Glycan-based biomarkers accurately detected epithelial ovarian cancer, suggesting they may have utility as a new diagnostic test, according to study results.
“This is one of many papers we’ve done to see if glycans can distinguish between women who have ovarian cancer and those who don’t,” Gary Leiserowitz, MD, chief of the division of gynecologic oncology at UC Davis Comprehensive Cancer Center, said in a press release. “So far, the results have been consistent and promising.”
Leiserowitz and colleagues obtained blood samples from the tissue bank of the Gynecologic Oncology Group. The samples came from 100 healthy female controls, 52 women with tumors classified as having low malignant potential and 147 women diagnosed with epithelial ovarian cancer. The researchers matched the cases and controls for age within ±5 years at the time of enrollment.
Investigators used mass spectroscopy to evaluate the samples for different quantities of glycan expression. They then used a two-phase procedure for biomarker detection and confirmation.
The researchers developed candidate classifiers of glycans that distinguished cases from controls, using a training set in the discovery phase. They later evaluated the classification ability of these candidate classifiers using independent test samples not used in the discovery phase.
The glycan patterns demonstrated excellent detection in differentiating ovarian cancer and low malignant potential tumor cases from control cases. The candidate glycan-based biomarkers formulated on a training set (86% sensitivity, 95.8% sensitivity for differentiating ovarian cancer from controls through “leave-one-out” cross-validation) verified their capability as a detection test utilizing an independent test set (70% sensitivity, 86.5% specificity).
“We take all these rigorous step‐wise approaches to eliminate the possibility of bias,” Leiserowitz said. “This paper establishes that these are consistent and reproducible findings. This is a real phenomenon.”
Disclosure: The researchers report no relevant financial disclosures.