| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Molecular Imaging |
1 From the Center for Molecular Imaging Research, Massachusetts General Hospital, Harvard Medical School, 13th St, Bldg 149, Rm 5408, Boston, MA 02129. Received May 28, 2003; revision requested August 7; revision received September 3; accepted October 14. Supported in part by National Institutes of Health grants P50-CA86355 and R24-CA92782. U.M. supported in part by an award from the Broad Medical Research Program of the Eli and Edythe L. Broad Foundation, Los Angeles, Calif. M.A.F. supported by a stipend from the Max Kade Foundation, New York. Address correspondence to U.M. (e-mail: mahmood@helix.mgh.harvard.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: An imaging device that was based on a miniaturized fiberoptic sensor (MIFS) was built in which images created with a 2.7-F fiberoptic catheter were relayed through a dichroic mirror, through a bandpass filter, and on two independent cameras. This system permitted simultaneous recording of white-light and fluorescent images. Spatial resolution, spectral transmissions, and sensitivity were determined in vitro. In vivo testing was performed in nude mice bearing intraperitoneal tumors that express green fluorescent protein and in a mouse model of ovarian carcinoma with enzyme-activatable near-infrared probes sensitive to tumoral protease activity. Signal intensity on images of tumors and that on images of normal tissue were measured and compared with t test.
RESULTS: The catheter, which was advanced through an 18-gauge sheath, showed resolution of 7 line pairs per millimeter and detection limit for fluorochrome Cy5.5 of 1–10 pmol. Detection of endogeneous green fluorescent protein gene expression was feasible in tumor nodules smaller than 1 mm in diameter (mean tumor signal intensity, 153.26 ± 26.45 [SD], compared with that of adjacent nontumoral tissue of 36.73 ± 11.69; P < .008). Similarly, activation of the near-infrared probe by tumoral proteases could be detected in peritoneal tumor seeds of ovarian cancer model with mean tumor signal intensity of 246.33 ± 7.77 compared with that of adjacent nontumoral tissue of 41.56 ± 18.64 (P < .001). Mean contrast-to-noise ratio in the near-infrared channel exceeded white-light contrast-to-noise ratio by a factor of 6.7 (P < .02).
CONCLUSION: With this system, in vivo MIFS imaging of gene expression, enzyme activity, and potentially other molecular events is feasible, through direct interventional access to several organs and body cavities and potentially through transvascular approaches.
© RSNA, 2004
Index terms: Animals • Catheters and catheterization, technology • Enzymes • Experimental study • Genes and genetics
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
These developments in probe design have been paralleled by the construction of highly sensitive reflectance imaging systems with direct illumination and fluorescence-mediatedtomographic systems (6,7). The adaptation of interventional methods can potentially further expand the versatility of optical imaging methods to certain clinical scenarios where high-spatial-resolution local imaging is desirable but organ accessibility for surface-based optical imaging is limited (8).
A wide spectrum of pathophysiologic mechanisms can already be addressed with molecularly specific optical probes. These mechanisms include tumor detection (9,10); monitoring of treatment response (11); and imaging of arteriosclerosis (12), inflammation (13), and thrombosis (14). Most of these investigations have been performed in mouse models and were demonstrated ex vivo after explantation of the diseased organ or in subcutaneous xenotransplants to achieve the necessary sensitivity and the desired submillimeter spatial resolution.
Such end-point procedures preclude one of the strengths of optical imaging: the repeated observation of the same target in time (15). In addition, several activatable ("smart") optical probes are almost ready for application in humans. Thus, existing exterior imaging systems must be adapted for the much larger volumes of the human body, and this adaptation will further complicate the issues of penetration depth and signal attenuation.
The previously mentioned observations indicate the growing need for fluorescence imaging systems optimized for high-resolution high-sensitivity local imaging. Ideally, such systems could be used to monitor the specific molecular effects of therapy in the target tissues, either with evaluation of optical reporter genes or assessment of disease-specific enzyme activity. Thus, the purpose of our study was to construct and evaluate an interventional catheter-based imaging system for the intravital monitoring of molecularly sensitive near-infrared fluorescent probes and optical marker genes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
The illumination source was a mercury vapor lamp (SUV-DC; Lumatec, Munich, Germany) attached to a liquid light guide with a 5-mm core diameter and a numeric aperture of 0.55. At the outer end of the light guide, the light was first collimated with an aspheric lens, then passed through the 670-nm excitation shortpass filter (Omega Optical), and finally refocused into the excitation fiber bundle of the imaging catheter. Lenses and mountings (Edmund Optics) were used as needed.
Optical Testing
Spectral transmission was determined (M.A.F.) for both the excitation and the emission paths (Fig 2, A) with a spectrophotometer (HP3000; Hitachi Instruments, Tokyo, Japan). The filter breakthrough (ie, the fraction of reflected excitation light that can pass the emission filter) was measured with the same instrument with both filters in place. The cutoff wavelengths of the filters were selected to provide a minimum attenuation of 106 (optical density, 6) for the entire sensitivity range of the charge-coupled device camera.
|
Sensitivity Tests
Sensitivity testing was performed (M.A.F.) to determine the minimal detectable concentration of free fluorophore after signal attenuation through the whole optical train and to determine the minimal volume (at a given fluorophore concentration) necessary to produce a sufficient fluorescent signal. Sensitivity was determined in dilution series of Cy5.5 and Cy7 monofunctional dyes (Cy5.5 and Cy7 N-hydroxysuccinimide ester; Amersham Biosciences, Uppsala, Sweden), with concentrations ranging from 30 nmol to 1 mmol (Fig 2, C).
The tip of the catheter was directly immersed, and the signal intensity of the resulting homogeneous diffused image was averaged for the field of view. Image acquisition times were 0.1 to 0.3 second. In addition, to correlate the fluorescent signal intensity with the illuminated volume of fluorophore at a given concentration, a 1 µmol/L solution of Cy5.5 in a black low-reflectance 300-µL well was imaged with a gradual retraction of the catheter tip from the bottom of the well in 0.25-mm increments. With this procedure, the thickness of the fluorochrome solution in front of the catheter was increased and contributed to the fluorescent signal. At each increment, the signalintensity of the homogeneously illuminated field of view was recorded, and the volume of the illuminated cone of dye that generated the signal was calculated.
The sensitivity was confirmed in vivo with injection of Cy5.5–N-hydroxysuccinimide ester covalently coupled to the amine groups of an aminated dextran-coated iron oxide nanoparticle (10 µg iron, corresponding to an amount of Cy5.5 of 0.16 nmol) and observation of microvascular fluorescence before and after injection. The dye-nanoparticle conjugate represents an intravascular contrast agent with minimal leakage into the extracellular space and a plasma half-life of more than 10 hours.
Animal Models
The study was approved by the institutional animal care committee. To avoid any intestinal autofluorescence in the near-infrared range, the animals were fed a chlorophyll-free diet for at least 7 days preceding near-infrared imaging. All animal procedures were performed by two of the authors (M.A.F., U.M.).
Group 1 consisted of four athymic mice (nu/nu; Taconic, Germantown, NY) that were injected intraperitoneally with 2 x 106 of 9L tumor cells stably transfected with green fluorescent protein (16). Group 2 consisted of four nude mice that were injected intraperitoneally with 3 x 106 of OVCAR-3 ovarian cancer cells (17) stably transfected with luciferase. After 2–4 weeks, multiple peritoneal tumor nodules of 1–2 mm in diameter developed. For the two experimental groups, two groups of four mice each not injected with tumor cells were investigated as the respective control groups (control groups 1 and 2). The sample size was taken into account in the statistical evaluation, as discussed later in this article. Twenty-four hours before imaging, in all animals, 2 nmol of the imaging probe was injected through the tail veins.
After anesthesia was induced with 90 mg per kilogram of body weight ketamine hydrochloride and 9 mg/kg xylazine hydrochloride administered intraperitoneally, catheter laparoscopy was undertaken (M.A.F., U.M.). A sheath (0.9-mm outside diameter) was introduced into the lower right quadrant of the abdomen, and 1–3 mL of air was gently introduced into the peritoneal cavity while the intraabdominal pressure was monitored digitally. The catheter was introduced through the sheath, and the peritoneum was inspected. By turning the animal sideways, the bowel loops could be moved to gain access to the deeper parts of the mesentery, the prerenal tissue, and the retroperitoneal vessels that were visible through the parietal peritoneum.
Mice were sacrificed with injection of 100 mg/kg pentobarbital sodium immediately after catheter-based imaging (control mice, n = 4; mice with tumors expressing green fluorescent protein, n = 4; and mice with peritoneal ovarian tumors 24 hours after near-infrared fluorescent probe injection, n = 4). Reflectance imaging was performed in the opened abdomen as previously described (9) for comparison with catheter-based imaging. In addition, fluorescent tumors were excised and reinvestigated ex vivo with the MIFS imaging system, as well as with a reflectance imaging system (7) for comparison. The correlation between tumor infiltration and fluorescent signal was confirmed histologically after we froze the excised tissue in liquid nitrogen, sectioned it, and examined alternating sections that were either native for near-infrared fluorescence (representing cathepsin B–activated probe) or stained with hematoxylin-eosin (M.A.F.).
For comparison of detection accuracy between white light and fluorescence, each abdomen was divided into four quadrants. Presence of tumorous tissue in each quadrant was verified with a dissecting microscope and with histologic analysis in ambiguous cases (U.M., M.A.F.).
Probe Design
Cathepsins are lysosomal proteases (18) overexpressed in a variety of malignant tumors (19). The design of a quenched cathepsin B–activatable near-infrared fluorescent probe was described previously (20). Briefly, to a poly-L-lysine backbone, multiple methoxy-polyethylene glycol chains (21) and multiple cathepsin B–specific substrate peptides are attached. After this conjugation, monoreactive Cy5.5 is attached to the N terminus of each substrate peptide. On average, each probe molecule contained 92 methoxy-polyethylene glycol chains, 11 molecules of Cy5.5, and 44 unmodified lysine residues on the backbone. In the native state, fluorescent emission is effectively inhibited by fluorescence resonance energy transfer caused by the proximity of the fluorochromes. After enzymatic cleavage of the peptide substrates, the fluorophores are liberated, and this activity results in a 15–30-fold increase in fluorescent emission in this probe design (20).
Image and Statistical Analysis
Quantitative measurements were performed on the acquired in vivo near-infrared images after injection of the cathepsin B–sensitive probe or in the green fluorescent protein channel for evaluation of green fluorescent protein–expressing tumors. The mean signal intensity and SDs of tumor nodules and of adjacent nontumoral tissue were measured on 12-bit gray-scale images by using proprietary software. The dependent variable was the mean signal intensity in regions of interest of constant size comprising 96 pixels. The tumorous regions of interest were placed with the center over the respective brightest part of the tumors; the nontumorous regions of interest were placed over nontumoral tissue at a distance of two region-of-interest diameters from the tumor center. All regions of interest were placed by one author (M.A.F.).
Mean values and SDs of the region-of-interest measurements were calculated for tumorous and nontumorous regions of interest. Signal intensity of tumorous tissue was compared with that of nontumorous tissue by using the Student t test for independent samples (SPSS 10.0; SPSS, Chicago, Ill), in a similar approach as was used previously (7) at a level of significance of .05. With the t test, the sample size is not taken into account; thus, a valid rejection of the null hypothesis can be performed even with small sample sizes if the P value is less than the level of significance (22).
The tumor-to-background ratio was calculated as follows: SIt/SIb, where SIt represents signal intensity of tumor and SIb represents signal intensity of background.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The depth of the layer that was imaged was correlated to the fluorescent signal generated in a logarithmic function. On the basis of the thickness of the half-value layer that contributed to the signal intensity and the corresponding volume of the illuminated cone of tissue, illuminated volumes of 0.1 and 0.8 µL were sufficient to produce 67% and 85%, respectively, of the peak signal intensity. This resulted in an approximate in vitro detection limit of 1–10 pmol of Cy5.5.
In vivo MIFS imaging was technically feasible in all animals. There were no deaths related to the procedure. Real-time white-light imaging allowed clear differentiation of the peritoneal organs, liver, gallbladder, stomach, and intestinal loops. Mesenteric blood vessels could be identified on the visceral mesentery regularly and be followed to ramifications of 0.2 mm in diameter both in the white light and after injection of the intravascular contrast agent in the fluorescent channel (Fig 3, A–D).
|
Exogenous Activatable Near-Infrared Fluorescent Probe
The foci of peritoneal carcinomatosis in group 2 (Fig 3, I–L) were brightly fluorescent after injection of the cathepsin-activatable fluorescent probe. After injection of the near-infrared probe, mean tumor near-infrared signal intensity was 246.33 ± 7.77, compared with mean signal intensity from adjacent nontumoral tissue of 41.56 ± 18.64 (P < .001). Mean contrast-to-noise ratio in the near-infrared channel exceeded white-light contrast-to-noise ratio by a factor of 6.7 (P < .02). The tumor-to-background ratio was 5.8, a number that is comparable with the tumor-to-background ratio of endogenously expressed genes, as discussed previously. In all cases, the locations of the individual lesions observed on MIFS images were correlated with the locations on epifluorescent images obtained after laparotomy.
At sectioning and microscopic investigation, each tumor nodule showed near-infrared fluorescence of the native sections, and these results correlated with the staining of tumor margins on the stained sections. Of 16 abdominal quadrants in this group, 11 contained tumorous tissue. With white-light imaging, we detected two of 11 quadrants, with two false-positive findings that proved to be distended bowel loops without histologically detectable tumors. In the fluorescent channel, all 11 quadrants with positive findings were correctly identified, without false-positive findings.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Optical imaging can fill a gap in present clinical imaging modalities by providing a combination of high resolution, high sensitivity, and molecular specificity (9). The success of optical imaging in a clinical environment, however, will largely depend on the development of suitable detection methods, including catheter-based imaging devices, for an interventional approach optimized for each organ that is investigated.
Presently, in vivo near-infrared fluorescent imaging can be performed with epifluorescent application, with the light source and the detector on the same side of the target. The excitation light is directed on the surface of the target, and the emitted fluorescence is recorded with a camera. Advantages of such systems are the high spatial and temporal resolution and ease and speed of use.
Because of the attenuation of near-infrared light in tissue, the detectability of fluorochrome activity is limited to depths of 0.5 cm; thus, in vivo applications have been limited to subcutaneous lesions. A different approach is fluorescence-mediated tomography (24), which has the ability to produce tomographic images of near-infrared probe distribution in whole bodies and to provide quantitative information about fluorochrome concentrations (6,25). Challenges remain, however, in the extension of the techniques for clinically relevant volumes with respect to spatial resolution and in the modeling of the changing air-tissue interfaces in body regions such as the abdomen.
We demonstrated the feasibility of imaging molecular events, such as specific protease activity, and the expression of an optical reporter gene with a novel near-infrared imaging system that allows minimally invasive access to internal organs, thereby overcoming the limited depth sensitivities of the present imaging systems that are based on epifluorescence and reflectance. Images were acquired at sensitivity levels comparable with those of epifluorescent systems that have been successfully used in optical imaging of smart probes in various organs. We were able to use this device to detect specific tumoral activation of such enzyme-sensitive probes in deep peritoneal organs in a catheter-based approach.
A feature not previously available in fluorescent imaging systems is the simultaneous real-time availability of a white light and a fluorescent channel. This feature was made possible with a high-bandwidth excitation filter, which blocks out only the near-infrared emission spectrum of the fluorochrome, with minimal influence on the color of the reflected visible excitation light. The simultaneous display allows orientation and additional spectral characterization during the imaging sessions.
The inherent light losses of the fiberoptic imaging train could be compensated for by several independent modifications to previous approaches, which add up to a net increase in sensitivity of a factor of 1,000 over previously described systems, which is reflected in an effective reduction of exposure time to 0.1–0.3 second. The main improvements were a higher-intensity light source and a more efficient coupling of the light into the excitation fibers of the sensor. The proximity of the illumination fibers to the target and the high numeric aperture of the sensor, which was 0.36, further contributed to increased sensitivity.
Employment of new interference filter–coating technologies with six-cavity design and improved flank steepness allowed the selection of 50% cutoff wavelengths for the excitation and emission filter separated by only 30 nm instead of 60 nm that was used in previous designs. These features allowed matching of a broader part of the excitation and emission spectra of the dye. Finally, a spectrally matched dichroic mirror and a cooled high-quantum-efficiency charge-coupled device camera with extended infrared sensitivity were employed.
With the present design, both the exposure time and the concentration sensitivity of the MIFS allow real-time video fluorescent imaging with sufficient contrast and signal-to-noise ratio for in situ disease detection and characterization. A limitation of the current study is the overlap between the emission spectra of Cy5.5 and chlorophyll, which resulted in one false-positive interpretation in the study (as discussed in Results). This may be overcome in future studies with conversion of the system and the probes to longer emission wavelengths in the near-infrared region, such as by using fluorochrome AF750 (Molecular Probes, Eugene, Ore) and thereby eliminating contamination of the molecularly specific information with chlorophyll fluorescence.
Future generations of this device will include options for multichannel near-infrared imaging, thereby allowing the evaluation of multiple probes or multichannel probes with different fluorochromes attached (7). With these constructs, several molecular parameters can be monitored simultaneously and potentially can be combined in real time (eg, perfusion, vascular permeability, and enzyme activity) on live video images. Real-time image processing (eg, imaging by using real-time ratio of fluorescence images to white-light images) can potentially facilitate detection and differentiation of such molecular markers for specific diseases.
In conclusion, we designed, constructed, and tested a catheter-based optical fluorescent imaging system that is capable of detecting expression of optical reporter genes and activatable fluorescent probes in the visible and near-infrared range at concentrations comparable with those found in previous ex vivo surface imaging experiments. The proposed method is minimally invasive, is nondestructive, and can be carried out repeatedly in the same animal. The system is directly transferable to human coronary and other vascular imaging, as well as laparoscopic and endoscopic applications. The penetration depth of the excitation and emission light is sufficient to image the vessel wall directly through the bloodstream and to image intraluminal processes of the stomach or the intestine through the organ walls with an intraperitoneal approach.
Inherent advantages of this method include precise anatomic orientation because of the real-time white-light video image; repeatable access to intraperitoneal and retroperitoneal organs in rodents, including large and small blood vessels, kidneys, lymph nodes, and pancreas; and the potential for simultaneous acquisition of emissions of multiple fluorophores and for direct biopsy verification of the fluorescent findings through the working channel.
The MIFS catheter design is directly translatable to human application. Minimally invasive visualization of optical probes could affect therapeutic decisions for a number of the most prevalent diseases today. Intravascular application of the MIFS can provide insights about the level of macrophage activity and thus the risk profile for atherosclerotic plaque rupture (12).
The catheter-based assessment of protease activity may be one of the earliest detectable signs in lymph nodal or distant metastatic processes. Since proteases currently represent an attractive target for antitumor therapies, information about activity of specific proteases is indispensable for the early evaluation of treatment responses, making MIFS imaging potentially an objective end point for the tailoring of therapies in a given individual (26). In the colon, MIFS imaging can potentially facilitate detection of precancerous lesions by providing quantitative information about the activity of molecular markers that have been linked to the degree of dysplasia in polyps (10).
|
FOOTNOTES |
|---|
Abbreviation: MIFS = miniaturized fiberoptic sensor
Author contributions: Guarantor of integrity of entire study, M.A.F.; study concepts and design, M.A.F., R.W., U.M.; literature research, M.A.F., U.M.; experimental studies, M.A.F., U.M.; data acquisition and analysis/interpretation, M.A.F., R.W., U.M.; statistical analysis, U.M., M.A.F.; manuscript preparation, definition of intellectual content, editing, revision/review, and final version approval, M.A.F., R.W., U.M.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  R. A. Sheth, J. M. Tam, M. A. Maricevich, L. Josephson, and U. Mahmood Quantitative Endovascular Fluorescence-based Molecular Imaging through Blood of Arterial Wall Inflammation Radiology, June 1, 2009; 251(3): 813 - 821. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Upadhyay, R. A. Sheth, R. Weissleder, and U. Mahmood Quantitative Real-time Catheter-based Fluorescence Molecular Imaging in Mice Radiology, November 1, 2007; 245(2): 523 - 531. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. T. Heverhagen Smart Near-Infrared Imaging Probes--A Quantum Leap for Early Detection of Colorectal Cancer? Radiology, July 1, 2007; 244(1): 1 - 2. [Full Text] [PDF]
|
|
|
|
|
|
H. Alencar, M. A. Funovics, J. Figueiredo, H. Sawaya, R. Weissleder, and U. Mahmood Colonic Adenocarcinomas: Near-Infrared Microcatheter Imaging of Smart Probes for Early Detection--Study in Mice Radiology, July 1, 2007; 244(1): 232 - 238. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. R. Kim, A. Yudina, J. Figueiredo, W. Reichardt, D. Hu-Lowe, A. Petrovsky, H. W. Kang, D. Torres, U. Mahmood, R. Weissleder, et al. Detection of Early Antiangiogenic Effects in Human Colon Adenocarcinoma Xenografts: In vivo Changes of Tumor Blood Volume in Response to Experimental VEGFR Tyrosine Kinase Inhibitor Cancer Res., October 15, 2005; 65(20): 9253 - 9260. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. O. Bonow Molecular Beacons Illuminate Subcellular Events Circulation, April 12, 2005; 111(14): 1730 - 1732. [Full Text] [PDF]
|
|
|
|
 |
|
 F. A. Jaffer and R. Weissleder Molecular Imaging in the Clinical Arena JAMA, February 16, 2005; 293(7): 855 - 862. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Jiang, E. S. Olson, Q. T. Nguyen, M. Roy, P. A. Jennings, and R. Y. Tsien Tumor imaging by means of proteolytic activation of cell-penetrating peptides PNAS, December 21, 2004; 101(51): 17867 - 17872. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Hoffman Can Optical Molecular Imaging Techniques with Catheter-based Approaches Be Used for Disease Detection? Radiology, June 1, 2004; 231(3): 609 - 610. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
