[1743] Automated Cellular Imaging System for Assessing HER-2 Status in Breast Cancer Specimens
DM Minot, BR Kipp, RM Root, A Nassar, AC Clayton. Mayo Clinic, Rochester, MN
Background: Previous studies suggest that automated immunohistochemistry (IHC) may improve the accuracy and reproducibility of HER2 IHC testing. The goal of this study was to assess the performance characteristics of the Automated Cellular Imaging System III (ACIS) for HER2 analysis using three scoring methodologies and comparing these results to both manual pathology interpretation and FISH analysis for HER2 amplification.
Design: This retrospective study was performed on 159 biopsy slides from patients who underwent routine HER2 testing between January and February, 2008. The three ACIS scoring methods included the Equal Distribution Score (EQD), Hot Spots Only (HSO), and Ten Region Score (TRS). The HSO method consisted of selecting six areas that appeared (by visual examination of the digitalized image) to have the most intense staining. The EQD method consisted of collecting six areas including 2 areas of high, 2 areas of moderate, and 2 areas of low intensity staining. The TRS method consisted of collecting 10 areas that the operator thought would give the best overall average intensity of the entire specimen. FISH for HER2 amplification was performed using the PathVysion
HER2 DNA probe kit.
Results: The number of IHC 0 or 1+/FISH positive cases was equivalent for all methods studied, yet the EQD and TRS method had significantly (P<0.001) fewer 2+ cases (n=16) and (n=18) respectively, versus the manual method (n=35). Both methods, EQD and TRS, had higher positive predictive values (PPV) (38%) and (33%) respectively, versus the manual method (20%). The percentage of patients with a (+) FISH result based on IHC scores is summarized in Table 1.
| 0 | 1+ | 2+ | 3+ | |
| Pathologist Manual Interpretation | 0/31 (0) | 2/84 (2) | 7/35 (20) | 8/9 (89) |
| ACIS HSO Method | 0/41 (0) | 2/69 (3) | 4/34 (12) | 11/15 (73) |
| ACIS EQD Method | 0/54 (0) | 2/80 (3) | 6/16 (38) | 9/9 (100) |
| ACIS TRS Method | 0/78 (0) | 2/54 (4) | 8/18 (33) | 9/9 (100) |
Conclusions: Both EQD and TRS methods had equivalent PPV's for the detection of FISH (+) HER2 cases and had fewer equivocal (2+) specimens, that require FISH analysis. The EQD and TRS methods may more accurately identify FISH amplified HER2 cases with fewer 2+ cases that would be reflexed to FISH analysis, when compared to the manual method. Future prospective studies are needed to determine if the ACIS can decrease the overall costs for patients undergoing HER2 testing while maintaining accurate test results.
Category: Techniques
Wednesday, March 11, 2009 1:00 PM
Poster Session VI # 247, Wednesday Afternoon