In the January 2016 issue of the JCI, Hedberg et al. use whole exome sequencing (WES) to identify genetic alterations that occur in metastatic and recurrent head and neck squamous cell carcinoma (HNSCC). In a small cohort of patients, this study examined the genetic profiles of patient-matched tumor pairs and identified newly arisen DDR2 mutations in several recurrent HNSCC cancers. These mutations may provide a new treatment target, as such mutations confer sensitivity to dasatinib, a SRC family kinase inhibitor, in this and other studies. This month, we had the opportunity to chat with Matthew Hedberg about this research, its impact on the management and treatment of HNSCC, and his training and career as an aspiring physician-scientist.
What was your role in developing this project?
WES has been performed in primary, metastatic, and recurrent disease in other cancers. This is the first cohort WES study in HNSCC due to some unique challenges inherent to head and neck cancer. It is less common than many of the cancers that have been more extensively studied to date and has a high degree of morbidity; and it can be difficult to obtain high-quality tumor samples of metastatic and recurrent disease due to things like necrosis that can occur in cystic lesions, prolonged clinical follow-up times, and treatment field effects from primary therapy that can compromise recurrent tumor samples, etc. We were uniquely able to undertake this study because of the prospective tissue banking that we undertake in the head and neck cancer SPORE program built under Dr. Grandis at Pittsburgh. A conversation between Dr. Grandis and Dr. Richard Lifton, while he was in Pittsburgh for a talk, resulted in a collaboration that allowed for this study utilizing our clinical samples and Dr. Lifton’s sequencing resources.
As data came out of Yale’s sequencing core, Gerald Goh of Dr. Lifton’s group curated the sequencing results, and we both carried out independent data analysis of the alterations that were identified. After comparing our conclusions, which were highly similar, I proceeded to design, optimize, and conduct the benchtop modeling experiments we present in the paper. I then analyzed the data and interpreted the results in the context of what our findings may mean for the clinical care of head and neck cancer patients specifically, and cancer patients generally.
What did you find most challenging or unexpected in this process or this project?
Several technical challenges were encountered. We had optimization issues with our early sequencing attempts. Some sequencing artifacts that Gerald detected caused us to question some of the initial primary data, and ultimately we went back and re-extracted DNA for resequencing. As is often the case with human cohorts, tracking down a lot of the clinical parameters that we wanted to examine was challenging at times — especially in the setting of patients for whom you are observing a treatment history of many years, with multiple encounters, related to both their head and neck cancers, as well as their other comorbidities. All while deciphering and normalizing input from different providers and different electronic and paper medical records in order to generate the most robust and up-to-date data set possible for our analyses.
Designing reasonable experimental models to test hypotheses in also challenging in HNSCC, as the impact of a single genetic modification can be difficult to ascertain the setting of a genetically heterogeneous cancer like HNSCC. But the experiments we ultimately pursued demonstrate an effect of DDR2 alterations on dasatinib response in the models we used, with the caveat that a lot of other contributing factors are likely to influence response in any one set of experimental HNSCC models and/or patients. Overall, I feel this research serves as a reasonable template for future efforts to probe the evolution of this disease over time within individual patients, and how we might adapt our treatment strategies to be more responsive to this evolution and hopefully improve outcomes.
How do you foresee this practically impacting diagnosis and treatment in head and neck cancer?
Quite frankly, this is an opening salvo into metastatic/recurrent HNSCC. With a cohort of this size, we are not going to immediately change the way we are managing this disease. But it serves as an example for similar sequencing projects and experimental designs that can be utilized as we build larger databases and develop an evidence base to inform our treatment strategies. Even in this small cohort, we were able to describe an expansive spectrum of genetic heterogeneity that exists between primary tumors and synchronous metastases versus primary tumors and metachronous recurrences. The heterogeneity we observed, especially in the setting of recurrent disease, argues in support of a personalized medicine approach not just for individual patients, but also for the same patient at the point of both primary and recurrent disease. The characteristics of recurrent disease likely evolve secondary to the treatment received, the respective genetics of the patient and primary tumor, and other comorbidities. Our findings can support such theories. But at this early stage, we need a larger database of sequencing information on recurrence and metastatic head and neck cancers to draw statistically significant conclusions. The cost-effectiveness of tailoring clinical approaches remains a challenge in many scenarios. But it is known to be beneficial in appropriately selected subsets of patients. And, at the end of the day, physicians treat patients directly and populations by extension. Any treatment that we render is going to be targeted to a single individual, informed, ideally, by its efficacy in larger cohorts. It’s my hope that our research can inform and improve these efforts.
What do you think was the role of DDR2 in this study? What is the normal role of the gene in head and neck cancer, and what are the implications of identifying this as a mutation in a subset of your cohort?
The role of DDR2 in head and neck cancer is just starting to be explored. Phenotypically, there is some recent evidence (1) suggesting that it may lead to increased invasion and other cancer-associated phenotypes in HNSCC. At the molecular level, a lot remains to be teased out. We know there have been certain DDR2 mutations associated with sensitivity to SFK inhibitors in other malignancies, which was consistent with the results we obtained in our experiments. The real question in HNSCC going forward is, “Are these genetic events enriched in recurrent disease compared to primary disease?” A larger proportion of the recurrent tumors in our cohort harbored DDR2 mutations than has been seen in primary disease. But our cohort is not large enough to determine if this is a statistically significant increase. As the scientific community expands our dataset and sequences more samples of recurrent disease, the results will have implications for the therapies we offer to patients who recur.
Would you foresee targeting DDR2 early on in treatment before the mutation has arisen to prevent such a mutation from occurring, or only once you discover such a mutation occurs in the tumor?
Monotherapy often fails over time; so the challenge is in figuring out the appropriate combination therapy. We did not measure intratumor genetic heterogeneity in this project, but others have, in head and neck cancer. There is an appreciable degree of intratumor genetic heterogeneity in HNSCC, with a variety of different subclones that can variably become predominant over time as the cancer evolves in the setting or absence of treatment (2). If you can identify the predominant subclones in a given tumor, and if there was a DDR2 alteration in one of them, I think it represents a reasonable target, ideally in combination with one or more additional targetable genetic alterations, such as PIK3CA mutation/amplification. This is likely to occur together with standard-of-care chemotherapy and/or radiation. Targeting these subclones could be a significant contribution to therapy, especially if it has an additive or synergistic effect with chemotherapy and radiation therapy. Again, kinase inhibitors, and monotherapy in general, eventually fail. A subclonal population will evolve, and resistance will arise. But if we can use targeted therapies in select patients to supplement other frontline therapies like radiation and chemo, and perhaps even immunotherapy, we may be able to overwhelm the tumor’s ability to evolve and survive. Optimizing combinations of these modalities really represents the forefront of translational research in HNSCC and other cancers.
How did you decide on your career as training as a physician-scientist?
The research bug bit me pretty young. I was involved with research starting in high school at the suggestion of my mother. She encouraged me to apply for this summer program where they trained high school students in microbiology lab techniques. At the end of the summer, my PI, Dr. James E. Metherell, offered me a part-time position, which I accepted. I ultimately continued research through high school and undergrad, where I had the opportunity to work under Drs. Malay Haldar and Mario Capecchi. Through these experiences, I found that I enjoyed research: the challenges, and the idea of attempting to answer questions that nobody has the answer to. However, I also had several struggles: projects that went on for months, or even years, that ultimately fizzled out and didn’t amount to much. Given the significant ups and downs of the research lifestyle, I decided: “Okay, if I'm going to commit the remainder of my professional efforts to this, I want to make sure that my research is something that will be applicable, impactful, and really make a change in somebody’s life.” Around that same time, I started shadowing some physicians and really enjoyed the clinical/human side of things. With these exposures, I just said: “You know, I seem to like all of this stuff, and if I could make it work together, that would be pretty cool.” The physician-scientist career just seemed like a natural fit from there that I grew into over time.
Do you see yourself as someone who enjoys the clinical side more now that you’re back in the clinic, or do you miss being in the lab?
You know, it is really hard to make those comparisons. The mantra of our MD/PhD training programs is that we are going to be the ones that bridge that gap and really bring those two parts of the Venn diagram closer together. But, in my limited experience, they really are two different worlds. Working with patients can be tremendously rewarding, especially on an immediate timescale. You can go in and help somebody with something that day — make a diagnosis, make a treatment recommendation, and know that you have improved their life right then and there. That is a pretty special feeling.
That being said, the excitement, and sometimes even the agony, of research can be just a whole different type of challenge. Tinkering away on a problem for which there is not a known answer is very different than being in the clinic. Science is a totally different animal, totally different drill, and to some degree, a different skill set, although there is definitely overlap. It’s just two different experiences, both of which are very, very rewarding, and I don’t know that I could pick a favorite child.
What factors so far have contributed to your success?
We talked already about my parents, who strongly suggested that I do that high school research program and always supported my academic pursuits. That catalyst, in combination with some great high school chemistry, biology, and physics teachers, really got me into the sciences. The major contributions to the limited success I’ve achieved thus far have been great mentors, incredible support from my family, and having the opportunity to work with some tremendous teams of researchers. Everybody on this paper has been fantastic. Dr. Goh was absolutely invaluable. Dr. Bauman has helped guide my clinical maturation, gain exposure to HNSCC patients, and translate this basic science into the clinic. Throughout my undergraduate and PhD training, Drs. Capecchi and Grandis really showed me how to think on a large, “big question science” level, and how to network. Dr. Grandis is especially skilled at working with other groups. The collaboration on this project, with all these different researchers, is a perfect example. Ultimately, I have been very fortunate that I have been able to be on some terrific teams of scientists and clinicians, and be a part of some great programs, such as the medical scientist training program and molecular pharmacology program here at the University of Pittsburgh. These programs have offered a tremendous amount of support to help guide me into good laboratories and make progress through what can be a very protracted training process.
(1) ) Xu J et al. Overexpression of DDR2 contributes to cell invasion and migration in head and neck squamous cell carcinoma. Cancer Biol Ther. 2014;15(5):612–622.
(2) Zhang XC et al. Tumor evolution and intratumor heterogeneity of an oropharyngeal squamous cell carcinoma revealed by whole-genome sequencing. Neoplasia. 2013;15(12):1371–1378.
About the First Author
Matthew Louis Hedberg, PhD,, is an MD/PhD candidate at the University of Pittsburgh-Carnegie Mellon University Medical Scientist Training Program. He received his BA in chemistry from the University of Utah, where he worked under Dr. Mario Capecchi. He received his PhD in molecular pharmacology and chemical biology from the University of Pittsburgh, under Dr. Jennifer R. Grandis and is currently a third-year medical student at the University of Pittsburgh School of Medicine. He holds a Ruth L. Kirschstein NRSA MD-PhD F30 predoctoral fellowship grant from the National Cancer Institute.
About the interviewers
Freddy T. Nguyen is an MD/PhD candidate at the University of Illinois at Urbana-Champaign. He is the founder of the American Physician Scientists Association and served on the Associate Member Council of the American Association for Cancer Research. His research interests currently lie at the intersection of biomedical optics and cancer research. He received his BS in chemistry and BA in mathematics from Rice University.
Evan Noch, MD, PhD, is a third-year resident in a research track position within the Department of Neurology at Weill Cornell Medical Center-New York Presbyterian Hospital in New York City. He currently serves as the Chief Resident for Research for the program. He studies glioblastoma metabolism in the lab of Lewis Cantley and plans to pursue a fellowship in neuro-oncology. Dr. Noch received his MD and PhD from the MD/PhD program at Temple University School of Medicine.
Matthew L. Hedberg, Gerald Goh, Simion I. Chiosea, Julie E. Bauman, Maria L. Freilino, Yan Zeng, Lin Wang, Brenda B. Diergaarde, William E. Gooding, Vivian W.Y. Lui, Roy S. Herbst, Richard P. Lifton, Jennifer R. Grandis