Risk for subsequent breast cancer after lobular carcinoma in situ: Do clinical factors matter?

Abstract

1593 Background: Increased breast cancer risk conferred by lobular carcinoma in situ (LCIS) is poorly understood. Molecular studies suggest a subset of LCIS may behave as precursors to invasive disease, yet criteria for risk stratification in the clinical management of women with LCIS are lacking. We reviewed our experience with LCIS over 25 yrs to identify clinical factors associated with a subsequent cancer diagnosis. Methods: Patients participating in surveillance following a diagnosis of LCIS were identified from institutional databases and clinical practice records (1984-2009). Those taking tamoxifen were excluded. Comparisons were made between patients who developed cancer (cases) and those without evidence of cancer after ≥ 96 months of followup (controls). Cases were further analyzed by time to cancer diagnosis, < 12 months vs. ≥ 12 months from LCIS. Results: 675 pts with LCIS w/out concurrent cancer at presentation were identified; 31 (5%) had bilateral LCIS. Average age at LCIS dx was 52 yrs (range 24-83). At a median of 58 months from LCIS dx (0-344mos), 78 pts (12%) developed 89 breast cancers (11 bilateral). Of 597 pts remaining cancer free, 152 (25%) have been followed ≥ 96 mos (controls); median followup for this gp 131 mos (96-344 mos). There was no difference in age at LCIS, FH breast cancer, finding of concurrent atypia, or number/site of biopsies after LCIS dx between cancer cases and controls. Cases were more likely to have had ≥ 1 MRI after LCIS dx compared to controls (73% vs. 46%, p < 0.001). Among cases, median time to cancer after LCIS dx was 39 mos, without a dominant histology (DCIS 35%, IDC 31%, ILC 25%, IDC w/ lobular fts 4.5%, and rare subtypes 4.5%). Comparing those dx with cancer < 12 mos (n = 15) vs. ≥ 12 mos (n = 63) after LCIS, there was no difference in histology, tumor size, stage, clinical risk factors or use of ≥ 1 MRI. Conclusions: In this large cohort of patients undergoing surveillance for LCIS, we observed a 2% annual incidence of breast cancer, yet we were unable to identify additional clinical risk factors associated with the development of cancer. Greater use of MRI among those diagnosed with cancer suggests a role for enhanced imaging in this cohort, however this finding requires confirmation in prospective studies. No significant financial relationships to disclose.