Recall rate in a semi-annual breast surveillance program for high-risk women.

Abstract

11084 Background: High recall rate and unnecessary biopsies are a concern with screening mammogram (MMG) and magnetic resonance imaging (MRI) in young women at high risk for breast cancer. The aim of this study is to investigate the overall recall rate (ORR) and its consequences in a semi-annual MRI surveillance protocol. Methods: A multi-modality surveillance program for women at high-risk for breast cancer was initiated in 2004. Yearly MMG was supplemented with semi-annual MRI. After 5 years of this protocol, an additional 5 years of follow-up with yearly MMG and MRI were offered only to mutation carriers. Defining positive screening as BIRADS 0, 4 and 5, ORR were analyzed and biopsies findings characterized. Fisher’s exact test was performed to evaluate association between the effect of MMG density, menopausal status, mutation status, age (<50y or > 50y), and previous breast cancer with recalls. Results: With a mean follow-up of 3.6 years, 226 patients (pts) underwent 1,467 MRI and 851 MMG. Among 56 abnormal MRI and 18 abnormal MMG, 62 pts (27%) were recalled for further tests (6 recalled at two screening episodes). The MRI and MMG ORR were 3.8% and 2.1%, respectively. Only six cases (8.8%) were abnormal for both tests. Overall, 50 additional breast ultrasound, 10 additional MMG, 8 additional MRI were performed. Of the 40 biopsies in 38 pts (2 pts were biopsied twice), 11 cancers were detected (ductal in situ: 3; invasive: 7 ductal and 1 lobular). Abnormal MRI findings led to 37 biopsies, of which 10 (27%) detected cancers. One high-grade ductal carcinoma in situ was exclusively detected by MMG. The first and second MRI screening rounds together showed 27/56 (66%) abnormal findings which led to 23/40 (58%) of the biopsies, and detected 5/11 (45%) of the tumors. No statistically significant predictors of recall were identified (p < 0.05). Conclusions: Semi-Annual MRI protocol did not show increased ORR when compared to annual MRI/MMG surveillance. Approximately two-thirds of all recalls were in the first two rounds of screening which is consistent with previous results. Ongoing work will evaluate clinical and MRI features associated with recall to improve practice. Clinical trial information: NCT00989638.