Imaging of Spontaneous Canine Mammary Tumors Using Fluorescent Contrast Agents

Photochemistry and Photobiology, 1999, 70(1): 87-94

Imaging of Spontaneous Canine Mammary Tumors Using Fluorescent Contrast Agents

Jeffery S. Reynolds', Tamara L. Troy', Ralf H. Mayerl, Alan B. Thompson', David J. Waters2, Karen K. Cornell*, Paul W. Snyder2 and Eva M. Sevick-Muraca*' 'School of Chemical Engineering and *School of Veterinary Medicine, Purdue University, West Lafayette, IN, USA

Received 21 December 1998; accepted 14 April 1999

ABSTRACT

We present near-infrared frequency-domain photon migration imaging for the lifetime sensitive detection and localization of exogenous fluorescent contrast agents within tissue-simulating phantoms and actual tissues. We employ intensity-modulated excitation light that is expanded and delivered to the surface of a tissue or tissue- simulating phantom. The intensity-modulated fluorescence generated from within the volume propagates to the surface and is collected using a gain-modulated image-intensified charge-coupled device camera. From the spatial values of modulation amplitude and phase of the detected fluorescent light, micromolar volumes of diethylthiatricarbocyanine iodide (T = 1.17 ns) and indocyanine green (ICG) (T = 0.58 ns) embedded 1.0 cm deep in a tissue phantom are localized and discriminated on the basis of their lifetime differences. To demonstrate the utility of frequency-domain fluorescent measurements for imaging disease, we image the fluorescence emitted from the surface of in vivo and ex vivo canine mammary gland tissues containing lesions with preferential uptake of ICG. Pathology confirms the ability to detect spontaneous mammary tumors and regional lymph nodes amidst normal mammary tissue and fat as deep as 1.5 cm from the tissue surface.

INTRODUCTION

Over the past decade, studies of biomedical optical imaging of deep (> 1 cm) tissues have focused upon the collection of diffusely propagated, near-infrared (NIR)? light. Interior optical property maps of tissue-simulating phantoms have been mathematically reconstructed from surface measurements using continuous wave (CW) light (1) and frequency-domain photon migration (FDPM) (2,3). However, the clinical suc-

*To whom correspondence should be addressed at: School of Chem- ical Engineering, Purdue University, West Lafayette, IN 47907, USA. Fax: 765-494-0805.

tAbbreviarions: CCD, charge-coupled device; CW, continuous wave; DTTCI, diethylthiatricarbocyanine iodide; FDPM, frequency-domain photon migration; HpD, hematophorphyrin derivative; ICG, indocyanine green; NIR, near-infrared; PpIX, protoporphyrin IX; RF, radio frequency.

0 1999 American Society for Photobiology 003 1-8655/99 $5.00+0.00

cess of biomedical optical imaging will rely upon the ability to conduct surface measurements quickly and with maximal signal-to-noise ratio.

Franceshini et al. (4) and Moesta et al. (5) have demonstrated FDPM measurements for the detection of breast cancer in humans based upon contrast due to tumor neovascu- larization and hemoglobin absorption. While those reports demonstrate the ability to detect breast tumors as small as 0.5 cm in vivo with FDPM, the success of optical mammography depends upon the ability to detect smaller lesions at earlier stages of tumor progression. Detection of smaller lesions may require the use of exogenous fluorescent contrast agents that selectively alter local tissue optical properties to provide enhanced detection. In previous work, we have demonstrated that the contrast imparted by exogenous fluorescent agents can exceed that from absorption when FDPM techniques are employed (6).

Fluorescence FDPM involves launching NIR modulated excitation light at the tissue-air interface and detecting the modulated fluorescent light that is generated from within the tissue and that propagates to the tissue surface. As the modulated excitation light propagates through the tissue, its average intensity and modulation amplitude are attenuated while its modulation phase is delayed relative to the incident light. The tissue absorption and scattering properties determine the excitation light's local average intensity, modulation amplitude and phase delay. When the excitation light encounters a fluorophore, modulated fluorescent light is generated with an average intensity, modulation amplitude and modulation phase that is governed by the local fluorescent properties (yield and lifetime). As the fluorescent light propagates to the surface to be detected, it suffers yet another round of attenuation and phase delay as a result of the optical properties of the intervening tissue. Thus, in order to account for the complicating factors of tissue absorption and scatter, a mathematical model for propagation must be used in an inverse algorithm to determine interior optical properties (including fluorescent properties) from exterior measurements. In this work FDPM measurements consist of (1) the phase delay, 8, induced by scattering and fluorescent lifetime, T ;

(2) the amplitude of the detected intensity-modulated wave, termed the AC signal; (3) the average of the detected intensity-modulated wave, termed the DC signal. The modulation ratio is then calculated as AC/DC.

Localization and imaging of interior fluorescent tissue vol-

87

88 Jeffrey S. Reynolds et al.

Figure 1. Schematic of FDPM multipixel instrument used in front illumination geometry.

umes from experimental FDPM (7) and CW (8) measurements have been demonstrated in tissue-mimicking phantoms with perfect uptake of dye within the tissue volume of interest and with no dye in the surroundings. In addition, researchers have demonstrated the tumor-localizing effects of fluorescent contrast agents using CW detection (9,lO). Mathematical image reconstruction of fluorescence lifetimes from synthetic FDPM measurements conducted under conditions of perfect (1 1) and imperfect (12-14) uptake have also been performed. Under these realistic conditions of imperfect uptake of dye into the diseased tissue volume of interest, fluorescence FDPM can exploit the added phase delay and modulation attenuation contrast induced by the changes in fluorescence decay kinetics as mediated by the local bio-

VENTRAL

chemical environment. Furthermore, if the contrast agent lifetime changes in the diseased tissue, then imaging contrast is not totally dependent on preferential uptake.

Fluorescent lifetime differences between endogenous and exogenous fluorophores have been used for discrimination of normal versus diseased superficial tissues in tumor-bear- ing mice treated with a diagnostic dose of a photodynamic therapeutic agent, hematoporphyrin derivative (HpD) (1 5) . The effective lifetime of fluorescence was also shown to be higher in mouse leukemia and fibrosarcoma as compared to normal tissue (16). Perhaps this was due to increased uptake of HpD resulting in contrast between the longer lifetime of HpD compared to the lifetime of tissue autofluorescence. More recently Cubeddu ef al. have shown that time-domain imaging of the fluorescence generated from protoporphyrin IX (PpIX) could produce lifetime maps with high contrast between normal tissue and basal cell carcinoma in human skin (17). The PpIX accumulates preferentially in the diseased tissue following the topical application of 6-aminole- vulinic acid (ALA). Indeed, time-domain or frequency-domain imaging of a fluorescent contrast agent whose lifetime is sensitive to a diseased tissue state could be a very powerful diagnostic tool for evaluating tumor viability and the efficacy of nonsurgical treatments. However, in order for this lifetime contrast to be exploited for noninvasive imaging deep within tissue, one first must use fluorescent contrast agents that are excited and emit in the NIR therapeutic window (-700-900 nm) where tissue absorption is low. Then the complicating effects of tissue scatter and absorption must be taken into account through use of inverse-imaging algorithms that numerically reconstruct interior optical properties based on exterior measurements.

Figure 2. Diagram showing of mammary chain images.

nomenclature for identifying location Figure 3. Schematic of the tissue phantom with front illumination.

Photochemistry and Photobiology, 1999, 70(1) 89

For successful biomedical optical imaging using numeri- cal reconstruction techniques, many independent measurements are required to alleviate the underdetermined, ill- posed nature of the reconstruction problem. Furthermore, clinical applications require that these measurements be made rapidly while maintaining a high signal-to-noise ratio on often weak optical signals. In order to meet these con- flicting requirements, we employ a gain-modulated image intensifier as a homodyned multipixel radiofrequency (RF) phase-sensitive camera. This use of a gain-modulated image intensifier was first introduced by Lackowicz and Berndt

Figure 4. The 128 X 128 pixel, 830 nm fluorescent FDPM images of (a) DC, (b) AC, (c) phase delay and (d) ACDC of the detected fluorescence generated from two 0.4 cmz vials containing 1.4 pW DTTCI (right) and 1.0 pW of ICG (left) in 0.5% Intral- ipid and approximately 1.0 cm from the surface of a 0.5% Intralipid solution. Front illumination with a 780 nm laser diode modulated at 100 MHz was used.

Figure 5. The 128 X 128 pixel, 830 nm fluorescent FDPM images of (a) DC, (b) AC, (c) phase delay and (d) ACDC of the detected fluorescence generated from the area of the left fifth mammary gland of Dog 1 ex vivo (cranial side down in images). Front illumination with a 780 nm laser diode modulated at 100 MHz was used.

(18) for nonscattering systems and subsequently applied to FDPM imaging of absorbers in laboratory scattering phantoms by Sevick et at'. (19). We have incorporated improve- ments in intensifier, charge-coupled device (CCD) and data acquisition technologies to improve the FDPM imaging of both absorptive and fluorescent objects in phantoms (20-22) and obtained preliminary images in animal models (23). Re- cently, Wagnikres ef al. demonstrated this technique for fluorescent surface imaging in vivo using UV excitation through an endoscope (24).

In this paper we experimentally confirm the altered con-

90 Jeffrey S. Reynolds eta/.

trast with changes in fluorescence lifetime using tissue-simulating phantom-containing dyes with different lifetimes. We then demonstrate the ability to detect exogenously induced fluorescence in diseased canine mammary tissue following injection of fluorescent indocyanine green (ICG). While ICG does not experience lifetime changes, the results nonetheless provide impetus for the development of lifetime sensitive dyes to enhance specificity and contrast for imaging deeply located diseased tissue.

METHODS AND MATERIALS Multipixel FDPM apparatus. The imaging apparatus consists of a modulated laser diode for illumination and a gain-modulated image intensifier coupled to a CCD camera for detection as shown in Fig. 1. Illumination is accomplished by a 20 mW, 780 nm laser diode whose RF modulation at 100 MHz is phase-locked to the RF modulation driving the image intensifier. This light is expanded with a lens and then used to illuminate the surface of the tissue or phantom. An imaging lens and an 830 nm bandpass filter placed in front of the image intensifier enables measurement of fluorescent light generated from within the tissue phantom. As the phase of the image intensifier’s gain modulation is stepped through 360” relative to the laser diode modulation, the CCD camera (TEKCD-5 12-EFT, Princeton Instruments, Inc., Trenton, NJ) records the steady-state, homodyned image at the output of the Generation III image intensifier (FS9910C. I’IT Industries, Roanoke, VA). The image intensifier output is efficiently coupled to the CCD array using a fiber optic taper. The RF modulation magnitude and phase for each image pixel is then efficiently calculated off-line using an algorithm de- scribed elsewhere (22). For a typical acquisition, a series of images are averaged at each phase delay to average small variations due to physiological motion such as pulse and respiration that results in a typical image acquisition time of approximately 30 s for these 128 X 128 pixel fluorescent images.

Measurement geometry. Front or transillumination geometries are employed for both phantom and tissue measurements. In the front illumination geometry a lens expands the beam directly from the laser diode to illuminate the surface being imaged. In the transillumination configuration, the beam is coupled into a multimode fiber for delivery to the tissue surface on the side opposite the imaged surface (23). The front illumination geometry is more suitable for fluorescence imaging than for absorption imaging because in fluorescence imaging the bandpass filter rejects the backscattered light that would normally overwhelm a front illumination absorption image. Front illumination is especially advantageous in cases where transillumination is not practical, as in the case of breast lesions close to the chest wall or possibly within axillary lymph nodes. In our experience, the front illumination geometry is more convenient for imaging the mammary chain of dogs because the mammary tissue volumes are often too small and too close to the abdominal wall to be imaged in a transillumination geometry.

Fluorescent contrast agent for mammary chain imaging. The fluorescence of ICG was used to detect spontaneous tumors within the canine mammary chain. Although ICG does not appear to exhibit lifetime sensitivity related to tissue state, it does have several desirable attributes for this preliminary clinical imaging. (1) The ICG has an excitation peak near 800 nm and an emission peak near 830 nm. Thus, both the excitation and emission light fall within the biological or therapeutic window where tissue absorption is small. These spec- tral characteristics are especially important when trying to image small lesions deep within tissues ( I 4 cm). (2 ) The ICG has a high quantum efficiency and is reasonably stable in the blood. The 0.58 ns lifetime is compatible with the modulation frequencies (30-150 MHz) and photon transport times commonly encountered while per- forming FDPM in tissue. (3) Because ICG is Food and Drug Ad- ministration approved as an intravenous diagnostic agent, its toxi- cology is relatively well understood. (4) The ICG appears to have some preferential uptake in tumors, owing to leakage from the tumor neovasculature.

Animal model and imaging. Eligible subjects were chosen from the pet dog population scheduled to undergo mammary tumor ex-

cision at the Purdue University Veterinary Teaching Hospital. All lesions were spontaneous, naturally occurring tumors of various types. Written owner consent was obtained and all procedures were approved by the Purdue Animal Care and Use Committee. Prior to imaging, each dog was anesthetized and the abdominal area shaved. A dose of 1 m a g body weight ICG was administered intravenously in a saline bolus and FDPM illumination and imaging commenced immediately. For the purpose of this paper, we have referred to the location of each of the 10 canine mammary glands as left 1-5 and right 1-5, where 1 is the most cranial gland and 5 is the most caudal gland (see Fig. 2). While imaging was conducted across the third, fourth and fifth mammary glands, the focus of the measurements was upon palpable lesions within the mammary tissue. Generally, imaging was conducted with only light compression of the tissues using a clear acrylic plate to both flatten the tissue over the image field-of-view and reduce motion artifacts. The area immediately caudal to the fifth mammary gland was also imaged in order to assess regional (inguinal) lymph node involvement. One hour of imaging time was allotted before the diseased sections of the mammary chain were surgically removed.

Following surgical removal, images of the excised mammary tissue were obtained by placing the tissue flat on a horizontal surface and using front illumination techniques to gather FDPM images. Following the ex vivo measurements, the excised tissue was frozen at -80°C. Frozen mammary tissues were sliced longitudinally with a band saw. Thawed tissues were photographed, fixed in 10% buff- ered formalin and processed for histopathological examination. All gross lesions within the mammary tissue were carefully measured with calipers and their dimensions recorded.

RESULTS Tissue phantom measurements

In order to assess the performance of the instrument and demonstrate the utility of FDPM imaging of fluorescence in tissue-like scattering media, we obtained multipixel FDPM images of a tissue-simulating phantom (Fig. 3) containing two 0.7 cm-diameter, 1 cm-long cylindrical volumes filled with fluorescent dyes diluted in 0.5% Intralipid solution. The phantom was a 10 cm-diameter cylinder filled with 0.5% Intralipid in water (from 20% Intralipida, Pharmacia & Up- john, Clayton, NC). Both heterogeneities were located 1.0 cm beneath the top (imaged) surface of the phantom and separated by 2 cm. They were imaged using front illumination with a 780 nm laser diode modulated at 100 MHz. One heterogeneity contained a 1 .O pJ4 ICG (Cardio-Green@, Becton Dickinson Microbiology Systems, Cockeysville, MD) solution while the second contained a 1.4 phf solution of the dye DTTCI (3,3’-diethylthiatricarbocyanine iodide, Eastman Kodak, Rochester, NY). The fluorophore concen- trations were adjusted to approximately equalize their fluorescent yields, but the lifetimes of the two dyes differ: DTTCI has a single fluorescence lifetime of 1.17 ns, while the lifetime of ICG is 0.58 ns (6).

The resulting FDPM images of DC, AC, ACDC and 8 are shown in Fig. 4. Although signals from both heterogeneities are distinctly visible, the two dyes cannot be identi- fied in the average intensity (DC) image of Fig. 4a. How- ever, because the modulation amplitude of generated fluorescence is proportional to 1/[1 + ( O T ) * ] ~ ’ ~ and the phase is proportional to tan-'(^^) (where 7 is the fluorescence lifetime and o is the modulation angular frequency), the two dyes are easily distinguished on the basis of lifetime in the modulation phase and modulation ratio (AC/DC) images in Fig. 4c and d. The fluorescent signal arising from DTTCI experiences a greater phase delay and loss of modulation


amplitude because of the longer fluorescent lifetime of DTTCI. As expected, there is an approximately 17" relative phase shift and a 15% change in modulation between the two dyes. These laboratory results provide validation for proposed optical imaging based on fluorophore lifetime (1 1- 14).

Animal model measurements

In the following results, we present a series of FDPM images from two female canine subjects, a 25 kg. 8 year-old boxer (Dog 1) and a 42 kg, 10 year-old black Labrador retriever (Dog 2). All images are fluorescent images with an 830 nm bandpass filter and 780 nm front illumination modulated at 100 MHz. The average intensity (DC) image, modulation amplitude (AC), modulation phase and modulation ratio ( A C E ) images are plotted for each location. The data are orientated with the cranial side of the image down. By imaging the fluorescence before, during and after ICG injection, the fluorescent signal from the ICG was determined to be typically more than three orders of magnitude greater than autofluorescence or backscattered excitation light that leaks through the bandpass filter.

Detection of mammary tumor and lymph node in Dog 1

Figure 5 represents the FDPM images of the fluorescence generated ex vivo in the left fifth mammary gland of Dog 1 using front illumination excitation of the lightly compressed tissue surface overlying a 1 cm-diameter palpable lesion within the mammary gland. The mammary tissue was excised 120 min after ICG administration. Several features are evident in the fluorescent images in Fig. 5a. The circular shadow in the upper right comer was caused by the nipple. The bright area down the left side of the image resulted from a fold in the tissue at the excision boundary. The bright (red) area below and to the left of the nipple originates from the palpable breast lesion, a malignant mixed mammary tumor verified by pathology. Two small (red) fluorescent areas fur- ther below the tumor originate from volumes associated with no palpable lesions. Pathological examination indicates that these signals originated from reactive regional lymph nodes located approximately 1.5 cm deep within the tissue, as can be seen in the histopathological slices in Fig. 6.

I n vivo and ex vivo detection of mammary tumor and lymph node in Dog 2

Figure 7 represents the in vivo FDPM images of the fluorescence generated from a palpable nodule within the left fourth mammary gland of Dog 2. Front illumination with a 780 nm laser diode modulated at 100 MHz was used with gentle compression at the surface 23 min after injection of ICG. The tumor can be clearly seen in all four data plots. Pathologic examination indicated that the fluorescent FDPM image represented in Fig. 7 correlated with a 1.2 cm (lon- gitudinal) by 0.5 cm (axial) papillary adenoma located approximately 1 cm deep within the mammary tissue.

Figure 8 is an in vivo image acquired from Dog 2 just cranial to the nipple of the left fifth gland 30 min after ICG injection. While this area was not associated with a palpable nodule, pathologic examination confirmed that the fluores-

B

Figure 6. Histology slices from the Left fifth mammary gland of Dog I (the area imaged in Fig. 5). The slices are approximately 5 mm thick. The skin is down in the first (most superficial) slice. Slice four represents a depth of approximately 1.5 cm.

cence is attributed to a blood vessel that bifurcates approximately 1 cm below the surface in the area just cranial of the regional lymph node.

For comparison, an ex vivo image combining the two previous areas is shown in Fig. 9. Here the bottom of the image corresponds to the area near the center of Fig. 7 and the top corresponds to Fig. 8. Note that the fluorescent spots are smaller in the ex vivo images because the tissue was excised 90 min after ICG injection, allowing more than three times as long for ICG washout in the ex vivo images. In addition, mild compression was used in the in vivo images and not in the ex vivo images.

Figure 10 represents the in vivo FDPM images of the fluorescence generated from the area of the right fifth mammary gland of Dog 2. Here, front illumination is used without compression 43 min after injection of the ICG. Pathologic examination showed that the fluorescent source in this image corresponds to the regional lymph node.

DISCUSSION While ICG is not known to experience lifetime sensitivity (or thus phase contrast), the FDPM images nonetheless demonstrate the ability for in vivo detection of frequency-domain fluorescent signals from an exogenous contrast agent. Based upon systematic pathologic examination of the excised mammary tissues, these signals appear to originate from mammary tumors that arise spontaneously within the mammary gland. Because the physiological differences between spontaneous and implanted tumors could yield vastly different optical characteristics, we believe that spontaneous breast cancer in dogs represents an important large animal model for evaluating contrast agents and clinical measurement procedures. In contrast with models involving induced or implanted tumors in rodents, this model allows realistic as- sessment of a technique's ability to detect mammary tumors amidst relevant volumes of normal mammary tissue and fat.

92 Jeffrey S. Reynolds et a/.

These encouraging results will hopefully accelerate the development of improved fluorescent contrast agents with many of the desirable qualities of ICG but that also exhibit high tumor specificity and fluorescent kinetics that are sensitive to important physiological parameters.

Although lifetime contrast using exogenous contrast has been demonstrated in superficial tumors (15-17). exploita- tion of lifetime contrast deep within tissue depends on the ability of inverse-imaging algorithms to separate the various propagation effects in the highly scattering tissue from variations in fluorescent intensity and lifetime changes. Com-

Figure 7. The 128 X 128 pixel. 830 nm fluorescent FDPM images of (a) DC, (b) AC, (c) phase delay and (d) AC/DC of the detected fluorescence generated from the area just caudal of the left fourth mammary gland of Dog 2 in vivo (cranial side down in images). Front illumination with a 780 nm laser diode modulated at 100 MHz was used.

Figure 8. The 128 X 128 pixel, 830 nm fluorescent FDPM images of (a) DC, (b) AC, (c) phase delay and (d) AClDC of the detected fluorescence generated from the area just cranial of the left fifth mammary gland of Dog 2 in vivo (cranial side down in images). Front illumination with a 780 nm laser diode modulated at 100 MHz was used.

bining rapid, high-resolution, multipixel FDPM measurements with inversion algorithms currently under development by others and us (7,ll-13) may enable imaging and contrast on the basis of fluorescent lifetime. While ICG may not provide the lifetime discrimination that has been shown using HpD and PpIX in superficial diseased tissue, it is note- worthy that slight differences between modulation ampli- tuddphase images and the corresponding average intensity images in Figs. 4-8 indicate that additional information is still contained in the frequency-domain images compared to CW measurements. This added information should prove


Figure 9. The 128 X 128 pixel, 830 nm fluorescent FDPM images of (a) DC, (b) AC, (c) phase delay and (d) ACDC of the detected fluorescence generated from between the left fourth and fifth mammary glands Dog 2 ex vivo (cranial side down in images). Front illumination with a 780 nm laser diode modulated at 100 MHz was used.

Figure 10. The 128 X 128 pixel. 830 nm fluorescent FDPM images of (a) DC, (b) AC, (c) phase delay and (d) ACDC of the detected fluorescence generated from a lymph node in the area of the right fifth mammary gland Dog 2 in vivo (cranial side down in images). Front illumination with a 780 nm laser diode modulated at 100 MHz was used.

useful for the often underdetermined mathematical reconstruction of interior optical properties from measurements conducted at the tissue surface. Furthermore, as is evidenced in Fig. 4, CW measurements do not contain sufficient information for lifetime determination independent of dye con- centration and quantum efficiency. Without frequency or time-dependent measurements the opportunity for reconstruction of fluorescence decay kinetics within tissues is lost.

It would seem natural to attempt to eliminate variations in illumination and skin absorption by referencing the fluorescent images to measurements of the backscattered exci-

tation light. We do regularly compare and normalize emission images to excitation images to be certain that we are not imaging illumination or absorption artifacts in our fluorescent images. However, while normalization may work well for superficial imaging and endoscopic applications, especially when using shorter wavelengths that do not propa- gate deeply in tissue, the deep penetration of NIR light and the inherent three-dimensional nature of deep imaging in scatter causes a nonlinear coupling of the measured fluorescent emission light to the backscattered excitation light. An illustrative example of this problem would be the case of a

94 Jeffrey S. Reynolds et a/.

uniform phantom with a small area of increased backscatter at the surface. This increased scatter would cause a bright spot in the backscatter image even though less light is prop- agating into the tissue to excite fluorescence. Thus, normal- izing the emission to the excitation image would in this case decrease the fidelity of the image.

Our ability to detect fluorescent signals apparently origi- nating from regional lymph nodes suggests that FDPM fluorescence imaging coupled with improved fluorescent dyes could also provide a valuable diagnostic method for access- ing regional lymph node involvement in human breast cancer. Lymph node status in breast cancer is a powerful pre- dictor of recurrence and survival and the number of lymph nodes with metastases provides critical prognostic information regarding the choice of adjuvant therapy (25). Currently, lymph node involvement is assessed by dissection and sub- sequent pathologic examination, but researchers are inves- tigating the use of other diagnostic modalities including magnetic resonance imaging, x-ray computed tomography and sonography (25). Recently, gamma-ray imaging of a technetium-99m sulfur colloid injected into the area of a known breast tumor has been used to identify the sentinel lymph nodes that can then be removed (26-27). Similarly, multipixel FDPM imaging after an injection of a fluorescent contrast agent with receptor-mediated uptake could yield a low-cost and noninvasive technique for detecting nodal involvement.

Acknowledgements-This work was supported in part by the National Institutes of Health (ROlCA67176 and K04CA68374) and by Mal- linckrodt, Inc. The authors acknowledge the assistance Ms. Amy Hon- komp, Dr. A. Maras and Dr. E. Jay, as well as helpful discussions with Dr. William H. Ralston of Mallinckrodt, Inc.

1.

2.

3.

4.

5 .

6.

7.

8.

Huabei, H., K. D. Paulsen and U. L. Osterberg (1996) Optical image reconstruction using DC data: simulations and experiments. Phys. Med. Biol. 41, 1-16. O’Leary, M. A.. D. A. Boas, B. Chance and A. G. Yodh (1995) Experimental images of heterogeneous turbid media by frequency-domain diffusing-photon tomography. Opt. Lett. 20, 42-28. Huabei, H., K. D. Paulsen, U. L. Osterberg, B. W. Pogue and M. S. Patterson (1996) Optical image reconstruction using frequency domain data: simulations and experiments. J. Opt. SOC. Am. A 13, 253-266. Franceshini, M. A,, K. T. Moesta, S. Fantini, G. Gaida, E. Grat- ton, H. Jess, W. W. Mantulin, M. Seeber, P. M. Schlag and M. Kaschke (1997) Frequency-domain techniques enhance optical mammography: initial clinical results. P roc. Nutl. Acad. Sci. USA 94, 6468-6473. Moesta, K., S. Fantini, H. Jess, S. Totkas, M. Franceschini, M. Kaschke and P. Schlag (1998) Contrast features of breast cancer in frequency-domain laser scanning mammography. J. Biomed.

Sevick-Muraca, E. M., G. Lopez, T. L. Troy, J. S. Reynolds and C. L. Hutchinson (1997) Fluorescence and absorption contrast mechanisms for biomedical optical imaging using frequency-domain techniques. Photochem. Photobiol. 66, 55-64. O’Leary, M. A., D. A. Boas, B. Chance and A. G. Yodh (1994) Reradiation and imaging of diffuse photon density waves using fluorescence inhomogeneities. J. Luminesc. 60-61, 281-286. Hull, E. L., M. G. Nichols and T. H. Foster (1998) Localization

Opt. 3, 129-136.

of luminescent inhomogeneities in turbid media with spatially resolved measurements of cw diffuse luminescence emittance.

9. Reddi, E., A. Segalla, G. Jori, P. K. Kerrigan, P. A. Lidell. A. L. Moore, T. A. Moore and D. Gust (1994) Carotenoporphryins as selective photodiagnostic agents for tumors. Br. J. Cancer 69, 4 0 4 5 .

10. Takemura, T., S. Nakajima and I. Sakata (1994) Tumor localizing fluorescent diagnostic agents without phototoxicity. Pho- tochem. Photobiol. 59, 366-370.

11. O’Leary, M. A,, D. A. Boas, X. D. Li, B. Chance and A. G. Yodh (1996) Fluorescence lifetime imaging in turbid media. Opt. Lett. 21, 158-160.

12. Paithankar, D. Y., A. Chen and E. M. Sevick-Muraca (1996) Fluorescence yield and lifetime imaging in tissues and other scattering media. Proc. SPlE 2679, 162-175.

13. Paithankar, D. Y., A. U. Chen, B. W. Pogue, M. S. Patterson and E. M. Sevick-Muraca (1997) Imaging of fluorescent yield and lifetime from multiply scattered light re-emitted from tissues and other random media. Appl. Opt. 36, 2260-2272.

14. Chang, J., H. L. Graber and R. L. Barbour (1997) Luminescence optical tomography of dense scattering media. J. Opt. SOC. Am.

15. Cubeddu, R., G. Canti, A. Pifferi, P. Taroni and G. Valentini (1993) Time-gated fluorescence imaging for the diagnosis of tumors in a murine model. Photochem. Photobiol. 57. 480-485.

16. Cubeddu, R., G. Canti, A. Pifferi, P. Taroni and G. Valentini (1997) Fluorescence lifetime imaging of experimental tumors in hematoporphyrin derivative-sensitized mice. Photochem. Pho- tobiol. 66, 229-237.

17. Cubeddu. R.. A. Pifferi. P. Taroni. A. Torricelli. G. Valentini.

Appl. Opt. 37, 2755-2765.

A 14, 288-299.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

F. Rinaldi and E. Sorbellini (1999) Time-gated and lifetime imaging techniques for the detection of skin tumors. Proc. SPIE. (In press) Lakowicz. J. and K. Berndt (1991) Lifetime-selective fluorescence imaging using an rf phase-sensitive camera. Rev. Sci. In- strum. 62, 1727- 1734. Sevick, E. M., J. R. Lakowicz, H. Szmacinski, K. Nowaczyk and M. Johnson (1992) Frequency-domain imaging of obscure absorbers: principles and applications. Photochem. Phorobiol.

Reynolds, I. S., T. L. Troy and E. M. Sevick-Muraca (1997) Multi-pixel imaging of interfering photon density waves. P roc. SPIE 2979, 122-128. Troy, T. L., J. S. Reynolds and E. M. Sevick-Muraca (1997) Photon migration imaging using multi-pixel measurements. Proc. SPIE 2979, 1 11-121. Reynolds, J. S., T. L. Troy and E. M. Sevick-Muraca (1997) Multipixel techniques for frequency-domain photon migration imaging. Biotechnol. Prog. 13, 669-680. Reynolds, J. S., T. L. Troy, D. J. Waters, K. K. Cornell and E. M. Sevick-Muraca (1998) Multi-pixel frequency-domain photon migration imaging of a fluorescent contrast agent. Biomed. Opt. Spectrosc. Diugn. Trends Opt. Photon. XXII, 5&53. Wagnitres, G., J. Mizeret, A. Studzinski and H. van den Berg ( 1997) Frequency-domain fluorescence lifetime imaging for endoscopic clinical cancer photodection: apparatus design and preliminary results. J. Fluoresc. 7. 75-83. Moore, M. P. and D. W. Kinne (1998) The surgical management of primary invasive breast cancer. Cu. Cancer J. Clin. 45, McMaster, K. M.. A. E. Giuliano, M. I. Ross, D. S. Reintgen, K. K. Hunt, V. S. Klimberg, P. W. Whitworth, L. C. Tafra and M. J. Edwards (1998) Sentinel-lymph-node biopsy for breast cancer-not yet the standard of care. N . Engl. J. Med. 339, 990- 995. Krag, D., D. Weaver, T. Ashikaga, F. Moffat, S. Klimberg, C. Shriver, S. Feldman, R. Kusminsky, M. Gadd, J. Kuhn, S. Har- low, P. Beitsch, P. Whitworh, R. Foster, Jr. and K. Dowlatshahi (1998) The sentinel node in breast cancer-a multicenter validation study. N. Engl. J. Med. 339, 941-946.

16, 169-185.

Imaging of Spontaneous Canine Mammary Tumors Using Fluorescent Contrast Agents

Documents