Abstract
The reliability of the traditional HER2 immunohistochemistry (IHC) scoring system for classifying low-expression cases remains uncertain, creating a barrier to effectively deploying new antibody-drug conjugates. To assess inter-observer agreement, 15 pathologists independently scored HER2 IHC in 460 gastric cancer specimens. We found the highest overall percent agreement (OPA) for HER2 3+ scores (0.17). The OPA for HER2 IHC 2+ cases was 0.11. However, consensus could not be reached for specimens scored as 0 or 1+. These findings suggest that conventional HER2 IHC scoring is relatively reliable for identifying patients with high HER2 expression but reveals significant inconsistency in classifying HER2-low versus HER2-zero cases. This highlights a critical limitation of the current system for guiding modern therapies and underscores the necessity for revised, precise scoring methodologies.