PURPOSE

Patients with anaplastic thyroid cancer (ATC) have a very poor prognosis. Immunotherapy is a potential treatment, while the current outcome is limited which may be due to the complicated tumor microenvironment (TME). Tumor associated macrophages (TAMs) is the most abundant cell in the TME of ATC. We aimed to clarify the novel indicators based on TAM in ATC.

METHODS

Transcriptome files were downloaded from the Gene Expression Omnibus (GEO) dataset. Weighted gene co-expression network analysis, cox regression, support vector machine, and random forest were utilized to identify TAM-related prognostic genes. Consensus clustering and principal component analysis were performed for integrated analysis. Moreover, external validation (Fudan University Shanghai Cancer Center cohort) was conducted in 23 ATC samples via immunohistochemistry.

RESULTS

ATC patients with an abundance of TAMs had a poorer prognosis. Four TAM related genes (FZD6, RBBP8, PREX1, HSD3B7) were identified and a TAM-related prognostic index (TAMRPI) was constructed with high area under the curve (AUC). Next, high TAMRPI was related to the higher level of TAM infiltration and upregulation of several pathways, such as E2F targets, IL6-JAK-STAT3, and G2M checkpoint. Immune checkpoint TIM-3 and CSF1R were positively associated with TAMRPI, and dysfunction of T cells was increased in high TAMRPI subset. Moreover, in the external validation of protein level, strong expression of TAM related genes was related to poorer prognosis, which was further supported by time-dependent AUC analysis.

CONCLUSION

TAM is negatively correlated to the prognosis of ATC. FZD6, RBBP8, PREX1, and HSD3B7 are potential biomarkers of ATC.