Endocrine Abstracts

Introduction: Most thyroid cancer (TC) patients have a good prognosis, but some of them show aggressive behavior and are resistant to treatment. Still no effective biomarkers are available to predict which patients will progress. Epigenetic alterations are a hallmark of cancer, and profiling DNA methylation has allowed a better understanding of the epigenetic landscape of TC and its influence in TC progression. However, bulk DNA methylomes include information not only from tumoral cells but also from other cell populations that compose tumor microenvironment (TME) and influence tumor progression, aggressiveness and treatment response.

Objective: Our objective is to characterize the TME of TC to better understand its role and clinical application.

Methods: We applied a deconvolution approach to bulk DNA methylation array data using PRmeth, an R-based method that allows deconvolution of partially available DNA methylation data. We profiled the methylome of a large cohort of thyroid samples (n =114), including normal tissue (NT), low-risk papillary and follicular TC (PTC and FTC, respectively), metastatic PTC and FTC (both radioiodine avid (RAI-A) and refractory (RAI-R)), and paired metastases, using EPIC arrays, and analyzed data from external sources which in addition to NT, PTC and FTC, included PDTC and ATC (n =1000). We also used publicly available data of 13 cell types known to be part of the TME as well as some TC cell lines as references.

Results : In total 19 cell populations were identified, including 1 for fibroblasts, 1 for endothelial cells, 11 for immune cells, and 1 for TC highly undifferentiated cells. We also identified 5 unknown groups present in all samples that would correspond to normal follicular cells and different cancer cells. Results showed different cell-type proportions according to histology. Interestingly, we found almost no immune infiltration in FTC (low-risk and metastatic), while infiltration was higher in PTC and increased along PTC progression. PDTC and ATC showed the highest levels of infiltration with more than half of ATC cells being immune cells. The comparison between RAI-A and RAI-R PTC, or RAI-A and RAI-R FTC, did not show differences in the immune infiltration but in the proportion of the different cancer cell populations.

Conclusion: We conclude that the TME of a TC is key for tumor progression, being more important in some histological subtypes than in others. The high immune infiltration in highly aggressive tumors makes them good candidates for immunotherapy.