Submit a Comment

Challenges Persist in Personalized Medicine

STOCKHOLM, SWEDEN –One of the goals of personalized medicine is that genomic sequencing of a patient’s tumor may identify actionable/druggable targets, and therapies aimed at those targets will improve outcomes and possibly cure for cancers. There have been some spectacular successes in personalized medicine – most notably Herceptin for HER2-positive breast cancer and Gleevec for chronic myelogenous leukemia and gastrointestinal stromal tumor (GIST).

But, according to Gordon Mills, MD, Institute for Personalized Cancer Therapy, M.D. Anderson Cancer Center, Houston TX, “The number of successes, and in most cases these are only found in small subpopulations of patients and are transient, are far outnumbered by spectacular failures…. We are far from overcoming the hurdles associated with the implementation of personalized cancer therapy.”

At the recently held European Multidisciplinary Cancer Congress (EMCC) in Stockholm, Sweden, Dr. Mills reviewed the many hurdles investigators face in personalized medicine research including:

  • Identifying actionable/druggable targets among the many genetic aberrations that are discovered through genomic sequencing.
  • Converting short-lived responses to targeted therapy to more durable ones.
  • Distinguishing between “passenger” mutations (ones that are along for the ride) and “driver” mutations (responsible for disease).
  • Overcoming resistance to targeted therapy that develops from activation of compensatory pathways.
  • Improving accuracy of genomic testing; 70% of tests are true positive, and 70% are true negative.
  • Genetic aberrations may change as cancer progresses. The primary tumor characteristics can be altered at relapse, and again at metastasis, suggesting that re-biopsy is necessary at each stage of disease to select appropriate therapy.
  • Regulatory issues, including drug approval and reimbursement;
  • Ethical concerns – do patients want to know their genetic findings?
  • The high cost of genomic sequencing, storing data, and targeted therapies weighed against benefits only in small subpopulations of cancer patients with a specific aberration and short-lived responses.

After reviewing these challenges in some depth, Dr. Mills described a pilot program initiated at the M.D. Anderson Cancer Center on personalized medicine. The program services about 30,000 patients per year in which a patient’s tumor undergoes genomic sequencing to identify actionable/druggable genetic aberration, and then if a drug is available for the identified target, a clinical trial of 1 is undertaken.

Mass array testing of 44 genes was undertaken in 996 patients to search for actionable mutations. The investigators identified 457 cancer-associated genes and discovered that 122 of them are frequently found in epithelial tumors. However, according to the investigators, most of the mutations were rare and less than 25% of those identified were actionable.  Dr. Mills said this percentage is much lower than has been reported in the literature.

The most common mutation was p53, but there is no targeted therapy for this aberration, he said. Fifteen percent of mutations were RAS, which are associated with resistance. He pointed out that most patients with advanced disease have the p53 mutation; but that most mutations are not “drivers,” but rather “passengers”. Investigators at M.D. Anderson are working on methods of distinguishing between passenger and driver mutations.

Dr. Mills also emphasized the difficulty in finding actionable mutations. For instance, out of the 16 most common mutations in acute myelogenous leukemia and the 237 most common mutations in lung cancer, less than 6 from both groups are actionable.

In addition, once an actionable/druggable target is identified, costs of targeted therapy can be quite high. Given that most targeted therapies achieve short-lived responses in a small subpopulation of patients, value for dollar spent becomes questionable for payers. Even though the cost of genomic sequencing has come down to under $10,000 per patient, the costs of bioinformatics or data storage and handling can be more expensive than the sequencing itself.

“It is estimated that even the eventual $1,000 genome sequencing will cost $100,000 to manage and interpret,” Dr. Mills stated.

He cautioned about all the hype surrounding personalized medicine and the false hopes this has created for patients. While there is incredible excitement about the potential implementation of personalized cancer therapy, he said “it is easy to contend that the excitement is massively overblown.”

Studies at EMCC Further Reflect the Challenges of Personalized Medicine

Several presentations with failed or suboptimal results reflected the inherent challenges in personalized medicine. In one study of patients with renal cell carcinoma, genomic sequencing and identification of single nucleotide polymorphisms (SNPs) failed to identify a subset of patients who would benefit from therapy with axitinib and also failed to show an effect on hypertension, which is a marker for response to axitinib. Axitinib is a selective inhibitor of VEGF receptors 1, 2, and 3.  (Abstract #7103, presented by Bernard Escudier, MD, Institut Gustave Roissy, Villejuif, France)

The PICCOLO study of patients with chemoresistant colorectal cancer found that the benefit of anti-EGFR therapy with panitumumab was not universal in patients with wild-type KRAS (the subset who should derive treatment benefit), because 29% of patients with wild-type KRAS were found to have other mutations that conferred resistance to the drug. These disappointing findings led study author Matt Seymour, MD, University of Leeds, U.K., to say, “There was no benefit and maybe even harm of anti-EGFR therapy in patients selected for therapy by KRAS wild-type status.” This study is an example of activation of compensatory pathways when a specific mutation is targeted, which appears to explain emergence of resistance.  (Abstract #6007)

BOLERO-2, a large Phase III trial of postmenopausal women with estrogen receptor-positive breast cancer resistant to the aromatase inhibitors letrozole or anastrozole, showed that everolimus added to another aromatase inhibitor, exemestane, was able to overcome hormonal resistance. Supposedly, the cross talk between everolimus (an mTOR inhibitor) and exemestane (an aromatase inhibitor) explains this result. (Presented by Jose Baselga, MD, Massachusetts General Hospital, Harvard Medical School, Boston, MA; Abstract #9LBA)

Further support of this thesis comes from another study presented at the meeting by G.S. Bhattacharyya, MD, Orchid Nursing Home, Kolata, India, showing that sirolimus (another mTOR inhibitor) given with tamoxifen overcomes resistance to tamoxifen alone (Abstract #16LBA).

In a separate study of sorafenib, a multi-targeted tyrosine kinase inhibitor (LBA27), enrichment of the study population for KRAS mutational status did not improve outcomes. Outcomes were similar for wild-type KRAS and mutated KRAS in patients with stage III or IV non small-cell lung cancer (NSCLC).  (Presented by A.M. Dingemans, MD, Maastricht University Medical Center, The Netherlands)

In a separate presentation on sarcoma, Jean-Yves Blay, MD, University of Lyon, France, explained that at least five genetic subtypes of sarcoma have been identified, along with complex genetic alterations. He said that, so far no targeted therapy has made substantial inroads in extending survival. However, the search continues. “Molecular subtypes will drive therapy [of sarcoma],” he predicted.

Similarly, genetic insights into ovarian cancer have led to improved understanding of the biology of ovarian cancer, but to date, no targets for small molecule inhibitors have been identified, said Michael Bookman, MD, University of Arizona Cancer Center, Tucson, AZ.

By Alice Goodman

You must be logged in to post a comment