Disrupting protein-protein interactions with small molecules is considered to be very challenging due to the fact that the interfaces between two proteins are often very extensive and relatively flat. If there exists a high-resolution structure of the PD1/PD-L1 complex, look for a "hot spot," a small region that is responsible for a large portion of the binding affinity, preferably with a "pocket" into a which a small molecule could bind.

Disrupting protein-protein interactions with small molecules is considered to be very challenging due to the fact that the interfaces between two proteins are often very extensive and relatively flat. If there exists a high-resolution structure of the PD1/PD-L1 complex, look for a "hot spot," a small region that is responsible for a large portion of the binding affinity, preferably with a "pocket" into a which a small molecule could bind.

Seconding what Adam says. There's no particular indication that we CAN'T develop a small molecule inhibitor of the PD1/PDL1 interaction, but there's every indication that it will be very difficult. Small molecule agents that block protein/protein interactions outside of narrow catalytic sites are few and far between. Even where they do exist, they tend to be large and hydrophobic, generally not-drug-like according to e.g. the Lipinski rules.

At the end of the day, you also have to question the value of the molecule. Antibodies are incredibly powerful when you have the right target (eg: an extracellular protein such as PD1 or PDL1). They tend to have very high affinity and tremendous selectivity compared to the average small molecule. They're not perfect - it's hard to imagine an orally available antibody, they do tend to be more expensive to produce, and they do have off targets (the average commercial antibody for Western Blot, for instance, will generally pick up 2-3 unintended bands). But still, this is simultaneously a perfect target for antibodies, and a very hard target for small molecules. Even if you tried to make a small molecule, the discovery cost might well render the final drug just as expensive as the antibody, at least until it was off patent.

Thank you to Adam and William's professional and insightful comments. Perhaps small molecules might not be an optimal/ economic drug R&D direction for PD1/PD-L1, at least for the present stage.