). No change was observed during the progesterone therapy(Fig. 4). These results can be compared with the results from normal postmenopausalendometrium, where combined therapy with 2 different combined continuous HRTregimens resulted in a decrease in ER expression. Even in endometrial carcinoma,better response to progesterone therapy has been shown in cells or tissues pretreatedwith estrogen [257].
The mean values of PR in overall evaluation and in the foci of minimal staining for PR(PR-min) did not change during the progesterone therapy, but a decrease of PR-maxwas observed during the therapy (Fig. 5). The mean value of PR staining was notsignificantly different in tumors of different grade, but PR-max in G3 tumors waslower compared with G1 and G2 tumors before the therapy (Table 2, paper III).The decrease in the PR-max observed during the therapy was separately evaluated intumors of different grade, and it turned out that the decrease was found for G1 and G2tumors but not for G3 tumors (Table 3, Fig. 5).
42
A
Biopsy 1 - 3
321
Mea
n ER
(ce
lls/1
00)
90
80
70
60
50
4030
B
Biopsy 1 - 3
321
Mea
n ER
(ce
lls/1
00)
90
80
70
60
50
4030
GRADE
1
2
3
C
Biopsy 1 - 3
32 1
Mea
n ER
, max
(cel
ls/1
00)
100
90
80
70
60 50
D
Biopsy 1 - 3
321
Mea
n ER
, max
(ce
lls/1
00)
100
90
80
70
6050
GRADE
1
2
3
Figure 4 ER and ER-max expression in 29 endometrial carcinomas before (biopsy 1), during (biopsy
2), and after (biopsy 3) medroxy-progesterone treatment.
A and B. Expression of ER was unchanged in the whole material as well as in tumors of
different grade separately.
C and D. ER-max stayed also in the same range and there was no interaction with tumor
grade.
In B and D the differences of the mean values between G3 tumors and G1–2 tumors are
clearly seen.
43
A
Biopsy 1 - 3
321
Mea
n PR
(ce
lls/1
00)
50
40
30
20
10
B
Biopsy 1 - 3
321
Mea
n PR
(ce
lls/1
00)
50
40
30
20
10
GRADE
1
2
3
C
Biopsy 1 - 3
321
Mea
n PR
, max
(cel
ls/1
00)
90807060504030
D
Biopsy 1 - 3
321
Mea
n PR
, max
(ce
lls/1
00)
90807060504030
GRADE
1
2
3
Figure 5 A and B. The decrease in PR expression during progesterone therapy was not significant.
B. The interaction of grade was seen in PR as G1 and G2 tumors show changes in a different
way, while PR in G3 tumors was unchanged.
C. PR-max was markedly decreased during the progesterone therapy and, although slightly
increased (not significant) after withdrawal, was still on a lower level compared with biopsy1.
D. The interaction of tumor grade and PR-max is on a borderline level (F2,52 =2.51; P =0.053)
but (in separate evaluation) different reaction patterns are seen in progesterone therapy in
tumors of different grades.
3.2.2. Comparison between benign and malignant endometrium
Relatively high expression of both ER and PR was observed in endometrial carcinoma,and, as shown above, the density of sex-steroid receptors is highly dependent on tumorgrade. It has previously been pointed out that ER expression is high in endometrialcarcinoma [271], even if loss of receptors is a part of carcinogenesis [80, 236, 239,271], and ER and PR are generally decreased in cancer compared with endometrialhyperplasia.
Since we had evaluated ER and PR expression in benign postmenopausal material andin endometrioid endometrial carcinoma grade by grade using the same quantitativemethod, we could compare these materials (Table 4).
Table 4ER and PR expression, proliferation indicated as Ki-67 index and apoptotic index, Ai, in
benign postmenopausal endometrial epithelium and in endometrioid endometrial carcinoma
of different grades. Both ER and PR expression were higher, while the Ki-67 index and Ai
were lower in benign endometrium compared with cancer of any grade (ANOVA; ad hoc
LS
ERexP shex
D).
N ER
Mean (SEM)
PR
mean (SEM)
Ki-67
mean (SEM)
Ai
mean (SEM)
Benign
endometrium
32 95.78 (2.0) 82.78 (4.5) 5.49 (0.7) 5.52 (0.6)
Carcinoma G1 13 79.85 (6.2) 46.08 (6.7) 20.23 (3.3) 13.92 (1.7)
Carcinoma G2 10 73.00 (6.3) 46.70 (9.0) 41.10 (4.0) 16.96 (2.2)
44
and PR in benign postmenopausal endometrium was higher than the receptorpression of endometrial carcinoma of any grade (ANOVA; P < 0.001, ad hoc LSD;< 0.05-0.001) (Table 4). The difference increased with tumor grade (results notown). This comparison shows that the results indicated previously (i.e., that thepression is decreased in carcinoma) could be seen in our material.
Carcinoma G3 6 39.83 (17.2) 21.17 (12.2) 44.00 (4.1) 14.67 (1.9)
ANOVA, sign. P < 0.001 P < 0.001 P < 0.001 P < 0.001
45
3.2.3. Random areas/specific areas
Both ER and PR were studied in at least 10 random areas of the specimen and in theareas of maximal and minimal receptor expression, as described in Methods.Further, for better evaluation of the relation between proliferation and the expressionof sex-steroid-receptors, an evaluation was done in specific areas as follows. Tenspecific areas were identified: 5 foci of maximal proliferation, indicated with highKi-67 index; and 5 foci of minimal proliferation. An evaluation of ER and PRexpressions in these 10 areas representing foci of maximal and minimal proliferationwas established in adjacent sections for 29 cases before and during the progesteronetherapy. Consequently, the changes during the therapy could be compared separatelyin the foci of maximal and minimal proliferation. The results from 10 areas in eachsection were analyzed with repeated-measures ANOVA.Evaluated in the biopsy before the treatment (biopsy 1), the areas representing the fociof high proliferation (Ki-max) expressed higher ER density in G2 tumors, and higherexpression of PR in G1 tumors compared with the foci of low proliferation. The ERand PR evaluated in the biopsy during the progesterone treatment (biopsy 2) showedhigher expression of PR in the foci of high proliferation compared with the foci of lowproliferation in G1 and G2 tumors (F1,12 = 14.28; P < 0.01 and F2,9 = 9.00; P < 0.05,respectively), but not in G3 tumors. The difference in the expression of ER betweenfoci of high and low proliferation was not significant. However, studied separately inG1 tumors, a difference in ER expression could be shown (F1,12 = 6.48; P < 0.05).
Since we had the same evaluation done before and during progesterone treatment,the changes in the mean ER and PR expression could be studied separately in the fociof maximal and minimal proliferation. The Wilcoxon signed rank test was used, and itshowed that the mean PR in the areas of maximal proliferation was significantlydecreased in G1 tumors (P < 0.01) during progesterone treatment, but the decrease inG2 tumors was not significant (P = 0.11), and there was no change at all in G3 tumors(P = 1.0). There was a trend toward decreased PR expression even in the areas ofminimal proliferation in G1 tumors (P = 0.075), but in tumors of any grade there wasno significant decrease from biopsy 1 to 2 in the mean ER expressions in the areas ofeither high or low proliferation (Table 3, paper IV).
The results from the evaluation of sex steroid receptors in the specific foci of maximaland minimal proliferation agree with the results from the evaluation of random foci:decrease of PR expression was revealed in both cases. The results from the study ofspecific foci of high and low proliferation reveal a covariation of the high expressionof sex steroid receptors and high proliferation in grade 1 and 2 tumors.Decrease in PR expression could be seen in the foci of maximal proliferation, and, aswill be shown in results later, a decrease of proliferation was marked in these foci.Both changes are probably effects of progesterone therapy in the same foci. Thecovariation of receptors and proliferation in tumors of grades 1 and 2 may help explainwhy these tumors respond to progesterone therapy [260, 261, 272]. High ER and PRexpression could also be an estrogen effect, and pretreatment with estrogen has been
46
shown to facilitate the progesterone effect in experimental studies [257].The tumors of premenopausal women are also mostly well-differentiated tumors withsex steroid receptor expression, and unopposed-estrogen etiology such as PCOsyndrome is suspected in these cases. These tumors have shown a good progesteroneresponse [262-264]. Receptor responsiveness may also be an indicator of the specificcarcinogenic pathway of the tumor, and high receptor expression and goodresponsiveness of the tumor to progesterone therapy may provide evidence for thehormone-dependent pathway of carcinogenesis [25-28, 93, 104, 233] and for betterprognosis generally associated with these qualities of the tumor [4, 40, 147, 152, 153,221, 273].
If the changes in receptor expression in this study of endometrioid carcinoma are againcompared with the effect of combined HRT in postmenopausal material, we can seeclear differences: in ca-material, ER is unchanged (↔) and PR decreased (↓) duringprogesterone therapy, while in postmenopausal material, under substitution of bothestrogen and gestagen, ER was decreased (↓) and PR unchanged (↔) with tendency toincrease (↑) (see summary in Table 5). Because the changes in receptor expressionseen in postmenopausal material during combined continuous HRT could beinterpreted as evidence that this hormonal therapy is more favorable than progesteronealone, it could be an argument for some combination of estrogen with progesterone asa therapy against endometrial carcinoma. The increase in progesterone effect bypretreatment with estrogen has been shown in an experimental study [257].
47
4. Homeostasis, indicated as proliferation and apoptosis
Proliferation and apoptosis together regulate the homeostasis of as well benign asmalignant tissues. Using the same methods in evaluation of both benign and malignanttissue, the changes in tissue homeostasis can be followed during internal or externalhormonal manipulation. It is, however, impossible to determine the relation indicatingsteady state situation by the approach we have used. Apoptosis is a quick phenomenaof about 4 hours, while Ki-67 staining indicates all phases of cell cycle except G0.
4.1. Benign endometrium
4.1.1. Cyclic endometrium
Proliferation, indicated as Ki-67 index, was studied in the superficial cyclicendometrium at the end of the menstrual cycle, during decreasing E2 and progesteroneserum levels. Four days prior to menstruation, about 10% of the cells were in activecell phase (Ki-67 index = 10) both in the epithelium and in the stroma. However,a different development of proliferation was seen thereafter with rapidly decreasingproliferation in the epithelium, while increasing proliferation was observed in thestroma until the first day of menstruation, followed by decrease until the secondmenstrual day (Fig. 1, paper I).
Apoptosis, which together with proliferation, regulates the homeostasis in theendometrium, was evaluated as well. The results showed that apoptotic cells were rareand scattered among ordinary cells both in glands and the stroma in sections fromendometrium taken 3 to 4 days prior to menstruation. Two days prior to the onset ofmenstruation, an increasing frequency of apoptotic cells was seen in the epithelium,while apoptotic cells in the stroma were still relatively rare. The apoptotic index inepithelial endometrium increased up to the day of the menstrual start and peaked onthe second day of menstruation. In the stroma an increase in the apoptotic index wasseen on the day of onset of menstruation. The maximal apoptotic index in the stromawas about 50% of that observed in the glandular epithelium.
The homeostasis of the epithelium during the decline in levels of sex steroid hormonesat the end of the luteal phase could be summarized as an involution process with highapoptotic activity and low proliferation. The stroma reveals a more complicatedreaction, with relatively high proliferation and, after that, quick cell turnover with bothhigh apoptotic index and high proliferation, until proliferation decreases on the secondmenstrual day.
The increased proliferation in the stroma in the late luteal phase has been describedpreviously as a second wave of proliferation [2]. The reaction in the stroma is delayedcompared with that in the glandular epithelium, and proliferation was still seen duringthe time of low serum progesterone values. If the reaction is progesterone-induced,
48
growth factors or other local regulators complete it. Another explanation could be thatischemia in the tissue prior to menstruation or endometrial damage during earlyinterstitial bleeding activates the reparation mechanism where the stroma is active, andproliferation is triggered by mechanisms other than hormonal induction. Ischemia isalso a possible inductor of apoptosis. In any event, apoptosis has been shown afterhormonal withdrawal in experimental studies [178-180], and during the luteal phase inother studies [13, 18-21, 23, 69, 162, 166, 172-174].
The decrease of ER and PR in the epithelium prior to and at the time of increasingapoptosis provides evidence for the hormonal regulation of apoptosis in theendometrium. Further, the increase of the cycle-specific apoptotic activity near to thestart of the menstrual flow shown in this study may provide evidence for its role in theinduction of menstruation. However, the factors regulating the vascular changes in thestroma are probably most important for the initiation of the bleeding, and the factorsgiving different hormonal responses in the epithelium and stroma are outside the scopeof these studies. An involution-like process in epithelial glands results in the narrowstraight glands of the early follicular phase. High cell turnover, indicated as a highapoptotic index and high proliferation at the same time, suggests an active role for thestroma in remodeling of the endometrium during menstruation.
4.1.2. Postmenopausal endometrium
The Ki-67 index in the epithelial part of the endometrium was 5.5 (%), and stayed onthe same level during combined continuous HRT (Tables 1 and 2, paper II).Both lower and higher rates have been reported previously [22, 132] in untreatedpostmenopausal endometrium. The proliferation index (Ki-67 = 5.5) in this study ofpostmenopausal women was higher than the Ki-67 index of 2 to 3 (%) found inepithelial glands during the 2 first days of menstruation at the time of minimal meanE2 level (123 pmol/l) of healthy menstruating women (paper I) (MWU test; P < 0.05).We have no E2 values in HRT material, but in the women with endometrialcarcinoma, the mean level of E2 was 89 pmol/l (Table 2), and in the group of healthypostmenopausal women, the levels of E2 should not be higher. Thus the decreasinglevels of both E2 and progesterone give lower proliferation in epithelial tissuecompared with continuous low levels, or during the menstruation process there mayexist other factors that counteract proliferation through different mechanisms.
The apoptotic index both in the epithelium and in the stroma was unchanged duringthe combined therapy. The apoptotic index of 5.2 before HRT is lower than the indexin 2 other studies [22, 274] but identical with the index in a control group of 4postmenopausal women in a third study [67]. The unchanged proliferation in theepithelium during the HRT indicates that progesterone was able to block theproliferative effects of E2 in glands. Thus, the homeostasis of the epithelial part wasunaffected as well as the endometrial thickness. In the stroma the combined effect ofthe therapy is increased proliferation (Tables 1 and 2, paper II).Again in postmenopausal endometrium as well as in cyclic endometrium, differentresponses to hormonal changes were seen in the epithelium and in the stroma. Increase
49
in stromal proliferation is seen during the year of HRT and a decreased incidence ofbreakthrough bleeding has been observed during the same period [210]. A certainvolume of stroma may be needed to support the vascular network and to counteractvascular fragility, as proposed in previous studies [212, 213, 217]. The reaction ofreceptors, with decrease of ER expression and a non-significant tendency towardincrease in PR, indicates that receptors have different sensitivity for combinedtherapies.
4.2. Endometrial carcinoma before, during, and after progesterone
therapy
Since the tumor tissue before progesterone therapy clearly showed a heterogeneousstaining pattern for Ki-67, the expression of Ki-67 was separately evaluated in the fociof maximal expression and in the foci of minimal expression as described in Methods.Only 1 tumor showed a totally homogenous staining for Ki-67 (Table 2, paper IV).The same method of evaluation was interpreted in biopsy 2 and 3. For summary, seeTable 5.
Increased frequency of apoptotic cells was observed near the necrotic areas, and theseareas were mostly excluded together with the necrotic foci. In vital tumor tissue,apoptotic cells were scattered among living cells. The apoptotic index (Ai) was notseparately counted in the areas of higher and lower frequency because > 2000 cellswere needed to state the apoptotic index and the foci of higher apoptotic activity weretoo small.
The Ai was the same across tumors of all grades (Table 3, paper III and Table 6),but a previous study has indicated a possible difference in the frequency of apoptosisaccording to tumor grade [249, 275]. One reason for this difference could be that weexcluded the necrotic areas (with increased apoptosis in and around them), which wereseen more frequently in tumors of high grade.
The Ki-67 index differed according to the tumor grade, with lower overall Ki-67 indexand Ki-max in the tumors of G1 compared with G2 and G3, (Table III, paper III) in thebiopsy before the progesterone therapy. The difference of Ki-67 index was also seen inthe mean values of all 3 biopsies according to tumor grade (Table 3 and 6). A higherproliferation rate, indicated with Ki-67 index in tumors of higher grade [36, 53], moremalignant phenotype [225, 236, 276] and worse prognosis [39, 224] has been observedin earlier studies, but the heterogeneity aspect has not been studied before.
Both Ki and Ki-max were decreased during the progesterone therapy, and thisdecrease was separately seen in tumors of G1 and G2 (Fig. 6), while tumors of grade 3showed unchanged proliferation during the therapy. Ki-min was also unchanged intumors of all grades. Thus the heterogeneity of Ki-67 staining was decreased duringthe study. A more heterogeneous pattern of Ki-67 expression was observed in biopsy 1compared with biopsy 2 and 3 as an effect of marked decrease in Ki-max duringprogesterone treatment.
50
The Ai was unchanged during the progesterone therapy and after both 2-day and 6-daywithdrawal of the therapy (Fig. 7). These results were in accordance with a previousstudy indicating that endometrial carcinoma does not respond with increased apoptosisto progesterone [277]. Another experimental study indicates that Ishikawa cell linefrom a well-differentiated endometrial adenocarcinoma may respond with apoptosis atthe beginning of progesterone therapy [278], but this kind of early reaction is notpossible to show in this study setup.
It was interesting to see that both grade 1 and 2 tumors expressed high ER and PRintensity before the therapy, they showed coexistence of receptors and proliferation insame foci, and they also showed a response to progesterone therapy. Both G2 and G3tumors showed higher proliferation compared with G1 tumors, but tumors of G1 andG2 both responded to the progesterone therapy, while G3 tumors stayed on a higherproliferation level during the therapy (Fig. 5 and Table 6). Higher proliferation rate hasbeen shown in tumors of women who have not used hormonal therapy [279].In contrast, however, lower proliferation rate, higher expression of sex steroidhormone receptors, and better response to the progesterone therapy may argue for theestrogen-dependent pathway of carcinogenesis [26, 28, 104, 244]. Great similaritieswere observed in G1 and G2 tumors in our material. The similar behavior of G1 andG2 tumors as well as the better prognosis in this group [39, 227, 280] compared withG3 tumors could support the previous proposal of using a 2-tiered instead of 3-tieredgrading system [281] for endometrial carcinoma. In this material, both G1 and G2tumors show changed homeostasis during progesterone therapy with decreasedproliferation rate and unchanged apoptosis, while the homeostasis of G3 tumors wasunaffected by progesterone (Table 5).
Apoptosis could counteract proliferation and modulate the homeostasis during thetherapy trial. However, apoptosis was unchanged during the therapy and no variationwas seen between the tumors of different grade. Thus, proliferation alone regulated thegrowth in this material and proliferation alone was responsible for the increasingaggressiveness of tumors with advancing grade.
51
A
Biopsy 1 - 3
321
Mea
n K
i-67
(cel
ls/1
00) 50
40
30
20
10
B
Biopsy 1 - 3
321
Mea
n K
i-67
(cel
ls/1
00) 50
40
30
20
10
GRADE
1
2
3
C
Biopsy 1 - 3
32 1
Mea
n K
i-67,
max
(cel
ls/1
00)
80 70 60 50 40 30 20
D
Biopsy 1 - 3
321
Mea
n K
i-67
max
(ce
lls/1
00)
80706050403020
GRADE
1
2
3
Figure 6
A. Ki-67 expression was decreased during progesterone therapy (biopsy 2) and after
withdrawal (biopsy 3) compared to biopsy 1.
B. There was an interaction of tumor grade with Ki-67 (F2,52 =3.49; P <0.05, repeated
measurements ANOVA). G3 tumors showed a different pattern of behavior, with unchanged
proliferation during the progesterone therapy and after withdrawal, while G1 and G2 tumors
showed decrease of proliferation during the therapy.
C. Proliferation, measured as Ki-67 expression in the areas of maximal proliferation (Ki-max)
was decreased during progesterone therapy and after withdrawal (biopsies 1 and 2).
D. Tumors of all grades show similar behavior through the biopsy series even if the decrease
of Ki-max was significant only in G1 and G2 tumors. Consequently, no interaction between
grade and Ki-max could be observed.
52
A
Biopsy 1 - 3
321
Mea
n A
i (c
ells
/100
0)20,0
15,0
10,0
B
Biopsy 1 - 3
321
Mea
n A
i (c
ells
/100
0)
20
18
16
14
12
10
GRADE
1
2
3
Figure 7
Apoptotic index (Ai) in biopsies before (1), during (2), and 2 or 6 days after (3) medroxy-
progesterone treatment in endometrial carcinoma.
A. All tumors (29 patients).
B. Tumors grouped by grade 1, 2, and 3 (13, 10, and 6 patients). Ai was in the same range
during the study period and in tumors of different grades. There was no interaction of
histopathological tumor grade and Ai. (Statistics by repeated measurements ANOVA.)
53
4.3. Predictive factors of progesterone therapy
The decrease in proliferation was the main effect of progesterone in this study. Thedecrease of proliferation, further referred to as delta-Ki, was therefore studied in thescope of factors illuminating the properties of the tumors before the therapy. Previousstudies have indicated the value of sex steroid receptor expression as a predictivefactor of tumor responsiveness to progesterone therapy [260, 261], but the responserates are not directly correlated with receptors. Since there also are indications fromexperimental studies that stromal factor may mediate progesterone effect inendometrial cancer cells (Ischikawa cells) [5, 282], we studied both epithelial andstromal factors as possible predictors of progesterone effect.
4.3.1. Epithelial factors
Each of ER, PR, and bcl-2 in the epithelial part of tumors showed correlation withdelta-Ki (Pearson's correlation coefficient, r = 0.61; P = 0.001 r = 0.51; P = 0.005, andr = 0.38; P = 0.045, respectively) (Figure 1, paper IV). Tumor grade also has greatimportance for the response. Attempts to fit a linear multiple regression test of thefactors were complicated by the intrinsic correlations between the factors. Nonlinearmodels were also limited because of the sample size. However, the receptor expressionmay reflect the effect of estrogen on the tumor. The correlations may therefore supportthe theory that progesterone's effect consists of inhibition of estrogen-inducedproliferation by activation of 17ß-HSD type 2, which converts the biologically activeEstradiol (E2) to less active estriol, E1. This enzyme has been shown to be present innearly one half of the endometrial carcinomas [95, 100, 101]. Further, the expressionof 17ß-HSD type 2 is inversely correlated with the age of the patients [101], and theseresults contribute to the positive outcome of progesterone therapy observed in thegroups of young patients with low-grade tumors [262-264].
The correlation of bcl-2 and delta-Ki was due to 4 patients who had bcl-2 negative(score 0) tumors with no response to progesterone. There has been earlier implicationfrom a study of breast carcinoma that bcl-2 could predict effects of hormonal therapy[283], but results from the studies with bcl-2 in endometrial carcinoma have not beenconclusive [73].
4.3.2. Stromal factors
The amount of stroma in tumors decreased with increasing tumor grade, as illustratedin Figure 8 (ANOVA, P = 0.003). In pairwise comparison it turned out that G3 tumorsdiffered from G1 and G2 tumors (P =0.001 and P = 0.03 respectively). This differencewas expected since the amount of stroma is indirectly included in the criteria of tumorgrade in endometrial carcinoma. Both ER and PR expression in the stroma was lowerthan in the epithelial part of the tumors (Wilcoxon signed-rank test, P < 0.001 for both)(Figure 9) and correlated to the amount of stroma (Pearson's correlation coefficient, r =0.63; P < 0.001 for ER and r = 0.72, P < 0.001 for PR, respectively). There was nosignificant correlation between ER or PR in stroma and delta-Ki. Thus, no direct
support was found in this material for the importance of stromal mediating factor forthe progesterone therapy.
F
F
40
54
PRER
100
80
60
40
20
0
Epith.
Str.
111
10953
1427
%
A. The amount of stroma in 29 tumors of
endometrial carcinoma.
The amount of stroma (% of total
volume) decreased with increasing tumor
grade (ANOVA, P = 0.003, ad hoc LSD
G1/G2, ns; G1/G3, P < 0.01; G2/G3, P <
0.05).
igure 8 – Amount of stroma.
B. Sex steroids receptor expression
in 29 endometrial carcinomas.
ER and PR expression in epithelial
part of the tumors was higher than in
stroma (Wilcoxon signed ranks test,
P < 0.001 for both).
igure 9 – Sex steroid receptors
Tumor grade 1-3
321
Amou
nt o
f stro
ma
(%)
30
20
10
0
55
5. Bcl-2 and P53
5.1. Bcl-2
In cyclic endometrium, bcl-2 was scored as relatively low both in the epithelium andin the stroma, and this observation was in accordance with previous studies (Fig. 1,paper I) [52, 69]. A slight increase of bcl-2 expression in the epithelium just prior tomenstruation, at the same time period with increasing apoptosis, cannot be explainedin this material without knowing the possible occurrence of bax.
The bcl-2 score in endometrial carcinoma showed no significant relation to tumorgrade, apoptosis, and proliferation or sex steroid receptors. Neither was there anychange in expression of bcl-2 during progesterone therapy or withdrawal. The 4tumors without any expression of bcl-2 showed no response (decrease of Ki-67 index)to progesterone therapy, in contrast to most tumors showing bcl-2 expression, whichresponded to progesterone with decreased proliferation.
5.2. P53
Deranged p53 gene expression in most cells of the tumor was observed in only 2 casesout of 29 tumors, and this result was expected, since a low incidence of p53 mutationin endometrioid endometrial carcinoma has been reported earlier [237, 241-243]. Theexpression of p53 is frequently seen in endometrial tumors of more aggressivesubtypes and, probably, with carcinogenesis independent of estrogen. However, alltumors of endometrioid endometrial carcinoma were not negative according to p53expression.
56
Table 5Summary of the studies in papers I, II, and IVER and PR, Ki-67 index, Ai, and bcl-2 were studied during hormonal changes. The hormonal
changes or provocation studied were as follows: intrinsic decrease of estradiol and
progesterone in paper I, combination of estrogen and progesterone therapy in paper II and
progesterone therapy and withdrawal in paper IV. The results after withdrawal are compared
with the results before the therapy. Symbols used: ↓ = significantly decreased, ↑ =
significantly increased, ↔ = unchanged, (↑) or (↓) = tendency to increase or decrease.
Study tissue Materials Changes of
steroid
level
Epithelium. Stroma
Fertile endom.
Paper I
37 women
75 biopsies
E↓
P↓
ER↓
PR↓
Ki-67↓
Ai↑
Bcl-2↓
ER↑↓
PR↑↓
Ki-67↑↓
Ai↑↓
Bcl-2↑↓
PM
endometrium
Paper II
92 pm women,
2 biopsies from
43 women
E↑
P↑
ER↓
PR↔(↑)
Ki-67↔
Ai↔
ER↓
PR↔ (↑)
Ki-67↑
Ai↔
Endometrial
Carcinoma
Paper IV:
Progesterone
therapy
29 patients
3 biopsies/pat.
P↑ ER↔
PR-max↓
Ki-67↓
Ki-max↓
Ai↔
Bcl-2↔
P52↔
No correl. betw.
stroma amount and
prog. effect in prolif.
Endom.
Carcinoma
Paper IV
Progesterone
withdrawal
29 patients
20 pat. 2 days
withdrawal
9 pat. 6 days
withdrawal
P↓ ER↔
PR-max↓
Ki-67↓
Ki-max↓
Ai↔
Bcl-2↔
P52↔
57
Table 6Summary of steroid receptor expression, proliferation, and apoptosis of endometrioid
endometrial carcinoma of grades 1–3 throughout the study period of biopsies 1–3. The
arrows in column biopsy 1 indicate higher (↑), lower (↓) or equal (↔) levels compared with
other grades. In columns for biopsies 2 and 3 the arrows indicate a change from the previous
biopsy.
Endometrioid
endometrial
Carcinoma,
grades 1-3
Biopsy 1
Before therapy
Biopsy 2
Changes during
progesterone therapy
Biopsy 3
Changes during
progesterone withdrawal
Grade 1 ER↑, ER-max↑
PR↑, PR-max
Ki-67↓
Ai↔
ER↔, ER-max↔
ER-het↔
PR↓, PR-max↓
PR-het↓
Ki-67↓
Ki-het↓
ER, ER-max↔
ER-het↔
PR↑, PR-max↑
PR-het↔
Ki-67↔, all variables
Grade 2 ER↑, ER-max↑
PR↑, PR-max↑
Ki-67↑
Ai↔
ER↔, ER-max↔
ER-het↔
PR↓, PR-max↓
PR-het↔
Ki-67↓
Ki-het↓
ER↔, all variables
PR↔, all variables
Ki-67↔, all variables
Grade 3 ER↓ ,ER-max↔
PR↓
Ki-67↑
Ai↔
ER↔, all variables
PR↔, all variables
Ki-67↔, all variables
Ai↔
ER↔, all variables
PR↔, all variables
Ki-67↔, all variables
Ai↔
58
SUMMARYProliferation and apoptosis, the main regulators of tissue homeostasis, as well as sexsteroid receptors and p53 and bcl-2, were evaluated in endometrium and inendometrial carcinoma during hormonal changes and manipulation as follows: 1. Fertile endometrium during declines in E2 and progesterone levels during the lateluteal phase and beginning of menstruation (Table 5); 2. Postmenopausal endometrium during low hormonal levels of postmenopause andstable increased estrogen and gestagen levels during HRT (Table 5); and 3. Endometrial carcinoma during postmenopausal levels of estrogens andprogesterone, during medroxy-progesterone therapy with 20 mg daily dose and duringfalling levels of medroxy-progesterone after the withdrawal of the therapy (Tables 5and 6).
The following effects were observed during the hormonal circumstances describedabove (See Tables 5 and 6):
Homeostasis (apoptosis/proliferation)The epithelial tissue showed increasing apoptosis and decreasing proliferation at theend of the luteal phase and during menstruation. This image of involution in theepithelium was in contrast with the development in the stroma, where highproliferation and increasing apoptosis were observed, indicating quick cell turnover.Combined continuous HRT for postmenopausal women induced no increase inproliferation of the endometrial epithelium, and apoptosis was unchanged as well.Increased proliferation was observed in the stroma. There was increasing proliferationin endometrial carcinoma with increasing tumor grade, while apoptosis did not varyaccording to the grade. Decreased proliferation and unchanged apoptosis in G1 and G2tumors were observed during progesterone therapy. No change was seen during thetherapy in G3 tumors, nor in tumors of any grade during the withdrawal of the therapy.
Hormonal sensitivityThe hormonal sensitivity illuminated as ER and PR receptors was decreased in theendometrial epithelium during decreasing serum levels of both E2 and progesterone. Inthe stroma an increase was observed prior to menstruation before the fall duringmenstruation. ER was decreased during combined continuous HRT with E2/NETAboth in the epithelium and in the stroma, while PR was not significantly changed.During progesterone therapy, endometrial carcinoma showed decrease of PR, whileER was unchanged. There was a covariation of the sex steroid receptors andproliferation in the same foci of G! and G2 tumors.P53 was rare in endometrial carcinoma of endometrioid type. This fact together withthe antiproliferative hormonal effects in G1 and G2 tumors argues for the hormone-dependent pathway of carcinogenesis of most endometrioid endometrial tumors ofgrades 1 and 2. The changes of the antiapoptotic factor bcl-2 are not alone conclusivein cyclic endometrium or in endometrial carcinoma.
59
CONCLUSIONSApoptosis is involved in the mechanism of menstruation. Both apoptosis andproliferation collaborate in endometrial remodeling during menstruation, when theepithelium returns to the status of narrow glands and stromal damage is repaired.
The balance between proliferation and apoptosis is maintained unchanged inpostmenopausal endometrial epithelium during HRT with 2 different regimens.This balance contributes to endometrial safety. Increased proliferation observed in thestroma with the same therapy may counteract breakthrough bleeding since stromalsupport of the vascular network was increased. It may also be a safety issue, andshould be further studied.
Increasing discrepancy between proliferation and apoptosis with increasing tumorgrade contributes to the more aggressive growth of grade 3 endometrial carcinomascompared with the G1 and G2 tumors.
The response to progesterone therapy in G1 and G2 tumors may be facilitated by thecoexistence of sex steroid receptors and high proliferation in the same foci.
Progesterone therapy of endometrial carcinoma works via decreased proliferation,but not via increased apoptosis.
The hormonal sensitivity, illuminated as ER and PR, is regulated by estradiol andprogesterone in the epithelial endometrium and in endometrial carcinoma of grades 1and 2, but in the stroma the effect is differently modulated by local factors, and thestimulatory effect of progesterone may also be suspected.
60
ACKNOWLEDGEMENTSThe studies contributing to these theses were carried out at the Department of ClinicalSciences, Obstetrics and Gynecology, Department of Oncology and Department ofPathology, Umeå University. I wish to express my warm and sincere gratitude to:
Torbjörn Bäckström for guiding me in the world of science from the first study tocompleting this thesis, for his endless enthusiasm, and for sharing his deep knowledgeof steroid hormones.
My second supervisors, Karin Boman and Stefan Cajander. Karin for dealing with herexperience of endometrial cancer, being my clinical teacher and coworker in the fieldof gynecological oncology. She and her wonderful family generously opened theirhome to me during the short and long visits in Umeå. Stefan for all the momentssitting at the double microscope, for the most enthusiastic supervision in themicroscopic world, and for the microphotographs.
My friend in science and sailing, Inga-Stina Ödmark, for giving discussions, herconstantly positive attitude, and support in many ways.
My coworker Patric Westin for teaching me the morphology of apoptosis.Anders Berg for giving me the opportunity to share the utilities of his laboratory.Ingalis Fransson, Birgitta Ekholm, and Elisabeth Dahlberg for their skillful technicalassistance in processing the cancer material.
Peter Pitkänen for letting me use his laboratory to process the biopsies of my firststudy, and Birgitta Höglin and Marianne Jäderling for their excellent technicalassistance.
Mats G Karlsson for coauthorship and Monica Sievert and Vivianne Sjökvist for theirkindly assistance with the postmenopausal material.
Björn Risberg for coauthorship.
Lars Berglund, the head of the Department of KK Sundsvall, for his positive attitudeand great support over the years. He never stopped believing in my abilities.
My colleagues in KK Sundsvall for all the support and good laughs together.
Hans Malker, the head of the FoU centre, Landstinget Västernorrland, for alwaysfighting for resources for scientific work.
Gunnar Nordahl and Tatjana Pavlenko for skillful statistical help, teaching and co-authorship, and Annika Dahl and Erling Englund for all statistical advice anddiscussions.
61
Lennart Bråbeck and his staff at the FoU unit, Per-Arne Lyre'n, Vivan Rönnqvist,Thomas Näsberg, Mats Sjöling, and Marika Augutis for all their practical help.
The staff at the Medical Library for friendly cooperation and excellent quick service.
The Dionysos Daughters, my neat group of female colleagues, for all good stories andfun together, while tasting the wines.
Nils-Gunnar for sharing the duties of parenthood.
Ann-Christine and Harri, my closest friends, for always being there and patientlywaiting for the party.
Boris for all his kind support.
Johan for bringing so much happiness into my life.
62
63
REFERENCES1. Satyaswaroop PG, Zaino RJ, Mortel R. Human endometrial adenocarcinoma
transplanted into nude mice: growth regulation by estradiol. Science1983;219(4580):58-60.
2. Clarke C, Sutherland R. Progestin regulation of cellular proliferation. EndocrRev 1990;11:266-72.
3. Casper F, R. Regulation of estrogen/progestogen receptors in the endometrium.Int J Fertil 1996;41(1):16-21.
4. Kauppila A, Friberg LG. Hormonal and cytotoxic chemotherapy forendometrial carcinoma. Steroid receptors in the selection of appropriatetherapy. Acta Obstet Gynecol Scand Suppl 1981;101:59-64.
5. Yang S, Fang Z, Gurates B, Tamura M, Miller J, Ferrer K, et al. Stromal PRsmediate induction of 17beta-hydroxysteroid dehydrogenase type 2 expression inhuman endometrial epithelium: a paracrine mechanism for inactivation of E2.Mol Endocrinol 2001;15(12):2093-105.
6. Imai T, Kurachi H, Adachi K, Adachi H, Yoshimoto Y, Homma H, et al.Changes in epidermal growth factor receptor and the levels of its ligands duringmenstrual cycle in human endometrium. Biol Reprod 1995;52(4):928-38.
7. Corleta H, Capp E, Strowitzki T. Cycle modulation of insulin-like growthfactor-binding protein 1 in human endometrium. Braz J Med Biol Res2000;33(11):1387-91.
8. Hata H, Hamano M, Watanabe J, Kuramoto H. Role of estrogen and estrogen-related growth factor in the mechanism of hormone dependency of endometrialcarcinoma cells. Oncology 1998;55 Suppl 1:35-44.
9. Kerr J, F, Wyllie A, H, Currie A, R. Apoptosis: a basic biological phenomenonwith wideranging implications in tissue kinetics. Br J Cancer 1972;26:239-59.
10. Evans-Storms R, B, Cidlowsky J, A. Regulation of apoptosis by steroidhormones. J Steroid Biochem Mol Biol 1995;53:1-8.
11. Tenniswood M, P, Guenette R, S, Lakins J, Mooibroek M, Wong P, Welsh JE.Active cell death in hormone dependent tissue. Cancer Metastasis Rev1992;11:197-220.
12. Walker N, I, Bennett R, E, Kerr J, F, R. Cell death by apoptosis duringinvolution of the lactating breast in mice and rats. Am J Anat 1989;185:19-32.
13. Spencer S, J, Cataldo N, A, Jaffe R, B. Apoptosis in the human femalereproductive tract. Obstet Gynecol Surv 1996;51(5):314-23.
14. Heermeier K, Benedict M, Li M, Furth P, Nenez G, Henninghausen L. Bax andBcl-xshort are induced at the onset of apoptosis in involuting mammaryepithelial cells. Mech Dev 1996;56:197-207.
15. Bartelmez G, W. Histological studies of the menstruating mucous membrane ofthe human uterus. Carnegie Institute Contrib. Embryol. 1933;24:141-86.
16. Cornillie F, J, Lauweryns JM, Brosens I, A. Normal human endometrium. Anultrastructural survey. Gynecol Obstet Invest 1985;20:113-29.
64
17. Coates P, J, Hales S, A, Hall P, A. The association between cell proliferationand apoptosis: studies using the cell cycle-associated proteins Ki67 and DNApolymerase alpha. J Pathol 1996;178:71-77.
18. Hopwood D, Levison D, A. Atrophy and apoptosis in the cyclical humanendometrium. J Pathol 1975;119:159-66.
19. Tabidzadeh S, Kong Q, F, Satyaswaroop P, G, Zupi E, Marconi D, Romanini C,et al. Distinct regional and menstrual cycle dependent distribution of apoptosisin human endometrium. Potential regulatory role of T cells and TNF-alfa.Endocr J 1994;2:87-95.
20. Tabibzadeh S. Signals and molecular pathways involved in apoptosis, withspecial emphasis on human endometrium. Hum Reprod Update 1995;1(4):303-23.
21. Verma V. Ultrastructural changes in human endometrium at different phases ofthe menstrual cycle and their functional significance. Gynecol Obstet Invest1983;15:193-212.
22. Morsi HM, Leers MP, Jager W, Bjorklund V, Radespiel-Troger M, el KabarityH, et al. The patterns of expression of an apoptosis-related CK18 neoepitope,the bcl-2 proto-oncogene, and the Ki67 proliferation marker in normal,hyperplastic, and malignant endometrium. Int J Gynecol Pathol2000;19(2):118-26.
23. Morsi HM, Leers MP, Radespiel-Troger M, Bjorklund V, Kabarity HE, Nap M,et al. Apoptosis, bcl-2 expression, and proliferation in benign and malignantendometrial epithelium: An approach using multiparameter flow cytometry.Gynecol Oncol 2000;77(1):11-7.
24. Bianchini F, Kaaks R, Vainio H. Overweight, obesity, and cancer risk. LancetOncol 2002;3(9):565-74.
25. Kurman RJ, Kaminski PF, Norris HJ. The behavior of endometrial hyperplasia.A long-term study of "untreated" hyperplasia in 170 patients. Cancer1985;56(2):403-12.
26. Sherman ME. Theories of endometrial carcinogenesis: a multidisciplinaryapproach. Mod Pathol 2000;13(3):295-308.
27. Boyd J. Estrogen as a carcinogen: the genetics and molecular biology of humanendometrial carcinoma. Prog Clin Biol Res 1996;394:151-73.
28. Jeyarajah A, Oram D, Jacobs I. Molecular events in endometrial carcinogenesis.Int J Gynecol Cancer 1996;6:425-38.
29. Tamm I, Schriever F, Dorken B. Apoptosis: implications of basic research forclinical oncology. Lancet Oncol 2001;2(1):33-42.
30. Wyllie A, H, Kerr J, F, R, Currie A, R. Cell Death: the significance ofapoptosis. Int Rev Cytol 1980;68:251-306.
31. Schwartzman RA, Cidlowsky JA. Apoptosis:the biochemistry and molecularbiology of programmed cell death. Endocr rev 1993;14(2):133-51.
32. Fesus L, Davies P, J, A, Piacentini M. Apoptosis: molecular mechanisms inprogrammed cell death. Eur J Cell Biol 1991;56:170-77.
33. English H, F, Kyprianou N, Isaacs jT. Relationship between DNAfragmentation and apoptosis in the programmed cell death in the rat prostate
65
following castration. Prostate 1989;15:233-50.34. Kerr JF, Searle J. Deletion of cells by apoptosis during castration-induced
involution of the rat prostate. Virchows Arch B Cell Pathol 1973;13(2):87-102.35. Speroff L, Glass R, Kase N. Hormone biosyntesis, methabolism, and
mechanism of action. In: C. Mitchell (ed.) Clinical gynecologic endocrinologyand infertility. 5 ed. pp. 31-92. Baltimore: Williams & Wilkins; 1994.
36. Yabushita H, Masuda T, Sawaguchi K, Noguchi M, Nakanishi M. Growthpotential of endometrial cancers assessed by a Ki-67 Ag/DNA dual-color flow-cytometric assay. Gynecologic Oncology 1992;44:263-67.
37. Nordstrom B, Strang P, Bergstrom R, Nilsson S, Tribukait B. A comparison ofproliferation markers and their prognostic value for women with endometrialcarcinoma. Ki-67, proliferating cell nuclear antigen, and flow cytometric S-phase fraction. Cancer 1996;78(9):1942-51.
38. Stendahl U, Strang G, Wagelius G, Bergström R, Tribukait B. Prognosticsignificance of proliferation in endometrial adenocarcinomas: an multivariateanalysis of clinical and flow cytometric variables. Int J Gynecol Pathol1991;10:271-84.
39. Salvesen HB, Iversen OE, Akslen LA. Identification of high-risk patients byassessment of nuclear Ki-67 expression in a prospective study of endometrialcarcinomas. Clin Cancer Res 1998;4(11):2779-85.
40. Boman K, Mäentausta O, Bäckström T, Strang P, Stendahl U. Sex steroidhormones and receptors in relation to S-phase fraction and ploidy level inendometrial carcinoma. Anticancer Res 1995;15:999-1002.
41. Lukes A, S, Kohler M, F, Pieper C, P, all e. Multivariable analysis of DNAploidy, p53, and HER-2/neu as prognostic factors in endometrial carcer. Cancer1994;73:2380-85.
42. Thornton J, G, Ali S, O`Donovan P, Griffin N, Wells M, MacDonald R, R.Flow cytometric studies of ploidy and proliferative indices in the Yorkshire trialof adjuvant progestogen treatment of endometrial cancer. Br J Obstet Gynecol1993;100:253-61.
43. Quarmby VE, Korach KS. The influence of 17 beta-estradiol on patterns of celldivision in the uterus. Endocrinology 1984;114(3):694-702.
44. Geum D, Sun W, Paik SK, Lee CC, Kim K. Estrogen-induced cyclin D1 andD3 gene expressions during mouse uterine cell proliferation in vivo: differentialinduction mechanism of cyclin D1 and D3. Mol Reprod Dev 1997;46(4):450-8.
45. Kato N, Watanabe J, Jobo T, Nishimura Y, Fujisawa T, Kamata Y, et al.Immunohistochemical expression of cyclin E in endometrial adenocarcinoma(endometrioid type) and its clinicopathological significance. J Cancer Res ClinOncol 2003;129(4):222-6.
46. Sutherland RL, Prall OW, Watts CK, Musgrove EA. Estrogen and progestinregulation of cell cycle progression. J Mammary Gland Biol Neoplasia1998;3(1):63-72.
47. Musgrove EA, Swarbrick A, Lee CS, Cornish AL, Sutherland RL. Mechanismsof cyclin-dependent kinase inactivation by progestins. Mol Cell Biol
66
1998;18(4):1812-25.48. Strang P, Stendahl U, Tribukait B. Prognostic significance of S-phase fraction
as measured by DNA flow cytometry in gynecologic malignancies. Ann N YAcad Sci 1993;677:354-63.
49. Tribukait B. Clinical DNA flow cytometry. Med Oncol Tumor Pharmacother1984;1(4):211-8.
50. Gerdes J, Schwab U, Lemke H, Stein H. Production of the mouse monoclonalantibody reactive with a human nuclear antigen associated with cellproliferation. Int J Cancer 1983;31:13-20.
51. Gerdes J, Lemke H, Baisch H, Wacker H-H, Schwab U, Stein H. Cell cycleanalysis of a cell proliferation-associated human nuclear antigen defined by themonoclonal antibody Ki-67. The journal of immunology 1984;133(4):1710-15.
52. Gompel A, Sabourin J, C, Martin A, Yaneva H, Audoin J, Decroix Y, et al. Bcl-2 expression in normal endometrium during the menstrual cycle. Am J Pathol1994;144:11951202.
53. Kokeguchi S, Hayase R, Sekiba K. Proliferative activity in normalendometrium and endometrial carcinoma measured by immunohistochemistryusing Ki-67 and anti-DNA polymerase alfa antibody, and flow cytometry. ActaMed Okayama 1992;46:113-21.
54. Gavrieli Y, Sherman Y, Ben-Sasson SA. Identification of programmed celldeath in situ via specific labeling of nuclear DNA fragmentation. J Cell Biol1992;119(3):493-501.
55. Wisman JH, Jonker RR, Kejzer R, Van de Velde CJH, Cornelisse CJ, vanDierendonck JH. A new method to detect apoptosis in paraffin sections: in situend labelling of fragmented DNA. J Histochem Cytochem 1993;41(1):7-12.
56. Yasuda M, Umemura S, Osamura RY, Kenjo T, Tsutsumi Y. Apoptotic cells inthe human endometrium and placental villi: pitfalls in applying the TUNELmethod. Arch Histol Cytol 1995;58(2):185-90.
57. Brändström A, Westin P, Bergh A, Cajander S, Damber J-E. Castration inducesapoptosis in the ventral prostate but not in an androgensensitive prostaticadenocarcinoma in the rat. Cancer Res 1994;54:3594-601.
58. Maciorowski Z, Klijanienko J, Padoy E, Mosseri V, Fourquet A, Chevillard S,et al. Comparative image and flow cytometric TUNEL analysis of fine needlesamples of breast carcinoma. Cytometry 2001;46(3):150-6.
59. Wyllie A, H. Glucocorticoid-induced thymocyte apoptosis is associated withendogenous endonuclease activation. Nature 1980;284:555--56.
60. Rosl F. A simple and rapid method for detection of apoptosis in human cells.Nucleic Acids Res 1992;20(19):5243.
61. Davoli MA, Ren Y, Zhu Y, Fourtounis J, Jones C, Robertson GS, et al.Automated analysis of global ischemia-induced CA1 neuronal death usingterminal UTP nick end labeling (TUNEL). J Neurosci Methods 2002;115(1):55-61.
62. Weibel ER. Stereological methods. London: Academic press inc. London LTD;1979.
67
63. Wilkinson EJ, Hendricks JB. Role of the pathologist in biomarker studies. JCell Biochem Suppl 1995;23:10-8.
64. Hockenbery D, Nunez G, Milliman C, Schreiber R, D, Korsmeyer S, J. Bcl-2 isinner mitochondrial membrane protein that blocks programmed cell death.Nature 1995;348:334-36.
65. LeBrun D, P, Warnke R, A, Cleary M, L. Expression of bcl-2 in fetal tissuessuggests a role in morphogenesis. Am J Pathol 1993;142:743-53.
66. Tao X-J, Tilly K, I, Maravei D, V, al. e. Different expression of members of thebcl-2 gene family in proliferative and secretory human endometrium: glandularepithelial cell apoptosis is associated with increased expression of bax. Journalof clinical endocrinology and metabolism 1997;82:2738-46.
67. Kokawa K, Shikone T, Otani T, Nishiyama R, Ishii Y, Yagi S, et al. Apoptosisand the expression of Bax and Bcl-2 in hyperplasia and adenocarcinoma of theuterine endometrium. Hum Reprod 2001;16(10):2211-8.
68. Koh E, A, Illingworth P, J, Duncan W, C, Critchley H, O, D.Immunolocalization of bcl-2 protein in human endometrium in the menstrualcycle and simulated early pregnancy. Hum Reprod 1995;10:1557-62.
69. Otsuki Y, Misaki O, Sugimoto,O, Ito Y, Tsjujimoto Y, Akao Y. Cyclic bcl-2gene expression in human uterine endometrium during menstrual cycle. Lancet1994;344:28-29.
70. Henderson GS, Brown KA, Perkins SL, Abbott TM, Clayton F. bcl-2 is down-regulated in atypical endometrial hyperplasia and adenocarcinoma. Mod Pathol1996;9(4):430-8.
71. Bozdogan O, Atasoy P, Erekul S, Bozdogan N, Bayram M. Apoptosis-relatedproteins and steroid hormone receptors in normal, hyperplastic, and neoplasticendometrium. Int J Gynecol Pathol 2002;21(4):375-82.
72. Pecci A, Scholz A, Pelster D, Beato M. Progestins prevent apoptosis in a ratendometrial cell line and increase the ratio of bcl-XL to bcl-XS. J Biol Chem1997;272(18):11791-8.
73. Giatromanolaki A, Sivridis E, Koukourakis MI, Harris AL, Gatter KC. Bcl-2and p53 expression in stage I endometrial carcinoma. Anticancer Res1998;18(5B):3689-93.
74. Sakuragi N, Ohkouchi T, Hareyama H, Ikeda K, Watari H, Fujimoto T, et al.Bcl-2 expression and prognosis of patients with endometrial carcinoma. Int JCancer 1998;79(2):153-8.
75. Levine AJ, Momand J, Finlay CA. The p53 tumour suppressor gene. Nature1991;351(6326):453-6.
76. Wang XW, Harris CC. p53 tumor-suppressor gene: clues to molecularcarcinogenesis. J Cell Physiol 1997;173(2):247-55.
77. Harper JW, Adami GR, Wei N, Keyomarsi K, Elledge SJ. The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1 cyclin-dependent kinases.Cell 1993;75(4):805-16.
78. Marchetti A, Doglioni C, Barbareschi M, Buttitta F, Pellegrini S, Bertacca G, etal. p21 RNA and protein expression in non-small cell lung carcinomas:evidence of p53-independent expression and association with tumoral
68
differentiation. Oncogene 1996;12(6):1319-24.79. Strang P, Nordström B, Nilsson S, Bergström B, Tribukait B. Mutant p53
protein as a predictor of survival in endometrial carcinoma. European journal ofcancer 1996;32A(4):598-602.
80. Li SF, Shiozawa T, Nakayama K, Nikaido T, Fujii S. Stepwise abnormality ofsex steroid hormone receptors, tumor suppressor gene products (p53 and Rb),and cyclin E in uterine endometrioid carcinoma. Cancer 1996;77(2):321-9.
81. Ito K, Watanabe K, Nasim S, all e. Prognostic significance of p53overexpression in endometrial cancer. Cancer research 1994;54:4667-70.
82. de la Calle-Martin O, Romero M, Fabregat V, Ercilla G, Vives J, Yague J. MspIpolymorphism of the human p53 gene. Nucleic Acids Res 1990;18(16):4963.
83. Speroff L, Vande Wiele RL. Regulation of the human menstrual cycle. Am JObstet Gynecol 1971;109(2):234-47.
84. Irwin JC, Utian WH, Eckert RL. Sex steroids and growth factors differentiallyregulate the growth and differentiation of cultured human endometrial stromalcells. Endocrinology 1991;129(5):2385-92.
85. Lebovic DI, Shifren JL, Ryan IP, Mueller MD, Korn AP, Darney PD, et al.Ovarian steroid and cytokine modulation of human endometrial angiogenesis.Hum Reprod 2000;15 Suppl 3:67-77.
86. Moller B, Rasmussen C, Lindblom B, Olovsson M. Expression of theangiogenic growth factors VEGF, FGF-2, EGF and their receptors in normalhuman endometrium during the menstrual cycle. Mol Hum Reprod2001;7(1):65-72.
87. Perrot-Applanat M, Ancelin M, Buteau-Lozano H, Meduri G, Bausero P.Ovarian steroids in endometrial angiogenesis. Steroids 2000;65(10 -11):599-603.
88. Moutsatsou P, Sekeris CE. Estrogen and progesterone receptors in theendometrium. Ann N Y Acad Sci 1997;816:99-115.
89. Perret S, Dockery P, Harvey BJ. 17beta-oestradiol stimulates capacitative Ca2+entry in human endometrial cells. Mol Cell Endocrinol 2001;176(1-2):77-84.
90. Lessey B, A, Killam A, P, Metzger D, A, Haney A, F, Greene G, L, McCarty K,S. Imunohistochemical anlysis of human uterine estrogen and progesteronereceptors throughout the menstrual cycle. J Clin Endocrinol Metab1988;67(2):334-40.
91. Hachisuga T, Hideshima T, Kawarabayashi T, Eguchi F, Emoto M, ShirakusaT. Expression of steroid receptors, Ki-67, and epidermal growth factor receptorin tamoxifen-treated endometrium. Int J Gynecol Pathol 1999;18(4):297-303.
92. Hulka BS. Links between hormone replacement therapy and neoplasia. FertilSteril 1994;62(6 Suppl 2):168S-75S.
93. Deligdisch L, Holinka CF. Endometrial carcinoma: two diseases? CancerDetect Prev 1987;10(3-4):237-46.
94. Vidal JD, Register TC, Gupta M, Cline JM. Estrogen replacement therapyinduces telomerase RNA expression in the macaque endometrium. Fertil Steril
69
2002;77(3):601-8.95. Utsunomiya H, Suzuki T, Kaneko C, Takeyama J, Nakamura J, Kimura K, et al.
The analyses of 17beta-hydroxysteroid dehydrogenase isozymes in humanendometrial hyperplasia and carcinoma. J Clin Endocrinol Metab2001;86(7):3436-43.
96. Miettinen MM, Mustonen MV, Poutanen MH, Isomaa VV, Vihko RK. Human17 beta-hydroxysteroid dehydrogenase type 1 and type 2 isoenzymes haveopposite activities in cultured cells and characteristic cell- and tissue-specificexpression. Biochem J 1996;314(Pt 3):839-45.
97. Mustonen MV, Isomaa VV, Vaskivuo T, Tapanainen J, Poutanen MH,Stenback F, et al. Human 17beta-hydroxysteroid dehydrogenase type 2messenger ribonucleic acid expression and localization in term placenta and inendometrium during the menstrual cycle. J Clin Endocrinol Metab1998;83(4):1319-24.
98. Casey ML, MacDonald PC, Andersson S. 17 beta-Hydroxysteroiddehydrogenase type 2: chromosomal assignment and progestin regulation ofgene expression in human endometrium. J Clin Invest 1994;94(5):2135-41.
99. Ferenzy A, Guralnic M. Endometrial microstructure: Structure-functionrelationship throughout the menstrual cycle. Semin Reprod Endocrinol1983;1(3):205-19.
100. Maentausta O, Boman K, Isomaa V, Stendahl U, Backstrom T, Vihko R.Immunohistochemical study of the human 17 beta-hydroxysteroiddehydrogenase and steroid receptors in endometrial adenocarcinoma. Cancer1992;70(6):1551-5.
101. Sasano H, Suzuki T, Takeyama J, Utsunomiya H, Ito K, Ariga N, et al. 17-beta-hydroxysteroid dehydrogenase in human breast and endometrial carcinoma. Anew development in intracrinology. Oncology 2000;59(Suppl 1):5-12.
102. Longcope C. Endocrine function of the postmenopausal ovary. J Soc GynecolInvestig 2001;8(1 Suppl Proceedings):S67-8.
103. Speroff L, Glass R, Kase N. Menopause and postmenopausal hormone therapy.In: C. Mitchell (ed.) Clinical gynecologic endocrinology and infertility. 5 ed.pp. 583-649. Baltimore: Williams & Wilkins; 1994.
104. Inoue M. Current molecular aspects of the carcinogenesis of the uterineendometrium. Int J Gynecol Cancer 2001;11(5):339-48.
105. Weiderpass E, Baron JA, Adami HO, Magnusson C, Lindgren A, Bergstrom R,et al. Low-potency oestrogen and risk of endometrial cancer: a case-controlstudy. Lancet 1999;353(9167):1824-8.
106. Weiderpass E, Adami HO, Baron JA, Magnusson C, Bergstrom R, Lindgren A,et al. Risk of endometrial cancer following estrogen replacement with andwithout progestins. J Natl Cancer Inst 1999;91(13):1131-7.
107. Stadberg E, Mattsson LA, Uvebrant M. 17 beta-estradiol and norethisteroneacetate in low doses as continuous combined hormone replacement therapy.Maturitas 1996;23(1):31-9.
108. van de Weijer PH, Scholten PC, van der Mooren MJ, Barentsen R, KenemansP. Bleeding patterns and endometrial histology during administration of low-
70
dose estradiol sequentially combined with dydrogesterone. Climacteric1999;2(2):101-9.
109. Wren BG, McFarland K, Edwards L, O'Shea P, Sufi S, Gross B, et al. Effect ofsequential transdermal progesterone cream on endometrium, bleeding pattern,and plasma progesterone and salivary progesterone levels in postmenopausalwomen. Climacteric 2000;3(3):155-60.
110. Brynhildsen J, Hammar M. Low dose transdermal estradiol/norethisteroneacetate treatment over 2 years does not cause endometrial proliferation inpostmenopausal women. Menopause 2002;9(2):137-44.
111. Bergeron C, Ferenczy A. Endometrial safety of continuous combined hormonereplacement therapy with 17beta-oestradiol (1 or 2 mg) and dydrogesterone.Maturitas 2001;37(3):191-9.
112. McLennan C, E, Rydell A, H. Extent of endometrial shedding during normalmenstruation. Obst Gynecol 1965;26:605-21.
113. Johannisson E, Holinka CF, Arrenbrecht S. Transdermal sequential andcontinuous hormone replacement regimens with estradiol and norethisteroneacetate in postmenopausal women: effects on the endometrium. Int J FertilWomens Med 1997;42(Suppl 2):388-98.
114. Moyer DL, Felix JC, Kurman RJ, Cuffie CA. Micronized progesteroneregulation of the endometrial glandular cycling pool. Int J Gynecol Pathol2001;20(4):374-9.
115. Oosterbaan HP, van Buuren AH, Schram JH, van Kempen PJ, Ubachs JM, vanLeusden HA, et al. The effects of continuous combined transdermal oestrogen-progestogen treatment on bleeding patterns and the endometrium inpostmenopausal women. Maturitas 1995;21(3):211-9.
116. Beato M, Klug J. Steroid hormone receptors: an update. Hum Reprod Update2000;6(3):225-36.
117. Taylor AH, Al-Azzawi F. Immunolocalisation of oestrogen receptor beta inhuman tissues. J Mol Endocrinol 2000;24(1):145-55.
118. Ariazi EA, Clark GM, Mertz JE. Estrogen-related receptor alpha and estrogen-related receptor gamma associate with unfavorable and favorable biomarkers,respectively, in human breast cancer. Cancer Res 2002;62(22):6510-8.
119. Witek A, Mazurek U, Paul M, Bierzynska-Macyszyn G, Wilczok T.Quantitative analysis of estrogen receptor mRNA in human endometriumthroughout the menstrual cycle using a real-time reverse transcription-polymerase chain reaction assay. Folia Histochem Cytobiol 2001;39 Suppl2:116-8.
120. Utsunomiya H, Suzuki T, Harada N, Ito K, Matsuzaki S, Konno R, et al.Analysis of estrogen receptor alpha and beta in endometrial carcinomas:correlation with ER beta and clinicopathologic findings in 45 cases. Int JGynecol Pathol 2000;19(4):335-41.
121. Matsuzaki S, Murakami T, Uehara S, Canis M, Sasano H, Okamura K.Expression of estrogen receptor alpha and beta in peritoneal and ovarianendometriosis. Fertil Steril 2001;75(6):1198-205.
71
122. Fujimoto J, Hirose R, Sakaguchi H, Tamaya T. Expression of oestrogenreceptor-alpha and -beta in ovarian endometriomata. Mol Hum Reprod1999;5(8):742-7.
123. Legler J, Zeinstra LM, Schuitemaker F, Lanser PH, Bogerd J, Brouwer A, et al.Comparison of in vivo and in vitro reporter gene assays for short-termscreening of estrogenic activity. Environ Sci Technol 2002;36(20):4410-5.
124. Coward P, Lee D, Hull MV, Lehmann JM. 4-Hydroxytamoxifen binds to anddeactivates the estrogen-related receptor gamma. Proc Natl Acad Sci U S A2001;98(15):8880-4.
125. Hawkins MB, Thornton JW, Crews D, Skipper JK, Dotte A, Thomas P.Identification of a third distinct estrogen receptor and reclassification ofestrogen receptors in teleosts. Proc Natl Acad Sci U S A 2000;97(20):10751-6.
126. Delmas PD. Treatment of postmenopausal osteoporosis. Lancet2002;359(9322):2018-26.
127. Pettersson K, Grandien K, Kuiper GG, Gustafsson JA. Mouse estrogen receptorbeta forms estrogen response element-binding heterodimers with estrogenreceptor alpha. Mol Endocrinol 1997;11(10):1486-96.
128. Pike AC, Brzozowski AM, Hubbard RE, Bonn T, Thorsell AG, Engstrom O, etal. Structure of the ligand-binding domain of oestrogen receptor beta in thepresence of a partial agonist and a full antagonist. Embo J 1999;18(17):4608-18.
129. Benson JR, Pitsinis V. Update on clinical role of tamoxifen. Curr Opin ObstetGynecol 2003;15(1):13-23.
130. Sakamoto T, Eguchi H, Omoto Y, Ayabe T, Mori H, Hayashi S. Estrogenreceptor-mediated effects of tamoxifen on human endometrial cancer cells. MolCell Endocrinol 2002;192(1-2):93-104.
131. Stackievicz R, Drucker L, Radnay J, Beyth Y, Yarkoni S, Cohen I. Tamoxifenmodulates apoptotic pathways in primary endometrial cell cultures. Clin CancerRes 2001;7(2):415-20.
132. Mourits MJ, Ten Hoor KA, van der Zee AG, Willemse PH, de Vries EG,Hollema H. The effects of tamoxifen on proliferation and steroid receptorexpression in postmenopausal endometrium. J Clin Pathol 2002;55(7):514-9.
133. Conneely OM, Mulac-Jericevic B, DeMayo F, Lydon JP, O'Malley BW.Reproductive functions of progesterone receptors. Recent Prog Horm Res2002;57:339-55.
134. Felix JC, Farahmand S. Endometrial glandular proliferation and estrogenreceptor content during the normal menstrual cycle. Contraception1997;55(1):19-22.
135. Bergeron C, Ferenzy A, Toft DO, Schneider W, Shyamala G.Immunocytochemical study of progesterone receptors in the humanendometrium during the menstrual cycle. Lab Invest 1988;59:862-69.
136. Matsuzaki S, Fukaya T, Suzuki T, Murakami T, Sasano H, Yajima A.Oestrogen receptor alpha and beta mRNA expression in human endometriumthroughout the menstrual cycle. Mol Hum Reprod 1999;5(6):559-64.
72
137. Mangal RK, Wiehle RD, Poindexter AN, 3rd, Weigel NL. Differentialexpression of uterine progesterone receptor forms A and B during the menstrualcycle. J Steroid Biochem Mol Biol 1997;63(4-6):195-202.
138. Karck U, Kommoss F. Does tamoxifen change oestrogen and progesteronereceptor expression in the endometrium and breast? Eur J Cancer 2000;36Suppl 4:S45-6.
139. Koshiyama M, Yoshida M, Takemura M, Yura Y, Matsushita K, Hayashi M, etal. Immunohistochemical analysis of distribution of estrogen receptors andprogesterone receptors in the postmenopausal endometrium. Acta ObstetGynecol Scand 1996;75(8):702-6.
140. Eckert RL, Katzenellenbogen BS. Human endometrial cells in primary tissueculture: modulation of the progesterone receptor level by natural and syntheticestrogens in vitro. J Clin Endocrinol Metab 1981;52(4):699-708.
141. West NB, Brenner RM. Progesterone-mediated suppression of estradiolreceptors in cynomolgus macaque cervix, endometrium and oviduct duringsequential estradiol-progesterone treatment. J Steroid Biochem 1985;22(1):29-37.
142. Kreitmann-Gimbal B, Bayard F, Nixon WE, Hodgen GD. Patterns of estrogenand progesterone receptors in monkey endometrium during the normalmenstrual cycle. Steroids 1980;35(4):471-9.
143. Punnonen R, Mattila J, Kuoppala T, Koivula T. DNA ploidy, cell proliferationand steroid hormone receptors in endometrial hyperplasia and earlyadenocarcinoma. J Cancer Res Clin Oncol 1993;119(7):426-9.
144. Bergeron C, Ferenzy A, Shyamala C. Distribution of Estrogen Receptors invarious cell types of normal, hyperplastic, and neoplastic human endometrialtissues. Lab Invest 1988;58(3):338-45.
145. Lindahl Pea. Prognostic value of flow cytometrical DNA-measurements instage I-II endometrial carcinoma: correlations with steroid receptorconcentration, myometrial invasion, and degree of differentiation. AnticancerRes 1987;7:791-98.
146. Geisinger KR, Marshall RB, Kute TE, Homesley HD. Correlation of female sexsteroid hormone receptors with histologic and ultrastructural differentiation inadenocarcinoma of the endometrium. Cancer 1986;58(7):1506-17.
147. Fukuda K, Mori M, Uchiyama M, Iwai K, Iwasaka T, Sugimori H. Prognosticsignificance of progesterone receptor immunohistochemistry in endometrialcarcinoma. Gynecol Oncol 1998;69(3):220-5.
148. Kohlberger PD, Kieback DG, Breitenecker F, Loesch A, Gitsch G, Kainz C, etal. Epithelial and stromal estrogen and progesterone receptor expression inendometrial cancer: true color computer-assisted image analysis versussubjective scoring. Gynecol Oncol 1997;64(3):404-10.
149. Lax SF, Pizer ES, Ronnett BM, Kurman RJ. Clear cell carcinoma of theendometrium is characterized by a distinctive profile of p53, Ki-67, estrogen,and progesterone receptor expression. Hum Pathol 1998;29(6):551-8.
150. Carcangiu ML, Chambers JT, Voynick IM, Pirro M, Schwartz PE.Immunohistochemical evaluation of estrogen and progesterone receptor content
73
in 183 patients with endometrial carcinoma. Part I: Clinical and histologiccorrelations. Am J Clin Pathol 1990;94(3):247-54.
151. Tornos C, Silva EG, el-Naggar A, Burke TW. Aggressive stage I grade 1endometrial carcinoma. Cancer 1992;70(4):790-8.
152. Creasman WT. Prognostic significance of hormone receptors in endometrialcancer. Cancer 1993;71(4 Suppl):1467-70.
153. Pertschuk LP, Masood S, Simone J, Feldman JG, Fruchter RG, Axiotis CA, etal. Estrogen receptor immunocytochemistry in endometrial carcinoma: aprognostic marker for survival. Gynecol Oncol 1996;63(1):28-33.
154. Grenman S, E, Klemi P, Toikkanen S, Irjala K, Laippala P, Vähä-Eskeli K, etal. Association of steroid hormone receptor content and flow cytometric DNAploidy in endometrial carcinoma. Ann Chir Gynecol 1994;83:10-14.
155. Zaino RJ, Clarke CL, Mortel R, Satyaswaroop PG. Heterogeneity ofprogesterone receptor distribution in human endometrial adenocarcinoma.Cancer Res 1988;48(7):1889-95.
156. Mortel R, Zaino R, Satyaswaroop PG. Heterogeneity and progesterone-receptordistribution in endometrial adenocarcinoma. Cancer 1984;53(1):113-6.
157. Charpin C, Andrac L, Habib MC, Vacheret H, Lavaut MN, Xerri L, et al.Immunocytochemical assays in human endometrial carcinomas: amultiparametric computerized analysis and comparison with nonmalignantchanges. Gynecol Oncol 1989;33(1):9-22.
158. Quinn MA, Cauchi M, Fortune D. Endometrial carcinoma: steroid receptors andresponse to medroxyprogesterone acetate. Gynecol Oncol 1985;21(3):314-9.
159. Runowicz CD, Nuchtern LM, Braunstein JD, Jones JG. Heterogeneity inhormone receptor status in primary and metastatic endometrial cancer. GynecolOncol 1990;38(3):437-41.
160. Friberg LG, Kullander S, Persijn JP, Korsten CB. On receptors for estrogens(E2) and androgens (DHT) in human endometrial carcinoma and ovariantumours. Acta Obstet Gynecol Scand 1978;57(3):261-4.
161. Tabibzadeh S. The signals and molecular pathways involved in humanmenstruation, a unique process of tissue destruction and remodelling. Molecularhuman reproduction 1996;2(2):77-92.
162. Ioffe OB, Papadimitriou JC, Drachenberg CB. Correlation of proliferationindices, apoptosis, and related oncogene expression (bcl-2 and c-erbB-2) andp53 in proliferative, hyperplastic, and malignant endometrium [see comments].Hum Pathol 1998;29(10):1150-9.
163. Nayak NR, Brenner RM. Vascular proliferation and vascular endothelial growthfactor expression in the rhesus macaque endometrium. J Clin Endocrinol Metab2002;87(4):1845-55.
164. Matsumoto Y, Iwasaka T, Yamasaki F, Sugimori H. Apoptosis and Ki-67expression in adenomyotic lesions and in the corresponding eutopicendometrium. Obstet Gynecol 1999;94(1):71-7.
74
165. Taguchi M, Kubota T, Aso T. Immunohistochemical localization of tenascinand ki-67 nuclear antigen in human endometrium throughout the normalmenstrual cycle. J Med Dent Sci 1999;46(1):7-12.
166. Vaskivuo TE, Stenback F, Karhumaa P, Risteli J, Dunkel L, Tapanainen JS.Apoptosis and apoptosis-related proteins in human endometrium. Mol CellEndocrinol 2000;165(1-2):75-83.
167. Parr M, B, Parr E, L. The implantation reaction. 2 ed. New York: Plenum Press;1989.
168. Smith SK. Regulation of angiogenesis in the endometrium. Trends EndocrinolMetab 2001;12(4):147-51.
169. Slayden OD, Rubin JS, Lacey DL, Brenner RM. Effects of keratinocyte growthfactor in the endometrium of rhesus macaques during the luteal-folliculartransition. J Clin Endocrinol Metab 2000;85(1):275-85.
170. Ferenczy A, Bertrand G, Gelfand MM. Proliferation kinetics of humanendometrium during the normal menstrual cycle. Am J Obstet Gynecol1979;133(8):859-67.
171. Kokawa K, Shikone T, Nakano R. Apoptosis in the human uterineEndometrium during the menstrual cycle. J Clin Endocrinol Metab1996;81(11):4144-47.
172. Richart R, Ferenczy A. Endometrial morphologic response to hormonalenviroment. Gynecol Oncol 1974;2:180-97.
173. Shikone T, Kokawa K, Yamoto M, Nakano R. Apoptosis of human ovary anduterine endometrium during the menstrual cycle. Horm Res 1997;48(suppl3):27-34.
174. Watanabe H, Kanzaki H, Narukawa S, Inoue T, Katsuragawa H, Kaneko Y, etal. Bcl-2 and Fas expression in eutopic and ectopic human endometrium duringthe menstrual cycle in relation to endometrial cell apoptosis. Am J ObstetGynecol 1997;176(2):360-8.
175. von Rango U, Classen-Linke I, Krusche CA, Beier HM. The receptiveendometrium is characterized by apoptosis in the glands. Hum Reprod1998;13(11):3177-89.
176. Gu Y, Jow G, M, Moulton B, C, Lee C, Sensibar JA, Park-Sarge O-K, et al.Apoptosis in decidual tissue regression and reorganization. Endocrinology1994;135:1272-79.
177. Moulton B, C. Transforming growth factor-beta stimulates endometrial stromaapoptosis in vitro. Endocrinology 1994;134:1055-60.
178. Rotello R, J, Lieberman R, C. Characterization of uterine epithelium apoptoticcell death kinetics and regulation by progesterone and RU 486. Am J Anat1992;140:449-56.
179. Sandow B, A, West N, B, Norman R, N, Brenner R, B. Hormonal control ofapoptosis in hamster uterine luminal epithelium. Am J Anat 1979;156:15-36.
180. Nawaz S, Lynch M, P, Galand P, Gerschenson L, E. Hormonal regulation ofcell death in rabbit uterine epithelium. Am J Pathol 1987;127:51-59.
75
181. Taga M, Sakakibara H, Suyama K, Ikeda M, Minaguchi H. Gene expression oftransforming growth factor-alpha in human endometrium duringdecidualization. J Assist Reprod Genet 1997;14(4):218-22.
182. Giudice LC, Mark SP, Irwin JC. Paracrine actions of insulin-like growth factorsand IGF binding protein-1 in non-pregnant human endometrium and at thedecidual-trophoblast interface. J Reprod Immunol 1998;39(1-2):133-48.
183. Osteen KG, Keller NR, Feltus FA, Melner MH. Paracrine regulation of matrixmetalloproteinase expression in the normal human endometrium. GynecolObstet Invest 1999;48(Suppl 1):2-13.
184. Rotello RJ, Lieberman RC, Purchio AF, Gerschenson LE. Coordinatedregulation of apoptosis and cell proliferation by transforming growth factor beta1 in cultured uterine epithelial cells. Proc Natl Acad Sci U S A1991;88(8):3412-5.
185. Tabidzadeh S, Zupi E, Babaknia A, Liu R, Marconi D, Romanini C. Site andmenstrual cycle-dependent expression of the proteins of tumor necrosis factor(TNF) receptor family, and BCL-2 oncoprotein and phase specific productionof TNF alfa in human endometrium. Hum Reprod 1995;10:277-86.
186. Nayak NR, Critchley HO, Slayden OD, Menrad A, Chwalisz K, Baird DT, et al.Progesterone withdrawal up-regulates vascular endothelial growth factorreceptor type 2 in the superficial zone stroma of the human and macaqueendometrium: potential relevance to menstruation. J Clin Endocrinol Metab2000;85(9):3442-52.
187. Hirai M, Nakagawara A, Oosaki T, Hayashi Y, Hirono M, Yoshihara T.Expression of vascular endothelial growth factors (VEGF-A/VEGF-1 andVEGF-C/VEGF-2) in postmenopausal uterine endometrial carcinoma. GynecolOncol 2001;80(2):181-8.
188. Bermont L, Lamielle F, Fauconnet S, Esumi H, Weisz A, Adessi GL.Regulation of vascular endothelial growth factor expression by insulin- likegrowth factor-I in endometrial adenocarcinoma cells. Int J Cancer2000;85(1):117-23.
189. Bermont L, Lamielle F, Lorchel F, Fauconnet S, Esumi H, Weisz A, et al.Insulin up-regulates vascular endothelial growth factor and stabilizes itsmessengers in endometrial adenocarcinoma cells. J Clin Endocrinol Metab2001;86(1):363-8.
190. Markee J, E. Menstruation in intraocularendometrial transplants in Rhesusmonkey. Contrib Embryol 1940;28:219.
191. Marsh MM, Findlay JK, Salamonsen LA. Endothelin and menstruation. HumReprod 1996;11 Suppl 2:83-9.
192. Baird DT, Cameron ST, Critchley HO, Drudy TA, Howe A, Jones RL, et al.Prostaglandins and menstruation. Eur J Obstet Gynecol Reprod Biol1996;70(1):15-7.
193. Salamonsen LA, Kovacs GT, Findlay JK. Current concepts of the mechanismsof menstruation. Baillieres Best Pract Res Clin Obstet Gynaecol1999;13(2):161-79.
76
194. Reynolds S, R, M. The physiologic basis of menstruation: a summary of currentconcepts. J Am Med Ass 1947;1:552-57.
195. Bartelmez G, W. The phases of the menstrual cycle and their interpretations interms of the pregnancy cycle. Am J Obstet Gynecol 1957;74:931-55.
196. Israel R, Misheli DR, Labudovich M. Mechanism of normal and dysfunctionaluterine bleeding. Clinical Obstetrics and Gynecology 1970;13:386 - 99.
197. Davie R, Hopwood D, A. Intercellular spaces and cell junctions in endometrialglands: their possible role in menstruation. Br J Obstet Gynecol 1977;84:467-76.
198. Nogales-Ortis F, Puerta J. The normal menstrual cycle. Chronology andmechanism of endometrial desquamation. Obstet Gynecol 1978;51:259-64.
199. Christiaens G, C, M, L, Sixma J, J, Haspels A, A. Morphology of haemostasesin menstrual endometrium. Br J Obstet Gynecol 1980;87:425-39.
200. Henzl MR, Smith RE, Boost G, Tyler ET. Lysosomal concept of menstrualbleeding in humans. J Clin Endocrinol Metab 1972;34(5):860-75.
201. Ferenzy A. Regeneration of the human endometrium. New York: Masson Publ.;1980.
202. Salamonsen LA, Woolley DE. Matrix metalloproteinases in normalmenstruation. Hum Reprod 1996;11 Suppl 2:124-33.
203. Tabibzadeh S, Sun XZ. Cytokine expression in human endometrium throughoutthe menstrual cycle. Hum Reprod 1992;7(9):1214-21.
204. Pike MC, Peters RK, Cozen W, Probst-Hensch NM, Felix JC, Wan PC, et al.Estrogen-progestin replacement therapy and endometrial cancer. J Natl CancerInst 1997;89(15):1110-6.
205. Beresford SA, Weiss NS, Voigt LF, McKnight B. Risk of endometrial cancer inrelation to use of oestrogen combined with cyclic progestagen therapy inpostmenopausal women. Lancet 1997;349(9050):458-61.
206. Gadducci A, Fanucchi A, Cosio S, Genazzani AR. Hormone replacementtherapy and gynecological cancer. Anticancer Res 1997;17(5B):3793-8.
207. Pukkala E, Tulenheimo-Silfvast A, Leminen A. Incidence of cancer amongwomen using long versus monthly cycle hormonal replacement therapy,Finland 1994-1997. Cancer Causes Control 2001;12(2):111-5.
208. Darj E, Axelsson O, Nilsson G, Nilsson S, Risberg B. Ki-67 immunostaining ofendometrial biopsies with special reference to hormone replacement therapy.Gynecol Obstet Invest 1995;39(2):120-4.
209. Cameron ST, Critchley HO, Glasier AF, Williams AR, Baird DT. Continuoustransdermal oestrogen and interrupted progestogen as a novel bleed-freeregimen of hormone replacement therapy for postmenopausal women. Br JObstet Gynaecol 1997;104(10):1184-90.
210. Odmark IS, Jonsson B, Backstrom T. Bleeding patterns in postmenopausalwomen using continuous combination hormone replacement therapy withconjugated estrogen and medroxyprogesterone acetate or with 17beta-estradioland norethindrone acetate. Am J Obstet Gynecol 2001;184(6):1131-8.
77
211. Bjorn I, Backsrom T. Drug related negative side-effects is a common reason forpoor compliance in hormone replacement therapy. Maturitas 1999;32(2):77-86.
212. Hampton AL, Salamonsen LA. Expression of messenger ribonucleic acidencoding matrix metalloproteinases and their tissue inhibitors is related tomenstruation. J Endocrinol 1994;141(1):R1-3.
213. Hickey M, Dwarte D, Fraser IS. Superficial endometrial vascular fragility inNorplant users and in women with ovulatory dysfunctional uterine bleeding.Hum Reprod 2000;15(7):1509-14.
214. Hickey M, Higham J, Sullivan M, Miles L, Fraser IS. Endometrial bleeding inhormone replacement therapy users: preliminary findings regarding the role ofmatrix metalloproteinase 9 (MMP-9) and tissue inhibitors of MMPs. FertilSteril 2001;75(2):288-96.
215. Vincent AJ, Zhang J, Ostor A, Rogers PA, Affandi B, Kovacs G, et al. Matrixmetalloproteinase-1 and -3 and mast cells are present in the endometrium ofwomen using progestin-only contraceptives. Hum Reprod 2000;15(1):123-30.
216. Salamonsen LA, Zhang J, Hampton A, Lathbury L. Regulation of matrixmetalloproteinases in human endometrium. Hum Reprod 2000;15 Suppl 3:112-9.
217. Salamonsen LA, Zhang J, Brasted M. Leukocyte networks and humanendometrial remodelling. J Reprod Immunol 2002;57(1-2):95-108.
218. Socialstyrelsen. Cancer Incidence in Sweden 1998. 1998;219. Abeler VM, Kjorstad KE, Berle E. Carcinoma of the endometrium in Norway: a
histopathological and prognostic survey of a total population. Int J GynecolCancer 1992;2(1):9-22.
220. Salvesen HB, Akslen LA. Molecular pathogenesis and prognostic factors inendometrial carcinoma. Apmis 2002;110(10):673-89.
221. Creasman WT, Soper JT, McCarty KS, Jr., McCarty KS, Sr., Hinshaw W,Clarke-Pearson DL. Influence of cytoplasmic steroid receptor content onprognosis of early stage endometrial carcinoma. Am J Obstet Gynecol1985;151(7):922-32.
222. Vecek N, Nola M, Marusic M, Ilic J, Babic D, Petrovecki M, et al. Prognosticvalue of steroid hormone receptors concentration in patients with endometrialcarcinoma. Acta Obstet Gynecol Scand 1994;73(9):730-3.
223. Nordström B, Strang P, Lindgren A, Bergström B, Tribukait B. Endometrialcarcinoma: the prognostic impact of papillary serous carcinoma (UPSC) inrelation to nuclear grade, DNA ploidy and p53 expression. Anticancer research1996;18:899-904.
224. Salvesen HB, Iversen OE, Akslen LA. Prognostic significance of angiogenesisand Ki-67, p53, and p21 expression: a population-based endometrial carcinomastudy. J Clin Oncol 1999;17(5):1382-90.
225. Cherchi PL, Marras V, Capobianco G, Ambrosini G, Piga MD, Fadda GM, etal. Prognostic value of p53, c-erb-B2 and MIB-1 in endometrial carcinoma. EurJ Gynaecol Oncol 2001;22(6):451-3.
78
226. Geisler JP, Geisler HE, Wiemann MC, Zhou Z, Miller GA, Crabtree W. Lack ofbcl-2 persistence: an independent prognostic indicator of poor prognosis inendometrial carcinoma. Gynecol Oncol 1998;71(2):305-7.
227. Geisler JP, Geisler HE, Miller GA, Wiemann MC, Zhou Z, Crabtree W. MIB-1in endometrial carcinoma: prognostic significance with 5-year follow-up.Gynecol Oncol 1999;75(3):432-6.
228. Westhoff C, Heller D, Drosinos S, Tancer L. Risk factors for hyperplasia-associated versus atrophy-associated endometrial carcinoma. Am J ObstetGynecol 2000;182(3):506-8.
229. Boyd J. Molecular biology in the clinicopathologic assessment of endometrialcarcinoma subtypes [editorial; comment]. Gynecol Oncol 1996;61(2):163-5.
230. Lynch HT, Lanspa S, Smyrk T, Boman B, Watson P, Lynch J. Hereditarynonpolyposis colorectal cancer (Lynch syndromes I & II). Genetics, pathology,natural history, and cancer control, Part I. Cancer Genet Cytogenet1991;53(2):143-60.
231. Vogelstein B, Fearon ER, Hamilton SR, Kern SE, Preisinger AC, Leppert M, etal. Genetic alterations during colorectal-tumor development. N Engl J Med1988;319(9):525-32.
232. Ignar-Trowbridge D, Risinger JI, Dent GA, Kohler M, Berchuck A, McLachlanJA, et al. Mutations of the Ki-ras oncogene in endometrial carcinoma. Am JObstet Gynecol 1992;167(1):227-32.
233. Boyd J, Risinger JI. Analysis of oncogene alterations in human endometrialcarcinoma: prevalence of ras mutations. Mol Carcinog 1991;4(3):189-95.
234. Enomoto T, Fujita M, Inoue M, Nomura T, Shroyer KR. Alteration of the p53tumor suppressor gene and activation of c-K-ras-2 protooncogene inendometrial adenocarcinoma from Colorado. Am J Clin Pathol1995;103(2):224-30.
235. Enomoto T, Fujita M, Inoue M, Rice JM, Nakajima R, Tanizawa O, et al.Alterations of the p53 tumor suppressor gene and its association with activationof the c-K-ras-2 protooncogene in premalignant and malignant lesions of thehuman uterine endometrium. Cancer Res 1993;53(8):1883-8.
236. Yamauchi N, Sakamoto A, Uozaki H, Iihara K, Machinami R.Immunohistochemical analysis of endometrial adenocarcinoma for bcl-2 andp53 in relation to expression of sex steroid receptor and proliferative activity[published erratum appears in Int J Gynecol Pathol 1996 Oct;15(4):369]. Int JGynecol Pathol 1996;15(3):202-8.
237. Bandera CA, Boyd J. The molecular genetics of endometrial carcinoma. ProgClin Biol Res 1997;396:185-203.
238. Berchuck A, Boyd J. Molecular basis of endometrial cancer. Cancer 1995;76(10Suppl):2034-40.
239. Satyaswaroop PG, Mortel R. Endometrial carcinoma: an aberration ofendometrial cell differentiation. Am J Obstet Gynecol 1981;140(6):620-3.
79
240. Okamoto A, Sameshima Y, Yamada Y, Teshima S, Terashima Y, Terada M, etal. Allelic loss on chromosome 17p and p53 mutations in human endometrialcarcinoma of the uterus. Cancer Res 1991;51(20):5632-5.
241. Risinger JI, Dent GA, Ignar-Trowbridge D, McLachlan JA, Tsao MS,Senterman M, et al. p53 gene mutations in human endometrial carcinoma. MolCarcinog 1992;5(4):250-3.
242. Naito M, Satake M, Sakai E, Hirano Y, Tsuchida N, Kanzaki H, et al. Detectionof p53 gene mutations in human ovarian and endometrial cancers bypolymerase chain reaction-single strand conformation polymorphism analysis.Jpn J Cancer Res 1992;83(10):1030-6.
243. Honda T, Kato H, Imamura T, Gima T, Nishida J, Sasaki M, et al. Involvementof p53 gene mutations in human endometrial carcinomas. Int J Cancer1993;53(6):963-7.
244. Kohler MF, Nishii H, Humphrey PA, Saski H, Marks J, Bast RC, et al.Mutation of the p53 tumor-suppressor gene is not a feature of endometrialhyperplasias. Am J Obstet Gynecol 1993;169(3):690-4.
245. Risinger JI, Umar A, Boyd J, Berchuck A, Kunkel TA, Barrett JC. Mutation ofMSH3 in endometrial cancer and evidence for its functional role inheteroduplex repair. Nat Genet 1996;14(1):102-5.
246. Risinger JI, Berchuck A, Kohler MF, Watson P, Lynch HT, Boyd J. Geneticinstability of microsatellites in endometrial carcinoma. Cancer Res1993;53(21):5100-3.
247. Kyushima N, Watanabe J, Hata H, Jobo T, Kameya T, Kuramoto H. Expressionof cyclin A in endometrial adenocarcinoma and its correlation with proliferativeactivity and clinicopathological variables. J Cancer Res Clin Oncol2002;128(6):307-12.
248. Nielsen AL, Nyholm HC. Proliferative activity as revealed by Ki-67 in uterineadenocarcinoma of endometrioid type: comparison of tumours from patientswith and without previous oestrogen therapy. J Pathol 1993;171(3):199-205.
249. Heathley M, K. Asociation between the apoptotic index and establishedprognostic parameters in endometrial carcinoma. Histopathology 1995;27:469-72.
250. Takai N, Miyazaki T, Nishida M, Nasu K, Miyakawa I. Survivin expressioncorrelates with clinical stage, histological grade, invasive behavior and survivalrate in endometrial carcinoma. Cancer Lett 2002;184(1):105-16.
251. Peiro G, Diebold J, Baretton GB, Kimmig R, Lohrs U. Cellular apoptosissusceptibility gene expression in endometrial carcinoma: correlation with Bcl-2,Bax, and caspase-3 expression and outcome. Int J Gynecol Pathol2001;20(4):359-67.
252. Crescenzi E, Criniti V, Pianese M, Tecce MF, Palumbo G. Differentialexpression of antiapoptotic genes in human endometrial carcinoma: bcl-XLsucceeds bcl-2 function in neoplastic cells. Gynecol Oncol 2000;77(3):419-28.
253. Kelley RM, Baker WH. The role of progesterone in human endometrial cancer.Cancer Res 1965;25(7):1190-2.
80
254. Podczaski E, Mortel R. Hormonal treatment of endometrial cancer: past, presentand future. Best Pract Res Clin Obstet Gynaecol 2001;15(3):469-89.
255. Podratz K, C, O´Brien P, C, Malkasian G, D, Decker D, G, Jefferies J, A,Edmonsson J, H. Effects of progestational agents in treatment of endometrialcarcinoma. Obstet Gynecol 1985;66(suppl. 1):106-10.
256. Podratz K, C. Hormonal therapy in endometrial carcinoma. Recent Results inCancer Research 1990;118:242-51.
257. Satyaswaroop PG, Zaino R, Clarke CL, Mortel R. Nude mouse system in thestudy of tumor biology, treatment strategies and progesterone receptorphysiology in human endometrial carcinoma. J Steroid Biochem 1987;27(1-3):431-8.
258. Murphy LC, Dotzlaw H, Alkhalaf M, Coutts A, Miller T, Wong MS, et al.Mechanisms of growth inhibition by antiestrogens and progestins in humanbreast and endometrial cancer cells. J Steroid Biochem Mol Biol 1992;43(1-3):117-21.
259. Piver M, S, Barlow J, J, Lurain J, R, Blumenson L, E. Medroxuprogesteroneacetate (Depoprovera) versus hydroxyprogesterone caproate (Delalutin) inwomen with metastatic endometrial adenocarcinoma. Cancer 1980;45:268-72.
260. Martin PM, Rolland PH, Gammerre M, Serment H, Toga M. Estradiol andprogesterone receptors in normal and neoplastic endometrium: correlationsbetween receptors, histopathological examinations and clinical responses underprogestin therapy. Int J Cancer 1979;23(3):321-9.
261. Kauppila A, Kujansuu E, Vihko R. Cytosol estrogen and progestin receptors inendometrial carcinoma of patients treated with surgery, radiotherapy, andprogestin. Clinical correlates. Cancer 1982;50(10):2157-62.
262. Farhi DC, Nosanchuk J, Silverberg SG. Endometrial adenocarcinoma in womenunder 25 years of age. Obstet Gynecol 1986;68(6):741-5.
263. Kim YB, Holschneider CH, Ghosh K, Nieberg RK, Montz FJ. Progestin aloneas primary treatment of endometrial carcinoma in premenopausal women.Report of seven cases and review of the literature. Cancer 1997;79(2):320-7.
264. Randall TC, Kurman RJ. Progestin treatment of atypical hyperplasia and well-differentiated carcinoma of the endometrium in women under age 40. ObstetGynecol 1997;90(3):434-40.
265. Urbanski K, Karolewski K, Kojs Z, Klimek M, Dyba T. Adjuvant progestagentherapy improves survival in patients with endometrial cancer afterhysterectomy. Results of one-institutional prospective clinical trial. Eur JGynaecol Oncol 1993;14(Suppl):98-104.
266. Lewis GC, Jr., Slack NH, Mortel R, Bross ID. Adjuvant progestogen therapy inthe primary definitive treatment of endometrial cancer. Gynecol Oncol1974;2(2-3):368-76.
267. Macdonald RR, Thorogood J, Mason MK. A randomized trial of progestogensin the primary treatment of endometrial carcinoma. Br J Obstet Gynaecol1988;95(2):166-74.
81
268. Vergote I, Kjorstad K, Abeler V, Kolstad P. A randomized trial of adjuvantprogestagen in early endometrial cancer. Cancer 1989;64(5):1011-6.
269. FIGO. Announcements. FIGO Stages - 1988 Revision. Gynecol Oncol1989;35:125-27.
270. Whitehead MI, Townsend PT, Pryse-Davies J, Ryder T, Lane G, Siddle NC, etal. Effects of Various Types and Dosages of Progestogens on thePostmenopausal Endometrium. The Journal of Reproductive medicine1982;27(8):538-48.
271. Satyaswaroop PG, Mortel R. Sex steroid receptors in endometrial carcinoma[editorial]. Gynecol Oncol 1993;50(3):278-80.
272. Thigpen JT, Brady MF, Alvarez RD, Adelson MD, Homesley HD, Manetta A,et al. Oral medroxyprogesterone acetate in the treatment of advanced or recurrent endometrial carcinoma: a dose-response study by the GynecologicOncology Group. J Clin Oncol 1999;17(6):1736-44.
273. McCarty KS, Jr., Barton TK, Fetter BF, Creasman WT, McCarty KS, Sr.Correlation of estrogen and progesterone receptors with histologicdifferentiation in endometrial adenocarcinoma. Am J Pathol 1979;96(1):171-83.
274. Mourits MJ, Hollema H, De Vries EG, Ten Hoor KA, Willemse PH, Van DerZee AG. Apoptosis and apoptosis-associated parameters in relation totamoxifen exposure in postmenopausal endometrium. Hum Pathol2002;33(3):341-6.
275. Heatley MK. A high apoptotic index occurs in subtypes of endometrialadenocarcinoma associated with a poor prognosis. Pathology 1997;29(3):272-5.
276. Lax SF, Pizer ES, Ronnett BM, Kurman RJ. Comparison of estrogen andprogesterone receptor, Ki-67, and p53 immunoreactivity in uterineendometrioid carcinoma and endometrioid carcinoma with squamous,mucinous, secretory, and ciliated cell differentiation. Hum Pathol1998;29(9):924-31.
277. Saegusa M, Okayasu I. Progesterone therapy for endometrial carcinoma reducescell proliferation but does not alter apoptosis. Cancer 1997;83:111-21.
278. Amezcua CA, Lu JJ, Felix JC, Stanczyk FZ, Zheng W. Apoptosis may be anearly event of progestin therapy for endometrial hyperplasia. Gynecol Oncol2000;79(2):169-76.
279. Nielsen A, L, Nyholm H, C, J, Engel P. Expression of MIB-1 (paraffin ki-67)and AgNOR morphology in endometrial adenocarcinomas of endometriod type.Int J Gynecol Pathol 1994;13:37-44.
280. Gassel AM, Backe J, Krebs S, Schon S, Caffier H, Muller-Hermelink HK.Endometrial carcinoma: immunohistochemically detected proliferation index isa prognosticator of long-term outcome. J Clin Pathol 1998;51(1):25-9.
281. Scholten AN, Creutzberg CL, Noordijk EM, Smit VT. Long-term outcome inendometrial carcinoma favors a two- instead of a three-tiered grading system.Int J Radiat Oncol Biol Phys 2002;52(4):1067-74.
82
282. Arnold JT, Lessey BA, Seppala M, Kaufman DG. Effect of normal endometrialstroma on growth and differentiation in Ishikawa endometrial adenocarcinomacells. Cancer Res 2002;62(1):79-88.
283. Gee JMW, Robertson JFR, Ellis IO, Willsher P, McCelland RA, Hoyle HB, etal. Immunocytochemical localization of bcl-2 protein in human breast cancersand its relationship to a series of prognostic markers and response to endocrinetherapy. Int J Cancer 1994;59:619-28.
