Learn how to evaluate combinations of human-specific checkpoint inhibitors in vivo, with double knock-in humanized drug target mouse models.
Testing Combination Immunotherapy Regimens
The future of immunotherapy is in combination regimens, both with other immuno-oncology agents and targeted treatments. To assess these combination opportunities in vivo, suitable models are needed.
For surrogate murine agents, models such as syngeneics or murine tumor homografts are available. But when you move to human-specific therapeutics, these mouse models are ruled out due to species specificity issues.
New preclinical models are now available, expressly designed to evaluate in vivo combinations of human-specific agents targeting common checkpoint proteins. These models feature pairs of knocked-in humanized drug targets (e.g. humanized PD-1 and CTLA-4) within a functional murine immune system.
Humanized Target Models
Before we discuss the double knock-in models, it’s best to take a step back and introduce the single humanized target models they’re developed from.
A robust platform of mouse models exists, with each model featuring stable expression of a human immuno-oncology target (e.g. PD-1, CTLA-4, OX40, CD40, etc). The models have been genetically engineered, replacing sections of murine genes of interest with the human counterpart. This generates chimeric human/murine proteins that are then targeted with human specific therapeutic antibodies.
Models are highly characterized to confirm human checkpoint protein expression on the appropriate immune cell population and validated to ensure model response to the human antibody.
This results in a range of models with a competent immune system, which are used to assess human-specific checkpoint inhibitors, or agonistic antibodies targeting the immune system. The models are ideal for early stage studies to:
- Test that a therapeutic antibody is engaging the right target in vivo.
- Rule out eventual toxicities resulting from the treatment.
Humanized drug target models are also great for “standard” preclinical efficacy studies. In these studies, tumor burden is evaluated over time, in control and treatment groups, and the percent tumor growth inhibition (TGI) by the agent is calculated.
Combination regimens can be assessed, but this would be limited to one human-specific agent and one murine analog (e.g. human-specific anti-PD-1 and murine OX40 for a humanized PD-1 model).
Developing Double Knock-In Models
When you want to assess combinations of two human-specific agents, a double knock-in model is required. These models feature two humanized checkpoint proteins within the same immunocompetent model, with models available for PD-1 and CTLA-4, PD-1 and PD-L1, and PD-1 and OX40, and additional models under validation.
Double knock-in models are developed by breeding two single knock-in models – for the PD-1/CTLA-4 dual model, individual humanized PD-1 and humanized CTLA-4 are bred together. Similarly, to the single knock-in models, characterization is performed to ensure both proteins are expressed. Validation with single agents and combination regimens also takes place, with some interesting results already seen for a PD-1/CTLA-4 double knock-in model.
Memory Response in Dual PD-1/CTLA-4 Humanized Target Model
Combination anti-PD-1 and CTLA-4 treatment resulted in complete tumor regression following model implantation with human PD-L1 expressing MC38 tumor cells. “Cured” mice remained disease free for up to days post-grouping and were subsequently used for a rechallenge study. Reimplantation with the same MC38 tumor line led to a memory response and no new tumor growth, compared with tumor development in treatment naïve controls.
All of this results in models with a functional murine immune system, expressing two humanized drug targets, amenable to the testing of checkpoint inhibitor combination strategies in vivo.
Summary
As we look to expand response to immunotherapy through novel combination regimes, more preclinical models are needed for in vivo, human-specific drug studies. Double knock-in humanized drug target models allow the assessment of human-specific combinations of the most common checkpoint inhibitors. They also provide a cost-effective alternative to using more complex, fully humanized models.
