In this interview we discuss health disparities in cancer care, how redesigning cancer clinical trials can address some of these disparities, and goals for the “next generation” of cancer health disparities research.
Four leading national cancer organizations recently released a joint position statement to guide the future of cancer health disparities research. The statement represents a unified strategy by the American Association for Cancer Research, the American Cancer Society, the American Society of Clinical Oncology, and the National Cancer Institute to foster cooperation across the cancer research community to ensure that all patients-regardless of social demographics, socioeconomic status, or the communities in which they live-benefit from cancer research.
1. What are some of the most well-understood or researched health disparities that exist in cancer care?
DR. POLITE: What we know in almost every single cancer is that there are differences in both incidence and survival when you look at the data by race-especially when we look at black vs white populations-and we also see it with regard to socioeconomic status. The field of health disparities has a focus on why some groups are dying from cancer at a higher rate than others, and trying to understand and unpack that question.
2. The statement discusses goals for the “next generation” of cancer health disparities research. How do you define the last or current generation of disparities research and how will that evolve?
DR. POLITE: Where the field has been focused for the last 15, 20, 25 years is on trying to describe the problem and really trying to understand, why are there differences in the incidence of cancer? Why are there differences in the treatment of cancer? Why are there differences in survival after treatment? We’ve been looking deep down at why differences occur and how best to define populations.
Where we believe the next generation has to move is in the direction of, how do we begin to find solutions to these problems? Many of us leading that effort believe that simply identifying and defining the problem cannot be the goal of the next generation of research. It has to be how we move the field forward and capitalize on the incredible science and work that has been done in health disparities research. That was why this group came together.
We are a group of people doing health disparities research, some for decades. All of us feel that the field of cancer health disparities has matured well and that the science has matured well. Now is the time to take everything we have learned, take that science, and start to be more rigorous about looking for solutions and making sure we have set up an infrastructure that will allow us to get solutions.
3. One of the listed priorities of the position statement is redesigning cancer clinical trials to acknowledge and address cancer disparities. Can you discuss this goal? Why was it chosen and how can it be achieved?
DR. POLITE: Often within cancer clinical trials the health disparity question is an afterthought; the trial is finished, a retrospective analysis is done, and if racial/ethnicity or socioeconomic data were collected, investigators use the data to discern whether or not there were any differences by race, ethnicity, or socioeconomic status. It is rarely a hypothesis-driven goal of the trial.
Our first recommendation is, in many trials, especially for cancers in which we know there are significant disparities, such as triple-negative breast cancer, that the disparity question has to be embedded in the trial and the trial has to be powered to detect it, just like any other endpoint. You wouldn’t look for a progression-free survival endpoint without adequately setting the trial to answer that question. We believe that if you are going to start asking questions on whether there are differences by race, ethnicity, or socioeconomic status, then you have to set up a hypothesis and make sure you are collecting the correct variables in a prospective fashion, so that at the end of the trial you can actually answer the question. That may require the trial to have specific targets for different racial, ethnic, or socioeconomic groups; for instance, a goal of the trial might be to enroll 1,000 patients with at least 15% to 20% being African-American. If the trial is under-enrolling African Americans, you could consider halting enrollment for white patients and consider oversampling African-American patients to achieve the goal, so that you can answer the disparity question that you proposed in your hypothesis.
We have to move away from post-hoc analyses to ones in which we embed the disparity question in the research design, the same way we are starting to embed questions of genomics, next-generation sequencing, etc. In clinical trials, we should be embedding the disparity question at the beginning and making sure that we are collecting the right data and powering the trial appropriately to answer the questions.
4. Another important stated goal is advancing community engagement in cancer research. Can you give an example of where such efforts have struggled in the past and how they can be improved?
DR. POLITE: There have been many successful attempts at this. Some groups have done a wonderful job. I will point to Electra Paskett and her team at Ohio State University working in Appalachia for many years. They have done it because they have a dedicated commitment from Ohio State and from funders to establish their relationships, fund junior faculty, and to be there year after year and decade after decade. Tremendous work has come out of their group as a result of that.
How do you get successful at doing this type of research? First, you can’t think in terms of grant cycles. If you go into a community and after 3 to 5 years the grant runs out and you leave, this will leave a lasting impression on the community that you are really not there for the long term.
Another issue that arises in academia is that if junior faculty want to engage in community research, they must spend the first several years building trust in the community, and listening and understanding what the needs are. If you look at a typical promotion clock, by year 7 you have to have enough published research in order to be promoted. If you spend your first 2 to 3 years developing relationships, you are behind the eight ball. It becomes a fool’s errand for people to go into community-engaged research if they want an academic career.
For example, John Carethers, who is the chairman of medicine at the University of Michigan, was on our committee and has been implementing a system in which the clock is extended by 2 years for those engaged primarily in community-based research. He believes we should give them the extra 2 years to establish those relationships, so that this type of research becomes something that people are willing to do.
It needs to be understood that doing research in the community does not just require coming in with a protocol, popping it in there, and then leaving. It requires building relationships, building trust, and committing to being there beyond a grant-funded cycle. If you can’t do that, then your ability to do truly deep, meaningful community-engaged research is going to be severely limited.
5. Finally, what should practicing clinicians know about this position statement? What, if anything, can they do now to help?
DR. POLITE: For those who are running clinical trials, obviously it is important to make sure that they use the tools that are out there to recruit as many patients as possible-especially in practices that have minority or low socioeconomic status populations. For practices that are operating in lower-resource environments or are serving large minority populations, and are not part of the clinical trial process, work should be done to get these populations involved.
One of the things we have seen in published studies is that we have a dramatic underrepresentation of genomic samples from minority populations. If you look at The Cancer Genome Atlas, we cannot investigate molecular differences in minority populations if these differences have allele mutation frequencies of 5% or 10%. We do not have enough samples from these groups to be able to even ask the questions.
We want clinicians who have access to these populations to be involved in things like The Cancer Genome Atlas, involved in collecting and sending blood and tumor samples-so that we can have a more representative collection of samples. The committee is very worried that as we move into the genomic era-with a lot of therapies being developed by pharma and decisions being made about what markers work and don’t work-that we are going to leave large segments of our population behind because we simply don’t have information on them. We don’t even know if they have mutations. We don’t know what the mutations are. If we don’t know these things as we move rapidly into personalized medicine, we will see disparities widen as we leave groups behind.
It becomes incumbent on all of us in the community to make sure, especially if we have populations that are not majority populations, that we think about ways to engage all populations in clinical trial research and that we look for opportunities to collect tissue samples, blood samples, etc, so that they get incorporated into these large biorepositories.
Financial Disclosure:Dr. Polite receives research funding from Merck and is a consultant for GLG and Pfizer.