Emerging Role for Biomarkers and Implications for Care

Kathy Boltz, PhD

Biomarkers and genomics were a major theme at the American Society of Clinical Oncology (ASCO) 2015 annual meeting, and the quest to find the right biomarker to match a patient with a therapy that has the best chance for success is on.

Announced at this year’s meeting, the clinical benefit from the PD-1 immune checkpoint inhibitor pembrolizumab was found to be predicted by mismatch repair (MMR) deficient status in patients with colorectal cancer.1 Data from the clinical trial showed patients having MMR deficient status led to an immune-related objective response rate (IR-ORR) of 40% (4 of 10 patients) and immune-related progression-free survival (IR-PFS) of 78% (7 of 9 patients). In contrast, patients with MMR proficient colorectal cancers had IR-ORR rates of 0% (0 of 18 patient) and IR-PFS of 11% (2 of 18 patients).

Basket trials—the sorting together of different kinds of cancers with the same mutation into the same treatment “basket”—are seeking to use biomarkers to choose targeted therapies, rather than disease site. ASCO and the National Cancer Institute announced two upcoming clinical studies that aim to expand the boundaries of precision medicine. One trial, TAPUR, will explore available targeted cancer drugs in a range of cancer types with a genomic variation known to be a drug target. The other trial, NCI-MATCH, will use over 20 different drugs or drug combinations, each targeting a specific gene mutation, to match patients with a therapy that targets a molecular abnormality in their tumor.

Finding Clinical Utility Through Biomarkers

“The one thing we don't want to do in oncology is to take a drug that can help some people and lose its activity in the noise,” said Richard Goldberg, MD, Physician-in-Chief, The James Cancer Center, The Ohio State University. “That can happen if you have a small subset of patients whose tumors are helped by that particular targeted therapy getting lost in a big denominator of people whose tumors have no chance of benefiting.“

“So, as we go forward, some of these studies where we're genotyping large numbers of patients, and then assigning them to various trials based on their abnormalities, is a very interesting paradigm shift from where we've been in the past,” said Goldberg.

“As the cost of sequencing has rapidly decreased over the past few years, from an economic and clinical perspective, it will make increasing sense to perform broader, clinical-grade, broad-based sequencing analysis,” said Robert Penny, MD, PhD, chief executive officer of Paradigm in an interview at ASCO. “The ability to interrogate a wide range of clinically actionable alterations enables both physicians and patients access not only to targeted therapies, but also enrolling in clinical trials using targeted agents.

Goldberg explained that he generally uses genomic sequencing for his patients that have advanced disease who are getting near the end of standard therapy.

“What I'm looking for is to try and find a phase one or a phase two trial for them that is a shot in the light rather than a shot in the dark. What I mean is not just a random trial that we put them on, but a trial that we're choosing based on the abnormalities that seem to be driving their tumors,” said Goldberg.

“If we can do that, I think the odds of a patient responding to a clinical trial are higher than the typically quoted 5% for phase one studies. That’s the big place for this to be applicable,” said Goldberg.

Goldberg’s approach is common, according to oncologists who participated in a survey at ASCO conducted by ZS Associates and M3 Global Research. Pan-cancer genomic testing is most commonly used at late stage or metastasis, according to the 42 oncologists who responded to this question. It is a “hope” to find something to help prolong life, rather than a standard upfront protocol to guide treatment. Among the 95 oncologists who described how frequently they used pan-cancer genomic testing, only 12% said that they always use pan-cancer testing, while 33% said often and 37% said sometimes.


 “It’s encouraging to see that nearly half (48%) of respondents claimed to test at diagnosis or early-stage disease,” said Jon Roffman, Principal of ZS Associates, a global firm focused on improving business performance through sales and marketing solutions. “Still, 52% are reserving pan-cancer genomic testing only for later stage patients who are running out of hope. There are several barriers that need to be removed to get more patients tested earlier to maximize the impact of genomic testing.”

The biggest barriers to genomic testing today are actionability and reimbursement. Ability to act on the test results ranked among the top barriers by 61% of oncologists, underscoring the need for more therapy options that will work against the newly understood genomic profile. Reimbursement related to the biomarker (ranked number 1 by 22% of the 95 oncologists responding) is also a major hurdle. Roffman commented that these survey results confirm the importance of adapting the treatment and reimbursement approach as our understanding of cancer evolves.

Questions of Payer Coverage

Molecular testing is now typically covered by insurance plans, explained David Spetzler, PhD, chief scientific officer of Caris Life Sciences in an interview at ASCO. He stated that payers are willing to cover testing for molecular information known to be relevant to a patient’s disease state, such as BRCA1/2 in breast cancer.

On the other hand, when it comes to cases like an abnormality in a patient’s tumor that is covered for melanoma but not for colon cancer, Goldberg described it as an uphill battle. “We’re going to have to confront that with the insurance industry to decide if we just allow patients to go on very low likelihood third, or fourth, or fifth line treatment, or do we let their genomics guide the way we treat them?”

“I hope we will let the genomics guide the way we treat people,” said Goldberg. When asked about how insurance companies and payers will get to the point of paying based on genomics rather than the site of the tumor, Goldberg responded, “It’s going to be a wrestling match.”

Over half of US oncologists surveyed (57%; n=95) at ASCO strongly agreed that genomic testing will become a part of standard treatment protocols in 5 years for most cancers. The same percentage (57%) strongly agreed that health insurance must reimburse genomic testing as a critical part of the treatment approach.


These physicians agreed that testing is on the rise across the board for a range of cancer types. The highest rates of testing and the most expected increases in testing were for non-small cell lung cancer, breast cancer, prostate cancer, and melanoma.


When asked what it will take to get oncologists to further integrate biomarker testing into their practice, Roffman said, “It will take treatment options that make the tests more actionable for a broad set of patients, and also more consistent reimbursement. There’s a bit of chicken and egg here. As more and more therapies get approved in niche subpopulations, we’ll eventually reach a tipping point where it makes sense to just do the test right away for all patients.”

The most popular biomarker to test for was EGFR, according to 35% of the 95 oncologists surveyed. The next most actionable were HER2 (ranked #1 by 23% and #2 by 21%), ALK (14%), and ER/PR (9%). These results highlight the fact that doctors view those biomarkers that already have FDA-approved therapeutic options to be the most actionable. PD-1/PD-L1 was considered most actionable by 11% of oncologists surveyed, and this may increase since new data was shared at ASCO.


Next Steps in Biomarkers

Biomarkers are continuing to move even beyond next generation sequencing of DNA toward expression information through data on RNA and protein. Yet dealing with increasing amounts of information and finding a meaningful signal through the noise is an increasing concern.

“We are already at the limit of gene panels now,” said Neil M. Barth MD, FACP, Chief Medical Officer, Agendia Inc in an interview at ASCO. "The current panels have 50 to 250 genes, the top 200 genes are identified, and these panels are giving us more variance than we can use."

The idea that the data are getting overwhelming was emphasized by Eliezer Van Allen, MD, Dana-Farber Cancer Institute, in his ASCO session on informatic strategies to turn sequence data into cancer patient management. Van Allen explained that the transition from hotspot genotyping, with about 100 data points per patient, to transcriptome data with whole-exome and whole-genome sequencing, has brought us to well over 100,000 data points per patient.

Barth emphasized the importance of looking beyond DNA and including RNA to understand gene expression. He explained that while identifying DNA mutations in BRCA 1/2 is the gold standard, RNA-seq expands the treatable patient population by about 30%. Though standards have not yet been developed for RNA-seq, also known as whole transcriptome shotgun sequencing, it is being done now, stated Barth.

An analysis of the I-SPY 2 trial, presented at ASCO 2015, found that the absence of a mutation does not indicate an inability to identify patients by expression, noted Barth. This exploratory analysis looked at 2 different gene signatures, PARPi7 and BRCA-like, among breast cancer patients who were HR+/HER2 negative.2 Over a third (38%) of these patients were DNA repair deficient, indicating they would likely benefit from PARP inhibitor therapy. Their benefit would be similar to that of patients with the BRCA1 or BRCA2 mutations, but these further patients would not be identified by only analyzing gene mutations. This suggests the value of this further analysis for identifying patients for targeted therapies.

Similarly, Penny explained that DNA sequencing alone will not provide all the information required to develop a refined precision approach to all therapies. He stated, “We must look to additional important information at the RNA, protein, and immunologic level.”

Interest at ASCO also was high in liquid biopsies, such as using circulating tumor DNA to monitor and profile clonal evolution and response to therapy. Spetzler mentioned that Caris offers this, along with a range of analysis techniques.

The next step may be examining circulating microvesicles, which Spetzler explained are detectable much earlier in the blood plasma than RNA. Circulating microvesicles are shed by the tumor and immune system, have a role in immune suppression, and occur at concentrations of 1010 to 1012 microvesicles per ml of blood plasma.

All Rare Diseases?

“Every single patient actually has a rare disease,” said Spetzler. Analysis must consider not just the mutations present but also the clonal subsets present.

Barth stated that in many respects, each disease in cancer does become an orphan disease. Penny felt that it depends.

“While it is true that technologies have enabled us to more effectively classify tumors at the genomic level for treatment, leading to more niche indications (e.g., ALK+ in NSCLC and BRAF V600E+ in melanoma), recent data presented in the immune oncology space point to the potential for the rebirth of blockbuster therapies with significant clinical impact across a range of tumor types.  However, even here, given the cost of therapies, it will be important to understand the biology of the responders and non-responders,” said Penny.

Some biomarker companies are making a play for whole exome sequencing, Barth said. He explained that further information can be found through examining epigenetics, such as methylation.

For immuno-oncology, RNA-seq is more helpful than DNA. Barth stated that gene expression pathways carry more weight than mutational analysis in immuno-oncology. As the number of mutations increases, immunogenicity increases and neo-antigenic expression is picked up by RNA-seq. While RNA-seq is still a few years away, Barth stated that it is coming.

Barth emphasized that we are scratching the surface at the DNA level but not yet at the RNA or protein level. While we are finding mutations for every tumor, it is still a challenge to link those to actionable targets.

“Today, it is probably an 80/20, where 80% of the focus should be on clinical discovery and 20% on actionability with approved therapies,” said Roffman. “I hope that ratio will be more like 60/40 in 5 years, where we’ll have more approved options to tailor therapy for cancer patients, while still having a lot of strong development of new targeted agents.”

Meet the Expert

Jon Roffman is a Principal with ZS in the Boston office and the leader of ZS’ Oncology Field Strategy practice within ZS’ Oncology vertical. He is focused on addressing sales and marketing issues for clients in the oncology marketplace, with an emphasis in launch planning, field force strategy and execution.

Jon has significant experience with oncology field teams including sales force, medical, payer and key account teams. Jon has also led many initiatives to design innovative customer engagement and deployment models in Oncology. Jon has advised more than a dozen Oncology companies on sales and marketing strategy issues. ranging from small biotechs launching their first product to large pharma companies building out their portfolios.

Prior to joining ZS in 2005, Jon earned a B.S. in Electrical Engineering from Cornell University and an MBA from New York University’s Stern School of Business. 

About the Contributors

ZS is the world’s largest firm focused exclusively on improving business performance through sales and marketing solutions, from customer insights and strategy to analytics, operations and technology. More than 3,000 ZS professionals in 21 offices worldwide draw on deep industry and domain expertise to deliver impact where it matters for clients across multiple industries. To learn more, visit www.zs.com or follow us on Twitter (@ZSAssociates) and LinkedIn.

M3 Global Research provides the most comprehensive and highest quality market research recruitment and support services available to the global healthcare industry with relationships with over 2.5 million physicians in more than 70 countries worldwide. M3 Global Research operates an ISO 26362 certified panel with the highest quality data collection and project management capabilities that cover the spectrum of quantitative and qualitative techniques utilized today.  In addition to market research, M3 Global Research provides valued services to today’s medical professionals that include medical education, ethical drug promotion, clinical development, job recruitment, and clinic appointment services.  M3 has offices in Tokyo, Washington D.C., Fort Washington, PA, Rochester, NY, San Mateo, CA, Oxford, London, and Seoul.


1. Le DT et al. PD-1 blockage in tumors with mismatch-repair deficiency. New England Journal of Medicne. 2015; doi:10.1056/NEJMoa1500596.

2. van’t Veer L. et al. DNA repair deficiency biomarkers identify HR+/HER2- breast cancer patients who may benefit from veliparib/carboplatin: results from I-SPY 2 trial. ASCO 2015 abstract 521.

Reprinted with permission from OBR at OBROncology.com