| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Experimental Therapeutics, Preclinical Pharmacology |
1 Lilly Research Laboratories, Indianapolis, Indiana;
2 St. Vincent’s Institute of Medical Research, Melbourne, Fitzroy, Australia;
3 Genzyme Corporation, Framingham, Massachusetts; and
4 Curagen Corporation, Branford, Connecticut
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In our study, we describe the use of an intraosseous model involving the 13762 syngeneic rat mammary carcinoma tumor line to induce osteolytic bone resorption. This model is consistent with our current understanding of tumor-bone interactions at a metastatic site, where tumor cells interact with osteoclasts. Tumor cells were locally implanted into the intramedullary space in the proximal tibiae of female Fisher 344 rats. The tumors and resulting bone destruction were evaluated using radiographic and computed tomography scanning modalities. Imaging was able to visually demonstrate and quantify the local destruction of trabecular, cortical, and whole bone by the implanted tumor. Tumor-implanted animals were evaluated for their response to bisphosphonate therapy. In addition to the local osteolytic response at the site of tumor implantation, the animals also demonstrated significant tumor seeding of the lungs, leading to pulmonary metastases. To compare the expression of PTHrP and TGF-ß1 when the tumor was implanted in different sites, immunohistochemical evaluation was performed. This evaluation confirmed a similar expression pattern regardless of tumor implant site. The tumor cells were also evaluated for their in vitro sensitivity to standard chemotherapeutic agents and bisphosphonates.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cell Proliferation Assay
13762 rat mammary carcinoma cells were plated at a concentration of 1.0 x 103 cells/well in 96-well plates. The cells were exposed to the following agents in complete media for 72 h: risedronate; pamidronate; alendronate; doxorubicin; cis-platinum; carboplatin; methotrexate; 5-fluorouracil; gemcitabine; vincristine; and paclitaxel (0.01–40 µM; Sigma-Aldrich, St. Louis, MO). After a 72-h incubation period, 20 µl of WST-1 cell proliferation reagent tetrazolium salt (Roche, Indianapolis, IN) were added to each well and allowed to incubate at 37°C in a 5% CO2 humidified atmosphere for 1 h. At that time, the optical absorbance of each well was read using a spectrophotometer set at 440 nm. An IC50 was determined for each compound tested. Values reported represent the average of three determinations.
Luminex Assay
13762 cells were grown in cell culture as described previously. The cells were grown in serum-free media for 24 h. At that time, the conditioned media were harvested, and the total number of cells in the flasks was determined using trypan blue exclusion assay. Cytokine levels were determined using a Luminex Multiplex assay. Luminex technology allows for the simultaneous detection and quantification of multiple proteins per sample by using antibodies bound to beads. The total amount of secreted TGF-ß1, vascular endothelial growth factor, basic fibroblast growth factor, and tumor necrosis factor 
in the media was determined and expressed as ng/106 cells. The multiplex kits were custom designed in house and manufactured by R&D Systems (Minneapolis, MN). Each 96-well filter plate (Millipore, Bedford, MA) was blocked with 100 µl of blocking buffer for 30 min and filtered by vacuum at 2 psi. The 25X multiplex bead mix was mixed using a vortex mixer for 1 min and then mixed using a sonic mixer for 30 s before diluting. The bead mix was diluted in wash buffer, and 50 µl were immediately added to each well. Wash buffer was filtered through vacuum at 2 psi.
Each standard was dissolved in media supplied in kits. The standards and samples were added to the filter plates containing the bead mix in 50 µl in triplicate. To measure TGF-ß1, 250 µl of sample were acidified with 50 µl of 1N HCl and incubated for 10 min. The samples were neutralized with 50 µl of 1.2N NaOH/0.5 M HEPES and further diluted with RD6M at a 1:4 dilution. The beads are light sensitive, and therefore the plates were covered with foil. The plates were placed on a shaker at 4°C overnight to allow binding of growth factors to the bead-bound antibodies. The following day, the media were vacuum-filtered at 2 psi, and 50 µl of detection antibody were added to each well. The wells were washed four times with 100 µl of wash buffer. Then, the plates were incubated on a shaker at room temperature for 1 h in the dark. The plates were again washed four times with 100 µl of wash buffer. Streptavidin-phycoerythrin (50 µl) was added to each well. The plates were shaken for 15 min at room temperature in the dark. The wells were washed four times with 100 µl of wash buffer. The beads were then resuspended in 150 µl of wash buffer for analysis. Plates were wrapped in aluminum foil and maintained at 4°C until analysis. Immediately before analysis, the plates were again shaken to ensure complete resuspension of beads. The fluorescence intensity of the detection antibody was determined using a FACScan Luminex 100 (Luminex Corp., Austin, TX). Fluorescence intensity readings for 100 beads/cytokine were collected for each standard and sample dilution. The lower limit of detection for each assay is 5.6 pg/ml.
13762 Tumor Bone Metastasis Model
Animals.
Female Fisher 344 rats (Taconic Farms, Germantown, PA) weighing between 140 and 160 g were used throughout the study. All animals were housed and handled in accordance with The Guide for the Care and Use of Laboratory Animals and Lilly Research Laboratories Institutional Animal Care and Use Committee guidelines. Animals were maintained in microisolator cages in temperature- and light-controlled rooms (12-h cycle). Animals had access to food and water ad libitum.
Tumor Implantation.
13762 cells were grown to 80% confluence, harvested using 0.25% Trypsin-EDTA (Life Technologies, Inc.), counted for cell viability using a trypan blue exclusion test, and then resuspended in serum-free media for intratibial or s.c. implantation. For intratibial tumor cell implantation, the rats were sedated using acepromazine maleate (Boehringer Ingelheim, St. Joseph, MO) at a dose of 2.5 mg/kg i.p. Anesthesia was induced and maintained with 2–3% isofluorane/100% oxygen. The stifle and proximal tibial region were clipped and surgically prepared. Using a 25-gauge needle, the proximal tibia was slowly penetrated via a medial approach to the bone. The approach required a distal-proximal trajectory for the needle to deliver the cells into the intramedullary space. A total of 2.0 x 105 cells suspended in plain DMEM (final volume, 50 µl) were implanted into the tibial intramedullary space. Animals were monitored through the recovery period and returned to their cages.
Radiology.
On a weekly basis, over a period of 4 weeks, animals were anesthetized using 2.5% isofluorane/100% oxygen and radiographed using a Faxitron unit (Faxitron X-Ray, Inc., Wheeling, IL). Animals were placed in ventral recumbancy, and radiographs were captured using Green light-sensitive 3M SE+ Veterinary X-ray film in 3M Asymetrix Fast Detail Green Light Emitting Screens. (3M, St. Paul, MN).
Tissue and Blood Collection.
At the end of each weekly radiography session, five tumor-bearing and five non-tumor-bearing animals were euthanized via CO2 asphyxiation. Tibiae, lungs, blood serum, and plasma were harvested from each animal. Serum was separated from blood collected in additive-free tubes. Tibiae were fixed in 10% buffered formalin for 24 h and then stored in 70% ethanol until scanned. For immunohistochemical evaluation, tibiae were decalcified over a 24–48-h period (Decalcifier I; Surgipath, Richmond, IL) before processing. At euthanasia, the lungs were perfused with 10% buffered formalin, kept in 10% formalin for 24 h, and then stored in 70% ethanol.
Drug Treatment.
A similarly tumor-implanted cohort of rats underwent a random assignment (5 days after implantation) into the following groups: untreated tumor-bearing controls; pamidronate (80 µg/kg/day); risedronate (80 µg/kg/day); and alendronate (10, 1.0, 0.1, and 0.01 µg/kg/day). The drugs were dissolved in 0.9% sterile saline and injected s.c. (daily on days 5–21). Age- and sex-matched normal non-tumor-bearing controls were similarly housed. These animals were radiographically evaluated on a weekly basis after tumor implantation for radiographic evaluation of osteolytic lesions. On day 21, animals were euthanized and processed as described previously.
S.c. Tumor Implantation.
To determine the histological expression of TGF-ß1 and PTHrP in s.c. implanted tumors, 2.0 x 106 13762 cells were implanted s.c. in the flank region of similarly housed rats. Cells were implanted in 0.2 ml of sterile PBS. The tumors were allowed to grow until they reached 2000 mm3, and at that time, the animals were sacrificed using CO2 asphyxiation followed by cervical dislocation; the tumor tissue was harvested and placed in 4% paraformaldehyde. The tissue was allowed to fix for 4 h while shaking before being transferred to 70% ethanol until further processing.
Blood Chemistry.
The collected serum samples were evaluated using an IDEXX clinical blood analyzer (IDEXX Laboratories, Wesbrook, ME). The following parameters were evaluated: total protein; alkaline phosphatase; alanine aminotransferase; blood urea nitrogen; creatinine; phosphorus; glucose; and Ca2+.
Immunohistochemistry.
Paraffin sections (5 µm) were deparaffinized in xylene and rehydrated through graded alcohol into distilled water. Antigen retrieval was performed by heating the sections, immersed in Target Retrieval Solution (Dako, Carpinteria, CA), to 96°C in a water bath for 20 min. Immunostaining was performed with the use of the labeled streptavidin-biotin technique. Endogenous peroxidase activity was first blocked with 3% (v/v) hydrogen peroxide in distilled water for 10 min at room temperature. All remaining steps were also performed at room temperature. After a 30-min treatment with 10% normal horse serum diluted in PBS, the sections were subsequently incubated for 60 min with polyclonal rabbit anti-PTHrP (peptides 1–16) antibody R87 (1:5000; Ref. 43
), sheep polyclonal anti-PTHrP (peptides 50–69) antibody 8094 (1:500), and monoclonal antibody TGF-ß1 (1:10; R&D Systems) prepared using the Dako Animal Research Kit. Antibody dilutions were made with PBS containing 1% normal horse serum. A biotinylated goat antirabbit antibody (1:200; Dako) for PTHrP R87 and a biotinylated rabbit antigoat antibody (1:400; Dako) for PTHrP 8094 were used as secondary antibodies. Sections were exposed to the secondary antibody for 10 min, followed by a 10-min incubation with a peroxidase-conjugated streptavidin (Dako). For TGF-ß1, after addition of the primary antibody, peroxidase-conjugated streptavidin was added for 10 min, and then biotinyl tyramide (New England Nuclear Life Science Products, Boston, MA) was added for 10 min, followed again by streptavidin-peroxidase for 10 min. All signals were visualized using the Dako liquid 3,3'-diaminobenzidine substrate-chromogen system. Negative controls included substitution of the primary antibody with PBS-1% normal horse serum.
Lung Metastasis.
Lungs from untreated, intratibially tumor-implanted animals were visually evaluated. The lungs were perfused with 10% buffered formalin, allowed to fix for 24 h, and then transferred to 70% ethanol. The total number of surface nodules was identified and counted (under 5x magnification). Each experimental group consisted of five animals.
MicroCT Analysis of Rat Tibiae.
Right proximal tibiae from rats were scanned with an MS-8 Small Specimen MicroCT Scanner (Enhanced Vision Systems Corporation, Ontario, Canada). This scanner uses cone-beam geometry to acquire the projection images for the entire volume by rotating through 180° plus the cone angle. The use of an area detector with 1:1 pitch allows the reconstruction of isotropic voxels. In this study, the voxel dimension was 22.6 µm. The scanner settings used were as follows: tube voltage = 55 kV; tube current = 141 µA; exposure time/projection = 1 s; gain on Hamamatsu detector = 63; number of consecutive frames integrated by the Hamamatsu detector array = 2; angle increment between views = 0.749; and number of views = 262. The bones were scanned in a 50% ethanol/saline solution, and an SB3 fanthom block was included in each scan. SB3 is a bone standard that mimics the X-ray attenuation properties of cortical bone. Each reconstructed volume was calibrated using the gray level values of ethanol/saline and SB3. The slice 3 mm from the growth plate was located by slicing through the volume to find the edge of the growth plate and then moving 133 slices from there. All analyses were derived from this slice. A semiautomated deformable boundary technique was used to detect the inner and outer cortical wall, allowing analysis of the trabecular region alone, the cortical region alone, and the whole bone slice. For these three regions, the bone mineral density (mg/cm3) was calculated. Statistical differences between groups were compared using one-way ANOVA with Tukey’s multiple comparison test.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Growth Factor Production.
Evaluation of cytokine release into cell culture medium from 13762 cells maintained for 24 h in serum-free medium demonstrated that tumor necrosis factor , vascular endothelial growth factor, and basic fibroblast growth factor were undetectable in the media. The TGF-ß1 concentration in the medium was 1.35 ng/106 cells. Plasma TGF-ß1 levels from non-tumor-bearing and tumor-bearing rats harvested on day 28 after implantation (n = 5) were an average concentration of 15.98 and 24.98 ng/ml, respectively. The difference between both groups was not statistically significant. In vivo circulating levels of PTHrP were measured but shown to be below the assay’s level of detection (data not shown).
Imaging Studies.
Radiographic evidence of changes consistent with tumor-induced osteolysis was evident 7 days after tumor implantation. On radiographs, the proximal tibiae show radiolucent lesions at the site of tumor implantation. The lesions increase in size in subsequent weeks throughout the experimental period (Fig. 1)
. MicroCT scanning of the tibia detected TBMD loss at day 21 after implantation (Fig. 2)
. For tumor-implanted bones, the mean TBMD was 380.5, 382.6, and 207.8 mg/cm3 on days 7, 14, and 21, respectively. TBMD at day 21 represents a 48% reduction over non-tumor-bearing controls. CBMD was 1204.5, 1232.5, and 1227.4 mg/cm3 on days 7, 14, and 21, respectively. WBMD was 703.0, 701.7, and 548.3 mg/cm3 on days 7, 14, and 21, respectively. This represents a 26% reduction in total WBMD over non-tumor-bearing control animals (Fig. 3)
. Tumor-bearing animals at day 28 after implantation did not offer enough mineral content in any of the three parameters tested to adequately quantify the available bone density.
|
|
|
. At 1, 0.1, or 0.01 µg/kg (daily on days 5–21), there was no statistically significant difference in TBMD between treated and tumor-implanted animals. Pamidronate (80 µg/kg, days 5–21) showed a 52% protection of TBMD when compared with the untreated non-tumor-bearing control. Risedronate (80 µg/kg, days 5–21) was able to maintain TBMD at a level of 93% when compared with non-tumor-bearing animals. This effect was significant (P < 0.001) when compared with tumor-implanted untreated rats. When WBMD was evaluated, alendronate-treated animals (10 µg/kg, days 5–21) showed a statistically significant preservation of bone density at 55% (P < 0.05) of the non-tumor-bearing animals. WBMD in alendronate-treated animals (1.0, 0.1, and 0.01 µg/kg, days 5–21) remained unchanged when compared with untreated controls. Pamidronate (80 µg/kg, days 5–21) did not show a statistically significant difference between the treated and untreated groups. Risedronate (80 µg/kg, days 5–21) was able to maintain WBMD to 95% of the levels seen in the non-tumor-bearing group. This level of mineral density protection was statistically significant (P < 0.001) when compared with tumor-implanted untreated rats. The use of bisphosphonate therapy in this model was able to demonstrate preservation of both TBMD and WBMD. CBMD remained unchanged for all groups.
|
Histopathology.
Tumors growing in several anatomical locations were evaluated for the expression of TGF-ß1 and PTHrP. Tumors grown s.c., intratibially or as a result of either hematogenous or lymphatic spread, consistently demonstrated the expression of both TGF-ß1 and PTHrP (Fig. 5)
. When grown in vivo, the tumor demonstrates an aggressive growth pattern, and the tumor tissue, regardless of anatomical site, shows broad areas of nondifferentiation as well as localized zones of glandular differentiation consistent with an epithelial tumor of mammary origin. Gross metastatic dissemination to the lungs was evident as early as day 14 and progressed until day 28 (Fig. 6)
.
|
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
When grown in cell culture and exposed to several routinely used chemotherapeutic agents, 13762 cells demonstrate sensitivity to many standard chemotherapeutic agents. In contrast, the cell line did not show biologically significant in vitro sensitivity toward alendronate, risedronate, pamidronate, or methotrexate when exposed for up to 72 h at concentrations of up to 40 µM. The relatively low activity of these agents against a mammary carcinoma is consistent with the relatively low cytotoxic profile of bisphosphonates when compared with more traditional cytotoxic agents. Generally, methotrexate is not considered an effective agent against breast tumors. The lack of cell-based cytotoxic activity of the bisphosphonates against 13762 cells in culture was not predictive of the in vivo results obtained. When the tumor-implanted rats were treated with bisphosphonates, significant bone preservation was maintained. In principle, this effect can be attributed to the direct modification of osteoclastic activity by the bisphosphonates. Although we have not excluded the possibility of a direct effect on the implanted tumor cells by bisphosphonates, the antiresorptive activity observed in this study is consistent with the more multifaceted effects of bisphosphonates on bone turnover and directly on osteoclast activity (53, 54, 55, 56) . Previous work has demonstrated elevated concentrations of bisphosphonates associated with osteoclastic surfaces in vivo (57) . This localization is limited to the active resorption space. In the 13762 model of tumor-driven osteoclastic resorption, it is likely that a similar localization occurs. This localization, by nature of the histological setting, does not appear to be directly adjacent to the tumor mass, thus exposure to significantly increased concentrations is unlikely to directly affect the tumor cells. The tumor itself is exposed to the lower circulating plasma concentrations of the agents and not the high local concentrations associated with the osteoclastic space.
When implanted into the tibiae of rats, 13762 tumor cells induced radiographic evidence of resorption, as noted by the presence of lytic lesions on radiographs taken at 7 days after implantation. These lesions, although apparent in radiographic film, did not become quantifiable until day 21 using MicroCT scanning. This discrepancy in detection can be attributed to the intrinsic differences between the imaging methods used. Radiography allows for visualization of the osteolytic process by evaluating the entire affected bone. The use of MicroCT technology permitted the quantification of bone parameters in the selected slice of bone being evaluated. The MicroCT slice selection established in this model identified a bone slice 3 mm distal to the tibial growth plate to determine BMD. This distance corresponds to 133 slices from the growth plate, a standard used throughout the study. Whereas the early stages of tumor implantation remain focally localized, as the invasive and osteolytic process progresses, more of the total bone volume is affected. It is at this point that the imaging techniques can clearly discriminate between the progress of the lytic lesion and the distally induced lesion induced at implantation. As seen in both the radiographs and the whole bone MicroCT images, the bulk of the proximal third of the tibia becomes lysed by the latter time points. It is at the later stages of progression when accurate and reproducible quantification of osteolysis can be achieved using MicroCT. This scanning method was able to accurately and reproducibly detect the progressive loss of both TBMD and WBMD. Due to its higher relative density, CBMD remained relatively unaffected during most of the study period. CBMD was only lost in animals after 28 days of tumor growth. On treatment with bisphosphonates, tumor-bearing animals show clear protection against resorption of TBMD and WBMD. The use of risedronate in this model was able to preserve the tibiae from tumor-induced osteolysis. Until the time of euthanasia, animals did not demonstrate external signs of tumor growth, lameness, or distress throughout the experimental period (up to day 21). Also important to this model is the relatively normal clinical profile of the animals throughout the experimental period. The lack of significant alterations to the measured clinical parameters allows an additional way to discriminate potential adverse effects of test compounds in this model. Although tumor-bearing animals demonstrated expression of PTHrP in the tumor tissue, no significant elevation of serum Ca2+ was observed. We hypothesize that this result could be attributed to the relatively low tumor burden carried by the animals, which could potentially produce insufficient quantities of PTHrP to cause systemic alterations. Another possible reason for the lack of systemic effects is the possibility that PTHrP is being actively degraded or sequestered in bone. PTHrP levels in circulation were measured but found to be negligible (data not shown). Also noted was the lack of an increase in circulating levels of TGF-ß1 even with a strong immunohistochemical signal from the tumors at the different implant sites. Again, we hypothesize that the effects and localization of the secreted TGF-ß1 remain compartmentalized to the affected site. Animals allowed to progress to day 28 did show an increase of anesthetic-induced complications due to the overwhelming pulmonary tumor burden. This burden produces a total decrease in respiratory capacity. Pulmonary nodules arising from this model can be severe. It remains unclear how these lesions originated. In principle, they could have arisen from one of two routes. First, lesion can arise by direct injection to the bone marrow, allowing for hematogenous/lymphatic spread. Second, the lesion can arise by direct metastatic spread from the tumor formed at the injection site. The 13762 tumor has been previously demonstrated to be spontaneously metastatic in rats, and this would suggest that either of the above-mentioned routes would be a likely method for its spread.
Breast cancer induced bone lysis remains a challenging unmet medical need for which the promise of mechanistic-based therapies remains high. This intraossseous syngeneic model represents a novel system to evaluate the activity of agents with activity against tumor-mediated osteolysis. This model offers an advantage over xenograft systems because it uses a syngeneic tumor model in conventional animals, allowing for ease of husbandry and manipulation. We have demonstrated that the model is consistent with our understanding of PTHrP-TGF-ß1-dependent tumor-induced osteoclastic bone resorption. This rat model offers the advantage that numerous relevant parameters can be evaluated simultaneously, in a pathophysiologically relevant manner. These parameters include total bone resorption, cytokine characterization, efficacy of antiresorptive therapies, and toxicity assessment. Development of targeted therapies for tumor-driven bone resorption will be enhanced by this novel multifunctional model.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Requests for reprints: Dr. Enrique Alvarez, CuraGen Corp., 322 East Main Street, Branford, Connecticut 06405. Phone: (203) 871-4319; Fax: (203) 315-3301; E-mail: EAlvarez{at}Curagen.com
5 The abbreviations used are: TGF, transforming growth factor; TBMD, trabecular bone mineral density; CBMD, cortical bone mineral density; WBMD, whole bone mineral density, PTHrP, parathyroid hormone-related protein; MicroCT, microcomputed tomography. 
Received 11/16/02; revised 3/19/03; accepted 3/29/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Commentary
This article has been cited by other articles:
|
|
 |
|
  C. M. Park, J. M. Goo, H. J. Lee, M. A Kim, H.-C. Kim, K. G. Kim, C. H. Lee, and J.-G. Im FN13762 Murine Breast Cancer: Region-by-Region Correlation of First-Pass Perfusion CT Indexes with Histologic Vascular Parameters Radiology, June 1, 2009; 251(3): 721 - 730. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. S. Mitsiades, C. Rouleau, C. Echart, K. Menon, B. Teicher, M. Distaso, A. Palumbo, M. Boccadoro, K. C. Anderson, M. Iacobelli, et al. Preclinical Studies in Support of Defibrotide for the Treatment of Multiple Myeloma and Other Neoplasias Clin. Cancer Res., February 15, 2009; 15(4): 1210 - 1221. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Green Bisphosphonates: Preclinical Review Oncologist, September 1, 2004; 9(suppl_4): 3 - 13. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Onyia, R. J. S. Galvin, Y. L. Ma, D. L. Halladay, R. R. Miles, X. Yang, T. Fuson, R. L. Cain, Q. Q. Zeng, S. Chandrasekhar, et al. Novel and Selective Small Molecule Stimulators of Osteoprotegerin Expression Inhibit Bone Resorption J. Pharmacol. Exp. Ther., April 1, 2004; 309(1): 369 - 379. [Abstract] [Full Text]
|
|
|
|
|
|
M. R. Smith Antitumor Activity of Bisphosphonates Clin. Cancer Res., November 15, 2003; 9(15): 5433 - 5434. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
