Recognition of serous ovarian tumors in human samples by multimodal nonlinear optical microscopy.

We used a multimodal nonlinear optics microscopy, specifically two-photon excited fluorescence (TPEF), second and third harmonic generation (SHG∕THG) microscopies, to observe pathological conditions of ovarian tissues obtained from human samples. We show that strong TPEF + SHG + THG signals can be obtained in fixed samples stained with hematoxylin and eosin (H&E) stored for a very long time, and that H&E staining enhanced the THG signal. We then used the multimodal TPEF-SHG-THG microscopies in a stored file of H&E stained samples of human ovarian cancer to obtain complementary information about the epithelium∕stromal interface, such as the transformation of epithelium surface (THG) and the overall fibrillary tissue architecture (SHG). This multicontrast nonlinear optics microscopy is able to not only differentiate between cancerous and healthy tissue, but can also distinguish between normal, benign, borderline, and malignant specimens according to their collagen disposition and compression levels within the extracellular matrix. The dimensions of the layers of epithelia can also be measured precisely and automatically. Our data demonstrate that optical techniques can detect pathological changes associated with ovarian cancer.