Objective
We sought to estimate cytologically benign endometrial cell (CBEC)-associated endometrial hyperplasia and cancer rates, and describe clinical and histologic outcomes.
Study Design
Medical records of women age >40 years with CBEC in 2005 through 2010 were reviewed for clinical characteristics; assessment with endometrial biopsy, ultrasound, or hysteroscopy; and consequent outcomes.
Results
Of 658 women, 281 (42.7%) were assessed: 39.4% of 330 premenopausal, and 46.0% of 328 postmenopausal women. Among these, cancer rate was 3.6% and differed between premenopausal (0.8%) and postmenopausal (6.0%) women ( P = .019). Hyperplasia rate was similar in premenopausal (3.9%) and postmenopausal (3.3%) women. Of 20 assessed women with endometrial pathology, 4 (1 premenopausal) women with cancer and 4 (2 premenopausal) women with hyperplasia had no abnormal bleeding.
Conclusion
Cancer was more common in postmenopausal women with assessed CBEC while hyperplasia was comparably distributed between premenopausal and postmenopausal women. Findings support CBEC assessment regardless of menopausal status or abnormal bleeding.
The finding of cytologically benign endometrial cells (CBEC) outside the proliferative phase of the menstrual cycle in premenopausal women or at anytime in postmenopausal women has been associated with endometrial malignancy. This association has been reported as uncommon in premenopausal women. Simsir et al reported 1 of 96 premenopausal women with CBEC and abnormal vaginal bleeding was found to have complex endometrial hyperplasia with atypia.
While most women with endometrial malignancy present with abnormal vaginal bleeding, 1 series reported that 2.1% of premenopausal women had CBEC as the only sign of endometrial pathology. Among postmenopausal women with CBEC, Simsir et al noted an 8% incidence of endometrial pathology and 5% incidence of endometrial cancer. Of the latter, 33% women were asymptomatic.
Some studies report an association between CBEC and hormone therapy, while others do not. Further, an increased incidence of endometrial malignancy has been shown in women with CBEC using tamoxifen.
There are few large studies focusing on the assessment and outcomes of CBEC. The largest series reported by Moroney et al had a 15% (370/2494) rate of endometrial sampling for CBEC. While several studies suggest endometrial sampling for all women age ≥40 years with CBEC, others support sampling for symptomatic premenopausal and all postmenopausal women only.
The 1991 Bethesda System for cytologic abnormalities recommended reporting CBEC in postmenopausal women only. The Bethesda 2001 System modified CBEC reporting to include all women ≥40 years because menopausal status and exogenous hormone use are often unknown or unclear in perimenopausal women. The Bethesda 2006 System affirmed CBEC reporting in women ≥40 years, noting that 0.5-1.8% of cervical cytology specimens from this group will include endometrial cells.
In our study, we estimate the rate of endometrial malignancy and hyperplasia in a large cohort of US women with assessment for CBEC. We describe the clinical presentation, current management, and outcomes of CBEC, thereby providing a view of clinical practice patterns.
Materials and Methods
Cervical cytology performed at Mayo Clinic from Jan. 2, 2005, through Oct. 20, 2010, and reported as “benign endometrial cells present in women age 40 and older” was reviewed. As part of clinical practice, the date of last menses was provided to the pathologist, allowing reporting of CBEC in women not currently menstruating (termed “out of phase”) and all postmenopausal women.
Women with a history of hysterectomy at the time of the cervical cytology and women who denied access to their medical records for research purposes were excluded. Medical record reviews were completed by 1 of 8 physician investigators. An initial subset of medical records was reviewed by multiple investigators for assessment of reliability and internal validity of the data collection instrument. The study was approved by the Mayo Clinic Institutional Review Board.
Medical records were abstracted for age, parity, race, body mass index (BMI), contraception, menopausal status, hormone therapy, presence and duration of vaginal bleeding, and clinical risk factors for endometrial pathology including diabetes, hypertension, and polycystic ovary syndrome at the time of first CBEC (index).
Women were classified as having abnormal bleeding if they were premenopausal and had reported bleeding other than menses; postmenopausal, receiving hormone therapy, and had reported bleeding other than withdrawal bleeding; or postmenopausal and receiving hormone therapy without cyclic withdrawal bleeding or not receiving hormone therapy and had reported any vaginal bleeding. Diagnostic assessment modalities and indicated interventions completed within 2 months before or at anytime following the index CBEC were abstracted. Assessment included ≥1 of the following: endometrial sampling, pelvic ultrasound, hysteroscopy, dilation and curettage, or hysterectomy. The 2-month preindex CBEC inclusion interval reasonably assured that only assessment pertinent to CBEC were included. Endometrial hyperplasia and cancer as well as findings pertinent to vaginal bleeding such as endometrial polyp, fibroids, and thickened endometrium were reported. Thickened endometrium was defined as >4-mm double-layer thickness in postmenopausal women only. There is no clear definition of thickened endometrium in premenopausal women, therefore it was not included among the abnormal findings for this group. In the subset of women with endometrial pathology, medical and surgical outcomes were collected along with a description of clinical symptoms including abnormal vaginal bleeding where present. The case of the less common nonendometrioid cancer received a confirmatory pathologic review. The subset of women with repeated CBEC was reviewed to describe evaluation and outcomes.
All information was recorded in a Research Electronic Data Capture (REDCap) database. REDCap is a secure World Wide Web–based application for capturing and managing research study data. The data were summarized using standard descriptive statistics, including frequency and percentage, mean, SD, median, and interquartile range (IQR). Patient characteristics at the time of the index CBEC were compared between postmenopausal and premenopausal women using the 2-sample t test for age and the χ 2 test for the categorical variables. The percentage of women with abnormal findings was compared between postmenopausal and premenopausal women using the χ 2 test. A 2-sided P value < .05 was considered statistically significant. Analysis was performed using a software package (SAS, version 9.2; SAS Institute Inc, Cary, NC).
Results
A total of 658 women met the study inclusion criteria. At index CBEC, 330 (50.2%) women were premenopausal and 328 (48.8%) were postmenopausal. Their clinical characteristics are summarized in Table 1 . Of the women included, 281 (42.7%) had assessment with ≥1 of the following: pelvic ultrasound, endometrial biopsy, or hysteroscopy. Of those with assessment, 130 women were premenopausal (39.4% of all premenopausal women) and 151 were postmenopausal (46.0% of all postmenopausal women). Based on these 281 women, the median interval from the index CBEC to start of assessment was 1 day (IQR, 0–20 days). The median interval to completion of assessment was 9 days (IQR, 0–39 days). An additional 111 women had a repeat cervical cytology only, 118 women had a clinical visit only, and 148 did not have further relevant health care provider contact at Mayo Clinic.
| Characteristic | Total (n = 658) | Premenopausal (n = 330) | Postmenopausal (n = 328) | P value a (n = 658) |
|---|---|---|---|---|
| Mean age, y (SD) | 53.5 (7.6) | 49.2 (3.8) | 57.9 (7.9) | < .001 |
| Age range, y | 40.0-91.8 | 40.0-58.4 | 40.5-91.8 | |
| Nulliparous, n (%) | 97 (14.7) | 51 (15.5) | 46 (14.0) | .60 |
| Race, n (%) | .86 | |||
| White | 617 (93.8) | 307 (93.0) | 310 (94.5) | |
| Non-white | 23 (3.5) | 11 (3.4) | 12 (3.7) | |
| Not recorded | 18 (2.7) | 12 (3.6) | 6 (1.8) | |
| BMI, kg/m 2 , n (%) | .083 | |||
| <25 | 243 (36.9) | 136 (41.2) | 107 (32.6) | |
| 25-29.9 | 202 (30.7) | 98 (29.7) | 104 (31.7) | |
| >30 | 207 (31.5) | 95 (28.8) | 112 (34.2) | |
| Not recorded | 6 (0.9) | 1 (0.3) | 5 (1.5) | |
| Contraception at index CBEC, n (%) | — | |||
| Yes | 161 (48.8) | |||
| Combined hormonal | 27/161 (16.8) | |||
| Progestin | 8/161 (5.0) | |||
| Nonhormonal | 110/161 (68.3) | |||
| Other | 16/161 (9.9) | |||
| No | 58 (17.6) | |||
| Not recorded | 111 (33.6) | |||
| HT at index CBEC, n (%) | 129 (39.3) | — | ||
| Systemic combined HT | 89/129 (68.9) | |||
| Systemic ET | 6/129 (4.7) | |||
| Local ET | 24/129 (18.6) | |||
| Tamoxifen | 8/129 (6.2) | |||
| Other HT | 2/129 (1.6) | |||
| Abnormal bleeding at index CBEC, n (%) b | .42 | |||
| Yes | 108 (16.4) | 49 (14.8) | 59 (18.0) | |
| No | 524 (79.6) | 260 (78.8) | 264 (80.5) | |
| Not recorded | 26 (4.0) | 21 (5.4) | 5 (1.5) | |
| Comorbidities, n (%) | ||||
| Diabetes mellitus | 41 (6.2) | 18 (5.5) | 23 (7.0) | .41 |
| Hypertension | 131 (19.9) | 41 (12.4) | 90 (27.4) | < .001 |
| PCOS | 5 (0.8) | 4 (1.2) | 1 (0.3) | .18 |
a Comparisons between 2 groups were evaluated based on 2-sample t test for age and χ 2 test for all other patient characteristics. Patients with “not recorded” values were excluded from each comparison
b Women were classified as having abnormal bleeding if they were premenopausal and had reported bleeding other than menses; postmenopausal, receiving HT, and had reported bleeding other than withdrawal bleeding; or postmenopausal and receiving HT without cyclic withdrawal bleeding or not receiving HT and had reported any vaginal bleeding.
The results of post-CBEC assessment are summarized in Table 2 . Of the 658 women, endometrial biopsy was obtained in 213 (32.4%), while 83 (12.6%) had hysteroscopy and 165 (25.1%) had ultrasound. All patients with endometrial pathology had endometrial biopsy obtained with or without hysteroscopic guidance. A comparable proportion of abnormal ultrasound and endometrial biopsy findings and a higher proportion of abnormal hysteroscopic findings were noted in the postmenopausal group as compared to the premenopausal group ( P = .037).
| Assessment type | Of 330 premenopausal women with CBEC | Of 328 postmenopausal women with CBEC | ||
|---|---|---|---|---|
| No. of patients a | No. (%) of patients with abnormal findings b | No. of patients a | No. (%) of patients with abnormal findings b | |
| Ultrasound | 84 | 29 (34.5) | 81 | 38 (46.9) |
| Hysteroscopy | 42 | 14 (33.3) | 41 | 23 (56.1) |
| Endometrial biopsy | 100 | 21 (21.0) | 113 | 23 (20.4) |
a Some patients had >1 evaluation modality;
b Includes thickened endometrium (>4 mm; for postmenopausal patients only), endometrial polyp, fibroid, endometrial hyperplasia, or endometrial cancer.
In the 330 premenopausal women, bleeding status was not recorded in 21. Of these, 6 had assessment, which was abnormal in 2 (33.3%) women. Another 260 premenopausal women had no abnormal bleeding as defined in the “Materials and Methods” section. Of these, 77 had assessment, which was abnormal in 27 (35.1%) women. Of the remaining 49 premenopausal women who presented with abnormal bleeding, 47 had assessment, which was abnormal in 20 (42.6%) women.
In the 328 postmenopausal women, bleeding status was not recorded in 5. Of these, 2 had assessment, which was abnormal in 1 woman. Another 264 postmenopausal women had no abnormal bleeding as defined in the “Materials and Methods” section. Of these, 94 had assessment, which was abnormal in 27 (28.7%) women. The remaining 59 postmenopausal women presented with abnormal bleeding. Of these, 55 had assessment, which was abnormal in 30 (54.6%) women.
The specific abnormal findings for the 49 premenopausal and 58 postmenopausal women with abnormal outcomes are summarized in Table 3 . Endometrial hyperplasia (n = 10) or cancer (n = 10) was identified in 6 premenopausal and 14 postmenopausal women with assessment as summarized in Table 4 . Four women with endometrial cancer had no abnormal vaginal bleeding at index CBEC, including 1 premenopausal woman. Four women with endometrial hyperplasia had no abnormal vaginal bleeding at index CBEC, including 2 premenopausal women. Therefore, 8 of 20 (40.0%) women, 3 of 6 (50%) premenopausal and 5 of 14 (35.7%) postmenopausal, with endometrial pathology had no abnormal vaginal bleeding at index CBEC. Within the subset of 171 women with assessment and no abnormal bleeding, 2.9% had endometrial pathology. Within the same subset, 3 (2.3%) of the 77 premenopausal and 5 (3.3%) of the 94 postmenopausal women had endometrial pathology.
| Outcome | Premenopausal, no. (% of 130) | Postmenopausal, no. (% of 151) | Total, no. (% of 281) |
|---|---|---|---|
| Endometrial cancer | 1 (0.8) | 9 (6.0) | 10 (3.6) |
| Hyperplasia | |||
| Complex, with atypia | 1 (0.8) | 3 (2.0) | 4 (1.4) |
| Complex, without atypia | 1 (0.8) | 1 (0.7) | 2 (0.7) |
| Simple, without atypia | 3 (2.3) | 1 (0.7) | 4 (1.4) |
| Thickened endometrium | N/A | 27 (17.9) | 27 (17.9) |
| Polyp | 20 (15.4) | 14 (9.3) | 34 (12.1) |
| Fibroids | 23 (17.7) | 3 (2.0) | 26 (9.3) |
| Overall | 49 (37.7) | 58 (38.4) | 107 (38.1) |
| Age, y | Menopause | BMI at index CBEC (kg/m 2 ) | Bleeding at index CBEC | Bleeding length (mo) | First bleeding episode | Abnormal bleeding a | HT | Outcome | Tumor | Treatment | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Hyperplasia | Grade | Stage | |||||||||
| 46 | Pre | 22.4 | Y | <1 | N | N | C atypia | 1 b | VH BSO | ||
| 50 | Pre | 30.6 | N | N | Simple | VH BSO | |||||
| 52.5 | Pre | 40.5 | Y | >12 | N | Y | Simple | Progestin | |||
| 53.9 | Pre | 32.3 | Y | 1-3 | Y | Y | Simple | D&C hysteroscopy (−) | |||
| 56 | Pre | 29.8 | Y | 1-3 | Y | Y | C no atypia | D&C (−) | |||
| 51.8 | Post | 19.4 | Y | 1-3 | N | Y | Y | C atypia | VH BSO | ||
| 53.5 | Post | — | Y c | <1 | Y | Y | N | C atypia | Robotic hysterectomy BSO | ||
| 56.1 | Post | 20.5 | Y | <1 | N | N | Y | C no atypia | Progestin | ||
| 56.9 | Post | 22.3 | Y | <1 | Y | N | Y | Simple | Progestin | ||
| 66.5 | Post | 26.4 | Y | 1-3 | Y | Y | Y | C atypia | VH BSO | ||
| Tumor type | |||||||||||
| 49.2 | Pre | 34.6 | N | N | Endometrioid | 2 | IV | TAH/BSO staging | |||
| 52.5 | Post | 27.9 | N | N | N | Endometrioid | 1 | 1A | TAH/BSO staging | ||
| 57.1 | Post | 41.3 | Y | >12 | N | Y | N | Endometrioid | 1 | 1B | TAH/BSO staging |
| 60.8 | Post | 32.1 | N | N | N | Endometrioid | 1 | 1A | TAH/BSO staging | ||
| 62.9 | Post | 37.8 | Y | <1 | Y | Y | N | Clear cell | 3 | 1A | TAH/BSO staging |
| 64.2 | Post | 30.2 | Y d | 1-3 | Y | Y | N | Endometrioid | 3 | 1A | TAH/BSO staging |
| 66.8 | Post | 42.5 | Y | <1 | Y | Y | N | Endometrioid | 1 | 1A | TAH/BSO staging |
| 67.1 | Post | 34.7 | Y | >12 | N | Y | N | Endometrioid | 2 | 1A | TAH/BSO staging |
| 68.3 | Post | 47.2 | N | N | N | Endometrioid | 3 | 1A | TAH/BSO staging | ||
| 77.6 | Post | 24.4 | Y | 4-6 | N | Y | N | Endometrioid | 1 | 1A | TAH/BSO staging |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
