Does levonorgestrel-releasing intrauterine system increase breast cancer risk in peri-menopausal women? An HMO perspective

Nava Siegelmann-Danieli, Itzhak Katzir, Janet Vesterman Landes, Yaakov Segal, Rachel Bachar, Hadas Rotem Rabinovich, Martin Bialik, Joseph Azuri, Avi Porath, Yossef Lomnicky

To evaluate the association between levonorgestrel-releasing intrauterine system (LNG-IUS) use and breast cancer (BC) risk. A cohort of all Maccabi Healthcare Services (MHS) female members aged 40-50 years between 1/2003 and 12/2013 was used to identify LNG-IUS users as "cases," and 2 age-matched non-users as "controls." Exclusion criteria included: prior BC diagnosis, prior (5 years pre-study) and subsequent treatment with other female hormones or prophylactic tamoxifen. Invasive tumors were characterized by treatments received (chemotherapy, hormonal therapy, trastuzumab, or combination thereof). The analysis included 13,354 LNG-IUS users and 27,324 controls (mean age: 44.1 ± 2.6 vs. 44.9 ± 2.8 years; p < 0.0001). No significant differences in 5-year Kaplan-Meier (KM) estimates for overall BC risk or ductal carcinoma in situ occurrence were observed between groups. There was a trend towards higher risk for invasive BC in LNG-IUS users (5-year KM-estimate: 1.06% vs. 0.93%; p = 0.051). This difference stemmed primarily from the younger women (40-45 years; 0.88% vs. 0.69%, p = 0.014), whereas in older women (46-50 years), it was non-significant (1.44% vs. 1.21%; p = 0.26). Characterization of invasive BC by treatment demonstrated that LNG-IUS users had similar proportions of tumors treated with hormonal therapy, less tumors treated with trastuzumab, (7.5% vs. 14.5%) and more tumors treated with chemotherapy alone (25.8% vs. 14.9%; p = 0.041). In peri-menopausal women, LNG-IUS was not associated with an increased total risk of BC, although in the subgroup of women in their early 40's, it was associated with a slightly increased risk for invasive tumors.