Benign Diseases of the Endometrium

  • Ricardo R. LastraEmail author
  • W. Glenn McCluggage
  • Lora Hedrick Ellenson
Living reference work entry


The endometrium and the myometrium are of mesodermal origin and are formed secondary to fusion of the müllerian (paramesonephric) ducts between the 8th and 9th postovulatory weeks. Likewise, the cervix is of müllerian origin. The squamous epithelial lining of the ectocervix and upper two-thirds of the vagina (müllerian vagina) are of vaginal müllerian origin, while the squamous epithelial lining of the lower third of the vagina (sinus vagina) develops from the urogenital sinus; the endocervical glandular epithelium has recently been shown to be of uterine müllerian origin (Fluhmann 1960; Fritsch et al. 2013). Until the 20th week of gestation, the endometrium consists of a single layer of columnar epithelium supported by a thick layer of fibroblastic stroma. After the 20th gestational week, the surface epithelium invaginates into the underlying stroma, forming glandular structures that extend toward the underlying myometrium. The uterus, which is made up of the uterine corpus and uterine cervix, measures approximately 4 cm in length at birth, and in newborns, the cervix makes up the majority of the uterus. At this stage, the endometrium measures less than 0.5 mm in thickness, and the surface and glands are lined by a low columnar to cuboidal epithelium devoid of either proliferative or secretory activity, which resembles the inactive endometrium of postmenopausal women.

Embryology and Anatomy

The endometrium and the myometrium are of mesodermal origin and are formed secondary to fusion of the müllerian (paramesonephric) ducts between the 8th and 9th postovulatory weeks. Likewise, the cervix is of müllerian origin. The squamous epithelial lining of the ectocervix and upper two-thirds of the vagina (müllerian vagina) are of vaginal müllerian origin, while the squamous epithelial lining of the lower third of the vagina (sinus vagina) develops from the urogenital sinus; the endocervical glandular epithelium has recently been shown to be of uterine müllerian origin (Fluhmann 1960; Fritsch et al. 2013). Until the 20th week of gestation, the endometrium consists of a single layer of columnar epithelium supported by a thick layer of fibroblastic stroma. After the 20th gestational week, the surface epithelium invaginates into the underlying stroma, forming glandular structures that extend toward the underlying myometrium. The uterus, which is made up of the uterine corpus and uterine cervix, measures approximately 4 cm in length at birth, and in newborns, the cervix makes up the majority of the uterus. At this stage, the endometrium measures less than 0.5 mm in thickness, and the surface and glands are lined by a low columnar to cuboidal epithelium devoid of either proliferative or secretory activity, which resembles the inactive endometrium of postmenopausal women.

During the prepubertal years, the endometrium remains inactive, and the cervix continues to comprise the major part of the uterus. In the reproductive years, the dimensions and weight of a normal uterus vary widely according to parity. In nulliparous women, the uterus measures approximately 8 cm in length, 5 cm in width at the level of the fundus, and 2.5 cm in thickness; most weigh between 40 and 100 g. Multigravid uteri are larger with increasing length and weight with increasing parity. The internal os, a fibromuscular junction, separates the muscular uterine corpus from the fibrous uterine cervix. The uterine corpus is divided into the fundus, body, and isthmus. The fundus is that part of the uterus above the orifices of the fallopian tubes, and the isthmus represents the lower uterine segment. The uterus is located between the rectum (posteriorly) and the urinary bladder (anteriorly); it is supported by the round ligaments and the utero-ovarian ligaments and covered by the pelvic peritoneum. The endometrium during the reproductive period undergoes cyclical morphological changes (described in detail below), which are particularly evident in the superficial two-thirds, the so-called functionalis layer. Morphological alterations are minimal in the deeper one-third, the so-called basalis layer. In postmenopausal women, the endometrial morphology is similar to that in the prepubertal years (see section “Postmenopausal Endometrium”).

Vascular Anatomy

The endometrium has an abundant vascular supply that originates from the radial arteries of the underlying myometrium. These arteries penetrate the endometrium at regular intervals and give rise to the basal arteries, which in turn divide into horizontal and vertical branches, the former providing the blood supply to the endometrial basalis and the latter to the overlying functionalis layer. The endometrial vessels in the functionalis layer are referred to as spiral arteries. Their development and arborization near the endometrial surface and their connections with the subsurface epithelial precapillary system, as well as extreme coiling during the menstrual cycle, are influenced by ovarian steroid hormones and prostaglandins.

A differentiating feature between the endometrial and myometrial arteries is the absence of subendothelial elastic tissue in the endometrial arteries, except for those in the basal layer, and its presence in the myometrial arteries. Veins and lymphatics are closely associated with the endometrial arteries and glands, respectively. Uterine lymphatics drain from subserosal uterine plexuses to the pelvic and para-aortic lymph nodes.

Congenital Defects

Congenital abnormalities of the uterus are uncommon. They may be secondary to the in utero effects of exogenous hormones, such as diethylstilbestrol (DES) (Kaufman et al. 1977), or imbalances in endogenous hormones associated with abnormal gonads and chromosomal defects. In utero exposure to DES is, of course, almost currently nonexistent. Genotypically normal females with normal gonads may also have müllerian duct abnormalities. These developmental aberrations, such as defects in the fusion of the müllerian ducts, are caused by errors in embryogenesis. The etiology of these developmental errors is mostly unknown, but hormonal imbalances or genetic abnormalities may be implicated. These disorders are frequently associated with malformations in the urinary system and the distal gastrointestinal tract. For practical purposes, müllerian duct abnormalities can be divided into two categories, namely, abnormalities of fusion and abnormalities caused by atresia.

Fusion Defects of the Müllerian Ducts

Normally, the upper two-thirds of the vagina and the uterus are formed by fusion of the paired müllerian ducts. After fusion, the intervening wall degenerates, forming the endometrial cavity and the upper vaginal canal. Nonfusion of the müllerian ducts results in a bicornuate uterus. If the ducts fuse but the wall between the two lumens persists, an abnormal septate uterus results. Occasionally, a carcinoma develops in one cavity, and only the other normal cavity is sampled during the investigation of abnormal uterine bleeding. If the defect is minor or confined to the fundus, the uterus is referred to as arcuatus. If the full length of the uterus and the upper vagina is divided by a septum, the condition results in uterus didelphys with a partially double vagina. These congenital anomalies may result in infertility or spontaneous abortion and in some cases require surgical correction.

Atresia of the Müllerian Ducts and the Vagina

Atresia of the müllerian ducts and the vagina may be partial or complete. The etiology of these conditions is obscure, although a genetic cause is suggested in families with multiple affected siblings. The pattern of inheritance may be autosomal recessive or dominant. In cases of bilateral müllerian duct atresia, the upper genital tract may consist of bilateral non-canalized muscular tissue located on the lateral pelvic walls. In Mayer–Rokitansky–Küster–Hauser syndrome, a severe defect characterized by müllerian and vaginal aplasia, patients may have urinary tract anomalies such as a pelvic kidney or anephria. Vertebral and other skeletal abnormalities may also be present, suggesting a more generalized morphogenetic abnormality. If only one of the müllerian ducts is involved, the affected side will show a portion of tubal fimbria and a small muscular mass at the pelvic sidewall. Occasionally, a rudimentary structure remains as an appendage attached to the unaffected side, giving rise to a uterus bicornis unicollis, which results in a pelvic mass and cyclic pelvic pain associated with menses. Patients with these conditions are endocrinologically normal with normal gonads. It has been postulated that activating mutations affecting the gene coding for antimüllerian hormone or its receptor may be related to the development of these syndromes (Lindenman et al. 1997). If the anomaly is associated with obstruction of the vagina and functional endometrial tissue is present, hydrocolpos may be present at birth, or patients may present with primary amenorrhea. A number of multiple malformation syndromes have been associated with müllerian or vaginal agenesis. Winter syndrome, a genetically inherited autosomal recessive disorder, is characterized by vaginal agenesis, renal agenesis, and middle ear anomalies (Winter et al. 1968). Management of patients with complete vaginal atresia requires surgery to create a neovagina. If the anomaly is isolated vaginal atresia, as most commonly occurs, the patient will usually be fertile if a normal uterus and fallopian tubes are present.

Normal Cyclical Endometrium

In the reproductive years, the endometrium is characterized by cyclical proliferation, differentiation, and shedding in response to estrogen and progesterone secretion by the ovaries. Endometrial morphology, as a consequence, is continually changing depending on the levels of estrogen and progesterone (Crum et al. 2003; Noyes et al. 1950). During the proliferative phase of the menstrual cycle, the endometrium has a relatively constant morphology, which does not differ significantly from day to day; as such, accurate dating is not possible in the proliferative phase. Following ovulation, the morphological appearances in the secretory phase have been considered relatively specific from day to day, such that it is possible to accurately date secretory phase endometrium to within 1 or 2 days. However, this view has been challenged with one study finding that traditional endometrial histological dating criteria are much less temporally distinct and discriminating than originally described (Murray et al. 2004). Presently, most endometrial biopsies are performed during the investigation of abnormal uterine bleeding, and it is relatively uncommon to be asked to date the endometrium, although formerly this was often requested in the investigation of infertility. The typical endometrial cycle is 28 days, although the length varies in individual women and between women. In general, the differences in cycle length are due to variation in the duration of the proliferative phase, the secretory phase usually being constant and lasting 14 days from the time of ovulation to the onset of menstruation. In the reproductive years, the endometrium is divided into two regions, namely, the superficial functionalis (stratum spongiosum) and the basalis (stratum basale). The former exhibits the greatest degree of hormonal responsiveness, while the latter is much more unresponsive, the morphology not varying greatly during the menstrual cycle; as such, a biopsy consisting entirely of basalis is not adequate for dating of the menstrual cycle. Usually the endometrial glands are regularly spaced and have a perpendicular arrangement from the basalis to the surface. The basalis abuts the myometrium and regenerates the functionalis following its shedding during menstruation. The basalis is composed of inactive appearing glands, cellular stroma, and spiral arteries that have thicker muscular walls than those in the functionalis. Accurate typing of the endometrium is, in general, not possible when polyps, endometritis, or other pathological lesions are present.

Proliferative Phase

The onset of menstruation is the first day of the menstrual cycle. Following menstruation, which varies in length, the uterus is lined by shallow basal endometrium and the deeper part of the functionalis. The endometrium begins to proliferate on the third or fourth day of the cycle, and during the proliferative phase, it increases in thickness up to 4 or 5 mm. Between the 5th and 14th days of a typical 28-day cycle, there is glandular, stromal, and vascular growth, with the endometrium progressively increasing in depth until ovulation. The endometrial glands are uniform and widely and regularly spaced, and have a simple tubular architecture which can be appreciated on cross section (Fig. 1). An occasional mildly dilated gland is a normal feature and of no significance. Mitotic figures are easily identified within the glands, and the presence of mitotic activity should be confirmed before labeling an endometrium as proliferative in type. The glandular epithelium is composed of pseudostratified cuboidal or low columnar cells with a moderate amount of basophilic cytoplasm. The nuclei are oval or rounded, may contain small nucleoli, and are orientated perpendicular to the basement membrane. Estrogenic activity during the proliferative phase often results in focal ciliation of the surface epithelial cells; thus, surface ciliated cells are a feature of normal proliferative endometrium, and this does not indicate ciliated or tubal metaplasia (see section “Endometrial Epithelial Metaplasia (Epithelial Cytoplasmic Change)”).
Fig. 1

Proliferative endometrium. Widely spaced tubular glands with low columnar cells, exhibiting mitotic activity

Proliferative activity is maximal between the 8th and 10th days of the cycle, and at this stage, the glandular epithelium becomes more stratified and mitoses are more frequently seen. In the late-proliferative phase, the glands become progressively more convoluted and tortuous and appear more variable in size and shape; however, they remain tubular in configuration. Occasional subnuclear vacuoles may be seen. During the proliferative phase, the endometrial stroma is usually densely cellular, and the stromal cells are small and oval with hyperchromatic nuclei and indistinct cytoplasm and cell borders. Mitotic figures are present within the stroma, although less numerous than within the glands. Scanty thin-walled stromal blood vessels are present. Lymphoid aggregates resembling follicles can be seen within the stroma in the proliferative phase of the cycle. The degree of mitotic activity within both the glands and stroma decreases in the late-proliferative phase; simultaneously, early stromal edema develops.

Secretory Phase

Secretory endometrium is characterized by glandular secretion, stromal maturation, and vascular differentiation occurring in response to progesterone produced by the postovulatory corpus luteum. The endometrium increases in thickness up to 7 or 8 mm. The secretory phase may be divided into three stages, namely, the early secretory phase (from the 2nd to 4th postovulatory day, days 16–18 of a normal 28-day cycle), the mid-secretory phase (from the 5th to 9th postovulatory day, days 19–23 of a normal 28-day cycle), and the late-secretory phase (from the 10th to 14th postovulatory day, days 24–28 of a normal 28-day cycle). These phases are continuous and not sharply demarcated, and some areas of the endometrium may appear at a slightly more advanced stage than others.

The major morphological features that occur in the endometrium during the secretory phase are described in Table 1. There is an interval of 36–48 h between ovulation and the first morphologically recognizable signs of early secretory activity. In the early secretory phase, the endometrial glands still have a tubular appearance, and mitotic activity may be identified. The initial morphological feature of ovulation is the appearance within the glandular epithelium of subnuclear vacuoles. These typically appear on the 16th day of the typical 28-day cycle, that is, the 2nd postovulatory day. Initially, subnuclear vacuoles are identified only within occasional cells and are irregularly distributed but are usually most obvious in the mid-zone of the functionalis. There is a progressive increase in the number and distribution of subnuclear vacuoles until they involve almost all cells within most glands in the functionalis (Fig. 2). Subnuclear vacuoles are maximal between the 17th and 18th day of the cycle (3rd and 4th postovulatory days). As stated, some areas of the endometrium may appear at a slightly more advanced stage than others, and in the early secretory phase, there may be an admixture of glands exhibiting proliferative and secretory activity; in fact, an individual gland may exhibit both mitotic activity and subnuclear vacuolation. It is generally assumed that ovulation has occurred when there are subnuclear vacuoles in at least 50% of the cells in at least 50% of the glands; scattered subnuclear vacuoles are not reliable evidence of ovulation and, as stated, may be seen in late-proliferative endometrium. In the early secretory phase, the stroma is indistinguishable from that of late-proliferative endometrium.
Table 1

Morphological landmarks in secretory endometrium according to cycle day

Cycle day


Morphological features


Early secretory

Irregular basal vacuoles

Pseudostratified nuclei

Numerous mitoses present


Early secretory

Even subnuclear vacuoles

Aligned nuclei

Occasional mitoses


Early secretory

Supranuclear (luminal) vacuoles

Pseudostratified nuclei

Mitoses rare



Straight glands

Few remaining supranuclear vacuoles

Intraluminal secretions

Absent mitoses



Angulated glands

Prominent intraluminal secretions

Absent mitoses

Increasing stromal edema



Peak stromal edema with “naked” stromal cells

Centrally locates inspissated luminal secretions



Decreasing stromal edema

Spiral arterioles prominent

Earliest predecidual change around spiral arterioles


Late secretory

Serrated glands with secretory exhaustion

Minimal or absent stromal edema

Prominent predecidual change around spiral arterioles with vascular groups bridged by predecidua


Late secretory

Focal predecidua under surface epithelium


Late secretory

Extensive predecidual change under surface epithelium

Occasional mitoses reappear

Mild stromal inflammatory infiltrate


Late secretory

Diffuse predecidual change

Increasing number of mitoses

Prominent stromal inflammatory infiltrate

Fibrin thrombi

Foci of hemorrhage

Fig. 2

Early secretory phase endometrium. Tubular glands exhibit subnuclear vacuolation

Between the 19th and 23rd day of a typical 28-day cycle (the mid-secretory phase), the degree of glandular secretion increases. Cytoplasmic vacuoles become supranuclear, and secretions are seen within the glandular lumina (Fig. 3); it is important to realize that secretory material within the glandular lumina is not specific to secretory endometrium, but may also be seen in proliferative, hyperplastic, and malignant endometria. Mid-secretory glands are usually angular in shape, and mitotic activity is no longer apparent. The glands in the superficial layers of the functionalis tend to exhibit less secretory activity, and, as a result, superficial biopsies may produce a false impression of poorly developed secretory activity. Stromal edema progressively increases, is most obvious in days 22 and 23, and is most prominent in the mid-zone (Fig. 4). Spiral arteries become apparent. At this stage, the stromal cells surrounding spiral arteries acquire a more conspicuous eosinophilic cytoplasm; these cells are referred to as predecidual cells.
Fig. 3

Mid-secretory endometrium. The glands contain supranuclear vacuoles with secretions within the glandular lumina

Fig. 4

Mid-secretory endometrium. There is marked stromal edema with “naked” stromal cells

In the late-secretory phase (days 24–28 of a typical 28-day cycle or 10th to 14th postovulatory days), there is typically diminution of glandular secretory activity (secretory exhaustion), and the glands become serrated. Predecidual stromal change increases, initially being most apparent in the cells surrounding the spiral arteries (Fig. 5). The predecidual change results in the formation of the so-called compact layer (stratum compactum) beneath the surface epithelium, the deeper layers of stroma exhibiting less predecidual change. Sometimes, the predecidual cells may have a spindle cell or even signet-ring morphology and may not be readily appreciated as predecidual cells. Occasional mitoses may reappear in the predecidual stromal cells on day 26 or 27. A stromal infiltrate of granulated lymphocytes (described in detail later) is now obvious, and occasional neutrophils may appear in the premenstrual phase; the presence of granulated lymphocytes and neutrophils should not be misinterpreted as evidence of an endometritis. In late-secretory endometrium, the glands may be closely packed (this impression can be exacerbated by tangential sectioning), and this can superficially resemble hyperplastic endometrium (Fig. 6); however, other features of a hyperplastic endometrium, such as mitotic activity, are absent. In some late-secretory endometria, the glands have a “hypersecretory” appearance, resembling Arias-Stella reaction. This should not, in itself, be interpreted as evidence of early pregnancy. In the immediate premenstrual days, apoptotic activity is seen within the glands, fibrin thrombi appear in the small blood vessels, and there is extravasation of red blood cells into the stroma.
Fig. 5

Late-secretory endometrium. There is predecidual change surrounding the spiral arteries and focally beneath surface epithelium

Fig. 6

Late-secretory endometrium. The glands may be closely packed superficially resembling a hyperplastic endometrium

Menstrual Phase

Menstruation occurs after the 28th day of the normal cycle (the onset of menstruation is the first day of the menstrual cycle) and is characterized by glandular and stromal breakdown. Menstruation usually lasts for about 4 days. The endometrial glands are serrated and collapsed. Some of the glands remain vacuolated, imparting their secretory appearance. The stroma is condensed and collapsed, and the stromal cells aggregate into tightly packed balls (stromal blue balls) and separate from the glands (Fig. 7); the presence of tightly aggregated balls of stromal cells with hyperchromatic nuclei may be worrisome to the unwary. The predecidual appearance of the stromal cells is lost. Other features include the presence of necrotic debris, neutrophil infiltration, interstitial hemorrhage, and fibrin deposition. Apoptotic bodies are identified within both the glands and the stroma (Fig. 8). When menstrual activity is well developed, little or no stromal tissue may remain, and the glands become closely packed, sometimes with a back-to-back appearance. To the inexperienced, this may result in consideration of a hyperplasia or carcinoma. Following breakdown, the endometrial glands assume a surface micropapillary architecture (Fig. 9). The term papillary syncytial metaplasia (discussed later) is used for this appearance, but this is an inaccurate term since this is not strictly speaking a metaplasia but rather a regenerative/degenerative process secondary to tissue breakdown. Mitotic figures may be present within the papillary proliferations. Papillary syncytial metaplasia is also seen following surface breakdown associated with non-menstrual conditions. On occasion, the micropapillary architecture is particularly striking and, if associated with mitotic activity, raises the possibility of a serous carcinoma. The appreciation of the accompanying features of breakdown and immunohistochemical staining with p53 may be of value. Most serous carcinomas exhibit diffuse, intense, nuclear immunoreactivity with p53, while papillary syncytial metaplasia exhibits a wild-type staining pattern, which has been described as weak and heterogeneous (Quddus et al. 1999).
Fig. 7

Menstrual phase endometrium. The stromal cells aggregate into “blue balls”

Fig. 8

Menstrual phase endometrium. Apoptotic bodies are present within the endometrial glands and stroma

Fig. 9

Papillary syncytial metaplasia. Following breakdown, the endometrial glands regenerate with a surface micropapillary architecture: neutrophils are often seen within the epithelium

Lower Uterine Segment Endometrium

Lower uterine segment or isthmic endometrium is poorly responsive to steroid hormones, and the morphology does not alter significantly during the menstrual cycle; as is the case with the endometrial basalis, lower uterine segment endometrium is not useful for dating the menstrual cycle. Lower uterine segment endometrium is composed of inactive poorly developed glands that are often ciliated. They are irregularly distributed, and some may be dilated. The stroma typically has a fibrous appearance, and the stromal cells are more elongate and “fibroblast-like” than in the corpus. Given these features, lower uterine segment endometrium may be mistaken for a polyp in a biopsy specimen. In the inferior part of the lower uterine segment, the glands merge with mucinous type glands from the upper endocervix, and the stroma becomes even more fibrous (Fig. 10).
Fig. 10

Lower uterine segment endometrium. This is composed of a mixture of ciliated and mucinous glands within a fibrous stroma

Steroid Hormone, Steroid Hormone Receptor, and Immunopeptide Interactions in the Endometrial Cycle

As detailed, the menstrual cycle in postmenarchal, premenopausal women follows a regular series of morphological and physiological events characterized by proliferation, secretory differentiation, shedding, and regeneration of the uterine lining. These alterations are controlled by the cyclical release of the steroid sex hormones estradiol (E2) and progesterone (P) from the ovaries; the endometrium is thus a highly sensitive indicator of the hypothalamic–pituitary–ovarian axis. Steroid hormone control of endometrial epithelial and stromal cells is mediated by estrogen receptors (ER) and progesterone receptors (PR). These steroid receptors are proteins concentrated in the nuclei of the endometrial epithelial and stromal cells that have high affinity to bind E2 and P. Because they are sex steroid hormone (ligand) specific, a particular receptor may display high affinity for a closely related class of hormones, and these classes may compete for available binding sites. For example, ER effectively binds not only E2 but also estrone (E1), as well as synthetic estrogens, such as diethylstilbestrol (DES).

Although E2 plays a crucial role in the proliferation of endometrial cells in vivo, E2 alone is not able to induce the proliferation of endometrial cells in primary culture. It has been suggested that the mitogenic action of E2 is mediated indirectly via a paracrine effect by the polypeptide growth factor, epidermal growth factor (EGF). EGF promotes the transition of cells from the G0 to G1 phase of the cell cycle (Taketani and Mizuno 1991). Human endometrial cells possess EGF receptors and mRNA for EGF. EGF-like immunoreactivity is seen in both the endometrial epithelial and stromal cells, with higher concentrations in the epithelium than the stroma, and parallels the fluctuation of cyclic sex steroid hormones during the menstrual cycle. It appears that the EGF receptor content is regulated by ovarian E2 and P secretion (autocrine control). Indeed, EGF alone fails to influence cell proliferation, but in combination with E2, it increases the mean glandular but not stromal cell counts by more than 50% in vitro. The immunolocalization of EGF in normal human endometrium and the stimulation of epithelial cell proliferation in culture by EGF and E2 provide support for a role of EGF in endometrial growth.

Similarly, endometrial stromal cells produce insulin-like growth factors 1 and 2 (IGF-1 and IGF-2) and high-affinity IGF binding proteins (IGFBP), which act on both epithelial cells and stromal cells via insulin-like growth factor receptors, stimulating proliferation, differentiation, and metabolic effect (Rutanen 1998). Estrogen stimulates IGF-1 gene expression in the endometrium and is presumed to mediate estrogen action, while IGF-2 gene expression is thought to be related to endometrial differentiation. IGFBP-1 is produced by decidualized endometrial stromal cells and inhibits the biological actions of IGF-1 and, as a result, inhibits the actions of estrogen. Therefore, clinical conditions characterized by the absence of IGFBP-1 may lead to unopposed IGF-1 action and subsequent endometrial proliferations and cancer development (Rutanen 1998). It has been shown that women with polycystic ovarian syndrome have increased endometrial expression of IGF-1 compared to controlled women, which may play a role in the higher incidence of endometrial carcinoma in this population (Shafiee et al. 2016).

The continuous dynamic remodeling of the endometrium results from a delicate balance of cellular proliferation and programmed cell death within specific subpopulations of stromal and epithelial cells, a process that is modulated by steroid hormones. A ladder pattern of DNA cleavage characteristic of apoptosis is seen in the late-secretory, menstrual, and early-proliferative phases (Shikone et al. 1997). Localization of apoptotic subpopulations using in situ assays for DNA breakage has shown that the majority of apoptotic cells represent glandular cells within the basalis and these cells increase in number throughout the secretory phase and peak during menses (Tabibzadeh et al. 1994). The apoptotic effects of steroid hormones are likely mediated through a complex network of inhibitors and initiators. Progestins have been shown to decrease endometrial secretion of the apoptosis inhibitor Bcl-2, a process that is reversed upon administration of antiprogestogenic agents (Critchley et al. 1999; Gompel et al. 1994). Progestins may also positively promote apoptosis by increasing levels of the apoptosis inducer gene BAK (Tao et al. 1998).

The concentrations of ER and PR in the normal endometrium vary during the normal menstrual cycle according to fluctuating plasma levels of E2 and P. The highest values of ER (approximately 400 fmol/mg protein) and PR (approximately 1,000 fmol/mg protein) occur during the mid-proliferative phase (8th–10th day of the cycle) and correspond to rising plasma levels of E2. E2 promotes the synthesis of both ER and PR, whereas P inhibits the synthesis of ER. Monoclonal antibodies to ER allow the precise intracellular localization of ER by means of immunohistochemistry. Most ER is localized in the nuclei rather than the cytoplasm of endometrial epithelial and stromal cells. Endothelial cells fail to stain with ER antibodies. The concept of the mechanisms of sex steroid hormone–receptor action in target cells includes the following major steps: (1) circulating and unbound steroid hormone molecules are taken up from the cytoplasmic membrane, presumably by cytoplasmic receptors; (2) the hormone molecules enter the nucleus, which contains most (90–95%) of the cellular receptors; (3) the intranuclear hormone molecules induce conversion of the inactive (nonfunctional) 4S form of receptor to the active (functional) 5S form of the receptor; (4) the hormonally activated 5S receptor binds to acceptor genes in the nucleus and influences gene expression by stimulating RNA polymerase and thus mRNA transcription; and (5) the newly formed mRNA is transported to the cytoplasm, where it is translated into proteins, including anabolic and catabolic enzymes, as well as new receptors (receptor replenishment). According to this concept, the most significant effect of sex hormones is intranuclear activation of receptors that in turn initiate a sequence of events, which results in alterations in the physiologic functions of target cells.

Immunohistochemistry of the Normal Endometrium

The normal endometrial glands and stroma are estrogen receptor (ER) and progesterone receptor (PR) positive (Fig. 11). Endometrial glands are diffusely positive with the anti-apoptotic protein Bcl-2 during the proliferative phase of the menstrual cycle. There is marked diminution in staining during the early and mid-secretory phases with reappearance during the late-secretory phase (Bozdogan et al. 2002; Gompel et al. 1994; Morsi et al. 2000). A minor population of endometrial epithelial cells exhibit nuclear immunoreactivity with p63; it has been speculated that these are reserve cells or basal cells and the origin of metaplastic endometrial epithelial cells (O’Connell et al. 2001). Endometrial glands are cytokeratin 7 positive and cytokeratin 20 negative. The normal endometrial stroma is CD10 positive (Fig. 12), Bcl-2 positive, and CD34 negative, in contrast to cervical stroma, which is largely CD10 and Bcl-2 negative and CD34 positive (Barroeta et al. 2007), although there may be immunoreactivity with CD10 of stromal cells surrounding endocervical glands (McCluggage et al. 2003a). This differing immunophenotype may be useful in assessing whether the stroma accompanying a neoplasm is endometrial or cervical in type and in assessing tumor origin when this is problematic. The immunophenotype of lower uterine segment stroma overlaps with that of the endometrial and the cervical stroma. Normal endometrial stroma may be smooth muscle actin positive but is desmin negative. The immunohistochemistry of endometrial hematopoietic cells is discussed in the next section.
Fig. 11

Estrogen receptor in the endometrium. The normal endometrial glands and stroma are positive with estrogen receptor

Fig. 12

CD10 in the endometrium. The normal endometrial stroma is CD10 positive

Hematopoietic Cells Within the Endometrium

The normal endometrium contains a variety of hematopoietic cells, the composition of which varies depending on the stage of the menstrual cycle and the menopausal status. Lymphocytes, including lymphoid aggregates and occasionally lymphoid follicles with germinal centers, are found at all stages of the menstrual cycle and in the postmenopausal endometrium. As mentioned above, in the normal menstrual cycle, lymphoid aggregates are most common in the proliferative phase. However, they are more commonly seen postmenopausally within the basal endometrium (Fig. 13); it is not clear whether this is due to lymphoid aggregates being more numerous postmenopausally or whether they are more obvious secondary to the glandular atrophy. Immunohistochemical studies have demonstrated that B lymphocytes (CD20 and CD79a positive) constitute approximately 1% of the normal lymphoid population of the endometrium and are present mainly in aggregates in the basalis and rarely as individual cells in the functionalis. T lymphocytes (CD3 positive) are more common and are present throughout the endometrial stroma, usually as individual cells, and are more numerous during the secretory phase (Bulmer et al. 1988; Disep et al. 2004). The distribution of B and T lymphoid cells is altered in endometritis (see section “Endometritis”). Granulated lymphocytes (CD56 positive) are present in large numbers in predecidualized endometrial stroma in the mid- and late-secretory phases (Bulmer et al. 1987, 1988); these were formerly designated endometrial stromal granulocytes and have a mononuclear or bilobed nucleus and abundant eosinophilic cytoplasm containing a variable number of granules. The function of granulated lymphocytes is not entirely clear, but they have the characteristics of natural killer cells and, as well as being positive with CD56, are immunoreactive with T-cell markers. Neutrophils are present in small numbers throughout the menstrual cycle but only become morphologically evident in large numbers in association with the tissue breakdown and necrosis associated with menses. In contrast to lymphocytes and neutrophils, plasma cells are not considered to be a component of the normal endometrium, although the presence of an occasional plasma cell is allowable in an otherwise morphologically normal endometrium. Plasma cells are, of course, a feature of endometritis (see section “Endometritis”). Rare mast cells, demonstrable with toluidine blue or Giemsa staining, may be found in the endometrium, primarily within the basalis. Mast cells are also found normally in the myometrium, in endometrial polyps, and in leiomyomas, often in large numbers in the latter. The number of mast cells in the endometrium and myometrium tends to decrease with advancing age. Histiocytes are also seen in the normal endometrium (see section “Contaminants and Other Elements in Endometrial Biopsies”). Eosinophils are not a component of the normal endometrium. Rarely, foci of extramedullary hematopoiesis are present in the endometrium, usually in association with an underlying hematopoietic disorder or occasionally representing remnants of fetal tissue (Creagh et al. 1995; Valeri et al. 2002).
Fig. 13

Lymphoid aggregates in basal endometrium. Lymphoid aggregates are a normal phenomenon within the endometrial stroma

Gestational Endometrium

Pregnancy is characterized by morphological changes involving the endometrial glands and stroma. The trophoblastic populations in an intrauterine gestation are covered in chapter “Gestational Trophoblastic Tumors and Related Tumor-Like Lesions”. Early in pregnancy, the endometrium displays hypertrophic and hypersecretory features that have been referred to as “gestational hyperplasia.” The endometrium is characterized by (1) glandular ferning with epithelial and intraluminal secretions, (2) stromal edema and vascular congestion, and (3) decidual reaction of the stromal cells. The changes are similar to but quantitatively exaggerated compared to those of nongestational endometrium at days 22–26 of the menstrual cycle. In the normal cycle, each of these alterations is prominent on a given day of the secretory phase, whereas during early pregnancy they occur simultaneously. The features described are not pathognomonic of early pregnancy but may occasionally be seen in other situations (see section “Arias-Stella Reaction”).

Immediately after implantation of the blastocyst on the endometrial surface, changes begin to occur in the endometrial glands and the stroma, although the overall morphology of late-secretory endometrium is maintained for several days. The early changes include glandular serration and distension, increase in glandular secretions, stromal edema, and a stromal predecidual reaction. The morphological features during the first 2 weeks of gestation are subtle, but after approximately 15 days, they become more characteristic with the formation of decidual cells. Compared to predecidual cells, decidual cells are larger with prominent cell membranes and more abundant eosinophilic cytoplasm that may contain small vacuoles (Fig. 14). The nuclei of decidual cells are round to ovoid with finely dispersed chromatin and indistinct nucleoli. Stromal granulated lymphocytes are present during early pregnancy, and spiral arteries become more apparent; they have thicker walls than in nongestational secretory endometrium. Some of the spiral arteries display acute atherosis with concentric intimal proliferation of myofibroblasts and accumulation of foamy cells. As the pregnancy progresses, decidual cells become widespread with better defined cell borders and develop an epithelioid appearance. Small numbers of granulated lymphocytes remain throughout the gestation. Some of the glands become atrophic, while others have a hypersecretory appearance. Four to 8 weeks after implantation, the glands often exhibit, at least focally, the Arias-Stella reaction, which is a response to the presence of trophoblastic tissue in the uterus or at an ectopic site. Histologically, the Arias-Stella reaction is characterized by cellular stratification, secretory activity, vacuolated cytoplasm, and enlargement of the epithelial cell’s nuclei and cytoplasm (Fig. 15). The nuclei may be enlarged up to three times normal and can exhibit considerable atypia and a hobnail appearance with bulging into the glandular lumina. Mitotic figures may be present, although these are rarely prominent, and there is a low MIB1 proliferation index. Atypical mitoses have rarely been described (Arias-Stella et al. 1994). The Arias-Stella reaction may be extensive, involving many glands, or focal with involvement of only a few glands or even part of a gland. The changes persist for at least 8 weeks following delivery. Several histological variants have been described, namely, atypical, early secretory pattern, hypersecretory pattern, regenerative pattern, and monstrous cell pattern (Arias-Stella 2002). However, there is considerable overlap between these patterns, and there is no value in attempting to subclassify Arias-Stella effect in the endometrium. The Arias-Stella reaction may also be seen in the glandular epithelium of the cervix and fallopian tube and involving endometriosis or vaginal adenosis (Nucci and Young 2004).
Fig. 14

Gestational endometrium. The stroma is expanded and composed of decidualized cells with abundant eosinophilic cytoplasm

Fig. 15

Arias-Stella reaction in pregnancy. There is cellular stratification, vacuolated cytoplasm, and enlargement of the epithelial cell nuclei

Apart from the Arias-Stella reaction, the endometrial glands may undergo other changes in the presence of trophoblastic tissue. These include abundant clear glycogen-rich cytoplasm; this overlaps with the Arias-Stella reaction, but the nuclear enlargement of the latter is not present. Another pregnancy-related change is the presence of optically clear nuclei within the endometrial epithelium (Fig. 16) (Mazur et al. 1983). This may occur in association with the Arias-Stella reaction or independently. This appearance is due to the intranuclear accumulation of biotin and may simulate the ground-glass nuclei of herpes simplex virus infection (Yokoyama et al. 1993). However, the nuclei lack the Cowdry type A eosinophilic intranuclear inclusions and nuclear molding characteristic of herpes simplex virus infection. Due to the presence of biotin within these nuclei, a falsely positive immunohistochemical stain might result if immunohistochemistry is attempted in these cases, secondary to nonspecific binding of avidin to the nuclear biotin (Matias-Guiu et al. 1994).
Fig. 16

Optically clear nuclei in pregnancy. Endometrial glands in pregnancy may contain cells with optically clear nuclei

Localized endometrial glandular proliferations have also been described during pregnancy, usually in first trimester gestations of women in their fourth or fifth decades. These rare focal lesions are characterized by glandular expansion, nuclear stratification, a cribriform architecture, and intraluminal calcifications (Genest et al. 1995). There is mitotic activity, but the cytology is bland. These appear to be benign lesions based on uneventful follow-up and an unusual response to pregnancy.

In early pregnancy, endometrial glands become strongly S100 positive (Nakamura et al. 1989). This immunoreactivity disappears after the 12th week of gestation. The reason for this S100 positivity is not clear.

It is emphasized that the described endometrial morphological changes of pregnancy may be seen with both an intrauterine gestation and an ectopic pregnancy. Confirmation of an intrauterine gestation requires the presence of trophoblast, either in the form of chorionic villi or a placental site reaction (described in chapter “Gestational Trophoblastic Tumors and Related Tumor-Like Lesions”). In the placental site, intermediate trophoblast infiltrates the decidua. On occasion, it may be difficult to distinguish between decidual cells and implantation site intermediate trophoblast, although intermediate trophoblastic cells are larger and more variable in size and shape, ranging from polygonal to spindle shaped. The nuclei are lobated, and hyperchromatic binucleate or multinucleate cells may be present. There may be prominent nucleoli and sharply defined cytoplasmic vacuoles. In contrast, the nuclei of decidualized stromal cells are uniform and round to oval with finely dispersed chromatin and inconspicuous nucleoli. In problematic cases, immunohistochemical staining may assist in distinguishing between intermediate trophoblastic cells at the placental site and decidualized stromal cells. Markers of value are discussed in chapter “Gestational Trophoblastic Tumors and Related Tumor-Like Lesions”, but trophoblastic cells are immunoreactive with broad-spectrum cytokeratins, cytokeratin 7, human placental lactogen, HLA-G, inhibin, and mel-CAM (CD146), while decidua is negative (Kurman et al. 1984; McCluggage et al. 1998; O’Connor and Kurman 1988).

Endometrium Associated with Ectopic Pregnancy

The morphology of gestational endometrium is discussed in the previous section. Endometrial changes also occur with an ectopic pregnancy. The features are variable, but by day 22–28 of the ectopic gestation, the endometrial glands usually exhibit secretory or hypersecretory features, sometimes with Arias-Stella reaction, at least focally. Occasionally, the endometrial glands are atrophic. If the ectopic trophoblast regresses, the endometrial glands have a variable appearance, ranging from proliferative to secretory to a picture identical to that seen in disordered proliferation or progestogen effect. The endometrial stroma in association with an ectopic gestation is usually decidualized and devoid of inflammation. Thick-walled spiral arteries are present. Trophoblastic tissue, in the form of chorionic villi or a placental site reaction, is absent.

Postmenopausal Endometrium

The age of menopause with cessation of ovulation and resultant diminution of hormone production by the ovaries is variable, but is usually around age 50. Postmenopausally, the endometrium becomes thin and atrophic, unless there is continuing estrogenic drive, either in the form of endogenous production or exogenous hormone use. When there is no estrogenic drive, the functionalis is absent and the endometrium is composed only of basalis, similar to the basalis of the reproductive years and of the premenarchal endometrium.

The histological appearance of the postmenopausal endometrium is variable. The endometrium is usually thin, and this is appreciated in hysterectomy or endometrial resection specimens. The glands do not exhibit proliferative activity and vary from consisting entirely of small widely spaced atrophic tubules (Fig. 17) to cystically dilated glands throughout (so-called cystic atrophy or senile cystic atrophy) (Fig. 18). A mixture of small tubules and cystically dilated glands may occur. The cystic glands seen in atrophic endometria in hysterectomy or resection specimens may not be observed in endometrial biopsies because tissue fragmentation during the procedure disrupts the glands. Usually, tubular glands are more prominent in the years immediately following menopause, while cystic atrophy is more common in older women, but this is variable. The glandular cells have small, dark, regular nuclei that may be round, ovoid, or low columnar, and in cystic glands, the nuclei are often compressed and attenuated. Sometimes, there is a degree of nuclear pseudostratification, which may result in a false impression of proliferative activity. The cytoplasm is usually scanty. High-power examination is required to confirm an absence of mitotic activity, and this is especially so in the distinction between proliferative endometrium and atrophy with small tubular glands and between cystic forms of atrophy and simple hyperplasia. The stroma in postmenopausal endometria may be densely cellular and composed of ovoid- to spindle-shaped cells with scant cytoplasm, or has a more fibrous appearance than in the premenopausal endometrium. With advancing age, the stroma tends to become more hypocellular and fibrous. This may be the direct cause of the cystic change because of blockage of the glands. Lymphoid aggregates are often prominent, more so than in premenopausal endometria. As mentioned, it is not clear whether this is due to an actual increase in the number of aggregates or due to them being more obvious because of the glandular atrophy. Occasional mitotic figures may be seen in the glands of postmenopausal endometria with no proven source of estrogenic stimulation. This may occur in women in whom menopause appears gradually and also with uterine prolapse. The postmenopausal endometrium is hormone receptor positive, like that of premenopausal endometria, and retains the capacity to respond to estrogenic stimulation.
Fig. 17

Atrophic endometrium. There are widely spaced small atrophic tubules within a fibrotic background

Fig. 18

Cystic atrophy. Cystically dilated glands are present within a fibrous stroma

The morphological features of postmenopausal endometria are similar to those of atrophic endometrium due to other causes, such as exogenous hormones, although there is often also stromal predecidualization in those patients taking progestogen-only compounds or combined preparations containing a progestogen. Atrophic endometrium may also occur in young patients with premature ovarian failure, either idiopathic or due to surgical removal, chemotherapy, or radiotherapy.

Endometrial Sampling

Most endometrial samples are taken during the investigation of abnormal uterine bleeding in pre-, peri-, or postmenopausal women. Traditionally, most endometrial samples were obtained by cervical dilatation and curettage (D&C). It was believed that this had a therapeutic effect in some cases of abnormal uterine bleeding, as well as producing a sample for histological examination. Presently, most endometrial samples are outpatient biopsies performed by pipelle or other methods of endometrial sampling. In contrast to the traditional D&C sample, pipelle biopsies do not require an anesthetic and are often performed in conjunction with ultrasound examination and/or hysteroscopy, both of which may identify focal lesions that could be missed by pipelle biopsy alone. The disadvantages of a pipelle biopsy are that often very scant tissue is obtained, especially in a postmenopausal woman with an atrophic endometrium, and focal lesions may be missed. Issues relating to adequacy of endometrial samples are discussed below. On occasions, a pipelle biopsy is followed by a D&C, for example, when there is a suspicion of malignancy and pipelle biopsy produces only a scant amount of tissue which is nondiagnostic, or when there are worrying features on the pipelle sample that are not diagnostic of malignancy. As stated earlier, endometrial sampling is currently not common in the investigation of infertility, but when this is the case, the timing of the biopsy is important; ideally, the biopsy should be taken in the mid-secretory phase between the 7th and 11th postovulatory days.

Endometrial polypectomy may be undertaken, and in such instances, sampling of the non-polypoid endometrium should also be performed. Hysteroscopic endometrial resection may be performed under a variety of circumstances, such as in the management of menorrhagia due to large or multiple polyps, and in cases of submucosal leiomyomas. This procedure produces a specimen consisting of multiple chippings, similar to prostatic chips. Chips should be weighed (this should also be done when biopsy or D&C yields a significant amount of tissue, but may be impractical with very scant specimens). All endometrial specimens should be submitted in their entirety for histological examination. The exception is large endometrial polyps where representative sampling may be undertaken; it should be remembered that endometrial polyps may be involved by small carcinomas or by serous endometrial intraepithelial carcinoma (serous EIC), and, as a rule of thumb, at least one block per cm should be taken of large endometrial polyps. Endometrial chips should be orientated if possible, since poorly orientated specimens, especially if tangentially sectioned, may result in a histological suspicion of adenomyosis; it is doubtful whether adenomyosis can be reliably diagnosed on a hysteroscopic endometrial resection sample, although this can be suspected on well-orientated specimens, especially when there is smooth muscle hypertrophy surrounding the islands of endometrial tissue within the myometrium or when there is endometrial tissue present on at least three sides of a fragment of myometrial tissue (Busca and Parra-Herran 2016). A related, although uncommon problem with endometrial resection specimens is the potential for overdiagnosis of myometrial invasion in cases of endometrial hyperplasia or adenocarcinoma; again, this is due to poor orientation and tangential sectioning.

In evaluating any endometrial specimen, an adequate clinical history is important, including the age of the patient and the reason for biopsy. Knowledge of the menopausal status as well as the date of onset of the last menstrual period (LMP) and the length of the menstrual cycle in premenopausal women should be provided. In some cases of “postmenopausal bleeding,” the patient is not actually postmenopausal but rather perimenopausal with a prolonged interval between periods, resulting in the clinician and the patient assuming that the woman is postmenopausal. Many women with abnormal uterine bleeding have been prescribed exogenous hormones (especially progestins) before biopsy to control the bleeding, and this information is not always conveyed to the pathologist. Other women may be taking hormone replacement therapy (HRT) or contraceptive agents. These hormonal compounds may alter the morphological appearance of the endometrium, and a knowledge that these and other relevant medications (such as tamoxifen) are being taken is paramount to the pathologist.

Criteria for Adequacy of Endometrial Sample

With the increasing trend to perform outpatient endometrial pipelle biopsies rather than formal curettage, pathologists are dealing with increasing numbers of endometrial specimens in which there is scant, or even no endometrial, tissue, especially when the endometrium is atrophic. These specimens may consist entirely of superficial strips or wisps of atrophic glands (Fig. 19), with little or no stroma, admixed with cervical mucus, ectocervical or endocervical tissue, and tissue from the lower uterine segment. Paradoxically, it often takes the pathologist longer to examine such specimens, since no underlying architecture is present, and the tissue must be examined carefully under high power to look for mitotic activity, which is abnormal in a postmenopausal endometrium. In specimens such as this, it is controversial as to what constitutes an adequate or inadequate specimen. Designation of a biopsy as inadequate may be of importance since this can have management and medicolegal implications. For example, some clinicians routinely perform a repeat biopsy when an earlier sample has been reported as inadequate while others do not. A biopsy reported as inadequate may suggest to some that the clinician is at fault or has not undertaken the biopsy procedure correctly. While this may be the case in some instances, in most it is not. In published studies, inadequate rates of outpatient endometrial biopsies range from 4.8% to 33% (Antoni et al. 1997; Archer et al. 1991; Clark et al. 2001; Gordon and Westgate 1999; Machado et al. 2003), but in most of these studies, the criteria for adequacy are not clear. A recent study showed that the presence of more than 10 strips of endometrium appeared sufficient to exclude a malignant process, with a negative predictive value close to 100%; this negative predicative value dropped to 81% in cases showing less than 10 strips of endometrium (Sakhdari et al. 2016). It is also worth noting that studies have shown that with an atrophic endometrium and no focal lesion, minimal tissue is the norm with a pipelle biopsy, and there is little chance of missing significant pathology (Bakour et al. 2000). Although it is difficult to recommend precise criteria for adequacy, caution should be exercised before categorizing an endometrial biopsy as inadequate or insufficient. In the majority of cases, the presence of only scant tissue in an endometrial specimen is not a reason for a repeat biopsy, provided the endometrial cavity has been entered, and at least some endometrial tissue is present in the biopsy specimen to confirm this, although theoretically endometrial-type glands with or without stroma could be derived from tuboendometrial metaplasia or endometriosis within the cervix.
Fig. 19

Scanty endometrial biopsy. Endometrial pipelle biopsy in postmenopausal woman where the specimen consists entirely of superficial strips of atrophic endometrial glands without accompanying endometrial stroma

It has been suggested that with an endometrial biopsy containing scant tissue for which the origin or differentiation cannot be determined, the term unassessable is more appropriate than inadequate or insufficient (Phillips and McCluggage 2005). In such cases, the gynecologist should correlate the biopsy results with the ultrasonographic and/or hysteroscopic findings. If, for example, there is a clinical suspicion of hyperplasia or malignancy, if there is recurrent postmenopausal bleeding, or if there are worrying ultrasonographic and/or hysteroscopic findings, then D&C should be performed. If the above investigations suggest an atrophic endometrium, rebiopsy is probably unnecessary.

Artifacts in Endometrial Biopsy Specimens

There are several common artifacts in endometrial biopsy specimens that have received scant attention in the literature (McCluggage 2006). Occasionally, these may be misinterpreted as an endometrial hyperplasia or even a carcinoma if not appreciated to be artifactual. Telescoping is common and refers to the presence of glands within glands (Fig. 20). This artifact seems to be a result of mechanical disruption and “snap back” of the glands during biopsy, resulting in a form of intussusception. Artifactual crowding and compression of glands are also common and may result in consideration of a complex endometrial hyperplasia. With this artifact, the glands often become “molded” together, and there is commonly tearing of the tissue around the glands, which is a clue to the artifactual nature of the glandular crowding (Fig. 21). An artifact that is especially common with, but not exclusive to, outpatient biopsies is the presence of superficial strips of endometrial epithelium, sometimes accompanied by minimal stroma, with a pseudopapillary architecture (Fig. 22). This may result in consideration of a wide range of papillary lesions, benign and malignant, which occur in the endometrium. Such superficial strips of pseudopapillary epithelium, which are generally atrophic, should be examined carefully under high power to look for proliferative activity and nuclear atypia. Crushed endometrial glands and stroma may be extremely cellular and can cause concern. Extensive crush artifact is more likely to occur in biopsies from atrophic endometrium in postmenopausal patients. As with the examination of other tissues, crushed elements should not be viewed in isolation. Problems associated with poorly orientated hysteroscopic endometrial resection specimens have already been described.
Fig. 20

Telescoping (glands within glands). This is a common artifact in endometrial biopsy specimens

Fig. 21

Glandular molding. A common artifact in endometrial biopsies is glandular “molding.” There is tearing of the tissue around the glands, which is a clue to the artifactual nature

Fig. 22

Pseudopapillary endometrium. Endometrial biopsy composed of superficial strips of endometrial epithelium with a pseudopapillary architecture

Hysteroscopic endometrial resection specimens might demonstrate vacuolation of the endometrial stromal cells secondary to cautery artifact, resulting in a signet-ring appearance, similar to the phenomenon of vacuolation sometimes observed in cervical stromal cells (McKenna and McCluggage 2008). Occasionally, vacuoles resembling adult fat might be seen in endometrial biopsy specimens, a finding termed pseudolipomatosis (Fig. 23). Since adipose tissue within an endometrial biopsy specimen is an alarming finding that would indicate iatrogenic uterine perforation (see section “Extrauterine Tissues in Endometrial Biopsy Specimens”), awareness of this finding is necessary to avoid overdiagnosing uterine perforation. In contrast to adipose tissue, which demonstrates relatively monotonous vacuoles with identifiable adipocyte nuclei and intervening capillaries, pseudolipomatosis shows variation in the size of the vacuoles and absence of both adipocyte nuclei and intervening capillaries (Deshmukh-Rane and Wu 2009). Pseudolipomatosis may be related to irritants used during sterilization or secondary to suction applied during the procedure.
Fig. 23

Pseudolipomatosis. Fatlike vacuoles can be distinguished from true adipose tissue by their variation in size and the absence of adipocyte nuclei or intervening capillaries

Contaminants and Other Elements in Endometrial Biopsies

Not uncommonly, fragments of tissue other than from the endometrium are present in endometrial biopsy or curettage specimens. Superficial myometrium is commonly seen, especially in vigorous curettage specimens and in postmenopausal women with an atrophic endometrial lining. It is very common to see cervical tissue as well as cervical mucus, often admixed with neutrophils, histiocytes, and giant cells (Fig. 24), in endometrial biopsy specimens. The cervical tissue usually takes the form of strips of endocervical glandular or squamous epithelium, sometimes with accompanying stroma. The squamous epithelium may be immature metaplastic in type. Usually the cervical origin is obvious, but occasionally this is not the case and diagnostic confusion may ensue. For example, if cervical glandular elements exhibiting microglandular hyperplasia are identified within an endometrial biopsy specimen, this may result in consideration of an endometrial hyperplasia or carcinoma, particularly in the postmenopausal setting. The confusion may be heightened by artifactual apposition such that it appears that the endometrial and cervical tissue are in continuity; assessment of whether the accompanying stroma is endometrial or cervical in type may assist in interpretation. Sometimes, dysplastic cervical squamous or glandular epithelium or tissue derived from a cervical neoplasm is present in an endometrial biopsy specimen. Fragments of fallopian tube epithelium may also be seen occasionally.
Fig. 24

Cervical mucus in endometrial biopsy specimen. In many endometrial biopsies, mucus derived from the cervix is present, often admixed with neutrophils, histiocytes, and giant cells

Occasionally, aggregates or sheets of histiocytes may be seen in an endometrial biopsy specimen, either free floating or within the endometrial stroma. Small numbers of histiocytes are not uncommon and are usually inconspicuous, but when present in large aggregates, this may result in consideration of an epithelial or stromal neoplasm (Fig. 25). Recognition of the characteristic lobated, reniform, or coffee-bean nucleus of histiocytes assists, and immunohistochemical staining for histiocytic markers, such as CD68 or lysozyme, may be of value. The histiocytes are probably a reaction to debris within the endometrial cavity, and, when present in large numbers, it has been referred to as nodular histiocytic hyperplasia (Fukunaga and Iwaki 2004; Kim et al. 2002). Occasionally, mitotic figures may be identified within the aggregates of histiocytes, and there may be prominent cell membranes. Rarely, the histiocytes have intracytoplasmic vacuoles and a signet-ring appearance (Iezzoni and Mills 2001), raising the possibility of a signet-ring carcinoma; staining for epithelial and histiocytic markers facilitates the diagnosis. Decidualized and predecidualized endometrial stromal cells may also contain intracytoplasmic vacuoles and simulate signet-ring cell carcinoma. Foamy histiocytes with abundant cytoplasm may also occur within the endometrium, either in association with an endometrial hyperplasia, carcinoma, or pyometra, or as a manifestation of xanthogranulomatous endometritis (see section “Endometritis”).
Fig. 25

Histiocytes in endometrial biopsy. Aggregate of histiocytes in an endometrial biopsy composed of cells with a coffee-bean nucleus and abundant eosinophilic cytoplasm

Extrauterine Tissues in Endometrial Biopsy Specimens

Rarely, extrauterine tissues are present in an endometrial biopsy specimen, and this raises the possibility that uterine perforation has occurred either during the current biopsy procedure or some time previously and that a fistulous tract is present. The most common extrauterine tissue is adipose tissue. Although this may potentially be derived from a uterine lipoleiomyoma, lipoma, hamartomatous lesion (McCluggage et al. 2000b), and other pathological lesions containing adipose tissue, or represent metaplastic adipose tissue within the endometrial stroma, in most instances the adipose tissue is derived from the pelvic soft tissues or omentum and indicates perforation. Occasionally, the adipose tissue is accompanied by fibrinous material and mesothelial cells; this is a reflection of underlying pelvic pathology with resultant mesothelial proliferation, which results in fixation of the uterus within the pelvis and makes perforation more likely. In some cases, the patient comes to no harm because of the perforation, presumably due to contraction of myometrial smooth muscle sealing off the defect, but in other instances, a pelvic and/or abdominal inflammatory process ensues and the patient becomes symptomatic. For this reason, the identification of adipose tissue in an endometrial biopsy specimen should prompt a phone call to the clinician. Rarely, other extrauterine tissues, such as intestinal mucosa, may be present in an endometrial biopsy specimen secondary to perforation.


Endometritis is a histological diagnosis based upon the identification within the endometrium of an abnormal pattern of inflammatory infiltrate; as such, it must be distinguished from the normal hematopoietic component of the endometrium (see section “Hematopoietic Cells Within the Endometrium”). Most cases of endometritis occur in the reproductive years, but sometimes postmenopausal women are affected. Presentation is typically with abnormal vaginal bleeding, most commonly intermenstrual bleeding or menorrhagia.

Endometritis may have both infective and noninfective etiologies. The endometrium is relatively resistant to ascending infection from the lower female genital tract because of the barrier created by the cervix and the cervical mucus. However, uncommonly, an endometritis occurs secondary to ascending infection, and this is often a component of pelvic inflammatory disease with inflammation elsewhere in the genital tract. Predisposing factors to endometritis include a recent pregnancy, the presence of an intrauterine device (IUD), cervical stenosis, and prior instrumentation. Endometritis may also accompany a pathological lesion within the uterus, such as an endometrial polyp, hyperplasia, carcinoma, or a leiomyoma. Usually, the morphological appearances are nonspecific, and, in the absence of an associated pathological lesion, an underlying cause cannot be determined, although occasionally the histological features suggest a particular etiology (see section “Specific Forms of Endometritis”).

Endometritis has traditionally been divided into acute and chronic forms, but these constitute a continuum, and often there is an admixture of acute and chronic inflammatory cells.

Endometritis may be focal or diffuse and can range from a subtle finding to a pronounced inflammatory reaction. Usually, the endometrial glands exhibit proliferative activity, and there may be mild glandular architectural distortion, in the form of occasional dilated glands. There is often associated surface breakdown with features identical to those seen in menstrual breakdown and breakdown due to non-menstrual causes. In some cases, an initial low-power clue to the diagnosis of endometritis is spindle cell alteration of the stroma (Fig. 26), although this feature is not specific and is not always present. In other cases, the stroma may be edematous. In acute endometritis, the predominant inflammatory cells are neutrophils, and collections of these may be seen within the glandular lumina, forming microabscesses (Fig. 27), or surrounding glands; neutrophils are often most easily seen just deep to the surface endometrium. In some cases, there is surface erosion with fibrinous debris and numerous acute inflammatory cells. By tradition, an unequivocal diagnosis of endometritis requires the presence of plasma cells (Fig. 28), since neutrophils are found normally in the endometrium just prior to and in association with menstruation, and lymphocytes, including lymphoid aggregates, are a normal component of the endometrial stroma. However, in acute forms of endometritis, plasma cells may be absent or few in number. Plasma cells are usually most numerous surrounding endometrial glands and just deep to the surface epithelium. A form of endometritis without plasma cells has been described and termed focal necrotizing endometritis (Bennett et al. 1999). The histological features of this are of focal, patchy inflammation comprising lymphocytes and neutrophils, centered on individual glands. Due to the focal nature, this form of endometritis can be easily overlooked. The clinical significance of focal necrotizing endometritis, if any, is not currently known.
Fig. 26

Endometritis. In some cases of endometritis, the endometrial stroma has a spindle cell appearance

Fig. 27

Acute endometritis. In acute endometritis, neutrophils are present, sometimes forming microabscesses within the glandular lumina

Fig. 28

Chronic endometritis. In cases of chronic endometritis, plasma cells are present within the endometrial stroma

Besides plasma cells, there are also increased numbers of lymphocytes in chronic endometritis, sometimes with prominent and unusually large lymphoid aggregates, and occasionally with the formation of lymphoid follicles. Other inflammatory cells, which may be a component of endometritis, include eosinophils and histiocytes. Usually, histiocytes are inconspicuous since they are admixed with other inflammatory cells, but occasionally large numbers of histiocytes, sometimes with abundant foamy cytoplasm, are present. When these are abundant, this is referred to as xanthogranulomatous endometritis (Fig. 29). Xanthogranulomatous endometritis may occur in association with an endometrial hyperplasia or carcinoma and secondary to cervical stenosis and obstruction.
Fig. 29

Xanthogranulomatous endometritis. Large numbers of foamy histiocytes are present within the endometrial stroma

In endometritis, reactive and metaplastic changes may involve the endometrial surface and glandular epithelium. Squamous, ciliated, eosinophilic, and other forms of epithelial metaplasia can occur, and there may be mild nuclear atypia with nuclear enlargement and prominent nucleoli. As stated, sometimes there are mild architectural changes with occasional dilated glands, but significant glandular crowding is not a feature of endometritis.

As stated, there may be problems in identifying plasma cells when they are few in number, especially in suboptimally stained sections. Endometrial stromal cells may have a plasmacytoid appearance, especially predecidualized cells in the mid-and late-secretory phase, and unequivocal plasma cells with eccentric nuclei and a perinuclear hof should be present. Occasional plasma cells may be seen in an otherwise normal endometrium, and, in the absence of at least some of the other features of endometritis described, a rigorous search for plasma cells is not justified. Plasma cells may be present in the stroma of an endometrial polyp, and this is not classified as an endometritis unless these are also seen in the non-polypoid endometrium. In problematic cases, histochemical or immunohistochemical stains may be of value in identifying plasma cells. Histochemical stains include methyl green–pyronin, and immunohistochemistry using plasma cell markers such as VS38 or syndecan (CD138) has been described (Bayer-Garner and Korourian 2001; Bayer-Garner et al. 2004; Leong et al. 1997). However, the clinical utility of this is limited. Immunohistochemistry or in situ hybridization for kappa and lambda immunoglobulin light chains may also help demonstrate plasma cells, but this is not routinely performed (Euscher and Nuovo 2002). Immunohistochemical staining with B lymphoid markers (CD20 and CD79a) may also assist in distinguishing between the physiological endometrial lymphocytic infiltrate and the inflammatory infiltrate of endometritis. Normally, the vast majority of lymphoid cells within the endometrial stroma are T cells (CD3 positive) with B lymphocytes accounting for about 1% of all endometrial leucocytes (Marshall and Jones 1988). B lymphoid cells are largely confined to lymphoid aggregates within the endometrial basalis, with occasional individual cells in the functionalis. In most cases of chronic endometritis, there are increased numbers of B lymphoid cells, and these also have an abnormal location being found outside lymphoid aggregates within the stroma and sometimes intraepithelially and within the glandular lumina (Disep et al. 2004). The number of T lymphocytes, histiocytes, and granulated lymphocytes in endometritis does not differ significantly from controls. It is emphasized that plasma cells do not usually stain with CD20 but are positive with CD79a.

Specific Forms of Endometritis

Chlamydia Trachomatis

Chlamydia trachomatis has been isolated from cases of both acute and chronic endometritis. However, it is unclear whether the organism is causative in such cases or an accompanying pathogen. Chlamydia trachomatis infection is relatively common in both the upper and lower female genital tracts and may be associated with pelvic inflammatory disease and infertility. Endometritis secondary to Chlamydia trachomatis has no specific histological features, although the inflammatory infiltrate may be intense and lymphoid follicles and large numbers of blasts may be seen (Paavonen et al. 1985). Stromal necrosis and reactive atypia of the endometrial epithelium may be present. Definitive diagnosis in most cases requires culture. However, Chlamydia trachomatis inclusion bodies have been identified within endometrial epithelial cells; these are extremely difficult to detect on hematoxylin and eosin-stained sections but can occasionally be recognized on Giemsa stain. Immunohistochemical staining may also be of value (Winkler et al. 1984), positivity being localized to the epithelial cells in the form of stippling within supranuclear intracytoplasmic vacuoles. Molecular investigations may also be of value in demonstrating Chlamydia trachomatis infection.


Cytomegalovirus (CMV) endometritis is rare. It is most common in immunosuppressed patients, but occasionally occurs in women with no underlying immune disorder. Typical intranuclear and cytoplasmic CMV inclusions are seen, mainly in epithelial cells but occasionally in stromal or endothelial cells. The other histological features are nonspecific; granulomas have been described in occasional cases (Frank et al. 1992).

Herpes Simplex Virus

Herpes simplex virus type II rarely results in an endometritis, usually secondary to ascending infection from the cervix and sometimes in women who are immunosuppressed. Typical herpes simplex virus inclusions are present within the glandular epithelium, and there are multinucleated cells with molded ground-glass nuclei. The other histological features are nonspecific, but patchy necrosis of the endometrial glands and stroma with an associated inflammatory infiltrate may occur (Duncan et al. 1989). Optically clear nuclei due to the accumulation of biotin, associated with the presence of trophoblast (see section “Gestational Endometrium”), may superficially resemble herpes simplex virus inclusions; while immunohistochemical staining may be of value in such instances, caution must be taken in their interpretation, as the biotin within these optically clear nuclei might result in a false-positive result.


Rarely mycoplasma organisms, usually Ureaplasma urealyticum, result in an endometritis. The inflammatory infiltrate is typically focal and has been termed “subacute focal endometritis” (Khatamee and Sommers 1989). The inflammatory cells comprise mainly of lymphocytes and histiocytes with few neutrophils and plasma cells, and they tend to be concentrated beneath the surface epithelium, adjacent to the glands, or around the spiral arteries. Granulomas have rarely been described.


The Gram-positive anaerobic bacterium Actinomyces may result in an endometritis, often in association with a long-term IUD (Gupta et al. 1976). The bacterial colonies form the so-called actinomycotic granules (AMGs), referred to as sulfur granules because of their tan to yellow color on gross examination. It is important to recognize Actinomyces, since the organism may result in ascending infection with resultant tubo-ovarian abscess formation and pelvic inflammatory disease. Histologically, AMGs are usually seen on the endometrial surface or within the superficial stroma as non-refractile granules with thin basophilic radiating filaments and sometimes a dense more eosinophilic granular core (Fig. 30). They are Gram positive on Brown and Brenn stain (Fig. 31) and are highlighted with Gomori methenamine silver stain. Although the diagnosis can be strongly suspected on morphological examination, culture is recommended for confirmation since other Gram-positive filamentous bacteria may be found in the gynecological tract. Because of the potential complications, Actinomyces must be distinguished from pseudoactinomycotic radiate granules (pseudo-sulfur granules) (Pritt et al. 2006). These are noninfectious lesions, most commonly seen in association with an IUD, but sometimes in non-IUD users. They consist of thick, irregular, club-like peripheral projections without a dense central core (Fig. 32). An associated inflammatory response may be present. With Brown and Brenn stains, there is diffuse, intense nonspecific staining, while silver stains are negative. Pseudoactinomycotic radiate granules are probably more common than Actinomyces (O’Brien et al. 1981), and occasionally the two coexist. The former may also be seen in association with pelvic inflammatory disease. The noninfectious nature of pseudoactinomycotic radiate granules is supported by microbiological and histochemical studies, as well as ultrastructural analysis. Their exact composition and nature is unknown, but they may represent an unusual response to foreign bodies (Splendore–Hoeppli phenomenon). Their main significance is that they may be mistaken for AMGs.
Fig. 30

Actinomycotic granules. Actinomycotic granules (AMGs) composed of thin basophilic radiating filaments with a more dense eosinophilic granular core

Fig. 31

Actinomycotic granule. Actinomycotic granules (AMGs) are Gram positive on Brown and Brenn stain

Fig. 32

Pseudoactinomycotic radiate granules. Pseudoactinomycotic radiate granules with thick, irregular, club-like peripheral projections without a central dense core

Fungi and Parasites

Fungi and parasites may rarely result in an endometritis, more commonly in underdeveloped countries. Blastomycosis (Blastomyces dermatitidis) and coccidioidomycosis (Coccidioides immitis) may result in an endometritis as part of a disseminated infection. Granulomas may be a component of the inflammatory infiltrate. There have been occasional reports of candidal and cryptococcal endometritis. Gomori methenamine silver and periodic acid–Schiff (PAS) stains are helpful in identifying these organisms.

Schistosoma, Enterobius vermicularis, and Echinococcus granulosus are rare causes of endometritis in developed countries, but schistosomiasis is endemic in some parts of the world. Schistosomal endometritis may be mild or severe and is characterized by granulomatous inflammation with lymphocytes, plasma cells, eosinophils, and histiocytes, sometimes closely simulating a tubercle. The endometrial surface may be ulcerated and replaced by granulation tissue. Diagnosis is made by identifying the ova in tissue sections or in smears of vaginal secretions. Toxoplasmosis (Toxoplasma gondii) evokes a nonspecific inflammatory reaction in the endometrium. The microorganism can be identified by immunofluorescence.


Malakoplakia may involve several organs, most commonly the urinary bladder, and is characterized by the presence of sheets of foamy histiocytes (von Hansemann’s histiocytes) containing Michaelis–Gutmann bodies. These are small, round, laminated calcospherites, which are present in the cytoplasm of the histiocytes and in an extracellular location. They contain calcium and can be demonstrated by von Kossa stain. The histiocytes are often admixed with other inflammatory cells, including plasma cells and neutrophils. Occasional examples have been reported in the endometrium (Thomas et al. 1978). Malakoplakia is a result of an abnormal immune response to bacteria, most commonly Escherichia coli, which are retained within the phagolysosomes of the histiocytes but are not digested; the Michaelis–Gutmann bodies are the result of mineral encrustation of incompletely digested bacteria.

Lymphoma-Like Lesion

The so-called lymphoma-like lesions are more common within the cervix, but have rarely been described in the endometrium (Young et al. 1985) (see chapter “Hematologic Neoplasms and Selected Tumor-Like Lesions Involving the Female Reproductive Organs”). Histologically, they are characterized by dense aggregates of lymphoid cells, often with large numbers of blasts and a starry-sky appearance, forming a superficial band-like infiltrate, this only being appreciated on a hysterectomy or hysteroscopic endometrial resection specimen. Lymphoid follicles with germinal centers, which may be large and ill-defined, are typically present. Along with germinal centers and large numbers of blasts, a mixed inflammatory infiltrate is present with small lymphocytes, plasma cells, neutrophils, and histiocytes. Lymphoma-like lesions represent an exaggerated form of chronic endometritis. The polymorphic nature of the infiltrate together with the presence of germinal centers and the superficial location of the inflammation (as stated, only appreciated on hysterectomy or endometrial resection specimens) help to distinguish lymphoma-like lesion from malignant lymphoma, as does the absence of a mass lesion grossly. Immunohistochemistry for kappa and lambda light chains or molecular investigations to demonstrate a polyclonal population may also be of value. Occasional cervical cases have been associated with Epstein–Barr virus infection (Young et al. 1985).

Endometrial Granulomas

Granulomas within the endometrium are rare. Worldwide, the most common cause is tuberculosis, and, although rare in developed countries, granulomatous endometritis should be considered as tuberculous in origin until proven otherwise. Tuberculosis of the endometrium usually occurs in premenopausal women and is rare after menopause. Caseous necrosis is characteristic of tuberculous granulomas, but due to the constant shedding associated with menstruation, endometrial granulomas in patients with tuberculosis are often noncaseating. Tubercle bacilli are seldom identified on Ziehl–Neelsen-stained sections, and culture should be undertaken in all cases in which histological examination raises the possibility of tuberculosis. Other infectious causes of granulomatous endometritis include various fungi, schistosomiasis, Enterobius vermicularis, and Toxoplasma gondii (see section “Fungi and Parasites”). As already discussed, granulomas are occasionally a feature of cytomegalovirus and mycoplasma endometritis.

Endometrial granulomas, especially when well circumscribed, also raise the possibility of sarcoidosis, and there have been rare reports of sarcoidosis involving the endometrium (Pearce and Nolan 1996). A granulomatous reaction to keratin may be seen in association with an endometrioid adenocarcinoma or atypical polypoid adenomyoma exhibiting squamous differentiation. Occasionally, such keratin granulomas may also be found on the surface of the ovaries, the fallopian tubes, or on the omentum or peritoneum; this is secondary to spread of keratin through the fallopian tubes and, in the absence of associated neoplastic cells, does not represent tumor spread. Foreign body granulomas in the endometrium may be secondary to talc or other substances and may be seen in association with an IUD. A palisading granuloma with fibrinoid material and a surrounding histiocytic and giant cell reaction, features resembling a rheumatoid nodule, may occur secondary to endometrial ablation; usually, the entire endometrium or much of the endometrium is affected, and well-circumscribed granulomas are not generally found (see section “Effects of Endometrial Ablation or Resection”). A similar picture may be seen secondary to endometrial resection. In rare instances, there is no obvious cause for endometrial granulomatous inflammation, so-called idiopathic granulomatous endometritis (Fig. 33).
Fig. 33

Granulomatous endometritis. A single granuloma is present within the endometrial stroma

Ligneous (Pseudomembranous) Endometritis

Ligneous (pseudomembranous) endometritis is discussed here, although the term endometritis is not strictly appropriate since there is often little or no inflammatory infiltrate. Ligneous disease is an inherited autosomal recessive condition characterized by absent or low plasminogen levels, resulting in the accumulation and deposition of fibrin. The histological features in the endometrium are identical to those of other affected sites, most commonly the conjunctiva. Ligneous disease is rare in the female genital tract and most commonly affects the cervix, vulva, or vagina. Rare endometrial cases have been described (Karaer et al. 2007; Scurry et al. 1993). When the endometrium is affected, this may result in dysmenorrhea and infertility. Histology shows amorphous eosinophilic material, somewhat similar to amyloid, which represents fibrin (Fig. 34); a Congo red stain would be negative. There may be associated mild inflammation, including occasional multinucleated giant cells. The inflammatory infiltrate may be more severe if there is surface ulceration.
Fig. 34

Ligneous (pseudomembranous) endometritis. Band of amorphous eosinophilic material represents fibrin deposition

Dysfunctional Uterine Bleeding

Dysfunctional uterine bleeding (DUB) is abnormal uterine bleeding in a premenopausal woman resulting from alterations in the normal cyclical changes of the endometrium, without an underlying specific pathological cause such as endometritis, polyps, exogenous hormones, hyperplasia, or carcinoma. In many cases, DUB is probably secondary to endogenous hormone imbalance. There are several morphological alterations of the endometrium that are characteristic of DUB, the most common being those associated with anovulatory cycles or luteal phase defects. These can be regarded as estrogen-related and progesterone-related, respectively, and are discussed in the next sections. Often DUB is managed by hormonal therapy, and a biopsy is only performed when symptoms persist; the hormone therapy may result in modification of the morphology. Importantly, DUB is not a pathological diagnosis but rather a clinical term.

A common, but not invariable feature of biopsies from patients with DUB is the presence of glandular and stromal breakdown. The features associated with this are not unique to DUB and are seen in menstrual endometrium and in bleeding associated with a variety of organic disorders. It is important to recognize the features of breakdown and to distinguish them from other pathological lesions. It is also important to realize that glandular and stromal breakdown is a nonspecific feature and that the intact endometrium must be assessed to evaluate the underlying abnormality. Glandular and stromal breakdown may also occur in an atrophic endometrium. The changes associated with glandular and stromal breakdown are described in the next paragraphs. In menstrual endometrium, the features of breakdown are diffuse and occur on a background of secretory endometrium. In contrast, in DUB the background endometrium is typically nonsecretory in type and breakdown is usually a focal phenomenon, resulting in a heterogeneous pattern with intact fragments of endometrium admixed with fragments exhibiting the features of breakdown. Furthermore, in menstrual endometrium the changes are acute and there are no features of chronic bleeding, such as hemosiderin deposition and accumulation of foam cells.

The morphological features associated with glandular and stromal breakdown are summarized in Table 2. An early feature is the accumulation of nuclear (apoptotic) debris in the basal cytoplasm of the glandular cells (Fig. 35) (Stewart et al. 1999). The stromal cells collapse and aggregate into tight clusters, which are separated by lakes of blood. These clusters of stromal cells, sometime called “stromal blue balls,” (Fig. 36) may form small polypoid extrusions or become detached from the surrounding tissue. They are characterized by tightly packed cells with hyperchromatic nuclei and scanty cytoplasm admixed with apoptotic debris. They can exhibit nuclear molding and may raise the possibility of small cell carcinoma. However, they are often covered by an epithelial lining, which may be flat or may exhibit the features of papillary syncytial metaplasia (see below), and are associated with other features of breakdown. Because of the stromal collapse, the endometrial glands become disrupted and crowded (Fig. 37). This glandular crowding may mimic hyperplasia or even adenocarcinoma, but recognition of the other features of breakdown facilitates the diagnosis. Fibrin thrombi are usually seen in small blood vessels, either in the spiral arteries or in superficial ectatic stromal vessels (Fig. 38). Another consistent feature of breakdown is papillary syncytial metaplasia (Fig. 9). Synonymous terms include eosinophilic syncytial change and surface syncytial change. The term papillary syncytial metaplasia is a misnomer since this is not a true metaplasia but has rather either been considered a regenerative reaction to surface breakdown or a degenerative or regressive process (Shah and Mazur 2008). Papillary syncytial metaplasia is characterized histologically by syncytial sheets of epithelial cells with indistinct cell borders that form micropapillary structures, sometimes with the small glandular lumina. The syncytia are devoid of stromal support and lack fibrovascular stromal cores. The cells usually have eosinophilic cytoplasm, and there is often a neutrophilic infiltrate. There may be mild nuclear atypia, and sometimes mitoses are identified.
Table 2

Features of endometrial breakdown

Glandular changes

Nuclear (apoptotic) debris in the basal cytoplasm of glandular cells

Papillary syncytial metaplasia

Glandular crowding

Stromal changes

Stromal collapse

Aggregates of stromal cells

Nuclear (apoptotic) debris in stroma

Fibrin thrombi

Hemosiderin pigment deposition

Foam cell accumulation

Fibrosis and hyalinization

Fig. 35

Glandular and stromal breakdown. An early feature is the accumulation of apoptotic debris in the cytoplasm of the glandular epithelial cells

Fig. 36

Endometrial breakdown. With breakdown, the stromal cells aggregate into stromal “blue balls”

Fig. 37

Endometrial breakdown. With breakdown, the endometrial glands become disrupted and crowded because of stromal collapse

Fig. 38

Endometrial breakdown. With breakdown, fibrin thrombi are typically seen within fibrin thrombi

Other features of breakdown, which are not present in all cases, include hemosiderin deposition, free within the stroma or within histiocytes (Fig. 39), and accumulation of foam cells, neither of which occurs in normal cyclical endometrium during the reproductive years. These features usually indicate chronic bleeding. The foam cells were initially thought to represent histiocytes, but it has been suggested that they may represent endometrial stromal cells, which become distended with lipid following erythrocyte breakdown (Fechner et al. 1979). Chronic bleeding may also occasionally result in focal stromal fibrosis and hyalinization.
Fig. 39

Chronic endometrial breakdown. With chronic breakdown, hemosiderin pigment may accumulate within histiocytes within the endometrial stroma

Estrogen-Related Dysfunctional Uterine Bleeding, Including Endometrium Associated with Anovulatory Cycles

This endometrial morphology is most common in the perimenopausal years, where it is usually secondary to anovulatory cycles with resultant absence of development of the corpus luteum and decrease in progesterone secretion; in fact, this is sometimes referred to as anovulatory endometrium. The term persistent proliferative endometrium has also been used. The underlying causes are complex, but many cases may be a result of hypothalamic dysfunction. The developing follicles persist for a variable period of time and produce estradiol before undergoing atresia, at which time estrogen withdrawal bleeding occurs. In other cases, estrogen breakthrough bleeding occurs when the persisting follicles produce estradiol resulting in the proliferating endometrium becoming thicker and outgrowing its blood supply. The usual presentation is with perimenopausal bleeding, but younger women may also be affected, for example, perimenarchal adolescents in whom regular ovulatory cycles are not established, those with polycystic ovarian syndrome (Stein–Leventhal syndrome), or older women taking unopposed estrogens or with an increase in endogenous estrogens, for example, secondary to obesity. Anovulatory cycles may also occur sporadically throughout the reproductive years. The histological features involve the entire endometrial compartment and are those of proliferative endometrium but with superimposed breakdown, as described above (Mutter et al. 2007). The extent of breakdown is highly variable and ranges from minute focal changes to a widespread phenomenon involving most of the specimen. There are often foci of ciliated (tubal) or other types of epithelial metaplasia that are randomly dispersed.

With chronic anovulatory cycles, there is abundant proliferative endometrium, and mild degrees of disorganization with dilated glands may occur. This results in a picture that is neither normal proliferative nor hyperplastic, which is referred to as disordered proliferative endometrium (Fig. 40). Occasional dilated glands within an otherwise normal proliferative endometrium do not warrant a diagnosis of disordered proliferation; in other words, there should be significant numbers of dilated glands and these should have a widespread distribution, although they are admixed with normal proliferative glands. In disordered proliferative endometrium, the normal gland to stroma ratio is largely maintained, although there may be focal mild glandular crowding and branching. In cases with significant numbers of dilated glands, the morphological appearances merge with and overlap with those of simple hyperplasia; in fact, disordered proliferative endometrium and simple hyperplasia almost certainly constitute a continuum, and it is likely that both represent a response to unopposed estrogens and not a true premalignant lesion, although there is a small increased risk of the development of an endometrioid adenocarcinoma. In simple hyperplasia, there is usually more glandular dilatation, with a paucity of normal proliferative glands, and the glands exhibit more budding and branching. However, there is significant interobserver variability in the distinction between disordered proliferative endometrium and simple hyperplasia, and, as stated, these form part of a continuum without sharply defined borders. Disordered proliferative endometrium may occasionally be confused with a polyp because of the glandular architectural distortion and dilatation; however, the fibrous stroma and thick-walled stromal blood vessels characteristic of a polyp are absent. Cystic atrophy may also enter into the differential diagnosis, but in this there is an absence of mitotic activity and the cells lining the glands are attenuated. Menstrual shedding following ovulation often results in disordered proliferative endometrium reverting to normal.
Fig. 40

Disordered proliferative endometrium. Occasional cystically dilated glands with abnormal glandular shapes are present within an otherwise typical proliferative endometrium

Progesterone-Related Dysfunctional Uterine Bleeding: Luteal Phase Defects

Luteal phase defects (also known as inadequate luteal phase, secretory insufficiency, or inadequate secretory phase) are a relatively common cause of DUB and also of ovulatory infertility. Although ovulation occurs, there is a relative or absolute insufficiency of progesterone secretion by the corpus luteum, which may either regress prematurely or fail to produce an adequate amount of progesterone. As a consequence, the luteal (secretory) phase does not develop appropriately, the secretory features in the endometrium being poorly developed. The underlying cause is unknown but is thought to be a result of hypothalamic or pituitary dysfunction, which results in decreased levels of follicle-stimulating hormone (FSH) and abnormal luteinizing hormone (LH) secretion. Because of inadequate progesterone secretion, there may be a lag in the histological date of the endometrium of at least 2 days compared to the actual postovulatory date. Other morphological features in some cases include discordance in development of the glands and stroma and different areas of the endometrium exhibiting marked variation in development; for example, some areas may exhibit early secretory activity while others show predecidual change. Alternatively, the glands may exhibit hypersecretory features while the stroma lacks predecidual change. Although the endometrium exhibits a secretory pattern, often this cannot be assigned to any day of the normal cycle. In addition to the features described, there is surface breakdown. The diagnosis can be made by a combination of repeated biopsies over several cycles and serial hormone measurements. In other cases, irregular shedding may be a result of a persistent corpus luteum with prolonged progesterone production. This is a poorly understood form of DUB, and, as such, the histological features are not well described.

Effects of Exogenous Hormonal Agents and Drugs

There are a plethora of exogenous hormonal agents in widespread use for a variety of indications including contraception, alleviation of menopausal symptoms, management of organic lesions or DUB, treatment of infertility, management of endometrial hyperplasia or carcinoma, and endometrial prophylaxis in patients with hyperestrogenic states or taking medications such as tamoxifen. The effects of the various hormonal agents on the endometrium are varied, although in many instances predictable, and depend on a number of factors, including the menopausal status of the patient, the exact composition of the hormonal preparation, and the dose and duration of administration. Not uncommonly, the endometrium is biopsied in patients taking exogenous hormones, for example, when abnormal bleeding occurs, when hormones do not correct suspected DUB, or when the effect of hormonal agents on the endometrium is assessed. Hysterectomy or endometrial resection may also be performed in patients taking exogenous hormones. Some nonhormonal agents may also result in endometrial morphological changes, although much less commonly than with hormonal compounds. In the following sections, the effects on the endometrium of the most common hormonal agents and of some nonhormonal medications are described. Full details regarding the preparations being taken are obviously of paramount importance to the pathologist in assessing the endometrium, but often the details are incomplete or not relayed to the pathologist at all. As such, the pathologist should always suspect the possibility of exogenous hormone use, especially when the endometrial morphology does not correspond to a normal cyclical or postmenopausal appearance.

Estrogen-Only Hormone Replacement Therapy (HRT)

There are various synthetic preparations of estrogens that are largely given to perimenopausal or postmenopausal women to treat menopausal symptoms. In women with a uterus, estrogen-only HRT (unopposed estrogen) is contraindicated due to the risk of endometrial proliferative lesions, including hyperplasia and endometrioid adenocarcinoma, and, as such, the use of estrogen-only preparations is unusual in a woman with a uterus. The morphological features in the endometrium vary, but there may be proliferative activity, a picture identical to disordered proliferation, any type of endometrial hyperplasia, or an endometrioid adenocarcinoma (The Writing Group for the PEPI Trial 1996). There may be associated surface breakdown and epithelial cytoplasmic change, including squamous and ciliated metaplasia. The risk of malignancy increases with the dose and duration of therapy; those adenocarcinomas that develop are usually, but not always, low grade and early stage. Unopposed estrogens result in endometrial hyperplasia in approximately 20% of women following 1 year of treatment. The “postmenopausal estrogen/progestin intervention (PEPI) trial” concluded that women taking estrogens alone had a high incidence of simple (27.7%), complex (22.7%), and atypical (11.7%) hyperplasia; this was significantly higher than in those taking placebos. The reported risk ratio for endometrial carcinoma in women taking unopposed estrogens has ranged from 2.3 to 10 (Grady et al. 1995; Paganini-Hill et al. 1989); the risk persists for many years after estrogen treatment is discontinued (Brinton and Hoover 1993; Shapiro et al. 1985). Estrogen-only preparations may also result in proliferative changes and the development of premalignant and malignant lesions in endometriosis; as such, caution should be exercised before prescribing unopposed estrogens following hysterectomy in a woman with known endometriosis.

Combined Estrogen and Progestin Hormone Replacement Therapy

Because of the potential adverse effects of unopposed estrogens, in most women with a uterus, an estrogen is combined with a progestin for HRT. Estrogen and progestin combinations may be given sequentially or continuously. Sequential (cyclic) regimes are variable but usually employ daily estrogens for the first 21 days, 25 days, or the whole of the month, with daily progestins added for the last 10–13 days; these regimes result in a withdrawal bleed. Continuous combined regimes use both estrogen and progestin daily. With continuous regimes, breakthrough bleeding may occur during the first 6 months, but this bleeding then usually stops. Patients receiving sequential or continuous regimes may undergo biopsy as part of routine surveillance or when unexpected bleeding occurs; there is no correlation between bleeding patterns and endometrial histology. With sequential regimes, the endometrium may exhibit atrophy, secretory, or weak proliferative activity, the latter especially if the biopsy is taken during the period of estrogen therapy. If the endometrium is biopsied during the period of progestin therapy, there may be poorly developed secretory activity in the glands with cytoplasmic vacuoles and scant luminal secretions. Focal glandular and stromal breakdown may also be seen. Sequential regimes do not completely eliminate the risk of carcinoma associated with unopposed estrogen therapy; the prevalence of endometrial hyperplasia associated with sequential HRT is 5.4% and that of atypical hyperplasia 0.7% (Sturdee et al. 2000). It should be remembered that HRT is most commonly taken by postmenopausal women, and in this age group, there is a background incidence of endometrial hyperplasia and carcinoma. With the continuous combined regimes, the endometrium is usually atrophic (Fig. 41) or exhibits weak secretory activity, and, in many instances, biopsies yield scanty material. There is no increased risk of endometrial proliferative lesions with continuous combined HRT (Nand et al. 1998), and, in fact, these regimes may protect against the development of endometrial hyperplasia and carcinoma and normalize endometria that have exhibited complex hyperplasia (Feeley and Wells 2001; Staland 1981; The Writing Group for the PEPI Trial 1996). However, data from the Women’s Health Initiative has shown an increased risk of breast and ovarian carcinoma in patients with long-term combined HRT (Anderson et al. 2003). Endometrial polyps are relatively common in women taking combined HRT and appear more common with sequential than continuous regimes (Feeley and Wells 2001).
Fig. 41

Endometrium associated with continuous combined hormone regimes. With continuous combined hormone regimes, the endometrium is usually atrophic

Progestin-Only Compounds

Various forms of synthetic analogues of progesterone, termed progestins, are in widespread use, either alone or in combination with an estrogen. Progestin-only hormonal compounds, taken either orally or systemically, are usually prescribed for abnormal uterine bleeding and result in suppression of ovulation and inhibition of endometrial growth. Progestins may also be given for the management of conditions such as endometriosis, for contraception, or for endometrial protection in patients taking tamoxifen. The effects of progestins on the endometrium are variable and depend on the degree of estrogen priming as well as the type of progestin and the dose and duration of therapy. They typically result in atrophy of the endometrial glands with predecidual change or decidualization (sometimes termed pseudodecidualization) of the stroma. The endometrial glands are most commonly small, tubular, and atrophic and lined by cuboidal cells with small round nuclei and scant cytoplasm, but sometimes exhibit poorly developed secretory activity. The stroma is usually expanded and composed of predecidualized or decidualized cells with abundant eosinophilic cytoplasm (Fig. 42) with infiltration by granulated lymphocytes; this may mimic endometritis, but plasma cells are absent. Marked decidualization is most common with high-dose progestins and may result in copious polypoid fragments of tissue being obtained at biopsy. These often exhibit superficial breakdown with associated neutrophil infiltration; again this may be mistaken for an endometritis. On occasions, some of the polypoid tissue fragments are totally necrotic, probably due to the stromal expansion with outgrowth of the blood supply. The decidualized stroma may, on occasion, contain cells exhibiting variation in nuclear size and shape with nuclear hyperchromasia; this may result in an alarming appearance, which may be exacerbated by stromal myxoid change and cytoplasmic vacuolation, resulting in a signet-ring appearance and mimicry of signet-ring carcinoma. When the progestin dose is low, the stromal cells may not exhibit predecidual or decidual change. The morphological effects associated with the Mirena coil, a progestin-containing intrauterine device, are described below (see section “Effects of Intrauterine Device (IUD)”).
Fig. 42

Endometrium associated with progestin-only compound. With progestin-only compounds, the stroma is typically expanded and composed of predecidualized cells

Progestin-Like Effect Without Exogenous Hormone Use

Occasionally, the endometrium exhibits changes identical to those occurring with progestins, namely, atrophic glands and stromal decidualization, in a patient not taking exogenous hormones. This may occur in both premenopausal and postmenopausal women (Clement and Scully 1988), and the etiology is, in most cases, poorly understood. Most other cases are probably secondary to a persistent functioning corpus luteum or a luteinized unruptured follicle where a follicle develops without ovulation and persists with luteinization of the granulosa and theca cells that produce progesterone. Alternatively, the changes may be a result of local mechanical factors rather than a response to progesterone-like hormones; mechanical stimulation, including biopsy or the presence of an intrauterine device, may rarely result in predecidual or decidual change in the endometrium. In very rare cases, the changes will be secondary to a progesterone secreting neoplasm in an ovary or elsewhere.

Gonadotropin-Releasing Hormone Agonists

Gonadotropin-releasing hormone agonists (GnRH agonists), including buserelin acetate, goserelin acetate, and leuprolide acetate, are commonly used in the management of uterine leiomyomas and endometriosis. After initial stimulation of the pituitary gland with increased production of luteinizing hormone (LH) and follicle-stimulating hormone (FSH), further administration results in desensitization of the pituitary to GnRH and a subsequent decreased production of LH and FSH. This results in decreased estrogen production by the ovaries and a hypo-estrogenic state. As a consequence, there is shrinkage of uterine leiomyomas, thus alleviating symptoms and potentially allowing myomectomy rather than hysterectomy or vaginal hysterectomy rather than abdominal hysterectomy. The endometrium in patients taking GnRH agonists is typically atrophic with small tubular glands, or sometimes the glands exhibit weak proliferative activity. When GnRH agonists are used in conjunction with a progestogen, the endometrium may exhibit decidualization of the stroma.


Several androgens are in widespread use, for example, danazol and tibolone. The main indications for these preparations are the treatment of endometriosis, but androgens may also be used in the management of menorrhagia or endometrial hyperplasia, or as HRT. In the early stages of androgen therapy, the glands may exhibit weak secretory activity, but with prolonged therapy, the endometrium becomes atrophic (Marchini et al. 1992).

Progesterone Receptor Modulators

Progesterone receptor modulators (PRMs) are synthetic compounds that interact with the progesterone receptor to inhibit or stimulate a downstream hormonal response. Compounds with progesterone receptor antagonist activity are used in contraception and in the management of uterine leiomyomas or endometriosis. The endometrial effects of PRMs have been described (Mutter et al. 2008). While in some cases the endometrium is inactive or has a normal cyclical appearance, in a subset of cases, there is asymmetry of stromal and epithelial growth, resulting in prominent cystically dilated glands with admixed estrogenic (mitotic) and progestogenic (secretory) activity. Vascular changes described include a chicken-wire vasculature and the presence of thick-walled and ectatic vessels. These novel changes are termed PRM-associated endometrial changes.


Tamoxifen is a nonsteroidal triphenylethyl compound that is widely used as adjuvant therapy in the treatment of breast cancer. It is a selective estrogen receptor modulator (SERM) and prolongs overall and disease-free survival in breast cancer, reduces the likelihood of disease in the contralateral breast, and may reduce the risk of development of breast cancer in asymptomatic women with a strong family history. Results from the National Surgical Adjuvant Breast and Bowel Project Breast Cancer Prevention study demonstrated that 5 years of tamoxifen use at a dose of 20 mg/day reduced breast cancer risk in high-risk women by 49% (Fisher et al. 1998).

The efficacy of tamoxifen in breast cancer is due to its antiestrogenic properties that are mediated by competitive binding to the estrogen receptor. However, tamoxifen may also exert a weak estrogenic effect and act on the human endometrium (Ismail 1994, 1999; Seidman and Kurman 1999), and the effects of tamoxifen on the endometrium appear to depend on the menopausal status as well as on the dose and duration of tamoxifen usage. Transvaginal ultrasonography has shown that the endometrium of tamoxifen-treated postmenopausal patients is significantly thicker than that of age-matched controls of women not taking tamoxifen (Cheng et al. 1997), and benign postmenopausal endometria of patients treated with tamoxifen exhibit a higher Ki-67 proliferation index than controls of patients not taking tamoxifen (Hachisuga et al. 1999). Tamoxifen therapy may result in a spectrum of endometrial proliferative lesions, including polyps; simple, complex, and atypical hyperplasia; and adenocarcinomas. Various other malignancies have also been described in association with tamoxifen. Since the majority of women taking tamoxifen are postmenopausal, most of the information regarding the endometrial side effects is related to this age group.

Nonneoplastic endometrium in postmenopausal patients taking tamoxifen is most commonly atrophic, while in other cases there is proliferative activity. The stroma is often fibrous, and, as a result, endometrial biopsies in patients taking tamoxifen are often very scanty. There may be glandular dilatation secondary to obstruction of the glands by stromal fibrosis. Stromal decidualization is usually secondary to simultaneous progestin administration (Cohen et al. 1996). One of the most characteristic and common endometrial lesions in women taking tamoxifen is polyps, which may be single or multiple and may occur on a background of hyperplasia, such that there is merging of polypoid and non-polypoid endometrium. There are no pathognomonic features of tamoxifen-associated endometrial polyps, but they tend to be larger than sporadic polyps. Periglandular stromal condensation, staghorn-shaped glands with intraglandular polypoid projections polarized along the long axis of the polyp, stromal edema and myxoid change, and epithelial metaplasias are all more common in tamoxifen-associated than sporadic polyps (Kennedy et al. 1999; Schlesinger et al. 1998), although all these features may be seen in sporadic polyps (Fig. 43). Cancers may develop in tamoxifen-associated polyps and can be of endometrioid or serous type. The presumed precursor lesion of serous carcinoma, serous endometrial intraepithelial carcinoma (serous EIC), may also involve tamoxifen-associated polyps (McCluggage et al. 2003b; Silva and Jenkins 1990). Rarely, a metastatic breast carcinoma, usually of lobular type, is identified within a tamoxifen-associated endometrial polyp (Houghton et al. 2003).
Fig. 43

Tamoxifen polyp. Staghorn-shaped glands with intraglandular polypoid projections, epithelial metaplastic changes, and focal stromal myxoid change are often present in tamoxifen polyps

As stated earlier, endometrial carcinomas may develop in association with tamoxifen. It is generally accepted that the risk of developing endometrial carcinoma in patients taking tamoxifen is two to three times higher than in an age-matched population of women not taking tamoxifen. It is probable that it is women who have been taking tamoxifen for a prolonged period of time and with a high cumulative dose who are at greatest risk. Since most endometrial cancers that develop in association with estrogenic stimulation are low-grade endometrioid in type, it might be expected that carcinomas associated with tamoxifen would have a similar profile. However, although some studies support this (Turbiner et al. 2008), it has also been suggested that high-grade endometrial cancers, including serous carcinomas and carcinosarcomas (in reality metaplastic carcinomas or carcinomas with sarcomatous metaplasia) (McCluggage 2002), are more common in patients taking tamoxifen and that the development of these cancers is not secondary to estrogenic stimulation but due to other mechanisms such as the formation of DNA adducts (McCluggage et al. 2000c; Shibutani et al. 2000).

Occasional uterine leiomyosarcomas, endometrial stromal sarcomas, adenofibromas, and adenosarcomas have been described in patients taking tamoxifen (Clement et al. 1996; Huang et al. 1996; McCluggage et al. 1996); it is possible that these complications arise simply by chance since tamoxifen is widely used. Adenomyosis has been shown to be more common in postmenopausal patients taking tamoxifen (Cohen et al. 1997), and this may exhibit unusual morphological features, such as stromal fibrosis, glandular dilatation, and epithelial metaplasias (McCluggage et al. 2000a). A rapid increase in size of uterine leiomyomas has been observed with tamoxifen therapy.

Other SERMs, such as raloxifene, have a similar efficacy to tamoxifen in the management of breast cancer. In contrast to tamoxifen however, these seem to be pure estrogen antagonists, lacking the weak estrogen agonist effects of tamoxifen, and do not result in endometrial proliferative lesions (Delmas et al. 1997).


Taxanes, such as paclitaxel, are commonly used chemotherapeutic agents as first-line treatment in ovarian, breast, and lung cancer. Taxanes act by the simultaneous promotion of tubulin assembly into microtubules and inhibition of microtubule disassembly. The morphological features in the endometrium have been rarely reported, and numerous mitoses in metaphase arrest with the formation of ring mitoses have been described (Irving et al. 2000). Similar morphological changes are seen in other organs, such as the gastrointestinal tract, in association with taxanes.

Endometrial Epithelial Metaplasia (Epithelial Cytoplasmic Change)

Endometrial epithelial metaplasias are nonneoplastic epithelial alterations (metaplasias may coexist with endometrial hyperplasia or carcinoma but, in themselves, are nonneoplastic) in which the normal endometrial epithelium is replaced, focally or diffusely, by another type of differentiated epithelium. The variety of epithelial metaplasias encountered within the endometrium reflects the capacity of epithelium derived from the müllerian ducts to undergo differentiation into any other form of epithelium found in the müllerian system. The various epithelial metaplasias commonly coexist. It has been suggested that metaplasia is an inappropriate term for some of the alterations as, strictly speaking, the term metaplasia refers to the replacement of one type of epithelium by another that is not normally found in that organ (Hendrickson and Kempson 1980). For this reason, some authors use the term “epithelial cytoplasmic change” rather than metaplasia.

Metaplasia usually involves nonsecretory endometrium and is often associated with hyperestrogenism. Metaplasias are common within endometrial polyps. Other associations include exogenous hormone therapy, especially but not exclusively unopposed estrogens, the presence of an intrauterine device, chronic endometritis, and pyometria; the latter two conditions are particularly associated with squamous metaplasia. In some cases, there is no obvious underlying cause. It has been suggested that progestin therapy given for endometrial hyperplasia or endometrioid adenocarcinoma may result in various epithelial metaplasias within the malignant or premalignant lesion (Wheeler et al. 2007). Metaplasia by itself is not associated with clinical symptoms, but if there is an associated endometrial hyperplasia or carcinoma, there may be abnormal bleeding (McCluggage 2003).

A particular problem with epithelial metaplasias is their tendency to be associated with endometrial hyperplasias (Carlson and Mutter 2008) or endometrioid adenocarcinoma. Squamous metaplasia and mucinous metaplasia are particularly common in endometrioid adenocarcinomas. In such instances, it may be difficult, at the lower end of the spectrum, to distinguish between a metaplasia with mild glandular complexity and a hyperplasia with coexistent metaplasia. It is important to make this distinction, since metaplasia by itself has no premalignant potential, and it is recommended that similar criteria are employed to those that are used in the diagnosis of endometrial hyperplasia without metaplasia (see chapter “Precursor Lesions of Endometrial Carcinoma”).

With some epithelial metaplasias, such as clear cell metaplasia and papillary syncytial metaplasia, the differential diagnosis may be between a metaplasia and a type 2 endometrial cancer or serous endometrial intraepithelial carcinoma (serous EIC). In such instances, immunohistochemistry may be of value in that serous EIC often exhibit diffuse, intense, nuclear p53 immunoreactivity while ER may show decreased expression. In contrast, most epithelial metaplasias are ER positive and exhibit a pattern of p53 immunoreactivity, which has been described as weak and heterogeneous (Quddus et al. 1999). A minor population of endometrial epithelial cells exhibit nuclear immunoreactivity with p63; it has been speculated that these are reserve cells or basal cells and the origin of the various epithelial metaplasias (O’Connell et al. 2001).

Squamous Metaplasia

Squamous metaplasia is one of the most common forms of endometrial epithelial metaplasia. Although usually a focal finding, there may occasionally be widespread squamous metaplasia with obliteration of the glandular lumina, such that it is difficult to assess the underlying glandular component. This is especially common when the squamous metaplasia is of morular type (see below). Squamous metaplasia is common in endometrioid adenocarcinoma and in endometrial hyperplasias; these should be excluded by careful examination of the glandular elements. It may also be seen in endometrial polyps.

There are two types of squamous metaplasia, namely, typical squamous metaplasia and morular metaplasia, although these sometimes coexist. Typical squamous elements are characterized by sheets of cells exhibiting obvious squamous differentiation in the form of intercellular bridges, prominent cell membranes, or keratinization (Fig. 44). Sometimes there is a histiocytic and giant cell reaction to keratin. Typical squamous metaplasia rarely involves much of or all of the endometrial surface, such that the endometrial cavity is extensively lined by squamous epithelium, a condition known as ichthyosis uteri. This condition most commonly develops secondary to longstanding cervical obstruction or chronic inflammation. Usually, in ichthyosis uteri, the squamous epithelium is of normal appearance, but rarely, it may have the features of a condyloma acuminatum or intraepithelial neoplasia (Stastny et al. 1995). It may rarely extend to involve the fallopian tubes and ovaries. Rarely an invasive squamous carcinoma develops in this manner (Takeuchi et al. 2012).
Fig. 44

Squamous metaplasia in the endometrium. Typical squamous metaplasia with obvious squamous differentiation in the form of prominent cell membranes

Squamous morules are morphologically distinct structures, which were named so because of their three-dimensional resemblance to mulberries (Dutra 1959). They are composed of rounded aggregates or syncytial sheets of cells that often fill the glandular lumina (Fig. 45). The constituent cells have central bland round, ovoid, or spindle shaped, evenly spaced nuclei, sometimes with small nucleoli. Some of these nuclei may contain optically clear biotin-rich inclusions. The cell borders are indistinct. Mitoses are rare or absent. There may be central necrosis. It is controversial whether morules actually exhibit squamous differentiation. Morphological features of overt squamous differentiation, such as keratinization, intercellular bridges, and prominent cell membranes, are typically absent in morules. Immunohistochemically, morules exhibit a somewhat different immunophenotype to that of typical squamous elements. Morules exhibit nuclear and cytoplasmic positivity with beta-catenin (Fig. 46) (Brachtel et al. 2005; Saegusa and Okayasu 2001), while in typical squamous elements, the “normal” membranous pattern of immunoreactivity is maintained (Houghton et al. 2008). Endometrial proliferative lesions with morules often exhibit beta-catenin gene mutation, resulting in the above-mentioned nuclear and cytoplasmic immunoreactivity. Morules are usually estrogen receptor (ER) and p63 negative (Chinen et al. 2004) and diffusely positive with CD10 (Chiarelli et al. 2006) and exhibit nuclear immunoreactivity with the intestinal transcription factor CDX2 (Fig. 47) (Houghton et al. 2008; Wani et al. 2008); it has been suggested that this is secondary to beta-catenin gene mutation. In contrast, typical squamous elements are usually positive with ER, p63, and CD10 and negative with CDX2. On the basis of the immunophenotype, it has been concluded that morules exhibit no firm immunohistochemical evidence of squamous differentiation, although immature squamous features cannot be excluded (Houghton et al. 2008). It has been suggested that the term morular metaplasia is used instead of squamous morules.
Fig. 45

Squamous morules in the endometrium. Squamous morules are composed of rounded aggregates or syncytial sheets of cells filling the glandular lumina

Fig. 46

Beta-catenin immunohistochemistry in squamous morules. Morules exhibit nuclear and cytoplasmic immunoreactivity with beta-catenin

Fig. 47

CDX2 immunohistochemistry in squamous morules. Morules exhibit nuclear immunoreactivity with the intestinal transcription factor CDX2

Mucinous Metaplasia

Mucinous metaplasia is a relatively uncommon form of endometrial epithelial metaplasia and is most commonly seen either in endometrial polyps or in association with a premalignant or malignant lesion. A diagnosis of mucinous metaplasia should be reserved for cases in which the endometrial epithelial cells are replaced by cells with abundant intracytoplasmic mucin, the cells resembling endocervical cells (Fig. 48). Normal endometrial epithelial cells contain a little intracytoplasmic mucin, especially with a luminal distribution, and so abundant intracytoplasmic mucin is required to diagnose mucinous metaplasia. Rarely, intestinal metaplasia has been described in the endometrium where the mucinous epithelium contains goblet cells (Wells and Tiltman 1989). Enteric-type mucin may be demonstrable with mucin stains in cases without morphological evidence of intestinal metaplasia (McCluggage et al. 1995).
Fig. 48

Mucinous metaplasia in the endometrium. Focally the cells have abundant mucinous cytoplasm

In mucinous metaplasia without an associated premalignant or malignant glandular proliferation, there are often small micropapillary projections. The nuclei are small and uniform, and mitoses are rare or absent. An important point with a florid mucinous proliferation of the endometrium is that mucinous adenocarcinomas, even those exhibiting myometrial invasion, can be cytologically bland with little in the way of mitotic activity. As such, complex mucinous proliferations of the endometrium present a particular diagnostic problem, especially, but not exclusively, in biopsy material (Vang and Tavassoli 2003). Mucinous proliferations of the endometrium have been divided into three categories depending on the degree of architectural complexity and association with underlying adenocarcinoma (Nucci et al. 1999). With any architecturally complex mucinous proliferation in an endometrial biopsy, a diagnosis of well-differentiated mucinous adenocarcinoma should be considered, and the term “complex endometrial mucinous proliferation” may be applied with a comment that there is a significant risk of a well-differentiated mucinous adenocarcinoma in the uterus. Rarely, mucinous metaplasia in the endometrium is accompanied by mucinous lesions elsewhere in the female genital tract, for example, in the cervix and the ovary, perhaps as a manifestation of a field-change effect (Baird and Reddick 1991), or in the context of patients with Peutz–Jeghers syndrome (Tantipalakorn et al. 2009).

Ciliated (Tubal) Metaplasia

Ciliated epithelial cells are normal on the endometrial surface, especially in the proliferative phase of the menstrual cycle. A diagnosis of ciliated metaplasia should be made only when one or more endometrial glands contain ciliated cells, which may be interspersed among non-ciliated cells, or may be extensive and line most of the gland (Fig. 49). The nuclei are cytologically bland, may be rounded and mildly stratified, and contain small nucleoli; the ciliated cells often have abundant eosinophilic cytoplasm. Ciliated metaplasia is particularly associated with estrogenic stimulation. As with other types of epithelial metaplasia, ciliated cells may be found in nonneoplastic, hyperplastic, and malignant endometria. The presence or absence of hyperplasia or adenocarcinoma is evaluated by the usual parameters.
Fig. 49

Ciliated metaplasia in the endometrium. The endometrial glands are lined by ciliated cells with abundant eosinophilic cytoplasm

Clear Cell Metaplasia

Clear cell metaplasia is rare and is characterized by replacement of endometrial epithelial cells by cells with abundant clear cytoplasm (Fig. 50). This may be a feature of pregnancy when the features overlap with Arias-Stella reaction. Clear cell metaplasia may be misdiagnosed as clear cell carcinoma, especially on a biopsy specimen. Distinction is based on the bland nuclear features and the fact that in clear cell metaplasia, the endometrial glands maintain a normal architecture and distribution. Other features favoring clear cell metaplasia over clear cell carcinoma include the focal nature of the lesion, absence of a grossly visible tumor, absence of stromal invasion, and strong estrogen receptor positivity. A significant number, but not all endometrial clear cell carcinomas, are estrogen receptor negative.
Fig. 50

Clear cell metaplasia in the endometrium. The endometrial glands are replaced by cells with abundant clear cytoplasm

Hobnail Cell Metaplasia

Hobnail cell metaplasia or change is rare and is characterized by the presence of cells with rounded apical blebs, which may involve the endometrial surface or protrude into the glandular lumina (Fig. 51). Hobnail cell metaplasia may be a reparative phenomenon following endometrial curettage or may be seen on the surface of a polyp. It may also occur in pregnancy. Hobnail cells are also a feature of some clear cell carcinomas, and this may enter into the differential diagnosis. Criteria useful in the distinction of hobnail cell metaplasia from clear cell carcinoma are similar to those used in distinction of the latter from clear cell metaplasia.
Fig. 51

Hobnail cell metaplasia in the endometrium. Hobnail cells are present on the surface of an endometrial polyp

Eosinophilic (Oxyphilic, Oncocytic) Metaplasia

Eosinophilic or oxyphilic metaplasia is relatively common and is characterized by the presence of epithelial cells with abundant eosinophilic cytoplasm (Fig. 52). The cytoplasm may be granular, in which case the term oncocytic metaplasia has been used. The term pink cell metaplasia has also been used. Ultrastructurally, abundant cytoplasmic mitochondria may be present, as is characteristic of oncocytes in other organs. Ciliated metaplasia is often characterized by abundant eosinophilic cytoplasm and overlaps with eosinophilic metaplasia. The epithelial cells in eosinophilic metaplasia can exhibit a significant degree of nuclear atypia; this is analogous to the degenerative nuclear atypia that is common in oncocytic cells in other organs. The main differential diagnosis is the eosinophilic or oxyphilic variant of endometrioid adenocarcinoma. Distinction from adenocarcinoma is based on the absence of a grossly visible lesion and maintenance of the normal glandular architecture.
Fig. 52

Eosinophilic metaplasia in the endometrium. The epithelial cells contain abundant eosinophilic cytoplasm

Papillary Syncytial Metaplasia

The term papillary syncytial metaplasia is a misnomer since this does not actually represent a metaplasia but rather a degenerative or reparative phenomenon associated with surface breakdown, either menstrual or non-menstrual in type. However, since the term papillary syncytial metaplasia is in widespread use, it is discussed here. Synonymous terms include eosinophilic syncytial change and surface syncytial change. It has been suggested that papillary syncytial metaplasia is a degenerative or regressive phenomenon based on a low proliferation and mitotic index (Shah and Mazur 2008). Papillary syncytial metaplasia is common and is characterized by small syncytia or micropapillary proliferations of endometrial epithelial cells, which may contain the small glandular lumina and which are devoid of stromal support, lacking fibrovascular stromal cores (Fig. 9). The cells usually have eosinophilic cytoplasm, and there is often a neutrophilic infiltrate. There may be mild nuclear atypia, and in a minority of cases, mitoses are present. The distinction between papillary syncytial metaplasia and serous adenocarcinoma or serous EIC has been discussed earlier. Another important consideration is that foci similar to papillary syncytial metaplasia may occur on the surface of some endometrioid adenocarcinomas. The distinction between papillary syncytial metaplasia and papillary adenocarcinomas of endometrioid or serous type is facilitated by recognition that papillary syncytial metaplasia is limited to the endometrial surface and is associated with other morphological features of breakdown such as apoptotic debris, neutrophils, and adjacent glandular and stromal breakdown.

Arias-Stella Reaction

Arias-Stella reaction (Arias-Stella effect or change) has been discussed earlier and is almost always associated with pregnancy, either intrauterine or ectopic, or with trophoblastic disease. It rarely occurs secondary to hormone therapy, especially progestins; occasionally there is no obvious cause (Dhingra et al. 2007). The most important differential diagnosis is clear cell carcinoma, but the diagnosis of Arias-Stella reaction is usually straightforward if the patient is known to be pregnant and if other morphological features of pregnancy are present, such as decidualization of the stroma. Arias-Stella reaction involves preexisting endometrial glands without evidence of stromal infiltration and there is no mass lesion. Although there is nuclear enlargement and atypia, a low nuclear-to-cytoplasmic ratio is maintained.

Papillary Proliferation of the Endometrium

The term hyperplastic papillary proliferation of the endometrium has been used for a lesion, usually occurring in postmenopausal women, characterized by the presence of papillae with fibrovascular stromal cores and variable degrees of branching and cellular tufting (Fig. 53). The papillae are lined by epithelial cells with bland nuclei (Lehman and Hart 2001). Although not strictly a metaplasia, the lesion is discussed here since epithelial metaplasias, most commonly mucinous, eosinophilic, or ciliated, are often also present. Sometimes the papillae are entirely intracystic (projecting into cystically dilated endometrial glands), while in other instances they involve the endometrial surface. Papillary proliferation is most commonly seen on the surface of an endometrial polyp, and, in some instances, the features are florid. There may be an association with hormonal preparations. A misdiagnosis of an adenocarcinoma of endometrioid or serous type is possible, especially if an underlying polyp is not present or not obvious. Awareness of this phenomenon and the realization that it often occurs in a polyp are clues to the diagnosis, although both endometrioid and serous adenocarcinomas may arise in and be confined to a polyp. The absence of nuclear atypia helps to exclude an adenocarcinoma. It has been considered that these papillary proliferations are a form of hyperplasia that is closely associated with epithelial metaplasia.
Fig. 53

Papillary proliferation of the endometrium. Papillary projections lined by bland epithelial cells on the surface of an endometrial polyp

These papillary proliferations can be subdivided into two separate categories, based on the degree of architectural complexity and proliferation (Ip et al. 2013). The first group, characterized by simple, short papillae without significant branching, is usually associated with a benign outcome, and the term “benign papillary proliferation of the endometrium” has been proposed to describe this finding. The second group, characterized by a more complex papillary proliferation, which is often more diffuse or multifocal, has an increased risk of recurrence or concurrent adenocarcinoma, and the term “complex papillary hyperplasia” is appropriate.

Endometrial Mesenchymal Metaplasia

Various forms of mesenchymal metaplasia, all of which are rare, may involve the endometrial stroma. There are two theories regarding the pathogenesis of the various mesenchymal metaplasias. They may result from metaplasia of the endometrial stroma or represent remnants of fetal tissue following abortion or instrumentation. They should be distinguished from heterologous mesenchymal components within a carcinosarcoma or another uterine neoplasm.

Smooth Muscle Metaplasia

This is the most common mesenchymal metaplasia in the endometrium. Given the common embryonic origin of endometrial stromal and smooth muscle cells, it is thought that a multipotential cell exists in the uterus, which has the capacity to differentiate into endometrial stroma and smooth muscle. This is in keeping with the observation that hybrid endometrial stromal and smooth muscle neoplasms exist. It is not uncommon to find small foci of smooth muscle within the endometrial stroma, and these foci have sometimes been referred to as intraendometrial leiomyomas. It is probable that some intraendometrial smooth muscle nodules are a result of the irregular nature of the normal endometrial–myometrial junction, but others reflect the capacity of endometrial stroma to differentiate into smooth muscle.

Cartilaginous and Osseous Metaplasia

Rarely, foci of benign cartilage or bone are found within endometrial biopsies, either within the endometrial stroma or free floating (Bahceci and Demirel 1996). In most cases, it is likely that this is of fetal origin, and this is especially likely if these tissues are found in the endometrium of a young woman with a past history of abortion. In other cases, the cartilage or bone is truly metaplastic. These tissues may also rarely be found within the stroma of an endometrial carcinoma. Benign cartilaginous or osseous metaplasia in an endometrial carcinoma should not be mistaken for the heterologous sarcomatous component of a carcinosarcoma. Rarely, endometrial ossification is associated with Asherman’s syndrome.

Glial Metaplasia

The presence of glial tissue in the endometrium is extremely rare. In most cases, glial tissue (confirmed if necessary by positive immunohistochemical staining with glial fibrillary acidic protein (GFAP)) is a consequence of a previous abortion. Support for this may come from the identification of other elements such as cartilage or bone.

Adipose Metaplasia

Metaplastic adipose tissue is rarely found within the endometrial stroma or within the stroma of an endometrial polyp. If identified in an endometrial biopsy or curettage specimen, the possibility of uterine perforation must be raised. Other explanations for the presence of adipose tissue in an endometrial biopsy specimen include derivation from a lipoma, a lipoleiomyoma, a uterine hamartomatous-like lesion containing adipose tissue (McCluggage et al. 2000b), or a carcinosarcoma.

Extramedullary Hematopoiesis

Rarely, foci of extramedullary hematopoiesis are present in the endometrium, usually in association with an underlying hematopoietic disorder or occasionally representing remnants of fetal tissue (Creagh et al. 1995).

Endometrial Polyps

Endometrial polyps are common and have been identified in between 2% and 23% of patients undergoing endometrial biopsy because of abnormal uterine bleeding (Schindler and Schmidt 1980). Polyps occur in pre- and postmenopausal women and are thought to be related in some way to hyperestrogenism, possibly originating as a localized hyperplasia of the endometrial basalis secondary to hormonal influences. There is an increased incidence of endometrial polyps with hormone replacement therapy, either estrogen-only HRT or combined preparations. Tamoxifen is also associated with an increased risk of the development of endometrial polyps (see section “Tamoxifen”). Molecular studies have demonstrated that many endometrial polyps represent monoclonal endometrial stromal overgrowths (commonly with abnormalities of chromosome 6) with secondary induction of polyclonal benign glands through undefined stromal–epithelial interactions (Dal Cin et al. 1992; Fletcher et al. 1992).

Polyps may be single or multiple, sessile or broad based, and pedunculated or attached to the endometrium by a slender stalk. They usually have a smooth surface, and small cysts may be seen on sectioning. They can arise anywhere in the endometrium, including the lower uterine segment, but are most common in the fundus. When large, they may fill the endometrial cavity and extend into the endocervical canal.

The pathological diagnosis is generally straightforward if the gynecologist is aware of the presence of a polyp, has conveyed this information to the pathologist, and has removed the polyp intact. On occasion, the gynecologist believes that a polyp is present, but histological examination shows a cyclical endometrium, often secretory in type, reflecting the fact that an abundant secretory endometrium may have a polypoid appearance. In many cases, the gynecologist is not aware of the presence of a polyp, which is removed piecemeal with the result that in biopsy material, fragments derived from the polyp are admixed with fragments of non-polypoid endometrium, making the diagnosis difficult. In biopsies performed because of abnormal uterine bleeding, the pathologist should always consider the possibility of a polyp. Under low-power examination, the initial clue to the diagnosis is often the admixture of fragments of normal cyclical or atrophic endometrium and fragments that are morphologically different. The histological features of a polyp, not all of which are present in every case, include the following:
  1. 1.

    Polypoid pieces of tissue lined by epithelium on three sides.

  2. 2.

    Glands set in a stroma that is qualitatively different than the endometrial stroma in the non-polypoid fragments. The stroma is often, but not always, more fibrous than that in the non-polypoid fragments and is sometimes markedly hyalinized (Fig. 54).

  3. 3.

    Glandular architectural abnormality with dilated glands and sometimes mild glandular crowding (Fig. 55).

  4. 4.

    Glands that appear different to those in the surrounding endometrium; for example, the glands of the non-polypoid endometrium may be secretory in type while the glands within the polyp are atrophic or exhibit poorly developed secretory or proliferative activity.

  5. 5.

    Collections of thick-walled stromal blood vessels (Fig. 56).

Fig. 54

Endometrial polyp. Dilated glands are set in a fibrous stroma

Fig. 55

Endometrial polyp. There may be a mild degree of glandular crowding within some endometrial polyps

Fig. 56

Endometrial polyp. Collections of thick-walled stromal blood vessels are a characteristic feature of endometrial polyps

The glands within a polyp are usually endometrioid in type, but not uncommonly exhibit metaplastic change, including ciliated, eosinophilic, mucinous, and squamous metaplasia. The epithelium may be atrophic, but often exhibits proliferative activity, even when the patient is postmenopausal and the surrounding endometrium is atrophic. The presence of proliferative activity in a polyp in a postmenopausal woman is of no clinical importance, although it is useful to comment in the pathology report on whether non-polypoid endometrium is also present and whether this exhibits proliferative activity. The stroma of a polyp is often more fibrous than that of the non-polypoid endometrium, but this is not invariable, and, in some polyps, the stroma is dense and cellular, resembling that of normal proliferative endometrium. As stated, collections of thick-walled stromal blood vessels are a characteristic feature of endometrial polyps, and ectatic thin-walled vessels are also sometimes seen.

Some authors have divided endometrial polyps into different types, such as proliferative/hyperplastic (proliferative glands with or without glandular crowding), atrophic (atrophic glands), and functional (glands resembling those in the surrounding cyclical endometrium). However, these patterns often overlap, and assignment to a specific type may be difficult; moreover, there is no clinical significance attached to the different types. Some polyps originate at the junction of the upper endocervix and lower uterine segment and contain both endocervical and ciliated lower uterine segment type glands.

The variety of morphological appearances affecting the epithelium or stroma that may be encountered in endometrial polyps can result in diagnostic difficulty. Papillary proliferations with fibrovascular cores (see section “Papillary Proliferation of Endometrium”) occasionally occur on the surface of an endometrial polyp or within cystically dilated glands. The epithelium on the surface of a polyp may exhibit a degree of atypia, often with degenerate-appearing nuclei (Fig. 57), and sometimes hobnail cell change. There may be focal surface glandular and stromal breakdown, and on occasions, polyps are extensively necrotic secondary to torsion or if they outgrow their blood supply (Fig. 58); vascular thrombosis may be seen in such cases. In some instances, this may result in the formation of a necrotic polypoid mass with only the surface epithelium or the ghost outlines of glands remaining. Variable amounts of stromal edema and occasionally myxoid change may be present as well as hemosiderin pigment and foamy histiocytes. Sex cord-like areas have rarely been described in the stroma of an endometrial polyp (De Quintal and De Angelo Andrade 2006). Some endometrial polyps contain bundles of smooth muscle within the stroma, often in close proximity to thick-walled blood vessels. This is usually a minor feature and of no significance. However, when the smooth muscle is prominent, the term adenomyomatous polyp has been used (see section “Adenomyomatous Polyp”). Stromal inflammatory cells, including plasma cells, may be present in endometrial polyps; this should not be interpreted as an endometritis unless plasma cells are also present in the non-polypoid endometrium. Stromal decidualization or pseudodecidualization may occur secondary to progestational compounds, but the degree of decidualization is typically less than in the surrounding endometrium.
Fig. 57

Endometrial polyp. The epithelium on the surface of a polyp may exhibit a degree of nuclear atypia with a degenerate appearance

Fig. 58

Necrotic endometrial polyp. Necrosis has occurred secondary to torsion

As discussed previously, there is an increased frequency of endometrial polyps in patients taking tamoxifen. The polyps may be single or multiple and may be large. There are no pathognomonic histological features of tamoxifen-associated endometrial polyps, but an increased incidence of epithelial metaplasias, periglandular stromal condensation, stromal edema, myxoid change, and staghorn-shaped glands with intraluminal polypoid projections, polarized along the long axis of the polyp, have been reported (Kennedy et al. 1999).

As stated, diagnosing an endometrial polyp is generally straightforward when the polyp is large and removed intact. However, when small and fragmented, the diagnosis is more difficult. Lower uterine segment endometrium may be mistaken for a polyp because of the irregular glandular architecture and fibrous stroma. The spindle cell alteration of the stroma seen in some cases of endometritis may resemble the fibrous stroma of a polyp. However, other morphological features of a polyp are absent, and there is a plasma cell infiltrate within the stroma; it should be remembered, however, that plasma cells may occur within the stroma of an endometrial polyp. In cases of large polyps with a degree of stromal condensation and increased cellularity around the glands, the differential diagnosis may include an adenosarcoma. Adenosarcoma typically has a leaf-like or club-like architecture, with broad papillae lined by surface epithelium, and intraglandular stromal projections, the overall architecture resembling a phyllodes tumor of the breast. In contrast, endometrial polyps usually have a smooth outline. The stroma in adenosarcoma is usually more cellular than in a benign polyp with increased mitotic activity and a degree of nuclear atypia, especially immediately surrounding the glands. With multiple recurrent endometrial polyps, a diagnosis of adenosarcoma should be suspected since the morphological features may be subtle. Adenofibroma may also be considered, but this is rare and exhibits a similar low-power architecture to adenosarcoma. In polyps with a stromal smooth muscle component, an atypical polypoid adenomyoma may be considered. However, the stroma of atypical polypoid adenomyoma exhibits more extensive smooth muscle differentiation, the glandular architecture is more complicated, and there is often extensive squamous morule formation (see section “Atypical Polypoid Adenomyoma”). Polyps in which the glands are mildly crowded and exhibit proliferative activity may be confused with an endometrial hyperplasia. In such cases, the identification of normal background endometrium facilitates the diagnosis, since hyperplasia is usually a diffuse process.

Endometrial Polyp with Atypical Stromal Cells

Rare endometrial polyps contain stromal cells with markedly atypical symplastic-like nuclei, resembling those seen in polypoid lesions elsewhere in the female genital tract, such as in fibroepithelial stromal polyps of the vulva and vagina (Tai and Tavassoli 2002). These atypical stromal cells are of no significance.

Adenomyomatous Polyp

As discussed, some endometrial polyps contain a minor component of stromal smooth muscle bundles, often in close proximity to thick-walled blood vessels. When the smooth muscle is prominent, the term adenomyomatous polyp has been used. This term and the term adenomyoma have also been used for a lesion in which endometrioid-type glands, sometimes with minor foci of ciliated, mucinous, or squamous metaplasia, are surrounded by endometrial stroma, which is in turn surrounded by smooth muscle (Fig. 59) (Gilks et al. 2000; Tahlan et al. 2006). These lesions, which may also be non-polypoid and located entirely within the myometrium, can be associated with underlying adenomyosis. They should not be confused with atypical polypoid adenomyoma; as mentioned above, the glands in adenomyomas are usually surrounded by endometrioid-type stroma, which is in turn surrounded by smooth muscle.
Fig. 59

Adenomyomatous polyp (adenomyoma). Endometrioid-type glands are surrounded by endometrial-type stroma, which is in turn surrounded by smooth muscle

Hyperplasia and Carcinoma Arising in an Endometrial Polyp

Occasionally, a hyperplasia or carcinoma arises within or involves an endometrial polyp (Carlson and Mutter 2008). A diagnosis of simple hyperplasia should not be made in a polyp, since a degree of glandular crowding and proliferative activity within dilated glands are features of many endometrial polyps. Complex and atypical hyperplasia is diagnosed in the same way as in non-polypoid endometrium. The hyperplasia may be confined to the polyp, but also involves the non-polypoid background endometrium in approximately 50% of cases (Kelly et al. 2007; Morsi et al. 2000). Both endometrioid and serous carcinomas (and occasionally other malignancies) may arise within or involve an endometrial polyp and sometimes be confined to this. Serous carcinomas have a particular tendency to arise in or involve endometrial polyps, as does the precursor lesion serous endometrial intraepithelial carcinoma (serous EIC) (Hui et al. 2005; Silva and Jenkins 1990). When serous EIC involves an endometrial polyp, there is partial replacement of the surface epithelium or sometimes the epithelium of the glands by cells with markedly atypical hyperchromatic nuclei, sometimes with prominent nucleoli, and prominent mitotic activity (Fig. 60). Immunohistochemical staining may assist in highlighting the serous proliferation, since the cells of serous EIC typically exhibit either diffuse, intense, nuclear immunoreactivity with p53 (Fig. 61) or complete absence of p53 expression, with a high MIB1 proliferation index; ER expression is usually decreased (see chapter “Precursor Lesions of Endometrial Carcinoma”). In contrast, the benign epithelium within the polyp exhibits a low MIB1 proliferation index and is ER positive, and only scattered nuclei are weakly positive with p53. p53 staining may reveal that the serous EIC is more extensive than is appreciated on initial morphological examination. Rarely, a carcinosarcoma arises in and is confined to an endometrial polyp. Occasional cases of metastatic carcinoma, especially breast lobular carcinoma, have been reported in endometrial polyps (Houghton et al. 2003).
Fig. 60

Endometrial polyp with serous endometrial intraepithelial carcinoma (serous EIC). Serous endometrial intraepithelial carcinoma may arise on the surface of an endometrial polyp

Fig. 61

p53 immunohistochemistry in serous endometrial intraepithelial carcinoma. Diffuse, intense, nuclear p53 immunoreactivity of serous endometrial intraepithelial carcinoma on the surface of an endometrial polyp

Atypical Polypoid Adenomyoma

Atypical polypoid adenomyoma is a biphasic polypoid lesion composed of endometrioid-type glands in a myomatous or fibromyomatous stroma (Mazur 1981; Young et al. 1986). Since the stroma may be fibromyomatous rather than overtly myomatous, some prefer the designation atypical polypoid adenomyofibroma (Longacre et al. 1996). Most patients are premenopausal or perimenopausal (average age 40 years) and present with abnormal uterine bleeding, usually in the form of menorrhagia. In some cases, the diagnosis is made during investigations for infertility. Occasional cases occur in postmenopausal women, and rare examples have been described in patients with Turner’s syndrome who have been prescribed unopposed estrogens (Clement and Young 1987). A single study has investigated molecular events in atypical polypoid adenomyoma and found MLH-1 promotor hypermethylation in some cases, a molecular alteration characteristic of some atypical hyperplasias and endometrioid adenocarcinomas (Ota et al. 2003).

Atypical polypoid adenomyoma is most commonly located in the lower uterine segment, although some cases involve the fundus, uterine body, or endocervix. In most cases, the lesion has an obvious polypoid gross appearance, in the form of either a sessile or broad-based polyp, but sometimes the polypoid nature is not grossly obvious, especially in smaller lesions.

The diagnosis may be made on endometrial biopsy, polypectomy, or at hysterectomy. Histology shows architecturally irregular endometrioid-type glands that may be widely separated and haphazardly arranged or somewhat crowded and arranged in groups, sometimes with a vaguely lobular pattern (Fig. 62). The endometrioid epithelium varies in appearance from cuboidal to low columnar to pseudostratified. The nuclei are usually round, sometimes with prominent nucleoli, and exhibit mild or, at the most, moderate cytological atypia. Occasional foci of ciliated or mucinous epithelium may be present. A characteristic histological feature that is present in most, but not all, cases is abundant squamous morule formation (Fig. 63); sometimes, the morules exhibit central necrosis. The glands are set in an abundant stroma, which varies from obviously smooth muscle in nature to fibromyomatous. Endometrial stroma is not present. The stromal cells are often arranged in short interlacing fascicles. Occasional mitotic figures may be identified within the stroma. The margin between the lesion and the underlying myometrium is usually rounded and well delineated, but occasionally there is merging with underlying adenomyosis.
Fig. 62

Atypical polypoid adenomyoma. Endometrioid-type glands are embedded within a myomatous stroma, displaying a vague lobular pattern

Fig. 63

Atypical polypoid adenomyoma. There is abundant squamous morule formation

In some cases, there is significant glandular crowding with a back-to-back architecture and stromal exclusion, such that there are foci that are virtually indistinguishable from, and which are best regarded as, grade I endometrioid adenocarcinoma. The term atypical polypoid adenomyoma of low malignant potential has been used for lesions with marked architectural complexity (Longacre et al. 1996), but this term is not recommended. Very rarely, there is underlying myometrial invasion, and/or an endometrioid adenocarcinoma is present in the surrounding endometrium.

Atypical polypoid adenomyoma is generally a benign lesion, but there is a risk of recurrence if curettage or polypectomy is undertaken. In one series, 45% of cases treated by curettage or polypectomy recurred (Longacre et al. 1996). Given this risk of recurrence and the small but definite risk of transition to endometrioid adenocarcinoma, which was estimated at 8.8% in one meta-analysis (Heatley 2006), hysterectomy is the treatment of choice if the diagnosis is made on biopsy or polypectomy. In a woman who wishes to retain her uterus and in whom a confident diagnosis of atypical polypoid adenomyoma has been made on biopsy or polypectomy, complete removal by curettage or polypectomy may be undertaken with close follow-up and imaging. Successful pregnancies have ensued in patients managed in this way. It has been suggested that recurrence is more likely in cases with marked architectural complexity.

The most important differential diagnosis of atypical polypoid adenomyomas is an endometrioid adenocarcinoma exhibiting myometrial invasion or with a prominent desmoplastic stroma, an obviously important distinction since most atypical polypoid adenomyomas exhibit a benign behavior with a potential for conservative management. Recognition of the polypoid nature of the lesion assists in establishing the diagnosis. Marked cytological atypia favors a myoinvasive adenocarcinoma since in atypical polypoid adenomyomas, there is usually no more than mild to moderate cytological atypia. The stromal component of atypical polypoid adenomyoma grows in short interlacing fascicles, in contrast to the elongated fibers of the normal myometrium. In curettings or biopsy from an atypical polypoid adenomyoma, there are usually also fragments of normal background endometrium, and with an endometrioid adenocarcinoma, it would be unusual on biopsy to obtain only myoinvasive neoplasm without free tumor fragments. Immunohistochemistry is of little value in distinguishing between atypical polypoid adenomyoma and a myoinvasive endometrioid adenocarcinoma, since the stromal component of atypical polypoid adenomyoma and myometrium infiltrated by carcinoma are both desmin and smooth muscle actin positive (Soslow et al. 1996). It has been suggested that CD10 may be of value, since this is negative in the stromal component of atypical polypoid adenomyoma, while the myoinvasive glands of endometrioid adenocarcinoma are typically surrounded by CD10-positive stromal cells (Ohishi et al. 2008). The differential diagnosis can also include a benign endometrial polyp in which there may be a minor component of smooth muscle within the stroma. So-called typical adenomyomatous polyps or adenomyomas (see section “Adenomyomatous Polyps”) also occur and are composed of benign endometrioid-type glands in a myomatous stroma (Gilks et al. 2000; Tahlan et al. 2006). Often, endometrial-type stroma surrounds the glands, and this in turn is surrounded by smooth muscle. Rarely, a carcinosarcoma enters into the differential diagnosis because of the admixture of epithelial and stromal elements. However, both the epithelial and mesenchymal components of a carcinosarcoma are obviously malignant and typically high grade.


Adenofibroma is a mixed tumor of the endometrium (and rarely also of the cervix) consisting of a benign epithelial and a benign mesenchymal component, both of which are integral components of the neoplasm. Adenofibromas most commonly occur in postmenopausal women, but the age range is wide. The most common presenting symptom is abnormal vaginal bleeding. Occasional cases have arisen in patients taking tamoxifen (Huang et al. 1996). Adenofibroma is rare and must be distinguished from the more common adenosarcoma with a subtle malignant stromal component (Zaloudek and Norris 1981) (discussed in chapter “Mesenchymal Tumors of the Uterus”). Grossly, adenofibroma occupies the uterine cavity, typically arising as a broad-based polypoid mass. The cut surface may be spongy or overtly cystic. Bland epithelium that is usually of endometrioid type but which may be mucinous, ciliated, or even squamous covers broad or fine papillary stromal fronds (Fig. 64). Cells of fibroblastic type, and more rarely endometrial stroma or smooth muscle, make up the mesenchymal element. The stroma may be cellular or fibrous, and the constituent cells are cytologically bland without nuclear pleomorphism. Mitotic figures are usually absent and should not exceed 1 per 10 high-power fields; greater mitotic activity than this warrants a diagnosis of adenosarcoma, which is more common. Periglandular cuffing by stromal cells should also result in consideration of adenosarcoma. Occasionally, an adenocarcinoma arises within an adenofibroma, but this is probably a coincidental association (Venkatraman et al. 2003).
Fig. 64

Adenofibroma of the endometrium. Adenofibroma containing benign glands with underlying bland fibrous stroma

While the adenofibroma is benign, hysterectomy is the most appropriate treatment. This is because adenosarcoma cannot be excluded in curettings or biopsy and also because recurrence of an adenofibroma treated by curettage or local excision may occur. It is for these reasons that some consider that adenofibroma does not exist, but rather is a well-differentiated form of adenosarcoma. At the heart of the controversy whether or not adenofibroma exists as an entity is the ability to distinguish it from adenosarcoma. The important features to be evaluated are the degree of mitotic activity in the stroma, the morphology of the stromal cells, and the presence of periglandular cuffing by stromal cells. Traditionally, a mitotic count greater than 1 per 10 high-power fields warrants a diagnosis of adenosarcoma, a diagnosis that should also be made if there is more than mild nuclear atypia or significant periglandular stromal cuffing. However, there can be significant morphological overlap between adenofibromatous endometrial polyps and adenosarcomas, and the presence of focal or poorly developed features can be allowed in a polypoid endometrial lesion, without the lesion behaving as an adenosarcoma (Howitt et al. 2015). Because of this considerable overlap, it is difficult to make a confident diagnosis of adenofibroma on curetted or avulsed material, and theoretically, an adenosarcoma cannot be excluded unless the whole tumor is available for examination. Thus, a hysterectomy is required to ensure that the tissue examined was not just the most benign area of an adenosarcoma. Cases have been reported in which repeated curettages have been carried out because the lesion was wrongly thought to be benign before a diagnosis of adenosarcoma has eventually been made (Clement and Scully 1990). It is also important to differentiate between adenofibroma and a benign endometrial polyp since the latter does not require further treatment. While the distinction can be difficult, adenofibroma should be considered if the lesion has a papillary surface and a stroma that is cellular. The stroma of endometrial polyps tends to be more hyaline than that of adenofibroma, although there is considerable overlap.

Effects of Intrauterine Device (IUD)

Intrauterine devices (IUDs) are in widespread use, mainly for contraceptive purposes. Older devices were usually composed entirely of plastic, while more modern devices are often composed of plastic with a copper coating. The effects of the Mirena coil, a progestin-containing IUD, are discussed in the next section (see section “Effects of Mirena Coil”).

The histological features in the endometrium in association with an IUD are largely due to local mechanical effects. The surface endometrium may take on the configuration of the IUD due to a direct pressure effect. There may be surface micropapillary formations and focal reactive changes with nuclear enlargement, mild nuclear atypia, small nucleoli, and cytoplasmic vacuolation (Fig. 65). Micropapillary fragments of epithelium exhibiting these features in an endometrial biopsy may result in diagnostic difficulty, and the clinical history of the presence of an IUD is helpful. The glandular epithelium may also exhibit epithelial cytoplasmic change, including squamous metaplasia, and there may be surface ulceration. The endometrial glands at the site of contact, and immediately surrounding this, may exhibit a different pattern of maturation from the rest of the endometrium. With long-term usage, the endometrium adjacent to the IUD may be fibrosed or simplified and composed of a monolayer. Rarely there is stromal microcalcification. Although the features may be subtle, a focal inflammatory infiltrate is commonly present consisting of neutrophils, lymphocytes, histiocytes, and plasma cells. Foreign body-type giant cells and granulomas may be a component of the inflammatory infiltrate, the severity of which may be related to the type of IUD and the duration of use. In most cases, the inflammation is superficial and largely confined to the site of contact, but in other instances it is more widespread. When confined to the IUD site, the inflammation is probably a consequence of irritation rather than infection, but when more widespread, it may be secondary to infection; in such cases, the inflammatory infiltrate, as well as being more widespread, is typically more intense. Infection appears more common with plastic devices than with copper-coated devices. Usually, a mixture of organisms is present. Long-term IUD use is associated with infection by the Gram-positive anaerobic bacterium Actinomyces (discussed earlier).
Fig. 65

Reaction to intrauterine device (IUD). Micropapillary formations are present, in keeping with a reaction to an intrauterine device

A rare but serious complication of IUD use is uterine perforation or laceration, most commonly occurring at the time of insertion. The risk of perforation is greatest in the postpartum period, when the tissues are soft and expanded. Displacement of the IUD into the pelvis with an associated inflammatory reaction may ensue secondary to perforation. Occasionally, spontaneous expulsion may also occur.

Effects of Mirena Coil

The Mirena coil is a levonorgestrel (a progestin)-releasing IUD that is in widespread use as a highly effective contraceptive. The Mirena coil is also licensed for the management of idiopathic menorrhagia and has been used to deliver progestin for endometrial protection in postmenopausal hormone replacement regimes. It has also been suggested that the Mirena coil can be used to prevent endometrial hyperplastic changes in patients taking tamoxifen or estrogen-only HRT.

The histological features of the endometrium in association with the Mirena coil include a low-power polypoid architecture (Fig. 66), probably secondary to a direct mechanical effect. Ulceration, surface micropapillary proliferations, and reactive atypia of the surface endometrium may also be present. The endometrial glands are usually small, tubular, and atrophic, but occasionally exhibit weak secretory activity. There is stromal expansion and decidualization or pseudodecidualization with infiltration by granulated lymphocytes (Fig. 67) (Phillips et al. 2003). Other histological features found in some cases include stromal myxoid or mucinous change, hemosiderin pigment, and glandular metaplastic changes. Stromal necrosis, infarction, and microcalcifications are found in a small percentage of cases. In some cases, plasma cells are present within the stroma, indicating a coexistent chronic endometritis, secondary to the presence of the IUD. Stromal hyaline nodules have also been described (Fig. 68) (Hejmadi et al. 2007). There may be associated progestational effects in the cervix, including microglandular hyperplasia and stromal decidualization.
Fig. 66

Mirena coil-associated endometrium. There is a low-power polypoid architecture

Fig. 67

Mirena coil-associated endometrium. There is stromal expansion and predecidualization. Hemosiderin pigment and inflammation are present within the stroma

Fig. 68

Mirena coil-associated endometrium. Stromal hyaline nodules may be seen in Mirena coil-associated endometrium

Radiation Effects on the Endometrium

The endometrial morphology can be altered by the effects of radiation, which may have been administered many years earlier. Radiation effect in the endometrium is characterized by surface endometrial epithelium or glands lined by cells with enlarged atypical, sometimes bizarre, hyperchromatic nuclei (Fig. 69). The nuclear chromatin may be smudged and indistinct, or there may be prominent nucleoli. Hobnail cells may be present. There is usually abundant cytoplasm that can be eosinophilic, clear, or vacuolated. The normal glandular architecture is typically maintained, or there may be glandular loss. The stroma can be fibrotic with associated vascular changes characteristic of radiation.
Fig. 69

Radiation effect on the endometrium. Glands are lined by cells with enlarged atypical nuclei

The presence of cells with enlarged atypical nuclei can raise the possibility of serous EIC. Knowledge of the history of prior radiation is obviously paramount in establishing the diagnosis; a lack of mitotic activity and a low nuclear-to-cytoplasmic ratio are other diagnostic clues. p53 immunohistochemistry may assist, since serous EIC usually exhibits diffuse, intense, nuclear immunoreactivity, while in radiation effect the nuclei exhibit weak heterogeneous staining. Prior radiation is associated with an increased risk of subsequent development of uterine malignancies. These may be of any morphological type, but carcinosarcomas are proportionally overrepresented.

Effects of Endometrial Ablation or Resection

Endometrial ablation is commonly undertaken as a nonsurgical procedure in the management of abnormal uterine bleeding, especially in premenopausal women where the suspicion of malignancy is low and preservation of fertility is not an issue. The object of endometrial ablation is destruction of the entire endometrium and superficial myometrium using thermal coagulation (electrosurgical rollerball), laser vaporization, or resection. Most patients become amenorrheic or hypomenorrheic following the ablation, but some have continuous bleeding and/or pain, and repeat endometrial biopsy or sometimes hysterectomy is undertaken either soon after the procedure or some time later.

The histological features depend on the time interval between the ablation and the subsequent biopsy or hysterectomy. In the early stages (up to 3 months), there may be complete or almost complete necrosis of the endometrium and superficial myometrium with replacement of this by a coagulum of necrotic fibrinoid material with a surrounding histiocytic and giant cell reaction (Fig. 70), the morphological features resembling that of a rheumatoid nodule or a necrotizing granulomatous process (Colgan et al. 1999; Ferryman et al. 1992). Spicules of thermally damaged tissue may be identified, and there is a variable degree of inflammation. Later, necrotic tissue is no longer present, but a giant cell and granulomatous reaction may persist, sometimes with pigment within giant cells. There is usually striking fibrosis, and there may be regeneration of the endometrium with the formation of a monolayer of simple cuboidal epithelium directly abutting the myometrium. In other cases, the endometrium is histologically relatively normal. Scarring may result in obstruction and the subsequent development of hematometra or pyometra. The histological features following endometrial resection are similar (McCulloch et al. 1995).
Fig. 70

Reaction to previous endometrial ablation. Endometrium is replaced by fibrinoid material with a surrounding histiocytic and giant cell reaction

Usually, an endometrial biopsy is taken just prior to ablation being performed. Rarely this contains an unsuspected carcinoma, and hysterectomy is performed soon after the ablation. In such instances, the endometrial carcinoma may have undergone total or extensive necrosis, secondary to the ablation procedure.

Effects of Curettage

Morphological changes may be seen in the endometrium secondary to recent curettage, especially if this has been vigorous. The changes are transient and will only be seen if repeat endometrial sampling or hysterectomy is undertaken a short time after curettage. There may be focal surface erosion with an associated mixed inflammatory infiltrate; in some cases, eosinophils are conspicuous (Miko et al. 1988). Following this, there is usually epithelial regeneration, sometimes with a micropapillary architecture, hobnail cell change, and mild reactive nuclear atypia. Usually, the changes are minor, result in no particular diagnostic difficulties, and return to normal within 10–14 days. A similar phenomenon has been described in the endocervix secondary to recent endometrial sampling and termed atypical reactive proliferation of the endocervix (Scott et al. 2006).

Asherman’s Syndrome

Asherman’s syndrome is characterized by focal or diffuse endometrial fibrosis and loss of distinction between the functionalis and basalis. The endometrium may be composed of a monolayer of epithelium with underlying fibrous tissue, and adhesions may form across the cavity. The endometrial stroma is fibrosed and may be calcified and, on rare occasions, even ossified. The endometrial glands are typically sparse and inactive and may be cystically dilated. Asherman’s syndrome can result in infertility, amenorrhea, or hypomenorrhea. Pregnancies may be complicated by premature labor, placenta previa, or placenta accreta. In some, but not all, cases, there is a known cause for the Asherman’s syndrome, which may be secondary to previous surgery, curettage or ablation, infection, miscarriage, or retained products of conception. Hysteroscopic lysis of adhesions can result in successful pregnancies.

Postoperative Spindle Cell Nodule

Postoperative spindle cell nodule represents an exaggerated reparative response at the site of previous surgery or biopsy and, in the female genital tract, is most common in the vagina. It has rarely been described in the endometrium (Clement 1988). Postoperative spindle cell nodule occurs within weeks or months of the inciting procedure and consists of a cellular proliferation of spindle-shaped cells with admixed vascular channels and inflammatory cells. There may be abundant mitotic activity, raising the possibility of a sarcoma, usually a leiomyosarcoma. However, the high mitotic activity contrasts with the bland nuclear features and lack of cytological atypia. The history of a recent procedure is obviously paramount in establishing the diagnosis. The constituent cells may be positive with smooth muscle actin, desmin, and, on rare occasions, cytokeratins.

Psammoma Bodies in the Endometrium

Calcified psammoma bodies may be seen in the endometrium in association with benign and malignant lesions and rarely in normal endometrium. They are present in up to one-third of uterine serous carcinomas, a lower frequency than in ovarian serous carcinomas (Hendrickson et al. 1982). More uncommonly, they are seen in other morphological types of endometrial malignancy, such as endometrioid carcinoma. Psammoma bodies are occasionally seen in normal endometria (usually atrophic or proliferative in type), sometimes in association with hormonal preparations, and in benign lesions, most commonly endometrial polyps (Herbold and Magrane 1986; Truskinovsky et al. 2008). They are often located within the glandular lumina, in which case it is likely that they represent calcification of inspissated secretions. In other cases, they are situated within the stroma where they may be secondary to prior inflammation or the presence of an IUD. In the absence of a malignant lesion, the occurrence of psammoma bodies within the glandular lumina or endometrial stroma is not an indication for the evaluation of the upper female genital tract to exclude malignancy. However, free-floating psammoma bodies without attachment to tissue may be an indication of an extrauterine serous carcinoma.

Emphysematous Endometritis

There have been occasional reports of emphysematous (pneumopolycystic) endometritis characterized by the presence of gas-filled cysts in the endometrial stroma (Perkins 1960; Val-Bernal et al. 2006). There may be simultaneous involvement of the cervix, or the condition may be confined to the endometrium. Histology shows empty cystic spaces of variable sizes and contour within the endometrial stroma lined by flattened stromal cells with occasional histiocytes and/or giant cells. Spontaneous resolution usually occurs. Emphysematous endometritis should be distinguished from sectioning artifact, dilated vascular spaces, and gas gangrene of the uterus, which is life-threatening and associated with tissue necrosis.

Benign Endometrial Stromal Proliferations

Endometrial stromal neoplasms are discussed in chapter “Mesenchymal Tumors of the Uterus”, as is the differential diagnosis of fragments of tissue composed entirely of endometrial stroma in an endometrial biopsy. Occasional cases of multifocal microscopic benign endometrial stromal proliferations confined to the endometrium without invasive growth have been reported (Stewart et al. 1998). These have been termed focal endometrial stromal hyperplasia and may mimic an endometrial stromal nodule or endometrial stromal sarcoma in biopsy samples. Rarely, markedly atypical stromal cells with a symplastic appearance are present within an otherwise normal endometrium (Fig. 71) (Usubutun et al. 2005).
Fig. 71

Atypical endometrial stromal cells. Rarely endometrial stromal cells with atypical symplastic-like nuclei are present in the normal endometrium. (Courtesy of Dr. Brigitte Ronnett)

Benign Trophoblastic Lesions

Benign lesions of intermediate trophoblast, namely, placental site nodule or plaque (PSNP) and exaggerated placental site, are discussed in chapter “Gestational Trophoblastic Tumors and Related Tumor-Like Lesions”. An endometrial PSNP may be identified in a biopsy, resection, or hysterectomy specimen many years following a pregnancy or abortion and rarely even in a postmenopausal woman. PSNP is characterized histologically by a well-circumscribed lesion composed of cells with degenerating large, often atypical, nuclei with an abundant eosinophilic cytoplasm (Fig. 72). The immunophenotype is discussed in detail later in chapter “Gestational Trophoblastic Tumors and Related Tumor-Like Lesions”.
Fig. 72

Placental site nodule or plaque. A well-circumscribed lesion is composed of epithelioid cells with abundant cytoplasm

Intravascular Endometrium

Occasionally in a hysterectomy specimen, menstrual endometrium is present within myometrial vascular channels (Fig. 73) (Banks et al. 1991). Rarely there is extensive vascular involvement and even invasion of parametrial vessels. Intravascular menstrual endometrium is of no significance other than it may be mistaken for neoplastic involvement of vascular channels. It should also be distinguished from intravascular foci of adenomyosis (Sahin et al. 1989).
Fig. 73

Intravascular menstrual endometrium. Menstrual endometrium within myometrial blood vessels is occasionally seen and is of no significance

Endometrial Autolysis

Hysterectomies are performed for a wide variety of benign and malignant conditions. Especially with large uteri, the formalin does not penetrate into the endometrial cavity, and, as a consequence, fixation is often not adequate and the endometrium may be markedly autolyzed. This not uncommonly results in problems in morphological assessment of the endometrium, which in some cases is so autolyzed that interpretation is impossible. This is important in all uteri but particularly so with endometrial neoplasms, where autolysis may result in problems in assessing tumor type and grade. Bisection of the uterus soon after surgery may improve fixation, but this often results in distortion of the specimen, making assessment of parameters such as the depth of myometrial invasion by tumor problematic. Packing the cavity with absorbable tissue paper following bisection and ensuring that the two halves of the uterus remain apposed minimizes the distortion. Uteri may also be injected with formalin using a needle and syringe directed alongside a probe, which is inserted through the external cervical os into the endometrial cavity (Houghton et al. 2004). This initiation results in significantly less endometrial autolysis.


  1. Anderson GL, Judd HL, Kaunitz AM et al (2003) Effects of estrogen plus progestin on gynecologic cancers and associated diagnostic procedures: the Women’s Health Initiative randomized trial. JAMA 290:1739–1748. Scholar
  2. Antoni J, Folch E, Costa J et al (1997) Comparison of cytospat and pipelle endometrial biopsy instruments. Eur J Obstet Gynecol Reprod Biol 72:57–61PubMedCrossRefPubMedCentralGoogle Scholar
  3. Archer DF, McIntyre-Seltman K, Wilborn WW Jr et al (1991) Endometrial morphology in asymptomatic postmenopausal women. Am J Obstet Gynecol 165:317–320; discussion 320–312PubMedCrossRefPubMedCentralGoogle Scholar
  4. Arias-Stella J (2002) The Arias-Stella reaction: facts and fancies four decades after. Adv Anat Pathol 9:12–23PubMedCrossRefPubMedCentralGoogle Scholar
  5. Arias-Stella J Jr, Arias-Velasquez A, Arias-Stella J (1994) Normal and abnormal mitoses in the atypical endometrial change associated with chorionic tissue effect [corrected]. Am J Surg Pathol 18:694–701PubMedCrossRefPubMedCentralGoogle Scholar
  6. Bahceci M, Demirel LC (1996) Osseous metaplasia of the endometrium: a rare cause of infertility and its hysteroscopic management. Hum Reprod 11:2537–2539PubMedCrossRefPubMedCentralGoogle Scholar
  7. Baird DB, Reddick RL (1991) Extraovarian mucinous metaplasia in a patient with bilateral mucinous borderline ovarian tumors: a case report. Int J Gynecol Pathol 10:96–103PubMedCrossRefPubMedCentralGoogle Scholar
  8. Bakour SH, Khan KS, Gupta JK (2000) Controlled analysis of factors associated with insufficient sample on outpatient endometrial biopsy. BJOG 107:1312–1314PubMedCrossRefPubMedCentralGoogle Scholar
  9. Banks ER, Mills SE, Frierson HF Jr (1991) Uterine intravascular menstrual endometrium simulating malignancy. Am J Surg Pathol 15:407–412PubMedCrossRefPubMedCentralGoogle Scholar
  10. Barroeta JE, Pasha TL, Acs G et al (2007) Immunoprofile of endocervical and endometrial stromal cells and its potential application in localization of tumor involvement. Int J Gynecol Pathol 26:76–82. Scholar
  11. Bayer-Garner IB, Korourian S (2001) Plasma cells in chronic endometritis are easily identified when stained with syndecan-1. Mod Pathol 14:877–879. Scholar
  12. Bayer-Garner IB, Nickell JA, Korourian S (2004) Routine syndecan-1 immunohistochemistry aids in the diagnosis of chronic endometritis. Arch Pathol Lab Med 128:1000–1003.<1000:rsiait>;2CrossRefPubMedPubMedCentralGoogle Scholar
  13. Bennett AE, Rathore S, Rhatigan RM (1999) Focal necrotizing endometritis: a clinicopathologic study of 15 cases. Int J Gynecol Pathol 18:220–225PubMedCrossRefPubMedCentralGoogle Scholar
  14. Bozdogan O, Atasoy P, Erekul S et al (2002) Apoptosis-related proteins and steroid hormone receptors in normal, hyperplastic, and neoplastic endometrium. Int J Gynecol Pathol 21:375–382PubMedCrossRefPubMedCentralGoogle Scholar
  15. Brachtel EF, Sanchez-Estevez C, Moreno-Bueno G et al (2005) Distinct molecular alterations in complex endometrial hyperplasia (CEH) with and without immature squamous metaplasia (squamous morules). Am J Surg Pathol 29:1322–1329PubMedCrossRefPubMedCentralGoogle Scholar
  16. Brinton LA, Hoover RN (1993) Estrogen replacement therapy and endometrial cancer risk: a meta-analysis. Obstet Gynecol 81:265–271PubMedPubMedCentralGoogle Scholar
  17. Bulmer JN, Hollings D, Ritson A (1987) Immunocytochemical evidence that endometrial stromal granulocytes are granulated lymphocytes. J Pathol 153:281–288. Scholar
  18. Bulmer JN, Lunny DP, Hagin SV (1988) Immunohistochemical characterization of stromal leucocytes in nonpregnant human endometrium. Am J Reprod Immunol Microbiol 17:83–90PubMedCrossRefPubMedCentralGoogle Scholar
  19. Busca A, Parra-Herran C (2016) The role of pathologic evaluation of endometrial ablation resections in predicting ablation failure and adenomyosis in hysterectomy. Pathol Res Pract 212:778–782. Scholar
  20. Carlson JW, Mutter GL (2008) Endometrial intraepithelial neoplasia is associated with polyps and frequently has metaplastic change. Histopathology 53:325–332. Scholar
  21. Cheng WF, Lin HH, Torng PL et al (1997) Comparison of endometrial changes among symptomatic tamoxifen-treated and nontreated premenopausal and postmenopausal breast cancer patients. Gynecol Oncol 66:233–237. Scholar
  22. Chiarelli S, Buritica C, Litta P et al (2006) An immunohistochemical study of morules in endometrioid lesions of the female genital tract: CD10 is a characteristic marker of morular metaplasia. Clin Cancer Res 12:4251–4256. Scholar
  23. Chinen K, Kamiyama K, Kinjo T et al (2004) Morules in endometrial carcinoma and benign endometrial lesions differ from squamous differentiation tissue and are not infected with human papillomavirus. J Clin Pathol 57:918–926. Scholar
  24. Clark TJ, Mann CH, Shah N et al (2001) Accuracy of outpatient endometrial biopsy in the diagnosis of endometrial hyperplasia. Acta Obstet Gynecol Scand 80:784–793PubMedCrossRefPubMedCentralGoogle Scholar
  25. Clement PB (1988) Postoperative spindle-cell nodule of the endometrium. Arch Pathol Lab Med 112:566–568PubMedPubMedCentralGoogle Scholar
  26. Clement PB, Scully RE (1988) Idiopathic postmenopausal decidual reaction of the endometrium. A clinicopathologic analysis of four cases. Int J Gynecol Pathol 7:152–161PubMedCrossRefPubMedCentralGoogle Scholar
  27. Clement PB, Scully RE (1990) Mullerian adenosarcoma of the uterus: a clinicopathologic analysis of 100 cases with a review of the literature. Hum Pathol 21:363–381PubMedCrossRefPubMedCentralGoogle Scholar
  28. Clement PB, Young RH (1987) Atypical polypoid adenomyoma of the uterus associated with Turner’s syndrome. A report of three cases, including a review of “estrogen-associated” endometrial neoplasms and neoplasms associated with Turner’s syndrome. Int J Gynecol Pathol 6:104–113PubMedCrossRefPubMedCentralGoogle Scholar
  29. Clement PB, Oliva E, Young RH (1996) Mullerian adenosarcoma of the uterine corpus associated with tamoxifen therapy: a report of six cases and a review of tamoxifen-associated endometrial lesions. Int J Gynecol Pathol 15:222–229PubMedCrossRefPubMedCentralGoogle Scholar
  30. Cohen I, Figer A, Altaras MM et al (1996) Common endometrial decidual reaction in postmenopausal breast cancer patients treated with tamoxifen and progestogens. Int J Gynecol Pathol 15:17–22PubMedCrossRefPubMedCentralGoogle Scholar
  31. Cohen I, Beyth Y, Shapira J et al (1997) High frequency of adenomyosis in postmenopausal breast cancer patients treated with tamoxifen. Gynecol Obstet Investig 44:200–205CrossRefGoogle Scholar
  32. Colgan TJ, Shah R, Leyland N (1999) Post-hysteroscopic ablation reaction: a histopathologic study of the effects of electrosurgical ablation. Int J Gynecol Pathol 18:325–331PubMedCrossRefPubMedCentralGoogle Scholar
  33. Creagh TM, Bain BJ, Evans DJ et al (1995) Endometrial extramedullary haemopoiesis. J Pathol 176:99–104. Scholar
  34. Critchley HO, Tong S, Cameron ST et al (1999) Regulation of bcl-2 gene family members in human endometrium by antiprogestin administration in vivo. J Reprod Fertil 115:389–395PubMedCrossRefPubMedCentralGoogle Scholar
  35. Crum CP, Hornstein MD, Nucci MR et al (2003) Hertig and beyond: a systematic and practical approach to the endometrial biopsy. Adv Anat Pathol 10:301–318PubMedCrossRefPubMedCentralGoogle Scholar
  36. Dal Cin P, De Wolf F, Klerckx P et al (1992) The 6p21 chromosome region is nonrandomly involved in endometrial polyps. Gynecol Oncol 46:393–396PubMedCrossRefPubMedCentralGoogle Scholar
  37. De Quintal MM, De Angelo Andrade LA (2006) Endometrial polyp with sex cord-like pattern. Int J Gynecol Pathol 25:170–172. Scholar
  38. Delmas PD, Bjarnason NH, Mitlak BH et al (1997) Effects of raloxifene on bone mineral density, serum cholesterol concentrations, and uterine endometrium in postmenopausal women. N Engl J Med 337:1641–1647. Scholar
  39. Deshmukh-Rane SA, Wu ML (2009) Pseudolipomatosis affects specimens from endometrial biopsies. Am J Clin Pathol 132:374–377. Scholar
  40. Dhingra N, Punia RS, Radotra A et al (2007) Arias-Stella reaction in upper genital tract in pregnant and non-pregnant women: a study of 120 randomly selected cases. Arch Gynecol Obstet 276:47–52. Scholar
  41. Disep B, Innes BA, Cochrane HR et al (2004) Immunohistochemical characterization of endometrial leucocytes in endometritis. Histopathology 45:625–632. Scholar
  42. Duncan DA, Varner RE, Mazur MT (1989) Uterine herpes virus infection with multifocal necrotizing endometritis. Hum Pathol 20:1021–1024PubMedCrossRefPubMedCentralGoogle Scholar
  43. Dutra FR (1959) Intraglandular morules of the endometrium. Am J Clin Pathol 31:60–65PubMedCrossRefPubMedCentralGoogle Scholar
  44. Euscher E, Nuovo GJ (2002) Detection of kappa- and lambda-expressing cells in the endometrium by in situ hybridization. Int J Gynecol Pathol 21:383–390PubMedCrossRefPubMedCentralGoogle Scholar
  45. Fechner RE, Bossart MI, Spjut HJ (1979) Ultrastructure of endometrial stromal foam cells. Am J Clin Pathol 72:628–633PubMedCrossRefPubMedCentralGoogle Scholar
  46. Feeley KM, Wells M (2001) Hormone replacement therapy and the endometrium. J Clin Pathol 54:435–440PubMedPubMedCentralCrossRefGoogle Scholar
  47. Ferryman SR, Stephens M, Gough D (1992) Necrotising granulomatous endometritis following endometrial ablation therapy. Br J Obstet Gynaecol 99:928–930PubMedCrossRefPubMedCentralGoogle Scholar
  48. Fisher B, Costantino JP, Wickerham DL et al (1998) Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst 90:1371–1388PubMedCrossRefPubMedCentralGoogle Scholar
  49. Fletcher JA, Pinkus JL, Lage JM et al (1992) Clonal 6p21 rearrangement is restricted to the mesenchymal component of an endometrial polyp. Genes Chromosomes Cancer 5:260–263PubMedCrossRefPubMedCentralGoogle Scholar
  50. Fluhmann CF (1960) The developmental anatomy of the cervix uteri. Obstet Gynecol 15:62–69PubMedPubMedCentralGoogle Scholar
  51. Frank TS, Himebaugh KS, Wilson MD (1992) Granulomatous endometritis associated with histologically occult cytomegalovirus in a healthy patient. Am J Surg Pathol 16:716–720PubMedCrossRefPubMedCentralGoogle Scholar
  52. Fritsch H, Hoermann R, Bitsche M et al (2013) Development of epithelial and mesenchymal regionalization of the human fetal utero-vaginal anlagen. J Anat 222:462–472. Scholar
  53. Fukunaga M, Iwaki S (2004) Nodular histiocytic hyperplasia of the endometrium. Arch Pathol Lab Med 128:1032–1034.<1032:nhhote>;2CrossRefPubMedPubMedCentralGoogle Scholar
  54. Genest DR, Brodsky G, Lage JA (1995) Localized endometrial proliferations associated with pregnancy: clinical and histopathologic features of 11 cases. Hum Pathol 26:1233–1240PubMedCrossRefPubMedCentralGoogle Scholar
  55. Gilks CB, Clement PB, Hart WR et al (2000) Uterine adenomyomas excluding atypical polypoid adenomyomas and adenomyomas of endocervical type: a clinicopathologic study of 30 cases of an underemphasized lesion that may cause diagnostic problems with brief consideration of adenomyomas of other female genital tract sites. Int J Gynecol Pathol 19:195–205PubMedCrossRefPubMedCentralGoogle Scholar
  56. Gompel A, Sabourin JC, Martin A et al (1994) Bcl-2 expression in normal endometrium during the menstrual cycle. Am J Pathol 144:1195–1202PubMedPubMedCentralGoogle Scholar
  57. Gordon SJ, Westgate J (1999) The incidence and management of failed Pipelle sampling in a general outpatient clinic. Aust N Z J Obstet Gynaecol 39:115–118PubMedCrossRefPubMedCentralGoogle Scholar
  58. Grady D, Gebretsadik T, Kerlikowske K et al (1995) Hormone replacement therapy and endometrial cancer risk: a meta-analysis. Obstet Gynecol 85:304–313. Scholar
  59. Gupta PK, Hollander DH, Frost JK (1976) Actinomycetes in cervico-vaginal smears: an association with IUD usage. Acta Cytol 20:295–297PubMedPubMedCentralGoogle Scholar
  60. Hachisuga T, Hideshima T, Kawarabayashi T et al (1999) Expression of steroid receptors, Ki-67, and epidermal growth factor receptor in tamoxifen-treated endometrium. Int J Gynecol Pathol 18:297–303PubMedCrossRefPubMedCentralGoogle Scholar
  61. Heatley MK (2006) Atypical polypoid adenomyoma: a systematic review of the English literature. Histopathology 48:609–610. Scholar
  62. Hejmadi RK, Chaudhri S, Ganesan R et al (2007) Morphologic changes in the endometrium associated with the use of the mirena coil: a retrospective study of 106 cases. Int J Surg Pathol 15:148–154. Scholar
  63. Hendrickson MR, Kempson RL (1980) Endometrial epithelial metaplasias: proliferations frequently misdiagnosed as adenocarcinoma. Report of 89 cases and proposed classification. Am J Surg Pathol 4:525–542PubMedCrossRefPubMedCentralGoogle Scholar
  64. Hendrickson M, Ross J, Eifel P et al (1982) Uterine papillary serous carcinoma: a highly malignant form of endometrial adenocarcinoma. Am J Surg Pathol 6:93–108PubMedCrossRefPubMedCentralGoogle Scholar
  65. Herbold DR, Magrane DM (1986) Calcifications of the benign endometrium. Arch Pathol Lab Med 110:666–669PubMedPubMedCentralGoogle Scholar
  66. Houghton JP, Ioffe OB, Silverberg SG et al (2003) Metastatic breast lobular carcinoma involving tamoxifen-associated endometrial polyps: report of two cases and review of tamoxifen-associated polypoid uterine lesions. Mod Pathol 16:395–398. Scholar
  67. Houghton JP, Roddy S, Carroll S et al (2004) A simple method for the prevention of endometrial autolysis in hysterectomy specimens. J Clin Pathol 57:332–333PubMedPubMedCentralCrossRefGoogle Scholar
  68. Houghton O, Connolly LE, McCluggage WG (2008) Morules in endometrioid proliferations of the uterus and ovary consistently express the intestinal transcription factor CDX2. Histopathology 53:156–165. Scholar
  69. Howitt BE, Quade BJ, Nucci MR (2015) Uterine polyps with features overlapping with those of Mullerian adenosarcoma: a clinicopathologic analysis of 29 cases emphasizing their likely benign nature. Am J Surg Pathol 39:116–126. Scholar
  70. Huang KT, Chen CA, Cheng WF et al (1996) Sonographic characteristics of adenofibroma of the endometrium following tamoxifen therapy for breast cancer: two case reports. Ultrasound Obstet Gynecol 7:363–366. Scholar
  71. Hui P, Kelly M, O’Malley DM et al (2005) Minimal uterine serous carcinoma: a clinicopathological study of 40 cases. Mod Pathol 18:75–82. Scholar
  72. Iezzoni JC, Mills SE (2001) Nonneoplastic endometrial signet-ring cells. Vacuolated decidual cells and stromal histiocytes mimicking adenocarcinoma. Am J Clin Pathol 115:249–255. Scholar
  73. Ip PP, Irving JA, McCluggage WG et al (2013) Papillary proliferation of the endometrium: a clinicopathologic study of 59 cases of simple and complex papillae without cytologic atypia. Am J Surg Pathol 37:167–177. Scholar
  74. Irving JA, McFarland DF, Stuart DS et al (2000) Mitotic arrest of endometrial epithelium after paclitaxel therapy for breast cancer. Int J Gynecol Pathol 19:395–397PubMedCrossRefPubMedCentralGoogle Scholar
  75. Ismail SM (1994) Pathology of endometrium treated with tamoxifen. J Clin Pathol 47:827–833PubMedPubMedCentralCrossRefGoogle Scholar
  76. Ismail SM (1999) Gynaecological effects of tamoxifen. J Clin Pathol 52:83–88PubMedPubMedCentralCrossRefGoogle Scholar
  77. Karaer A, Mert I, Akinsu F et al (2007) Ligneous inflammation involving the female genital tract. J Obstet Gynaecol Res 33:581–584. Scholar
  78. Kaufman RH, Binder GL, Gray PM Jr et al (1977) Upper genital tract changes associated with exposure in utero to diethylstilbestrol. Am J Obstet Gynecol 128:51–59PubMedCrossRefPubMedCentralGoogle Scholar
  79. Kelly P, Dobbs SP, McCluggage WG (2007) Endometrial hyperplasia involving endometrial polyps: report of a series and discussion of the significance in an endometrial biopsy specimen. BJOG 114:944–950. Scholar
  80. Kennedy MM, Baigrie CF, Manek S (1999) Tamoxifen and the endometrium: review of 102 cases and comparison with HRT-related and non-HRT-related endometrial pathology. Int J Gynecol Pathol 18:130–137PubMedCrossRefPubMedCentralGoogle Scholar
  81. Khatamee MA, Sommers SC (1989) Clinicopathologic diagnosis of mycoplasma endometritis. Int J Fertil 34:52–55PubMedPubMedCentralGoogle Scholar
  82. Kim KR, Lee YH, Ro JY (2002) Nodular histiocytic hyperplasia of the endometrium. Int J Gynecol Pathol 21:141–146PubMedCrossRefPubMedCentralGoogle Scholar
  83. Kurman RJ, Young RH, Norris HJ et al (1984) Immunocytochemical localization of placental lactogen and chorionic gonadotropin in the normal placenta and trophoblastic tumors, with emphasis on intermediate trophoblast and the placental site trophoblastic tumor. Int J Gynecol Pathol 3:101–121PubMedCrossRefPubMedCentralGoogle Scholar
  84. Lehman MB, Hart WR (2001) Simple and complex hyperplastic papillary proliferations of the endometrium: a clinicopathologic study of nine cases of apparently localized papillary lesions with fibrovascular stromal cores and epithelial metaplasia. Am J Surg Pathol 25:1347–1354PubMedCrossRefPubMedCentralGoogle Scholar
  85. Leong ASY, Vinyuvat S, Leong FJWM et al (1997) Anti-CD38 and VS38 antibodies for the detection of plasma cells in the diagnosis of chronic endometritis. Appl Immunohistochem 5:189–193CrossRefGoogle Scholar
  86. Lindenman E, Shepard MK, Pescovitz OH (1997) Mullerian agenesis: an update. Obstet Gynecol 90:307–312. Scholar
  87. Longacre TA, Chung MH, Rouse RV et al (1996) Atypical polypoid adenomyofibromas (atypical polypoid adenomyomas) of the uterus. A clinicopathologic study of 55 cases. Am J Surg Pathol 20:1–20PubMedCrossRefPubMedCentralGoogle Scholar
  88. Machado F, Moreno J, Carazo M et al (2003) Accuracy of endometrial biopsy with the Cornier pipelle for diagnosis of endometrial cancer and atypical hyperplasia. Eur J Gynaecol Oncol 24:279–281PubMedPubMedCentralGoogle Scholar
  89. Marchini M, Fedele L, Bianchi S et al (1992) Endometrial patterns during therapy with danazol or gestrinone for endometriosis: structural and ultrastructural study. Hum Pathol 23:51–56PubMedCrossRefPubMedCentralGoogle Scholar
  90. Marshall RJ, Jones DB (1988) An immunohistochemical study of lymphoid tissue in human endometrium. Int J Gynecol Pathol 7:225–235PubMedCrossRefPubMedCentralGoogle Scholar
  91. Matias-Guiu X, Pons C, Munoz J et al (1994) Biotin-containing intranuclear inclusions in tumor cells: possible cause for misinterpreting nuclear antigen immunostaining. Am J Clin Pathol 102:706–707PubMedCrossRefPubMedCentralGoogle Scholar
  92. Mazur MT (1981) Atypical polypoid adenomyomas of the endometrium. Am J Surg Pathol 5:473–482PubMedCrossRefPubMedCentralGoogle Scholar
  93. Mazur MT, Hendrickson MR, Kempson RL (1983) Optically clear nuclei. An alteration of endometrial epithelium in the presence of trophoblast. Am J Surg Pathol 7:415–423PubMedCrossRefPubMedCentralGoogle Scholar
  94. McCluggage WG (2002) Malignant biphasic uterine tumours: carcinosarcomas or metaplastic carcinomas? J Clin Pathol 55:321–325PubMedPubMedCentralCrossRefGoogle Scholar
  95. McCluggage WG (2003) Metaplasias in the female genital tract. In: Lowe D, Underwood J (eds) Recent advances in histopathology, vol 20. Royal Society of Medicine Press Ltd, LondonGoogle Scholar
  96. McCluggage WG (2006) My approach to the interpretation of endometrial biopsies and curettings. J Clin Pathol 59:801–812. Scholar
  97. McCluggage WG, Roberts N, Bharucha H (1995) Enteric differentiation in endometrial adenocarcinomas: a mucin histochemical study. Int J Gynecol Pathol 14:250–254PubMedCrossRefPubMedCentralGoogle Scholar
  98. McCluggage WG, Varma M, Weir P et al (1996) Uterine leiomyosarcoma in patient receiving tamoxifen therapy. Acta Obstet Gynecol Scand 75:593–595PubMedCrossRefPubMedCentralGoogle Scholar
  99. McCluggage WG, Ashe P, McBride H et al (1998) Localization of the cellular expression of inhibin in trophoblastic tissue. Histopathology 32:252–256PubMedCrossRefPubMedCentralGoogle Scholar
  100. McCluggage WG, Desai V, Manek S (2000a) Tamoxifen-associated postmenopausal adenomyosis exhibits stromal fibrosis, glandular dilatation and epithelial metaplasias. Histopathology 37:340–346PubMedCrossRefPubMedCentralGoogle Scholar
  101. McCluggage WG, Hamal P, Traub AI et al (2000b) Uterine adenolipoleiomyoma: a rare hamartomatous lesion. Int J Gynecol Pathol 19:183–185PubMedCrossRefPubMedCentralGoogle Scholar
  102. McCluggage WG, Abdulkader M, Price JH et al (2000c) Uterine carcinosarcomas in patients receiving tamoxifen. A report of 19 cases. Int J Gynecol Cancer 10:280–284PubMedCrossRefPubMedCentralGoogle Scholar
  103. McCluggage WG, Oliva E, Herrington CS et al (2003a) CD10 and calretinin staining of endocervical glandular lesions, endocervical stroma and endometrioid adenocarcinomas of the uterine corpus: CD10 positivity is characteristic of, but not specific for, mesonephric lesions and is not specific for endometrial stroma. Histopathology 43:144–150PubMedCrossRefPubMedCentralGoogle Scholar
  104. McCluggage WG, Sumathi VP, McManus DT (2003b) Uterine serous carcinoma and endometrial intraepithelial carcinoma arising in endometrial polyps: report of 5 cases, including 2 associated with tamoxifen therapy. Hum Pathol 34:939–943PubMedCrossRefPubMedCentralGoogle Scholar
  105. McCulloch TA, Wagner B, Duffy S et al (1995) The pathology of hysterectomy specimens following trans-cervical resection of the endometrium. Histopathology 27:541–547PubMedCrossRefPubMedCentralGoogle Scholar
  106. McKenna M, McCluggage WG (2008) Signet ring cells of stromal derivation in the uterine cervix secondary to cauterisation: report of a previously undescribed phenomenon. J Clin Pathol 61:648–651. Scholar
  107. Miko TL, Lampe LG, Thomazy VA et al (1988) Eosinophilic endomyometritis associated with diagnostic curettage. Int J Gynecol Pathol 7:162–172PubMedCrossRefPubMedCentralGoogle Scholar
  108. Morsi HM, Leers MP, Jager W et al (2000) 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 Pathol 19:118–126PubMedCrossRefPubMedCentralGoogle Scholar
  109. Murray MJ, Meyer WR, Zaino RJ et al (2004) A critical analysis of the accuracy, reproducibility, and clinical utility of histologic endometrial dating in fertile women. Fertil Steril 81:1333–1343. Scholar
  110. Mutter GL, Zaino RJ, Baak JP et al (2007) Benign endometrial hyperplasia sequence and endometrial intraepithelial neoplasia. Int J Gynecol Pathol 26:103–114. Scholar
  111. Mutter GL, Bergeron C, Deligdisch L et al (2008) The spectrum of endometrial pathology induced by progesterone receptor modulators. Mod Pathol 21:591–598. Scholar
  112. Nakamura Y, Moritsuka Y, Ohta Y et al (1989) S-100 protein in glands within decidua and cervical glands during early pregnancy. Hum Pathol 20:1204–1209PubMedCrossRefPubMedCentralGoogle Scholar
  113. Nand SL, Webster MA, Baber R et al (1998) Bleeding pattern and endometrial changes during continuous combined hormone replacement therapy. The Ogen/Provera Study Group. Obstet Gynecol 91:678–684PubMedPubMedCentralGoogle Scholar
  114. Noyes RW, Hertig AT, Rock J (1950) Dating the endometrial biopsy. Fertil Steril 1:3–25CrossRefGoogle Scholar
  115. Nucci MR, Young RH (2004) Arias-Stella reaction of the endocervix: a report of 18 cases with emphasis on its varied histology and differential diagnosis. Am J Surg Pathol 28:608–612PubMedCrossRefPubMedCentralGoogle Scholar
  116. Nucci MR, Prasad CJ, Crum CP et al (1999) Mucinous endometrial epithelial proliferations: a morphologic spectrum of changes with diverse clinical significance. Mod Pathol 12:1137–1142PubMedPubMedCentralGoogle Scholar
  117. O’Brien PK, Roth-Moyo LA, Davis BA (1981) Pseudo-sulfur granules associated with intrauterine contraceptive devices. Am J Clin Pathol 75:822–825PubMedCrossRefPubMedCentralGoogle Scholar
  118. O’Connell JT, Mutter GL, Cviko A et al (2001) Identification of a basal/reserve cell immunophenotype in benign and neoplastic endometrium: a study with the p53 homologue p63. Gynecol Oncol 80:30–36. Scholar
  119. O’Connor DM, Kurman RJ (1988) Intermediate trophoblast in uterine curettings in the diagnosis of ectopic pregnancy. Obstet Gynecol 72:665–670PubMedPubMedCentralGoogle Scholar
  120. Ohishi Y, Kaku T, Kobayashi H et al (2008) CD10 immunostaining distinguishes atypical polypoid adenomyofibroma (atypical polypoid adenomyoma) from endometrial carcinoma invading the myometrium. Hum Pathol 39:1446–1453. Scholar
  121. Ota S, Catasus L, Matias-Guiu X et al (2003) Molecular pathology of atypical polypoid adenomyoma of the uterus. Hum Pathol 34:784–788PubMedCrossRefPubMedCentralGoogle Scholar
  122. Paavonen J, Aine R, Teisala K et al (1985) Chlamydial endometritis. J Clin Pathol 38:726–732PubMedPubMedCentralCrossRefGoogle Scholar
  123. Paganini-Hill A, Ross RK, Henderson BE (1989) Endometrial cancer and patterns of use of oestrogen replacement therapy: a cohort study. Br J Cancer 59:445–447PubMedPubMedCentralCrossRefGoogle Scholar
  124. Pearce KF, Nolan TE (1996) Endometrial sarcoidosis as a cause of postmenopausal bleeding. A case report. J Reprod Med 41:878–880PubMedPubMedCentralGoogle Scholar
  125. Perkins MB (1960) Pneumopolycystic endometritis. Am J Obstet Gynecol 80:332–336PubMedCrossRefPubMedCentralGoogle Scholar
  126. Phillips V, McCluggage WG (2005) Results of a questionnaire regarding criteria for adequacy of endometrial biopsies. J Clin Pathol 58:417–419. Scholar
  127. Phillips V, Graham CT, Manek S et al (2003) The effects of the levonorgestrel intrauterine system (Mirena coil) on endometrial morphology. J Clin Pathol 56:305–307PubMedPubMedCentralCrossRefGoogle Scholar
  128. Pritt B, Mount SL, Cooper K et al (2006) Pseudoactinomycotic radiate granules of the gynaecological tract: review of a diagnostic pitfall. J Clin Pathol 59:17–20. Scholar
  129. Quddus MR, Sung CJ, Zheng W et al (1999) p53 immunoreactivity in endometrial metaplasia with dysfunctional uterine bleeding. Histopathology 35:44–49PubMedCrossRefPubMedCentralGoogle Scholar
  130. Rutanen EM (1998) Insulin-like growth factors in endometrial function. Gynecol Endocrinol 12:399–406PubMedCrossRefPubMedCentralGoogle Scholar
  131. Saegusa M, Okayasu I (2001) Frequent nuclear beta-catenin accumulation and associated mutations in endometrioid-type endometrial and ovarian carcinomas with squamous differentiation. J Pathol 194:59–67. Scholar
  132. Sahin AA, Silva EG, Landon G et al (1989) Endometrial tissue in myometrial vessels not associated with menstruation. Int J Gynecol Pathol 8:139–146PubMedCrossRefPubMedCentralGoogle Scholar
  133. Sakhdari A, Moghaddam PA, Liu Y (2016) Endometrial samples from postmenopausal women: a proposal for adequacy criteria. Int J Gynecol Pathol 35:525–530. Scholar
  134. Schindler AE, Schmidt G (1980) Post-menopausal bleeding: a study of more than 1000 cases. Maturitas 2:269–274PubMedCrossRefPubMedCentralGoogle Scholar
  135. Schlesinger C, Kamoi S, Ascher SM et al (1998) Endometrial polyps: a comparison study of patients receiving tamoxifen with two control groups. Int J Gynecol Pathol 17:302–311PubMedCrossRefPubMedCentralGoogle Scholar
  136. Scott M, Lyness RW, McCluggage WG (2006) Atypical reactive proliferation of endocervix: a common lesion associated with endometrial carcinoma and likely related to prior endometrial sampling. Mod Pathol 19:470–474. Scholar
  137. Scurry J, Planner R, Fortune DW et al (1993) Ligneous (pseudomembranous) inflammation of the female genital tract. A report of two cases. J Reprod Med 38:407–412PubMedPubMedCentralGoogle Scholar
  138. Seidman JD, Kurman RJ (1999) Tamoxifen and the endometrium. Int J Gynecol Pathol 18:293–296PubMedCrossRefPubMedCentralGoogle Scholar
  139. Shafiee MN, Seedhouse C, Mongan N et al (2016) Up-regulation of genes involved in the insulin signalling pathway (IGF1, PTEN and IGFBP1) in the endometrium may link polycystic ovarian syndrome and endometrial cancer. Mol Cell Endocrinol 424:94–101. Scholar
  140. Shah SS, Mazur MT (2008) Endometrial eosinophilic syncytial change related to breakdown: immunohistochemical evidence suggests a regressive process. Int J Gynecol Pathol 27:534–538. Scholar
  141. Shapiro S, Kelly JP, Rosenberg L et al (1985) Risk of localized and widespread endometrial cancer in relation to recent and discontinued use of conjugated estrogens. N Engl J Med 313:969–972. Scholar
  142. Shibutani S, Ravindernath A, Suzuki N et al (2000) Identification of tamoxifen-DNA adducts in the endometrium of women treated with tamoxifen. Carcinogenesis 21:1461–1467PubMedCrossRefPubMedCentralGoogle Scholar
  143. Shikone T, Kokawa K, Yamoto M et al (1997) Apoptosis of human ovary and uterine endometrium during the menstrual cycle. Horm Res 48(Suppl 3):27–34PubMedCrossRefPubMedCentralGoogle Scholar
  144. Silva EG, Jenkins R (1990) Serous carcinoma in endometrial polyps. Mod Pathol 3:120–128PubMedPubMedCentralGoogle Scholar
  145. Soslow RA, Chung MH, Rouse RV et al (1996) Atypical polypoid adenomyofibroma (APA) versus well-differentiated endometrial carcinoma with prominent stromal matrix: an immunohistochemical study. Int J Gynecol Pathol 15:209–216PubMedCrossRefPubMedCentralGoogle Scholar
  146. Staland B (1981) Continuous treatment with natural oestrogens and progestogens. A method to avoid endometrial stimulation. Maturitas 3:145–156PubMedCrossRefPubMedCentralGoogle Scholar
  147. Stastny JF, Ben-Ezra J, Stewart JA et al (1995) Condyloma and cervical intraepithelial neoplasia of the endometrium. Gynecol Obstet Investig 39:277–280CrossRefGoogle Scholar
  148. Stewart CJ, Michie BA, Kennedy JH (1998) Focal endometrial stromal hyperplasia. Histopathology 33:75–79PubMedCrossRefPubMedCentralGoogle Scholar
  149. Stewart CJ, Campbell-Brown M, Critchley HO et al (1999) Endometrial apoptosis in patients with dysfunctional uterine bleeding. Histopathology 34:99–105PubMedCrossRefPubMedCentralGoogle Scholar
  150. Sturdee DW, Ulrich LG, Barlow DH et al (2000) The endometrial response to sequential and continuous combined oestrogen-progestogen replacement therapy. BJOG 107:1392–1400PubMedCrossRefPubMedCentralGoogle Scholar
  151. Tabibzadeh S, Kong Q, Satyaswaroop P et al (1994) Distinct regional and menstrual cycle dependent distribution of apoptosis in human endometrium. Potential regulatory role of T cells and TNF-alpha. Endocr J 2:87–95Google Scholar
  152. Tahlan A, Nanda A, Mohan H (2006) Uterine adenomyoma: a clinicopathologic review of 26 cases and a review of the literature. Int J Gynecol Pathol 25:361–365. Scholar
  153. Tai LH, Tavassoli FA (2002) Endometrial polyps with atypical (bizarre) stromal cells. Am J Surg Pathol 26:505–509PubMedCrossRefPubMedCentralGoogle Scholar
  154. Taketani Y, Mizuno M (1991) Evidence for direct regulation of epidermal growth factor receptors by steroid hormones in human endometrial cells. Hum Reprod 6:1365–1369PubMedCrossRefPubMedCentralGoogle Scholar
  155. Takeuchi K, Tsujino T, Yabuta M et al (2012) A case of primary squamous cell carcinoma of the endometrium associated with extensive “ichthyosis uteri”. Eur J Gynaecol Oncol 33:552–554PubMedPubMedCentralGoogle Scholar
  156. Tantipalakorn C, Khunamornpong S, Lertprasertsuke N et al (2009) Female genital tract tumors and gastrointestinal lesions in the Peutz-Jeghers syndrome. J Med Assoc Thail 92:1686–1690Google Scholar
  157. Tao XJ, Sayegh RA, Tilly JL et al (1998) Elevated expression of the proapoptotic BCL-2 family member, BAK, in the human endometrium coincident with apoptosis during the secretory phase of the cycle. Fertil Steril 70:338–343PubMedCrossRefPubMedCentralGoogle Scholar
  158. The Writing Group for the PEPI Trial (1996) Effects of hormone replacement therapy on endometrial histology in postmenopausal women. The Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial. JAMA 275:370–375CrossRefGoogle Scholar
  159. Thomas W Jr, Sadeghieh B, Fresco R et al (1978) Malacoplakia of the endometrium, a probable cause of postmenopausal bleeding. Am J Clin Pathol 69:637–641PubMedCrossRefPubMedCentralGoogle Scholar
  160. Truskinovsky AM, Gerscovich EO, Duffield CR et al (2008) Endometrial microcalcifications detected by ultrasonography: clinical associations, histopathology, and potential etiology. Int J Gynecol Pathol 27:61–67. Scholar
  161. Turbiner J, Moreno-Bueno G, Dahiya S et al (2008) Clinicopathological and molecular analysis of endometrial carcinoma associated with tamoxifen. Mod Pathol 21:925–936. Scholar
  162. Usubutun A, Karaman N, Ayhan A et al (2005) Atypical endometrial stromal cells related with a polypoid leiomyoma with bizarre nuclei: a case report. Int J Gynecol Pathol 24:352–354PubMedCrossRefGoogle Scholar
  163. Val-Bernal JF, Villoria F, Cagigal ML et al (2006) Pneumopolycystic endometritis. Am J Surg Pathol 30:258–261PubMedCrossRefPubMedCentralGoogle Scholar
  164. Valeri RM, Ibrahim N, Sheaff MT (2002) Extramedullary hematopoiesis in the endometrium. Int J Gynecol Pathol 21:178–181PubMedCrossRefPubMedCentralGoogle Scholar
  165. Vang R, Tavassoli FA (2003) Proliferative mucinous lesions of the endometrium: analysis of existing criteria for diagnosing carcinoma in biopsies and curettings. Int J Surg Pathol 11:261–270PubMedCrossRefPubMedCentralGoogle Scholar
  166. Venkatraman L, Elliott H, Steele EK et al (2003) Serous carcinoma arising in an adenofibroma of the endometrium. Int J Gynecol Pathol 22:194–197PubMedCrossRefPubMedCentralGoogle Scholar
  167. Wani Y, Notohara K, Saegusa M et al (2008) Aberrant Cdx2 expression in endometrial lesions with squamous differentiation: important role of Cdx2 in squamous morula formation. Hum Pathol 39:1072–1079. Scholar
  168. Wells M, Tiltman A (1989) Intestinal metaplasia of the endometrium. Histopathology 15:431–433PubMedCrossRefPubMedCentralGoogle Scholar
  169. Wheeler DT, Bristow RE, Kurman RJ (2007) Histologic alterations in endometrial hyperplasia and well-differentiated carcinoma treated with progestins. Am J Surg Pathol 31:988–998. Scholar
  170. Winkler B, Reumann W, Mitao M et al (1984) Chlamydial endometritis. A histological and immunohistochemical analysis. Am J Surg Pathol 8:771–778PubMedCrossRefPubMedCentralGoogle Scholar
  171. Winter JS, Kohn G, Mellman WJ et al (1968) A familial syndrome of renal, genital, and middle ear anomalies. J Pediatr 72:88–93PubMedCrossRefPubMedCentralGoogle Scholar
  172. Yokoyama S, Kashima K, Inoue S et al (1993) Biotin-containing intranuclear inclusions in endometrial glands during gestation and puerperium. Am J Clin Pathol 99:13–17PubMedCrossRefPubMedCentralGoogle Scholar
  173. Young RH, Harris NL, Scully RE (1985) Lymphoma-like lesions of the lower female genital tract: a report of 16 cases. Int J Gynecol Pathol 4:289–299PubMedCrossRefPubMedCentralGoogle Scholar
  174. Young RH, Treger T, Scully RE (1986) Atypical polypoid adenomyoma of the uterus. A report of 27 cases. Am J Clin Pathol 86:139–145PubMedCrossRefPubMedCentralGoogle Scholar
  175. Zaloudek CJ, Norris HJ (1981) Adenofibroma and adenosarcoma of the uterus: a clinicopathologic study of 35 cases. Cancer 48:354–366PubMedCrossRefPubMedCentralGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  • Ricardo R. Lastra
    • 1
    Email author
  • W. Glenn McCluggage
    • 2
  • Lora Hedrick Ellenson
    • 3
  1. 1.Department of PathologyUniversity of ChicagoChicagoUSA
  2. 2.Department of PathologyRoyal Group of Hospitals TrustBelfastUK
  3. 3.Department of Pathology and Laboratory MedicineWeill Cornell Medical College and New York Presbyterian HospitalNew YorkUSA

Personalised recommendations