Image Credits: Sebastian Kaulitzki/shutterstock.com
Molecules scatter photons of light in two ways, either elastically or inelastically. With elastic scattering the frequency and wavelength of light does not change significantly. Some photons are scattered inelastically, because they transfer some energy to intramolecular bonds and in the process shift to another frequency, a phenomenon called the Raman effect.
The Raman scattering can be measured using a highly sensitive spectrometer and narrowband laser excitation. The Raman spectrum of each molecule shows how each bond interacts with the light. The Raman spectral peaks are a function of specific vibrational motion modes that correlate with a particular type of bond. Some of these modes are stretching, bending or scissoring. In the H-C=C-bond, for instance, the H-C=bond shows stretching vibrational motion, as seen in biological tissue which contains a lot of fat. In other words, the Raman spectra help to identify the chemical composition of a tissue, by producing a vibrational fingerprint consistent with the sum of the individual vibrational spectra produced by the protein, fat and nucleic acids present in it.
Raman Spectroscopy and the Brain
In the brain, Raman spectra produced by tumor cells are different from those acquired from normal cells. This allows accurate identification of brain tumors both pre- and intra-operatively. The fact that normal and edematous brain tissue could be distinguished by a difference in Raman spectra was known as far back as 1990 using the O-H stretching mode. However, this technique lacks sensitivity and there is considerable background interference which hinders its use.
Vital brain tissue can be differentiated from necrotic tissue quite easily and with almost complete accuracy by Raman spectroscopy. This enables guided brain biopsy in real-time, and also helps define the margins of excision of brain tumors. This distinction is because of the high cholesterol levels in necrotic cells which affect their Raman spectra. Invasive tumor margins can be differentiated from the actual tumor as well using this technique.
Other techniques which depend upon the use of Raman spectroscopy for brain studies include color mapping depending upon the relative intensity of each chemical component, such as red for a high protein concentration. These are able to differentiate between white matter, gray matter and tumor cells. While relatively simple, this multichannel color imaging technology can diagnose tumors with up to 90% accuracy.
Today fiberoptic probes are used for interoperative Raman spectroscopy using a handheld instrument that is held against the brain surface or resected tissue cavity. The spectral peaks are analyzed to tell whether the tissue in the next millimeter of brain tissue below the surface in contact with the probe is normal or invaded by tumor cells, provided 15% or more invasion is present. This promises to be quite accurate (92%) with a high sensitivity and specificity of above 91%.
Raman spectroscopy has some important limitations such as the small fraction of inelastic scattering, which in turn requires a long period of image acquisition and results in low resolution, as well as increased artifact. To optimize the use of this technique, coherent Raman spectroscopy (CRS) microscopy is used in tandem to increase the intensity of the generated signals, using coherent anti-Stokes Raman scattering (CARS) and stimulated Raman scattering (SRS).
These provide a second beam to produce coherent vibrational movement of the chemical bonds. The result is stronger signals, over 10 000 orders of magnitude more than spontaneous Raman scattering, so that the final resolution is much higher, up to submicron scales, and the speed of imaging increases to near-video rates.
Image Credits: HQuality/shutterstock.com
CRS microscopy allows the specimen to be examined without thin slicing because the nonlinear manner of biomolecular excitation results in inherent 3D slicing. It allows image selection based on different vibrational modes, while the various spectral peaks distinguish lipid-rich from protein-rich cells. Broadband CARS is even more rapid and sensitive but makes use of a wider part of the spectrum within the characteristic molecular fingerprint region.
The advantages are the ability to perform the analysis even in the presence of water, the compatibility of the fiberoptic probe with a curved path so that a spectrum can be acquired without destruction of tissue, and finally, a rapid and accurate real-time method of tissue assessment for intraoperative decision making. This is possible because the sample does not need special preparation or labeling, and can thus be assessed during surgery without interrupting the surgery; the diagnosis is made rapidly; and the procedure is completely non-invasive. Another advantage is that it reduces the total number and expense of testing to arrive at the right diagnosis by other means, such as FISH, electron microscopy or immunohistochemical staining. It has been validated to show perfect concurrence with standard histopathology results following H&E stains.
All these techniques are important in reducing the incidence and death rate from brain tumors by revealing the infiltrative margin. This enables more accurate tumor removal without endangering neurological function and even the patient’s life by unnecessarily removing healthy tissue. At the same time, it avoids residual tumor cells which could lead to a relapse, unless it is impossible to achieve complete tumor excision surgically.
Another use of Raman spectroscopy is to guide a second biopsy following a failed first one with no tumor cells picked up. It can also locate the right site for radiation therapy. It can also provide a very accurate means of diagnosing malignancy intraoperatively during surgery for pediatric brain tumors.