Background: Papillary thyroid microcarcinoma exhibits an indolent clinical course and could be a candidate for active surveillance in the appropriate setting. It remains unknown whether papillary microcarcinoma is biologically different from larger papillary carcinoma >1 cm.
Methods: We analyzed clinicopathological information and transcriptome data of papillary thyroid cancer samples from The Cancer Genome Atlas. Propensity-score matching was used to construct a matched cohort consisting of 29 microcarcinomas and 58 carcinomas. Principal component analysis and unsupervised hierarchical cluster analysis were carried out to investigate the similarity of gene expression profiles.
Results: After adjustment for differences in baseline clinicopathological and genetic factors, transcriptome could be grouped mainly on the basis of tumor class (BRAF-like vs RAS-like) and tumor size (microcarcinoma vs carcinoma). The gene set enrichment analysis showed that extracellular matrix-associated pathways were enriched in the MSigDB database.
Conclusion: Papillary thyroid microcarcinomas display a distinct gene expression pattern different from the corresponding carcinomas. We hypothesize that tumor microenvironment may play a role in the microcarcinoma/carcinoma phenotypic divergence.
Keywords: microcarcinoma; propensity score matching; thyroid carcinoma; transcriptome.
© 2019 Wiley Periodicals, Inc.