The Failed Promise of Gene Based Tests for Diagnosing and Treating Cancer

The Human Genome Project was a $3 billion, 13-year project initiated by the US Department of Health and the National institutes of Health (NIH). Completed in 2003 with help from international collaborators, it identified and mapped the 20,000-25,000 genes in human DNA. This information was gathered to better understand which genes are involved in which diseases and medical conditions and to spur medical innovation.[1]

When the Human Genome Project released its first “draft” in 2000, many scientists believed it would revolutionize medical research. President Bill Clinton claimed that genetic diagnosis (the ability to tell who has a disease after looking at the genes) would “revolutionize the diagnosis, prevention and treatment of most, if not all, human diseases.” In particular, many hoped it would serve as a way to treat and prevent cancer. At that time, the director of the genome agency at the National Institutes of Health claimed genetic diagnosis of diseases would be accomplished within 10 years.[2]

Eleven years later, we are still waiting for progress in fighting cancer. Researchers

hoped to find gene patterns that could be used to tell who has the disease, who is likely to get the disease, and what the likely outcome of the disease will be. This would allow for better intervention and treatment.[3] However, this has simply not been the case. Although many disease-causing gene mutations have been discovered, disease is often the result of an interaction of many genes, not just a few. As a result, gene-based cancer tests only function well when a few genes are primarily responsible for the disease. Furthermore, it is a long process from when researchers first discover the responsible gene to actually creating a test or a drug.[4] Although the ability to map the human genome has great promise, a decade later it still hasn’t yielded good methods for diagnosing cancer. Even more disappointing: recent scandals and severe product flaws have cast doubt on gene-based research as a whole.

The BRCA1 and BRCA2 Case: A Bright Spot on the Horizon for Genetic Testing?

Despite problems with gene-based cancer research, some successful genetic tests have been developed. The tests for the BRCA1 and BRCA2 gene mutation, which place women at much higher risk for breast cancer, demonstrate the potential for genetic testing as well as the technology’s limitations.

In the early 1990s researchers discovered that mutations in the BRCA1 and BRCA2 greatly increased a women’s risk of developing breast and ovarian cancer. In the general population, 12% of women develop breast cancer.[5] In contrast, women carrying mutations of the BRCA1 or BRCA2 have a 60% chance of developing breast cancer, a five-fold increase. Similarly, the general population has a 1.4% chance of developing ovarian cancer while 15-40% of women with the harmful mutation develop the disease.

Doctors can now test women for the disease and offer different treatment options based on the results. For example, if a woman has the BRCA1 or BRCA2 gene, she can undergo increased cancer screening including mammograms, clinical breast and ovarian cancer exams as well as ultrasounds. Some individuals even decide to remove at-risk tissue (including healthy breasts, fallopian tubes, and ovaries) in order to reduce the risk of developing cancer. Since not all of the tissue can be removed, some women who choose to do this still develop breast and ovarian cancer later on.

Testing for the BRCA1 and BRCA2 genes are the most successful cases of genetic testing, but their role in preventing and treating breast and ovarian cancers is limited by a number of factors. The BRCA1 and BRCA2 account for only 5-10% of breast cancers and 10-15% of ovarian cancer, a fraction of overall breast and ovarian cancer cases.[6] Moreover, not all women carrying BRCA1 or BRCA2 genes will develop breast or ovarian cancer. So while a positive test CAN tell a woman her risk of developing certain cancers, it CAN’T tell which women among those who have the mutation will actually develop cancer or when they will develop cancer. Part of the reason these test are as useful as they are is that they are single gene test. Gene-based tests are much less accurate when multiple genes need to be tested—the case for most diseases.

Furthermore, gene-based tests are only useful if implemented correctly. However, many doctors do not follow the proper genetic counseling guidelines when testing women for hereditary breast and ovarian cancer. One study found that too many average risk women receive BRCA1 and BRCA2 tests, while too few high-risk women receive testing. This can be dangerous for not only the high-risk women that went untested but also the average-risk patients whom genetic testing is not recommended for, due to the risk of unnecessary preventative measures.[7] In order to use the BRCA1 and BRCA2 tests safely and effectively physicians need to follow the guidelines and assess these women’s risks accurately.

Less Successful Cases: Ovasure and Duke

Because of the relative success of the BRCA1 and BRCA2 genetic tests, there have been many similar (and mostly failed) attempts to develop genetic tests, especially for the most common and deadly types of cancer. However, recent product flaws and scandals have cast doubt on the promise of gene-based research and have focused many researchers and the public on its limitations.

The genetic test developed at Yale University OvaSure was introduced in 2008 to detect ovarian cancer in an early treatable stage. This test initially generated a lot of excitement because when ovarian cancer is detected early, over 90% of women live more than five years. If on the other hand, it is detected late, only 30% of women live more than five years. Over 15,000 women die from the disease every year.[8]

This test not only failed to live up to its potential but the FDA declared in 2008 that it may actually harmful to the public health.[9] A 2011 randomized control trial run by Saundra Buys of the University of Utah found that not only did screening with OvaSure NOT reduce ovarian cancer mortality, but many women who did not really have ovarian cancer (false positives) had unnecessary surgery as a result. These surgeries sometimes caused serious complications and even death. Out of the study’s 78,000 participants, 3,285 received false positives due to OvaSure. Out of these participants 1,085 underwent unnecessary biopsy surgery. In 15% of cases, the surgery caused serious complications. In addition, the test was not good at identifying who actually had the disease; many people also had false negatives.[10]

In some cases, the potential of gene-based research has resulted in fraudulent studies. In one of the more famous cases, a team of investigators from Duke, led by oncologist and genetics researcher Anil Potti, spent many years publishing papers, receiving funding, and treating patients based on looking for gene patterns that would supposedly determine which drugs would best shrink an individual patient’s tumor. In theory this would help eliminate experimentation and allow doctors to zero-in on the treatment most likely to benefit a particular cancer patient. After the results of Potti’s four prominent peer-reviewed journal articles and three clinical trials containing falsified data, Potti resigned. The articles were retracted and the trials shut down. Although two statisticians repeatedly brought up problems with their research, no one paid attention until Potti was found in the summer of 2010 to have falsified his credentials.[11]

Dr. Nevins, who conducted research alongside Dr. Potti, claimed the team of researchers continued their trials because they didn’t know problems with their research were due to manipulated and falsified data.[11]

The inability of the peer-review process to catch major errors and misrepresentation of data in these studies is disconcerting. If scientists could not catch these errors for years, it implies that other studies could have similar errors that also went unnoticed. In the Potti case, many members of the research team were not aware that data manipulation had taken place. Furthermore, the increasing complexity of studies and complicated data sets increases the chances of cases like these happening in the future, especially as researchers focus on the more complicated area of multi-gene cancer tests. Even more troublesome, these incidences happened at prominent institutions normally noted for excellent research and validity of data.

The Bottom Line

While gene-based cancer research has promise it also has many limitations. Because diseases often are the result of many genes, not just a few, it is extremely difficult for a single test to effectively analyze the interaction of many genes. Even single gene based tests whose effectiveness has been proven (such as BRCA1 and BRCA2) are only able to tell who is at increased risk for a disease—they can’t tell for sure who will actually develop it.

All articles on our website have been approved by Dr. Diana Zuckerman and other senior staff. 


1. Human Genome Project Information.

2. Wade, Nicholas. A Decade Later, Genetic Map Yields Few New Cures. The New York Times. June 12, 2010.

3. Human Genome Project Information.

4. Wade, Nicholas. A Decade Later, Genetic Map Yields Few New Cures. The New York Times. June 12, 2010.

5. National Cancer Institute. Fact Sheet. BRCA1 and BRCA2: Cancer Risk and Genetic Testing.

6. National Cancer Institute. Fact Sheet. BRCA1 and BRCA2: Cancer Risk and Genetic Testing.

7. Trivers, K. F., Baldwin, L.-M., Miller, J. W., Matthews, B., Andrilla, C. H. A., Lishner, D. M. and Goff, B. A. Reported Referral for Genetic Counseling of BRCA ½ Testing Among United States Physicians. Cancer. July 2011.

8. Pollack, Andrew. Cancer Test for Women Raises Hope, and Concern. The New York Times. August 25, 2008.

9. Libby’s HOPE. Lost in Translation? FDA Believes that LapCorp’s Ovarian Cancer Early Detection Test (Ovasure) Lacks Adequate Clinical Validation. August 23, 2008.

10. Buys S.S., et al. Effects of Screening on Ovarian Cancer Mortality: The Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Randomized Controlled Trial. The Journal of the American Medical Association. July 2011; 2011 (616):1.

11. Nelson, Roxanne. ‘Data Gone Wrong’: Unreproducable Cancer Genomics Studies. Medscape News Today. June 2011.