Oxford Myeloma Workshop 2025 | Modeling novel drug and management strategies in multiple myeloma

Yesterday I was able to give a talk on a kind of novel modeling of myeloma and our goal, you know, overall, it’s been a decade of work kind of put together with a lot of individuals, not just myself. But we went with the concept of how can we change the way we think about studying myeloma. So we want to make sure that we’re really no longer using… we want to use patient specimens, patient data to derive kind of how myeloma really works in individuals. 

And so, what we designed and what I walked through yesterday was kind of the journey from our ability to take patient specimens, so volunteers who then donate consented specimens for us that look at the myeloma cells themselves and how they respond to different drugs. And with this, we’re able to create avatars so we can then test multiple drugs on the patient’s specimen, the patient’s tumor cells, and see how they respond. And then again, using a number of kind of tools and mathematical models, we can then actually predict how these patients would respond in real time, clinically. 

And so the goal was to set up a situation where we could have as many patients as we could, so offer to everybody, and then have a quick turnaround. So within eight days, have a readout or a model saying, this is how this patient will respond to this drug or drug combination. And so that’s when we started literally over a decade ago, and it’s grown immensely. 

And our goal eventually was, or at this time, was to make this a tool we could give to everybody. But it’s a little bit more complicated than we wanted it to be. So instead of being the perfect tool, as we thought, we’ve kind of then changed it a little bit. And thankfully for the same samples, we’re able to then get really highly granular molecular data. And so we’ve had what’s called RNA sequencing, where we can get the expression of all the genes in the myeloma in these patients, as well as what we call whole exome sequencing, so the mutational patterns of these same patients. 

And then what’s really interesting, what we think is really interesting, was then we can then take these individual patient specimens, their molecular data, and we can see how every drug response or drug combination response correlates with this data. So now we can come up with these profiles of drug response. And the goal is then we can then apply that to patients. So that becomes the new biomarker to help us derive how the patients would do in therapy. And so that’s kind of the short and then the long-term goal of what yesterday’s conversation was. 

And then now with the data that’s going on around the world through lots of collaborations, as we’re getting more and more able to identify what the structures of the myeloma genome are. And then our hope is using new technologies, whether it be modeling technologies or truly artificial intelligence, can we start layering all these pieces on top to really come up with personalized predictors for patients. So it’s not just, you know, this drug or that drug, but what really can we do with all this information, right? The detailed granular stuff we get from the patient data. We want to tie in what’s called the immune microenvironment. We do the same thing with the cells that aren’t the tumor within the patient. And then again, we have lots of clinical data. We can pull that data in at the same time and use all of this information to really come up with a here is how this patient or this group of patients should behave, so kind of a much more personalized medicine than we have today. Again, it’s future thinking, but really we are to the point in myeloma where one size doesn’t fit all we know we need to start breaking people into groups and we think this is a really good way to do it and again it’s going to take lots of collaboration from a lot of individuals that was kind of the summary of what we talked about yesterday.

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