Biomarker Discoveries Bring New Choices and Challenges

Click here to view as PDF.

The ability to predict how the course of cancer will likely play out in an individual patient based on genetic factors is already changing the way certain cancers are approached, but it will take unprecedented levels of cooperation between researchers, federal regulators, drug developers, patient groups and clinical oncologists to continue advances in the field.

The collective knowledge of practicing oncologists is becoming increasingly important to advances in biomarker discovery, according to Shawnmarie Mayrand-Chung, PhD, JD, Program Director, representing the National Institutes of Health to the Biomarker Consortium (www.biomarkersconsortium. org). Dr. Mayrand-Chung explained the purpose of the Biomarker Consortium, formally launched in 2006, at the recent Biomarker World Conference in Philadelphia.

“Doctors are a frontline link to patients,” Dr. Mayrand-Chung said. “Involving practicing physicians in the process of biomarker research will provide insight into their day-to-day experiences with patients and contribute to designing effective clinical trials.” If individual clinical oncologists report discoveries in their patients to the Consortium, for example, collective results can provide the basis of new discoveries. “Research is inherently slow,” Dr. Mayrand-Chung said. “If we can speed it up by aggregating the observations we make through clinical interactions, it just makes sense.”

How Reliable Are Biomarkers in Cancer Treatment and Drug Development?

The discovery of how likely a cancer treatment will work in people with certain types of cancer based on the genetic makeup of either the person or the tumor has already changed how certain forms of aggressive breast cancer are treated, but many oncologists remain skeptical whether such predictors will eventually be 100% accurate in all patients and in all cancers.

Dr. Scott Patterson, executive director of medical sciences for Amgen, is a biomarker believer. He cited the Food and Drug Administration’s (FDA) “Critical Path Initiative”, which encourages researchers to utilize cutting-edge scientific tools and discoveries like biomarkers to measure the progress of a disease, or the effects of a particular treatment, quicker and more accurately than the traditional “phase 1-to-phase 4” method does.

Dr. Patterson cites the example of the KRAS biomarker to demonstrate the validity of using other biomarkers for targeted drug development and cancer therapy. Mutation of the KRAS gene can lead to overexpression of epidermal growth factor receptors (EGFR) and continued tumor growth.

“Even though biologists have known about the presence of a mutant KRAS gene in the tumors of patients with pancreatic, colon, and lung cancers, there has been a long standing-debate over whether this is indicative of a poor prognosis,” he said. Skeptics site preclinical studies in mice and studies conducted with relatively small patient samples, which seemed to indicate some patients with KRAS mutations may respond to chemotherapeutic agents that target EGFRs.

Amgen examined its archive of tumor samples from patients enrolled in their phase II studies with metastatic colorectal cancer (mCRC) in an attempt to come to a more definitive conclusion about the KRAS mutation’s role and then went on to conduct studies in more than 400 patients with metatastic colorectal cancer (mCRC) in its pivotal trial to compare their response to the best supportive care only (in this study it was FOLFIRI) and FOLFIRI combined with the Amgen drug panitumumab (Vectibix).

“The best supportive care along with panitumumab gave people with the wild type (normal) KRAS gene the best chance of objective response,” Dr. Patterson said, citing data from the study. In patients with a wild-type KRAS gene, 34% achieved a stable disease state, 17% showed a partial response to the combination therapy, and 36% went on to progressive disease. In comparison, 70% of patients with mCRC and a confirmed KRAS mutation in the study saw their disease worsen and none had a partial response to therapy, although disease stabilized in 12% of the patients.

Dr. Patterson said while these findings are promising, the treatment of individual patients is best left in the hands of their oncologists and he hopes for the day when there is a clearer understanding by patients and physicians on how and when to collect and interpret cancer-related biomarkers like KRAS. Amgen is preparing to study this subject further, Dr. Patterson said, to “ultimately generate a companion diagnostic for first- and secondline therapies.”

The reality, according to Dr. Patterson, is the example of panitumumab and the KRAS genetic marker “will not be unique.” Indeed, he said, “oncologists on both sides of the Atlantic are hoping treatment can eventually be determined at primary diagnosis. That’s what I’m hearing, and that’s what we’re pushing for.”

Pharma Companies Find Biomarker Research to Be a Lonely Road

Recognizing the multitude of hurdles researchers and pharmaceutical companies face in bringing biomarkers from discovery to clinical practice, and with hundreds of drugs with oncology indications in various stages of development, the National Institutes of Health (NIH) has partnered with both public and private groups to form “The Biomarker’s Consortium.” With an executive committee consisting of members from the Pharmaceutical Research and Manufacturers of America (PhRMA), the Biotechnology Industry Association as well as the FDA, Centers for Medicare & Medicaid Services (CMS) and the Foundation for the National Institutes of Health, one of the main goals of the Consortium is to create a “precompetitive environment” where the groups can pool resources for the “discovery development and regulatory acceptance” of biomarkers important to the treatment of cancer and several other disease states.

“Pharmaceutical companies’ investment in biomarker research is a difficult business model,” Dr. Mayrand-Chung said. “Clinical trials are expensive and difficult, and require high risk and input on the company’s part, so if we can share the risk and expense re- quired for biomarker development and qualification with other pharmaceutical companies and NIH and FDA, it just makes sense.” While members of the Consortium agree there will be no “intentional generation of intellectual property” Dr. Mayrand-Chung said companies are expected and encouraged to use the publicly available precompetitive information from the Consortium in product development and commercialization.

The oncology steering committee is co-chaired by Anna D. Barker, PhD, deputy directory of the National Cancer Institute of the NIH and Barbara Weber, MD, vice president, translational medicine, genetics and oncology for GlaxoSmithKline. Other committee members include representatives from pharmaceutical companies, academic research centers, and patient advocacy groups like the American Cancer Society.

Currently, the oncology committee is studying the use of Fluorodeoxyglucose-Positron Emission Tomography (FDG-PET) as a biomarker for clinical trials conducted in non-Hodgkin’s lymphoma and non—small cell lung cancer. The goal is to examine whether FDGPET imaging can help access the effect of traditional cytotoxic drugs in clinical outcomes and survival in lung cancer and lymphoma. This, according to the Committee, could have an “enormous affect on patient management by validating a tool that can measure responses to treatments and enable more rapid drug development.”

The studies hope to determine how effective FDG-PET imaging is in measuring the effect of conventional cytotoxic drugs in patients with either lung cancer or lymphoma. The qualification of the FDG-PET as a biomarker could, according to the Committee, potentially impact patient care by providing a tool to measure treatment response earlier and perhaps lead to more rapid drug development. Findings could also affect the regulatory review process for these cancers and “provide a basis on which CMS could make reimbursement decisions for cancer treatments,” Dr.Mayrand-Chung said.

Biomarkers Used to Predict Safety, Toxicity, and Dosage in Patients With Certain Genetic Markers

The correlation between certain genetic markers and specific cancer therapies is, in some cases, becoming strong enough to warrant relabeling of drugs from the FDA to reflect whether they are likely to work in an individual patient or not.

As pointed out by Dr. Alan H.B. Wu, professor of laboratory medicine at the University of California in San Francisco, the FDA has already changed the labeling on the colorectal cancer drug irinotecan (Campostar) instructing oncologists to test their patients for the presence of a UGT1A1 enzyme mutation before initiating the treatment. While the number is small, patients with this type of enzymatic mutation have a high risk of severe toxic reaction to the drug.

Breast cancer treatment and prophylactic tamoxifen (Nolvadex), taken orally, may be the next cancer drug to be relabeled. Increasing numbers of studies indicate the drug shows increased efficacy in women and men with a wildtype CYP2D6 enzyme, although survival rates between those without the wildtype 2D6 and those with the mutation are unknown. If a 2D6 mutation is detected, an oncologist may instead consider treating the patient with aromatase inhibitors such as Zoladex or Lupron, especially in pre-menopausal patients.

The University of California in San Francisco, where Dr. Wu is based, is running a study where people with breast cancer being treated with tamoxifen are invited to join an educational “tamoxifen pharmacogenomics program.” Participants are given the option of being genotyped in a certified lab; clinicians tell the patient’s oncologist if their patient has genetic indicators of an aggressive form of the disease. So far, 65 women have voluntarily enrolled in the program. Among other things, the study is examining whether oncologists change their management of a patient’s disease if they know their patient has a high-risk genotype.

Dr. Wu thinks it is still premature for the vast majority of clinical oncologists to routinely genotype their patients. “I don’t recommend [oncologists] go out and genotype their patients without a plan on how they’re going to use the information,” he said. “The tests are expensive, they are not widely available, and there are no guidelines to interpret the results.” Ultimately, Dr. Wu said, “implementation will only be successful if testing leads to clinically actionable decisions.”

Patients Weigh in on ‘Personalized’ Medicine, With an Emphasis on ‘Quality of Life’ Factors

People with cancer and their advocates, according to at least one genomic research company, are beginning to add their voices to the biomarker discovery debate. In some cases, they are already influencing how this information is discovered and applied to their current cancer treatment options.

Patrick Terry, co-founder and director of industry relations for Genomic Health, Inc., a developer of genomic-based clinical laboratory services, said time spent in “engagement mode” with patients with breast cancer and their advocates helped his company design their flagship genomic test, Oncotype DX. Given to patients with early-stage breast cancer, the test is credited with predicting whether the disease is likely to metastasize within 10 years of initial diagnosis.

“What we heard from the breast cancer community is that uncertainty has an impact on quality of life,” Mr. Terry said, “so the patient community helped us to apply the technology to measure this. We discovered it was not so much a ‘targeted therapy’ question they wanted answered, but a ‘likelihood of recurrence’ question. If we can tell patients the likelihood of disease recurrence they can make a much more informed care decision.”

Patients then wanted to take the information gathering even further: could a test be developed to ascertain a “mid-range” risk of recurrence?

“This is a difficult and highly controversial area,” Mr. Terry said. “But patient advocates said they wanted to help patients who were deciding whether to treat—or not to treat—their conditions.” In response to this demand, Genomic Health is now conducting trials to find that out.

“Whether we can spare women chemotherapy is the universal question behind this trial.” Bucking the trend of slow clinical trial enrollment, Mr. Terry anticipates this trial, which will be open for two years, will be ahead of schedule because of immense patient interest.

Access to survival and quality-of-life information is changing the relationship between oncologists and their patients at a core level. “We’re seeing a wholesale shift where oncologists are ‘partnering’ with patients. Patients are demanding this type of information,” Mr. Terry said. “For the clinical oncologist, the dilemma is how to present this information so the patient can weigh [her] treatment options appropriately.”

However, patients and oncologists are not the only ones interested in their likelihood of surviving cancer or having to contend with recurrence somewhere down the road. Different ‘stakeholders’ may come to very different conclusions as to what is “appropriate” therapy based on a patient’s genetic makeup.

“There are some highly integrated care teams, along with the patient and the patient’s family, so you may have a highly interactive and dynamic situation,” Mr. Terry said. “Throw in the hospital administrator, insurance company and the lawyers and it complicates the mix even more.”

Mr. Terry predicted detailed conversations between oncologists and patients regarding which therapy to choose based on individual genetic findings may still be 10 to 20 years away, but said oncologists should start demanding rigorous data so they can have improved confidence in their decision making for patient care and who will be on the patient care team.

“Delegating these responsibilities to the cancer advocate community is another aspect of what clinicians can do,” he said. “There is the trauma of a cancer diagnosis, but there is also the empowerment of knowledge.”

Will the Pull Toward Biomarker Identification Push More People Out of Clinical Trials?

With more oncology drugs in the research pipeline than ever before, the need for clinical trial enrollment by both patients and control groups is reaching critical levels.

“Demand for patients is expected to increase, but problems in recruitment will cause delays in drug development,” said Geert Kolvenbag, MD, PhD, executive director of development for emerging oncology and infection brands at AstraZeneca. Although genotyping can mean the quicker development of more targeted therapies, the added step of preselecting groups on a molecular basis may exacerbate enrollment problems.

Some of the reasons for low clinical trial participation are well known: patients’ lack of awareness of clinical trials, failure of physicians to refer patients to clinical trials, and patients afraid of becoming “guinea pigs.” Other reasons are less obvious.

Many clinical trial enrollment and retention problems, Dr. Kolvenbag said, are on the professional side. “I think it’s disappointing a lot of investigators don’t recruit the number of patients they promise us. In recent phase III programs, we found about 20% of sites recruited 80% of the patients.” Bigger sites don’t necessarily mean better trial enrollment, either. “It’s not always how large the site it, but how well organized it is.”

Community oncologists and enrollment sites may be less inclined to recruit patients for trials because of monetary factors, Dr. Kolvenbag said. Many believe the time and effort invested in recruitment or even awareness building is not cost effective for their practice or their center.

Adding biomarker identification to the mix will presumably only make enrollment problems worse, with already hard-to-get patients loathe to agree to anything more than a blood draw. “On top of all the other factors, tissue collection may be an extra burden,” he said. “Language is also a problem: genetic testing is a loaded term. Placebo is also a loaded term, and patients may think getting a placebo means they’re ‘getting nothing.’ So we’re trying to find an educational process which explains what clinical trials are, how they work, and what the benefits are to patients.”

Today’s oncologists, with their direct access to patients, are becoming key partners in clinical trial enrollment. “We’re working on suggestions for appropriate messages and words doctors can use to communicate with their patients,” Dr. Kolvenbag said. “These newly targeted therapies are bringing us to the brink of a new era in cancer treatment. But it won’t happen by itself.”