Applying Molecular Profiling to Clinical Practice: Promises and Challenges 

Get Permission

A “new kind of pathology,” with anatomy and histology being supplemented by molecular etiology, has been emerging over the past decade and promises better response rates among patients with cancer, as genomic alterations continue to be identified and treated with targeted therapies.

“The list of genomic alterations in cancer that can be therapeutically targeted is growing rapidly,” according to Shridar Ganesan, MD, PhD, Associate Director of Translational Science at Rutgers Cancer Institute of New Jersey, in a presentation on molecular profiling at the recent Best of ASCO meeting in Chicago. The presentation incorporated information from education sessions presented at the 2013 ASCO Annual Meeting.1-5

Critical Targets

Using non–small cell lung cancer as an example, Dr. Ganesan noted that while there are a few histologic subtypes, molecular analysis reveals “a large, surprising hetereogeneity of different molecular driver events that are seen in histologically identical cancers.” These molecular driver events “are probably critical for the development of that cancer and are critical for their continued growth and proliferation, and so they are critical targets that are now being appropriately hit by small molecules,” Dr. Ganesan added.

Overall, the kinds of alterations that can be targeted in cancer cells include “classical activating oncogenes like ABL, ALT, RET, and JAK. But now we are also starting to be able to target loss of function in tumor suppressors like BRCA1, BRCA2, and PALB2. Loss of function of those genes leads to an underlying repair defect that can be targeted by certain classes of standard chemotherapy like the platinum salts as well as new small molecules like the PARP inhibitors,” Dr. Ganesan stated.

Impact of IMPACT Study

“Maybe the reason we have not gained as much as we would like from large, randomized trials based on histologically defined cancer subtypes is that common cancers are really caused by a large set of underlying genomic aberrations. Targeted agents and perhaps even standard chemotherapy work only in a subset of cancers that have the appropriate underlying sensitizing mutations,” Dr. Ganesan said.

The IMPACT (Initiative for Molecular Profiling in Advanced Cancer Therapy) study showed that response rates can increase when patients receive appropriate molecular therapy. Patients eligible for phase I trials at The University of Texas MD Anderson Cancer Center in Houston had tumor specimens analyzed for the presence of genomic alterations and were matched to appropriately targeted phase I trials if available, or to other phase I trials.

Of 2,282 patients with adequate tissue available to be analyzed, 52.2% had at least one molecular aberration. About 20% of these patients who were matched to appropriate molecular therapies had no progression for at least 6 months, including 2% with complete response and 17% with partial response by RECIST criteria.

Patients with matched therapy also had improved failure-free and overall survival. Patients who had no molecular match and were assigned to other trials had a 0% complete response rate and a 14% partial response rate.

“This suggests that if you have molecular targeted therapy, you can see a response signal very early, even in phase I data,” Dr. Ganesan noted.

These data include a response rate of 77% using imatinib (Gleevec) in patients with chronic myelogenous leukemia and a 57% response rate for crizotinib (Xalkori) in patients with non–small cell lung cancer and the EML4-ALK fusion biomarker.

Challenges Amid Advances

Along with an increasing number of molecular therapies, the technology for gene sequencing is rapidly improving. This has led to “the overall schema of personalized medicine we were hoping for,” Dr. Ganesan said. In this schema, “when a person has a cancer diagnosis, not only do you get your standard histologic diagnosis, but the tumor is also sent for multiplexed genomic sequencing, leading ultimately to a readout of what genomic alterations are present or absent in each patient. That would lead to a genomic profile for each patient, which hopefully will allow us to select targeted therapy.” Dr. Ganesan explained.

“Although this is very exciting, there are a lot of challenges or obstacles to implementing this kind of personalized oncology, and they include logistical problems as well as biologic challenges,” he noted.

“One of the big problems involves tissue collection and testing. There are no established guidelines for what makes an adequate tissue sample for molecular or genomic analysis and what exact platform and type of analysis should be done,” Dr. Ganesan said. “If we are really going to use this technology to dictate patient care, we need to carefully evaluate the positive and negative predictive value of each test,” he added. “Right now, we don’t know the limitations of many of these assays.”

Collaboration Needed

To overcome these problems, Dr. Ganesan suggested collaboration on guidelines, such as the collaborative efforts ASCO participated in to develop guidelines for HER2 testing in breast cancer. “Ultimately, we will have to develop some sort of standardized method for evaluating and validating these assays,” Dr. Ganesan said. “We need to have overall national guidelines.”

Another problem concerns determining which gene alterations “are truly actionable,” Dr. Ganesan noted. That determination should be based on the strength of the evidence that a genomic alteration is driving the growth of the cancer, and the size of the effect to be achieved by intervening.

“To establish that, we need to have standardized clinical decision support,” Dr. Ganesan said. As an example, he cited a computer search form that would allow the clinician to enter data on the molecular alterations and the tumor site of origin and then get current data on clinical implications and what agents are available to treat the patient.

Access to Drugs

“Another big problem, of course, is access to new targeted therapies,” Dr. Ganesan added. One of the proposed solutions is to create a national formulary of targeted agents against common aberrations along with a registry of administered treatment and patient outcomes.

A national pharmacy exchange could benefit patients who receive a targeted agent matched to their molecular profile, as well as physicians who receive guidance in treatment recommendations and access to drugs, drug manufacturers who receive data on drug use and outcomes, payers who receive data to inform future coverage decisions, and regulators who receive data on extent and outcomes of off-label drug and test use and real-world safety data.

Biologic Challenges

There are also biologic challenges to applying molecular profiling in clinical practice, Dr. Ganesan explained. These include variations in how mutations signal in different contexts, intratumor heterogeneity, and rapid emergence of resistance, compensating mutations, and pathway adaptation.

Among recommendations for preventing resistance are the following strategies: target drivers of genomic instability and clonal diversity, combine targeted therapy with immunotherapy and/or chemotherapy, conduct frequent tumor biopsies to monitor emergence of new aberrations, and eradicate significant subclones early. As with many aspects of life, timing is also important and treatment should begin “as early as possible before diversity explodes.”

Encourage Tissue Donation

In a question-and-answer session following Dr. Ganesan’s presentation, the meeting moderator, Douglas Yee, MD, of the University of Minnesota Masonic Cancer Center in Minneapolis, said that patients should be encouraged to donate tissues to be analyzed for genomic aberrations as part of “a movement to have better-annotated specimens available for research. In my practice, I think most patients would be willing to contribute,” Dr. Yee noted.

In an interview with The ASCO Post, Dr. Yee was asked about the National Institutes of Health’s recent decision to restrict its financing of research on the HeLa genome because of family concerns about privacy if the full genomic sequence of HeLa cells was published. “The HeLa cells issue is from a completely different era and a very egregious case of not informing the patient about what was being done to her, and I think there is no excuse for that,” Dr. Yee said. In the current situation, patients would be fully informed and understand that specimens would be used for research about clinical outcomes.

Patients with cancer often ask how they can contribute to cancer control efforts, Dr. Yee noted. “When we talk about how to help, we always talk about advocacy and money and donations, but I think there is another component where patients can help by making their tissues available for study,” he said. ■

Disclosure: Drs. Ganesan and Yee reported no potential conflicts of interest.


1. Kurzrock R: Applications of profiling to the clinical setting. 2013 ASCO Annual Meeting Education Session. Presented June 3, 2013.

2. Schilsky RL: Removing barriers to personalized cancer care. 2013 ASCO Annual Meeting Education Session. Presented June 1, 2013.

3. Tsimberidou AM: Rewards and challenges of individualized therapy in clinical trials. 2013 ASCO Annual Meeting Education Session. Preseneted June 1, 2013.

4. McShane LM: Designing prospective trials in the era of molecular profiling. 2013 ASCO Annual Meeting Education Session. Presented June 3, 2013.

5. Swanton C: Heterogeneity of tumors and how to approach it. 2013 ASCO Annual Meeting Education Session. Presented June 3, 2013.