.png)
I had the pleasure of sitting down with Ronac Mamtani, MD, MSCE, section chief for genitourinary cancers and researcher at the University of Pennsylvania and Penn’s Perelman School of Medicine at Reuters Digital Health 2026. In the last decade of partnership with Flatiron, Dr. Mamtani has been asking some of oncology's most important questions, mentoring the next generation of physician-scientists at Penn Medicine, and pushing the boundaries of what real-world evidence can tell us about how patients actually fare in the clinic, not just in clinical trials. He has a gift for making complex research feel urgent and deeply human, and he is one of the most thoughtful voices I know on what meaningful research and mentorship look like in practice.
Partnering with oncologists enables Flatiron to push forward our mission to improve and extend lives by learning from the experience of every person with cancer. Leveraging AI, machine learning, and human abstraction, Flatiron curates data from unstructured formats—clinician notes, pathology reports, imaging—to generate research-grade data at scale. Today, this means researchers can learn from the experience of 6M+ patients represented in our network. And as our network continues to grow, so does the opportunity and statistical power to study populations that have historically been too small to analyze.
What follows is a conversation about how that work happens, why it matters, and where it's going.
You are an experienced leader and academic researcher. When you have a research question, how do you approach answering it?
Dr. Mamtani: Historically, cancer researchers like me were limited to three general sources of data, and each one told us part of the story, but not the whole story. Cancer registry data like SEER gave us rich cancer information but limited treatment detail. Medicare claims gave us treatment information but not the cancer context. And our own EHR data gave me everything I needed, but only within the walls of one institution.
Flatiron combines cancer stage, histology, biomarker data, genomic profiling, performance status, granular treatment information, and clinically meaningful, validated endpoints like real-world progression, response, and adverse events. It isn't about having more data. It’s about having the right data.
How has this changed what Penn can do as a research institution?
Dr. Mamtani: Three things. First, it allowed us to use real-world data to complement our clinical trial program, not replace it. Clinical trials tell us how a drug performs in ideal scenarios with ideal patients. Flatiron data tells us how a drug performs in routine care, in a broader, more representative population. Those are different and complementary questions, and you need both.
Second, it lets us move beyond institution-specific observations to population-level insight, going from months or years of data collection to rapid hypothesis testing.
And third, it's democratized access to high-quality, research-ready data for the full spectrum of investigators at Penn. Our med students. Our residents. Our fellows. Watching that next generation get their hands on this kind of data has been really remarkable.
What does the patient impact actually look like?
Dr. Mamtani: The most common question I get in clinic every day is "Will this treatment work for me?" As an oncologist treating patients with advanced cancer, I turn to the clinical trial, but sometimes it's an imperfect fit, because we're trying to apply a result from a trial population to the specific person sitting in front of me.
With Flatiron, we take what we learn from the trial, then use real-world data to build models that account for the individual characteristics of millions of patients across the US. That tells us how a drug performs in patients who actually look like the people I treat.
One study that captures this well: Lova Sun, MD, an assistant professor at Penn, used Flatiron data to ask a question that had gone unanswered for far too long—how long should a patient with controlled cancer remain on immunotherapy? In clinical trials, treatment was capped at two years. But in the real world, patients are scared to stop. They don't want their cancer to come back.
Dr. Sun's study, published in JAMA Oncology, used rigorous statistical methods to compare patients who stopped immunotherapy at two years against those who continued. The finding was reassuring: stopping at two years didn't lower survival. Those results immediately changed conversations in clinic. That's a high-impact publication, and an answer that matters to real patients.
Where do you see the partnership going from here?
Dr. Mamtani: I'm most excited about moving beyond effectiveness. My job isn't just to help patients live longer, it's to help them live better. That means we need to understand real-world safety: who is most at risk for toxicity, when does it occur, and how can we intervene earlier. I think that territory is genuinely paradigm-shifting.
Also, when you talk about a larger sample size to a clinical investigator, we immediately think about subgroups—the populations that are consistently underrepresented in clinical trials: women, underrepresented minorities, elderly patients. We now have the sample sizes to study those groups and answer questions with greater precision. That's been a long time coming.
If you’re an academic researcher or health system leader looking to do more with real-world oncology data, we’d love to connect. Reach out to us.
