Individualized Prediction Model in Myeloma & Genome Sequencing | Francesco Maura, MD | ASH 2022

So my name is Francesco Maura. I'm an assistant professor at the University of Miami Silvester Cancer Center. And during this ASHE, we presented several oral communications. And the one that I personally presented were three. So in one, we developed the first individualized prediction model for outcome in multipromethoma. All the clinical, all the prognosis code in multipromethoma developed so far, like ISS, RISS, or R2ISS, are basically used as a relative score. So they don't assign to patient a risk. They just assign patients to a group with a certain risk, which is different. We need to develop tools that allow us to identify the risk of your, of this single patient, the next patient you see in your ambulatory. So the model integrates genomics, clinical data, and treatment. So you can think about this model something like you have a new patient coming in your ambulatory. You screen all this information. You put all this information in the model. And the model will tell you which is the best treatment for your patient in terms of transplant or transplant, maintenance, no maintenance, or which is the best treatment combination the patient should get. So the model counts 2,000 patients and the performance are massively higher than all the existing prognosis score. So this is a new way to develop a prognostication in multipromethoma and develop individualized treatment. Not all patients should receive the same therapy and not all therapy should be used in the same way. So we need to understand the heterogeneity, decipher it, and use it to treat our patient in the best way and also the most tolerated and safest way. And that's, I think, is one of the first steps that is going to change the current landscape where patients are treated all in the same way. The next order we present is a whole genome sequencing investigations of patients treated with Dara2-up carfizome and dexamethasone or with carfizome and dexamethasone. So the patients were enrolled in two clinical trials. One is the Manhattan trial published by Ola Langer in 2021 on JAMA oncology and the other one was published by Nea Corde and Ola Langer on American Journal of Hematology in 2021. So the analysis to perform whole genome sequencing allow us to detect all the different genomic features at structural variants, mutations, copy number that you can have in a tumor. Using this comprehensive analysis, we identify several genomic features that can be identified at baseline at diagnosis that are highly predictive of poor outcome or failure to achieve a sustained MRD negativity in unit diagnosis multiple malabation. We validate our observation on independent series and what is emerged is that the complexity of resistant multiple meloma is very dense. We need the comprehensive technology like whole genome sequencing integrated with the large data set. This is the first step, but definitely whole genome sequencing show that there is a complex network of genomic features that cooperate probably together, inducing creating cells that are more resistant to immunotherapy, more resistant to Dara2-up or carfizome or revlimid. We found new lesions also like Icarus 3D lesion where patients tend to relapse pretty quickly, including in the compass data because probably this lesion associates with the linalidomide resistance, a drug that is used today in most of the patients in the frontline. So this analysis, as I said before, is a first step. We need to expand our court. We need to do more genome sequencing, integrate the data and use that to understand the patients that despite getting a quadruplex therapy with Dara2-up, carfizome, or revlimid exometason, still relapse early or fail to achieve sustained MRD negativity. The third study is a genomic study that little bit complicated, but can be described in a simple term in this way. Myeloma is known to have initiating events like translocation on the IGH and hyperdeploy. These events are called initiating because those are the first events that we think occur in the tumor cell. So the tumor cells before the MGUS, that was not a tumor, you have a B cell, that is transformed and from that point it will acquire multiple attrition over time till become multiple myeloma. So we consider IGH translocations and hyperdeployed as the first event. In this study, we show for the first time that in a fraction of patients that is not big but provide a proof of concept, we were able to identify alterations acquired before these initiating events. These alterations are not necessarily myeloma specific, they look more like B cell tumor specific, and though this suggests the potential existence of entities with alterations that can go in different tumor types according to the second needs. It is suggested multiple myeloma may be initiated, yes, by one of those events, but those events like IGH and hyperdeployed can occur on an already established clone with alterations already acquired. And this of course can affect the heterogeneity, the temporal evolution, and eventually also the response to our treatment. So that of course opens a lot of interest in early development of myeloma and in understanding what happened and why a B cell tumor with these features turned to multiple myeloma and not for example on other lymphoma or maybe stay quiet for decades.