IR spectral imaging of secreted mucus: a promising new tool for the histopathological recognition of human colonic adenocarcinomas.

AIMS: During colonic carcinogenesis, mucin-type glycoproteins are known to undergo quantitative and qualitative alterations. The aim of this study was to determine the value of infrared (IR) spectral histology for the histopathological recognition of colonic adenocarcinomas based on mucin-associated IR spectral markers. METHODS AND RESULTS: Paraffin-embedded tissue sections of normal human colon and adenocarcinomas were analysed directly by IR-microspectroscopy (IR-MSP), without prior chemical dewaxing. IR-MSP imaging combined with multivariate analysis permitted the construction of IR colour-coded images of the tissue sections providing spatially resolved biochemical information. This allowed localization of mucin-rich areas and provided label-free spectral-based staining of secreted mucus related to the biochemical heterogeneity of its mucin content. IR images of secreted mucus display the same spectral clusters in both normal and adenocarcinomatous colonic tissues, but with significant differences in surface percentages. Such differences allow a distinction between these two tissue types. Spectral variations associated with changes of mucin secondary structure were the most accurate mucus spectral marker for discriminating between normal colon and adenocarcinomas in the sample set. CONCLUSIONS: IR-MSP imaging provides a new type of histology, independent of visual morphology, presenting tremendous possibilities for discovery and clinical monitoring of cancer markers.

IR spectral imaging for histopathological characterization of xenografted human colon carcinomas.

This study aims to develop IR imaging of tumor tissues for generating an automated IR-based histology. Formalin-fixed paraffin-embedded xenografts of human colon carcinomas were analyzed. Chemometric and statistical multivariate treatments of spectral data permitted to probe the intrinsic chemical composition of tissues, directly from paraffinized sections without previous dewaxing. Reconstructed color-coded spectral images revealed a marked tumor heterogeneity. We identified three spectral clusters associated to tumoral tissues, whereas HE staining revealed only a single structure. Nine other clusters were assigned to either necrotic or host tissues. This spectral histology proved to be consistent over multiple passages of the same xenografted tumor confirming that intratumoral heterogeneity was maintained over time. In addition, developing an innovative image analysis, based on the quantification of neighboring pixels, permitted the identification of two main sequences of spectral clusters related to the tissue spatial organization. Molecular attribution of the spectral differences between the tumor clusters revealed differences of transcriptional activity within these tumor tissue subtypes. In conclusion, IR spectral imaging proves to be highly effective both for reproducible tissue subtype recognition and for tumor heterogeneity characterization. This may represent an attractive tool for routine high throughput diagnostic challenges, independent from visual morphology.

Combination of FTIR spectral imaging and chemometrics for tumour detection from paraffin-embedded biopsies.

FTIR spectral imaging was applied on formalin-fixed paraffin-embedded biopsies from colon and skin cancerous lesions. These samples were deposited onto different substrates (zinc selenide and calcium fluoride respectively) and embedded using two types of paraffin. Formalin fixation followed by paraffin embedding is the gold standard in tissue storage. It can preserve molecular structures and it is compatible with immunohistochemistry. However, paraffin absorption bands are significant in the mid-infrared region and can mask some molecular vibrations of the tissue. Direct data processing was applied on spectral images without any chemical dewaxing of the tissues. Extended Multiplicative Signal Correction was used to correct the spectral contribution from paraffin. For this purpose, the signal of paraffin was modelled using Principal Component Analysis and paraffin spectra were removed from the raw images based on an outlier detection. Then, pseudo-colour images were computed by K-means clustering in order to highlight histological structures of interest. This robust chemometrics methodology was applied on the two samples. Tumour areas were successfully demarcated from the rest of the tissue in both colon and skin independently of the embedding material and of the substrate.

Fiber-optic probes for in vivo Raman spectroscopy in the high-wavenumber region.

In vivo Raman spectroscopy, using fiber-optic probes is hindered by the intense background signal, which is generated in the fused-silica fibers, in the fingerprint region of the Raman spectrum (approximately 0-2000 cm(-1)). Optical filtering is necessary to obtain tissue spectra of sufficient quality. The complexity of fiber-optic probes for fingerprint Raman spectroscopy, in combination with size constraints and flexibility requirements for in vivo use have been a major obstacle in the development of in vivo diagnostic tools based on Raman spectroscopy. A setup for remote Raman spectroscopic tissue characterization in the high-wavenumber region ( approximately 2400-3800 cm(-1)) is presented. It makes use of a single, unfiltered, optical fiber for guiding laser light to the sample and for collecting scattered light and guiding it back to a spectrometer. Such a simple configuration is possible because the fused-silica core and cladding of the fiber present almost no Raman background signal at these wavenumbers. Several commercially available optical fibers were tested with respect to Raman signal background, to determine their suitability for in vivo Raman spectroscopy measurements in the high-wavenumber region. Different fiber core, cladding, and coating materials were tested. Silica core-silica clad fibers, with an acrylate coating and a black nylon jacket, proved to be one of the best candidates. In vitro measurements on brain tissue of a 6-month-old pig were obtained with a remote high-wavenumber Raman setup. They illustrate the low background signal generated in the setup and the signal quality obtained with a collection time of 1 s.