Data Release to Enable Smarter Trials for Pediatric Cancer: Q&A With John Maris, MD

With the release of over 200 genomic tumor models spanning 25 different types of childhood cancer, researchers may now have the ability to skip lengthy preclinical work in their development of novel treatments. With funding from Alex’s Lemonade Stand Foundation (ALSF), the Pediatric Preclinical Testing Consortium (PPTC) announced their data sets will now be made available to any qualified academic petitioner — a move that John Maris, MD, oncologist at Children’s Hospital of Philadelphia Cancer Center and principal investigator of the PPTC’s CHOP site, believes is the first of its kind.

The project will enable more precise clinical trials by not only allowing scientists to draw on previous analyses (so as to identify the most impactful drugs to take to clinical trials and avoid repeating failed experiments), but also by providing researchers with a wealth of data on genetic targets rather than broad disease types.

“One of the goals of my laboratory and the PPTC is to really increase the probability that a child is going to benefit from a treatment because we have a strong predictor of whether or not they’re going to respond,” Dr. Maris said. “We know that CAR T-cells against CD-19 won’t work if CD-19 is not there. Many other drugs in development likely have as robust a biomarker as CD-19 for CAR T-cells; we just don’t know what they are yet. So, this really is about moving towards treating pediatric cancer not as diseases, but as molecular entities. Five patients who have neuroblastoma and their tumors look identical under the microscope, may actually be genetically very distinct.”

With information about both patients’ genetic data and their response to various drugs at their fingertips, scientists are better positioned to design and conduct clinical trials based on specific genetic targets. We sat down with Dr. Maris to learn more about the impact that such a data release could make on improving outcomes for children with cancer.

Let’s start with the PPTC: Can you briefly describe its goal and the data set it holds?

For the last 15 years, I have been part of a National Cancer Institute-funded program called the Pediatric Preclinical Testing Consortium. The major issue that we’re trying to address is prioritizing which new drugs to take to clinical testing. There are, thankfully, fewer childhood cancer patients than there are adult patients, and so we can’t just test every new drug as it comes along. The goal of the PPTC is thus to work with industry to select new drugs for testing in so-called patient-derived xenograph (PDX) models of childhood cancer. A PDX model is when we get a gift from the child of a tumor biopsy, and it comes over into the lab and goes right into a mouse rather than being manipulated in a test tube. We’ve shown in the past that those models are much more faithful recapitulations of the tumors of the patients in the clinic. We then develop large cohorts of these mice and run clinical trials just like we would in the clinic. And those [drugs] that score can move forward, and those that don’t get de-prioritized.

Over the years, we’ve developed over 250 of these models. And because what we were faced with is that many of the new drugs are not necessarily developed for neuroblastoma or breast cancer, but again specific to a genetic target, in the PPTC we decided to do a very extensive genetic characterization of all of our models.

Why did you and ALSF and the PPTC decide to release the data?

We have developed a very large set of data, and we really think this data is important to anyone who wants to do experiments. The traditional way things are done in academia is that you do your experiment, you analyze the data, you write a paper, and then when you publish the paper, you release the data. But we wanted to release the data as soon as it was ready, and these models are open to anyone. I think the significance of this is that, if a company has a great drug against a specific mutant protein, we can now say, these models have [that mutation] and these models don’t, and therefore we can do much more rigorous preclinical testing and hopefully be much more efficient in our timeline of testing and getting these into the clinic.

How does the traditional timeline of preclinical testing compare to what is possible with this data release?

The process used to be, if you test a drug and maybe some tumors responded and some didn’t, then we’d have to go through the exercise of trying to figure out why. And that could take months or even years. Now, the explanation may be right in front of us because it was either the exact hypothesis we were testing there that had the mutation or not, or even if our hypothesis was wrong, we have this enormous amount of data at our disposal to ask: Why did some respond and why did some not? There’s not a single answer to your question, but I think that we’ve streamlined significantly.

How do you see this project growing, or what do you see in the future for it?

Cancer is complex, and there are many things you can determine about a cancer from the types of assays we did, which are largely genetic sequencing. But there are other very important changes that are silent to the sort of technology that we did, so we’re planning a proteomics project for all of these where — instead of sequencing the DNA and RNA — we examine the quantities of key proteins that are involved in cancer growth and metastasis. And then we continue to generate models. So we view this as sort of version one, and the proteomics will probably be version two, and I think the version three would be assaying the new models and making sure that they have the same sort of characterization.

Why would you describe the ALSF/PPTC data release as a unique tool?

This is the first time that I’m aware of that an academic consortium has teamed up to generate data to guide drug development. It’s not surprising that some of the larger drug companies have done similar things with their own drugs, but as far as I’m aware, this is the only release of childhood cancer data where multiple types of different cancers are all analyzed together, and it’s not just the genetic data — that’s there and that’s important — but it’s also the response data to the different drugs. Making that available broadly will allow other researchers to either not repeat experiments that are doomed to fail or extend experiments that look very interesting. I think that’s a very unique resource.

Learn more about Dr. Maris’ neuroblastoma work on Cornerstone.