Abstract:

Objective To analyze the concordance of Claudin18.2 expression with different antibody clone numbers in gastric adenocarcinoma tissues. Methods A total of 56 newly diagnosed cases of gastric adenocarcinoma （including oesophagogastric junction adenocarcinoma） who were tested using Claudin18.2 clone number EPR19202 at Cancer Hospital of Chinese Academy of Medical Sciences， Peking Union Medical College， Shenzhen Branch from December 2023 to September 2024 were selected as the research subjects. All patients were then tested with clone number 43-14A， and the positive detection rates of the two different clone number antibodies were compared. The intraclass correlation coefficient （ICC） and Cohen's Kappa coefficient were adopted to evaluate the consistency of the interpretation results of positive cell proportions with different intensities after staining with the two antibodies. Additionally， the Spearman correlation coefficient was employed to assess the correlation of expression levels detected by the two antibodies. Results The positive staining of Claudin18.2 detected by both antibodies EPR19202 and 43-14A was exhibited membranous staining in normal gastric mucosal epithelial cells and tumor cells. EPR19202 demonstrated brown to yellow-brown staining， while 43-14A showed brown to dark brown staining. Regarding the positive detection rate， there were 13 positive cases and 43 negative cases for EPR19202， resulting in a positive rate of 23.21%. For 43-14A， there were 24 positive cases and 32 negative cases， yielding a positive rate of 42.86%. The positive rate of 43-14A was significantly higher than that of EPR19202， indicating a statistically significant difference between the two positive rates （χ²=4.04， P=0.045）. The ICC assessment results showed that in the EPR19202 stained sections， the ICC values for the three physicians' interpretation of the percentages of moderately and strongly positive cells in this clone antibody stained sections were 0.87 and 0.78 respectively； in the 43-14A stained sections， the corresponding ICC values were 0.80 and 0.91， respectively. The complete consistency rate of the three physicians' interpretation of the positive/negative judgment of the two antibody stained sections was 84% to 93%， with a Cohen's Kappa value ranging from 0.56 to 0.86. According to the Landis & Koch standard， the consistency of the three physicians' interpretation reached a moderate to almost complete consistent level， and the consistency of 43-14A interpretation was better. The Spearman correlation analysis results showed that the two clone numbers were moderately to highly positively correlated in the proportion of moderately positive cells （r=0.49， P<0.001； r=0.60， P<0.001； r=0.65， P<0.001）， and moderately positively correlated in the proportion of strongly positive cells （r=0.41， P=0.002； r=0.51， P<0.001； r=0.49， P=0.001）， suggesting that there was a moderate correlation in their expression levels； the Cohen's Kappa value of the three physicians' joint interpretation was 0.57. Conclusions Both Claudin18.2 antibodies， EPR19202 and 43-14A， show accurate positive staining localization， making them suitable for detecting Claudin18.2 in gastric adenocarcinoma and related precancerous lesions. The two clones show moderate to high correlation in detection results， but certain discrepancies are observed in weakly positive and highly heterogeneous cases. Clone 43-14A demonstrates a higher positivity rate in gastric adenocarcinoma tissues and better consistency in physicians' interpretation.

Key words: Stomach neoplasms, Tight junction proteins, Antibodies, monoclonal, Immunohistoche-mistry, Consistency analysis