| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
* Department of Surgery, Division of Surgical Research, Children’s Hospital, and
Department of Ophthalmology, Harvard Medical School, Boston, Massachusetts, USA;
Unidad de Investigacion, Hospital de Leon. Altos de Nava S/N. Leon Spain 24071; and
Department of Hepato-Biliary Surgery, University of Hamburg, University Hospital Hamburg, Germany
1Correspondence: *Department of Surgery, Division of Surgical Research, Enders-1022, Children’s Hospital, Harvard Medical School, 300 Longwood Ave., Boston, MA 02115, USA. E-mail: robert.damato{at}tch.harvard.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: angiogenesis • dormant tumors • occult cancers
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. However they appear to be common and are harbored at various stages of cancer. Occult tumor cells can persist in a state of dormancy for many years in individuals treated for cancer (2)
. Upon relapse, dormant cancer cells may emerge as a rapidly expanding tumor mass (3
, 4)
. Individuals with ‘unknown primaries’ may present with metastatic cancer in the absence of a detectable primary tumor. Because of their small size, these primary tumors may not be found unless they grow to a detectable size (5
, 6)
.
Occult primary cancers are also found incidentally during autopsies of individuals who did not die of cancer (trauma, for example). These tumors are usually small (< 1 mm) and are harbored at a rate much higher than the number of cancers that are diagnosed. These microscopic tumors may therefore serve as a reservoir in which a small subset gives rise to large solid tumors, whereas the majority of the tumors do not become manifest (1)
.
Due to the small size of occult microscopic tumors and the lack of suitable models, the biology of these tumors and their transition to expanding solid tumors are not well understood. We hypothesized that some transplanted tumors that fail to grow logarithmically remain viable and persist as small tumor clusters that can easily be overlooked during routine visual inspection. In this respect, such tumors may resemble occult, dormant human cancers. We therefore examined whether green fluorescence protein (GFP)-expressing human tumor lines that fail to grow in experimental animals could be found in vivo. Small clusters of tumor cells were found to persist, and these foci were characterized with respect to angiogenesis, proliferation, and apoptosis in their dormant state and on their transitioning to expanding solid tumors.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The MG-63 (8)
and the p53-/- SAOS-2 (9)
osteogenic sarcoma lines were obtained from the American Type Culture Collection (ATCC; Rockville, MD).
Cell transfection
Cells were plated in 6-well tissue culture plates at a density of 2 x 105 cells/well and allowed to attach overnight. Before transfection, the adherent cells were washed twice in serum-free media. For each well, the cells were transfected with 2.5 µg of DNA and 12 µL of GenePORTERTM (Gene Therapy Systems, San Diego, CA) in 1 mL of serum-free DMEM at 37°C. After 9 h, an equal volume of complete DMEM containing 14% FBS was added to the transfection media. After 15 h, the media was changed and cells were cultured for an additional 24 h. The transfected cells were selected for stable transformants in complete media containing the appropriate antibiotic (0.5 mg/mL G418 or 200 µg/mL hygromycin) for 2 wk.
Fluorescence microscopy
A Leica fluorescence stereo microscope (model MZ FLIII) equipped with a 100 W high-pressure mercury vapor arc lamp was used. GFP was excited using a 480/40 nm excitation filter (GFP2) and emitted fluorescence was visualized through a 510 nm barrier filter. Images were collected using a DAGE-MIT DC-330 color camera and processed using Openlab software (Improvision, Coventry, UK).
Histology
Formalin-fixed, paraffin-embedded tissues were cut into 4 µm sections and baked on slides for 20 min at 65°C. The sections were deparaffinized and stained in Harris’ hematoxylin and eosin Y (H&E, Fisher, Pittsburgh). For CD31 staining, deparaffinized sections were treated with 20 µg/mL proteinase K in PBS at room temperature for 30 min. The slides were washed twice in PBS and treated with 3% hydrogen peroxide in PBS for 5 min at room temperature. The sections were incubated in blocking buffer containing 10% goat serum in PBS for 10 min at room temperature. The sections were then incubated with 5 µg/mL primary antibody to CD31 (MEC13.1 antibody, PharMingen, San Diego, CA) diluted in blocking buffer for 1 h at room temperature in a humidified chamber. The slides were rinsed three times in PBS and incubated with secondary biotinylated antibody to rat (Vector Laboratories, Burlingame, CA) diluted 1/200 in PBS containing 10% goat serum. After 30 min at room temperature in a humidified chamber, the slides were developed using a tyramide amplification kit (NEN) and visualized with DAB substrate (Vector). The sections were counterstained with Gill’s hematoxylin. Sections were stained for proliferating cells using an antibody to PCNA (proliferating cell nuclear antigen; Zymed, San Francisco, CA). Apoptotic cells were stained using the Apop TagTM kit (Intergen, Purchase, NY) and counterstained with methylene green.
To quantitate proliferation, apoptosis, and CD31 staining, the sections were first scanned at low power (100x) to avoid counting in areas of necrosis. The number of proliferating and apoptotic cells was counted at high power (400x) and expressed as a percentage of the total number of cells. CD31 staining was quantitated by counting the number of squares occupied by a vessel in a 20 x 20 square grid at 400x power.
Reagents
The activated c-Ha-ras gene from EJ bladder carcinoma (10)
, cloned into the pSV2Neo vector, was kindly provided by Dr. H. N. Ananthaswammy (U.T. M.D. Anderson Cancer Center, Houston, TX). The pSV2Neo vector containing the G418 resistance marker was used as a control. A sequence containing the open reading frame of activated c-Ha-ras was amplified by reverse transcriptase-polymerase chain reaction using the upstream primer 5'-gaattccatgacggaatataagctggt-3' and the downstream primer 5'-ctgcagtcaggagagcacacacttgc-3'. The amplified DNA was sequence verified and cloned in frame into the EcoRI-PstI site of the vector pEGFPC1 (Clontech, Palo Alto, CA) containing the G418 resistance marker. The construct, GFP-ras, generates a chimeric protein in which activated c-Ha-ras is fused to the carboxyl terminus of GFP. The empty vector expressing GFP alone was used as a control. An expression vector for the human VEGF165 isoform containing the hygromycin resistance marker was kindly provided by Dr. Shay Soker (Children’s Hospital, Boston). An ELISA assay (R&D Systems, Abingdon, UK) was used according to manufacturer’s directions to quantitate VEGF protein in tumor-conditioned media.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A, B). Careful examination under the area of hair growth revealed a small white focus, which by histology was found to contain tumor cells. Culturing the resected focus and reestablishing a cell line that morphologically resembled the parental line demonstrated viability of the foci. The reestablished cell line expressed a human epitope that was recognized by an antibody specific to the human alpha 3 integrin receptor (not shown). This confirmed that the tumor foci in the SCID mice were derived from the injected human tumor cells. Figure 1C
shows the widths of the tumor foci 3 to 4 months after subcutaneous (s.c.) injection. The histogram shows that the tumors persisted for several months within a limited size distribution and did not grow beyond a threshold size. The dormant tumor lesions were found by immunohistochemistry to contain proliferating cells (Fig. 1D
) as well as cells undergoing apoptosis (Fig. 1E
). This suggested that the inability of the tumors to grow beyond a threshold size was not due to lack of tumor cell proliferation, but instead was attributed to proliferating cells in balance with cells undergoing apoptosis.
|
These observations prompted us to investigate the fate of other tumor lines that failed to form visible tumors in vivo. The SAOS-2 tumor line, as reported by others, did not form tumors in the s.c. space of immunosuppressed mice (ATCC). To reliably identify small tumor cell clusters in vivo, the tumor cells were stably transfected with the gene for GFP. Mice injected with these tumor cells were examined under a fluorescent microscope. Even though these mice did not develop visible solid tumors, foci of GFP-expressing tumor cells were localized noninvasively from the skin surface by blue light (488 nm) epi-ilumination (Fig. 2
).
|
When the mouse was killed, a small white tumor (which fluoresced green under blue light) was attached to the underside of the skin. H&E-stained histological sections of these tumors revealed that most of the healthy cells were near the outer perimeter whereas the central region was necrotic (Fig. 2B
). Immunostaining for PCNA revealed that the tumor foci contained both proliferating and apoptotic cancer cells (Fig. 2B
, Table 1
). To reveal microvessels, the sections were stained for CD31, a marker expressed on endothelial cells. Rare CD31-positive staining vessels were found scattered in the dormant tumor lesions.
|
To determine whether the lack of tumor growth in the s.c. space was a site-specific phenomenon, the GFP-expressing SAOS-2 tumor cells were injected intravenously by tail vein injection, which frequently results in tumor colonies in the lungs. Examining the lungs under a dissecting fluorescence microscope revealed colonies of tumor cells on the surface (Fig. 3
A). Tumor clusters in the lungs 
25–50 µm in diameter were found on histological sections 3 wk after tail vein injection (Fig. 3B
). By immunohistochemistry, the tumor clusters were in close proximity to vessels (CD31-positive) and contained proliferating (PCNA-positive) cells but virtually no apoptotic (TUNEL-positive) cells, which indicated that the tumors were expanding in mass. After 10 wk, tumor cell clusters 0.5 mm in diameter were found in histological sections (Fig. 3B, C
). These tumor clusters also contained proliferating cells but, in contrast to earlier time points, contained more apoptotic cells. Anti-CD31 antibodies stained vessels in the lungs, but few vessels were detected within the tumors. Even after 17 wk, the mice appeared healthy and showed no signs of weight loss or labored breathing. The growth rate of the tumors had begun to plateau, since the average size of the tumors in the lungs had increased to only 0.7 mm (Fig. 3C
). The tumors still contained a high percentage of proliferating cells, however, which indicated that the decreased rate of mass increase was not due to lack of proliferation, but instead was due to an increase in the fraction of cells undergoing apoptosis.
|
A third tumor line that failed to form tumors in immunodeficient mice was characterized. The gastric tumor line ST-2 failed to form visible tumors when injected into the s.c. space, the peritoneum, or the tail vein. However, in rare instances (<3%=1/33) mice inoculated with this tumor in the s.c. space, after remaining apparently tumor free for many months (>8 months), developed solid tumors. This demonstrated that some tumor cells persisted below the threshold of palpable detection for many months. In the majority of mice, tumor cells were not found even under a dissecting scope. To reliably address the frequency at which ST-2 formed persistent foci, the tumor cells were engineered to stably express GFP before injecting into mice. Despite the lack of apparent tumor growth 3 months after inoculation, all of the mice (10/10) harbored microscopic GFP-expressing tumor cell clusters (Fig. 4
A). By histology, dormant gastric tumor foci were found to contain proliferating as well as apoptotic tumor cells (Table 1
and Fig. 6
). The inability of ST-2 to form a visible mass therefore did not appear to be due to lack of proliferating cells, but was due to a balance of cells undergoing proliferation and apoptosis. To determine whether the lack of progressive tumor growth was specific to the s.c. space, the GFP-expressing ST-2 tumor cells were also injected into the peritoneum. After 8 wk, the mice showed no signs of weight loss or ascites that would indicate a tumor mass in the peritoneum. Under a fluorescent dissecting microscope, however, small clusters of tumor cells were found attached to tissue in the peritoneal cavity (Fig. 4B
).
|
|
Having established the persistence of small tumor foci in vivo, we sought to induce loss of tumor dormancy by introducing genes that could promote tumor expansion. In Fig. 5
, MG-63 tumor cells were transfected with either the gene for GFP, or ras fused to the 3' terminus of GFP (GFP-ras). The transfected tumor lines (106 cells) were injected into the s.c. space of SCID mice. As expected, the MG-63 tumor cells transfected with GFP did not expand in mass. After 4 months, the mice injected with MG-63-GFP were killed and the skin removed. We found thin, white tumor foci attached to the facia that contained relatively few vessels and were fluorescent green under blue light. These tumors have never been observed to escape even after 8 months of observation. In contrast, the MG-63 tumor cells transfected with GFP-ras formed solid tumors after 1 month. These tumors were richly vascularized and reddish in appearance under bright field. The tumors were fluorescent green when excited with blue light, which demonstrated that the GFP-Ras chimera was expressed in the tumor cells. Abundant, nonfluorescent vessels were clearly visible against the background of the fluorescent tumor. Loss of dormancy induced by activated ras (40/40) correlated with increased levels of VEGF in the conditioned media relative to control tumor cells (Table 2
). Immunohistochemistry of MG-63 tumor cells transfected with activated ras showed a robust angiogenic response by CD31 staining accompanied by increased proliferation and decreased apoptosis of tumor cells (Table 1
, Fig. 6
).
|
|
Transfection of ras into ST-2 cells resulted in a loss of dormancy similar to that observed in MG-63 (35/35 tumors, Table 2
). The tumors exhibited increased VEGF production (Table 2)
and a 20-fold increase in microvessel counts, accompanied by increased proliferation and decreased apoptosis. Transfection of ras into SAOS-2 surprisingly did not result in a loss of dormancy. In an attempt to elicit an explanation for the difference in this line from the other lines, we examined the effect of ras on VEGF levels in SAOS-2 cells. ras transfection failed to induce a significant increase in VEGF levels in these cells. The failure of ras transfection to induce VEGF expression and loss of dormancy was not due to lack of activated ras expression, since the tumor cells expressed the protein by Western blot and exhibited increased urokinase activity (not shown).
To directly address the role of VEGF on loss of dormancy, we transfected SAOS-2 cells with VEGF. This transfection resulted in an eightfold increase in VEGF production. Loss of dormancy was induced in VEGF165-overexpressing tumors (4/10 tumors), which was not observed in the activated ras transfected tumor cells (0/15) or in the parental tumor line (0/12). Histological examination of the tumors confirmed a 10-fold increase in microvessel count, which was accompanied by a decrease in the percentage of cells undergoing apoptosis without a change in the fraction of proliferating cells. Transfection of ST-2 cells with VEGF165 produced similar results as the SAOS-2 cells with regard to increased angiogenesis and decreased apoptosis. Thus, in both cases VEGF165 transfection alone was able to induce an escape from dormancy albeit not as striking as that observed with ras transfected MG-63 and ST-2. For unknown reasons, the transfected VEGF165 cDNA was not expressed at elevated levels in MG-63 tumor cells in two separate transfections. Since elevated protein levels could not be detected, the effect of transfecting VEGF165 in Table 1
and 2
was not included for MG-63.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In patients who have been treated for primary cancer, relapse after a long ‘disease-free’ period is attributed to the emergence of occult, dormant cancer cells. In breast cancer, a small percentage of women surgically treated are at risk for recurrence even 15–20 years after being ‘cured’ of cancer (2)
. Long-term persistence has also been reported for other cancers such as occult melanoma (3)
, osteogenic sarcoma (11)
, and gastric cancer (12)
. It can be speculated that individuals treated for cancer may harbor a large reservoir of disseminated dormant cancers. In a study investigating the microvessel density of melanoma metastases, micrometastases were found to have low microvessel density, whereas macrometastases were found to have high microvessel density (13)
. This suggests that neovascularization plays an important role in the transition from dormant metastases to rapidly growing tumors.
Occult primary cancers are commonly discovered during autopsies of individuals who did not die of cancer (trauma, for example). These tumors are usually small (often < 1 mm in diameter) and are found at relatively high frequencies. For example, microscopic breast cancers were found in 30% of women autopsied (14)
, although the clinical incidence of breast cancer is only 1%. Microscopic thyroid cancers were found in > 98% of individuals autopsied (15)
, but the rate of thyroid cancer that is diagnosed is only 0.1%. Other occult primary cancers discovered incidentally include prostate cancer (16)
, brain cancer (17)
, pancreatic cancer (18)
, and neuroblastoma (19)
. In ‘unknown primary’ cases, individuals presenting with metastatic cancer harbor a primary tumor that, due to its small size, may remain undetectable for many years (5
, 6)
. Carriers of genetic mutations in genes (brca1, brca2, and cdh1, for example) that predispose individuals to cancer can harbor multiple foci of microscopic tumors on histological examination of prophylactically removed organs, which may appear normal in gross appearance and by palpation (20
, 21)
. The large increase in risk for developing familial cancer, compared to sporadic forms, may be directly related to the large increase in frequency of microscopic foci that develop in the effected organs.
Some studies suggest that angiogenesis plays an important role in regulating the transition from dormancy to progressive tumor growth. For example, in an experimental model for tumor dormancy, tumor cells implanted in the vitreous of the eye were unable to undergo neovascularization and did not expand in volume. Moreover, these tumors contained proliferating cells, which demonstrated that lack of expansion was not due to lack of proliferation. If the dormant tumors came in contact with the retina, they became vascularized and rapidly expanded in volume (22)
. More recently, angiogenesis inhibitors produced endogenously in vivo or administered exogenously induced tumor dormancy that was defined by a balance of proliferation and apoptosis (23
24
25)
. Clinical dormancy of thin, early-stage human melanomas could be recapitulated experimentally by injecting tumor lines derived from such tumors into nude mice. Overexpressing VPF/VEGF121 in the nontumorigenic cutaneous melanoma was sufficient to induce high grade tumorigenic competence in nude mice (26)
.
In a mouse model for ß-islet cell tumorigenesis (RIP-Tag model), all islet cells in a transgenic mouse line express the large T antigen at birth. By 12 wk, 75% of islets progress to small foci of proliferating cells but only 4% are angiogenic. The number of angiogenic islets closely correlated with the incidence of tumor formation, which suggested that an ‘angiogenic switch’ played an important regulatory role in the transition from microscopic foci to solid cancer (27)
. The angiogenic factors aFGF and VEGF were detected before and throughout the angiogenic switch, which suggested that down-regulation of angiogenesis inhibitors as well as induction of proangiogenic factors was required for neovascularization (28)
. The defined temporal and histological changes that occur in the RIP-Tag model are consistent with the multistep paradigm for tumorigenesis of human cancers (29)
. The high incidence of occult human cancers suggests that an angiogenic switch, as in the mouse RIP-Tag model, may be a relatively late event that plays a significant role in the transition from microscopic foci to macroscopic tumors in sporadic and perhaps heritable forms of cancer.
Several studies have shown that angiogenesis may correlate with the growth of human tumors in immunodeficient mice. For example, high and low expression of FGF-3, a member of the fibroblast growth factor family, correlated with tumorigenic and nontumorigenic human colon carcinoma clones (30)
. Also, VEGF189 expression was reported to correlate with human tumor growth in immunodeficient mice (31)
. Such studies did not examine the fate of the transplanted tumor cells. Achilles et al. recently demonstrated that a solid human tumor is heterogeneous with respect to angiogenic activity and tumor growth. The majority of clones derived from a parental mixed population induced angiogenesis and formed rapidly growing tumors. A small subpopulation of clones exhibited little or no angiogenic activity and produced dormant tumors (32)
. A low angiogenic potential was proposed as a mechanism for ‘no take.’
In our experimental system, human tumor lines injected into mice formed small, often microscopic, persistent foci. These tumor lines may represent a relatively homogeneous population of cells that exhibit a dormant phenotype. The tumor clusters were composed of proliferating (from 10–15% proliferation index) and apoptotic cells. It is estimated that 1 million tumor cells having a 10% proliferation index would, in the absence of cell death, produce a 40 cm3 tumor after 120 days (4 months). Instead, the three human tumors studied were found to persist with an average diameter of <0.05 cm (<10-4 cm3), <1/100,000-fold the volume predicted in the absence of apoptosis or cell death.
Some tumor lines that fail to grow in mice can be stimulated to grow by injecting them in the presence of a laminin-enriched basement membrane matrix (Matrigel). Matrigel or a component of it has been observed to promote angiogenesis, so it has been speculated that Matrigel facilitates tumor growth by enhancing tumor cell survival (33)
. We have injected the dormant tumor lines used in our studies in the presence of Matrigel but did not see an effect in vivo with respect to loss of dormancy. Such studies suggest inherent differences in mechanisms leading to an angiogenic switch in different tumor cell lines. Although Matrigel did not induce loss of dormancy, overexpressing VEGF165 induced a robust angiogenic response and loss of dormancy in 30% of animals injected with ST-2 and 40% of animals injected with SAOS-2. Loss of dormancy was accompanied by reduced apoptosis and, in contrast to transfecting ras, was not accompanied by a significant increase in proliferation. Thus, angiogenesis induced by VEGF was sufficient to induce loss of dormancy by reducing apoptosis. Reduction in the apoptotic fraction accompanying angiogenesis may be due to increased perfusion, but may also be attributed to endothelial cell derived factors that promote tumor growth and survival (34)
.
The greater potency of activated ras on tumor growth with respect to VEGF165 is consistent with reports that overexpressing VEGF alone did not recapitulate the rapid growth phenotype induced by activated ras in other experimental tumor models (35
36
37)
. The greater potency of ras can be attributed to both a direct as well as an indirect effect. Activated ras can directly stimulate tumor proliferation (38)
and confer resistance to apoptosis (39)
. Indirectly, activated ras can stimulate angiogenesis and tumor growth by inducing angiogenesis factors such as VEGF (40)
and by down-regulating angiogenesis inhibitors such as thrombospondin (41
, 42)
. Activated ras may promote neovascularization by altering the expression of multiple stimulators and inhibitors of angiogenesis. Similar to ras, other oncogenes and anti-oncogenes indirectly affect tumor growth by an angiogenic mechanism (43)
. Examples include myc, which down-regulates thrombospondin expression (44)
, and bcl-2, which induces VEGF expression (45)
. The p53 tumor suppressor gene regulates the expression of thrombospondin (41)
and other uncharacterized inhibitors of angiogenesis (46)
. Introducing p53 into a mouse fibrosarcoma elevated expression of thrombospondin and induced angiogenesis-restricted dormancy (47)
.
Models for persistent, dormant cancer may be useful for uncovering mechanisms that control the transition from dormancy to progressive growth. In vivo models for human tumor dormancy would also be useful in the design of therapies to treat occult cancers to prevent relapse or primary cancer, particularly in individuals who are at high risk (20
, 21
, 48)
. Since the rate of proliferation and apoptosis in dormant cancers are balanced, agents that induce a decrease in the proliferation rate or an increase in the apoptotic fraction may result in their eradication. GFP-expressing dormant tumor cells, which can be detected noninvasively in vivo through the skin, would be useful in evaluating treatment efficacy over time.
|
ACKNOWLEDGMENTS |
|---|
Received for publication October 18, 2001.
Revision received April 26, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Folkman Tumor Suppression by p53 Is Mediated in Part by the Antiangiogenic Activity of Endostatin and Tumstatin Sci. Signal., September 26, 2006; 2006(354): pe35 - pe35. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. N. Naumov, E. Bender, D. Zurakowski, S.-Y. Kang, D. Sampson, E. Flynn, R. S. Watnick, O. Straume, L. A. Akslen, J. Folkman, et al. A model of human tumor dormancy: an angiogenic switch from the nonangiogenic phenotype. J Natl Cancer Inst, March 1, 2006; 98(5): 316 - 325. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Udagawa, M. Puder, M. Wood, B. C. Schaefer, and R. J. D'Amato Analysis of tumor-associated stromal cells using SCID GFP transgenic mice: contribution of local and bone marrow-derived host cells FASEB J, January 1, 2006; 20(1): 95 - 102. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Facchini, F. Fiore, M. Caraglia, R. D'Angelo, G. Nasti, F. Caponigro, R. Formato, A. Budillon, and R. V. Iaffaioli A liver angioma colonized by colon cancer cells in a patient with two primitive localizations by colon adenocarcinoma: biologic, diagnostic and therapeutic implications Ann. Onc., December 1, 2005; 16(12): 1980 - 1981. [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Streck, P. V. Dickson, C. Y.C. Ng, J. Zhou, J. T. Gray, A. C. Nathwani, and A. M. Davidoff Adeno-Associated Virus Vector-Mediated Systemic Delivery of IFN-{beta} Combined with Low-Dose Cyclophosphamide Affects Tumor Regression in Murine Neuroblastoma Models Clin. Cancer Res., August 15, 2005; 11(16): 6020 - 6029. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Epstein Maintenance Therapy to Suppress Micrometastasis: The New Challenge for Adjuvant Cancer Treatment Clin. Cancer Res., August 1, 2005; 11(15): 5337 - 5341. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Wood, T. Udagawa, Y. Hida, and R. J. D'Amato X-Linked Dominant Growth Suppression of Transplanted Tumors in C57BL/6J-scid Mice Cancer Res., July 1, 2005; 65(13): 5690 - 5695. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. C. Graham, L. A. Wirtzfeld, L. T. MacKenzie, C. O. Postenka, A. C. Groom, I. C. MacDonald, A. Fenster, J. C. Lacefield, and A. F. Chambers Three-dimensional High-Frequency Ultrasound Imaging for Longitudinal Evaluation of Liver Metastases in Preclinical Models Cancer Res., June 15, 2005; 65(12): 5231 - 5237. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Lu and A. J. Schroit Vascularization of Melanoma by Mobilization and Remodeling of Preexisting Latent Vessels to Patency Cancer Res., February 1, 2005; 65(3): 913 - 918. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Goto, B. Yang, D. Petersen, K. A. Pepper, P. A. Alfaro, D. B. Kohn, and C. P. Reynolds Transduction of green fluorescent protein increased oxidative stress and enhanced sensitivity to cytotoxic drugs in neuroblastoma cell lines Mol. Cancer Ther., September 1, 2003; 2(9): 911 - 917. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||
