Spatial heterogeneity within cholangiocarcinoma tumors can complicate the evaluation of claudin 18 expression when clinical assessments rely on limited tissue samples, according to a study published in BMC Cancer. Researchers examined how staining variations across tumor sections impact biomarker classification, which is crucial because claudin 18-directed therapeutic strategies are currently entering clinical trials for cholangiocarcinoma.
Claudin 18.2 has emerged as a promising therapeutic target in gastrointestinal malignancies, prompting increased interest in its application to biliary tract cancers. Determining eligibility for targeted therapies often depends on identifying a specific proportion of tumor cells displaying moderate-to-strong membranous staining. However, because distribution of the biomarker can vary throughout a tumor, researchers aimed to characterize intratumoral heterogeneity of claudin 18 immunohistochemical expression to determine whether limited biopsy sampling could lead to misclassification.
To investigate these spatial variations, the study examined 85 whole-section tissue blocks obtained from 38 patients diagnosed with cholangiocarcinoma. This selected patient cohort was enriched for tumors exhibiting prior partial or strong claudin 18 expression, drawn from a previously published cholangiocarcinoma series. By analyzing multiple whole-tissue sections per patient, the investigators sought to capture variations that might be missed during routine single-sample evaluations.
The research team quantified claudin 18 expression by calculating the percentage of tumor cells demonstrating moderate-to-strong, defined as 2+ or 3+, membranous staining. Cases were categorized using a threshold of at least 75 percent positive tumor cells. In addition to measuring staining percentages, the researchers annotated spatial expression patterns across sections and compared whole-section patient classifications with prior tissue-microarray results.
Across the entire cohort, whole-section analysis revealed that 11 of the 38 patients were uniformly above the 75 percent positivity threshold across all evaluated blocks. Conversely, 23 patients were uniformly below the threshold, while four patients demonstrated heterogeneous expression across different tumor areas. Looking at all 85 individual tumor blocks analyzed, 29 blocks, or 34.1 percent, met the 75 percent cutoff.
Among the subset of 29 patients who had at least two assessable tumor blocks available for evaluation, interblock heterogeneity was identified in four cases, representing 13.8 percent of that group. In these four multiblock cases, claudin 18 staining crossed the 75 percent cutoff between different blocks from the same patient. Staining remained consistent across blocks in the majority of evaluated multiblock tumors.
The researchers also evaluated how well limited sampling techniques mirror broader tumor pathology by comparing whole-section classifications with previous tissue-microarray data. A single tissue-microarray core served as a model for limited biopsy sampling. Classifications between the microarray cores and whole-tissue sections agreed in 30 of 38 patients, corresponding to a concordance rate of 78.9 percent and a Cohen’s kappa coefficient of 0.57.
To evaluate the reliability of the scoring methods, independent reassessments were conducted on the tissue blocks. These evaluations demonstrated high reproducibility for block-level threshold classification, achieving an agreement rate of 92.9 percent with a kappa value of 0.85. Categorization of spatial expression patterns also demonstrated substantial agreement between observers, showing 81.2 percent concordance and a kappa value of 0.67.
The study concluded that while claudin 18 staining was consistent across sampled blocks in most patients within this enriched cohort, limited sampling models such as single tissue cores do not always reflect broader tumor status. The authors highlighted that sampling heterogeneity represents a significant practical consideration for claudin 18 biomarker assessment in cholangiocarcinoma, suggesting that evaluating spatial patterns alongside staining percentages could provide valuable diagnostic context.
Subject of Research: Intratumoral spatial heterogeneity of claudin 18 immunohistochemical expression in cholangiocarcinoma
Article Title: ​Spatial heterogeneity challenges limited sampling of claudin 18 expression in cholangiocarcinoma
Article References: Srivastava, A., Haug, L., Beldermann, O., Bein, J., Wiegering, A., Zeuzem, S., Wild, P. J., Walter, D., & Kinzler, M. N. (2026). ​Spatial heterogeneity challenges limited sampling of claudin 18 expression in cholangiocarcinoma. BMC Cancer. https://doi.org/10.1186/s12885-026-17319-x
Image Credits: AI Generated
DOI: 10.1186/s12885-026-17319-x
Keywords: cholangiocarcinoma, claudin 18, claudin 18.2, spatial heterogeneity, intratumoral heterogeneity, immunohistochemistry, biomarker assessment, tissue microarray, whole-section histology, gastrointestinal oncology, biliary tract cancer, membranous staining, interblock heterogeneity, biopsy sampling, pathology concordance, targeted cancer therapy, clinical trials, diagnostic pathology
News Source: Nathaniel Bowman. (November 16, 2026). Limited Tissue Sampling May Misclassify Claudin 18 Expression in Cholangiocarcinoma. Scienmag.



