Effect of formalin fixation on the near-infrared Raman spectroscopy of normal and cancerous human bronchial tissues.

Raman spectroscopy is a vibrational spectroscopic technique that can be used to probe molecular changes associated with tissue malignancy. In this report, the effect of formalin fixation on human bronchial tissues was studied by near-infrared (NIR) Raman spectroscopy to determine if the variations of Raman spectra caused by formalin fixation would affect the potential diagnostic ability for lung cancer detection. A rapid dispersive-type NIR Raman system with an excitation wavelength of 785 nm was used for this study. Bronchial tissue samples were obtained from six patients with known or suspected malignancies of the lung. Raman spectra of fresh normal and tumor tissue were compared with spectra of formalin-fixed normal and tumor tissue. Changes of the ratios of Raman intensities at 1445 to 1655 cm(-1) and 1302 to 1265 cm(-1) versus formalin fixing times varying from 2 to 24 h were also examined. The major tissue Raman peaks at 1265, 1302, 1445, and 1655 cm(-1) were found in both fresh and fixed bronchial tissues. However, bronchial tissue preserved in formalin showed a progressive decrease in overall intensities of these Raman peaks. Raman contaminations due to formalin were also found in the 980-1100, and 1480-1650 cm(-1) ranges with notable formalin peaks (1041 and 1492 cm(-1)) appearing in the fixed normal and tumor tissues. The results showed that NIR Raman spectra of human bronchial tissues were significantly affected by formalin fixing and tissue hydration. Diagnostic markers at the 980-1100, and 1500-1650 cm(-1) regions derived from fixed tissues do not appear to be applicable for in vivo lung cancer detection. To yield valid Raman diagnostic information for in vivo applications, fresh tissue should be used. If only fixed tissue is available, thorough rinsing of specimens in phosphate-buffered saline (PBS) before spectral measurements may help reduce the formalin fixation artifacts on tissue Raman spectra.