This article is more than 5 years old. Information may no longer be current.
Research initiative to assess link between cancer’s genetic composition, radiation response
Click Here to Manage Email Alerts
A researcher at Cleveland Clinic received a $2 million grant to investigate the relationship between a cancer’s genetic composition and the efficacy of radiation therapy.
Mohamed Abazeed, MD, PhD, assistant professor of medicine and radiation oncologist at Cleveland Clinic, will strive to identify genetic biomarkers that will allow clinicians to treat patients with more tailor-made radiation treatments.
The studies he proposes, which combine new computational and statistical approaches with large-scale biomarker-driven assessments of patients and patient-derived xenografts, will aim to assess genetic regulators of radiation resistance and further delineate the associated pathways.
Abazeed’s group completed the largest profiling effort of survival after radiation in cancer cell lines, comprising a diverse collection of 533 genetically annotated tumor cell lines from 26 cancer types. The proposed work seeks to translate some of these findings toward clinical use.
HemOnc Today spoke with Abazeed about the evolution of his idea, the aims of the study, and what subsequent research may entail.
Question: How did this research initiative come about?
Answer: The inception of this work can be traced back to my days as a radiation oncology trainee. My most vivid memory of treatment response heterogeneity was over the course of my 3 months on the thoracic service at Massachusetts General Hospital. It became obvious to me that, despite delivering similar doses of radiation to patients, there was remarkable variation in tumor responses. Although there was a lot of hand-waving and conjecture about the basis of this variation, no one really had a clue why this was the case. I decided that this topic is important because it suggested a means to personalize radiation treatments, and I resolved to study it.
Q: It seems like such a simple and straightforward idea . W hy had no one considered this before?
A: I was at the right place at the right time. The field of cancer genomics was entering hyper-drive mode. As I transitioned into the research phase of my training, there were many concurring developments. The Cancer Genome Atlas projects were either well underway or near completion, we were beginning to understand more about the genetic landscapes of the cancers, and the cost of next-generation sequencing technology was precipitously declining. Also, my soon-to-be post-doctoral mentor, Matthew Myerson, MD, PhD, along with others, had identified targetable EGFR mutations in lung cancer. This demonstrated that targetable alterations found in cancers of epithelial origin can be targeted and that genetic profiling can result in improved patient responses. The field of radiation biology, which measures the interaction between ionizing radiation and biologic material, has made significant contributions to our understanding of disparate topics, such as DNA damage response, tumor hypoxia, and experimental models of tumor kinetics and repopulation. However, genetic determinants of response to radiation had not be investigated on a large scale. The training, resources and mentorship that I received at the Broad Institute and Dana-Farber Cancer Institute is what made my investigations at the time possible.
Q: Can you elaborate on what you will do with this initiative?
A: In many ways, my transition to Cleveland Clinic in 2013 and my scientific development as a member of the faculty here mimics the evolution of our work from a genetic biomarker discovery approach toward clinical translation. The clinic is a wonderful place to conduct studies for the purpose of translation into routine clinical practice. In the years since my arrival, we’ve been diligently assembling the requisite computational and experimental expertise, along with an extensive amount of cell-, mouse- and human-based material. Due to the transdisciplinary nature of our program, this initiative represents an integrated laboratory effort that could help develop a new standard for the validation of radiotherapeutic biomarkers in cancer. Our program has several key features. They include in vitro studies that use a large panel of cell lines that represent multiple genetic backgrounds; mechanistic studies that indicate biological relevance; the use of mathematical and mouse models that better reflect inter- and intratumoral heterogeneity of the original tumor; and large-scale correlative clinical studies in two distinct and large populations of patients with non-small cell lung cancer. We believe that implementation of this program could accelerate the discovery, translational speed and success of clinical studies that ultimately will use biomarker-driven stratification, tailor radiation prescriptions and schedules, and/or employ more precise drug or radiation treatments.
Q: So you are laying a foundation on which others can build?
A: Yes. Clinical trial success in oncology is unacceptably low. We need a better preclinical program to optimize the design and quality of clinical studies. For example, with the exception of HPV-positive tumors in head and neck cancers, clinical studies alone have not helped to stratify patients on the basis of whether they are more or less likely to respond to radiotherapy. We believe the answer is first to select biomarkers using an inductive approach — without bias — followed by rigorous validation in several complementary preclinical models. As it pertains to the latter, I’d like to highlight one critical feature of our program. That is the use of patient avatars, also known as patient-derived xenografts. There are several lines of evidence that these models faithfully recapitulate their tumors of origin. Our group has generated a substantial repository and we plan to effectively utilize them in our proposed studies. We think these models are poised to serve as powerful tools to enhance clinical trial success.
Q: How deep into dosing levels will you dive? Will you come away with specific numbers, or just a general idea about response or nonresponse?
A: Our goal is to incorporate biological predictors of response to radiotherapy. That has several implications, including radiation dosing and other minutia. I imagine prediction estimates similar to the way we predict the weather. We’ve made substantial improvements in the atmospheric sciences over several decades. This is attributed to more data and the better modeling of a complex system. As a result, you and I now have some degree of confidence whether it’s going to rain tomorrow. I’m hopeful that similar successes can be achieved in radiation oncology. Our goal is to provide the field with the information capability to make predictions about the extent that a patient before you will respond to therapy. The ability to predict can have significant implications for clinical care, including dose escalation or de-escalation, recommendations for alternative therapies — for example, surgery — and/or the use of particular combinations of drugs with radiation on the basis of individual tumor genetic features.
Q: What are your expectations for the results? Are you making any predictions ?
A: The irony about our effort to develop models for prediction is that we personally don’t make any predictions, at least initially. We infer genetic associations using an unbiased selection of candidate biomarkers. We decrease our false-discovery rate using several statistical approaches and we establish causation through a rigorous target validation process. Therefore, we are guided by our data rather than by our suppositions. That is generally a good thing because humans have well-documented cognitive biases — including availability heuristics, the post hoc ergo propter hoc fallacy and the anecdote fallacy — that frequently result in poor rationale for lines of scientific investigation or, worse yet, unscientific recommendations.
Q: When do you expect results?
A: It is a 5-year award with yearly milestones and deliverables. The focus in the first 2 years will be on retrospective and prospective biomarker validation in human tissue and blood. In the last 3 years, our focus will shift toward mouse preclinical and co-clinical trials to sensitize lung tumors that we predict will be the most resistant to radiotherapy. – by Rob Volansky
For more information:
Mohamed Abazeed, MD, PhD, can be reached at Cleveland Clinic, 9500 Euclid Ave., Cleveland, OH 44195; email: abazeem@ccf.org.
Disclosure: Abazeed reports no relevant financial disclosures.