Objectives: The presence of target receptors in non-target tissues can reduce the therapeutic window for a promising therapy due to toxicity in non-target tissues. Bispecific drugs engineered to bind to highly expressed protein in the desired tissue could broaden the therapeutic window. However, the relative affinities to the targeting protein and disease-related receptor must be tuned to balance the effect of tissue targeting and competitive binding. We explore the effects of this balance with mechanistic pharmacokinetics/receptor occupancy (PK/RO) modeling.
Methods: A mechanistic PK/RO model of a bispecific tissue targeting drug was developed. The model included 4 compartments: a central compartment, a peripheral compartment (to capture antibody PK), an efficacy compartment with both receptor and targeting protein expressed, and a toxicity compartment with only receptor expressed.
Two assumptions, that the drug can or cannot bind to targeting protein and receptor simultaneously, were tested. The effect of varying the affinities of the drug to targeting protein and receptor on receptor occupancy (RO) in toxicity and efficacy compartments for single and repeated doses were predicted.
Results: In the absence of simultaneous binding to targeting protein and receptor, the model predicts that tight binding to targeting proteins results in high local drug concentration but low RO in the target tissue for single doses and requires a large number of repeated doses to achieve RO saturation and steady state. In the presence of this binding, tighter binding to targeting proteins decreases the dose to achieve high RO and extends the amount of time that receptors are saturated for single doses while having little effect on the projected efficacy of repeated doses in the Kd range tested.
Conclusion: Computational models of tissue targeting bispecifics can facilitate drug design and minimize risk in patients by providing useful information to optimize therapeutic window.