Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

In-vivo optical detection of cancer using chlorin e6 – polyvinylpyrrolidone induced fluorescence imaging and spectroscopy

In-vivo optical detection of cancer using chlorin e6 – polyvinylpyrrolidone induced fluorescence... Background: Photosensitizer based fluorescence imaging and spectroscopy is fast becoming a promising approach for cancer detection. The purpose of this study was to examine the use of the photosensitizer chlorin e6 (Ce6) formulated in polyvinylpyrrolidone (PVP) as a potential exogenous fluorophore for fluorescence imaging and spectroscopic detection of human cancer tissue xenografted in preclinical models as well as in a patient. Methods: Fluorescence imaging was performed on MGH human bladder tumor xenografted on both the chick chorioallantoic membrane (CAM) and the murine model using a fluorescence endoscopy imaging system. In addition, fiber optic based fluorescence spectroscopy was performed on tumors and various normal organs in the same mice to validate the macroscopic images. In one patient, fluorescence imaging was performed on angiosarcoma lesions and normal skin in conjunction with fluorescence spectroscopy to validate Ce6-PVP induced fluorescence visual assessment of the lesions. Results: Margins of tumor xenografts in the CAM model were clearly outlined under fluorescence imaging. Ce6-PVP-induced fluorescence imaging yielded a specificity of 83% on the CAM model. In mice, fluorescence intensity of Ce6-PVP was higher in bladder tumor compared to adjacent muscle and normal bladder. Clinical results confirmed that fluorescence imaging clearly captured the fluorescence of Ce6-PVP in angiosarcoma lesions and good correlation was found between fluorescence imaging and spectral measurement in the patient. Conclusion: Combination of Ce6-PVP induced fluorescence imaging and spectroscopy could allow for optical detection and discrimination between cancer and the surrounding normal tissues. Ce6-PVP seems to be a promising fluorophore for fluorescence diagnosis of cancer. Page 1 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 A number of fluorochromes such as fluorescein, toluidine Background As with most cancers, early diagnosis is critical to achieve blue, cyanine dyes and indocyanine green have been favorable prognosis. Currently, random surveillance biop- described with variable stabilities, quantum efficiencies, sies are the existing gold standard for the identification of and ease of synthesis. However, most of the fluoro- lesions in pre-neoplastic conditions. However this chromes are not tumor specific and are rapidly eliminated method is prone to sampling error, time-consuming, sub- from the organism. Chemically and endogenously synthe- jective and cost-inefficient. A diagnostic method that sized fluorochromes such as porphyrin based photosensi- could provide rapid, automated classification of cancer tizers have properties that may be utilized both lesions would increase the efficiency and comprehensive- experimentally and clinically. Porphyrins have been ness of malignancy screening and surveillance procedures. known to naturally localize in malignant tissue where A variety of optical techniques have recently been utilized they emit light when irradiated at certain wavelengths, for the diagnostic study of cancerous tissue. These include providing a means to detect tumor by the location of its fluorescence spectroscopy [1], Raman spectroscopy [2], fluorescence. However, one of the major limitation is its light scattering spectroscopy [3], and Fourier-transform slow clearance from tissues and long period of skin pho- infrared spectroscopy [4]. These optical spectroscopic totoxicity. Moreover, the porphyrin's core absorbs wave- techniques are capable of providing biochemical and lengths of light too short for optimal penetration in tissue. morphological information in short integration times, As such, by reducing a pyrrole double bond on the por- which can be used for automated diagnosis of intact tis- phyrin periphery, a chlorin core compound can be gener- sue. However, in order to be useful as a comprehensive ated with a high absorption at longer wavelengths of 660 screening procedure, the optical technique must allow – 670 nm that can penetrate deeper in human tissue than rapid real time imaging of a large area of tissue rather than those of porphyrins. Of particular interest among the eval- point by point measurement, such that suspicious regions uated chlorins is the naturally occurring chlorin e6 (Ce6) could be identified accurately and biopsied for his- [11]. Ce6 has improved efficacy and has decreased side topathologic correlation [5]. effects compared to first generation photosensitizers from hematoporphyrin derivatives. Due to the importance of With the advent of molecular probes, imaging methods Ce6's characteristic fluorescence properties, there is a need such as ultrasound, microCT (Computed Tomography), to identify new formulations that are stable, exhibit ease microMRI (Magnetic Resonance Imaging), and microPET in manufacturing and selectively deliver the photosensi- (Positron Emission Tomography) can be conducted not tizer to target tissue in an efficient manner. Hence, we only to visualize gross anatomical structures, but also to have investigated the use of Ce6 in combination with the visualize substructures of cells and monitor molecule polymer polyvinylpyrrolidone (Ce6-PVP). Polyvinylpyr- dynamics [6]. Imaging of endogenous or exogenous fluor- rolidone is one of the most important excipient used in ochromes has several important advantages over other modern pharmaceutical technology. We have previously optical approaches for tumor imaging. This imaging tech- described the selective localization and photodynamic nique relies on fluorochrome induced fluorescence, activity of Ce6-PVP in nasopharyngeal and lung carci- reflectance, absorption or bioluminescence as the source noma models that provided rationale for its use as a ther- of contrast, while imaging systems can be based on diffuse apeutic agent for photodynamic therapy [12,13]. By optical tomography, surface-weighted imaging, phase- employing a chick chorioallantoic membrane model, array detection, intensified matrix detector and charged- Ce6-PVP was shown to selective accumulate in bladder coupled device camera detection, confocal endomicros- tumors xenografts and had a faster clearance from normal copy, multiphoton imaging, or microscopic imaging with CAM when administered topically compared to system- intravital microscopy [7,8]. Fluorescence ratio imaging is atic administration [14]. The uptake ratio of Ce6-PVP was a method widely used for optical diagnosis of cancer after found to have a 2-fold increase across the CAM when administration of a photosensitizer [9]. Enhanced con- compared to that of Ce6, indicating that PVP was able to trast between tumor and adjacent normal tissue can be facilitate diffusion of Ce6 across the membrane [15]. Fur- obtained based on calculating the ratio between red inten- thermore, Ce6-PVP had less in vivo systemic phototoxic sity of the photosensitizer (600–700 nm) over the blue/ effect compared to Ce6 alone after light irradiation in green intensity of the back-scattered excitation light or tis- photodynamic therapy in mice bearing tumors [16]. sue autofluorescence (450–550 nm). Many investigations Using a chemical fluorescence extraction technique and have confirmed good agreement with the histopathologi- cuvette-based spectrofluorimetry, our data demonstrated cal extent of the tumor, implying that this technique can that the distribution of Ce6-PVP drug were much lower in be applied as a useful tool for indicating tumor boundary normal organs such liver, spleen, kidney, brain, heart and [10]. lung compared to Ce6 delivered using dimethylsulfoxide (DMSO) [17]. We also postulated that the extent of tumor necrosis post Ce6-PVP mediated photodynamic therapy Page 2 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 (PDT) was dependent on the plasma concentration of window was resealed to avoid evaporation of the drug Ce6-PVP, implying a vascular mediated cell death mecha- solution from the CAM. After 30 min incubation, macro- nism [18]. scopic fluorescence imaging was performed at 0.5, 1, 2, 3, 4, and 5 h post drug administration using a commercially In this study, we have evaluated the usefulness of Ce6-PVP available fluorescence endoscopic system (Karl Storz, Tut- to accurately define the margin of the tumor from its nor- tlingen, Germany). A modified xenon short arc lamp (D – mal adjacent tissue in the chick choriallantoic membrane Light system in blue light mode, Karl Storz) filtered by a (CAM) tumor model. We also presented visual informa- band pass filter (380 – 450 nm) was used for excitation of tion on Ce6-PVP induced fluorescence in tumor and gross photosensitizer in tissue. Fluorescence was captured via a anatomical structure of normal organ of a murine model. sensitive CCD camera (Tricam SL PAL, Karl Storz) Fluorescence spectroscopy measurements were also per- attached to an endoscope integrated with a long pass filter formed to characterize emission spectra from these tissue (cut-off wavelength 470 nm). This observation LP filter of samples as well as to corroborate results from fluorescence the endoscope only minimally transmits the diffuse back- images. Finally, a pilot trial was carried out to validate the scattering excitation light with a peak at 450 nm (blue use of Ce6-PVP as a clinically relevant diagnostic photo- light), while has a transmission of over 98% in the 470 – sensitizer using both imaging and spectroscopy modality 800 nm range. The red channel registered the photosensi- for differentiation of normal and tumor tissue in a patient. tizer's fluorescence and the blue channel captured the dif- fusely back-scattered excitation light. A short exposure of Methods the surface of tissue to the excitation light (10 s) was per- Photosensitizer formed to avoid excessive photobleaching effects. White The formulation of Ce6-PVP, also known as Fotolon or light imaging was used to correlate the boundaries of Photolon was supplied by HAEMATO-science GmbH, tumors and organs. All procedures involving preparation Germany. It is a co-lyophilisate of Ce6 sodium salt and and administration of the photosensitizer were conducted PVP (a pharmaceutical grade polymer, molecular mass ≈ under low ambient lighting. 12,000 g/mol) in a 1:1 mass ratio [19]. Murine tumor model Cell culture A total of 10 Balb/c athymic nude mice and C57 mice, 6– MGH (European Collection of Cell Cultures), a poorly 8 weeks of age, weighing an average of 24 g were obtained differentiated human bladder carcinoma cells were grown from the Animal Resource Centre, Western Australia and as a monolayer in RPMI-1640 medium supplemented Centre for Animal Resources, National University of Sin- with 10% fetal bovine serum, 1% non-essential amino gapore respectively. Before inoculation, the cell layer was acids (Gibco, USA), 1% sodium pyruvate (Gibco, USA), washed with phosphate-buffered saline, trypsinized, and -1 6 100 units mL penicillin-streptomycin (Gibco, USA) and counted using a hemocytometer. Approximately 3.0 × 10 incubated at 37°C, 95% humidity and 5% CO . MGH cells suspended in 150 μl of Hanks' Balanced Salt Solution (Gibco, USA) were injected subcutaneously into Chick choriallantoic membrane tumor model the lower flanks of Balb/c athymic nude mice. The ani- Fertilized chicken eggs were incubated at 37°C in a mals were used for experiments when the tumors meas- humidified atmosphere inside a hatching incubator ured around 7 – 10 mm in diameter. This ensured that the equipped with an automatic rotator (Octagon 20, Brinsea, tumor sizes were kept consistent to minimize variations Somerset, UK). At embryo age (EA) 7, a window of about due to the degree of vascularization of the implants. Mice 1.5 cm was opened in the eggshell to detach the shell were injected with a dose of 5 mg/kg of Ce6-PVP via tail membrane from the developing CAM. Then, the window vein injection. At 1, 3 and 6 h, mice were sacrificed and was sealed with sterilized parafilm to avoid contamina- the skin overlaying the tumor was carefully removed to tion and the eggs were returned to the static incubator for expose the tumor and normal peritumoral muscle for flu- further incubation until the day of experiment. On EA 9, orescence imaging. C57 mice were used for imaging and approximately 5–10 × 10 MGH cells were inoculated on spectroscopy of normal organs. All procedures were the CAM. The window of the eggs were resealed with ster- approved by the Institutional Animal Care and Use Com- ile parafilm and returned to the static incubator. Grafted mittee, SingHealth, Singapore, in accordance with inter- cells were allowed to grow on the CAM for up to 5 days. national standards. On EA 14, Ce6-PVP was dissolved in 0.9% sodium chlo- Spectroscopic measurement using fiber optics-based ride to constitute a stock solution of 1 mg/mL. The stock solution was further diluted to obtain a volume of 500 μL fluorescence spectrometer containing a dose of 1 mg/kg body weight of the chick's The spectral measurement was performed on mice sacri- embryo. The photosensitizer was applied on the entire ficed at 1 and 3 h post Ce6-PVP administration. A fiber surface of the CAM and left to incubate for 30 min. The optics-based fluorescence spectrometer (Spex SkinSkan, Page 3 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 JY Inc., Edison, NJ, USA) was used for the measurement of cedure significantly reduced the within-class variances. fluorescence intensity of Ce6-PVP. A monochromator Spectral data from the various organs and tissues were with a 150-W Xenon lamp was used as the excitation light analyzed to determine spectral line shape and the peak source. The excitation light (400 nm) was guided to illu- fluorescence intensities at region of 660 – 690 nm. minate samples by one arm of a Y-type quartz fiber bun- dle, and the emission fluorescence was collected by Results Fluorescence imaging of bladder tumor xenografts on the another arm of the fiber bundle, guided to another motor- controlled monochromator. The resulting emission spec- CAM model tra were recorded from 650 to 750 nm, in 1 nm incre- Fluorescence was not observed from the tumors under ments, collected using the DataMax version 2.20 blue light illumination before drug administration. After (Instruments SA, Inc.) software package. The optical fiber topical administration of Ce6-PVP, an intense red fluores- tip was placed on the measuring sites and fluorescence cence in the bladder tumor xenografts was observed, sug- intensity spectra were measured. After each measurement, gesting selective localization in the malignant cells. the optical fiber tip was carefully cleaned to remove the Fluorescence in the normal CAM tissue was lower com- possible remaining drug on the tip. pared to fluorescence in the tumor tissue, suggesting either a lower uptake or faster clearance rate from normal Human subject tissue of the CAM (Figure 1). The fluorescence retention After informed consent, 1 patient with histologically from 1 to 5 h post topical administration of Ce6-PVP in proven angiosarcoma was recruited in this pilot case bladder tumor xenografts on CAM was tabulated using study. One tumor was located on the scalp, and 2 at the the red to blue ratio algorithm and fitted into a ROC curve temperomandibular joint. These tumors had been previ- to validate the ability of Ce6-PVP to discriminate tumor ously treated with Ce6-PVP photodynamic therapy. The from adjacent normal CAM membrane. By applying a cut- patient was intravenously administered with Ce6-PVP off value to these ratios as a diagnostic criterion, it allows with a dose of 2.0 mg/kg for repeated photodynamic ther- the generation of sensitivity and specificity values to dis- apy on existing and a new angiosarcoma lesion on the tinguish tumor from healthy CAM. A cut-off red to blue scalp and face. Before light irradiation, fluorescence imag- ratio of > 1.08 gives the highest combined sensitivity and ing and spectroscopy were performed on 3 angiosarcoma specificity were 70.8% (95% CI 48.9% to 87.4%) and lesions, normal scalp and skin at 1 and 3 h post drug 83.3% (95% CI 62.6% to 95.3%) respectively (Figure 2). administration. The patient had to remain in subdued Raising the value to > 1.33 gives the sensitivity and specif- light throughout the imaging period. This study was icity values of 62.5% (95% CI 40.59% to 81.20%) and approved by the National Cancer Centre Singapore's Insti- 91.2% (95% CI 73.0% to 99.0%) respectively. tutional Review Board. Fluorescence imaging of tumor and normal organs in mice Data analysis models To evaluate the quality of discrimination between healthy Fluorescence imaging was performed on the bladder and tumor tissues of the fluorescence images in the CAM tumor xenograft, peritumoral muscle, and normal blad- model, the red to blue ratio algorithm was applied. Such der at 1 and 3 h post intravenous injection of Ce6-PVP in algorithm is independent of the geometries of excitation/ mice (Figure 3). Representative fluorescence images of collection of signals and the power of excitation during skin and various internal organs taken at 1 h post Ce6-PVP the fluorescence imaging process. The sensitivity and the administration are presented in Figure 4. Overall, tumor specificity of the classifier were calculated using the fluorescence was observed to be more intense compared receiver-operator characteristics (ROC) curves by plotting to the adjacent peritumoral muscle. Fluorescence inten- the fluorescence intensity of tumor against the fluores- sity in bladder tumor was also higher compare to fluores- cence intensity of normal CAM tissue using the GraphPad cence of normal bladder tissues. The internal organs were software (GraphPad Prism™ Version 4.03, San Diego, also found to yield substantial fluorescence especially the USA). The ROC curve illustrates the trade-off between sen- gall bladder, liver, stomach, small and gastrointestinal sitivity and specificity for the different threshold of red to tracts. Minimal fluorescence was observed in the heart blue ratios to distinguish healthy from tumor tissue. Next, and spleen. The fluorescence intensity in muscle, skin, the cut-off value corresponding to the highest combined lung, liver, and bladder dropped at 3 h post drug admin- sensitivity and specificity was chosen and evaluated. For istration. There was little or no fluorescence remaining in fluorescence spectroscopy, the emission spectra data pre- heart and spleen. In contrast, the tumors showed sus- sented in the results are the absolute fluorescence intensi- tained fluorescence intensity at 3 h post drug administra- ties of the tissues after pre-processed using Fourier tion. By 6 h post drug administration, minimal transformation to decrease noise levels and normalizing fluorescence was detected in the gastrointestinal tract, the spectra to baseline at 700 nm. The normalization pro- liver and bladder (Figure 5). Page 4 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 (A) Tumor Normal CAM (B) Tumor Egg shell Normal CAM (C) Tumor Egg shell Normal CAM F Figure 1 luorescence imaging of MGH human bladder tumor xenografted on the CAM model Fluorescence imaging of MGH human bladder tumor xenografted on the CAM model. (A) White light image of the tumor before drug administration, (B) Ce6-PVP induced red fluorescence in tumor imaged under blue light illumination at 3 h post drug administration. Minimal fluorescence was observed in the adjacent normal CAM. (C) By displaying the fluores- cent image in a pseudo color using simple image processing technique, a clear discrimination of the tumor border can be visu- alized. Page 5 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 > 1.082 1.0 0.5 0.0 Tumor Adjacent normal tissue A p Figure 2 o scatter plot comparing st topical drug administr the ation fluorescence intensity in tumor and the adjacent normal CAM tissue was compiled from 1 – 5 h A scatter plot comparing the fluorescence intensity in tumor and the adjacent normal CAM tissue was com- piled from 1 – 5 h post topical drug administration. The points on the scatter plot are normalized individual measure- ments from 24 eggs. The dotted line is the cut-off fluorescence intensity threshold derived from the ROC curve to classify tumor from normal tissue with a sensitivity and specificity of 70.8% (95% CI 48.9% to 87.4%) and 83.3% (95% CI 62.6% to 95.3%) respectively. In vivo fiber optic spectrofluorometric measurement red fluorescence intensity had visibly higher spectral peak, Typical fluorescence emission spectra from tumor, adja- thus suggesting that that the macroscopic fluorescence cent peritumoral muscle and normal bladder after i.v. imaging were reproducible. administration of Ce6-PVP are shown in Figure 6. In gen- Fluorescence imaging and spectrofluorometric eral, the peak fluorescence intensities of tumor were higher than those of normal sites. The greatest intensity measurement in a patient occurred in the region between 660 – 670 nm. When the Fluorescence imaging and spectroscopy carried out on 3 spectra were normalized to baseline value, changes in tumors in an angiosarcoma patient showed that the peak intensity became evident. The line-shape differences tumors developed maximum fluorescence emission were predominantly due to increased Ce6-PVP accumula- intensity at 3 h post Ce6-PVP administration (Figure 8). tion of tumor in the red region (emission peak at 665 No observable variations were found for the intensity of nm). Fluorescence emission spectra of skin, heart, lung, the fluorescence between tumors. The fluorescence kinet- gall bladder, liver, spleen, kidney and gastrointestinal ics study findings from this consistent to those of our ear- tract are shown in Figure 7. Except for the gall bladder, all lier results [17]. Spectra from tumor areas show a clear other organs showed a decreased of fluorescence emission distinct Ce6-PVP induced fluorescence spectrum that dis- of Ce6-PVP at 3 h compared to 1 h post drug administra- criminate between the tumor and normal skin, with tion. Essentially, fluorescence images that showed greater Page 6 of 13 (page number not for citation purposes) Red to blue intensity ratio BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 1 h post Ce6-PVP administration White light Fluorescence Muscle (B) (A) Muscle Tumor Tumor (D) (C) Bladder Bladder 3 h post Ce6-PVP administration White light Fluorescence Tumor (E) (F) Muscle Muscle Tumor (G) (H) Bladder Bladder Mac tion of 5 mg Figure 3 roscopic fluorescence imaging /kg Ce6-PVP in bladder tumor xenografts and normal bladder at 1 h and 3 h post-intravenous administra- Macroscopic fluorescence imaging in bladder tumor xenografts and normal bladder at 1 h and 3 h post-intrave- nous administration of 5 mg/kg Ce6-PVP. Ce6-PVP induced fluorescence can be characterized by red fluorescence. At both time points, higher fluorescence intensity was observed in the tumor compared to the adjacent muscle and normal blad- der of the mice. Page 7 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 White light Fluorescence (B) (A) skin skin (D) (C) spleen spleen (E) (F) heart heart (G) (H) lung lung (I) (J) liver liver gall bladder gall bladder intestines intestines (K) (L) Mac a Figure 4 t 1 h post roscop-in ic white travenous a light and fluorescence imagi dministration of 5.0 mg ng /kg Ce6-PVP in skin, heart, lung, gall bladder, liver, spleen, kidney and gastrointestinal tract Macroscopic white light and fluorescence imaging in skin, heart, lung, gall bladder, liver, spleen, kidney and gastrointestinal tract at 1 h post-intravenous administration of 5.0 mg/kg Ce6-PVP. Page 8 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 Fluorescence White light (A) (B) bladder bladder (D) (C) liver liver gall bladder gall bladder (F) (E) intestines intestines Macro tr Figure 5 act at s 6 h post-intr copic white light ( avenou A, C, E) s admi and f nistra luores tion of 5.0 mg/kg Ce6-PVP cence (B, D, F) imaging in normal bladder, liver, gall bladder, and gastrointestinal Macroscopic white light (A, C, E) and fluorescence (B, D, F) imaging in normal bladder, liver, gall bladder, and gastrointestinal tract at 6 h post-intravenous administration of 5.0 mg/kg Ce6-PVP. tumor showing higher fluorescence emission intensity Discussion compared to normal tissue. Fluorescence imaging approaches are increasingly being used as a medical diagnostic procedure to assess tissue malignancy over conventional methods because they do Page 9 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 1 h post Ce6-PVP 3 h post Ce6-PVP 0.007 0.007 Tumor Tumor 0.006 0.006 Bladder Bladder 0.005 0.005 Muscle Muscle 0.004 0.004 0.003 0.003 0.002 0.002 0.001 0.001 0.000 0.000 650 660 670 680 690 700 650 660 670 680 690 700 Wavelengt h, nm Wavelengt h nm Co a Figure 6 dm mpar inistr isaon o tion of Ce6-PVP using f emission spectra of 400 nm bladde ex r tcitation umor xenograft, normal bladder and muscle of the murine model at 1 and 3 h post Comparison of emission spectra of bladder tumor xenograft, normal bladder and muscle of the murine model at 1 and 3 h post administration of Ce6-PVP using 400 nm excitation. The spectral signatures showed a peak at the wavelength 665 – 670 nm in tumor while the fluorescence intensity of normal bladder and muscle is weaker than that of the tumor tissue. not use potentially harmful ionizing radiation [20]. In sit- Results from the CAM experiments have provided the uations where discrimination of suspicious lesion is clin- motivation to examine Ce6-PVP fluorescence distribution ically problematic, fluorescence imaging may provide in bladder tumor xenograft as well as in various normal added advantage in demarcating abnormal tissue. The organs of a murine model. Macroscopic fluorescence development of photosensitizer based fluorescence imag- imaging showed that there was considerable distinction in ing is hindered by problems such as skin photosensitivity, the localization of fluorescence in tumor compared to poor selectivity of the photosensitizer, and formulation other organs that could enable discrimination between issues. For these reasons, Ce6 was formulated with PVP to tumor and normal organs. Organs of elimination and address these issues. Formulations using biocompatible detoxification such as skin, gall bladder and gastrointesti- polymers such as PVP are increasingly being used in the nal tracts were characterized by high photosensitizer accu- pharmaceutical industry for enhancing drug solubility mulation efficiency. In contrast, all other normal organs and bioavailability. Complexation of Ce6 with PVP was such as muscle and bladder had much lower photosensi- found to prevent Ce6 aggregation in aqueous media and tizer accumulation at 1 h post drug administration. Blood led to an enhancement of Ce6 fluorescence quantum vessels growing on the tumor can also be observed yield, while keeping the quantum yield of the intersystem because they contrast with the fluorescence of the tumors. crossing essentially unchanged [21]. In this study we have At 3 – 6 h post drug administration, a decrease of fluores- further examined the potential clinical use of Ce6-PVP in cence intensity became evident on all normal organs, con- cancer imaging and diagnosis. We first tested out the fea- firming that Ce6-PVP has fast clearance rate from normal sibility of this photosensitizer as an exogenous fluoro- organs. In some instances, we have observed variability of phore on the CAM tumor model. We were able to observe fluorescence intensity on the surface tissue such as stom- the red fluorescence emitted by the tumor tissue excited ach and lung. This is possibly attributed to the variations using a filtered xenon lamp excitation, which enabled of the tissue optical properties of the organs given by their clear determination of the tumor margin. The sensitivity color, density and composition. of Ce6-PVP was more than 80%, however, the specificity remained low. We have recently reported that the new for- The key issue in fluorescence imaging is that the emitted mulation of Ce6-PVP (> 95% purity level of active Ce6) fluorescence intensity measured from a tissue surface is demonstrated a higher sensitivity (98%) and specificity not necessarily proportional to the fluorophore concen- (82%) on the CAM model [15]. tration because the light is altered by the tissue's intrinsic Page 10 of 13 (page number not for citation purposes) Fluorescence intensity, a.u. Fluorescence intensity, a.u. BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 0.010 0.010 0.010 1 h 1 h 1 h Skin Heart Lung 3 h 3 h 3 h 0.007 0.007 0.007 0.005 0.005 0.005 0.002 0.002 0.002 0.000 0.000 0.000 650 660 670 680 690 700 650 660 670 680 690 700 650 660 670 680 690 700 W avelengt h, nm W avelengt h, nm Wavelengt h, nm 0.010 0.010 0.010 1 h 1 h Gall bladder GI tract 1 h Liver 3 h 3 h 3 h 0.007 0.007 0.007 0.005 0.005 0.005 0.002 0.002 0.002 0.000 0.000 0.000 650 660 670 680 690 700 650 660 670 680 690 700 650 660 670 680 690 700 W avelength, nm Wavelengt h, mn W avelengt h, nm 0.010 0.010 0.010 1 h 1 h 1 h Stomach Kidney Spleen 3 h 3 h 3 h 0.007 0.007 0.007 0.005 0.005 0.005 0.002 0.002 0.002 0.000 0.000 0.000 650 660 670 680 690 700 650 660 670 680 690 700 650 660 670 680 690 700 Wavelengt h, nm Wavelengt h, nm W avelength, nm Compar Ce6-PVP administra Figure 7 ison of emission tion spectra in the 650 – 700 nm region using 400 nm excitation in various normal organs at 1 and 3 h post Comparison of emission spectra in the 650 – 700 nm region using 400 nm excitation in various normal organs at 1 and 3 h post Ce6-PVP administration. Except for skin and gall bladder, it is evident that the emission spectra of nor- mal organs were lower compared to the emission spectra of tumor. absorption and scattering properties. Hence we have tizer and tissue properties, this method can potentially employed the utility of spectrometric point fluorescence improve the assessment of cancer location and its extent detection as a complementary technique. Spectra meas- within the local-regional area. While fluorescence point urements were carried out at 1 and 3 h post drug admin- spectroscopy studies are promising, it has several draw- istration to correlate the tumor intensity ratios obtained backs as a screening tool as it can only interrogate a small with fluorescence imaging to the tumor fluorescence spec- volume of tissue (typically, 0.5 – 1 mm ) directly beneath tral signal of the tissue. All the macroscopic fluorescence the probe tip. Point measurements inevitably involve a images correlated well to the spectra measurement. The degree of random sampling, which may not allow identi- Ce6-PVP induced spectra emission after normalization fication of early stage disease [22]. Hence, the combina- demonstrated a good separation to differentiate malig- tion techniques of fluorescence imaging and spectroscopy nant tumors from normal tissues. Besides measuring have been proven in good agreement with the actual physical parameters such as concentration of photosensi- Page 11 of 13 (page number not for citation purposes) Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 (A) White light at 3 h post Ce6-PVP (B) Fluorescence at 3 h post Ce6-PVP (C) (D) 0.025 0.025 Lesion at 3 h Lesion at 1 h Normal at 3 h Normal at 1 h 0.020 0.020 0.015 0.015 0.010 0.010 0.005 0.005 0.000 0.000 650 660 670 680 690 700 650 660 670 680 690 700 W avelengt h, nm W avelengt h, nm Fl n Figure 8 o uorescence i rmal skin on the scalp at 1 h maging and spectro and 3 h p fluorometri ost-ic measurement o ntravenous administr f fluo ation of 2 rescence em .0 mg is/kg Ce6-PVP sion in angiosarcoma lesion and surrounding Fluorescence imaging and spectrofluorometric measurement of fluorescence emission in angiosarcoma lesion and surrounding normal skin on the scalp at 1 h and 3 h post-intravenous administration of 2.0 mg/kg Ce6- PVP. tumor boundary found by histopathological mapping fluorescence imaging clearly captures the fluorescence in and early stage of disease [23,24]. angiosarcoma and good correlation was found between fluorescence imaging and spectral measurement in the Recently, we have reported photodynamic therapy – acti- patient. This is in agreement with other reports, that fluo- vated immune response against distant untreated rescence ratio imaging in combination with relative spec- tumours in recurrent angiosarcoma [25] and preferential tral measurement of the photosensitizer might be a viable accumulation of Ce6-PVP in angiosarcoma compared to method for the optical diagnosis of cancer [27]. Further- normal skin following intravenous administration in 3 more, in vivo and real-time determination of the time patients [17]. High dose PDT carried out at a high fluence course of photosensitizer's fluorescence could potentially rate resulted in local control of the disease for up to a year; be a crucial pre-irradiation screening tool to determine the however, the disease recurred and PDT had to be repeated exact location and extent of the tumor before photody- [26]. During the repeat PDT session, we measured the flu- namic therapy. orescence of 3 different lesions using fluorescence imag- ing followed by spectroscopy. The results confirmed that Page 12 of 13 (page number not for citation purposes) Fluorescence em ission, a.u. Fluorescence em ission, a.u. BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 13. Chin WW, Heng PW, Olivo M: Chlorin e6 – polyvinylpyrro- Conclusion lidone mediated photosensitization is effective against It is shown that Ce6-PVP has a rapid accumulation in the human non-small cell lung carcinoma compared to small cell tumor, and a relatively short half-life in normal organs. lung carcinoma xenografts. BMC Pharmacol 2007, 7:15. 14. Chin WW, Lau WK, Bhuvaneswari R, Heng PW, Olivo M: Chlorin When excited by blue light, Ce6-PVP accumulating cells e6-polyvinylpyrrolidone as a fluorescent marker for fluores- can be visualized and located in the tissue by virtue of its cence diagnosis of human bladder cancer implanted on the chick chorioallantoic membrane model. Cancer Lett 2007, fluorescence. The main advantage of Ce6-PVP induced 245(1–2):127-133. fluorescence imaging is its increased tumor selectivity 15. Chin WW, Heng PW, Lim PL, Lau KO, Olivo M: Membrane trans- with the ability to clearly define the tumor margin. Com- port enhancement of chlorin e6-polyvinylpyrrolidone and its photodynamic efficacy on the chick chorioallantoic model. bination of Ce6-PVP induced fluorescence imaging and Journal of Biophotonics 2008, 1(5):395-407. spectroscopy could allow detection and discrimination 16. Chin WW, Lau WK, Heng PW, Bhuvaneswari R, Olivo M: Fluores- between cancer and the surrounding normal tissues. cence imaging and phototoxicity effects of new formulation of chlorin e6-polyvinylpyrrolidone. J Photochem Photobiol B 2006, 84(2):103-110. Competing interests 17. Chin WW, Heng PW, Thong PS, Bhuvaneswari R, Hirt W, Kuenzel S, Soo KC, Olivo M: Improved formulation of photosensitizer The authors declare that they have no competing interests. chlorin e6 polyvinylpyrrolidone for fluorescence diagnostic imaging and photodynamic therapy of human cancer. Eur J Authors' contributions Pharm Biopharm 2008. 18. Chin WW, Heng PW, Bhuvaneswari R, Lau WK, Olivo M: The WWC and PST conceived of the study and carried out all potential application of chlorin e6-polyvinylpyrrolidone for- the experimental study. RB and KCS participated in the mulation in photodynamic therapy. Photochem Photobiol Sci 2006, 5(11):1031-1037. clinical study. PWH and MO participated in the coordina- 19. Isakau HA, Trukhacheva TV, Zhebentyaev AI, Petrov PT: HPLC tion of the study. All authors read and approved the final study of chlorin e6 and its molecular complex with polyvi- manuscript nylpyrrolidone. Biomed Chromatogr 2007, 21(3):318-325. 20. Brindle K: New approaches for imaging tumour responses to treatment. Nat Rev Cancer 2008, 8(2):94-107. Acknowledgements 21. Isakau HA, Parkhats MV, Knyukshto VN, Dzhagarov BM, Petrov EP, WW Chin and PS Thong are the recipients of the Singapore Millennium Petrov PT: Toward understanding the high PDT efficacy of chlorin e6-polyvinylpyrrolidone formulations: Photophysical Foundation scholarship. and molecular aspects of photosensitizer-polymer interac- tion in vitro. J Photochem Photobiol B 2008, 92(3):165-174. References 22. Stringer M, Moghissi K: Photodiagnosis and fluorescence imag- 1. DaCosta RS, Wilson BC, Marcon NE: Fluorescence and spectral ing in clinical practice. Photodiagnosis and Photodynamic Therapy imaging. ScientificWorldJournal 2007, 7:2046-2071. 2004, 1(2):9-12. 2. Wong Kee Song LM, Marcon NE: Fluorescence and Raman spec- 23. Wagnieres GA, Star WM, Wilson BC: In vivo fluorescence spec- troscopy. Gastrointest Endosc Clin N Am 2003, 13(2):279-296. troscopy and imaging for oncological applications. Photochem 3. Perelman LT: Optical diagnostic technology based on light Photobiol 1998, 68(5):603-632. scattering spectroscopy for early cancer detection. Expert Rev 24. Mang T, Kost J, Sullivan M, Wilson BC: Autofluorescence and Med Devices 2006, 3(6):787-803. Photofrin-induced fluorescence imaging and spectroscopy in 4. Schultz CP: The potential role of Fourier transform infrared an animal model of oral cancer. Photodiagnosis and Photodynamic spectroscopy and imaging in cancer diagnosis incorporating Therapy 2006, 3(3):168-176. complex mathematical methods. Technol Cancer Res Treat 2002, 25. Thong PS, Ong KW, Goh NS, Kho KW, Manivasager V, Bhuvaneswari 1(2):95-104. R, Olivo M, Soo KC: Photodynamic-therapy-activated immune 5. Ell C: Improving endoscopic resolution and sampling: fluores- response against distant untreated tumours in recurrent cence techniques. Gut 2003, 52(Suppl 4):iv30-33. angiosarcoma. Lancet Oncol 2007, 8(10):950-952. 6. Kherlopian AR, Song T, Duan Q, Neimark MA, Po MJ, Gohagan JK, 26. Thong PS, Olivo M, Kho KW, Bhuvaneswari R, Chin WW, Ong KW, Laine AF: A review of imaging techniques for systems biology. Soo KC: Immune response against angiosarcoma following BMC Syst Biol 2008, 2:74. lower fluence rate clinical photodynamic therapy. J Environ 7. Chen Y, Gryshuk A, Achilefu S, Ohulchansky T, Potter W, Zhong T, Pathol Toxicol Oncol 2008, 27(1):35-42. Morgan J, Chance B, Prasad PN, Henderson BW, et al.: A novel 27. Kopriva I, Persin A, Zorc H, Pasic A, Lipozencic J, Kostovic K, approach to a bifunctional photosensitizer for tumor imag- Loncaric M: Visualization of basal cell carcinoma by fluores- ing and phototherapy. Bioconjug Chem 2005, 16(5):1264-1274. cence diagnosis and independent component analysis. Photo- 8. Thong PS, Olivo M, Kho KW, Zheng W, Mancer K, Harris M, Soo KC: diagnosis and Photodynamic Therapy 2007, 4(3):190-196. Laser confocal endomicroscopy as a novel technique for flu- orescence diagnostic imaging of the oral cavity. J Biomed Opt Pre-publication history 2007, 12(1):014007. 9. Andersson-Engels S, Klinteberg C, Svanberg K, Svanberg S: In vivo The pre-publication history for this paper can be accessed fluorescence imaging for tissue diagnostics. Phys Med Biol 1997, here: 42(5):815-824. 10. Zheng W, Olivo M, Soo KC: The use of digitized endoscopic imaging of 5-ALA-induced PPIX fluorescence to detect and http://www.biomedcentral.com/1471-2342/9/1/prepub diagnose oral premalignant and malignant lesions in vivo. Int J Cancer 2004, 110(2):295-300. 11. Kostenich GA, Zhuravkin IN, Furmanchuk AV, Zhavrid EA: Photo- dynamic therapy with chlorin e6. A morphologic study of tumor damage efficiency in experiment. J Photochem Photobiol B 1991, 11(3–4):307-318. 12. Ramaswamy B, Manivasager V, Chin WW, Soo KC, Olivo M: Photo- dynamic diagnosis of a human nasopharyngeal carcinoma xenograft model using the novel Chlorin e6 photosensitizer Fotolon. Int J Oncol 2005, 26(6):1501-1506. Page 13 of 13 (page number not for citation purposes) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Medical Imaging Springer Journals

In-vivo optical detection of cancer using chlorin e6 – polyvinylpyrrolidone induced fluorescence imaging and spectroscopy

Loading next page...
 
/lp/springer-journals/in-vivo-optical-detection-of-cancer-using-chlorin-e6-J7kZ43lfpw

References (54)

Publisher
Springer Journals
Copyright
Copyright © 2009 by Chin et al; licensee BioMed Central Ltd.
Subject
Medicine & Public Health; Imaging / Radiology
eISSN
1471-2342
DOI
10.1186/1471-2342-9-1
pmid
19133127
Publisher site
See Article on Publisher Site

Abstract

Background: Photosensitizer based fluorescence imaging and spectroscopy is fast becoming a promising approach for cancer detection. The purpose of this study was to examine the use of the photosensitizer chlorin e6 (Ce6) formulated in polyvinylpyrrolidone (PVP) as a potential exogenous fluorophore for fluorescence imaging and spectroscopic detection of human cancer tissue xenografted in preclinical models as well as in a patient. Methods: Fluorescence imaging was performed on MGH human bladder tumor xenografted on both the chick chorioallantoic membrane (CAM) and the murine model using a fluorescence endoscopy imaging system. In addition, fiber optic based fluorescence spectroscopy was performed on tumors and various normal organs in the same mice to validate the macroscopic images. In one patient, fluorescence imaging was performed on angiosarcoma lesions and normal skin in conjunction with fluorescence spectroscopy to validate Ce6-PVP induced fluorescence visual assessment of the lesions. Results: Margins of tumor xenografts in the CAM model were clearly outlined under fluorescence imaging. Ce6-PVP-induced fluorescence imaging yielded a specificity of 83% on the CAM model. In mice, fluorescence intensity of Ce6-PVP was higher in bladder tumor compared to adjacent muscle and normal bladder. Clinical results confirmed that fluorescence imaging clearly captured the fluorescence of Ce6-PVP in angiosarcoma lesions and good correlation was found between fluorescence imaging and spectral measurement in the patient. Conclusion: Combination of Ce6-PVP induced fluorescence imaging and spectroscopy could allow for optical detection and discrimination between cancer and the surrounding normal tissues. Ce6-PVP seems to be a promising fluorophore for fluorescence diagnosis of cancer. Page 1 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 A number of fluorochromes such as fluorescein, toluidine Background As with most cancers, early diagnosis is critical to achieve blue, cyanine dyes and indocyanine green have been favorable prognosis. Currently, random surveillance biop- described with variable stabilities, quantum efficiencies, sies are the existing gold standard for the identification of and ease of synthesis. However, most of the fluoro- lesions in pre-neoplastic conditions. However this chromes are not tumor specific and are rapidly eliminated method is prone to sampling error, time-consuming, sub- from the organism. Chemically and endogenously synthe- jective and cost-inefficient. A diagnostic method that sized fluorochromes such as porphyrin based photosensi- could provide rapid, automated classification of cancer tizers have properties that may be utilized both lesions would increase the efficiency and comprehensive- experimentally and clinically. Porphyrins have been ness of malignancy screening and surveillance procedures. known to naturally localize in malignant tissue where A variety of optical techniques have recently been utilized they emit light when irradiated at certain wavelengths, for the diagnostic study of cancerous tissue. These include providing a means to detect tumor by the location of its fluorescence spectroscopy [1], Raman spectroscopy [2], fluorescence. However, one of the major limitation is its light scattering spectroscopy [3], and Fourier-transform slow clearance from tissues and long period of skin pho- infrared spectroscopy [4]. These optical spectroscopic totoxicity. Moreover, the porphyrin's core absorbs wave- techniques are capable of providing biochemical and lengths of light too short for optimal penetration in tissue. morphological information in short integration times, As such, by reducing a pyrrole double bond on the por- which can be used for automated diagnosis of intact tis- phyrin periphery, a chlorin core compound can be gener- sue. However, in order to be useful as a comprehensive ated with a high absorption at longer wavelengths of 660 screening procedure, the optical technique must allow – 670 nm that can penetrate deeper in human tissue than rapid real time imaging of a large area of tissue rather than those of porphyrins. Of particular interest among the eval- point by point measurement, such that suspicious regions uated chlorins is the naturally occurring chlorin e6 (Ce6) could be identified accurately and biopsied for his- [11]. Ce6 has improved efficacy and has decreased side topathologic correlation [5]. effects compared to first generation photosensitizers from hematoporphyrin derivatives. Due to the importance of With the advent of molecular probes, imaging methods Ce6's characteristic fluorescence properties, there is a need such as ultrasound, microCT (Computed Tomography), to identify new formulations that are stable, exhibit ease microMRI (Magnetic Resonance Imaging), and microPET in manufacturing and selectively deliver the photosensi- (Positron Emission Tomography) can be conducted not tizer to target tissue in an efficient manner. Hence, we only to visualize gross anatomical structures, but also to have investigated the use of Ce6 in combination with the visualize substructures of cells and monitor molecule polymer polyvinylpyrrolidone (Ce6-PVP). Polyvinylpyr- dynamics [6]. Imaging of endogenous or exogenous fluor- rolidone is one of the most important excipient used in ochromes has several important advantages over other modern pharmaceutical technology. We have previously optical approaches for tumor imaging. This imaging tech- described the selective localization and photodynamic nique relies on fluorochrome induced fluorescence, activity of Ce6-PVP in nasopharyngeal and lung carci- reflectance, absorption or bioluminescence as the source noma models that provided rationale for its use as a ther- of contrast, while imaging systems can be based on diffuse apeutic agent for photodynamic therapy [12,13]. By optical tomography, surface-weighted imaging, phase- employing a chick chorioallantoic membrane model, array detection, intensified matrix detector and charged- Ce6-PVP was shown to selective accumulate in bladder coupled device camera detection, confocal endomicros- tumors xenografts and had a faster clearance from normal copy, multiphoton imaging, or microscopic imaging with CAM when administered topically compared to system- intravital microscopy [7,8]. Fluorescence ratio imaging is atic administration [14]. The uptake ratio of Ce6-PVP was a method widely used for optical diagnosis of cancer after found to have a 2-fold increase across the CAM when administration of a photosensitizer [9]. Enhanced con- compared to that of Ce6, indicating that PVP was able to trast between tumor and adjacent normal tissue can be facilitate diffusion of Ce6 across the membrane [15]. Fur- obtained based on calculating the ratio between red inten- thermore, Ce6-PVP had less in vivo systemic phototoxic sity of the photosensitizer (600–700 nm) over the blue/ effect compared to Ce6 alone after light irradiation in green intensity of the back-scattered excitation light or tis- photodynamic therapy in mice bearing tumors [16]. sue autofluorescence (450–550 nm). Many investigations Using a chemical fluorescence extraction technique and have confirmed good agreement with the histopathologi- cuvette-based spectrofluorimetry, our data demonstrated cal extent of the tumor, implying that this technique can that the distribution of Ce6-PVP drug were much lower in be applied as a useful tool for indicating tumor boundary normal organs such liver, spleen, kidney, brain, heart and [10]. lung compared to Ce6 delivered using dimethylsulfoxide (DMSO) [17]. We also postulated that the extent of tumor necrosis post Ce6-PVP mediated photodynamic therapy Page 2 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 (PDT) was dependent on the plasma concentration of window was resealed to avoid evaporation of the drug Ce6-PVP, implying a vascular mediated cell death mecha- solution from the CAM. After 30 min incubation, macro- nism [18]. scopic fluorescence imaging was performed at 0.5, 1, 2, 3, 4, and 5 h post drug administration using a commercially In this study, we have evaluated the usefulness of Ce6-PVP available fluorescence endoscopic system (Karl Storz, Tut- to accurately define the margin of the tumor from its nor- tlingen, Germany). A modified xenon short arc lamp (D – mal adjacent tissue in the chick choriallantoic membrane Light system in blue light mode, Karl Storz) filtered by a (CAM) tumor model. We also presented visual informa- band pass filter (380 – 450 nm) was used for excitation of tion on Ce6-PVP induced fluorescence in tumor and gross photosensitizer in tissue. Fluorescence was captured via a anatomical structure of normal organ of a murine model. sensitive CCD camera (Tricam SL PAL, Karl Storz) Fluorescence spectroscopy measurements were also per- attached to an endoscope integrated with a long pass filter formed to characterize emission spectra from these tissue (cut-off wavelength 470 nm). This observation LP filter of samples as well as to corroborate results from fluorescence the endoscope only minimally transmits the diffuse back- images. Finally, a pilot trial was carried out to validate the scattering excitation light with a peak at 450 nm (blue use of Ce6-PVP as a clinically relevant diagnostic photo- light), while has a transmission of over 98% in the 470 – sensitizer using both imaging and spectroscopy modality 800 nm range. The red channel registered the photosensi- for differentiation of normal and tumor tissue in a patient. tizer's fluorescence and the blue channel captured the dif- fusely back-scattered excitation light. A short exposure of Methods the surface of tissue to the excitation light (10 s) was per- Photosensitizer formed to avoid excessive photobleaching effects. White The formulation of Ce6-PVP, also known as Fotolon or light imaging was used to correlate the boundaries of Photolon was supplied by HAEMATO-science GmbH, tumors and organs. All procedures involving preparation Germany. It is a co-lyophilisate of Ce6 sodium salt and and administration of the photosensitizer were conducted PVP (a pharmaceutical grade polymer, molecular mass ≈ under low ambient lighting. 12,000 g/mol) in a 1:1 mass ratio [19]. Murine tumor model Cell culture A total of 10 Balb/c athymic nude mice and C57 mice, 6– MGH (European Collection of Cell Cultures), a poorly 8 weeks of age, weighing an average of 24 g were obtained differentiated human bladder carcinoma cells were grown from the Animal Resource Centre, Western Australia and as a monolayer in RPMI-1640 medium supplemented Centre for Animal Resources, National University of Sin- with 10% fetal bovine serum, 1% non-essential amino gapore respectively. Before inoculation, the cell layer was acids (Gibco, USA), 1% sodium pyruvate (Gibco, USA), washed with phosphate-buffered saline, trypsinized, and -1 6 100 units mL penicillin-streptomycin (Gibco, USA) and counted using a hemocytometer. Approximately 3.0 × 10 incubated at 37°C, 95% humidity and 5% CO . MGH cells suspended in 150 μl of Hanks' Balanced Salt Solution (Gibco, USA) were injected subcutaneously into Chick choriallantoic membrane tumor model the lower flanks of Balb/c athymic nude mice. The ani- Fertilized chicken eggs were incubated at 37°C in a mals were used for experiments when the tumors meas- humidified atmosphere inside a hatching incubator ured around 7 – 10 mm in diameter. This ensured that the equipped with an automatic rotator (Octagon 20, Brinsea, tumor sizes were kept consistent to minimize variations Somerset, UK). At embryo age (EA) 7, a window of about due to the degree of vascularization of the implants. Mice 1.5 cm was opened in the eggshell to detach the shell were injected with a dose of 5 mg/kg of Ce6-PVP via tail membrane from the developing CAM. Then, the window vein injection. At 1, 3 and 6 h, mice were sacrificed and was sealed with sterilized parafilm to avoid contamina- the skin overlaying the tumor was carefully removed to tion and the eggs were returned to the static incubator for expose the tumor and normal peritumoral muscle for flu- further incubation until the day of experiment. On EA 9, orescence imaging. C57 mice were used for imaging and approximately 5–10 × 10 MGH cells were inoculated on spectroscopy of normal organs. All procedures were the CAM. The window of the eggs were resealed with ster- approved by the Institutional Animal Care and Use Com- ile parafilm and returned to the static incubator. Grafted mittee, SingHealth, Singapore, in accordance with inter- cells were allowed to grow on the CAM for up to 5 days. national standards. On EA 14, Ce6-PVP was dissolved in 0.9% sodium chlo- Spectroscopic measurement using fiber optics-based ride to constitute a stock solution of 1 mg/mL. The stock solution was further diluted to obtain a volume of 500 μL fluorescence spectrometer containing a dose of 1 mg/kg body weight of the chick's The spectral measurement was performed on mice sacri- embryo. The photosensitizer was applied on the entire ficed at 1 and 3 h post Ce6-PVP administration. A fiber surface of the CAM and left to incubate for 30 min. The optics-based fluorescence spectrometer (Spex SkinSkan, Page 3 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 JY Inc., Edison, NJ, USA) was used for the measurement of cedure significantly reduced the within-class variances. fluorescence intensity of Ce6-PVP. A monochromator Spectral data from the various organs and tissues were with a 150-W Xenon lamp was used as the excitation light analyzed to determine spectral line shape and the peak source. The excitation light (400 nm) was guided to illu- fluorescence intensities at region of 660 – 690 nm. minate samples by one arm of a Y-type quartz fiber bun- dle, and the emission fluorescence was collected by Results Fluorescence imaging of bladder tumor xenografts on the another arm of the fiber bundle, guided to another motor- controlled monochromator. The resulting emission spec- CAM model tra were recorded from 650 to 750 nm, in 1 nm incre- Fluorescence was not observed from the tumors under ments, collected using the DataMax version 2.20 blue light illumination before drug administration. After (Instruments SA, Inc.) software package. The optical fiber topical administration of Ce6-PVP, an intense red fluores- tip was placed on the measuring sites and fluorescence cence in the bladder tumor xenografts was observed, sug- intensity spectra were measured. After each measurement, gesting selective localization in the malignant cells. the optical fiber tip was carefully cleaned to remove the Fluorescence in the normal CAM tissue was lower com- possible remaining drug on the tip. pared to fluorescence in the tumor tissue, suggesting either a lower uptake or faster clearance rate from normal Human subject tissue of the CAM (Figure 1). The fluorescence retention After informed consent, 1 patient with histologically from 1 to 5 h post topical administration of Ce6-PVP in proven angiosarcoma was recruited in this pilot case bladder tumor xenografts on CAM was tabulated using study. One tumor was located on the scalp, and 2 at the the red to blue ratio algorithm and fitted into a ROC curve temperomandibular joint. These tumors had been previ- to validate the ability of Ce6-PVP to discriminate tumor ously treated with Ce6-PVP photodynamic therapy. The from adjacent normal CAM membrane. By applying a cut- patient was intravenously administered with Ce6-PVP off value to these ratios as a diagnostic criterion, it allows with a dose of 2.0 mg/kg for repeated photodynamic ther- the generation of sensitivity and specificity values to dis- apy on existing and a new angiosarcoma lesion on the tinguish tumor from healthy CAM. A cut-off red to blue scalp and face. Before light irradiation, fluorescence imag- ratio of > 1.08 gives the highest combined sensitivity and ing and spectroscopy were performed on 3 angiosarcoma specificity were 70.8% (95% CI 48.9% to 87.4%) and lesions, normal scalp and skin at 1 and 3 h post drug 83.3% (95% CI 62.6% to 95.3%) respectively (Figure 2). administration. The patient had to remain in subdued Raising the value to > 1.33 gives the sensitivity and specif- light throughout the imaging period. This study was icity values of 62.5% (95% CI 40.59% to 81.20%) and approved by the National Cancer Centre Singapore's Insti- 91.2% (95% CI 73.0% to 99.0%) respectively. tutional Review Board. Fluorescence imaging of tumor and normal organs in mice Data analysis models To evaluate the quality of discrimination between healthy Fluorescence imaging was performed on the bladder and tumor tissues of the fluorescence images in the CAM tumor xenograft, peritumoral muscle, and normal blad- model, the red to blue ratio algorithm was applied. Such der at 1 and 3 h post intravenous injection of Ce6-PVP in algorithm is independent of the geometries of excitation/ mice (Figure 3). Representative fluorescence images of collection of signals and the power of excitation during skin and various internal organs taken at 1 h post Ce6-PVP the fluorescence imaging process. The sensitivity and the administration are presented in Figure 4. Overall, tumor specificity of the classifier were calculated using the fluorescence was observed to be more intense compared receiver-operator characteristics (ROC) curves by plotting to the adjacent peritumoral muscle. Fluorescence inten- the fluorescence intensity of tumor against the fluores- sity in bladder tumor was also higher compare to fluores- cence intensity of normal CAM tissue using the GraphPad cence of normal bladder tissues. The internal organs were software (GraphPad Prism™ Version 4.03, San Diego, also found to yield substantial fluorescence especially the USA). The ROC curve illustrates the trade-off between sen- gall bladder, liver, stomach, small and gastrointestinal sitivity and specificity for the different threshold of red to tracts. Minimal fluorescence was observed in the heart blue ratios to distinguish healthy from tumor tissue. Next, and spleen. The fluorescence intensity in muscle, skin, the cut-off value corresponding to the highest combined lung, liver, and bladder dropped at 3 h post drug admin- sensitivity and specificity was chosen and evaluated. For istration. There was little or no fluorescence remaining in fluorescence spectroscopy, the emission spectra data pre- heart and spleen. In contrast, the tumors showed sus- sented in the results are the absolute fluorescence intensi- tained fluorescence intensity at 3 h post drug administra- ties of the tissues after pre-processed using Fourier tion. By 6 h post drug administration, minimal transformation to decrease noise levels and normalizing fluorescence was detected in the gastrointestinal tract, the spectra to baseline at 700 nm. The normalization pro- liver and bladder (Figure 5). Page 4 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 (A) Tumor Normal CAM (B) Tumor Egg shell Normal CAM (C) Tumor Egg shell Normal CAM F Figure 1 luorescence imaging of MGH human bladder tumor xenografted on the CAM model Fluorescence imaging of MGH human bladder tumor xenografted on the CAM model. (A) White light image of the tumor before drug administration, (B) Ce6-PVP induced red fluorescence in tumor imaged under blue light illumination at 3 h post drug administration. Minimal fluorescence was observed in the adjacent normal CAM. (C) By displaying the fluores- cent image in a pseudo color using simple image processing technique, a clear discrimination of the tumor border can be visu- alized. Page 5 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 > 1.082 1.0 0.5 0.0 Tumor Adjacent normal tissue A p Figure 2 o scatter plot comparing st topical drug administr the ation fluorescence intensity in tumor and the adjacent normal CAM tissue was compiled from 1 – 5 h A scatter plot comparing the fluorescence intensity in tumor and the adjacent normal CAM tissue was com- piled from 1 – 5 h post topical drug administration. The points on the scatter plot are normalized individual measure- ments from 24 eggs. The dotted line is the cut-off fluorescence intensity threshold derived from the ROC curve to classify tumor from normal tissue with a sensitivity and specificity of 70.8% (95% CI 48.9% to 87.4%) and 83.3% (95% CI 62.6% to 95.3%) respectively. In vivo fiber optic spectrofluorometric measurement red fluorescence intensity had visibly higher spectral peak, Typical fluorescence emission spectra from tumor, adja- thus suggesting that that the macroscopic fluorescence cent peritumoral muscle and normal bladder after i.v. imaging were reproducible. administration of Ce6-PVP are shown in Figure 6. In gen- Fluorescence imaging and spectrofluorometric eral, the peak fluorescence intensities of tumor were higher than those of normal sites. The greatest intensity measurement in a patient occurred in the region between 660 – 670 nm. When the Fluorescence imaging and spectroscopy carried out on 3 spectra were normalized to baseline value, changes in tumors in an angiosarcoma patient showed that the peak intensity became evident. The line-shape differences tumors developed maximum fluorescence emission were predominantly due to increased Ce6-PVP accumula- intensity at 3 h post Ce6-PVP administration (Figure 8). tion of tumor in the red region (emission peak at 665 No observable variations were found for the intensity of nm). Fluorescence emission spectra of skin, heart, lung, the fluorescence between tumors. The fluorescence kinet- gall bladder, liver, spleen, kidney and gastrointestinal ics study findings from this consistent to those of our ear- tract are shown in Figure 7. Except for the gall bladder, all lier results [17]. Spectra from tumor areas show a clear other organs showed a decreased of fluorescence emission distinct Ce6-PVP induced fluorescence spectrum that dis- of Ce6-PVP at 3 h compared to 1 h post drug administra- criminate between the tumor and normal skin, with tion. Essentially, fluorescence images that showed greater Page 6 of 13 (page number not for citation purposes) Red to blue intensity ratio BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 1 h post Ce6-PVP administration White light Fluorescence Muscle (B) (A) Muscle Tumor Tumor (D) (C) Bladder Bladder 3 h post Ce6-PVP administration White light Fluorescence Tumor (E) (F) Muscle Muscle Tumor (G) (H) Bladder Bladder Mac tion of 5 mg Figure 3 roscopic fluorescence imaging /kg Ce6-PVP in bladder tumor xenografts and normal bladder at 1 h and 3 h post-intravenous administra- Macroscopic fluorescence imaging in bladder tumor xenografts and normal bladder at 1 h and 3 h post-intrave- nous administration of 5 mg/kg Ce6-PVP. Ce6-PVP induced fluorescence can be characterized by red fluorescence. At both time points, higher fluorescence intensity was observed in the tumor compared to the adjacent muscle and normal blad- der of the mice. Page 7 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 White light Fluorescence (B) (A) skin skin (D) (C) spleen spleen (E) (F) heart heart (G) (H) lung lung (I) (J) liver liver gall bladder gall bladder intestines intestines (K) (L) Mac a Figure 4 t 1 h post roscop-in ic white travenous a light and fluorescence imagi dministration of 5.0 mg ng /kg Ce6-PVP in skin, heart, lung, gall bladder, liver, spleen, kidney and gastrointestinal tract Macroscopic white light and fluorescence imaging in skin, heart, lung, gall bladder, liver, spleen, kidney and gastrointestinal tract at 1 h post-intravenous administration of 5.0 mg/kg Ce6-PVP. Page 8 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 Fluorescence White light (A) (B) bladder bladder (D) (C) liver liver gall bladder gall bladder (F) (E) intestines intestines Macro tr Figure 5 act at s 6 h post-intr copic white light ( avenou A, C, E) s admi and f nistra luores tion of 5.0 mg/kg Ce6-PVP cence (B, D, F) imaging in normal bladder, liver, gall bladder, and gastrointestinal Macroscopic white light (A, C, E) and fluorescence (B, D, F) imaging in normal bladder, liver, gall bladder, and gastrointestinal tract at 6 h post-intravenous administration of 5.0 mg/kg Ce6-PVP. tumor showing higher fluorescence emission intensity Discussion compared to normal tissue. Fluorescence imaging approaches are increasingly being used as a medical diagnostic procedure to assess tissue malignancy over conventional methods because they do Page 9 of 13 (page number not for citation purposes) BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 1 h post Ce6-PVP 3 h post Ce6-PVP 0.007 0.007 Tumor Tumor 0.006 0.006 Bladder Bladder 0.005 0.005 Muscle Muscle 0.004 0.004 0.003 0.003 0.002 0.002 0.001 0.001 0.000 0.000 650 660 670 680 690 700 650 660 670 680 690 700 Wavelengt h, nm Wavelengt h nm Co a Figure 6 dm mpar inistr isaon o tion of Ce6-PVP using f emission spectra of 400 nm bladde ex r tcitation umor xenograft, normal bladder and muscle of the murine model at 1 and 3 h post Comparison of emission spectra of bladder tumor xenograft, normal bladder and muscle of the murine model at 1 and 3 h post administration of Ce6-PVP using 400 nm excitation. The spectral signatures showed a peak at the wavelength 665 – 670 nm in tumor while the fluorescence intensity of normal bladder and muscle is weaker than that of the tumor tissue. not use potentially harmful ionizing radiation [20]. In sit- Results from the CAM experiments have provided the uations where discrimination of suspicious lesion is clin- motivation to examine Ce6-PVP fluorescence distribution ically problematic, fluorescence imaging may provide in bladder tumor xenograft as well as in various normal added advantage in demarcating abnormal tissue. The organs of a murine model. Macroscopic fluorescence development of photosensitizer based fluorescence imag- imaging showed that there was considerable distinction in ing is hindered by problems such as skin photosensitivity, the localization of fluorescence in tumor compared to poor selectivity of the photosensitizer, and formulation other organs that could enable discrimination between issues. For these reasons, Ce6 was formulated with PVP to tumor and normal organs. Organs of elimination and address these issues. Formulations using biocompatible detoxification such as skin, gall bladder and gastrointesti- polymers such as PVP are increasingly being used in the nal tracts were characterized by high photosensitizer accu- pharmaceutical industry for enhancing drug solubility mulation efficiency. In contrast, all other normal organs and bioavailability. Complexation of Ce6 with PVP was such as muscle and bladder had much lower photosensi- found to prevent Ce6 aggregation in aqueous media and tizer accumulation at 1 h post drug administration. Blood led to an enhancement of Ce6 fluorescence quantum vessels growing on the tumor can also be observed yield, while keeping the quantum yield of the intersystem because they contrast with the fluorescence of the tumors. crossing essentially unchanged [21]. In this study we have At 3 – 6 h post drug administration, a decrease of fluores- further examined the potential clinical use of Ce6-PVP in cence intensity became evident on all normal organs, con- cancer imaging and diagnosis. We first tested out the fea- firming that Ce6-PVP has fast clearance rate from normal sibility of this photosensitizer as an exogenous fluoro- organs. In some instances, we have observed variability of phore on the CAM tumor model. We were able to observe fluorescence intensity on the surface tissue such as stom- the red fluorescence emitted by the tumor tissue excited ach and lung. This is possibly attributed to the variations using a filtered xenon lamp excitation, which enabled of the tissue optical properties of the organs given by their clear determination of the tumor margin. The sensitivity color, density and composition. of Ce6-PVP was more than 80%, however, the specificity remained low. We have recently reported that the new for- The key issue in fluorescence imaging is that the emitted mulation of Ce6-PVP (> 95% purity level of active Ce6) fluorescence intensity measured from a tissue surface is demonstrated a higher sensitivity (98%) and specificity not necessarily proportional to the fluorophore concen- (82%) on the CAM model [15]. tration because the light is altered by the tissue's intrinsic Page 10 of 13 (page number not for citation purposes) Fluorescence intensity, a.u. Fluorescence intensity, a.u. BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 0.010 0.010 0.010 1 h 1 h 1 h Skin Heart Lung 3 h 3 h 3 h 0.007 0.007 0.007 0.005 0.005 0.005 0.002 0.002 0.002 0.000 0.000 0.000 650 660 670 680 690 700 650 660 670 680 690 700 650 660 670 680 690 700 W avelengt h, nm W avelengt h, nm Wavelengt h, nm 0.010 0.010 0.010 1 h 1 h Gall bladder GI tract 1 h Liver 3 h 3 h 3 h 0.007 0.007 0.007 0.005 0.005 0.005 0.002 0.002 0.002 0.000 0.000 0.000 650 660 670 680 690 700 650 660 670 680 690 700 650 660 670 680 690 700 W avelength, nm Wavelengt h, mn W avelengt h, nm 0.010 0.010 0.010 1 h 1 h 1 h Stomach Kidney Spleen 3 h 3 h 3 h 0.007 0.007 0.007 0.005 0.005 0.005 0.002 0.002 0.002 0.000 0.000 0.000 650 660 670 680 690 700 650 660 670 680 690 700 650 660 670 680 690 700 Wavelengt h, nm Wavelengt h, nm W avelength, nm Compar Ce6-PVP administra Figure 7 ison of emission tion spectra in the 650 – 700 nm region using 400 nm excitation in various normal organs at 1 and 3 h post Comparison of emission spectra in the 650 – 700 nm region using 400 nm excitation in various normal organs at 1 and 3 h post Ce6-PVP administration. Except for skin and gall bladder, it is evident that the emission spectra of nor- mal organs were lower compared to the emission spectra of tumor. absorption and scattering properties. Hence we have tizer and tissue properties, this method can potentially employed the utility of spectrometric point fluorescence improve the assessment of cancer location and its extent detection as a complementary technique. Spectra meas- within the local-regional area. While fluorescence point urements were carried out at 1 and 3 h post drug admin- spectroscopy studies are promising, it has several draw- istration to correlate the tumor intensity ratios obtained backs as a screening tool as it can only interrogate a small with fluorescence imaging to the tumor fluorescence spec- volume of tissue (typically, 0.5 – 1 mm ) directly beneath tral signal of the tissue. All the macroscopic fluorescence the probe tip. Point measurements inevitably involve a images correlated well to the spectra measurement. The degree of random sampling, which may not allow identi- Ce6-PVP induced spectra emission after normalization fication of early stage disease [22]. Hence, the combina- demonstrated a good separation to differentiate malig- tion techniques of fluorescence imaging and spectroscopy nant tumors from normal tissues. Besides measuring have been proven in good agreement with the actual physical parameters such as concentration of photosensi- Page 11 of 13 (page number not for citation purposes) Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. Fluorescence intensity, a.u. BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 (A) White light at 3 h post Ce6-PVP (B) Fluorescence at 3 h post Ce6-PVP (C) (D) 0.025 0.025 Lesion at 3 h Lesion at 1 h Normal at 3 h Normal at 1 h 0.020 0.020 0.015 0.015 0.010 0.010 0.005 0.005 0.000 0.000 650 660 670 680 690 700 650 660 670 680 690 700 W avelengt h, nm W avelengt h, nm Fl n Figure 8 o uorescence i rmal skin on the scalp at 1 h maging and spectro and 3 h p fluorometri ost-ic measurement o ntravenous administr f fluo ation of 2 rescence em .0 mg is/kg Ce6-PVP sion in angiosarcoma lesion and surrounding Fluorescence imaging and spectrofluorometric measurement of fluorescence emission in angiosarcoma lesion and surrounding normal skin on the scalp at 1 h and 3 h post-intravenous administration of 2.0 mg/kg Ce6- PVP. tumor boundary found by histopathological mapping fluorescence imaging clearly captures the fluorescence in and early stage of disease [23,24]. angiosarcoma and good correlation was found between fluorescence imaging and spectral measurement in the Recently, we have reported photodynamic therapy – acti- patient. This is in agreement with other reports, that fluo- vated immune response against distant untreated rescence ratio imaging in combination with relative spec- tumours in recurrent angiosarcoma [25] and preferential tral measurement of the photosensitizer might be a viable accumulation of Ce6-PVP in angiosarcoma compared to method for the optical diagnosis of cancer [27]. Further- normal skin following intravenous administration in 3 more, in vivo and real-time determination of the time patients [17]. High dose PDT carried out at a high fluence course of photosensitizer's fluorescence could potentially rate resulted in local control of the disease for up to a year; be a crucial pre-irradiation screening tool to determine the however, the disease recurred and PDT had to be repeated exact location and extent of the tumor before photody- [26]. During the repeat PDT session, we measured the flu- namic therapy. orescence of 3 different lesions using fluorescence imag- ing followed by spectroscopy. The results confirmed that Page 12 of 13 (page number not for citation purposes) Fluorescence em ission, a.u. Fluorescence em ission, a.u. BMC Medical Imaging 2009, 9:1 http://www.biomedcentral.com/1471-2342/9/1 13. Chin WW, Heng PW, Olivo M: Chlorin e6 – polyvinylpyrro- Conclusion lidone mediated photosensitization is effective against It is shown that Ce6-PVP has a rapid accumulation in the human non-small cell lung carcinoma compared to small cell tumor, and a relatively short half-life in normal organs. lung carcinoma xenografts. BMC Pharmacol 2007, 7:15. 14. Chin WW, Lau WK, Bhuvaneswari R, Heng PW, Olivo M: Chlorin When excited by blue light, Ce6-PVP accumulating cells e6-polyvinylpyrrolidone as a fluorescent marker for fluores- can be visualized and located in the tissue by virtue of its cence diagnosis of human bladder cancer implanted on the chick chorioallantoic membrane model. Cancer Lett 2007, fluorescence. The main advantage of Ce6-PVP induced 245(1–2):127-133. fluorescence imaging is its increased tumor selectivity 15. Chin WW, Heng PW, Lim PL, Lau KO, Olivo M: Membrane trans- with the ability to clearly define the tumor margin. Com- port enhancement of chlorin e6-polyvinylpyrrolidone and its photodynamic efficacy on the chick chorioallantoic model. bination of Ce6-PVP induced fluorescence imaging and Journal of Biophotonics 2008, 1(5):395-407. spectroscopy could allow detection and discrimination 16. Chin WW, Lau WK, Heng PW, Bhuvaneswari R, Olivo M: Fluores- between cancer and the surrounding normal tissues. cence imaging and phototoxicity effects of new formulation of chlorin e6-polyvinylpyrrolidone. J Photochem Photobiol B 2006, 84(2):103-110. Competing interests 17. Chin WW, Heng PW, Thong PS, Bhuvaneswari R, Hirt W, Kuenzel S, Soo KC, Olivo M: Improved formulation of photosensitizer The authors declare that they have no competing interests. chlorin e6 polyvinylpyrrolidone for fluorescence diagnostic imaging and photodynamic therapy of human cancer. Eur J Authors' contributions Pharm Biopharm 2008. 18. Chin WW, Heng PW, Bhuvaneswari R, Lau WK, Olivo M: The WWC and PST conceived of the study and carried out all potential application of chlorin e6-polyvinylpyrrolidone for- the experimental study. RB and KCS participated in the mulation in photodynamic therapy. Photochem Photobiol Sci 2006, 5(11):1031-1037. clinical study. PWH and MO participated in the coordina- 19. Isakau HA, Trukhacheva TV, Zhebentyaev AI, Petrov PT: HPLC tion of the study. All authors read and approved the final study of chlorin e6 and its molecular complex with polyvi- manuscript nylpyrrolidone. Biomed Chromatogr 2007, 21(3):318-325. 20. Brindle K: New approaches for imaging tumour responses to treatment. Nat Rev Cancer 2008, 8(2):94-107. Acknowledgements 21. Isakau HA, Parkhats MV, Knyukshto VN, Dzhagarov BM, Petrov EP, WW Chin and PS Thong are the recipients of the Singapore Millennium Petrov PT: Toward understanding the high PDT efficacy of chlorin e6-polyvinylpyrrolidone formulations: Photophysical Foundation scholarship. and molecular aspects of photosensitizer-polymer interac- tion in vitro. J Photochem Photobiol B 2008, 92(3):165-174. References 22. Stringer M, Moghissi K: Photodiagnosis and fluorescence imag- 1. DaCosta RS, Wilson BC, Marcon NE: Fluorescence and spectral ing in clinical practice. Photodiagnosis and Photodynamic Therapy imaging. ScientificWorldJournal 2007, 7:2046-2071. 2004, 1(2):9-12. 2. Wong Kee Song LM, Marcon NE: Fluorescence and Raman spec- 23. Wagnieres GA, Star WM, Wilson BC: In vivo fluorescence spec- troscopy. Gastrointest Endosc Clin N Am 2003, 13(2):279-296. troscopy and imaging for oncological applications. Photochem 3. Perelman LT: Optical diagnostic technology based on light Photobiol 1998, 68(5):603-632. scattering spectroscopy for early cancer detection. Expert Rev 24. Mang T, Kost J, Sullivan M, Wilson BC: Autofluorescence and Med Devices 2006, 3(6):787-803. Photofrin-induced fluorescence imaging and spectroscopy in 4. Schultz CP: The potential role of Fourier transform infrared an animal model of oral cancer. Photodiagnosis and Photodynamic spectroscopy and imaging in cancer diagnosis incorporating Therapy 2006, 3(3):168-176. complex mathematical methods. Technol Cancer Res Treat 2002, 25. Thong PS, Ong KW, Goh NS, Kho KW, Manivasager V, Bhuvaneswari 1(2):95-104. R, Olivo M, Soo KC: Photodynamic-therapy-activated immune 5. Ell C: Improving endoscopic resolution and sampling: fluores- response against distant untreated tumours in recurrent cence techniques. Gut 2003, 52(Suppl 4):iv30-33. angiosarcoma. Lancet Oncol 2007, 8(10):950-952. 6. Kherlopian AR, Song T, Duan Q, Neimark MA, Po MJ, Gohagan JK, 26. Thong PS, Olivo M, Kho KW, Bhuvaneswari R, Chin WW, Ong KW, Laine AF: A review of imaging techniques for systems biology. Soo KC: Immune response against angiosarcoma following BMC Syst Biol 2008, 2:74. lower fluence rate clinical photodynamic therapy. J Environ 7. Chen Y, Gryshuk A, Achilefu S, Ohulchansky T, Potter W, Zhong T, Pathol Toxicol Oncol 2008, 27(1):35-42. Morgan J, Chance B, Prasad PN, Henderson BW, et al.: A novel 27. Kopriva I, Persin A, Zorc H, Pasic A, Lipozencic J, Kostovic K, approach to a bifunctional photosensitizer for tumor imag- Loncaric M: Visualization of basal cell carcinoma by fluores- ing and phototherapy. Bioconjug Chem 2005, 16(5):1264-1274. cence diagnosis and independent component analysis. Photo- 8. Thong PS, Olivo M, Kho KW, Zheng W, Mancer K, Harris M, Soo KC: diagnosis and Photodynamic Therapy 2007, 4(3):190-196. Laser confocal endomicroscopy as a novel technique for flu- orescence diagnostic imaging of the oral cavity. J Biomed Opt Pre-publication history 2007, 12(1):014007. 9. Andersson-Engels S, Klinteberg C, Svanberg K, Svanberg S: In vivo The pre-publication history for this paper can be accessed fluorescence imaging for tissue diagnostics. Phys Med Biol 1997, here: 42(5):815-824. 10. Zheng W, Olivo M, Soo KC: The use of digitized endoscopic imaging of 5-ALA-induced PPIX fluorescence to detect and http://www.biomedcentral.com/1471-2342/9/1/prepub diagnose oral premalignant and malignant lesions in vivo. Int J Cancer 2004, 110(2):295-300. 11. Kostenich GA, Zhuravkin IN, Furmanchuk AV, Zhavrid EA: Photo- dynamic therapy with chlorin e6. A morphologic study of tumor damage efficiency in experiment. J Photochem Photobiol B 1991, 11(3–4):307-318. 12. Ramaswamy B, Manivasager V, Chin WW, Soo KC, Olivo M: Photo- dynamic diagnosis of a human nasopharyngeal carcinoma xenograft model using the novel Chlorin e6 photosensitizer Fotolon. Int J Oncol 2005, 26(6):1501-1506. Page 13 of 13 (page number not for citation purposes)

Journal

BMC Medical ImagingSpringer Journals

Published: Jan 8, 2009

There are no references for this article.