| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© Oxford University Press 2006.
NEWS |
Experts Wrestle With Problems Developing Biomarkers, Search For New Tests
Biomarkers have a variety of uses in preclinical development: screening for promising drugs, determining the dosing and scheduling of a drug, predicting the risk of developing disease, predicting a patient's chance of responding to a drug or of having an adverse reaction to a drug, or serving as surrogate endpoints in a trial. And many researchers argue that the proper identification and validation of biomarkers will be important for the future of medicine.
"Only 30% of the patients in clinical trials and patient care have a measurable positive response. The remaining 70% are exposed to risks with no benefit," said Michael Phelps, Ph.D., chair of the Department of Molecular and Medical Pharmacology and director of the Institute for Molecular Medicine at the University of California, Los Angeles. "We have to develop more personalized medicine so that drugs are used by people who stand to benefit. This would mean transforming molecular diagnostics into a measurement science that can separate individual patients by their biological signatures."
|
Phelps and other researchers and clinicians from academia, industry, and the government spent two and a half days at the National Academy of Sciences in Washington, D.C., in mid-March discussing roadblocks to the development of cancer biomarkers and ways to overcome them (see box). The group was convened by the Institute of Medicine's National Cancer Policy Forum.
Biomarkers play a key role in personalizing medicine, and their kinds and uses vary. The biological material can be genes, proteins, or messenger RNA, and markers may be analyzed singly or in groups. Each biomarker type, however, will require individual testing for validation, and, depending on the biomarker type, that validation test may vary. The estrogen receptor and human epidermal growth factor receptor 2 (HER-2) are examples of biomarkers that are currently used to determine treatment for breast cancer.
But discovering more biomarkers that can reliably predict whether a patient will respond to a drug has been a slow process for many reasons. Chief among them, speakers at the meeting noted, is the difficulty of obtaining tissue samples to use in testing biomarkers.
Helen Francis Lang, Ph.D., a scientist at Affymetrix in Santa Clara, Calif., has spent months searching through the databases of several institutions for appropriate samples. "Many of the patient records have incomplete medical histories, or the sample size is too small for statistical significance, or the databases are not searchable. The methods of collecting data are different for each institution—one center measures 15 variables and another only 10. There is no consistency." She added that, in many European countries, researchers have easier access to patient material than those in the United States because those countries have centralized health care systems with standardized methods of collecting and storing data.
Intellectual property rights and informed-consent agreements that restrict the future use of samples are other factors that may interfere with access to biological samples.
| Suggestions for Improving Cancer Biomarker Development In mid-March, a group convened by the National Cancer Policy Forum discussed several ways to improve cancer biomarker development. Possible solutions that emerged from the half-dozen breakout groups that met during the meeting were as follows:
|
Another often-cited roadblock to biomarker development is the difficulty of getting investors to back biomarker discovery. Besides the risk that a biomarker may not work, the returns on a diagnostic test—which likely will be used only once per patient—may be less than a therapeutic drug prescribed over a given period. Furthermore, once a diagnostic test enters the market, competitors may be quick to develop a better test.
Besides the few incentives for private industry to take on this work, the National Institutes of Health has not traditionally funded this sort of translational work, and the academic career reward structure does not encourage biomarker development, noted Scott Ramsey, M.D., Ph.D., from the Fred Hutchinson Cancer Center and a faculty member at the School of Medicine at the University of Washington in Seattle.
Several participants noted that the lethargy in this field is also due to a lack of clear regulatory procedures for developing a particular diagnostic or an articulated pathway for FDA approval of a biomarker. For example, there is no business model for moving microarrays to the market and no standardized method of collecting data for proteomics. Many speakers at the March forum bemoaned the fact that biomarker testing isn't routinely integrated into clinical trials and that valuable data from completed clinical trials may be lost.
Despite the substantial obstacles and efforts required to overcome them, several participants pointed out that it was important keep positive stories in mind. Some new diagnostic tests have recently entered the clinic. One, Oncotype DX, marketed by Genomic Health in Redwood City, Calif., predicts the likelihood of breast cancer recurrence in women with newly diagnosed node-negative, estrogen receptor–positive breast cancer who will be treated with tamoxifen. The test, which amplifies 21 genes (using reverse transcription–polymerase chain reaction) from a paraffin-embedded breast tumor tissue sample, may predict the likelihood of recurrence as well as which women are likely to benefit from chemotherapy. Stored clinical samples from two large clinical trials were used to test its clinical validity, although the value of the test in making clinical decisions is still being studied.
Another bright spot was the 1999 creation of the Microarray Gene Expression Data Society (MGED) in the United Kingdom. MGED is a grass-roots organization whose mission is to produce standards that allow people to share and understand the data produced by microarray experiments. In 2000, they published a paper in Nature Genetics in 2000 including a guide to authors, editors, and reviewers of microarray gene expression papers. They also established a data exchange format.
FDA also said in a recent report, Critical Path Report and Opportunities List, that biomarker development is a critical area of focus, and, to that end, the agency is developing the process for regulatory acceptance of biomarkers used with cancer therapies. In February, FDA, NCI, and the Centers for Medicare and Medicaid Services announced an Oncology Biomarker Qualification Initiative to speed the development and evaluation of cancer therapies. Their first project will be to validate and standardize the use of fluorodeoxyglucose–positron emission tomography (PET) scanning as a predictor of tumor response in patients being treated for non-Hodgkin lymphoma. The scanner acts like a molecular camera, which uses the fluorine isotope tag on glucose to visualize the higher metabolic rate of glucose in cancer cells compared with that in normal cells.
Another ongoing effort is the creation of the Multiple Myeloma Research Consortium, which brings together leading academic institutions to accelerate drug development in the treatment of multiple myeloma (see sidebar, p. 573). Current projects involve preclinical efforts to validate key targets in the disease and collaborations with biotech partners to test promising therapies in early trials.
Looking back on the two and a half days, Richard A. Frank, M.D., Ph.D., vice president of clinical and medical strategy at GE Healthcare in Princeton, N.J., thought the main point to remember was to stay focused on the goal—developing products to help patients. Biomarkers hold the promise of detecting disease early, staging it accurately, testing drug response, assessing the treatment benefit, and refining treatment regimens, Frank said.
A committee appointed by the Institute of Medicine will use the proceedings of the meeting to help develop recommendations for the field of biomarker development. The final report of the committee is expected in early 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||