ISAL 2025 | Serially transplantable, genetically modified PDX models of AML

Yes, so indeed what we did over the last few years is that we transplanted primary AML patient cells into immunocompromised NSG mice, as a lot of groups do, and use these PDX models. But what we aimed to do is to generate models that we can serially passage in those mice to then have very reliable models that we can functionally and also genetically characterize. So the aim is to have then at hand PDX models that we can really select for specific projects, like you would select a cell line for a specific project with defined mutations, defined chromosomal alterations. And what we see is that if it comes to the serial repassaging, if it comes to reliable and also fast engraftment, nobody wants to work with cell line or PDX models that need like years in mice, we see that only very aggressive primary patient cells show the serial and reliable engraftment. So in the end we end up with around about 20 percent of the samples that show this reliable and fast engraftment that can be genetically modified to express a luciferase or also like for example shRNA and we have a clear bias towards very aggressive relapsed leukemias but it’s not that it’s that we find highly defined mutations so we see that that we have a bias towards KRAS and WT1 or ETV6 mutations. But importantly, what we see is that these PDX cells really resemble primary patient cells much better than cell lines. So the most commonly seen mutations in PDX cells, we can also model with, that we see in patient cells, we can model with the PDX cells. And some cannot be modeled with cell line cells, for example, IDH1 or 2 mutations cannot be modeled with cell lines but with PDX. And for cell lines, for example, there’s a huge bias towards TP53 mutations, which we don’t see with the PDX cells. And also if it comes to chromosomal alterations, cell lines have a clear bias towards very highly complex cytogenetics, while with PDX cells we can also model cytogenetically normal AMLs or AMLs with single isolated chromosomal alterations. And what we then do is to use these models that represent primary patient cells a lot to use them for molecular trials like CRISPR screens, but also for a lot of different therapy trials. And by expressing luciferase, we can monitor the disease in real time and see both the outgrowth of leukemia but also the response of leukemia. And what we have seen with this in vivo imaging is that often we see a reduced outgrowth of the cells after therapy, for example, after cytarabine or azacitidine treatment. And this is something that with standard models you would not see if you would not express a luciferase. So in the end we can say that these models can be used for diverse preclinical trials and that we can really tailor these PDX models always to the projects that they are needed for. And these models are also shared with a lot of different collaboration partners who then also use these cells for their studies.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.