Abstract
We aimed to compare the assay performance characteristics of HER2 status in mucinous epithelial ovarian cancer (EOC) by ToGA (Trastuzumab for Gastric Cancer) biopsy versus ToGA surgical specimen methods. Forty-nine tissue microarray (TMA) samples of mucinous EOC from Asian women were analyzed by immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) tests using ToGA trial HER2 scoring methods. The overall concordance between IHC and FISH by the ToGA surgical specimen method is 97.56% and by the ToGA biopsy specimen method is 97.14%. The agreements of HER2 IHC results under both biopsy and surgical specimen methods were nearly perfect (weighted kappa = 0.845). Additionally, the percentage of Her2 FISH amplification showed increasing trend with increasing HER2 IHC ordinals (negative, equivocal, positive) by both TOGA biopsy (P<0.001) and surgical specimen method (P<0.001). After excluding equivocal cases, the sensitivity (100%), PPV (88.89%) and NPV (100%) of HER2 IHC were unchanged under either surgical specimen method or biopsy method. However, the specificity (96.97%) and accuracy (97.56%) of HER2 IHC was slightly higher under the surgical specimen method than those (specificity 96.30%, accuracy 97.14%) under the biopsy method. Of the total 49 cases, the number (n = 14) of HER2 IHC equivocal results under the ToGA biopsy method was 1.75-fold higher than those (n = 8) under the ToGA surgical specimen method (28.57% vs. 16.32%). Therefore, compared to ToGA surgery specimen method, the ToGA biopsy method caused more equivocal IHC cases to be referred to FISH testing and did not increase the detection rates of Her2 FISH amplification.