How would you interpret a strong binding signal with unusually fast dissociation kinetics?

Why

In a biosensor experiment the response is the amount bound, and the kinetics come from fitting a model to the curve. Several artifacts can produce the described pattern. Too much ligand on the surface, or too much analyte, gives a large signal but distorts kinetics, because binding becomes limited by how fast analyte reaches the surface rather than by chemistry.

Myszka and colleagues showed that transport influences surface-based binding assays and that models must account for it. Nonspecific binding to the chip or the capture reagent also adds a rapid rise and a rapid fall that is not the interaction at all, which is why a reference surface and double referencing are standard. A bivalent antibody binding a surface-immobilized antigen can rebind with its second arm, which usually slows apparent dissociation, so a surprisingly fast off-rate may indicate the opposite problem, that the fit is being driven by a different process.

A 1:1 model fitted to data generated by heterogeneous ligand or bivalent analyte gives numbers that look precise and mean little. Good practice is to vary surface density and flow rate and see whether the fitted constants move.