3 Questions on…
Answers straight from the experts on the latest news and topics in oncology
Friday, February 12, 2016
With MICHAEL KATTAN, PHD, MBA, of Cleveland Clinic and the AJCC's Precision Medicine Core
The current cancer staging system (TNM) was conceived to codify the anatomic extent of disease at diagnosis to provide an accurate prognostic factor for solid tumors—in a way that was simple, easy-to-use, and could be used worldwide. TNM assess cancer progression in patients based on three criteria: the local extent of the cancer within the site of origin (T), the degree of metastatic involvement of the regional lymph nodes (N), and the presence or absence of distant metastatic disease.
But, in the current era of precision medicine, experts say the system is too simple.
So, in 2014 the American Joint Committee on Cancer—the group that developed and maintains the cancer staging system—formed a committee of experts, the Precision Medicine Core (PMC), to develop new criteria to evaluate cancer risk calculators to enhance the current staging system and to determine which ones should be endorsed by the AJCC. Last month that committee published guidelines containing these new criteria, which they say will promote more accurate and individualized cancer predictions, guide more precise treatments, and improve patient survival rates and outcomes. The guidelines are published in a paper online ahead of print in CA: A Cancer Journal for Clinicians (DOI: 10.3322/caac.21339).
The next step will be the various cancer disease management teams of the AJCC reviewing the statistical prediction models that have been created and published in the literature in their respective areas—and making recommendations to the PMC. The PMC will make the final decisions on which models will be endorsed.
In an email interview, PMC member and lead author of the guidelines Michael Kattan, PhD, MBA, Chair of the Department of Quantitative Health Sciences at Cleveland Clinic's Lerner Research Institute, elaborated on why these guidelines are important and why the current staging system needs revamping.
1. What are the limitations of the current cancer staging system?
"It lumps patients within a stage with respect to their prognosis. All patients who are stage II would share the same prognosis—and that's not a very accurate approach. A "bad II" might have a worse prognosis than a "favorable III," but the staging system won't allow this [variability]. A statistical prediction model won't lump patients like that into groups."
2. These new guidelines do not replace the current staging system, right? Could you explain how they will be used?
"Right. The guidelines will permit the addition of statistical prediction models to accompany [cancer] staging systems. The models do not replace or affect the staging systems.
"However, there may come a day when the staging system is no longer useful in the presence of a good statistical prediction model. It is not inconceivable to, at some point, have only statistical prediction models without staging systems because the models predict outcomes more accurately."
3. How are better statistical prediction models and cancer staging systems related to precision medicine?
"All of this essentially means predictions that are better tailored to the individual patient. We move away from considering all patients within a particular stage to be the same. Instead, we take everything we know about you and make the best prediction that we can.
"The next step is reviewing the literature to find existing statistical prediction models that meet requirements. Those [models] will be endorsed. Next we challenge researchers to build more that satisfy our requirements."
Friday, January 29, 2016
With JOHN J. WHYTE, MD, MPH, of FDA’s Center for Drug Evaluation and Research
In oncology, what is the role of race in terms of variability in response to drugs? How many African Americans versus Asians versus Caucasians should be enrolled in a clinical trial? What about age? What about sex?
Those questions form the basis of what would be a very interesting discussion in oncology, explained John J. Whyte, MD, MPH, Director of Professional Affairs and Stakeholder Engagement in the Center for Drug Evaluation and Research of the U.S. Food and Drug Administration. And it was those types of questions that prodded the FDA to launch its new Drug Trials Snapshots online database, he added. “The first step was really just putting all of that information out there.”
For every new molecular entity approved since January 1, 2015 the Snapshots database includes information about who participated in the clinical trials that supported the FDA approval of that drug, highlighting any differences in the benefits and side effects among sex, race, and age groups.
In a phone interview, Whyte elaborated on why the FDA created the online database and what it means for the future of drug development.
1. How did the creation of this database come about?
“In July 2012 President Obama signed the FDA Safety and Innovation Act that required the FDA report to Congress by 2013 on the diversity of participants in clinical trials and the extent to which safety and effectiveness data is based on factors such as sex, age, race, and ethnicity.
“So there has been this effort and directive to be transparent in terms of demographic data in clinical trials. And really it was the vision of the FDA’s Center for Drug Evaluation and Research, Dr. Janet Woodcock, to implement this Drug Trial Snapshots program, a place where anyone can go and find—in an easy-to-read and easy-to-analyze format—all of the participants in clinical trials, based and grouped by sex, race, and age. And anyone can notice any differences in safety and efficacy based on this demographic data.”
2. Is that information all reported in clinical trials anyway? Was all of that information already publicly available?
“The information that’s being captured in Snapshots general existed in different documents and databases that were available to the public, but it might have been hard to find.
“For instance, all of the Medical Officer Reviews for new molecular entities are on the [FDA] website, but those data might be in very lengthy documents. So now the Drug Trials Snapshot provides you that information in text and in pictorial format that may not have existed elsewhere.
3. How do you expect the database to be used? Will it change physicians’ day-to-day practice in terms of what drugs they prescribe to patients?
“This [initiative] is about transparency. There have been a lot of advocacy groups that have said the number of women in all clinical trials should be 50 percent, or that the number of African Americans should be proportionate to prevalence of the disease in the population. What is the right number of people in clinical trials to make statements about safety and efficacy based on sex, race, and age? The first step is really putting all of that information out there.
“You’re asking—should we change our actions because there aren’t enough Asians enrolled in a trial for breast cancer or thyroid cancer? We’re not saying that. We’re saying here is one more piece of information that you want to consider. This information is intended to spur the debate and the discussion about what is the right number of people.
“And we know we likely do need to enroll more women and more African Americans in clinical trials. But, what we’re really interested in right now is to what endpoint? How do we know we’ve reached the right number?
“We’re saying let’s continue to study it to find out what is the right number. And in the meantime physicians and patients should look at this information and take it in the broader context of does it matter if only two percent of a trials participants were African Americans. It’s not a simple answer.”
Tuesday, January 19, 2016
With ROBIN T. ZON, MD, FACP, FASCO, Chair of ASCO’s Task Force on Clinical Pathways
“When appropriately designed and implemented, oncology pathways are detailed, evidence-based treatment protocols for delivering quality cancer care for specific patient presentations, including the type and stage of disease.”
So asserts the American Society of Clinical Oncology’s recently released policy statement on “Clinical Pathways in Oncology.” But the discussion about clinical pathways for cancer care is not so simple, also begging consideration of the potential harms and the challenges to implementing and using the treatment management tools, along with how they might streamline cost-effective care delivery.
In a phone interview, Robin T. Zon, MD, FACP, FASCO, Chair of ASCO’s Task Force on Clinical Pathways, co-author of the policy statement, and a Vice President and Senior Partner at Michiana Hematology-Oncology, P.C., told OT why ASCO determined it was necessary to make the nine recommendations in the statement and what’s ahead for clinical pathways in oncology.
1. What is a clinical oncology pathway—and why do they have the potential to improve the value of cancer care?
“To briefly explain, it’s important to distinguish pathways from clinical practice guidelines. Guidelines are recommendations to optimize patient care that are developed through a very onerous, systematic review of evidence and assessment of the benefits and harms for various care options. They tend to be very broad and rather flexible—much more so than clinical pathways. And guidelines typically don’t incorporate cost information into their process. The guideline are developed through a robust, transparent, peer-reviewed process and are usually held to very high standards.
“Clinical pathways on the other hand factor in treatment costs and they often tend to be much narrower in scope in terms of offering (versus clinical guidelines), often offering fewer options for oncology patients. A clinical guideline for a specific stage of a cancer may offer ten options and may not talk about cost, while a clinical pathway for that very same stage in cancer may offer only two or three options—using cost to determine those preferred care options.
“Pathways are intended to provide quality care and reduce costs—and there’s a lot of potential for pathways. It isn’t an issue of what’s wrong or what’s good about the pathways—it’s really a question of what’s the greatest potential of what can be achieved with the pathways. Currently many pathways are focused only on the treatment, but there’s great potential to improve the entire cancer care continuum for the patient from day one from evaluation and diagnosis, right on through the end of life—and that in fact is one of [ASCO’s] major recommendations.”
2. What are the concerns about the use of clinical pathways in oncology that prompted ASCO to form this Task Force and issue its recommendations?
“There was concern raised by many [ASCO] members that they were having to comply with multiple pathways in their practice—some oncologists reporting they had to adhere to eight or more pathways for exactly the same type and stage of cancer, with different payers putting forth different requirements. There’s quite a bit of redundancy in terms of the pathways out there and lack of transparency in the design and implementation of these tools.
“Also, an informal survey by ASCO of the Society’s state affiliates showed that up to 86 percent of the respondents using clinical oncology pathways said that they limit treatment options—and I personally would agree with that.
“Pathways need to respect that there is patient variability and autonomy; and to be 100 percent compliant with pathways is actually dangerous. Many times the patients that we are taking care of on a daily basis don’t look anything like the patients who are on the clinical research trials that ultimately led to the evidence that formed clinical guidelines and pathways. So there has to be allowance for variability.”
“We do believe at the end of the day, clinical pathways can be extremely valuable in oncology. But, we need to assure that these pathways are benefitting all the stakeholders involved in the cancer care continuum—the provider, the payer, and most importantly, the patient.”
3. Should there be more oversight or regulation of clinical pathways in oncology?
“That’s part of the issue now—that clinical pathways are being developed by companies, vendors, and even by physicians themselves, their provider groups, and individual institutions. And the problem is there is no standardized way in which pathways are being developed to ensure that they are encompassing the high-quality, high outcome benefit for patients, while reducing costs, as they are intended to do.
“More oversight or even going as far as having standards for clinical pathways in oncology is something that we, the Task Force, talked about as a possibility. And our nine recommendations [in the policy statement] are going to form the basis of work where we intend to consult with payers, vendors, and others who are developing pathways to ensure we are promoting high quality care, as the next step.”
Thursday, December 17, 2015
With BRIAN DRUKER, MD, Director of the Oregon Health and Science University Knight Cancer Institute
NEW YORK—“I want to be able to sit at a computer with the genome of a cancer patient and ask, ‘Are there any other patients like my patient around the world? And, what are their outcomes?’ I want a million cancer patients’ genomes,” said Brian Druker, MD, Director of the Oregon Health and Science University Knight Cancer Institute, speaking as part of a panel of experts from academia, industry, and advocacy during a session on the future of data sharing in medicine—both personal data and big data—here at the 2015 Partnering for Cures meeting (sponsored by FasterCures, a center of the Milken Institute).
Oncology is doing that—i.e. moving in the direction of making that type of data sharing possible, Druker explained. But there are several challenges still including shifting away from the tendency to hold on to data in an effort (or hope) of making a profit off of that data, as well as creating a secure system where patients’ privacy is guaranteed to be protected, he said. And the panel agreed other challenges for big and small data sharing include legislative barriers, realistic business models, and quality assurance
Along with Druker, who is also the JELD-Wen Chair of Leukemia Research at OHSU and a Howard Hughes Medical Institute Investigator, the panel included: Linda Avey, Co-founder and CEO of We Are Curious, Inc.; Stephen Friend, President of Sage Bionetworks; Ben Heywood, Co-founder and President of PatientsLikeMe; and Michael Milken, Chairman of the Milken Institute and Founder of FasterCures. The panel was moderated by Gillian Tett, U.S. Managing Editor and Columnist at the Financial Times.
The panel discussed the next steps in breaking down the silos across the full health care industry to allow more data sharing—big and small. And in an interview after the panel, Druker told OT more about what makes data sharing so interesting and about a new precision medicine analytics platform OHSU is working with Intel to develop.
1. What would you say was most interesting about this discussion today between academia, industry, and advocacy about data sharing?
“To me the most interesting idea is that we can turn medicine from a transaction-oriented to patient-oriented profession, where patients have more control of their data.
“We’re starting to see that change—patients want their data shared. And if we can get past this view that there’s money to be made from the data as opposed to patients can be helped by sharing data, I think that we can actually accelerate progress.”
2. During the panel, you mentioned that oncology data would be sharable by 2020?
“It’s happening. We’re working on a project with Intel now to allow more access to data—the Collaborative Cancer Cloud (CCC). Our goal is by 2020 to be able to have enough data that we can analyze a patient’s genome and then have a treatment regimen designed based on that data, and have enough data aggregated in this kind of system that we can do that.
“Will that be all the cancer data? I don’t know. But if we reach a tipping point where we have numerous large institutions sharing data, everyone’s going to want to get on board. Now, whether that’s 2020 or 2022 I wouldn’t say. But our goal is to progressively scale this over the next five years.”
3. What are the next steps to make this type of data sharing scalable?
“Data sharing is happening in a lot of different places, and we need to aggregate and integrate. Part of the issue is deciding how to standardize data shared across multiple institutions, ASCO’s CancerLinQ, and other databases. We need to be able to share data in a standardized manner.
“And this is going to start with cancer—but it can be expanded into cardiovascular disease, Alzheimer’s research, you name it.”
Thursday, December 10, 2015
With SAMUEL APARICIO, BM, BCh, PhD, FRCPath, of The University of British Columbia
SAN ANTONIO—Genomic research has opened a lot of doors when it comes to cancer research, but in a plenary lecture here at the 2015 San Antonio Breast Cancer Symposium, a molecular oncologist explained the research is actually just getting started.
“Drug combinations have been found, in a way, through trial and error. People have tested things and found what works. Those combinations have been arrived at empirically. But, it’s actually a clonal evolution taking place,” explained Samuel Aparicio, BM, BCh, PhD, FRCPath, Professor in the Department of Pathology and Laboratory Medicine at the University of British Columbia, who delivered the lecture “Clonal Dynamics and Breast Cancer Subtypes.”
“We’re trying to learn how to be predictive. We really want to be able to have a better diagnostic assessment and to better understand how to manage patients longitudinally. But, it’s still early days for clonal dynamics,” he said.
He explored several breast cancer examples during the lecture—and explained in an interview after the talk what the field of clonal dynamics is and how it will play a role in all cancer research.
1. To start, how would you define ‘clonal dynamics’?
“So, a clone is just cells that are related to each other by dissent from a unitary cell. And dynamics means that one group of constituents, or a tribe—think of a tumor as being composed of many tribes—might take over another tribe so that the other tribe diminishes. The tribe is a clonal group, and dynamics simply means these clonal groups are getting bigger or smaller.
“The fundamental idea behind the clonal theory of cancer is that cells with similar properties can diverge from one another. And we can use that theory to describe how some populations of cells exist.
“Clonal dynamics is a way of describing the notion that malignant cells don’t all behave the same way—that you can have some malignant cells outgrowing others.”
2. What does ‘clonal dynamics’ mean for breast and other cancer research—and ultimately, for treating patients with cancer?
“What we’re really trying to do is understand why drug resistance emerges and how it emerges, so that we can predict what resistance is going to emerge—and eventually what combinations of treatments you might use to suppress this phenomenon.
“There’s an analogy that’s quite useful from infectious diseases. HIV used to be an incurable, life-limiting disease, until people figured out that in order to beat the virus you had to use not one drug or two drugs, but in fact three drugs that suppress the tendency of the virus to evolve. If you take a patient who has an infection and treat that patient with an antibiotic, you only partially treat the infection. You get resistant bacteria.
“It’s the same process in cancer. Cancer evolves by mutation and clonal dynamics. It’s just that the science that’s underpinning our ability to understand this process in patients’ tumors is just now beginning to emerge. In the last ten years or so a lot of the genomics research has now given us a way to measure all of that genetic complexity that’s associated with cancers. We now have a way of identifying which ‘tribes’ are in there—and which ones are increasing or decreasing in proportion.
“So, it’s still early days in the field, but where we’re heading is that by making more accurate measures of constituent parts of the patient’s tumor, and following that dynamically over time, we may be able to anticipate resistance and even predict patterns of resistance in a way that we aren’t currently able to.
“And there are other important consequences of the fact that cancers have clone structure. First—and I think every practicing clinician will be aware that cancer is a very dynamic disease—treatment interventions that are identified at one point in a patient’s trajectory may become inappropriate later or indeed new things may become appropriate. And increasingly it will become necessary to properly manage patients by re-biopsying disease more frequently because the disease you’re dealing with at a relapse may very well be a completely different looking disease from the one you’re dealing with when the patient was first referred.
“And with respect to the diagnostics, it’s important to keep in mind that limited tissue samples may only sample a fraction of the clones in the tumor, so you could biopsy a small area of the tumor and arrive at one set of conclusions about its molecular status and biopsy the other end and reach a different conclusion—a challenge because a lot of molecular diagnostics are done off of a single biopsy. That may be okay for small tumors, but for larger tumors one has to worry about the fact that there could be internal heterogeneity inside the tumor that would lead to possibly contradictory diagnostic findings. And that’s all a consequence of the clonal structures of cancers.”
3. So, right now, what type of research in this field is currently happening and what’s next?
“We already know a bit about single agent drug resistance, and we know that with molecularly targeted agents—virtually every molecularly targeted agent that’s been used to date from hormone therapy all the way to targeted agents all the way to kinase inhibitors—eventually in some number of patients, resistance will emerge. And that resistance tends to emerge in the form of singular mutations that happen in the target genes themselves leading to the loss of activity of drug in those cells.
“So what naturally happens is you get a clonal dynamic—the cells containing those mutations amplify because they’ve got nothing to suppress them. And that drug isn’t effective anymore—that’s clonal dynamics in action.
“So, the more drugs we layer the more complicated the genetic interactions are—and that’s what we are beginning to understand. And that’s really where the field is going.”