Estrogen Receptor (ER) Expression in Breast Carcinoma: Comparison of Antibodies (Ab) SP1 and 6F11.
JL Deschenes, S Mitchell, R Linford, S Fahim, T Hiruki. Cross Cancer Institute, University of Alberta, Edmonton, AB, Canada
Background: ER status assessment is critical for the optimal management of breast cancer (BrC) patients. We compared the new SP1 rabbit monoclonal Antibody (Ab) with Ab 6F11, which had been validated in our institution, through clinical trials, Q/A programs and comparison with biochemical assay values.
Design: Following the manufacturers' instructions, we stained with SP1 and 6F11 quadruplicate TMAs of 115 consecutive invasive BrC. Also SP1 was applied to 511 randomly selected whole sections from BrC samples (mostly invasive BrC in excision specimens) for which the orignal 6F11 IHC slides were reviewed and both were classified according to their ER status: negative (score 0, no or less than 1% cells with mild staining); positive, score 1 (less than 10% cells stained), and 2 (more than 10%), with a separate evaluation of staining intensity and background staining. Slides from 7 non-breast benign tissue samples were also stained in parallel with both Abs. All borderline (including all 0 and 1) results were reviewed and confirmed by two pathologists (JD, TH, or SF).
Results: All 115 TMA cases had the same ER expression pattern with both Abs (81 positive, 19 negative). 4/110 6F11-negative cores were positive with SP1, and 6F11 was positive in 1/107 SP1-negative cores. Of the 511 reviewed whole sections, concordance between SP1 and 6F11 was noted in 97.04 % of cases (kappa=0.90). Most (99.5 %, 2/414) of 6F11 positive cases were also positive with SP1. Also, 13/97 (13.4 %) cases scored negative by 6F11 were found to be positive by SP1 (Fisher's p=0.0035). 12 of these 13 cases had a score of 1. Significant differences in staining within the same score occured in 18.6 % of cases (n=95), and favored either SP1 (n=59) or 6F11 (n=36). Signal to noise ratio was described as better in SP1 (n=40), or 6F11-stained slides (n=11) or similar (rest), although SP1 generally tended to be crisper. In benign tissues, expected normal staining patterns were generally more obvious with SP1; however, enhanced staining of stromal cells and lymphocytes related to this Ab was seen as a real potential pitfall, also noted in BrC cases.
Conclusions: SP1 and 6F11 are comparable, but in BrC samples, SP1 provides slightly increased sensitivity and equal or better signal to noise ratio. SP1 will strongly stain stromal cells and lymphocytes in BrC samples, a pitfall to be aware of when assessing ER status in breast carcinoma samples. Most discordant SP1+/6F11 neg. cases have less than 10 % of cells stained with SP1.
Wednesday, March 11, 2009 9:30 AM
Poster Session V # 13, Wednesday Morning