Endometrial Dating

Endometrial dating refers to the determination of how closely the histologic characteristics of the endometrium match what is expected on the corresponding day of the menstrual cycle. In the past, this approach was one of the standard tests in an investigation of causes of infertility and pregnancy loss. However, the accuracy of this test has been questioned because abnormal results can be observed in cycles that eventually prove to result in a viable pregnancy.

Endometrial dating can be performed both before and after ovulation. Preovulatory phase (proliferative phase) endometrial dating is described as menstrual days, early follicular, midfollicular, and late follicular, but is not very precise. Postovulatory phase (secretory phase) endometrial dating has become the standard and is usually reported within a 2-day range, although the accuracy of this dating methodology is the subject of some debate.

For secretory phase endometrial dating, ovulation is used as the primary reference point. Originally, ovulation was assumed to occur on day 14 of the menstrual cycle and the days after ovulation numbered accordingly. Some pathologists refer to ovulation as “day 0” and report the postovulation day as the number of days after ovulation.

Most clinicians perform endometrial biopsy in the midluteal phase at about the time implantation is thought to occur. However, the original reports on secretory endometrial dating recommend a biopsy 3 days before the expected menses. Histologic criteria are then used to determine where the endometrial response would be in relation to ovulation (Table 8-1 and Figs. 8-1 and 8-2).

Table 8-1 Criteria for Histologic Dating

Figure 8-1 Midluteal phase endometrium has early spiral arteriolar formation (center) arising in edematous stroma. Endometrial glands are tortuous with basal nuclei, absence of mitotic activity, and intraluminal secretions.

Figure 8-2 Late luteal phase endometrium has large areas of coalescent predecidua and a conspicuous lymphoid infiltrate, giving the stroma a dimorphic appearance.

Several assumptions in the original concept of dating the endometrium besides the assumption of ovulation on day 14 increased the variation found with this method. For example, another assumption was that the length of the luteal phase is 14 days.¹ In reality, there is a normal variation in luteal phase length of several days. In addition, the original description fixed the time of ovulation with the onset of the period after the biopsy. The onset of ovulation can be more accurately determined using current modalities, such as determination of the midcycle urinary luteinizing hormone surge or ultrasonic identification of the collapse of a follicle. The accuracy of the former is approximately 85%; of the latter, 95%. Additional intrinsic inaccuracies of dating resulted from intraobserver and interobserver variation. This variation is typically about 2 days. For these reasons, a 2-day difference between the histologic estimation and the actual interval since ovulation is considered within normal limits.

Recently a detailed analysis on endometrial dating demonstrated that the histologic criteria used are not as temporally distinct as originally thought, and thus do not provide an accurate method to detect a luteal phase defect. In one study, approximately 20% of fertile couples had a delay of more than 2 days.² A between-cycle variation of more than 2 days was found in 30% to 60% of patients if the biopsy was performed between day 6 and day 13 after ovulation.

Endometrial Response to Exogenous Hormones

The endometrium responds to exogenous hormones with specific morphologic changes. With oral contraceptives, the progestin effect dominates because progestins decrease estrogen receptors. As a result, endometrial glands atrophy over time. The appearance of the stroma depends on the dose of progestin. Typically, a weak pseudodecidual response will occur (Fig. 8-3).

Figure 8-3 The microscopic appearance of the endometrium exposed to oral contraceptives varies, depending on the dose of progestin. Typically, stroma having a weak progestin effect surrounds simple, straight, widely spaced, and inactive-appearing endometrial glands.

Exposure to progestins only (e.g., medroxyprogesterone acetate) will show similar gland pattern but usually a more prominent stromal pseudodecidual response due to higher progestin dose. The selective estrogen receptor modulator tamoxifen is anti-estrogenic at some sites but has a weak estrogenic agonist effect on the endometrium. The endometrium usually atrophies, but variable hyperplasia and, rarely, adenocarcinoma can occur due to the weak estrogen agonist effect.

Endometritis

Chronic endometritis is sometimes demonstrated in endometrial biopsies for infertility or recurrent pregnancy loss. This is characterized by infiltration of plasma cells (Fig. 8-4). Chronic pelvic inflammatory disease, intrauterine devices, and postabortion complications are also associated with chronic endometritis. Although rare in the United States, endometrial tuberculosis may sometimes be found during a biopsy for infertility and is characterized by granulomas (Fig. 8-5).

Figure 8-4 The presence of plasma cells defines chronic endometritis. Typically, the endometrium resembles a proliferative-type endometrium. The stroma often has a spindled appearance.

Figure 8-5 Tuberculosis endometritis causes granulomatous inflammation. In reproductive age women, granulomas tend to be cellular but not markedly necrotic. The case illustrated here has a well-formed nonnecrotizing granuloma (right) adjacent to an endometrial gland. In postmenopausal patients, tuberculous endometritis causes striking necrotizing granulomatous endometritis.

Abnormal Uterine Bleeding

The biopsy of an endometrium for dysfunctional uterine bleeding in women may show a range of abnormalities of the endometrium (Fig. 8-6). In the reproductive age group organic lesions such as polyps and leiomyomas or a normal endometrium are often found. Pregnancy-related endometrial changes or premalignant or malignant changes may be found in this age group.

Figure 8-6 This endometrial biopsy specimen illustrates gland–stromal asynchrony. The endometrial glands have consistent and conspicuous subnuclear vacuoles corresponding to approximately postovulatory day 3 (cycle day 17). In contrast, the endometrial stroma has striking stromal edema corresponding to approximately postovulatory day 8 (cycle day 22). This appearance of gland–stromal asynchrony is the most common secretory phase morphologic abnormality.

In adolescent and perimenopausal age patients, anovulatory cycles are one of the most common causes of abnormal uterine bleeding. During anovulatory cycles, the endometrium is proliferative and may exhibit some breakdown. Thrombosis of vascular sinusoids also commonly occurs. Anovulatory cycles create a milieu of unopposed estrogen; therefore, hyperplasia and carcinoma can result.

Pregnancy-related Endometrial Changes

Early pregnancy, both intrauterine and ectopic, is characterized by hypersecretory endometrium (Fig. 8-7). However, hypersecretory endometrium is not specific for pregnancy, and similar changes can be seen with persistent corpus luteum cyst, double corpora lutea, or rarely as a drug effect. By the end of the first trimester, endometrial glands involute and the stroma shows a prominent decidual reaction (Fig. 8-8). Other histologic changes associated with gestation include the Arias-Stella reaction (Fig. 8-9) and optically clear nuclei (Fig. 8-10).

Figure 8-7 Hypersecretory endometrium has closely packed endometrial glands with marked intraglandular budding (ferning) and ongoing glandular secretions with secretory vacuoles. Hypersecretory endometrium characterizes early pregnancy; however, this change is not specific for gestation.

Figure 8-8 By the end of the first trimester, the hypersecretory glandular changes resolve. Subsequently, gestational endometrium has mostly simple widely spaced endometrial glands, separated by stroma with a prominent progestin effect (true decidual change), as seen in this example.

Figure 8-9 This photomicrograph illustrates the Arias-Stella reaction. The Arias-Stella reaction is a cellular and nuclear change characterized by nuclear enlargement and hyperchromatisma resulting from polyploidy. The enlarged hyperchromatic nuclei are typically arranged in an apical portion of the endometrial glandular cells. This nuclear change can be mistaken for malignancy; however, the Arias-Stella reaction, unlike adenocarcinoma, typically affects only scattered cells within the endometrial glands and mitotic figures are absent or at most infrequent.

Figure 8-10 This endometrial gland has the optically clear nuclear appearance seen with pregnancy (center top).

The Arias-Stella reaction refers to cytonuclear changes, including voluminous, vacuolated cytoplasms surrounding enlarged, hyperchromatic, polyploid nuclei that includes massively enlarged forms. The Arias-Stella reaction occurs almost exclusively in gestational endometrium associated with pregnancy or gestational trophoblastic disease; however, this reaction can rarely occur as a response to hormonal therapy in nonpregnant patients. The optically clear nuclei associated with gestation can resemble herpes virus inclusions.