Checkpoint inhibitors, anti-PD1/anti-PD-L1 activities are channeled via CD28 co-stimulation

Two new back-to-back studies in Science showed that T cell recovery activity attributed to checkpoint inhibitors, anti-PD1/anti-PD-L1, is mediated via CD28 co-stimulation. One of the study focused on analysis of biochemical events leading to PD1 signaling in an in vitro reconstitution model and another study provided corroborating data using mouse model. 

Summary results from both studies suggest that anti-PD1/PD-L1 activity is lost when T cells lack co-stimulatory molecule CD28. However, readouts here are more complicated since PD-L1 itself could bind to CD28 ligand B7 molecule. Also, I noticed that by default T cells with inducible CD28 deficiency (CD28^f/f Cre^ERT2+) show less accumulation and/or survival (low cell density) compared to WT counterparts.

Another confusion has to do with the fact that at least in one of the clinical trials anti-CD28 antibodies produced severe cytokine release syndrome and program was discontinued. It could be that anti-PD1/PDL1 therapy activates only certain type of T cells while anti-CD28 could have targeted much larger T cell population.

David Usharauli