I am always very apprehensive when it comes to targeting like this. As to your question specifically...I haven't looked deeply but from what I remember:
"Programmed death 1 (PD-1), an inhibitory receptor expressed on activated lymphocytes, regulates tolerance and autoimmunity..."
http://jem.rupress.org/content/203/4/883.full
I would think the higher risk would be in autoimmunity given that the counts would indicate an immune response to infection and administration could be delayed to avoid diminishing that response. Given the random nature of somatic recombination and PD-1 expression in the parenchyma, I'm not sure what indications could be used.
I'll include this just in case anyone interested hasn't seen it:
Safety and Activity of Anti–PD-L1 Antibody in Patients with Advanced Cancer
The data presented at two recent meetings I attended showed that this drug does have defiinite clinical anti-cancer activity in renal cancer, non-small cell lung cancer and melanoma which is more than anti-CTLA 4 antibody but the auto-immune side effects were definitely less serious than CTLA4