Skip to main content
Premium Trial:

Request an Annual Quote

Metabolic Signature Reported to Distinguish Early-Stage Ovarian Cancer

NEW YORK (GenomeWeb) – A Georgia Tech-based team of researchers has uncovered a metabolic signature that they believe distinguishes women with early-stage ovarian cancer from healthy women.

Using high-performance mass spectrometry, Georgia Tech's John McDonald and his colleagues studied the serum metabolomes of women with and without ovarian cancer. Through this, they homed in on 16 metabolites that could accurately identify women with ovarian cancer, as they reported today in Scientific Reports.

"We've tested it with different kinds of statistical tests and we consistently get high accuracy, and the metabolites that we've identified as diagnostic are ones that have previously been implicated in cancer," McDonald told GenomeWeb. "All of these things suggest that we're on the right track."

Ovarian cancer is often asymptomatic until its late stages, but when it's caught late, it has a five-year relative survival rate of less than 44 percent. If it's detected early, the five-year survival rate is about 90 percent.

In their search for a metabolic signature of early-stage ovarian cancer, the researchers turned to negative ion mode UPLC-MS to interrogate the serum metabolomes of 46 patients with early-stage serous epithelial ovarian cancer and 49 age-matched healthy controls.

This, they reported, detected more than 4,000 spectral features, though they whittled that number down to 255 through manual filtering that eliminated inconsistent and ambiguous features.

With the remaining features, the researchers built a discriminant linear support vector to find a model that could predict ovarian cancer. Through a recursive feature elimination approach in which features were removed from the model to gauge how they affected its predictive powers, McDonald and his colleagues zeroed in on a set of 16 metabolic features that could, in a separate group, predict ovarian cancer with 100 percent sensitivity and 100 percent specificity, an accuracy that McDonald noted was shocking.

He added that they validated this high predictive accuracy through other statistical approaches, including orthogonal partial least squares-discriminant analysis using a range of cross-validation approaches.

The researchers teased out the metabolite identities of 11 of the 16 features. These metabolites included a number of lipids and fatty acids, and they noted that changes in lipid and fatty acid metabolism has been linked previously to the onset and progression of ovarian cancer.

"In many cases, the feature we identified had previously been implicated in ovarian cancer, which is cool because the approach is completely unbiased in terms of what it's identifying as informative features," McDonald said.

Two features were identified as lysophospholipids —lysophosphatidylethanolamine and lysophosphatidylinositol — the serum levels of which increase in ovarian cancer. Lysophosphatidylinositol, for instance, binds and activates a G-protein coupled receptor to initiate the proliferation and growth of ovarian cancer cells.

Another feature, phosphatidylinositol, is one of many inositol membrane phospholipids that help recruit the serine/threonine kinase Akt to the plasma membrane and aids in its phosphorylation and activation. Phosphorylation of inositol phospholipids, the researchers noted, is carried out by PI3K, which has been associated with a range of cancer-related functions, including proliferation, cell adhesion, apoptosis, and transformation.

"It's comforting to know that it's not coming up with things that are completely out in left field," he added.

McDonald pointed out, though, that this study was mostly a proof-of-concept project and needs to be extended. While it incorporated patients from a range of geographical areas, he said future studies need to include many more patients from varying ethnic backgrounds.

He also said that he and he colleagues are setting up collaborations to test this metabolite signature in a cohort of women at high risk of developing ovarian cancer, such as women with BRCA mutations, to see whether it can catch the disease in its earliest stages.