| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
1 Center for Biomedical Electron Paramagnetic Resonance Spectroscopy and Imaging, Davis Heart and Lung Research Institute and 2 Division of Cardiovascular Medicine, Department of Internal Medicine, The Ohio State University College of Medicine, Columbus, Ohio; and Departments of 3 Gastrointestinal Oncology and 4 Clinical Laboratory, Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka, Japan
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
There is evidence that oxygen radical generation and altered tissue redox state are key factors in the process of tumor promotion and transformation (7 , 8) . In recent years, it has also become increasingly clear that oxygen radicals have a key role as second messengers in regulating the process of cellular growth and cell death (9) . Recently, it has been reported that a number of cancer-linked genes are associated with alterations in cellular redox state and free radical generation (10) . It is also clear that a number of modalities of cancer treatment are based on induction of high levels of oxygen radical formation that lead to tumor necrosis or apoptosis. Classical radiation therapy induces oxygen-dependent formation of hydroxyl radical, leading to cell death. A number of chemotherapeutic agents, including Adriamycin and bleomycin, trigger oxygen radical formation that is involved in their mechanisms of tumor killing (11 , 12) . Therefore, there has been a great need for techniques to accurately assess cellular redox state and altered radical generation in the process of oncogenesis, tumor biology, and tumor treatment.
In vivo electron paramagnetic resonance imaging (EPRI) methods have been developed to enable noninvasive determination of the spatial distribution of free radicals and of alterations in free radical metabolism associated with disease (13, 14, 15) . EPRI is a powerful technique able to measure free radical metabolism and redox state and can provide high sensitivity, specificity and spatial resolution. It has been extensively used to noninvasively measure the distribution and kinetics of transient/stable free radicals in biological tissues (15, 16, 17, 18) . The electron paramagnetic resonance (EPR) spectrum is a fingerprint of free radicals with their unique hyperfine structures. This technique is extremely sensitive, and it can detect free radicals in concentrations as low as 10 nmol/L with an in vivo imaging sensitivity that extends down to µmol/L levels (19 , 20) .
Furthermore, in recent years, this technique, along with appropriate nitroxide spin probes, has been extensively applied to provide information about the redox state within organs of living animals and alterations associated with disease (17 , 21) . These techniques have been applied to measure and map the distribution and metabolism of free radicals in a number of organs and tissues including the heart, brain, gastrointestinal tract, and, most recently, in tumors (16 , 21, 22, 23) . These paramagnetic nitroxide probes are bioreduced to the corresponding EPR silent hydroxylamines by mitochondrial reduction and intercellular antioxidants, and this reduction is markedly enhanced under hypoxic conditions (24, 25, 26) . Hypoxic tissues can be readily distinguished from normoxic tissues based on the rate of nitroxide reduction. Many solid tumors contain hypoxic regions (27, 28, 29) , and it has been shown previously (21) that the low pO2 in implanted tumors compared with that in normal tissue results in a more rapid reduction of the nitroxides, and this process has been measured previously and mapped in in vivo fibrosarcomas in mice by EPR spectroscopy and imaging.
Therefore, the present study was performed in an effort to determine whether EPR spectroscopy and imaging techniques can be applied to provide identification of gastric tumors with differentiation from surrounding tissue as well as information regarding alterations in cellular redox state associated with these neoplastic cells. Experiments were performed in a MNNG-induced model of gastric carcinoma with EPRI and localized EPR spectroscopy to measure and map the distribution and metabolism of a nitroxide radical redox probe. We observe that gastric tumors exhibit a distinctly different radical distribution and redox state than normal stomach tissue, and this enables their detection and spatial localization by EPRI techniques.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Induction of Gastric Carcinoma in Rats.
Six-week-old Wistar rats were purchased from Japan SLC (Hamamatsu, Shizuoka, Japan) and housed in suspended wire-bottomed metal cages in animal quarters maintained at a temperature of 21°C to 22°C and a humidity of 30% to 50% with a 12-hour light/dark cycle. Starting at 3 weeks of age, rats were given drinking water containing MNNG at a concentration of 50 µg/mL for 25 weeks. The MNNG solution was supplied from bottles covered with aluminum foil to prevent photolysis and refilled every other day as reported previously (30)
. After 25 weeks, MNNG administration was stopped, and normal tap water was supplied. Rats had free access to regular chow pellets (Nihon-Norsan, Yokohama, Japan). At 65 to 69 weeks, animals were used for EPRI and EPR spectroscopy experiments. Normal rats given normal tap water instead of MNNG solution served as tumor-free controls. Eight rats with gastric cancer tumors of 2 to 8 mm in diameter and six normal rats were used. The tumors could be readily identified by external palpation of the firm nodules in the stomach, and this was subsequently visually confirmed when the stomach was incised, with further verification by histology.
Electron Paramagnetic Resonance Imaging Measurements.
Three-dimensional EPRI measurements were performed in a manner similar to that described previously (31
, 32)
, with only slight modifications using EPRI instrumentation built at The Ohio State University, Center for Biomedical EPRI, consisting of a 750 MHz EPR spectrometer, three sets of water-cooled gradient coils attached to the magnet pole pieces, and a specially designed bridged loop-gap surface resonator customized for gastric imaging. This resonator has two-bridged gaps, a loop diameter of 25 mm with a length of 28 mm, and is surrounded by a cylindrical shield of 50 mm in diameter (Fig. 1A)
. A detailed description of this resonator design is to be reported separately.5
The EPRI parameters were as follows: frequency, 750 MHz; microwave power, 50 mW; modulation amplitude, 1 G; scan width, 25.6 G, scan time, 5 seconds; time constant, 40 milliseconds; gradient field, 8 G/cm. The sensitive imaging region provided by the resonator was mapped with three-dimensional EPRI measurements of a cylindrical phantom filled with the sharp single-line triarylmethyl radical. From this image, a map of H1 field magnitude was obtained and used to correct for any H1 inhomogeneity in the image space. EPRI of the phantom showed that the radical distribution could be measured within a diameter of 29 mm in the plane of the resonator surface and to a depth of 20 mm, adequate to image the uplifted stomach. Image data are displayed as either three-dimensional surface-rendered images or as a series of two-dimensional slices continuously cut with 0.25 mm thickness along the ZX plane.
|
. The rat with the holder was then placed on the positioning stage of EPRI instrument (Fig. 1A)
, and then the resonator was lowered until the stomach was optimally placed within or adjacent to the resonator surface and fixed in place. EPRI measurements of the stomach were performed immediately after administration of a 2-minute intravenous load of 1.5 mL of 300 mmol/L 3-CP. A continuous intravenous infusion of 300 mmol/L 3-CP was then continued at a rate of 0.035 mL/min. This maintenance dose provided a near constant EPR signal within the stomach for the duration of the imaging experiments. To provide verification that the immobilization process/positioner does not alter perfusion to the normal and tumor regions of the stomach, laser Doppler flow measurements were performed using a floLAB laser Doppler blood flow monitor (Moor Instruments Limited, Devon, United Kingdom) with a MP2 probe of 30 mm in length and 6 mm in diameter. No alterations in flow were seen with the immobilization process for exteriorization and EPRI.
Localized Electron Paramagnetic Resonance Spectroscopy Measurements.
Localized EPR spectroscopy measurements of the distribution of 3-CP in tumors and normal gastric tissue were performed using a surface coil resonator with a diameter of 10 mm and a sensitive depth of 7 mm, and an L-band microwave bridge. EPR variables were as follows: microwave power, 32 mW; modulation amplitude, 1 G; scan width, 60 G; scan time, 15 seconds; and time constant, 80 milliseconds.
In vivo Indocyanine Green–Near-Infrared Reflectance Imaging of Perfusion.
Gurfinkel et al. (33)
showed that ICG acts as a blood pool or blood persistence agent without selective uptake or efflux mechanism by in vivo NIR images of the phamacokinetic delivery of ICG. Because the rates of capillary permeability and blood pool or blood persistence are determined by blood flow, using NIR imaging with ICG, the differences in blood flow were examined by comparison of the rates of ICG distribution in tumor and normal tissue. Next, with ICG-NIR images, we examined whether blood was delivered in equivalent amounts to normal mucosa versus tumor regions. Furthermore, we examined whether the results might be affected by cutting the stomach. That is, the stomach was exteriorized and opened along the greater curvature, according to the methods described in the EPRI experiment and where there are few blood vessels. The opened stomach was then maintained, mucosa up, in a lifted position by a plastic holder. After intravenous administration of 5.0 mg/kg ICG to rats, a 25-mW tungsten halogen light fitted with an 805 nm band pass interference filter illuminated the surface of the mucosa tissue. With the video camera, recording of the ICG-NIR of the mucosa surface generated from within the tissue volume as great as 2 mm was started on a nonintensified videotape 2 minutes after injection of ICG. The frame, on which the first maximal ICG-NIR was recorded, was studied for the two-dimensional image of the ICG uptake rate. Numerical analysis of the pseudo-color images was performed according to the manufacturer’s specifications (Olympus Optical Co., Ltd., Tokyo, Japan).
Histology.
After the experiments, each stomach was sectioned for routine histologic examination. The stomach was opened along the greater curvature, pinned flat on a mat, fixed with 10% formalin (pH 7.4), and stained with hematoxylin and eosin. Adenocarcinomas were defined as neoplastic tumors with involvement of the submucosa or deeper layers such as the muscle layer and serosa (34)
.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The EPR image appeared similar to that of the visual image of the stomach. In contrast, three-dimensional EPRI of the treated animals with palpable gastric tumors exhibited prominent holes at locations corresponding to the gastric cancer tumors (Fig. 3)
. As illustrated in Fig. 3C
, the stomach was incised just above the greater curvature after EPRI measurement. The mucosal surface was exposed, revealing a tumor of about 8 mm (Fig. 3D
, dotted circle) present in the antrum. As shown in the diagram, the tumor was present along the greater curvature in the antrum and curved along the XY plane inward toward the abdomen. The three-dimensional EPR image of the tumor-bearing stomach in Fig. 3B
exhibited a prominent hole measuring about 6 mm in diameter at the location in the antrum on the greater curvature of the stomach. At this location, a hole devoid of nitroxide radical signal was seen, which was surrounded by an annular ring of low signal intensity, whereas the image of the surrounding stomach was otherwise homogeneous. The location of this hole in the EPR image corresponded to the location of the palpated tumor labeled in Fig. 3C
that was visualized after subsequent incision and exposure of the luminal surface. To better visualize the tumor, a series of two-dimensional slices are shown from the three-dimensional image data. These images were corrected to compensate for resonator H1 inhomogeneity as described previously (35)
. These image cuts along the Y axis clearly depict the relative absence of signal intensity at the location of the tumor (Fig. 4)
. Thus, EPRI was able to distinguish and map the location of gastric cancer tumor based on the relative absence of nitroxide signal within the tumor.
|
|
|
In Fig. 5
, measurements of the observed signal intensity are shown from the stomachs of three animals. The first contained a large tumor with a diameter of 8 mm that was almost large enough to fill the sensitive volume of the surface coil. The second contained three small adjacent tumors, each of only 2 mm in diameter, that only filled a small portion of the sensitive volume, and the third was of a similar antral region of a control non-tumor-containing stomach. The first point in this graph shows the signal intensity seen immediately after the 3-CP bolus infusion. In the stomach with the large tumor, a much lower signal magnitude was seen, whereas in the presence of smaller tumors, only a modest decrease was seen. In general, it was observed that the decrease in signal paralleled the portion of the sensitive volume of the resonator filled with tumor rather than normal tissue.
|
. Because the ratio of tumor volume to normal tissue within the resonator paralleled the observed decrease in signal and the decay kinetics were identical to that in normal tissue, this suggested that the observed signals in the tumor-containing stomachs were derived almost entirely from the normal gastric tissue surrounding the tumor. Thus, it appears that the observed signal arises almost entirely from normal stomach tissue and is absent from the tumor. This confirms the EPRI observations of a hole with an apparent lack of nitroxide signal at the tumor sites within the stomach and suggests that either the tumor tissue exists in a highly reduced state or there is less uptake of 3-CP within the tumor.
In an effort to determine whether the relative absence of signal in the tumor was due to the presence of a highly reduced state and not simply to lack of nitroxide uptake in the tumor, studies were performed in an incised stomach preparation with topical application of the 3-CP nitroxide solution. The stomach was rapidly removed from the rat, opened along the greater curvature, and fixed on a sample holder plate with the mucosal surface oriented upward. Similar to the results seen in the intact closed stomach, after topical application of the nitroxide, a void in the image was seen at the location of the tumor, indicating a much more rapid rate of reduction than in the surrounding normal tissue (Fig. 6)
. With topical application of the oxidizing agent ferricyanide (20 mmol/L), which converts the reduced EPR-silent hydroxylamine back to EPR-detectable nitroxide, the lower signal level in the tumor was reversed, as shown in Fig. 6C
.
|
Histology of Gastric Cancer.
Histologic studies indicated that all of the tumors occurred in the antral mucosa and that gastric tumors were all histologically well differentiated adenocarcinomas. The sample corresponding to the tumor-containing stomach whose EPRI data are presented in Figs. 3
and 4
is shown in Fig. 6
. Within the tumor, central islands of necrosis and surrounding fibrosis were seen (Fig. 6E
, arrow).
In vivo Indocyanine Green—Near-Infrared Images of Perfusion in Normal Stomach and Tumor.
In an effort to determine whether the relative absence of signal in the tumor was due to the presence of a highly reduced state and not simply a lack of perfusion and compound uptake in the tumor, studies were performed using in vivo ICG-NIR imaging of tumor-bearing stomach. We examined whether the immobilization process that was required for exteriorization of the stomach altered perfusion to the normal tissue and tumor; furthermore, we examined whether the agent was delivered in equivalent amounts to normal mucosa and tumor. The stomach was pulled out from the abdomen, opened, and put on the holder under anesthesia as shown in Fig. 7A
, and video recording of the ICG-NIR signal within the tissue was started 2 minutes after intravenous administration of ICG. Fig. 7B
shows the image of ICG uptake rates at the first maximum uptake. The rate of ICG uptake in the tumor (the white solid circle, except for halation spots) was determined as 85% of the rate in the normal tissue (the white dotted circle). The pseudo-color image of the rate of ICG uptake in Fig. 7C
also exhibited no significant difference in the color between the tumor (white solid circle) and the surrounding normal tissue. From these results, we confirmed that there was no significant difference in ICG uptake rate between tumor and normal tissue after the stomach was immobilized on the holder, and we also confirmed that ICG was delivered in equivalent amounts to normal and tumor regions after intravenous administration. Furthermore, it was confirmed that there was no significant difference in perfusion between tumor and normal tissue in the injured stomach. The histology of the ICG-treated stomach of Fig. 7
is shown in Fig. 7D
and proves that adenocarcinoma was present and highly differentiated, containing severe necrosis similar to the examples used for the EPRI experiments. Therefore, we concluded that there are no significant differences in either the perfusion or compound distribution rates between normal and tumor regions of the stomach.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
It is believed that the administered nitroxide radicals are readily reduced through the complex system of tissue-reducing enzymes and couples to hydroxylamines that are EPR silent (36) . As for the reaction between pro-oxidants such as superoxide and the applied nitroxide radicals, whereas electron transfer to the nitroxide could also occur, the lifetime of these oxidants is very short, and the rate of this reaction is relatively slow, so this pathway is not considered a major contributor to the overall process of reduction in the tissue (37) . However, the more rapid reduction of the probe could also reflect a higher flux of electrons through both reducing and pro-oxidant producing pathways.
The EPR image data indicated that the 3-CP nitroxide label was absent from the gastric tumors, although it was otherwise homogeneously distributed over the gastric wall after intravenous administration to the rat. Consistent with this result, pharmacokinetic studies using localized EPR spectroscopy suggested that the lifetime of 3-CP in the tumors was too short to be detected compared with that in the normal stomach tissue, so that only 3-CP remaining in the normal stomach tissue was detected after the 2-minute infusion of 3-CP. We have observed previously that the half-life of 3-CP in normal leg muscle is 14.6 ± 0.5 minutes (21) . In this study, a somewhat shorter value of 8.7 ± 0.9 minutes was observed from the normal stomach. In view of reports that the concentrations of antioxidants are high in the stomach, with levels above those of other tissues (38) , it is expected that the half-life of 3-CP in the normal stomach tissue would be shorter than that in normal skeletal muscle.
Assuming that the lifetime of the nitroxide is determined in vivo by the metabolism and distribution of nitroxide (17 , 39) , the short lifetime of 3-CP in the cancer tumors could be explained by a combination of the following three factors: (a) lower delivery of 3-CP by the circulation, (b) increased metabolic conversion of the nitroxide to the EPR-inactive form by the endogenous reducing agents, and (c) enhancement of bioreduction in the tumor under hypoxic conditions. It has been reported that the levels of endogeneous reducing agents such as thiols and ascorbic acid, which can rapidly reduce the nitroxide, are increased in gastric cancer (40 , 41) . It has also been reported that the blood flow of tumors such as carcinosarcoma can be lower than that in the normal stomach (42) . When the infusion rate is low in the endogenous reducing agent-rich tissue, the amount of delivered nitroxide and its contact time with the endogenous reducing agent would be increased, resulting in an increase in the amount of the reduced nitroxide. In addition, many tumors contain a high hypoxic fraction, in which bioreduction is enhanced (26 , 27 , 39) . In prior EPRI studies of implanted fibrosarcoma, it was reported that 3-fold lower pO2 values are present within the tumor and are accompanied by much more rapid nitroxide reduction in the tumor compared with that in normal tissue (21) . The reduced state of tumors could also be due to intermittent hypoxia (43 , 44) . In our model, NIR imaging of tissue perfusion demonstrated that there was no significant difference in perfusion of the tumor and the normal tissue, which suggests that the rapid 3-CP clearance observed was due primarily to metabolic reduction.
Alterations in oxygen radical generation and tissue redox state are key factors in the process of carcinogenesis (45) . There is evidence that enhanced oxidant stress is a key factor in tumor promotion and oncogenesis (46) . With regard to gastric carcinoma, it has been shown in the model we used that MNNG treatment induces increased oxygen radical generation, and this is thought to be an important process in triggering cellular transformation leading to tumor formation (7) . It is also clear from the literature that marked increases in oxygen radical generation as triggered by radiation treatment or certain types of chemotherapy including anthracyclines are effective in inducing tumor cell death (11 , 12) .
There have been a number of studies reporting that after oxidant stress, there is an increased reduction of nitroxides in biological tissues (25 , 47) . This phenomenon has been reported in a number of tissues including the lungs, brain, and heart. It is possible that after oxidant stress, there is a cellular compensatory response resulting in induction of antioxidative enzymes and increased levels of antioxidants. It is possible in tumors that such an antioxidant response with a higher reduced state could render resistance to radiation or other therapies. In addition to enhanced levels of antioxidants or antioxidative enzymes, it is possible that pro-oxidant enzyme pathways could be suppressed. Indeed, it has been reported previously that a number of human tumors are associated with an absence of p53, and we have observed previously that p53-induced genes are involved in oxidant formation leading to cellular apoptosis (10) . Thus, one can hypothesize that after a prolonged sublethal oxidant exposure, there could be altered gene expression with transformation of the cellular phenotype to a highly reduced antioxidative state. This, in turn, would tend to protect the cell from regulatory pathways that would otherwise lead to apoptotic cell death. Thus, the ability to assess the cellular redox state of transformed cells and tumors could be of particular importance in identifying the cellular phenotype and optimizing therapeutic approaches for tumor cell killing.
Whereas EPRI measures free radical metabolism and overall redox state, it is not specific for the cause of alterations in redox state. With inflammatory reactions without limited oxygen supply, the redox state would be expected to shift to a more oxidized condition, and one would expect the rate of nitroxide reduction to be slowed. Prior in vivo and ex vivo examples in the literature of this have been reported, including the effect of ultraviolet light on the skin of human volunteers or in the postischemic rat heart (17 , 48) . In the current study, EPRI was applied to measure the redox state of tumor versus normal tissue by measuring the reduction rate of administered nitroxide free radicals. By comparing the reduction rate in different spatial locations, the tissue redox state can be differentiated. In the case of mucosal injury and inflammation, the redox state may be different from that in the tumor area, so that EPRI can potentially be used to differentiate the tumor tissue versus the injured and inflamed mucosa. However, in our imaging study, there was no apparent inflammatory infiltration in the tumor area of the stomach shown from the histologic slices.
The detection of the redox state of tumors by EPRI could be particularly useful with regard to the choice and optimization of cancer chemotherapy or radiation treatment regiments as well as in the assessment of the efficacy of this treatment. Recent clinical studies (28) have demonstrated unequivocally that the extreme reduced state in solid tumors causes resistance to some anticancer drugs (49, 50, 51) and causes the extracellular pH to decrease (52) , which also renders some drugs ineffective (53 , 54) . However, other drugs remain effective in killing hypoxic cancer cells. Low oxygenation has also been reported to accelerate malignant progression and metastasis (29 , 55) . The ability of EPR methods to not only measure redox state but also provide measurement and spatial mapping of tissue oxygenation should also be very useful in this regard. Thus, EPR spectroscopy and imaging techniques could be of particular use in selection and monitoring of cancer treatment.
As such, the ability to noninvasively and nondestructively identify and image tumor tissue based on its altered redox state and oxygenation should be of particular value both in defining the basic mechanisms of oncogenesis and in optimizing therapies to treat cancer. The application of EPRI reported here is an early attempt to assess the ability of this recently developed technique to provide this type of critical information. Whereas the results obtained here are encouraging in that, in the gastric cancer model studies, the tumor could be visualized, and its redox state could be differentiated from that of the surrounding normal tissue, there are a number of further advances and developments that will be required to realize the full potential of this technique and to advance its eventual use in the important areas of cancer diagnosis and treatment. The major advances required include the development and testing of suitable clinically applicable redox or oximetry paramagnetic probes, as well as further instrumentation advances enabling higher sensitivity detection on larger samples.
At present, applications of in vivo EPRI have been restricted mostly to studies in small animals. These limitations have been based on the magnet geometries used, as well as limitations of the radiofrequency that can penetrate larger biological objects and related sensitivity concerns. Recently, we have performed the first applications of this technology to humans with the performance of topical EPR measurements and imaging of nitroxides in normal human volunteers (17) . These studies have successfully provided measurement and imaging of the tissue redox state in small, defined regions close to the surface. To date, EPRI has been used for imaging studies of human volunteers by several groups (17 , 18 , 56 , 57) . Future clinical applications may be facilitated by localized EPR spectroscopy and imaging with surface resonators or implantable resonators. This approach can optimize local sensitivity and limit the amount of required EPR probe, facilitating applications. The unique ability of EPR spectroscopy and imaging to measure tissue oxygenation and redox state has been proposed to be particularly useful in peripheral vascular disease, radiation and other tumor treatment, and tumor diagnosis, with related efforts under investigation in several groups to advance this technique from the basic research laboratory to clinical application.
In conclusion, our results demonstrate redox differences between normal and tumor cells in vivo. Based on this, we observe that EPR spectroscopy and imaging techniques enable identification and visualization of gastric carcinoma in the stomach of rats. These tumors possess a highly reduced state. The ability of this methodology to provide spatial mapping of tissue redox state as well as tissue oxygenation should be particularly useful in studies assessing the role of oxidant-mediated induction of cancer and the efficacy of anticancer therapies. With further technical and pharmacological advances, in the future this methodology could potentially provide valuable information regarding the clinical diagnosis and treatment of cancer.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Note: T. Mikuni and G. He contributed equally to this work.
Requests for reprints: Jay L. Zweier, 110G Davis Heart and Lung Research Institute, 473 West 12th Avenue, Columbus, OH 43210-1252. E-mail: zweier-1{at}medctr.osu.edu
5 S. Petryakov, G. He, P. Kuppusamy, J. L. Zweier. Bridged loop-gap resonator design for topical electron paramagnetic resonance imaging at 750 MHz, manuscript in preparation. 
Received 2/ 9/04. Revised 6/28/04. Accepted 7/ 9/04.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
