Exogenous Hormones and their Effects on the Endometrium
Exogenous hormonal agents represent one of the most commonly prescribed medications in women. Hormonal therapies are used for a wide range of indications, including birth control, postmenopausal hormone replacement, dysfunctional uterine bleeding, endometriosis, infertility, and the treatment of malignant and premalignant lesions of the endometrium and breast. These drugs can be administered by many methods, including oral, parenteral, transdermal, transvaginal, and subcutaneous, vaginal, or intrauterine implants. Consequently, a significant proportion of endometrial specimens that the surgical pathologist evaluates show the effects of these exogenous hormonal agents. At first glance, the spectrum of changes seen with hormone therapy appears excessively broad, and it is certainly true that nearly any endometrial appearance can be attributed to therapy with hormones. With a basic understanding of the effects of estrogens and progestins on the normal endometrium, however, substantial order emerges from this confusing pathologic array. Essentially, it is possible to predict the common appearances of hormone therapy based on the known effects of the endogenous steroid hormones on normal endometrium.
The basic effect of estrogens on the endometrium is to induce proliferation of the endometrial glands and stroma, including vascular endothelium. The degree of proliferation can vary in proportion to the estrogenic stimulus. Very low levels of estrogen or a very weak estrogen will lead to an inactive or atrophic endometrium. Higher levels lead to proliferation, which when excessive can produce a hyperplastic endometrium (Figure 15.1). The degree of proliferative activity can usually be assessed by the mitotic activity in both the glandular epithelium and the stroma. Women who are many years postmenopausal demonstrate profound endometrial atrophy, secondary to lack of estrogen, but even atrophic endometrium remains estrogen responsive to quite advanced age (Figure 15.2). A number of estrogenic drugs on the market are listed in Table 15.1.1
|Generic Name||U.S. Brand Name|
|Estradiol valerate||Delestrogen, Gynogen, Valergen|
|Estradiol (transdermal)||Estraderm, Climara, Fempatch|
|Conjugated equine estrogens||Premarin|
|Synthetic conjugated estrogens||Cenestin, Enjuvia|
Persistent exposure to a significant estrogenic stimulus, such as occurs in anovulation, leads to a pattern of continued, unrelieved proliferation. The endometrium cannot support such continued growth. Whether the estrogen source is endogenous or exogenous, the histologic consequence is the same: a combination of proliferative endometrium with episodic coexisting stromal breakdown, or shedding, which accounts for the clinical presentation of irregular bleeding (Figure 15.3). This pattern is also called ‘anovulatory bleeding’ or ‘anovulatory shedding.’ Because the endometrial vessels become abnormally large, bleeding can also be quite severe. Some pathologists mistakenly refer to the shedding endometrium as ‘menstrual,’ but it is important to distinguish this appearance from the appearance of menstrual endometrium occurring at the end of a normal cycle. In normal endometrium, the endometrial glands transform soon after ovulation to a secretory pattern and by the time menstrual bleeding begins will still show some residual changes, most commonly secretory exhaustion (see Chapter 14). This change is absent in anovulatory-type bleeding or where the drug administered is progestin poor. Another common difference is the presence of large fibrin thrombi in the vessels of anovulatory shedding endometrium (Figure 15.4). They are absent from menstrual endometrium because of the marked fibrinolytic activity typical of normal menstruation.
Figure 15.3 ‘Persistent’ proliferative endometrium with unopposed estrogen effect and secondary breakdown. Present is proliferative endometrium with scattered cysts and stromal breakdown forming stromal balls and collapsed eosinophilic epithelium. Often associated with fibrin thrombi, this pattern differs from a physiologic late menstrual–early proliferative endometrium in nonuniform distribution of breakdown throughout a thickened, dense, nonspindly stroma.
With prolonged estrogen exposure, the proliferating glands tend to lose their uniformity of size, shape, and distribution, leading to so-called disordered proliferative endometrium. Cystic dilatation of glands and ‘tubal’ metaplasia are commonly present. With continued, longer term, unopposed estrogen exposure, a high proportion of patients will ultimately develop non-atypical endometrial hyperplasia. A small number will also go on to develop endometrial intraepithelial neoplasia (EIN) (also called ‘atypical hyperplasia’) or even adenocarcinoma (see Chapters 17 and 18).
The effect of progestins on the endometrium depends on ‘priming’ by estrogen, which induces progesterone receptors in the endometrial cells. One important feature of progestins is that they act to downregulate estrogen and progesterone receptors; in other words, they reduce the sensitivity of the endometrium to both of these hormones. Prolonged exposure to progestins can thus lead to a histologic picture that is paradoxically similar to the atrophy seen in a postmenopausal or hormone-suppressed patient. Some commonly used progestins are shown in Table 15.2.1
Adapted from Schimmer and Parker (2012).1
|Generic Name||Brand Name|
|Medroxyprogesterone||Provera, Prodrox, Cycrin, Amen, Curretab|
|Levonorgestrel||Norplant, Plan B|
In the short term, however, progestins induce secretory differentiation in endometrial glands and decidual-type change in the stroma, i.e., the classic changes of a normal secretory endometrium. The glands develop large glycogen vacuoles that are then secreted into the increasingly complex gland lumina (Figures 15.5 and 15.6). At the same time, the stromal cells enlarge strikingly, acquire an abundant cytoplasm, and appear relatively cohesive (Figure 15.7). Even when pathology is present, the stroma may demonstrate pseudodecidual changes, but progesterone responsiveness is often diminished in disease states such as endometritis, and in polyps.
Figure 15.5 Secretory endometrium, demonstrating exogenous progesterone effect. The appearance is indistinguishable from a normal cycle. The woman, who was older, had received progesterone for 20 days.
With continued exposure or repeated cycles, the receptor downregulation causes the glands to lose their sensitivity to estrogens and progestins, leading to progressively less secretory change (‘secretory exhaustion’), such that they ultimately appear atrophic. At this stage, which only occurs after prolonged exposure or multiple cycles, the appearance is that of decidualized stroma with widely dispersed atrophic glands, often referred to as a ‘pill’ endometrium (Figure 15.8). More slowly, the decidualized stroma also begins to be suppressed by the receptor downregulation, and over months to years will become attenuated and atrophic, ultimately losing most of its decidual features. At this point, the appearance resembles endometrial atrophy due to menopause or hormone suppression.
Exposure to high-dose progestins (often given for dysfunctional uterine bleeding and usually with a good initial response) can, paradoxically, cause secondary necrosis of the superficial endometrium, often in the form of wedge-shaped infarcts, and renewed bleeding. The hysteroscopic appearances are often alarming. Both the continued bleeding and the hysteroscopic findings lead the clinician to perform a dilatation and curettage to exclude hyperplasia or malignancy and yielding abundant, macroscopically suspicious tissue fragments. Occasionally, patients will spontaneously pass large intact sheets of decidualized endometrium (‘decidual casts’).
Oral contraceptives (OCs) are one of the most widely used medications in the developed world. They are used as treatment for several common medical problems in addition to contraception, including dysfunctional uterine bleeding and endometriosis. A variety of formulations and administration schedules have been developed (Table 15.3).1 Sequential OCs contain only estrogen during the first half of the cycle, with progestin added during the second half. Because of the increased risk of endometrial cancer with some sequential formulations,2 the most commonly used types today are combined estrogen–progestin and progestin-only formulations administered over 21 consecutive days of the cycle, followed by 7 days of placebo tablets yielding a withdrawal week. Combined pills can be monophasic, in which there is a fixed dose of estrogen and a progestin in combination for the 21 days of each cycle, or they can be biphasic or triphasic, depending on whether the progestin dose is altered once or twice during the cycle. In contrast to some sequential OCs and estrogen-only hormone replacement regimens, which increase the risk of endometrial carcinoma, combined OCs are protective, with the degree of protection increasing with the length of therapy. Patients with 10 years of OC use have about a 75% reduction in endometrial carcinoma.3,4 OC use is also associated with a 30–50% decrease in the risk of ovarian carcinoma; this lowered risk persists for at least 20 years after cessation of their use and is also seen in BRCA1 and BRCA2 mutation carriers.5 Recently, a variety of extended cycle (91 day) and even continuous cycle (365 day) OCs have been introduced that appear to have similar effectiveness to traditional OCs.6 The histologic effects of these newer formulations have not been well studied.
Adapted from Schimmer and Parker (2012).1
The histologic appearance of the endometrium in a patient on combined OCs is extremely variable, but is dominated by the progestin effects. It depends on various factors, some poorly understood, including the type of pill, the duration of therapy, precise dose of the individual pill, levels of compliance with the regimen, and endogenous hormone synthesis and metabolism.7–9 Certain generalizations are possible, however. Within the first several cycles of a combined OC, there is a mixture of proliferative and secretory features seen in the endometrium (‘asynchronous’ or ‘discordant’ endometrium). The glands tend to remain relatively straight and narrow, resembling proliferative endometrial glands but with cuboidal rather than columnar epithelium, and with very minimal mitotic activity. Subnuclear vacuoles may be seen, especially during the first 2 weeks of the cycle. Decidual change can be seen early but is usually more evident after several cycles have been completed (Figures 15.9 and 15.10). Despite the stromal changes, spiral arterioles do not develop normally.10 With prolonged therapy over many cycles the glands develop secretory exhaustion and become increasingly small, inactive, and ultimately atrophic. As this occurs, the stromal decidual changes become increasingly well developed, until the endometrium is uniformly decidualized, with only rare, atrophic-appearing glands (Figures 15.11 and 15.12). This is the classic appearance of the so-called ‘pill’ endometrium. Prominent ectatic thin-walled veins are also seen, which often contain thromboses in patients having breakthrough bleeding (Figure 15.13). The classic pill endometrium is not seen in all patients, and there is a spectrum from well-developed stromal changes to complete atrophy, with only a very thin endometrium showing little or no identifiable decidual change. The atrophic changes are seen more frequently with some of the lower dose regimens.8
Formulations that include short estrogen-only periods have been used in an attempt to lower the overall dose of hormones. Mircette, for example, has 21 days of combination treatment followed by a 2 day, hormone-free interval and 5 days of unopposed estrogen. Technically, this is a sequential OC, but to date it has not had the safety issues associated with the older, higher dose sequential therapies. This formulation has a distinctly different histologic pattern than the combined OCs. Biopsies taken during the estrogen-only portion of the cycle show proliferative endometrium while biopsies taken during the combined portion of the cycle show secretory endometrium.11
In contrast to the combined OCs, the pure progestin OC (the mini-pill, 300 μg norethisterone) is taken daily without interruption during the cycle. The mechanism of contraception differs from combined OCs in that ovulation is not consistently suppressed. Instead, contraception is due to the production of relatively thick cervical mucus that is impermeable to sperm, and to the atrophic endometrium that will not support implantation. The endometrial changes seen with the mini-pill are not distinctly different from those seen with combined OCs, but marked atrophy is more common and biopsy frequently yields only scanty material. In addition, proliferative endometrium is not infrequently seen, and this correlates with abnormal bleeding.12
Several systems have been developed for long-term, progestin-only contraception. Medroxyprogesterone acetate microcrystals in an aqueous solution (Depo-Provera) are used as an intramuscular injection. The slow dissolution of the crystals maintains effective progestin levels for several months, allowing for an injection schedule of every 3 months. Levels of progestin are typically higher than with the mini-pill. Early after injection, some women develop exaggerated hypersecretory changes that resemble gestational endometrium, including the presence of Arias-Stella reaction. By 3–6 months these changes have resolved and stromal decidual changes develop, resembling those seen with combined OCs as described earlier. Long-term treatment can result in atrophy or pure decidua-like change (Figure 15.14), just as with combined OCs.8,13
The Norplant system consists of several Silastic tubes containing levonorgestrel. The tubes are slightly permeable to steroids, and the progestin is slowly released for a period of 3–5 years. As in other progestin-only regimens, ovulation occurs in about one-third of cycles, and contraception is by production of thick cervical mucus and an endometrium inimical to implantation. There are relatively few studies on the histologic findings with Norplant,14,15 but the findings seem to be similar to other progestin-only regimens, with atrophy being the most common finding. Proliferative endometrium can also be seen, and correlates with irregular bleeding patterns.16
Devices that supply progestins directly to the endometrium are also used. The most common of these uses a slow-release formulation of levonorgestrel in combination with an intrauterine device (Mirena). Although physically located within the endometrium, absorption is systemic and they can be used for treatment of extrauterine disease such as endometriosis as well as abnormal uterine bleeding.17–19 Histologically, the endometrium shows extensive decidualization.20
Three general types of hormone replacement therapy (HRT) have been used clinically: unopposed estrogen, cyclic estrogen and progestin, and combined estrogen and progestin formulations. Because of the markedly increased risk of endometrial cancer with unopposed estrogen, only the last two are in common use today. While the exact agents used and their dosage can vary, within each of these groups the histologic findings tend to be similar, and thus they will be discussed as categories.
The number of patients receiving long-term (>5 years) HRT dropped precipitously after large studies showed an increase in cardiovascular risk rather than the predicted protective effects.21 In the United States, the number of prescriptions for HRT in the national Medicaid program fell by 57% between 2002 and 2004.22 Nonetheless, HRT continues to be commonly used as short-term therapy for symptoms related to menopause.
Historically, estrogen alone was given as hormone replacement for postmenopausal women, and the beneficial effects of this therapy in preventing osteoporosis and cardiovascular disease have been well documented. Then a series of case–control studies subsequently demonstrated a markedly increased risk of endometrial adenocarcinoma in women using long-term estrogens unopposed by progestins. Patients receiving unopposed estrogens for 5 years or more have an approximately sixfold increase in endometrial carcinoma. With the addition of progestins, the risk of endometrial carcinoma fell to near control levels.23–25 The risk of endometrial carcinoma is dose related, but low-dose estrogens (0.3 mg conjugated estrogens or equivalent) continue to be associated with a risk of carcinoma if not combined with a progestin.26