NMR metabolomics study of follicular fluid in women with cancer resorting to fertility preservation

Abstract

Purpose: The purpose of this study was to evaluate the possible application of metabolomics to identify follicular fluid changes in cancer patients undergoing fertility preservation. Although metabolomics have been applied already in cancer studies, this is the first application on follicular fluid of cancer patients.

Methods: We selected for the study ten patients with breast cancer and lymphoma who resorted to oocyte cryopreservation to preserve fertility and ten healthy women undergoing in vitro fertilization treatments. Follicular fluid was collected at the time of oocytes retrieval. Metabolomic analysis of follicular fluids was performed by ¹H-nuclear magnetic resonance (NMR) spectroscopy in combination with multivariate analysis to interpret the spectral data. Univariate statistical analysis was applied to find correlations between patients' features and metabolites identified by NMR.

Results: Partial least squares discriminant analysis allowed to discriminate samples from cancer patients and healthy controls. Univariate statistical analysis found significant correlations between patients' features and metabolites identified by NMR. This finding allowed to identify biomarkers to differentiate both healthy controls from cancer patients and the two different classes of oncological patients.

Conclusion: The follicular fluids of cancer patients display significant metabolic alterations in comparison to healthy subjects. NMR-based metabolomics could be a valid prognostic tool for identifying and selecting the best cryopreserved oocytes and improving the outcome prediction in cancer women undergoing in vitro fertilization.

Keywords: Biomarkers; Cancer; Fertility preservation; Follicular fluid; Metabolomics; NMR.

Fig. 1

PLS-DA score plot based on ¹H-NMR spectra of follicular fluid samples from women examined in this study: a healthy control (N = 10, dots) and breast cancer (N = 6, triangles); b healthy control (N = 10, dots) and lymphoma (N = 4, crosses); c healthy control (N = 10, dots) and all oncological patients (N = 10, circles); d cross-validation of the PLS-DA model reported in (c) by permutation testing (20 permutations)