Analyses of the peripheral immunome following multiple administrations of avelumab, a human IgG1 anti-PD-L1 monoclonal antibody

Table 4 Effect of avelumab on 89 refined immune cell subsets

Post	Subset	Increase	Minimal change	Decrease	Unadjusted P-value	Direction	Holm adjusted P-value
1 cycle	PD-1⁺ ICOS⁺ CD4	12 (63%)	5 (26%)	2 (11%)	0.0181	↑	0.5611
	PD-1⁺ Tregs	12 (63%)	5 (26%)	2 (11%)	0.0141	↑	0.0705
	Functional intermediate NK	10 (53%)	6 (31%)	3 (16%)	0.0401	↑	0.5213
3 cycles	Functional intermediate NK	1 (7%)	4 (29%)	9 (64%)	0.0295	↓	0.4130
9 cycles	CTLA-4⁺ EM CD8	9 (56%)	6 (38%)	1 (6%)	0.0250	↑	0.6250
	Functional intermediate NK	2 (12%)	4 (25%)	10 (63%)	0.0386	↓	0.5018
	PD-1⁺ pDC	10 (63%)	2 (12%)	4 (25%)	0.0335	↑	0.1005
	CD16⁺ MDSC	9 (56%)	5 (31%)	2 (13%)	0.0092	↑	0.1288
	gMDSC	11 (69%)	2 (12%)	3 (19%)	0.0131	↑	0.1703
	CD16⁺ gMDSC	10 (62%)	3 (19%)	3 (19%)	0.0155	↑	0.1705
	PD-1⁺ lin neg MDSC	10 (62%)	4 (25%)	2 (13%)	0.0182	↑	0.1820

The frequency of 89 refined immune cell subsets was examined pre-therapy and post-1 cycle (n=19), 3 cycles (n=14), and 9 cycles (n=16) of avelumab. Table displays subsets that met criteria as a potentially biologically relevant trend. Results are displayed as the number of patients (percentage of total patients) with an increase of more than 25%, minimal change of less than 25%, and a decrease of more than 25% compared to pre-therapy. Unadjusted p-values (direction of change compared to pre-therapy) were calculated using the Wilcoxon matched-pairs signed rank test, and Holm adjustment was made for the number of subsets within the classic subsets with a frequency above 0.01% of PBMC
EM effector memory, gMDSC granulocytic MDSC, ICOS inducible T cell co-stimulator, lin neg MDSC lineage negative MDSC, MDSC myeloid derived suppressor cell, NK natural killer, pDC plasmacytoid DC, PD-1 programmed cell death protein 1, Tregs regulatory T cells