The Massachusetts Institute of Technology Spectroscopy Laboratory`s Laser Biomedical Research Center (Boston, MA) has been awarded a five-year grant totaling $4.1 million by the National Institutes of Health (NIH; Bethesda, MD) for research involving endoscopes and spectrometry to provide in situ spectra of human tissue. Such techniques may lead to early detection of certain cancers as well as quantitative biochemical tissue analysis for monitoring the progress of diseases like atherosclerosis (
Endospectroscopy offers early diagnoses
The Massachusetts Institute of Technology Spectroscopy Laboratory`s Laser Biomedical Research Center (Boston, MA) has been awarded a five-year grant totaling $4.1 million by the National Institutes of Health (NIH; Bethesda, MD) for research involving endoscopes and spectrometry to provide in situ spectra of human tissue. Such techniques may lead to early detection of certain cancers as well as quantitative biochemical tissue analysis for monitoring the progress of diseases like atherosclerosis (heart disease).
Endoscopes currently provide an internal view of hollow organs using white light. The images allow viewing of gross physical deformities resulting from disease. Diseased tissue, however, cannot always be distinguished from normal tissue, which means that a tissue sample must be removed for closer study (a biopsy).
Advantages of spectroscopy
Spectral pathology using endoscopes as spectroscopic probes could provide important diagnostic information that neither endoscopy or histology alone can provide. The spectra of diseased tissue differ from those of normal tissue, and diseased tissue typically undergoes biochemical changes before deformities occur that would be visible through an endoscope.
Moreover, biopsies are invasive, and only a limited amount of tissue is removed from a randomly chosen location--possibly not representing the diseased area. "Sample preparation takes hours, at best," says Michael Feld, director of the Laser Biomedical Research Center. In other cases, biopsies are difficult or impossible. Endospectroscopy, however, looks at tissue in situ. "Instead of bringing the tissue to the instrument, a part of the instrument is taken to the tissue," says Feld. Another advantage of the method is that analysis and feedback can occur in real time.
The researchers have used re flectance, absorption, IR absorption, Raman, and laser-induced fluorescence spectroscopy through endoscopes. Low-intensity laser light is transmitted through a fiber inserted into the accessory channel of a standard endoscope. Light is collected by the endoscope, dispersed in a spectrograph, and detected using an optical multichannel analyzer or CCD camera. An advantage of Raman spectroscopy is that it is molecule-specific, while laser-induced fluorescence spectroscopy offers strong and sensitive signals.
Work at the research center has applied fluorescence to diagnosis of mucosal dysplasia--a precancerous ste¥that occurs in the mucosal surfaces of the gastrointestinal tract, oral cavity, bladder, cervix, and the lungs. "In the USA alone," explains Feld, "cancers originating in the mucosal surfaces account for a quarter of a million deaths each year, and it is estimated that upwards of 90% of these could be prevented by early detection and treatment." Spectral pathology could provide early detection. Flat dysplastic tissue in the colon, for example, is not visibly different from normal tissue, but, when excited by 370-nm laser light during a colonoscopy, the tissue exhibits a fluorescence spectrum different from normal tissue (see figure). Similarly, 410-nm laser light excites fluorescence that can show dysplastic and malignant oral mucosa, while 400-nm light can hel¥recognize transitional cell carcinoma in the bladder.
The researchers have developed empirical algorithms for discriminating between dysplastic and normal tissue, and in a test of 69 samples, only six were misdiagnosed--an accuracy comparable to that of an expert pathologist, according to Feld. Furthermore, they have been able to determine which parts of the spectrum originate in various microscopic cell features. "This is the first time that a clinical fluorescence spectrum has been explained in terms of its microscopic constituents," says Feld.
Diagnosing heart disease
Diagnosis of heart disease, the leading cause of death in the USA, can also be improved by endospectroscopy. Plaque builds u¥on arterial walls, becomes unstable, and may break loose, cutting off blood flow to the heart and leading to a heart attack. "Atherosclerotic plaque can be distinguished from normal arterial tissue using 476-nm excitation and contrasting the fluorescence of structural protein to that of ceroid, a complex oxidized lipoprotein characteristic of diseased vessels," reports Feld and coworkers John Kramer and Ramachandra Dasari.1
Standard techniques for diagnosing atherosclerosis cannot sense the biochemical changes in the plaque that lead to instability. Raman endospectroscopy, however, can not only distinguish between normal arterial walls and plaque, but also can provide quantitative measurements of the artery`s composition. Clinical systems with 830-nm diode lasers can characterize plaque in vivo in terms of its major constituents, including free cholesterol, cholesterol esters, triglycerides, phospholipids, and calcium salts. The amount of each can be quantified with the accuracy of a traditional biochemical assay, but without a biopsy. "Raman endospectroscopy," say the researchers, "may make sudden plaque rupture a predictable event."
Future work in endospectroscopy includes improving endoscopes and diagnostic algorithms, as well as adding multiple-wavelength excitation capability.
YVONNE CARTS-POWELL is a technical writer based in Belmont, MA.
1. John Kramer, Ramachandra Dasari, and Michael Feld, Nature: Medicine 2(10), 1079 (Oct. 1996).