Equivalent Outcomes for Accelerated Partial Breast Irradiation for Invasive Lobular Carcinoma: A Matched Pair Analysis

Abstract

The 2009 ASTRO consensus guidelines considers invasive lobular histology (ILC) a cautionary feature in patient selection for accelerated partial breast irradiation (APBI). Patients with ILC have either been excluded or underrepresented in previous studies. The aim of this retrospective study is to review an institutional review board-approved single institution prospectively maintained database to compare outcomes in patients with ILC treated with APBI or Whole Breast Irradiation (WBI). The database was queried to examine all patients with ILC histology who underwent breast conserving therapy (BCT) at a single institution from 1981-2017. Patients were excluded if node positive, positive margins (considered to be tumor on ink for invasive or < 2mm for non-invasive tumors), received neoadjuvant chemotherapy or were undergoing retreatment. The patients were stratified by treatment type (APBI vs WBI). APBI techniques included Mammosite, Contoura, SAVI, insterstitial needle placement and 3D conformal radiation. WBI includes conventional and hypofractionated irradiation. Clinicopathologic characteristics were compared between groups. Patient age, laterality, surgical tumor size, grade, final margin status, presence or absence of extensive intraductal component (EIC), lymph-vascular invasion (LVSI), multifocality, ER positivity, and the use of adjuvant hormonal therapy or chemotherapy were reviewed. A propensity score match on a 1:1 basis for each of the above features was performed. Crude rates and Kaplan Meier estimates were obtained for LR, RR, DM, DFS, CSS or OS for each group. Median follow-up was 6.09 years for the APBI cohort (N=26) and 8.49 years for the WBI cohort (N=232). APBI patients tended to be older (69 +/− 12 vs 63 +/− 11 (years +/− sd)) and have smaller tumors (12.3 +/− 8.8 vs 16.7 +/− 11 (mm +/− sd), p = 0.05). The propensity score match successfully found a cohort of patients treated with WBI that was comparable to the APBI cohort in the factors listed in the methods (N=52). There were no significant differences in outcome in LR, RR, DM, DFS, CSS or OS. This match pair analysis further suggests that APBI is an effective alternative in well-selected patients with ILC histology. Patients with ILC should be considered for enrollment on randomized clinical trials to further clarify the role of APBI in patients with ILC.