Reliability of prognostic and predictive factors evaluated by needle core biopsy for breast cancer in large tumors.

Abstract

1104 Background: Preoperative needle core biopsy (NCB) in breast cancer allows both invasive carcinoma diagnosis and the assessment of the main prognostic/predictive markers. However, large tumors, which may be candidates for preoperative chemotherapy, are potentially more heterogeneous than small ones, and the reliability of these factors measurements by NCB in these cases has been less documented than in whole surgical specimens (SS). The aim of this study was to evaluate the correlation for the main histological and immunohistochemical (IHC) features between NCB and SS in large tumors (>2cm), and to assess major discordances that may impact the therapeutic decisions. Methods: All patients treated in our center for newly diagnosed early breast cancer between January 2008 and December 2011 were retrospectively screened. Large tumors for which both NCB and SS were available were included. Patients treated by preoperative chemotherapy were excluded. The assessed histological and IHC parameters were histological type, SBR grade, estrogen (ER) and progesterone receptors (PR) expression, HER2 status (completed by CISH if needed), intrinsic subtype, and proliferation markers (mitotic activity index (MAI), Ki67). All samples were independently read twice. Comparisons were performed using Kappa test. Results: 163 pairs of NCB-SS were analyzed. Average pathological tumor size was 3.2cm. Correlation was excellent for ER and HER2 (k=0.98 and 0.91 respectively), good for histological type (k=0.74), PR (k=0.79) and intrinsic subtype (k=0.73), but poor for Ki67 (k=0,60), SBR grade (k=0.29) and MAI (k=0.24). Among major discordances, 6 tumors were graded SBR I on NCB but SBR III on SS; 1 tumor was negative for ER on NCB but positive on SS. Importantly, 3 of the 21 HER2 positive cases (14%) were negative after CISH on NCB but positive on SS. Conclusions: Diagnostic NCB in large early breast tumors allows reliable determination of hormonal receptors expression, histological type and intrinsic subtype. SBR grade may however be deeply underestimated by this method, and false negative evaluation of the HER2 status in 14% of HER2 positive patients would have led to a detrimental lack of trastuzumab administration.