Expert Praises Predictive Role of Radiomics in Sarcoma

Seth Pollack, MD

Several challenges remain in the treatment landscape of soft tissue sarcoma, a disease whose diagnosis can range from one that is quite indolent for patients, to a more life-threatening, metastatic, aggressive cancer.

One such challenge is accurately predicting individual patient outcomes. Radiomics, an emerging research field, may serve to alleviate that struggle. Radiomics quantifies complex aspects of tumor images related to tumor biology. Research has validated the use of this technique to predict patient outcomes in lung and head and neck cancers, and it is now quickly making headway in the treatment of patients with sarcoma.

OncLive: What is the role of radiomics in sarcoma?

So what exactly was the research on radiomics that was presented here at CTOS?

In an interview with OncLive at the Connective Tissue Oncology Society (CTOS) conference, Seth Pollack, MD, assistant member, Fred Hutchinson Cancer Research Center, assistant professor, University of Washington Medical Center, medical oncologist, Seattle Cancer Care Alliance, discusses the evolving role that radiomics will play in the prediction of patient outcomes in sarcoma. He also sheds some light on the exciting future of immunotherapy in the treatment of this disease.Pollack: Radiomics is really a brand new field. It’s something that we’re only just starting to scratch the surface of. And it’s not something that has a currently established role in the management of patients. But it’s something that’s really exciting. Radiomics is essentially a way to look at images that actually takes the viewer out of it, and lets the computer look for elements to the radiologic files that a person wouldn’t be able to see.Sarcoma may be an ideal place to study radiomics because so many of our patients have MRIs, which are one of the best ways to do radiomics research. So in this study, patients with primary soft tissue sarcomas—who were treated at the University of Washington and Seattle Cancer Care Alliance—had their images analyzed through this list of radiomic features, and a model was made to predict patients’ outcomes.

What we found was that that model was really, really accurate. It actually did better than all of the normal ways that you predict outcomes of patients, based on tumor grade, tumor size, and other clinical features. That’s why we’re really excited about it. But I do think that the data need to be validated on other data sets and other patient cohorts.

What impact do you expect that this will have on clinical practice?

What would you like the community oncologist to ultimately take away from these findings?

What do you hope to see in the treatment landscape of this disease in the next 5 to 10 years?

We’re already doing another internal validation with another cohort at our institution, but I think we’ll also ultimately need to validate this at an external institution in order to really demonstrate that this is a good way to predict patient outcomes.I think having a better way to predict patient outcomes is really promising. Once we have that tool validated in a way that is really reliable, then I think we’ll be able to do the really exciting work, of seeing whether these features predict outcomes to treatments. And once we start doing that, then maybe we can use these things to choose which treatment a patient is best suited for.I think for the community oncologist, radiomics is a new field, and it’s not something that’s going to be affecting their practice today, but maybe it will tomorrow. So I think for the community oncologist, this is something to put on their radar, and kind of start watching because it’s going to be really exciting.From my perspective, in terms of the radiomics work, I should mention that Matt Spraker is a radiation oncology resident who really did all of the work, and we have a great biophysics team led by Matt Nyflot at the University of Washington, and it’s been really exciting for me as a medical oncologist specializing in sarcoma to be involved in this project.

But my main focus has been immunotherapy, and that’s where I’m really excited for the field to move toward in the next 5 to 10 years. I thought the exciting data from SARC028 that we saw at this meeting was really great, where we saw these patients with pleomorphic undifferentiated sarcoma that achieved responses with pembrolizumab (Keytruda). I think that’s really exciting. At our center, we have a trial looking at doxorubicin plus pembrolizumab; I think that’s going to be exciting.

Could you talk a little bit more about the trial looking at doxorubicin plus pembrolizumab?

I think these NY-ESO-1 targeted therapies for patients with synovial sarcoma look very promising. The trial from immune design with an NY-ESO-1 vaccine, and the PD-L1 inhibitor atezolizumab, that’s a really exciting trial. There’s so much happening right now, and I think we’ll learn a lot about all of this in the next 5 years.It just opened, so we don’t have any data yet, but I think it’s exciting for a few reasons. First of all, doxorubicin is basically standard frontline therapy. I think this is an opportunity to move immunotherapy into the frontline. I think this will be one of the first chances to really see that in sarcoma.

Secondly, doxorubicin does a lot of things to manipulate the tumor immune microenvironment that we think will synergize with pembrolizumab. It kills inhibitory immune cells, myeloid-derived suppressor cells, tumor-associated macrophages, and it does cause a little bit of lymphoid depletion, but it’s much more myelotoxic than it is lymphotoxic. So to a large degree, it leaves the lymphoid compartments relatively intact, while depleting these inhibitory myeloid cells, and I think that may give pembrolizumab a chance to really activate these T-cells without any other inhibition.

Lastly, by killing the tumor at the same time that you’re activating immunotherapy, there are a lot of good data where, by killing the tumor, you give the antigen presenting cells, the dendritic cells, and antigen presenting macrophages a chance to pick up this tumor debris and display these tumor-specific antigens to T-cells, and that causes them to be activated. I’m really hopeful that we see synergy.

What about olaratumab (Lartruvo) and its recent FDA approval? What impact do you expect that to have on this treatment landscape?

Having said that, this is a phase I/II trial, and we’re really getting into the safety component of it. I’m always cautious. We have to really confirm the safety of this combination, as it’s a novel combination, and then we can really start to look at these exciting questions of efficacy.I think it’s going to be huge. I think everyone’s going to start using it. I think it’s a great drug, and I’ve used it a lot myself. I think that we want to see the results of the phase III trial. Unfortunately we have, in sarcoma, had phase II trials that were positive, and then phase III trials that were negative. I don’t think that’s going to be the case with this drug, but you never know. You’ve got to be cautious.

There’s also some other ways to look at olaratumab that are coming down the pike. There’s a trial testing it in a combination with gemcitabine and docetaxel that I think is really exciting. I think that it would be great to have it more as a kind of drug you could combine with other agents. It doesn’t have a lot of toxicity, so I think it’ll be exciting to test it in even more combinations.