Virtual Tissue Microarrays for Validating Digital Biomarker Analysis in Colorectal Carcinoma

Area of Science:

• Oncology
• Pathology
• Biomarker Discovery

Background:

• Tissue microarrays (TMAs) are vital for high-throughput biomarker analysis.
• TMAs may not capture the full heterogeneity of whole tissue sections (WTS), potentially leading to inaccurate biomarker quantification.
• Understanding the representativeness of TMAs is crucial for reliable biomarker validation.

Purpose of the Study:

• To determine the minimum number of tissue cores required in virtual TMAs to achieve biomarker quantification precision comparable to WTS.
• To assess the impact of core number and tissue region (tumor center vs. invasive margin) on biomarker expression accuracy.
• To evaluate the utility of virtual TMAs for optimizing TMA construction for biomarker analysis in colorectal cancer.

Main Methods:

• Immunohistochemical staining for immune cells and fibroblasts in 50 colorectal cancers and 36 microsatellite unstable cases.
• Digitization of WTS and creation of virtual TMAs with varying core numbers.
• Analysis using Bland-Altman plots to compare biomarker quantification between virtual TMAs and WTS.

Main Results:

• Substantial disagreement between TMAs and WTS was observed, decreasing with more cores but remaining significant even with 8 cores.
• TMAs, especially with 3-4 cores, tended to underestimate biomarker expression in the tumor center and for specific cell types.
• Despite underestimation, 3 cores were sufficient for categorizing biomarkers into low and high expression groups.
• Microsatellite unstable tumors exhibited high heterogeneity, exacerbated by BRAF variants.

Conclusions:

• Virtual TMAs are effective for determining optimal core numbers for biomarker analysis.
• TMA representativeness varies by biomarker, tissue region, and tumor subtype.
• Thorough validation of TMAs against WTS for specific biomarkers is essential before large-scale clinical studies.