This article is more than 5 years old. Information may no longer be current.
Cancer research, treatment move toward individualized care
Genomics plays an increasingly critical role in the clinic.
Click Here to Manage Email Alerts
The gap between basic science and clinical outcomes in cancer therapy is shrinking rapidly as a result of increasing research into the genetic makeup of tumors, cancer patients and those predisposed to developing cancers. The outcome of this research is ongoing, dramatic changes in the notion of what constitutes individualized care.
A patient’s age was once considered the key factor in making personalized clinical decisions. Comorbidities that accompanied age were considered, as were the history and severity of the disease, drug tolerability issues, patient preferences and family history.
All of those factors continue to play a role in individualized care and factor into treatment outcomes. However, a debate on the quality of outcomes has become another component to the discussion. There are those who argue unequivocally that OS is the only standard by which cancer treatment should be measured, and those who believe other endpoints should be weighed equally.
|
Photo courtesy of Memorial Sloan-Kettering Cancer Center |
One point that is generally agreed upon is the need for better prevention and diagnostics. Increasing research on what is happening genetically at or before diagnosis may yield even greater insight into the pathology of all cancers as they pertain to the human genome.
“This is a terrific and important subject right now,” Martin R. Weiser, MD, associate attending surgeon at Memorial Sloan-Kettering Cancer Center, told HemOnc Today. “Not all diseases are the same. We need to understand them at the molecular level.”
Weiser said age, comorbidities, lifestyle and, in his particular case, bowel function goals — he specializes in sphincter preservation procedures — are critical elements to personalized care in the clinic, but as new information continues to emerge, everything from national guidelines to patient needs will evolve.
HemOnc Today spoke to a cross-section of experts about the ongoing evolution of personalized care. Most agreed that cutting-edge genomics is the area that needs to be explored most thoroughly.
BATTLE studies
John V. Heymach, MD, PhD, associate professor in the department of thoracic/head and neck medical oncology in the division of cancer medicine at The University of Texas MD Anderson Cancer Center, presented data on behalf of researchers from the Biomarker-integrated Approaches of Targeted Therapy for Lung Cancer Elimination (BATTLE) trial at the 102nd Annual Meeting of the American Association for Cancer Research.
The trial investigated gene expression signatures that may predict outcomes and tumor mutations in patients with refractory non–small cell lung cancer. The specific focus involved developing novel markers in 101 patients from the BATTLE cohort who bore wild-type epidermal growth factor receptor disease who may benefit from erlotinib (Tarceva, OSI Pharmaceuticals).
The researchers said the EGFR and KRAS pathways and epithelial-to-mesenchymal transition (EMT) have been linked to response and resistance to EGFR inhibitors. Gene signatures for these pathways were developed and tested using independent data from resected NSCLC patients and cell lines. Mutation signatures for both pathways were compared with those from three different institutions and validated in vivo and in vitro.
Another signature from the EMT genes (E-cadherin, vimentin, N-cadherin, FN-1) — those which the researchers said had bimodal distribution — was derived using 54 NSCLC cell lines. This signature also was validated independently.
The researchers derived a novel five-gene signature in erlotinib-treated patients and divided them based on 8-week disease control levels. The aim was to determine whether the signatures could predict disease control and tumor mutations.
Results indicated that the EGFR signature predicted EGFR mutations (area under the curve [AUC]=0.72 by receiver operating characteristic [ROC] analysis; P=.03) and the KRAS signature predicted KRAS mutations (AUC=0.67; P=.01).
Among patients with wild-type EGFR and KRAS mutations, the EMT and five-gene signatures were linked to improved disease control in patients treated with erlotinib (EMT signature: 64% for epithelial vs. 10% mesenchymal groups; P=.02; five-gene: 83% vs. 0%, P<.001). The EMT and five-gene signatures also were associated with improved PFS in these patients.
The EGFR or KRAS signatures were not associated with improved disease control in patients with wild-type EGFR and KRAS mutations who were treated with erlotinib.
Significant links between in vitro sensitivity to erlotinib in NSCLC cell lines and the EGFR, EMT and five-gene signatures also were observed. A section of the five-gene signature, known as LCN2/NGAL, was linked to epithelial phenotype. Analysis of the EMT signature yielded insight into potential therapeutic targets associated with mesenchymal phenotype, including Axl, according to the results.
“Non–small cell lung cancer is a leading killer, but we have a limited ability to match the right drug with the right patient,” Heymach said. “The best information we have now is in patients with the EGFR mutation.”
Heymach said tumors driven by the EGFR pathway have benefited from targeted therapy. “Patients without this mutation can benefit from these drugs, as well, but we don’t know which patients will or will not benefit,” he said. “We are attempting to develop markers for the remainder of these patients, those with wild-type or non-mutated EGFR.”
The study is unique in that all patients had a tumor biopsy that allowed clinicians to profile their disease before treatment, according to Heymach.
“The adaptive randomized design can help us match patients with drugs more quickly,” he said. “The result is that we came away with two strong candidate biomarkers.”
Prospective biopsies
Another analysis from the BATTLE cohort was a prospective, biopsy-mandated, biomarker-based, adaptively randomized study involving 255 pretreated patients with NSCLC who were randomly assigned one of the following treatments: erlotinib, vandetanib (AstraZeneca), erlotinib plus bexarotene (Targretin, Eisai), or sorafenib (Nexavar, Bayer). Treatments were administered based on relevant biomarkers taken from fresh core needle biopsy samples.
The final analysis involved 244 patients. The 8-week disease control rate in this final group was 46%. These patients achieved a median PFS of 1.9 months (95% CI, 1.8-2.4), a median OS of 8.8 months (95% CI, 6.3-10.6) and a 1-year survival rate of 35%.
Results of an 8-week landmark analysis indicated that the median survival was 9.6 months (95% CI, 7.4-12.5) in patients with 8-week disease control and 7.5 months (95% CI, 4.2-9.2) in patients without 8-week disease control (P=.018).
Erlotinib was associated with an overall 8-week disease control rate of 34%. This rate was 33% for vandetanib, 50% for erlotinib plus bexarotene and 58% for sorafenib.
The researchers defined 0.8 posterior probability of exceeding a disease control rate of 30% as measures of effective treatment marker group pairings. Results of that analysis indicated that erlotinib was effective in the VEGF/VEGF receptor 2 group; vandetanib was effective in the EGFR group; erlotinib plus bexarotene was effective in the EGFR, retinoid X receptor/Cyclin D1 and no-marker groups; and sorafenib was effective in the KRAS/BRAF, VEGF/VEGFR-2 and no-marker groups.
Edward S. Kim, MD, associate professor of medicine, chief of the section of head and neck medical oncology and director of clinical research operations in the department of thoracic/head and neck medical oncology at MD Anderson Cancer Center, said the critical element of the BATTLE trial was that tissue was collected prospectively, rather than after treatment was completed.
|
|
“We tested patients who had previously received chemotherapy but who were not doing well for one reason or another,” Kim said. “We mandated a biopsy. Not a diagnostic biopsy, but one for the specific purpose of identifying molecular abnormalities, done in real time to help determine the best therapy.”
The aim was feasibility, as a study of this kind previously had not been conducted in lung cancer. He said phase 2 or 3 trials sometimes do not meet goals, and often, only a small percentage of tissue is collected after the fact.
“Our patients were willing to be re-biopsied,” Kim said. “Essentially, we had a great deal of extra core tissue that allowed us to test certain markers. This allows for the study to have a strong discovery component. We are developing new technology and new markers all the time, so this allows us to see how the actual tumors respond to those advances.”
Beyond genetics
Given the complex algorithm of treatment toxicity, tumor profile and life expectancy, age was traditionally the most important talking point when the clinician, the patient and the patient’s family were outlining treatment goals. However, many clinicians said the age-based approach is declining in utility.
“Age is key, but even with that, there are gray areas,” Weiser said. “For example, there are ‘young’ 80-year-olds and ‘old’ 80-year-olds. So, again, it comes down to patient preference.”
Nicholas J. Petrelli, MD, Bank of America Endowed Medical Director of Christiana Care’s Helen F. Graham Cancer Center and professor of surgery at Thomas Jefferson University, put it succinctly. “Nothing supplants the discussion between doctor and patient,” he said.
Petrelli said the overwhelming amount of information that is available online can confuse patients and their loved ones, adding that one of the duties of clinicians is to ensure those in their care understand the right information and where to find it.
“Of course we are familiar with the research, and we want our patients to be familiar with it also,” Petrelli said. “However, national guidelines and clinical trials are based on thousands of patients. When a patient walks into our facility, we are only dealing with that one patient. It is up to us to take into account everything from their job to their personality, their anxiety level to the nature of their tumor.”
Weiser highlighted the personal nature of most cancers, particularly his field. “We can save the sphincter in many patients,” he said. “But in others we cannot. This is a decision that can impact the life of an individual in a very personal way. Some patients are eager to avoid colostomy, others are not.”
The decisions Weiser’s patients face come down to a basic argument of quantity vs. quality of life. He said these definitions can vary from patient to patient, and for any given disease or therapy, the benchmark for the standard of care can be based on data regarding time to progression, PFS, quality of life, OS or some combination thereof.
|
|
According to Petrelli, endpoints have diversity, but the standard should continue to be set higher. “The only way we are going to make a dent in this disease is to improve OS,” he said. “However, you do not want to extend OS and have them in a hospital the whole time. That is no way to live. We have to improve, especially in advanced stages. Treatments that improve OS by 2 to 3 weeks do not cut it.”
Most clinicians would agree that OS is the benchmark outcome for any treatment method, but there has been much debate about the role of other outcomes. Although past and current generations of drugs have saved and extended the lives of cancer patients for decades, a growing number of experts in the oncology community say other solutions are necessary.
Back to the lab
Jeff Boyd, PhD, vice president of Molecular Medicine, Robert C. Young, MD, Chair in Cancer Research, executive director of the Institute for Personalized Medicine, chief in the division of molecular pathology and professor and leader of the Women’s Cancer Program at Fox Chase Cancer Center, said, “Cytotoxic chemotherapy and radiation exposure have had a limited impact on the human cancer problem.”
Fox Chase is one of the few places in world that will be sequencing whole exome of tumors, according to Boyd. “More importantly, we are changing from a research and development facility to a clinical care model,” Boyd said. “We are using sequencing information to impact the therapeutic approach to treating patients.”
The focus of the approach will be to identify mutations in the gene that are susceptible to drugs, using compounds that carry fewer adverse events than chemotherapy and radiation. “We are looking for all manner of prototypes,” Boyd said. “The mutations themselves are druggable, as are the pathways. These can be turned off by inhibitors that are far less harmful than current therapies.”
Boyd said a strong argument against this approach is the cost. “Right now, it is between $10,000 and $15,000 per tumor analysis, but we expect that to decrease to around $1,000 within a few years,” he said. “Despite this, we are seeing that patients are willing to pay for the sequencing — and may continue to be so — even on a cash basis.”
Petrelli said, on a macro level, the cost argument against such high-level personalized medicine is largely invalid. “How much do we spend on any given cancer?” he said, noting that there are costs associated with his specialty.
“We spend so much money on colorectal cancer, wouldn’t it be nice to prevent it?” he said. “You can’t put a price on prevention. I can’t think of a single, intelligent reason not to spend on prevention.”
Boyd said he is hopeful for the future of his field. “In 10 years, this will be the standard of care at academic centers,” he said. “In 15 years, it will be the standard of care at community centers.”
|
|
Clinical guidelines
The backdrop of this discussion is the ongoing evolution of American Society of Clinical Oncology and National Comprehensive Cancer Network (NCCN) clinical care guidelines. Although it is understood by clinicians that these guidelines are simply a road map for care that is to be interpreted openly, patients may believe that they are more concrete.
Kim brought the question back to outcomes. “There are endpoints that we are trying to achieve with our patients, and then there are regulatory endpoints recognized by the FDA,” he said.
Kim said it is the clinician’s job to reconcile differences that may exist between these two goals. Moreover, even in light of data from thousands of patients reviewed by the decision-making bodies, uncertainty remains.
“What does it take to get a drug indicated by the FDA?” he said. “What is beneficial to the patient? If the tumor shrinks, it is beneficial. But sometimes the FDA disagrees if the patient does not live longer. If you can delay progression, you may have succeeded. The FDA struggles with these endpoints.”
Of particular interest to many clinicians and researchers is the increasing presence of genetic information in the ASCO and NCCN guidelines. The BRAF, EGFR, HER-2 and KRAS mutations have all been recognized on some level, and with the number of genome-based trials under way, that list is likely to increase.
Looking ahead
Although the research is heading in the direction of the genome, barriers to conducting these trials exist. In a paper in the Journal of Clinical Oncology, Laura MacConaill, PhD, senior scientist at Harvard Medical School and scientific director at the Center for Cancer Genome Discovery, and
Levi A. Garraway, MD, PhD, of the department of medical oncology at the Dana-Farber Cancer Institute, wrote that, “The impetus for widespread tumor genomic profiling could impose significant new infrastructural demands on leading academic cancer centers.”
They said implementation of genomic profiling could affect patient informed consent standards, as well as laboratory, pathologic and surgical procedures. “Universal standards and approaches to educate physicians regarding the appropriate interpretation and use of this information must be developed,” they wrote. “Complex issues surrounding reimbursement for both the diagnostic approaches and the resulting therapeutic implications must be addressed, especially in cases in which tumor genomic information might suggest off-label treatment avenues.”
Despite these obstacles, MacConaill and Garraway said the genomic view of cancer will forever shift the clinical oncology paradigm. “The development of advanced diagnostic tests capable of reading multifaceted genomic information in an efficient and cost-effective manner should prove highly enabling for individualized cancer treatment,” they wrote. “It is likely that clinical tumor mutation profiling will proceed in phases, with the initial efforts focused on a framework for interrogating the low-hanging fruit of actionable mutations and with subsequent efforts concentrated on maximizing the informative tumor genomic information that can be gleaned from tumor material in a cost-effective manner.”
MacConaill and Garraway wrote that the pace of this innovation is moving rapidly enough that complete cancer genome characterization may be feasible on a large scale.
“Massively parallel sequencing represents a promising technology for comprehensive tumor genomic profiling,” they wrote. “Personalized cancer genomic information, once widely available and thoughtfully implemented, holds considerable promise for improving the lives of many patients with cancer.” – by Rob Volansky
For more information:
- Kim ES. Cancer Discovery. 2011;doi:10.1158/2159-8274.CD-10-0010.
- Macconaill LE. J Clin Oncol. 2010;28:5219-5228.
Disclosures: Drs. Boyd, MacConaill, Petrelli and Weiser report no financial disclosures. Dr. Garraway reports consultant or advisory work with Novartis and Foundation Medicine; stock ownership in Foundation Medicine; and research funding from Novartis. Dr. Heymach reports research support from AstraZeneca and Bayer, and being on the advisory boards of AstraZeneca, Genentech and Bayer. Dr. Kim reports receiving commercial research support from AstraZeneca and Genentech and served as a consultant for AstraZeneca, Bayer and Genentech.
What are the physician-level implications of the move toward genomic-based individualized care?
The first thing to consider is the patient standpoint. Some feel quantity of life is important, and others say quality of life is important. These approaches can optimize patient care in terms of efficacy and toxicity, and that is where we are going.
|
|
The larger question is what do we as a society think is important. As far as I can tell, the FDA has not spent too much time asking what patients want. Moreover, we do not necessarily know what the FDA thinks. They seemed to be in favor of PFS, but now that may be changing.
There has to be room for all endpoints. If the drug improves survival, there should be room for it, but there is also a place for drugs that prolong the time to progression. What we need to do is develop the wisdom to use each of these therapies appropriately.
At the moment, we are probably not using drugs as effectively as we can, and every clinician needs to recognize this on a fundamental level, even before we consider what is required to adopt the next generation of treatment.
We are clearly heading in the right direction with regard to the genomic-based approach, but studies to prove it have not been done, and likely will never be done. For starters, they are expensive. The other point is that to do an optimal trial, it would require many patients, particularly in the control arm, and advocacy or PR people from the pharmaceutical companies make this very difficult. They want people on the currently approved drugs.
Jimmy Hwang, MD, is a gastrointestinal specialist at Georgetown Lombardi Comprehensive Cancer Center. Disclosure: Dr. Hwang serves as a consultant for and a member of the speaker bureau for Roche/Genentech.
We realize that there is a lot of information for clinicians to keep up on, so the Vanderbilt-Ingram Cancer Center launched a website (www.MyCancerGenome.org or www.vicc.org/mycancergenome) to enable a genetically informed approach to cancer medicine.
The website provides information on what mutations make cancers grow and what the implications are in those cancers, and we identify clinical trial information for patients according to specific mutation.
We think this approach is clinically promising because of an increasing number of clinical advances. For example, EGFR mutations in lung cancer are linked to sensitivity to EGFR tyrosine kinase inhibitors, ALK fusions in lung cancer are linked to sensitivity to ALK inhibitors, and BRAF mutations in melanoma are linked to sensitivity to BRAF inhibitors. But, some mutations in these genes are associated with increased drug sensitivity, while others are associated with decreased sensitivity. That is, not all mutations are the same just because they occur in the same gene. Beyond these mutations, the website provides information on other potential mutations to be targeted for which drugs are being developed.
|
|
We hope that MyCancerGenome can help facilitate and accelerate accrual on clinical trials with targeted agents as well as accelerate adoption of this kind of approach to cancer medicine.
As for when this approach will become the standard of care, the onus is on us in the research community to demonstrate the benefits of this kind of approach compared to the traditional way of just using tumor histology to prioritize treatment without molecular information.
We realize this kind of approach is still going to require a culture change in cancer medicine. One component of that is simply getting tumor samples and processing them quickly. When we do our tumor genotyping, we can do tests within 8 days of getting sample, but sometimes getting the sample from the outside hospital can take up to a month, which is unacceptable. Until we can make that more efficient, it is going to be difficult to implement these approaches on a large scale.
William Pao, MD, is an associate professor of Medicine and Ingram Associate Professor of Cancer Research Director, Personalized Cancer Medicine, at Vanderbilt-Ingram Cancer Center. Disclosure: Dr. Pao has consulted for Molecular MD, AstraZeneca, Bristol-Myers Squibb and Symphony Evolution, and has received research funding from Enzon, Xcovery and AstraZeneca.