RESUMO
OBJECTIVES: In a RCT, we have previously shown that the levonorgestrel intrauterine system (LNG-IUS, Mirena) produces a decidual response protecting the endometrium at one year follow-up. We here report on the long-term follow-up of this group of women, to test the hypothesis that a LNG-IUS could prevent the pro-proliferative uterine responses of tamoxifen for up to 4.5 years. METHODS: A randomised-controlled trial of postmenopausal women who had taken at least one year of adjuvant tamoxifen therapy. RESULTS: One hundred twenty-two women were recruited. Nine were found to be ineligible after randomisation. The average duration of follow-up was 26.25 months (IQR 14.5-36 months) in the surveillance group and 24.2 months (IQR 13.75-32.5 months) in the LNG-IUS group. Women with LNG-IUS in situ at the time of final assessment had decidualised endometrium, and no polyps. In the surveillance group new polyps arose in 8 cases. There were 3 new polyps in the group initially randomised to LNG-IUS, one in a patient who did not have the device inserted and 2 occurred in patients following the removal of the LNG-IUS. Univariate Cox proportional hazards regression models identified only endometrial thickness at trial entry as a statistically significant variable (HR 1.12, 95% CI 1.02 to 1.22, p=0.01) for the development of polyps. CONCLUSION: This study confirms that LNG-IUS induces benign endometrial changes and prevents endometrial polyps but only during its use in women taking tamoxifen. Endometrial thickness is a risk factor for the development of polyps.