New approaches to cervical cancer screening

By Anne-Thérèse Vlastos, MD, Rebecca Richards-Kortum, PhD, Andres Zuluaga, PhD, and Michele Follen, MD, PhD

While colposcopy remains an essential diagnostic tool in the hands of an experienced clinician, the five new technologies outlined here may eventually replace or at least supplement this important procedure.

Most epithelial cancers—including those of the cervix, urinary bladder, respiratory tract, oral mucosa, and skin—are preceded by precancerous changes. Yet despite the presence of these precursor lesions, cancer morbidity and mortality remains significantly high. Cervical cancer, for instance, is the third most common malignancy in women and one of the leading causes of death in women worldwide.¹ The disease takes an especially heavy toll on women younger than 45 years, with 44% of cases occurring in this age group.²

Early detection provides the best opportunity to improve patient survival and quality of life. In fact, the 5-year survival rate for early cervical cancer is currently 91%, compared with less than 5% for stage IV lesions.²

The introduction of Pap smear-based screening programs more than 50 years ago, led to a substantial decrease in cervical cancer mortality.^3,4 The main disadvantages of Pap screening, however, are the high rate of false positives, the need for an experienced cytologist, and the long wait for results. In addition, in cases of abnormal findings, colposcopic evaluation becomes the standard of care. And while this procedure is accurate, it adds to patient expenses and anxiety and requires considerable skills to perform.⁵

The morphologic alterations that characterize preneoplastic changes, including increased nuclear size, increased nuclear/cytoplasmic ratio, hyperchromasia, and pleomorphism, are well known. New devices that can better detect these changes yet lower detection costs are needed. During the last decade, the potential of such diagnostic optical imaging techniques as fluorescence spectroscopy, reflectance spectroscopy, Raman spectroscopy, optical coherence tomography (OCT), and confocal microscopy has generated considerable interest. These techniques are based on biophysical properties of light and tissue that can be manipulated or observed to predict diagnosis. All are currently being evaluated to determine their ability to provide information similar to or supplemental to that provided by methods such as colposcopy. These new technologies are also being investigated to determine if they can serve as immediate, real-time diagnostic tools. Researchers are aiming to make these new techniques user-friendly enough to be operated by health-care providers other than physicians, an approach that should help broaden their use and reduce health-care costs.

Fluorescence spectroscopy

Fluorescence spectroscopy has the potential to improve the accuracy and efficacy of screening programs for cervical intraepithelial neoplasia.^6-12 A safe, noninvasive, and real-time technique, laser-induced fluorescence spectroscopy assesses the autofluorescence of a tissue by measuring the amounts of naturally present fluorophores like NADH, FADH, elastin, and collagen, and absorbers such as hemoglobin.^9,13,14 The shape of the fluorescence spectrum differs among normal, preneoplastic, and neoplastic lesions, depending on the number of fluorophores in the tissue (Figure 1).^9,14

Click here to view full-size graphic

In practice, a monochromatic light of either ultraviolet or visible wavelengths is applied to the cervix through optical fibers. The emitted fluorescence, which represents a fingerprint of the tissue's biochemical properties, is then collected and measured quantitatively as a function of the wavelength used.¹⁵ Both ultraviolet and visible light are adequate for detection of intraepithelial lesions, but their penetration is limited to only several hundred microns.¹⁶ With little or no provider training, fluorescence spectroscopy could identify lesions before they are visible to the naked eye or by white-light scoping.⁵

Fluorescence spectroscopy uses emission spectra collected at excitation wavelengths of 337, 380, and 460 nm. These data in turn are incorporated into multivariate statistical algorithms that have been developed for differentially diagnosing squamous intraepithelial lesions (SILs) in vivo in the colposcopy clinic (Figure 2).^7-10 The resulting technique can be performed without a priori clinical information, and its sensitivity (86%) and specificity (74%) compare favorably with those of colposcopy (96% and 48%) and Pap smear (62% and 68%) (Figures 3­5, Table 1).^6,10,17-21

 

TABLE 1 Comparing Pap smear, colposcopy, fluorescence spectroscopy, and Raman spectroscopy
Techniques	Sensitivity (%)	Specificity (%)
Pap smear	62	68
Colposcopy	96	48
Fluorescence spectroscopy	86	74
Raman spectroscopy	82	92

Click here to view full-size graphic

Click here to view full-size graphic

Click here to view full-size graphic

Click here to view full-size graphic

Fluorescence spectroscopy also appears to work well when screening a population with a low prevalence of disease.^12,22 The fluorescence spectra of colposcopically normal cervical sites differ in the presence or absence of SIL.²² Yet, despite these differences, Brookner and associates found fluorescence spectroscopy to have sensitivities of 87% for squamous epithelium, 96% for columnar epithelium, and 78% for the transformation zone in the screening setting.²² These results surpass those achieved in the study in which these algorithms were developed and are comparable to the sensitivity of colposcopy by an experienced colposcopist.^5,7-10

Recent meta-analyses have confirmed that fluorescence spectroscopy in both screening and diagnostic settings is better than colposcopy, cervicography, human papillomavirus (HPV) testing, and cervicoscopy.^11,12 Receiver operating characteristic (ROC) curves and areas under ROC curves have been used to evaluate and compare medical technologies. The ROC curve can be used in both diagnostic settings (high prevalence of disease) (Figure 6) and screening settings (low prevalence of disease) (Figure 7).^23-26

Click here to view full-size graphic

Click here to view full-size graphic

In a meta-analysis by Mitchell and colleagues, ROC curves calculated from published reports for colposcopy, Pap smear screening, cervicography, speculoscopy, and HPV testing were compared with ROC curves for diagnostic fluorescence spectroscopy generated from measurements performed with a research device at M. D. Anderson Cancer Center.¹¹ Compared with other traditional methods, colposcopy produces the largest areas under the ROC curve. Having a larger area under the ROC curve is an indication of better performance across many screening and diagnostic settings. This means that the greater the area under the ROC curve, the better the device.

Therefore, it is important to recognize that fluorescence spectroscopy appears to produce an even higher area under the curve than colposcopy in the diagnostic setting and than Pap smear testing in the screening setting.^11,12,23-26 This would permit a large number of trained personnel, including nurse practitioners or even non-degreed health-care workers in developing countries, to use the technique. Moreover, fluorescence spectroscopy would provide immediate diagnosis, allowing easy evaluation and treatment at a single visit.

The fluorescence emission spectra are subjected to an analysis that reduces the data set to the biologically relevant data (a process called principal component analysis). In our studies, each fluorescent EEM is compared with blinded readings of biopsy specimens taken from the same sites that are measured. The promising algorithms discussed here are currently being tested in the context of a large clinical trial comparing a prototype device with colposcopy in experienced hands. Several groups have attempted to use fluorescence spectroscopy to diagnose preneoplasias and neoplasias of the colon, lung, and cervix.^{6,9,18,19,27-34} Once good results are confirmed, a challenge to industry will be to build a low-cost prototype that would permit widespread clinical application.³⁵

Raman spectroscopy

One shortcoming of fluorescence spectroscopy is the fact that fluorescence spectra may not easily differentiate between benign abnormalities like inflammation or metaplasia and preneoplastic lesions.^8,29 This is why vibrational Raman spectroscopy is considered a way to enhance the specificity of spectroscopic diagnosis.

The technique is based on Raman, or inelastic, scattering, which arises from molecule- induced vibrational or rotational transitions.³⁶ When a photon of light makes contact with a molecule inside tissue, energy is transferred from one to the other. The scattered photon has a different vibrational energy than the incident photon, that is, the photon entering the tissue. Raman spectroscopy analyzes the scattered intensity as a function of the energy difference between the incident and scattered photons. The characteristic spectra that result represent shifts in wave numbers from the incident frequency (Figure 8).³⁶ (Wave numbers are measured as cm^-1 and represent the inverse of the wavelength.)

Click here to view full-size graphic

While most biological molecules are not fluorescent, they are all Raman active and have fingerprint spectra in the near infrared. Mahadevan-Jansen and associates have used near infrared Raman (NIR) spectroscopy for the differential diagnosis of cervical precancers in vitro and in vivo.^36-38 With the help of empirical intensities and multivariate statistical algorithms, these investigators used in vitro NIR spectroscopy to diagnose samples of preneoplastic cervical lesions adjacent to inflamed, metaplastic, or normal areas of the cervix. The technique had a sensitivity and specificity of 82% and 92% or better, respectively.³⁷ The success of these in vitro Raman studies led the same group to develop a fiber-optic probe to measure the NIR spectra of cervical tissue in vivo. The in vivo Raman spectra could be obtained within 90 seconds and appeared similar to the in vitro Raman spectra.³⁸

Raman spectroscopy has also been investigated for its ability to detect cancers in the breast, brain, colon, bladder, uterus, and ovary.^39-41 Larger studies are needed to validate this promising technique.

Reflectance spectroscopy

Reflectance spectroscopy uses the biochemical composition and structure of a tissue to determine the tissue's specific optic spectrum. Because pathological tissues exhibit significant architectural changes at the cellular level, we can use reflectance spectroscopy to screen for and diagnose disease (Figure 9).⁴² In theory, if incident photon energy is unaltered after collision with a molecule, the frequency of the scattered photon is the same as the incident photon. This is referred to as Rayleigh, or elastic, scattering.^43,44

Click here to view full-size graphic

The elastic-scattering properties of tissue may give structural and morphological information about the tissue. When performed with such contact probes as optic fibers, reflectance spectroscopy is closely related to elastic-scattering spectroscopy.⁴⁵ Probes have been developed to measure tissue reflectance and to take measurements at specified pressures.⁴⁶

Reflectance spectroscopy has been investigated in skin and cervix.^47-52 The driving principle behind its use in the cervix is that chromatin within cells reflect light in a predictable manner: the denser the chromatin, the greater the scattering.⁵² Because chromatin content increases as cells become more dysplastic, measuring chromatin is important. This technique is being investigated in a large prospective clinical trial.

Optical coherence tomography

Another noninvasive method for improving on colposcopy is optical coherence tomography (OCT). This technique images tissue cross-sections by measuring their optical reflections. (An analogous technique is ultrasonography, which uses sound waves (Figure 10).⁵³) An image of the internal or subsurface tissue can then be reconstructed by tomography. Basically, a low-coherence light (e.g., = 852 nm) illuminates a Michelson interferometer—which consists of a light source split by a beam splitter into a measurement arm and a reference arm—and a light detector.^54-56 The system measures the amplitude and longitudinal delay of the backscattering of the light. A two-dimensional map of reflection is constructed by assembling multiple longitudinal one-dimensional scans of the sample.^54,57-61 OCT is able to detect very faint reflections of laser light directed into a tissue and determine at what depth these reflections occur.

Click here to view full-size graphic

The only factor limiting OCT resolution is the coherence length of the light source; however, high-depth resolution can be maintained even through a small aperture.⁵⁴ Thus, this technique holds promise for many medical uses, particularly in endoscopic imaging.^54,62-64 The image corresponding to the relative reflectivity of the tissue is related to the properties of individual cells, as well as to the overall structure of the tissue, both of which change in the presence of disease.

Clinically viable OCT systems are currently being evaluated. By and large, they are capable of imaging to a depth of 1 millimeter at a resolution of 10 to 20 µm, making them of great interest for screening epithelial lesions.^54,65-69 To investigate cervical intraepithelial neoplasia, Zuluaga and colleagues have built a transportable, completely self-contained instrument to obtain OCT images in vivo during colposcopic examinations.⁵⁴ The instrument, a 250-mm fiber-optic probe with a rigid end piece 18 mm in diameter, is placed in contact with the cervix. In 3 seconds, an image is obtained with or without chemical agents such as 6% acetic acid, toluidine blue dye, or Lugol's iodine solution. The images obtained show the subsurface morphology of an area of cervical tissue 2 mm in diameter (Figure 11).

Click here to view full-size graphic

OCT's potential for performing optical biopsy, combined with attractive features like noninvasive probing, short acquisition time for image construction, and high spatial resolution, has attracted considerable attention from researchers and physicians.⁶⁴ Many researchers, for instance, are interested in studying OCT's use in the female reproductive tract.^70-72 The potential disadvantages of OCT are its small penetration depth and the poor contrast it provides between structures with similar backscattering coefficients. If these shortcomings can be overcome, OCT may become a powerful tool for screening and detecting cervical SILs.

Confocal microscopy

Confocal microscopy is a potentially useful tool for diagnosis because it provides near-histologic and three-dimensional images of in vivo sites without the need for removing tissue. Owing to its ability to reject light from outside the focal volume, confocal microscopy is especially well suited for the noninvasive evaluation of thick tissue and assessment of cell morphology using reflected light.^73-80 With a spatial resolution of 1 to 2 µm, the technique can create images of single cells and their nuclei. To localize reflected light in three dimensions, confocal microscopy employs a pinhole placed on a conjugate image plane (Figure 12).⁷⁹ To visualize intracellular detail, the microscope provides contrast by changes in refractive index. Similar to histological analysis in concept, this technique is of special interest for screening and diagnosing of SILs.⁸⁰

Click here to view full-size graphic

Refractive index fluctuations provide the primary source of contrast for imaging cell structure. To date, in vivo confocal imaging has focused on the skin and the eye, where melanin is the source of contrast.^{73-75,77-79,81,82} Although amelanotic cells have inherently less contrast on confocal imaging than do melanotic cells, treating epithelial cells or tissue with a contrast agent like weak acetic acid can increase nucleus-related backscattering, thus providing additional contrast on confocal images.^52,80,83,84

As most ob/gyns know, within 2 to 3 minutes of applying a 3% to 6% acetic acid solution, SILs turn whitish, a phenomenon known as acetowhitening of the cervix.⁸⁵ As Collier and associates have recently shown in vitro, this can increase contrast on confocal images.⁸⁰ In fact, cell nuclei can be visualized throughout the entire epithelium within seconds after acetic acid application. The resulting image contrast is sufficient to visualize nuclear segmentation.⁸⁶ Nuclear deformity as well as the nuclear/ cytoplasmic ratio can then be calculated.^52,86 Since early preneoplastic changes are characterized by changes in the nuclear size and nuclear/cytoplasmic ratio, confocal imaging following application of acetic acid could eliminate the need for painful biopsies. Since acetic acid is already routinely used in colposcopy, this method could be easily implemented for cervical screening in the clinic (Figure 13).

Click here to view full-size graphic

Confocal microscopy has long been used in vitro, and applications for its endoscopic and in vivo use have long been under development.^87-89 However, many epithelial sites at risk for SIL need to be assessed by fiber-optic endoscopy instead of conventional confocal microscopy. Consequently, several fiber-optic confocal microscopes have already been developed for this purpose, though all of them suffer from a lack of reflected light and thus lack sufficient contrast in vivo.^90-94 Acetic acid could potentially increase the backscattered light from the nuclei, which might improve the usefulness of these real-time fiber-optic confocal endoscopes.

Drezek and colleagues recently reported the use of a video rate confocal microscope with micron resolution to create images of cervical cells plus colposcopically normal and abnormal cervical biopsy samples.⁹⁵ Images were obtained before and after application of a 6% acetic acid solution. The confocal imaging system resolved subcellular detail throughout the entire epithelial thickness. Normal and dysplastic tissues could be clearly differentiated. The addition of acetic acid dramatically enhanced the nuclear signal in all acquired images. As this study shows, high-contrast reflected light images of cervical tissue are attainable in near real time.

Conclusions

While we have made great strides over the decades in reducing cervical cancer morbidity and mortality, to make further inroads, we need new technologies that allow the identification, immediate diagnosis, and treatment of cervical cancer or its precursors in a single outpatient visit. These new technologies should be able to outperform or keep pace with colposcopy, the standard of care.

Effective colposcopy requires an experienced colposcopist, a high volume of patients, and good interaction between pathologist and primary-care provider. Future technologies should aim to provide real-time diagnosis through inexpensive and easy-to-use devices. Fluorescence spectroscopy, Raman spectroscopy, reflectance spectroscopy, OCT, and confocal microscopy are promising because they are reliable and easy to use. Nevertheless, they need to be subjected to clinical trials and cost-effectiveness analyses. Because all five technologies have different biological bases, they may complement each other in use. All have the potential to play a key role in the future diagnosis and prevention of cervical SIL.

Drs. Follen and Richards-Kortum hold patents for some of the products discussed in this article, including patents licensed by Lifespex. None are yet commercially available.

REFERENCES

1. Parkin DM, Pisani P, Ferlay J. Estimates of the worldwide incidence of 25 major cancers in 1990. Int J Cancer. 1999;80:8274-8241.

2. Ries LA, Kosary CL, Hankey BF, et al. SEER Cancer Statistics Review, 1973-1996. National Cancer Institute, Bethesda, MD, 1999.

3. Koss LG. The Papanicolaou test for cervical cancer detection. A triumph and a tragedy. JAMA. 1989;261:737-743.

4. Kurman RJ, Henson DE, Herbst AL, et al. Interim guidelines for management of abnormal cervical cytology. The 1992 National Cancer Institute Workshop. JAMA. 1994;271:1866-1869.

5. Mitchell MF, Schottenfeld D, Tortolera-Luna G, et al. Colposcopy for the diagnosis of squamous intraepithelial lesions: a meta-analysis. Obstet Gynecol. 1998;91:626-631.

6. Mahadevan A, Mitchell MF, Silva E, et al. Study of the fluorescence properties of normal and neoplastic human cervical tissue. Lasers Surg Med. 1993;13:647-655.

7. Ramanujam N, Mitchell MF, Majadevan A, et al. Development of a multivariate statistical algorithm to analyze human cervical tissue fluorescence spectra acquired in vivo. Lasers Surg Med. 1996;19:46-62.

8. Ramanujam N, Mitchell MF, Mahadevan A, et al. Spectroscopic diagnosis of cervical intraepithelial neoplasia (CIN) in vivo using laser-induced fluorescence spectra at multiple excitation wavelengths. Lasers Surg Med. 1996;19:63-74.

9. Ramanujam N, Mitchell MF, Mahadevan A, et al. In vivo diagnosis of cervical-intraepithelial neoplasia using 337-nm-excited laser-induced fluorescence. Proc Natl Acad Sci USA. 1994;91:10193-10197.

10. Ramanujam N, Mitchell MF, Mahadevan-Jansen A, et al. Cervical precancer detection using a multivarite statistical algorithm based on laser-induced fluorescence spectra at multiple excitation wavelengths. Photochem Photobiol. 1996;64:720-735.

11. Mitchell MF, Cantor SB, RamanuJam N, et al. Fluorescence spectroscopy for diagnosis of squamous intraepithelial lesions of the cervix. Obstet Gynecol. 1999;93:462-470.

12. Mitchell MF, Cantor SB, Brookner C, et al. Screening for squamous intraepithelial lesions with fluorescence spectroscopy. Obstet Gynecol. 1999;94:889-896.

13. Brookner CK, Agrawal A, Trujillo EV, et al. Safety analysis: relative risks of ultraviolet exposure from fluorescence spectroscopy and colposcopy are comparable. Photochem Photobiol. 1997;65:1020-1025.

14. Richards-Korturn R, Mitchell MF, Ramanujam N, et al. In vivo fluorescence spectroscopy: potential for noninvasive, automated diagnosis of cervical intraepithelial neoplasia and use as a surrogate endpoint biomarker. J Cell Biochem Suppl. 1994;19:111-119.

15. Lakowicz JR. Principles of fluorescence spectroscopy. New York, NY: Plenum Press; 1983.

16. Andersson-Engels S, Johansson J, Svanberg K, et al. Fluorescence imaging and point measurements of tissue: applications to the demarcation of malignant tumors and atherosclerotic lesions from normal tissue. Photochem Photobiol. 1991;53:807-814.

17. Fahey MT, Irwig L, Macaskill P. Meta-analysis of Pap test accuracy. Am J Epidemiol. 1995;141:680-689.

18. Mitchell MF. Accuracy of colposcopy. Consult Obstet Gynecol. 1994;6:70-73.

19. Ramanujam N, Mitchell MF, Mahadevan A, et al. Fluorescence spectroscopy: a diagnostic tool for cervical intraepithelial neoplasia. Gynecol Oncol. 1994;52:31-38.

20. Tumer K, Ramanujam N, Ghosh J, et al. Ensembles of radial basis function networks for spectroscopic detection of cervical precancer. IEEE Trans Biomed Eng. 1998;45:953-961.

21. Utzinger U, Trujillo EV, Atkinson EN, et al. Performance estimation of diagnostic tests for cervical precancer based on fluorescence spectroscopy: effects of tissue type, sample size, population, and signal-to-noise ratio. IEEE Trans Biomed Eng. 1999;46:1293-1303.

22. Brookner CK, Utzinger U, Staerkel G, et al. Cervical fluorescence of normal women. Lasers Surg Med. 1999;24:29-37.

23. Metz CE. Basic principles of ROC analysis. Semin Nucl Med. 1978;8:283-298.

24. Littenberg B, Moses LE. Estimating diagnostic accuracy from multiple conflicting reports: a new meta-analytic method. Med Decis Making. 1993;13:313-321.

25. Swets JA. Measuring the accuracy of diagnostic systems. Science. 1988;240:1285-1293.

26. Moses LE, Shapiro D, Littenberg B. Combining independent studies of a diagnostic test into a summary ROC curve: data-analytic approaches and some additional considerations. Stat Med. 1993; 12:1293-13 16.

27. Richards-Korturn R, Rava RP, Petras RE, et al. Spectroscopic diagnosis of colonic dysplasia. Photochem Photobiol. 1991;53:777-786.

28. Cothren RM, Richards-Korturn R, Sivak MV Jr, et al. Gastrointestinal tissue diagnosis by laser-induced fluorescence spectroscopy at endoscopy. Gastrointest Endosc. 1990;36:105-111.

29. Schomacker KT, Frisoli JK, Compton CC, et al. Ultraviolet laser-induced fluorescence of colonic tissue: basic biology and diagnostic potential. Lasers Surg Med. 1992;12:63-78.

30. Hung J, Lam S, LeRiche JC, et al. Autofluorescence of normal and malignant bronchial tissue. Lasers Surg Med. 1991;11:99-105.

31. Lam S, MacAulay C, Hung J, et al. Detection of dysplasia and carcinoma in situ with a lung imaging fluorescence endoscope device. J Thorac Cardiovasc Surg. 1993; 105:1035-1040.

32. Lam S, MacAulay C, Palcic B. Detection and localization of early lung cancer by imaging techniques. Chest. 1993;103(l suppl):12S-14S.

33. Glassmen WS, et al. Ultraviolet excited fluorescence spectra from non-malignant tissues of the gynecological tract. Lasers Life Sci. 1992;5:49-58.

34. Lohmann W, Mussmann J, Lohmann C, et al. Native fluorescence of the cervix uteri as a marker for dysplasia and invasive carcinoma. Eur J Obstet Gynecol Reprod Biol. 1989;31:249-253.

35. Cantor SB, Mitchell MF, Tortolero-Luna G, et al. Cost-effectiveness analysis of diagnosis and management of cervical squamous intraepithelial lesions. Obstet Gynecol. 1998;91:270-277.

36. Mahadevan-Jansen A, Richards-Korturn R. Raman spectroscopy for cancer detection: A review. Paper presented at: Proceedings of the 19th International Conference of IEEE/EMBS, 1997.

37. Mahadevan-Jansen A, Mitchell MF, Ramanujam N, et al. Near-infrared Raman spectroscopy for in vitro detection of cervical precancers. Photochem Photobiol. 1998;68:123-132.

38. Mahadevan-Jansen A, Mitchell MF, Ramanujam N, et al. Development of a fiber optic probe to measure NIR Raman spectra of cervical tissue in vivo. Photochem Photobiol. 1998;68:427-431.

39. Frank CJ, McCreery RL, Redd DC. Raman spectroscopy of normal and diseased human breast tissues. Anal Chem. 1995;67:777-783.

40. Manoharan R, Shafer K, Perelman L, et al. Raman spectroscopy and fluorescence photon migration for breast cancer diagnosis and imaging. Photochem Photobiol. 1998;67:15-22.

41. Liu CH, Das BB, Sha Glassman WL, et al. Raman, fluorescence, and time resolved light scattering as optical diagnostic techniques to separate diseased and normal biomedical media. J Photochem Photobiol B. 1992;16:187-209.

42. Dunn A, Richards-Korturn R. Three-dimensional computation of light scattering from cells. IEEE J Selected Topics in Quantum Electronics. 1996;2: 898-905.

43. Colthup NB, Daly LH, Wiberley SE, et al. Introduction to infrared and Raman spectroscopy. Burlington, Mass: Academic Press; 1990.

44. Bigio IJ, Mourant JR. Ultraviolet and visible spectroscopies for tissue diagnostics: fluorescence spectroscopy and elastic-scattering spectroscopy. Phys Med Biol. 1997;42:803-814.

45. Sato N, Matsumura. T, Shichiri M, et al. Hemoperfusion, rate of oxygen consumption and redox levels of mitochondrial cytochrome c (+c1) in liver in situ of anesthetized rat measured by reflectance spectrophotometry. Biochim Biophys Acta. 1981;634:1-10.

46. Ono K, Kanda M, Hiramoto J, et al. Fiber optic reflectance spectrophotometry system for in vivo tissue diagnosis. Applied Optics. 1991;30:98-105.

47. Anderson RR, Parrish JA. The optics of human skin. J Invest Dermatol. 1981;77:13-19.

48. Marchesini R, Cascinelli N, Brambilla, M, et al. In vivo spectrophotometric evaluation of neoplastic and non-neoplastic skin pigmented lesions. II: Discriminant analysis between nevus and melanoma. Photochem Photobiol. 1992;55:515-522.

49. Dawson JB, Barker DJ, Ellis DJ, et al. A theoretical and experimental study of light absorption and scattering by in vivo skin. Phys Med Biol. 1980;25:695-709.

50. Feather JW, Hajizadeh-Saffar M, Leslie G, et al. A portable scanning reflectance spectrophotometer using visible wavelengths for the rapid measurement of skin pigments. Phys Med Biol. 1989;34:807-820.

51. Hajizadeh-Saffar M, Feather JW, Dawson JB. An investigation of factors affecting the accuracy of in vivo measurements of skin pigments by reflectance spectrophotometry. Phys Med Biol. 1990;35:1301-1315.

52. Drezek RA, Collier T, Brookner CK, et al. Laser scanning confocal microscopy of cervical tissue before and after application of acetic acid. Am J Obstet Gynecol. 2000;182:1135-1139.

53. Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography. Science. 1991;254:1178-1181.

54. Zuluaga A, et al. Optical coherence tomography: a new tool for detecting sub surface epithelial abnormalities in vivo. Paper presented at: Eurogin 2000: Fourth International Multidisciplinary Congress; April 5-9,2000,2000; Paris.

55. Masters BR. Early development of optical low-coherence reflectometry and some recent biomedical applications. Journal of Biomedical Optics. 1999;4:236-247.

56. Fercher AF, Hitzenberger CK, Drexler W, et al. In vivo optical coherence tomography. Am J Ophthalmol. 1993;116:113-114.

57. Swanson EA, Izatt JA, Hee MR, et al. In vivo retinal imaging by optical coherence tomography. Optics Letters. 1993;18:1864-1866.

58. Schmitt JM, Knuttel A, Bonner RF. Low coherence interferometry for imaging microstructures within optically turbid tissues. Advances in Optical Imaging and Photon Migration: Technical Digest. 1994;190.

59. Bouma B, Tearney GJ, Fujimoto JG. High-resolution optical coherence tomographic imaging using a mode-locked Ti:AI203 laser source. Optics Letters. 1995;20:1486-1488.

60. Izatt JA, Hee MR, Owen GH, et al. Optical coherence microscopy in scattering media. Optics Letters. 1994;19:590-592.

61. Knuttel A, Boehlau-Godau M. Spatially confined and temporally resolved refractive index and scattering evaluation in human skin performed with optical coherence tomography. J Biomed Opt. 2000;5:83-92.

62. Feldchtein FI, Sergeev AM, Gelikonov V, et al. In vivo endoscopic OCT imaging of precancer and cancer states of human mucosa. Optics Express. 1997;1:432-440.

63. Sergeev A, et al. Melanin effect on light scattering in tissues: from electrodynamics of living cell to OCT imaging. SPIE. 1997;2981:58-63.

64. Izatt JA, et al. Ophthalmic diagnostics using optical coherence tomography. Proc Soc Photo Opt Instrum Eng. 1993;1877:136.

65. Swanson EA, Izatt JA, Hee MR. In vivo retinal imaging by using optical coherence tomography. Optics Letters. 1993;18:1864-1866.

66. Izatt J, Hee MR, Swanson EA. Micron-resolution biomedical imaging with optical coherence tomography. Optics Photonics News. 1993;4:14-18.

67. Clivaz X, Marquis-Weible F, Salathe RP, et al. High-resolution reflectometry in biological tissues. Optics Letters. 1992;17:4-6.

68. Yougquist RC, Caff S, Davies DEN. Optical coherence domain reflectometry: a new optical evaluation technique. Optics Letters. 1987;12:158.

69. Fujimoto JG, Boppart SA, Tearney GJ, et al. High resolution in vivo intra arterial imaging with optical coherence tomography. Heart. 1999;82:128-133.

70. Pitris C, Goodman A, Boppart SA, et al. High- resolution imaging of gynecologic neoplasms using optical coherence tomography. Obstet Gynecol. 1999;93:135-139.

71. Boppart SA, Goodman A, Libus J, et al. High resolution imaging of endometriosis and ovarian carcinoma with optical coherence tomography: feasibility for laparoscopic-based imaging. Br J Obstet Gynaecol. 1999; 106:1071-1077.

72. Herrmann JM, Brezinski ME, Bourna BE, et al. Two- and three-dimensional high resolution imaging of the human oviduct with optical coherence tomography. Fertil Steril. 1998;70:155-158.

73. Bertrand C, Corcuff P. In vivo spatio-temporal visualization of the human skin by real-time confocal microscopy. Scanning. 1994;16:150-154.

74. Masters BR, Aziz DJ, Gmitro AF, et al. Rapid observation of unfixed, unstained human skin biopsy specimens with confocal microscopy and visualization. J Biomed Opt. 1997;2:435-437.

75. Masters BR, Thar AA. Real-time scanning slit confocal microscopy of the in vivo human cornea. Applied Optics. 1994;33:695-701.

76. Jester JV, Andrews PM, Petroll WM, et al. In vivo, real-time confocal imaging. J Electron Microsc Tech. 1991;18:50-60.

77. Jester JV, Petroll WM, Garana RM, et al. Comparison of in vivo and ex vivo cellular structure in rabbit eyes detected by tandem scanning microscopy. J Microsc. 1992;165:169-181.

78. Corcuff P, Gonnord G, Pierard GE, et al. In vivo confocal microscopy of human skin: a new design for cosmetology and dermatology. Scanning. 1996;18:351-355.

79. Rajadhyaksha M, Grossman M, Esterowitz D, et al. In vivo confocal scanning laser microscopy of human skin: melanin provides strong contrast. J Invest Dermatol. 1995;104:946-952.

80. Collier T, Richards-Kortum R, Shen P, et al. Video confocal microscopy of amelanotic tissue: dynamics of aceto-whitening enable nuclear segmentation. Optics Express. 2000;6:40-48.

81. Gonzalez S, Rajadhyaksha M, Gonzalez-Serva A, et al. Confocal reflectance imaging of folliculitis in vivo: correlation with routine histology. J Cutan Pathol. 1999;26:201-205.

82. Rajadhyaksha M, Gonzalez S, Zavislan JM, et al. In vivo onfocal scanning laser microscopy of human skin 11: advances in instrumentation and comparison with histology. J Invest Dermatol. 1999;113:293-303.

83. Smithpeter C, Dunn A, Drezek R, et al. Near real time confocal microscopy of cultured amelanotic cells. sources of signal, contrast agents and limits of contrast. J Biomed Opt. 1998;3:429-436.

84. Smithpeter CL, Dunn AK, Richards-Kortum R, et al. Penetration depth limits of in vivo confocal reflectance imaging. Applied Optics. 1998;37:2749-2754.

85. Cartier R. Practical Colposcopy. 2nd ed. Stuttgart; New York, NY: Gustav Fischer; 1984.

86. Thiran JP, Macq B. Morphological feature extraction for the classification of digital images of cancerous tissues. IEEE Trans Biomed Eng. 1996;43:1011-1020.

87. Pawley JB. Handbook of Biological Confocal Microscopy. 2nd ed. New York, NY: Plenum Press; 1995.

88. Webb RH, Hughes GW, Delori FC. Confocal laser scanning ophthalmoscope. Applied Optics. 1987; 26:1492-1499.

89. White WM, Rajadhyaksha M, Gonzalez S, et al. Non-invasive imaging of human oral mucosa in vivo by confocal reflectance microscopy. Laryngoscope. 1999;109:1709-1717.

90. Gmitro AF, Aziz D. Confocal microscopy through a fiber-optic imaging bundle. Optics Letters. 1993;18:565-567.

91. Juskaitis R, Wilson T, Watson TF. Real-time white light reflection confocal microscopy using a fiber-optic bundle. Scanning. 1997;19:15-19.

92. Dabbs T, Glass M. Microscopy: fiber-optic confocal microscope: FOCON. Applied Optics. 1992;31:3030-3035.

93. Giniunas L, Juskaitis R, Shatalin SV. Endoscope with optical sectioning capability (Optical Activities in Industry). Applied Optics. 1993;32:2888-2890.

94. Dickensheets DL, Kino GS. Micromachined scanning confocal optical microscope. Optics Letters 1996;21:764-766.

95. Drezek R, et al. Laser scanning confocal microscopy of cervical tissue before and after application of acetic acid. Paper presented at: Eurogin 2000: Fourth International Multidisciplinary Congress; April 5-9, 2000; Paris.

Dr. Vlastos is from the Département de Gynécologie et Obstétrique, Laboratoire d'Hormonologie, Hôpitaux Universitaires de Genève, Geneva, Switzerland. A visiting scholar when this work was performed, Dr. Vlastos was at the Department of Gynecologic Oncology, The University of Texas M. D. Anderson Cancer Center, supported by the Swiss National Science Foundation, the Swiss Cancer League, the Cancer and Solidarity Foundation, and the Novartis Foundation.

Drs. Zuluaga and Richards-Kortum are from the Department of Biomedical Engineering, The University of Texas at Austin, Austin, Tex.

Dr. Follen is Professor of Gynecologic Oncology, Director of the Biomedical Engineering Center, and holder of the Boone Pickens Distinguished Professorship for Early Cancer Detection, The University of Texas M. D. Anderson Cancer Center, Houston, Tex. She also serves on the faculty of the Department of Obstetrics, Gynecology, and Reproductive Sciences. The University of Texas Health Science Center at Houston, Houston, Tex.

The authors wish to acknowledge the valuable contributions made to this article by Rebecca Drezek, PhD, Urs Utzinger, PhD, Sung K. Chang, MS, and Tom Collier, PhD, from the Department of Biomedical Engineering, The University of Texas at Austin, Austin, Tex.

Key points

• While the sensitivity and specificity of the traditional Pap smear are 62% and 68% respectively, the ratings for fluorescence spectroscopy are 86% and 74%; for Raman spectroscopy, it's 82% and 92%.
• Optical coherence tomography is capable of imaging to a depth of 1 mm at a resolution of 10 to 20 µm, making it potentially useful for screening epithelial lesions.
• Because of its ability to reject light from outside the focal volume, confocal microscopy is well suited for noninvasive evaluation of thick tissue and assessment of cell morphology.