Routine Endometrial Biopsy ‘of Limited Value’ in Tamoxifen Users

ATLANTA—A prospective study, presented at the 35th Annual Meeting of the American Society of Clinical Oncology, has found the utility of routine endometrial biopsy among women with breast cancer treated with tamoxifen (Nolvadex) to be “limited at best.” Another study presented at the meeting finds sonography to be inadequate as a substitute for endometrial biopsy in healthy women receiving tamoxifen prophylaxis.

Carolyn D. Runowicz, MD, professor and director, Division of Gynecologic Oncology, Albert Einstein College of Medicine and Montefiore Medical Center, New York, reported on a subgroup of 251 patients from the Breast Cancer Prevention Trial (BCPT) who fulfilled inclusion criteria (having both endometrial biopsies and transvaginal sonograms). Patients were examined by community gynecologists for a mean follow-up of about 4 years. Among these patients, 145 had received tamoxifen and 106 placebo.

Endometrial biopsy identified 17 women with endometrial events. Only six of these women (35%) had abnormal endometrial echoes greater than 6 mm, and eight (47%) had normal endometrial echoes 5 mm or less. Three patients had indeterminate sonography findings.

Of these 17 patients, 16 were diagnosed with endometrial hyperplasia (10 receiving tamoxifen and 6 placebo). One patient in the placebo arm had invasive cancer with bleeding. There were no cancers in the tamoxifen group.

Dr. Runowicz noted that eight patients with hyperplasia or cancer were found to have normal endometrial echoes, for a sensitivity of 42.9% and a specificity of 40.8%. The positive predictive value for sonography was 4.9%, and the negative predictive value was 90.9%. She concluded that the findings do not support substituting ultrasound for the more invasive endometrial biopsy. However, she said, the study was limited by the biopsy being done prior to the ultrasound.

Richard R. Barakat, MD, assistant attending surgeon, Memorial Sloan-Kettering Gynecology Service, presented data from a 7-year prospective study of the effect of tamoxifen on the endometrium in women with breast cancer. The objective, he said, was not to calculate the incidence of endometrial cancers in this population, but “to determine, in a prospective manner, what was happening to the endometrium over time by performing serial endometrial biopsies.”

The study was spurred by earlier trials that had found about a six- to sevenfold increased risk of endometrial cancer in breast cancer patients on tamoxifen. In the NSABP B-14 trial, the rate of endometrial cancer per 1,000 women per year was 0.2 in the placebo arm and 1.6 in the tamoxifen arm. “In any study that looks at screening for endometrial abnormalities in this patient population, the number of adverse outcomes, specifically endometrial cancer, is actually very low,” he noted.

After all exclusions, 111 patients were included for evaluation, with a total of 635 endometrial biopsies, all performed in the office by Dr. Barakat. Every patient had a baseline biopsy and biopsies every 6 months for 2 to 3 years. Pap smears were also performed during the visits. The mean number of biopsies per patient was six, and the median surveillance time was 36 months.

Patients were removed from the study if they developed endometrial cancer or atypical endometrial hyperplasia, which Dr. Barakat noted is “the only hyperplastic condition of the endometrium that has a real progression rate to endometrial cancer (about 27%).”

Although simple hyperplasia has rarely if ever progressed to endometrial cancer, the researchers monitored patients somewhat more frequently if they were found to have simple hyperplasia.

Sensitivity Is High

The sensitivity of endometrial biopsy in picking up endometrial abnormalities is very high, Dr. Barakat observed. “It has been documented to be over 95%,” he said. In this study, D&C was performed when a patient continued to be symptomatic and to bleed despite a normal biopsy. The Memorial Sloan-Kettering team also performed several D&Cs when patients were found to have adnexal masses during the follow-up procedure.

Dr. Barakat said that D&Cs were ultimately performed, for abnormal endometrial biopsy or persistent bleeding, in about 10% of the patients. Of this group, 3 patients (2.7%) had significant pathology requiring hysterectomy. None of the patients developed endometrial cancer, although one patient did have complex hyperplasia. “The progression rate of complex hyperplasia to endometrial cancer is only about 3%,” Dr. Barakat said.

One patient ultimately had a fatal high-grade leiomyosarcoma of the uterus. “It is controversial as to whether tamoxifen is or is not associated with the development of uterine sarcomas,” Dr. Barakat said, adding that the current literature is conflicting, and the condition is very rare. “It is such a rare condition, there is probably no way that we will ever be able to answer this question,” he noted.

One of the problems with screening, Dr. Barakat noted, is that it leads to many unnecessary procedures. “The uterus in women with breast cancer who are on tamoxifen is, in a sense, a battered organ,” Dr. Barakat commented. The women are often put through myriad tests, including transvaginal sonography, regular endometrial biopsies, and sometimes D&Cs and office hysteroscopies. This is partially driven by patient anxiety and, to some degree, by physician anxiety over medical-legal outcomes, he said.

Dr. Barakat concluded that there is no proven benefit of performing these diagnostic procedures to decrease endometrial cancer mortality. “The rarity of endometrial cancer in this study shows office endometrial biopsy as a sequential screening procedure to be of limited value,” he said.