Interobserver and interassay concordance in HER2 immunohistochemistry: implications on HER2-Low breast cancer

Abstract

Adopting a HER2-specific treatment approach for HER2-low breast cancer has been suggested after DESTINY-Breast04 (phase 3) trials. Hence, accurate pathologic evaluation gained higher importance, making interobserver agreement and interassay agreement questionable in this regard. To evaluate these, a cohort of 116 invasive breast cancer cases were stained with Dako A0485 and Ventana 4B5. Cases were evaluated by 5 observers independently and results were compared, together with in situ hybridization findings. Interobserver agreement (Kappa) for Ventana 4B5 and Dako A0485 were 0.62 and 0.63, respectively. Agreement was more prominent for IHC 3+ cases, followed by IHC 0 cases. Comparing the assays with one another, 74 cases had the same score in Ventana 4B5 and Dako A0485, while 37 and 5 cases had higher scores in Dako A0485 and Ventana 4B5, respectively. The results of the study demonstrate that pathologists are relatively unanimous on distinguishing 3+ cases (Kappa - 0.76: Ventana/0.78: Dako), but have less consistency in 1+ or 2+ cases (Kappa - 0.59: Ventana/0.60: Dako). It also illustrates that the case can obtain different HER2 IHC scores with different assays being used. Therefore, it is suggested for pathologists to recognize the pitfalls in HER2 IHC, be aware of the assay that is being used in their laboratory and hence evaluate the results in the accurate context.