Esophageal cancer is a disease with high mortality. In order to improve the 5 year survival rate after cancer treatment, it is important to develop a method for early detection of the cancer and for therapy support. There is increasing evidence that Raman spectroscopy, in combination with chemometric analysis, is a powerful technique for discriminating pre-cancerous and cancerous biochemical changes. In the present study, we used Raman spectroscopy to examine early-stage (stages 0 and I) esophageal cancer samples ex vivo. Comparison between the Raman spectra of cancerous and normal samples using a t-test showed decreased concentrations of glycogen, collagen, and tryptophan in cancerous tissue. Partial least squares regression (PLSR) analysis and self-organization maps (SOMs) discriminated the datasets of cancerous and normal samples into two groups, but there was a relatively large overlap between them. Linear discriminant analysis (LDA) based on Raman bands found in the t-test was able to predict the tissue types with 81.0% sensitivity and 94.0% specificity.

Subcutaneous adipose tissue (SAT), visceral adipose tissue (VAT), and fat beneath the dermis layer were investigated using a ball lens top hollow optical fiber Raman probe (BHRP). Hamsters were fed with trilinolein (TL) and tricaprin (TC) for six weeks and measurements were carried out every two weeks. The BHRP with an 800 μm diameter fused-silica ball lens was able to obtain information on the subcutaneous fat in a totally non-invasive manner. Changes in the concentration of TL and TC during the treatment were analyzed, and the relationship between fat accumulation and dietary fat was studied. It was found that SAT had, in general, a higher degree of unsaturation than VAT. The accumulation rate of TC found in SAT and VAT was 0.52 ± 0.38 and 0.58 ± 0.4%, respectively, while the TL accumulation rate was 4.45 ± 1.6 and 4.37 ± 2.4%, respectively. The results suggest different metabolic pathways for TC, a typical medium-chain fatty acid, and TL, a long-chain unsaturated fatty acid. Raman subsurface spectra were successfully obtained and used to analyze the subcutaneous fat layer. The accumulation rates of TL and TC found in skin fat were 5.01 ± 3.53% and 0.45 ± 0.36%, respectively. The results demonstrate the high feasibility of Raman spectroscopy for non-invasive analysis of adipose tissue.

The purpose of the present study was to investigate molecular compositions of lipid droplets changing in live hepatic cells stimulated with major fatty acids in the human body, i.e., palmitic, stearic, oleic, and linoleic acids. HepG2 cells were used as the model hepatic cells. Morphological changes of lipid droplets were observed by optical microscopy and transmission electron microscopy (TEM) during co-cultivation with fatty acids up to 5 days. The compositional changes in the fatty chains included in the lipid droplets were analyzed via Raman spectroscopy and chemometrics. The growth curves of the cells indicated that palmitic, stearic, and linoleic acids induced cell death in HepG2 cells, but oleic acid did not. Microscopic observations suggested that the rates of fat accumulation were high for oleic and linoleic acids, but low for palmitic and stearic acids. Raman analysis indicated that linoleic fatty chains taken into the cells are modified into oleic fatty chains. These results suggest that the signaling pathway of cell death is independent of fat stimulations. Moreover, these results suggest that hepatic cells have a high affinity for linoleic acid, but linoleic acid induces cell death in these cells. This may be one of the causes of inflammation in nonalcoholic fatty liver disease (NAFLD).

The aim of the present study is to evaluate the capability of a miniaturized Raman endoscope (mRE) system to monitor the advancement of colorectal tumors in model mice as a method that is noninvasive to the tumor itself. Nevertheless, the endoscope is narrow enough to observe the inside of the mouse colon in such a way that is semi-noninvasive to the animal. However, the mRE system allowed the visualization and Raman spectral measurement of any targeted point within the colorectal tumor in model mice under anesthesia, without damaging the tissue (i.e., noninvasively). Continuous monitoring of the same tumor allowed the observation of alterations in its molecular composition and size, along with its advancement. The tumor lesion was discriminated from normal tissues of the control mouse with an accuracy of 86.8%. We succeeded in visualizing and performing Raman spectral observations on a shrinking polyp type tumor. The Raman analysis suggested that it was not cured but supposedly transformed to another tumor type.

An in vivo non-invasive quantitative analysis technique was introduced for evaluating the fat composition of living marine diatoms by using Raman spectroscopy in conjunction with a chemometric method. This technique enabled the observation of real-time variations in individual lipids in diatom cells without specific treatment or fat extraction. A confocal Raman spectroscope was used to measure the marine centric diatom Thalassiosira ( T.) pseudonana that was cultured under six stress conditions, and the spectral data of accumulated fatty acids were obtained. A model-based chemometrics technique, ordinary least square was then used to extract specific signals from Raman spectra obtained for a mixture of fatty acids. The levels of four major lipid moieties from diatoms were extracted simultaneously, including myristic acid, palmitic acid, palmitoleic acid, and eicosapentaenoic acid from the Raman spectra. These results indicate that Raman spectroscopy in conjunction with a chemometrics method is reliable for the quantitative determination of the lipid composition accumulated in the cells of marine diatoms.

A miniaturized Raman endoscope (mRE) system was employed to study the effects of anticancer treatment on colorectal tumors in a live murine model. The endoscope is narrow enough to observe the inside of the mouse colon under anesthesia. It has a channel for a ball lens mounted on a hollow fiber Raman probe (BHRP) to measure any targeted point under the visual control of the endoscope. Colorectal cancer tissue was observed to study the alterations of the tissue in response to anticancer drug treatment. Three anticancer drugs, 5-fluorouracil (5-FU), cisplatin (CDDP), and docetaxel, were employed. Although no alteration was recognized in the endoscopic visual observations at 2 weeks after the drug treatment, the Raman spectra obtained in the live mouse colon indicated that molecular changes of lipids and proteins were observed. This study demonstrates that in situ Raman analysis is highly sensitive for detecting the effects of anticancer drugs.