Why PDX Models Improve Radiopharmaceutical Biodistribution Studies

Denis Beckford Vera, Head of Radiopharmacology, Champion Oncology, makes a case for incorporating patient-derived xenografts (PDXs) into the early stages of radiopharmaceutical development. He argues that doing so leads to more accurate, clinically relevant results and fewer costly setbacks down the line.

At the core of Vera's argument is the value of biodistribution studies conducted in PDX models rather than traditional immortalized cell lines. PDX tumors more faithfully replicate the architecture, vascularity, and receptor density of real patient tumors. They give researchers a far truer picture of how a radiopharmaceutical will actually behave in the body. Vera explains, "If you start early in the development using these patient-derived xenografts to evaluate the biodistribution of your radiopharmaceutical, even to choose lead compounds, you're going to have the real assessment of how your radiopharmaceutical is behaving."

One of the distinctive advantages of radiopharmaceuticals, is their inherent visibility. Researchers can directly observe where a compound travels and how it accumulates, making biodistribution an especially powerful tool for optimization. This real-time insight can streamline the path from preclinical research to first-in-human trials.

A major concern he raises is the tendency to overestimate first-in-human doses when relying on conventional cell line models, which often exhibit artificially high and homogeneous receptor expression. PDX models, by contrast, capture the heterogeneity seen across and within patients. Vera explains, "If you, from the beginning, use clinically relevant, preclinical models like patient-derived xenografts, you will have a real understanding of how much this drug is going to the tumor, and then that will directly impact the first-in-human dose estimation."

Ultimately, Vera's message is one of efficiency and precision, building on clinically relevant models from the outset reduces the likelihood of having to return for costly re-optimization later in development.

Transcript:

Editor's note: This transcript is a lightly edited rendering of the original audio/video content. It may contain errors, informal language, or omissions as spoken in the original recording.

I'm Denis Beckford Vera. I am the head of radiopharmacology at Champion Oncology. This is very related to clinical translation. So if you start early in the development using these patient-derived xenografts to evaluate the biodistribution of your radiopharmaceutical, even to choose lead compounds, you're going to have the real assessment of how your radiopharmaceutical is behaving. Moving from biodistribution, I think biodistribution should be the first type of assessment that you do preclinically in these, clinically relevant models, to understand where target engagement, for instance, in the tumor. How this tumor heterogeneity, and how this, tumor architecture, that mimics the patient's tumor, can really affect the biodistribution of the radiopharmaceutical, which is very tied to vascularity, for example. It is different in immortalized cell lines than in real patient samples. The receptor density and how the receptors distribute between the tumor and between, different patients and the lesions between the same patient as well. So you're going to get an idea of how this can really distribute. More importantly, you can really save time because with radiopharmaceuticals, you have something that you don't have with other modalities. You can see, you can see your real molecule. So if you are able to see where it goes and how it behave where it goes. So then you can optimize the process for later efficacy studies in this same tumor model. Even further, if there is a biologically relevant species, you can also identify off-target toxicity as well doing biodistribution studies, which is very important, very relevant to, to these types of treatments. From the beginning, you are evaluating your compound in clinically relevant models. You later don't have to come back and do optimization, what happens many times. So this is what happens, when you use biodistribution to assess a first-in-human studies and to estimate first-in-human doses. So many times, this dose is overestimated, based on these regular cell line-derived xenografts. We have high receptor expression, in many cases, has more homogeneous receptor expression. And then you're saying, you are getting a wrong understanding on how much your drug is really accumulating in those tumors. So if you, from the beginning, use clinically relevant, preclinical models like patient-derived xenografts, you will have a real understanding of how much this drug is going to the tumor, and then that will directly impact the first-in-human dose estimation. So you will not have to go back and do, or you will have the probability that you will have to go back and do so much optimization to improve in human doses and things like that, will be less likely, than if you went right.