Table 1.

Biomarkers of immune TME in HNSCC and their advantages and disadvantages.

Biomarkers	Method	Expression	Mechanism	Mode of action	Advantages	Disadvantages
PD-L1/PDCD1	CPS/TPS(IHC)	Increased	Suppressed T cell; Escaped immune killing	Exosome	Mature target therapy with clinical predictive biomarkers	The prediction efficiency is not enough, and supplemented biomarkers are needed
				Ligand-receptor
IFN-γ	Sequencing	Increased	Upstream of PD1	Ligand-receptor	Gene signatures have 95% of negative predictive value	No differentially expression found in baseline samples
					Potential biomarker to exclude immunotherapy
CXC family	Sequencing	Increased in ICI response	Introduced lymphocyte infiltration into the lesion and inhibited tumor growth	Ligand-receptor	Potential predictive biomarker by pan-cancer analysis	Need verification in untreated HNSCC
TILs	Microarray	Increased in patients with longer DFS, OS, and better locoregional control	Tumor cytolysis and maintained immune surveillance	Cell-to-Cell	Predictive model of TILs and PD-L1	Need verification in untreated HNSCC
	RNA sequencing
CAFs	Sequencing	Dependent on subgroup	HNCAF-1 Immunosuppression	Cell-to-Cell	HNCAF-0/3 predict PD-1 therapy response	Need verification in larger number of untreated HNSCC
			HNCAF-0/3 stimulated Trem and cytotoxic T cells	Ligand-receptor	Immunosuppressive HNCAF-1 specific in HNSCC
				Exosome