association of tamoxifen and uterine sarcoma Association of Tamoxifen and Uterine Sarcoma

Association of Tamoxifen and Uterine Sarcoma

  1. Carolyn D. Runowicz
  1. NSABP Operations Center, Pittsburgh, PA
  2. NSABP Biostatistical Center, Pittsburgh, PA
  3. University of Southern California School of Medicine, and Norris Comprehensive Cancer Center, Los Angeles, CA
  4. St Luke’s Roosevelt Hospital Center, New York, NY, This letter was published online ahead of print at www.jco.org.

To the Editor:

Over the last 20 years, tamoxifen has become an important component of treatment for both early and advanced hormone receptor–positive breast cancer. In the late 1990s, clinical trial findings reported by the National Surgical Adjuvant Breast and Bowel Project (NSABP) demonstrated the utility of tamoxifen in ductal carcinoma-in-situ1 and in risk reduction for women at high risk for developing breast cancer.2 A consistent finding in our trials has been an increased risk of developing uterine (endometrial) cancer in women taking tamoxifen,3 a risk that seems to increase the longer the drug is administered. The mechanism for this increased risk is thought to be the estrogen agonist activity of tamoxifen on the uterus, which is similar to the mechanism for increased incidence of uterine cancer in women who take unopposed estrogen as hormone replacement therapy.4,5

Uterine sarcoma is a rare form of uterine malignancy, occurring in 2% to 5% of all patients with uterine malignancy,6,7 with an incidence of approximately one to two cases per 100,000 women in the general population. Its signs and symptoms are similar to those of endometrial cancer. There are three general types of uterine sarcoma: malignant mixed mullerian tumors (MMMT, also known as malignant mixed mesodermal sarcoma or carcinosarcoma), which contain both epithelial and stromal elements (50%), and leiomyosarcomas (40%) and stromal cell sarcomas (10%), both of which lack epithelial elements. Whereas estrogens have been shown to play a causal role in the pathogenesis of endometrial adenocarcinomas, such a link has not been established for uterine sarcomas in general, although it may exist for the MMMT variant.8-13 Although the prognosis for uterine sarcoma is primarily dependent on the stage and histology of the disease at the time of diagnosis, this malignancy, particularly the MMMT variant, tends to present at a more advanced stage and may carry a worse prognosis in terms of disease-free and overall survival.

In NSABP studies in which we categorized second malignancies according to organ site, we found a small number of uterine sarcomas (predominantly MMMT) among the endometrial cancers that occurred.3 It was unclear whether the incidence of these malignancies was increased in women taking tamoxifen. Because a recent case-control study14 has suggested that this may be the case, we updated data on the incidence of uterine malignancy that occurred in our patients and reviewed additional information from a Surveillance, Epidemiology, and End-Results (SEER)–based case-control study and from the manufacturer of tamoxifen, AstraZeneca (Wilmington, DE). Our aim was to further evaluate the possible association of tamoxifen use and an increased risk of uterine sarcoma.

In NSABP trials B-09, B-14, B-21, B-23, B-24, and P-1, more than 17,000 women were randomized to take tamoxifen or placebo (Table 1). The duration of treatment was 2 years in B-09 and 5 years in the other trials. On protocol B-14, patients who completed 5 years of tamoxifen had the option to be randomly assigned to receive either up to an additional 5 years of tamoxifen or placebo. The NSABP continues to obtain follow-up information on these patients, including data on the incidence of second malignancies, breast cancer recurrence and/or incidence, and survival. Table 2 shows the incidence of uterine malignancy by treatment group and by histologic type (adenocarcinoma or sarcoma) in these studies as of September 30, 2001. From these data, it seems that the incidence of both common adenocarcinoma and uterine sarcoma is increased in women taking tamoxifen, with sarcomas making up approximately 10% of total uterine malignancies in these patients.

View this table:

 NSABP Tamoxifen Trials

View this table:

 Incidence of Uterine Malignancies in NSABP Treatment Trials and in Breast Cancer Prevention Trial P-1

To examine this question further, staff at AstraZeneca reviewed all available data on tamoxifen in its global drug safety database through July 11, 2001, for the occurrence of uterine malignancies. Based on an average daily dose of 20 mg, the estimated worldwide patient exposure to Nolvadex (tamoxifen citrate), since its first market introduction, is more than 12 million patient years (information on file, AstraZeneca). AstraZeneca’s database contains worldwide literature reports of adverse events, serious adverse events from clinical trials, and all spontaneous postmarketing adverse event reports. From these, 942 uterine malignancies were identified, approximately 48% of them in the United States. Uterine cancer (including endometrial adenocarcinoma) was noted in 85% of the reports (802 reports) and uterine sarcoma in 15% (140 reports). All 140 reported cases of uterine sarcoma were further stratified according to histologic type. Seventy-three percent of these sarcomas were MMMT, approximately one third of which had a fatal outcome. In the NSABP data, nine of the 12 sarcomas found were MMMT or carcinosarcomas.

In a population-based series of 324 women diagnosed with endometrial cancer after breast cancer, identified by four SEER registries, the proportion of women with sarcomas was similar among those who had taken tamoxifen (11 of 146, 7.5%) and those who had not (12 of 178, 6.7%).15 Women with sarcomas had worse endometrial cancer–specific survival than did those with adenocarcinomas (P < .0001). However, the prognosis of women with uterine sarcomas or adenocarcinomas who had taken tamoxifen was not worse than that of women not exposed to tamoxifen; in fact, for both histologic subgroups, the probabilities of survival were greater among women treated with tamoxifen (median follow-up, 85 months).

This study, like the NSABP data and the report by Bergman et al,14 indicates that tamoxifen treatment is associated with an increased risk of both endometrial adenocarcinoma and uterine sarcoma. While the NSABP data suggest that the proportional increase in risk of sarcoma due to tamoxifen could be greater than that for adenocarcinoma, the SEER study data do not support that conjecture, nor is a biologic explanation for such a phenomenon readily apparent.

How should this information be incorporated into treatment decisions for women considering tamoxifen therapy? For women with hormone receptor–positive invasive breast cancer, tamoxifen has been demonstrated to improve relapse-free and overall survival. For women with ductal carcinoma-in-situ or for those who are at high risk for breast cancer for whom survival benefits have not yet been documented, the risks and benefits of tamoxifen therapy should be thoroughly assessed for each patient. Because patients with uterine sarcomas have been included in estimates of risk for endometrial cancer, the detailed risk-benefit statistics previously published16 can be used to determine which women may benefit from tamoxifen for reduction of breast cancer risk. Physicians should be aware that a small proportion of uterine malignancies that occur in women who take tamoxifen may represent uterine sarcomas. In deciding if tamoxifen therapy is warranted, all potential life-threatening adverse events associated with tamoxifen should be considered, including endometrial adenocarcinoma or uterine sarcoma, thromboembolic events, and stroke. When tamoxifen therapy is recommended, women at risk for endometrial cancer should undergo annual gynecologic examinations and should be counseled to seek prompt medical attention if they experience any gynecologic symptoms such as menstrual irregularities, vaginal bleeding, change in vaginal discharge, or pelvic pain or pressure.