The Effect of Tamoxifen on the Endometrium
The Effect of Tamoxifen on the Endometrium
Tamoxifen (Nolvadex), a nonsteroidal antiestrogen, was first approved by the FDA for the treatment of patients with breast cancer in 1978. Large clinical trials have demonstrated a recurrence-free and overall survival benefit in both pre- and postmenopausal women. Long-term adjuvant tamoxifen is the endocrine treatment of choice for selected patients with breast cancer, and large-scale trials are currently underway to evaluate its role as a chemopreventive agent in healthy women at risk for breast cancer. Consequently, a large number of women will be subjected to both the benefits and potential risks of long-term tamoxifen therapy. One of the most significant potential complications is the development of endometrial cancer. The estimated annual risk of endometrial cancer in tamoxifen-treated patients is approximately 2 per 1,000 women. Most of these cancers will be detected at an early stage when they are highly curable. The potential benefit of tamoxifen treatment in breast cancer patients outweighs this risk; however, all patients receiving tamoxifen should undergo regular gynecologic evaluations.
Tamoxifen (Nolvadex), a nonsteroidal antiestrogen, has been widely used for the management of patients with breast cancer since the 1970s. It is believed to exert its main effect by blocking the binding of estrogen to the estrogen receptor. Although primarily an antiestrogen, tamoxifen may also exhibit some mild estrogenic effects. Since the initial report by Killackey et al  suggesting a possible link between tamoxifen use and the development of endometrial carcinoma in three patients, approximately 133 additional cases of tamoxifen-associated uterine cancer have been reported [2-13]. Most recently, the results of the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14 trial were published . This randomized trial of tamoxifen versus placebo in women with estrogen receptor (ER)-positive breast cancer confined to the breast with negative axillary nodes revealed a 7.5-fold increase in the risk of developing endometrial cancer in the tamoxifen-treated group.
The indications for tamoxifen use have broadened to include long-term adjuvant therapy as well as preventive therapy for selected high-risk women. Consequently, a large number of women, including healthy young patients with no history of cancer, will be subjected to the long-term effects of tamoxifen. This review evaluates the current literature regarding tamoxifen use in breast cancer patients and associated uterine neoplasia, and discusses the role of screening for endometrial cancer in tamoxifen-treated breast cancer patients.
Several preclinical studies have indicated that tamoxifen may indeed exert an estrogenic effect on the endometrium. Satyaswaroop et al  transplanted both ER-positive and ER-negative human endometrial cancer cell lines into nude mice and evaluated the effect of both tamoxifen and 17-beta-estradiol on tumor growth. Although the ER-negative tumor grew rapidly, there was no difference in the rate of tumor growth between tamoxifen- and estrogen-treated animals, compared with controls. In contrast, the rate of growth of the ER-positive cell line was significantly accelerated in the tamoxifen-treated animals, compared with controls, although to a lesser degree than was seen with estrogen treatment. In addition, tamoxifen increased the levels of functional progesterone receptors, providing further evidence for the estrogenic potential of tamoxifen.
Gottardis et al  demonstrated in athymic mice the contrasting actions of tamoxifen on the growth of ER-positive breast and endometrial cancer cell lines. While stimulating the growth of the endometrial tumor, tamoxifen had no effect on breast tumor growth and had an antagonistic action on the estradiol-stimulated growth of breast tumors.
The estrogenic effect of tamoxifen on the female genital tract has also been demonstrated clinically. Boccardo et al  investigated the estrogen-like effect of tamoxifen on the vaginal epithelium. Hormonal evaluation of the female patient can be determined by means of exfoliative vaginal cytology using the karyopyknotic index (KPI), which is the relation of mature superficial cells to intermediate cells. Postmenopausal estrogen replacement therapy generally induces increased cellular maturity and, consequently, increases the KPI. After at least 8 weeks of tamoxifen treatment, treated patients had a significantly higher KPI than did untreated patients, suggesting an estrogenic effect of tamoxifen on the vaginal epithelium.
Several recent reports have implicated tamoxifen treatment in the development of endometrial polyps and hyperplasia (Table 1 ). Lahti et al  performed transvaginal sonography, hysteroscopy, and endometrial curettage on 103 postmenopausal breast cancer patients; 51 had received tamoxifen at 20 to 40 mg/day for a median of 30 months, while 52 had received no hormonal treatment. The patients were matched with respect to age, parity, time since menopause, body mass index, and concomitant medical conditions. Compared with control subjects, the tamoxifen-treated patients had a significantly thicker mean endometrial width (10.4 mm versus 4.2 mm) and a larger uterine volume, as determined by transvaginal sonography. However, they also had a higher incidence of uterine fibroids.
The percentage of endometrial polyps was significantly higher in the tamoxifen-treated group, compared with controls (36% vs 10%). Of the 43 patients in the tamoxifen-treated group who had an endometrial thickness of 5 mm or more, 17 (39.5%) had endometrial polyps; however, 22 (51.2%) had no abnormal pathologic findings. Two patients in the tamoxifen group had endometrial hyperplasia, and one patient had endometrial cancer at the time of curettage. Two patients in the control group also had endometrial cancer. All significant endometrial pathology was found in asymptomatic patients with an endometrial width greater than 5 mm on transvaginal sonography. If transvaginal sonography were used as a screening method for endometrial cancer with the abnormal cutoff limit set at 5 mm, approximately 50% of patients would undergo endometrial sampling unnecessarily.
Since endometrial polyps often contain hyperplastic endometrial glands, their etiology is thought to be the presence of endogenous or exogenous estrogenic activity. The authors therefore concluded that long-term tamoxifen use is associated with estrogenic side effects manifested by an increased occurrence of polyps. This study failed to demonstrate any increase in precancerous or cancerous endometrial lesions associated with tamoxifen treatment.
Other Studies—Other authors have also reported a higher incidence of endometrial polyps and hyperplasia in tamoxifen-treated breast cancer patients. Corley et al  reported on four tamoxifen-treated patients who developed unusual endometrial polyps characterized by cystically dilated glands and stromal decidualization. These findings were not confirmed in the 17 cases of endometrial polyps reported by Lahti et al .
Gal et al  reported an 18% incidence of endometrial hyperplasia in random endometrial biopsies obtained from 38 breast cancer patients who had received tamoxifen at a dosage of 20 mg/day for at least 12 months. Thirty-two patients (46%) who were initially eligible for study could not participate, as endometrial samples could not be obtained due to atrophic changes. In the second phase of the study, the authors performed endometrial biopsies prospectively on 11 patients at 4- to 6-month intervals. Three (28%) of these patients developed endometrial hyperplasia following 6, 10, and 12 months of tamoxifen use. Again, 40% of the patients initially eligible for study could not participate due to inability to undergo an endometrial biopsy. The 20% overall incidence of endometrial hyperplasia reported by Gal et al  was higher than the 4% incidence noted by Lahti et al  in 51 tamoxifen-treated breast cancer patients.
Gibson et al  recently reported the results of a retrospective review of endometrial pathology found at dilatation and curettage (D&C) performed in 240 breast cancer patients. Seventy-five (31%) of these patients had received tamoxifen, 20 mg/day, for a mean duration of 26 months as therapy for their breast cancer. Patients in this study were stratified as symptomatic (abnormal bleeding) or asymptomatic (curettage performed as part of another procedure or due to the presence of a thickened endometrial stripe on ultrasound examination).
Among tamoxifen users who were symptomatic at the time of D&C, the incidence of endometrial polyps was 15%; hyperplasia, 2%; and adenocarcinoma, 11%; compared with an incidence of 13%, 4%, and 11%, respectively, for nonusers. Among asympto- matic patients receiving tamoxifen, the incidence of polyps, hyperplasia, and adenocarcinoma was 9%, 0%, and 0%, while nonusers had an incidence of 5%, 4%, and 0%, respectively. All of the cases of endometrial carcinoma were associated with abnormal vaginal bleeding. The 11% incidence of endometrial carcinoma in this study was comparable to what has been reported for all patients presenting with postmenopausal bleeding in the general population .
The mean duration of tamoxifen use in those patients found to have adenocarcinoma was 44 months, compared with 18 months in those patients with polyps, and 19 months in the patient with hyperplasia. This difference was not statistically significant, possibly due to small numbers. The authors concluded that short-term tamoxifen use at the 20 mg/day dosage level was not associated with a higher incidence of abnormal pathology, as detected at the time of D&C in breast cancer patients.
Breast Cancer Prevention Trials—Recently, large-scale randomized trials of tamoxifen as a chemopreventive agent have been initiated in England and the United States for women considered to be at high risk for developing breast cancer based on family history. Kedar et al  recruited a randomized cohort of 111 postmenopausal women from the Pilot Breast Cancer Prevention Trial at the Royal Marsden Hospital, to study the effect of tamoxifen on the uterus and ovaries. Sixty-one patients had received tamoxifen at 20 mg/day for a median of 22 months, while 50 had received placebo for a median of 24 months. Eight patients in each group were receiving estrogen replacement therapy at the time of study.
All patients underwent endovaginal sonography with color Doppler imaging, after which an endometrial biopsy was performed. The tamoxifen group was found to have a significantly larger uterine volume and a lower impedance to blood flow in the uterine arteries. Patients were not, however, stratified as to the presence of uterine fibroids. In addition, the tamoxifen-treated group had a significantly thicker mean endometrial diameter (9.1 mm versus 4.8 mm). On endometrial biopsy, 10 patients (16%) in the tamoxifen-treated group were found to have atypical endometrial hyperplasia, while five (8%) had endometrial polyps. None of the patients in the control group had hyperplasia on endometrial biopsy, and only one patient had an endometrial polyp. No patient in either group had evidence of endometrial carcinoma. The predictive value of an endometrial thickness 8 mm or larger for atypical hyperplasia or polyps was 100% (16/16), suggesting a possible screening role for transvaginal sonography in breast cancer patients receiving tamoxifen.