Fellows Spotlight: Andrea Arfe

MORE ABOUT ANDREA:

1. Can you describe your background and where you came from before this fellowship?
My research career began at the University of Milano-Bicocca (Milan, Italy), where I worked for about 5 years as a researcher in pharmacoepidemiology. During my time at the Bicocca University, I collaborated with physicians, pharmacologists, and computer scientists to study the safety and effectiveness of commonly-prescribed drugs—including non-steroidal anti-inflammatory drugs and anti-diabetic medications.
After my experience at the University of Milano-Bicocca, I completed a PhD program in Statistics at the Bocconi University of Milan. There, my doctoral research focused on developing new statistical methods useful for the analysis of data on cancer patients and therapies. I conducted this research as a visiting scholar at the Dana-Farber Cancer Institute in Boston, where I have started several collaborations with oncologists to validate different metrics to predict response to breast cancer therapies or novel drugs in immuno-oncology.

2. Why did you choose to pursue this fellowship opportunity?
My research goals are to contribute to projects that can ultimately promote the health of patients, as well as to develop innovative statistical methods useful for clinical trials in oncology. These goals cannot be fully realized without effective collaboration with medical researchers and consideration of the regulatory processes that guide the development of new cancer therapies. As part of the CRS, I am able to interact and collaborate with other experts in oncology and regulatory issues, providing me the necessary know-how and support to progress in my research projects.

3. What was the focus of your research during the fellowship, and what did you learn?
In my methodological research at the CRS, I use tools from Data Science and Machine Learning to improve the way we test new cancer therapies in clinical studies. In my approach, the task of designing one such experiment can be considered, in part, a data analysis problem. In fact, previous clinical trial data from patients that received a specific class of treatments can be fundamental to plan experiments that are faster, less expensive, and more likely to advance innovative treatments. For example, I am studying how data from past studies can help improve the design of early-phase trials in immuno-oncology. Designs that leverage data from completed studies of other therapies in the same class can help predict which doses of a new drug will be associated with a favorable balance of therapeutic benefits and risk of adverse events.
Doing research at the Center for Regulatory Science, I also became interested in the statistical problems posed by the development of new therapies for children with cancer. Progress in drug development has steadily improved survival from childhood malignancies over past decades, but there is still a tremendous need for safe and effective drugs to treat pediatric cancers. The development of new cancer therapies for children is more difficult than for adults because of several issues, including the difficulty of enrolling young patients in clinical trials and the rarity of specific tumor types in pediatric populations. In my research, I aim to understand what are the potential roadblocks for clinical trials of novel therapies for childhood tumors, as well as to propose novel approaches to circumvent them. For example, I am studying clinical trial designs that could make it feasible to develop novel targeted cancer therapies for young patients—a very important issue in light of recent reforms of the Food and Drug Administration.

4. What is your favorite memory from your time at CRS?
I enjoy the talks that are offered at CRS, which typically cover different issues related to the development of new drugs or medical devices. The speakers often have a very different background than mine, such as medicine or computer science. I find this variety in perspectives to be very stimulating, and also useful to get a sense of the “big picture”.

5. What advice do you have for someone just getting started in the regulatory science field?
Regulatory science is a big multi-disciplinary field, involving legal scholars, bioethicists, statisticians, and biomedical researchers. If you don’t think you can contribute, think again!

6. Can you describe a challenge you have faced and what you learned from it?
Sharing of patient-level data from clinical trials is still limited and finding datasets useful to develop my projects was a major roadblock for my research. To make progress, I learned how to use some tools that allow to reconstruct data from the figures and summary results reported in trial publications.

7. What are you doing next?
Starting September 2021, I will join the faculty of the Memorial Sloan Kettering Cancer Center Department of Biostatistics and Epidemiology in New York. There, I will continue developing the research topics started at CRS and collaborate with oncologists and other cancer researchers in applied research projects. My experience as a post-doctoral fellow at CRS has greatly prepared me for this position, thanks to the research support, mentoring, and access to training it provided me.

8. Is there anything else you'd like to share?
If you have data from a clinical trial, please share it online if you can. You are going to make a statistician very happy (and maybe help improve future trials too).