Scientists at Johns Hopkins have successfully detected ovarian cancer using a blood test for DNA shed by tumors. The test is based on digital analysis of single nucleotide polymorphisms, known as SNP or "snips," in which investigators separate the two strands of code found in every gene to search for imbalances that are a hallmark of cancer cell DNA.
According to Ie-Ming Shih, pathologist and director of this study for the Kimmel Cancer Center at Johns Hopkins, "Digital SNP appears to detect ovarian cancers very well and is far more precise than other available tests."
With 54 blood samples from late- and early-stage ovarian cancer patients, the Hopkins team used digital SNP analysis to find so-called "allelic imbalance" in 87 percent (13 out of 15) of early-stage ovarian cancers and 95 percent (37 out of 39) with late-stage disease. No allelic imbalance was detected in 31 blood samples from healthy individuals. The researchers also compared the type of allelic imbalance found in 17 of the samples with the corresponding tumor tissue and found that 15 of these had matching allelic imbalance patterns.
Details of the initial studies of the test are published in the Nov. 20 issue of the Journal of the National Cancer Institute.
Although digital SNP is too costly and labor intensive at present to serve as a general screening test, it might be useful for women at high risk, Shih says. The Johns Hopkins group also is investigating ways of achieving the same accurate detection rate with a less costly, more efficient test that could be used on a broader scale for ovarian and a variety of other cancers.
DNA released from dying cells has long been detectable in blood samples, using sensitive molecular technology. But to distinguish normal from cancerous DNA, Kimmel Cancer Center scientists analyzed both sets of genetic code in DNA sequences. The individual sets of code are called alleles. In normal cells, DNA's two alleles--one derived from the maternal copy of the gene and the other from the paternal copy--are balanced in their basic building blocks. Tumor cells, on the other hand, have an unequal ratio of maternal and paternal alleles. Digital SNP analysis counts the alleles present in each blood sample.
In the Johns Hopkins study, investigators first measured the total amount of DNA in blood samples taken from 44 healthy individuals; 122 patients with a variety of cancers ranging from head and neck cancers to brain cancer, as well as the 54 ovarian cancer patients; and 164 patients with noncancerous diseases such as diabetes and hypertension. They found that, compared with blood samples of healthy individuals, average amounts of total DNA more than doubled for those with noncancerous disease (7 ng/mL) but were eight times greater in samples from all cancer patients (59 ng/mL).
"The problem with using the total amount of DNA in the blood without performing digital SNP is that they are not specific for cancer, as elevated DNA levels can be found in blood samples from patients without cancer," Shih says.
Next, singling out the ovarian cancer samples, Shih and his team found high total amounts of DNA in only 47 percent (7 of 15) early-stage ovarian cancers and 56 percent (22 of 39) with late-stage disease. Adding another test, employing a standard ovarian cancer protein marker (CA125) currently used to monitor disease, did little to improve detection rates, Shih reported.
"A test based on digital SNP holds promise for improved detection in a wide range of cancers, as well as ovarian cancer, which is currently detected almost always when it is in late stages and difficult to treat," Shih says. Ovarian cancer will strike an estimated 23,000 U.S. women and cause approximately 14,000 deaths this year. It ranks fifth in cancer deaths among women. Generally, ovarian cancer is "silent" until the cancer has spread.
Funding for this research was provided by the National Cancer Institute and the Richard TeLinde Research Fund.
In addition to Shih, Johns Hopkins participants in this research are Hsueh-Wei Chang, Shing M. Lee, Steven N. Goodman, Gad Singer, Sarah K. R. Cho, Lori J. Sokoll, Fredrick J. Montz, Richard Roden, Zhen Zhang, Daniel W. Chan and Robert J. Kurman.