Study: Mass spec blood sera test accurately detects ovarian cancer
A new test involving mass spectrometry of a single drop of blood serum correctly identified women with ovarian cancer in 100 percent of the patients tested, according to a study published Aug. 10 in Cancer Epidemiology, Biomarkers, & Prevention Research.
A new mass spectrometric technology was integrated with a novel support vector machine computational method and used in ovarian cancer diagnostics by John McDonald, PhD, chief research scientist at the Ovarian Cancer Institute and professor of biology at Georgia Institute of Technology (Georgia Tech) in Atlanta, and colleagues.
The researchers profiled relative metabolite levels in sera from 44 women diagnosed with serous papillary ovarian cancer (stages I-IV) and 50 healthy women or women with benign conditions.
The measurement step in the test used a single drop of blood serum, which was vaporized by hot helium plasma and was developed by the research group of Facundo M. Fernandez, PhD, associate professor in the school of chemistry and biochemistry at Georgia Tech.
As the molecules from the serum became electrically charged, a mass spectrometer was used to measure the relative abundance of metabolites in the serum. Machine learning techniques developed by Alexander G. Gray, PhD, assistant professor in the college of computing and the center for the study of systems biology at Georgia Tech were then used to sort the sets of metabolites that were found in cancerous plasma from the ones found in healthy samples. McDonald's lab mapped the results between the metabolites found in both sets of tissue to discover the biological meaning of these metabolic changes.
The assay did "well" in initial tests involving 94 subjects. In addition to being able to generate results using only a drop of blood serum, the test distinguished between the cancer and control groups with an 99 percent to 100 percent accuracy (100 percent sensitivity and 100 percent specificity by the 64-30 split validation test; 100 percent sensitivity and 98 percent specificity by leave-one-out cross-validations). In addition, it registered neither a single false positive nor a false negative, according to McDonald and colleagues.
The group is currently in the midst of conducting the next set of assays, this time with 500 patients. "The caveat is we don't currently have 500 patients with the same type of ovarian cancer, so we're going to look at other types of ovarian cancer," said Fernandez. "It's possible that there are also signatures for other cancers, not just ovarian, so we're also going to be using the same approach to look at other types of cancers."
In addition to having a relatively low prevalence (0.04 percent), ovarian cancer is also asymptomatic in the early stages. The ability to accurately and inexpensively diagnose ovarian cancer will have a significant positive effect on ovarian cancer treatment and outcome, concluded the researchers.
A new mass spectrometric technology was integrated with a novel support vector machine computational method and used in ovarian cancer diagnostics by John McDonald, PhD, chief research scientist at the Ovarian Cancer Institute and professor of biology at Georgia Institute of Technology (Georgia Tech) in Atlanta, and colleagues.
The researchers profiled relative metabolite levels in sera from 44 women diagnosed with serous papillary ovarian cancer (stages I-IV) and 50 healthy women or women with benign conditions.
The measurement step in the test used a single drop of blood serum, which was vaporized by hot helium plasma and was developed by the research group of Facundo M. Fernandez, PhD, associate professor in the school of chemistry and biochemistry at Georgia Tech.
As the molecules from the serum became electrically charged, a mass spectrometer was used to measure the relative abundance of metabolites in the serum. Machine learning techniques developed by Alexander G. Gray, PhD, assistant professor in the college of computing and the center for the study of systems biology at Georgia Tech were then used to sort the sets of metabolites that were found in cancerous plasma from the ones found in healthy samples. McDonald's lab mapped the results between the metabolites found in both sets of tissue to discover the biological meaning of these metabolic changes.
The assay did "well" in initial tests involving 94 subjects. In addition to being able to generate results using only a drop of blood serum, the test distinguished between the cancer and control groups with an 99 percent to 100 percent accuracy (100 percent sensitivity and 100 percent specificity by the 64-30 split validation test; 100 percent sensitivity and 98 percent specificity by leave-one-out cross-validations). In addition, it registered neither a single false positive nor a false negative, according to McDonald and colleagues.
The group is currently in the midst of conducting the next set of assays, this time with 500 patients. "The caveat is we don't currently have 500 patients with the same type of ovarian cancer, so we're going to look at other types of ovarian cancer," said Fernandez. "It's possible that there are also signatures for other cancers, not just ovarian, so we're also going to be using the same approach to look at other types of cancers."
In addition to having a relatively low prevalence (0.04 percent), ovarian cancer is also asymptomatic in the early stages. The ability to accurately and inexpensively diagnose ovarian cancer will have a significant positive effect on ovarian cancer treatment and outcome, concluded the researchers.