| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
Cancer Research UK Centre for Cancer Therapeutics, Institute of Cancer Research, Sutton, Surrey SM2 5NG [P. A. C., M. L. G., S. E., P. W.]; Department of Gastrointestinal Oncology, Royal Marsden Hospital, Sutton, Surrey [D. C., M. E. H., D. M. T.]; and Department of Surgery, Mayday Hospital, Croydon, Surrey [M. L. G., R. I. S.], United Kingdom
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We identified a collection of serial samples, taken pre- and during treatment, from a consecutive series of patients with locally advanced rectal cancer. Using this collection we set out to explore the potential for the application of gene expression profiling by microarray to the measurement of molecular pharmacodynamic responses after treatment with a single dose of MMC4 and during a continuous infusion of 5FU. In this study, we report for the first time gene expression profiling in cancer patients before, and critically, during the period of exposure to chemotherapy. We have demonstrated that the approach is feasible, and we have detected a novel molecular response that would not have been predicted from in vitro studies and that would have otherwise been missed by conventional approaches. The results also suggest a possible new therapeutic approach. Overall our observations suggest that gene expression profiling in response to treatment could greatly benefit clinical studies of conventional or novel chemotherapeutic agents and provide encouragement for additional applications of the technology in the molecular pharmacology of cancer therapy.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Scientific and Ethical Committee approval was obtained. Pelvic MRI was used to determine T stage, and systemic spread was assessed by CT (5)
. All of the patients were T3 or T4 and had a significant risk of incomplete surgical clearance. Protracted venous infusional 5FU (300 mg/m2/day) was delivered via a Hickman line for 12 weeks. MMC (7 mg/m2) was administered as a single dose at the start of treatment and at 6 weeks. Chemoradiotherapy was undertaken for an additional 6 weeks with 5FU (200 mg/m2/day) and radiotherapy using a CT-planned, conformally blocked, three-field technique. Phase I delivered 25 fractions of 1.8 Gy to a total of 45 Gy, with Phase II aiming to deliver a boost to the tumor of 9 Gy in 5 fractions. After chemoradiotherapy 11 patients underwent surgery 4–10 weeks after completion of radiotherapy (median 6 weeks).
|
Tumors were biopsied by rigid sigmoidoscope, under direct vision, and were immediately snap-frozen and stored in liquid nitrogen. Biopsies were performed 6 weeks into the continuous infusion of 5FU and before the second dose of MMC and subsequent chemoradiotherapy. Paired pre- and during-treatment samples were obtained from 15 patients. One patient was limited to a pretreatment sample (patient 14), and 2 additional patients, treated with irinotecan (360 mg/m2; patient 13) or raltitrexed (5.1 mg/m2; patient 17) as a single dose every 3 weeks, were also included.
RNA Extraction and Microarray Analysis.
mRNA was extracted, and radiolabeled single-strand cDNA was prepared and hybridized to gene arrays carrying 4132 I.M.A.G.E/LLNL cDNA clones spotted on nylon membranes (Invitrogen, Carlsbad, CA) as described previously (7)
. Hybridization signals were detected by phosphorimaging, and the quality of signal for each spot was confirmed by visual inspection. Poor quality spots were flagged and removed from the analysis. Signal intensity was normalized using all of the data points and a pooled reference mRNA from 3 different normal bowel samples, because synchronous normal bowel samples were not available. The yield of mRNA was sufficient for a single array run per biopsy. Therefore, a strategy of filtering the data only to include genes altered in multiple patients was used (e.g. Refs. 3
, 8
). This approach eliminates genes altered in single samples and removes genes unaffected by treatment that will not contribute to the identification of differences or similarities between patients. In this study we used a filter of a 2-fold change in gene expression in either at least 2 of 17 pretreatment profiles or in 4 or more of 34 across all of the patient samples. Unsupervised analysis, which relied on expression data alone, used a K means algorithm and a visual output generated using TreeView (8)
. A supervised approach was also used in which prior knowledge of response or treatment was incorporated into the data analysis used a t test with a two-tailed distribution and unequal variance or SAM (9)
. The SAM algorithm identifies genes with statistically significant changes in expression by assimilating a set of gene-specific t tests and assigns a score based on degree of change of gene expression relative to SD; repeated permutations of the data are then used to identify a false detection rate (9)
. To aid interpretation of the gene expression patterns, expression profiles of the pooled normal bowel sample, and tissue cultured colorectal cancer cells (HCT116, HT29, and SW480), T (Jurkat) and B (DOHH2) lymphocytes, endothelial cells (HUVECs), and fibroblasts were included.
Protein Analysis by Immunohistochemistry or Western Blotting.
Frozen 12 µM sections were cut from a portion of the tumor biopsy and stained with H&E or with 2 µg/ml of a mixture of two antibodies that recognize wild-type and mutant conformations of p53 (NeoMarkers, Fremont, CA). Antigen-antibody complexes were detected using an APAAP system (Vector Laboratories, Burlingame, CA). Slides were also stained with a proprietary detection reagent for the senescence-associated ß-galactosidase or a positive control detection solution for endogenous ß-galactosidase (Trevigen Inc., Gaithersburg, MD) and counterstained with Nuclear Fast Red. In some instances, remaining tumor biopsies were analyzed by immunoblotting as described previously (7)
with antibodies specific to TS (a generous gift from Dr. Wynne Aherne, Institute of Cancer Research) or p53 (NeoMarkers).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Treatment was a 12-week protracted venous infusion of 5FU, a fluoropyrimidine antimetabolite (10)
. MMC, a bioreductive anticancer agent that induces DNA damage after enzymatic activation, was administered at the start of treatment and again at 6 weeks. This was followed by chemoradiotherapy for an additional 6 weeks with a reduced dose of infusional 5FU and CT planned radiotherapy. Pretreatment tumor biopsies were taken before the initial dose of MMC, and during-treatment biopsies were obtained 6 weeks into the primary chemotherapy treatment of a continuous infusion with 5FU, but before the second dose of MMC and before the initiation of subsequent chemoradiotherapy.
5FU is activated by metabolism and its metabolites affect multiple molecular targets by: (a) inhibition of TS by FdUMP; (b) incorporation of FdUTP into DNA; and (c) incorporation of FUTP into RNA (10
, 11) . A number of factors influence these mechanisms of action in vivo. The schedule of delivery is important in determining the mechanism of action, as daily or weekly pulse treatments at a high dose are thought to predominantly effect RNA metabolism, whereas a prolonged continuous infusion of a lower dose is thought to result in TS inhibition (10)
. 5FU has also been reported to induce p53 in vitro and in normal mouse intestine in vivo (12)
. Coincubation of 5FU with thymidine or uridine has demonstrated that the induction of p53 in vivo resulted from incorporation of FUTP into RNA (12)
. Induction of p53 will result in apoptosis, and/or cell cycle arrest and senescence. Because these factors would potentially influence gene expression we initially examined TS and p53 expression before expression profiling. We had sufficient material from 8 of the patients to allow immunoblot analysis of TS and p53 expression in pre- and during-treatment biopsies. An inactive ternary complex of TS:FdUMP:5–10 methylenetetrahydrofolate migrating at a slightly higher molecular weight than TS was detected in the during-treatment biopsies, but not the pretreatment samples (Fig. 1A
; Table 1
). Different levels of TS and the ternary complex were detected; however, in this small series of patients there was no correlation between response or outcome and levels of TS or ternary complex (data not shown). Analysis of p53 in pre- and during-treatment samples did not reveal any consistent evidence for p53 induction during treatment (Fig. 1A)
. These observations suggested that the drug treatment regimen resulted in TS inhibition but did not induce a prolonged activation of p53. There was also no evidence for a significant increase in poly(ADP)ribose polymerase cleavage by the apoptotic protease cascade (data not shown) or for the induction of the senescence-associated ß-galactosidase, after treatment (Fig. 1, B and C)
. In addition, the examples shown of p53-negative and p53-positive staining illustrate the general observations that (a) each biopsy consists mainly of tumor; (b) p53 staining, when detected, was restricted to tumor cells; and (c) consistent with the immunoblotting data there was no evidence for prolonged induction of wild-type p53 in normal tissue or tumor after treatment (Fig. 1, D–G
; Table 1
).
|
. Analysis of pretreatment versus during-treatment profiles in each individual patient demonstrated evidence for altered gene expression (Fig. 1, H–K, M, and N
). Comparison of the control versus control scatter plot (Fig. 1L)
to the pretreatment versus during-treatment plots (Fig. 1, M and N)
demonstrated clearly that both the number and degree of altered gene expression was considerably higher in the latter. The number of genes that decreased >2-fold during treatment (median of 219) generally exceeded those that increased >2-fold (median of 53; Fig. 1O
; Table 1
). However, there was no significant association between response to initial chemotherapy or overall response and the number genes changed, or the median or maximum fold-change in gene expression (P > 0.05; t test with a two-tailed distribution and unequal variance).
Analysis of Pretreatment Gene Expression by Microarray.
Analysis of pretreatment gene expression relative to the pooled normal bowel reference mRNA demonstrated many similarities with studies that have previously compared expression profiles of normal and colorectal cancer tissue either using glass slide arrays or serial analysis of gene expression (Fig. 2A
; Refs. 13, 14, 15, 16
). The normal bowel exhibited a higher level of expression of a number of classes of genes of which the products are associated with normal functions of colonocytes (e.g. FABP1, FXYD3, SEPP1 and SELENBP1), hematopoietic cells (e.g. immunoglobulin genes IGL and ILL1), and smooth muscle (e.g. ACTG2, MYLR2, MYL6, and CNN1). Among the genes overexpressed in tumors compared with normal bowel were several that had been reported previously (13, 14, 15, 16)
, such as the c-MYC oncogene, whereas MXI1, a repressor of c-MYC, was decreased (Fig. 2A)
.
|
. An alternative, supervised approach was also used. In this case the data were initially separated into responders or nonresponders and then analyzed using the SAM algorithm to identify genes associated with response (9)
. This supervised approach comparing expression to response detected 7 genes of which the pretreatment expression levels differed by >2-fold and were significantly associated with response to chemotherapy (Fig. 2C)
. One gene, aldhehyde dehydrogenase 7 (ALDH7), the product of which is involved in detoxification, particularly lipid peroxidation, was higher in nonresponders. Six genes (EEF1G, EIF4A1, CYB5-M, ATF4, RPS29, and ATP6S14) were lower in responders. There was no obvious mechanistic link to treatment for any of these genes, although expression of EEF1Ghas been reported previously to be increased in gastrointestinal cancers (17
, 18)
. However, follow-up studies in a larger cohort of patients will be required to validate the significance of these potential markers.
Gene Expression Profiling Detects Altered Gene Expression After Chemotherapy.
The principle objective of the study was to assess the feasibility of microarray profiling to investigate gene expression changes induced by therapy in the tumor tissue of cancer patients. Fig. 3
shows that the expression profiles from the pre- and during-treatment biopsies were divided between two major branches after unsupervised hierarchical clustering. Branch 1 (Fig. 3
A, left side, and Fig. 3
B, bottom) contained 13 pretreatment samples, 3 treatment biopsies, and also the B (DOHH2) and T (Jurkat) cell lines that were included for comparison. Branch 2 (Fig. 3
A, right side, and Fig. 3
B, top) contained 14 during treatment samples, 4 pretreatment biopsies, normal bowel, endothelial (HUVEC) cells, and fibroblasts. Clusters of genes of which the products were associated with lymphocytes, stroma, muscle, or normal bowel were identified (Nodes 1–4, respectively, Fig. 3A
). With one exception, a node with some genes associated with normal colon function, there was no major difference in the distribution of the genes associated with normal tissue types between the pre- and during-treatment tumor expression profiles for each patient. This indicated that any differences in gene expression profile during treatment were unlikely to result from altered content of normal tissue types in the tumor biopsies. The exception was the expression of a subset of genes we had shown to be enriched in the normal bowel pool that were decreased in branch 2 (predominantly during treatment samples; Fig. 2A
; Node 2, Fig. 3A
).
|
). Within this node 297 genes exhibited significantly different expression between branch 1 and branch 2 (P < 0.05; t test with a two-tailed distribution and unequal variance). These included 40 genes encoding products involved in RNA synthesis and processing, 79 genes involved in protein synthesis and processing, and 53 genes involved in metabolism, particularly ATP production. Interestingly, 90 of the 297 genes in this "synthesis and metabolism" gene cluster had been reported previously in expression profiling studies as having increased expression in response to c-MYC expression (19, 20, 21, 22)
. The array experiments detected a 1.8-fold median decrease (range, 0.8–4.5) in c-MYC expression in patient samples during treatment, and this was confirmed by real-time PCR of remaining RNA from 2 patients (data not shown).
The dataset was then divided into pretreatment and during-treatment samples from the 15 patients treated with MMC/5FU for whom we had pre- and during-treatment profiles, and analyzed using the SAM algorithm (9)
. This analysis did not reveal any genes with significantly increased expression during treatment, but did identify a set of genes that exhibited significantly decreased expression during treatment that were encompassed by the "synthesis and metabolism" cluster described above (Fig. 3A
; Table 2
). In this case 247 individual genes enriched with those encoding products involved in RNA synthesis (33), protein synthesis (80), and metabolism (38) showed significantly decreased expression by >2-fold during treatment; of these 80 had been reported previously as c-MYC regulated genes (Table 2
; Refs. 19, 20, 21, 22
). These also included genes encoding the molecular chaperones HSP90ß and GRP94, and also a number of their cochaperones that stimulate chaperone activity (p23, AHA1, CDC37, and HSC71; Table 2
). The ratios of treatment to pretreatment expression levels for these 247 genes across all of the patients were also analyzed by hierarchical clustering (Fig. 3, C and D)
. Thirteen of the 15 patients treated with MMC/5FU clustered closely together. The decrease in expression of the "synthesis and metabolism" genes, although detected, was not as pronounced in patient 8, the only patient that exhibited progressive disease during this study. Two patients (3 and 13) gave a different expression pattern in response to treatment (Fig. 3
C, bottom, and Fig. 3
D, left). Although care has to be taken, as these were single expression profiles, possible explanations for their different gene signatures were considered. Patient 13 was treated with irinotecan, a topoisomerase I inhibitor, suggesting that a difference in treatment regime may be responsible for the different expression profile. Patient 3 was treated with MMC/5FU, but interestingly had lower pretreatment levels of c-MYC compared with the other patients, which may account for the different profile (Fig. 2A)
. Also of note, the single patient treated with the specific TS inhibitor raltitrexed (patient 17) gave a clustering pattern similar to patients treated with MMC/5FU, consistent with a view that this altered gene expression pattern may be associated with TS inhibition.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Interestingly, a major "synthesis and metabolism" cluster of genes was detected in the tumor samples, the expression of which was decreased during treatment. Two thirds of these genes encoded products involved in RNA, and protein synthesis and processing, and also many involved in cellular metabolism, particularly ATP production. Notably, a third of these genes had been reported in previous expression mRNA and protein profiling studies as being positively regulated by c-MYC (19, 20, 21, 22, 23) . Consistent with this observation decreased c-MYC expression was also observed and validated. There was also some overlap between the "synthesis and metabolism" cluster that we describe here to be down-regulated by the chemotherapy regimen and genes reported to be decreased in expression after differentiation of Caco-2 cells, a process associated with a cell cycle arrest and decreased proliferation (24) . This observation would be consistent with our "synthesis and metabolism" cluster potentially being a proliferation-regulated set of genes. However, Mariadason et al. (24) also reported increased expression of genes encoding ribosomal protein genes during differentiation. This was in contrast with the decreased expression of ribosomal protein genes that we detected here, and that have been reported in mRNA and protein expression profiling of c-MYC-regulated genes (19, 20, 21, 22, 23) . Inhibition of c-MYC expression or activity results in reduced proliferation of tumor cells (25, 26, 27, 28) . Therefore, the expression pattern that we detected could reflect: (a) MMC/5FU treatment inhibiting the activity or expression of c-MYC, which then subsequently results in reduced proliferation, or alternatively; b) the expression pattern may be a direct result of reduced proliferation or loss of cells from the S phase of the cell cycle that leads to reduced c-MYC expression or activity; or (c) some combination of the two. Treatment of colon cancer cell lines with 5FU in vitro results in an accumulation of cells in early S phase (data not shown); however, it is unclear from the literature whether the same occurs in tumor cells after 5FU treatment of patients. The expression profiling did not reveal altered expression of genes encoding products that regulate cell cycle progression or are expressed in a particular phase of the cell cycle that might indicate accumulation or exit of cells from a particular phase of the cell cycle. Therefore, future experiments, including expression profiling in vitro after treatment with c-MYC interfering RNA or peptide inhibitors of c-MYC, are now underway to examine and distinguish between the influence of c-MYC and proliferation on the expression of genes in the "synthesis and metabolism" cluster that we have identified.
Interestingly, the single patient treated with the specific TS inhibitor raltitrexed, but not the patient treated with the topoisomerase I inhibitor irinotecan, clustered with the MMC/5FU-treated patients. Clearly care has to be taken in the interpretation of the data from these single patients. However, it is not unreasonable to speculate that the clustering pattern suggests that the gene expression changes seen with 5FU treatment may represent a downstream consequence of TS-inhibition resulting in reduced MYC activity. Because the irinotecan-treated patient was responsive to the drug, it can be tentatively concluded that the gene expression signature observed in during MMC/5FU treatment might not be a consequence of a general antiproliferative effect. Also, consistent with the gene expression profile being associated with TS inhibition, we detected the inactive ternary complex of TS:FdUMP:5–10 methylenetetrahydrofolate in tumor tissue during MMC/5FU treatment.
The nature of the mechanistic link between MMC/5FU treatment and c-MYC in vivo is not clear. However, there is evidence that patients with colorectal cancers that have amplified c-MYC have significantly improved outcome after adjuvant treatment with 5FU/levamisole (29) . It is also of interest to note a recent observation that brief inactivation of c-MYC is sufficient to induce sustained tumor regression in a c-MYC driven in vivo tumor model (30) . This raises the interesting possibility that decreased activity or expression of c-MYC or c-MYC-regulated genes could participate directly in the antitumor mechanism of action of MMC/5FU treatment in vivo.
To our knowledge this is the first clinical study to successfully profile global gene expression specifically in response to chemotherapy by comparing patient biopsies taken before and during drug exposure. The major objective of this relatively small study was to demonstrate the feasibility of the approach and in addition to obtain an initial characterization of the changes in gene expression associated with the chemotherapy regimen. Both of these objectives were achieved. The effects were detected irrespective of response, suggesting that they were not a consequence of treatment outcome, but were more likely a consequence of drug action at its target or a secondary downstream event after target modulation. Interestingly, although we detected evidence for TS inhibition in the tumor tissue, a mechanism consistent with the infusional route of 5FU administration (10) , we also detected decreased expression of genes required for RNA synthesis and processing. This suggests that infusional 5FU could exert effects on RNA metabolism by a mechanism distinct from the RNA-dependent effects caused by the incorporation of 5FUTP into RNA (10, 11, 12) . Importantly, expression profiling facilitated the novel discovery that 5FU treatment caused down-regulation of genes described previously as being c-MYC-regulated. Thus, the treatment regimen was reverting transcriptional changes associated with expression of a known oncogene. This effect was unlikely to have been detected by conventional approaches and illustrated that the hypothesis-generating power of gene expression profiling may be valuable in the clinical setting, as well as in preclinical studies. Clearly, the observed changes need to be followed up in prospective studies with larger numbers of patients where the relationship with response could also be addressed in greater detail.
In addition to suggesting potential important molecular effects of 5FU, our array profiling study also points to a possible new combination treatment for colorectal cancer. In particular, the down-regulation of the molecular chaperone HSP90 in the tumor biopsies by 5FU, and in addition the reduced expression of various cochaperones, such as the HSP90 activating protein AHA1 (31) suggests that the combination of 5FU with an HSP90 inhibitor such as 17-allyamino-,17-demethoxygeldanamycin (currently undergoing clinical trials; Ref. 7 ) could be effective. Additional studies are required to investigate this interesting possibility.
Gene expression profiling studies of the present type have to accommodate logistical and ethical considerations, particularly those associated with the need for repeat tumor biopsy. However, our initial observations have demonstrated both the feasibility and utility of gene expression profiling as a valuable hypothesis-generating tool for studying the molecular pharmacology of cancer drugs in the clinical setting. With judicious study design, such as the inclusion of samples to examine effects of treatment on gene expression in normal tissue and of earlier time points to detect more transient responses to drug exposure, global expression profiling has the potential to enhance the development of both established and novel anticancer agents, and also to facilitate the understanding of their mode of action in the clinic.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org). 
2 Supported by Cancer Research UK (to P. C. and P. W.; grant number SP2330, and P. W. is a Cancer Research UK Life Fellow). S. E. is supported by a Cancer Research UK Studentship.
3 To whom requests for reprints should be addressed, at Cancer Research UK Centre for Cancer Therapeutics, Institute of Cancer Research, Haddow Laboratories, 15 Cotswold Road, Sutton, Surrey SM2 5NG, United Kingdom. Phone: 44-020-8722-4131; Fax: 44-020-8770-7899; E-mail: Paul.Clarke{at}icr.ac.uk
4 The abbreviations used are: MMC, mitomycin c; 5FU, 5-fluorouracil; SAM, significance analysis of microarrays; TS, thymidylate synthetase; MRI, magnetic resonance imaging; CT, computed tomography; HUVEC, human umbilical vascular endothelial cell; FdUMP, 5-fluoro-dUMP; HSP, heat shock protein; FdUTP, 5-fluoro-dUTP; FUTP, 5-fluoro-UTP.
Received 3/13/03. Revised 7/14/03. Accepted 7/29/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
protein in colorectal cancer. Cancer (Phila.), 82: 816-821, 1998.
mRNA in gastric carcinoma. Cancer (Phila.), 75: 1446-1449, 1995.
This article has been cited by other articles:
|
|
 |
|
  A. Maloney, P. A. Clarke, S. Naaby-Hansen, R. Stein, J.-O. Koopman, A. Akpan, A. Yang, M. Zvelebil, R. Cramer, L. Stimson, et al. Gene and Protein Expression Profiling of Human Ovarian Cancer Cells Treated with the Heat Shock Protein 90 Inhibitor 17-Allylamino-17-Demethoxygeldanamycin Cancer Res., April 1, 2007; 67(7): 3239 - 3253. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Grade, P. Hormann, S. Becker, A. B. Hummon, D. Wangsa, S. Varma, R. Simon, T. Liersch, H. Becker, M. J. Difilippantonio, et al. Gene Expression Profiling Reveals a Massive, Aneuploidy-Dependent Transcriptional Deregulation and Distinct Differences between Lymph Node-Negative and Lymph Node-Positive Colon Carcinomas Cancer Res., January 1, 2007; 67(1): 41 - 56. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Boyer, W. L. Allen, E. G. McLean, P. M. Wilson, A. McCulla, S. Moore, D. B. Longley, C. Caldas, and P. G. Johnston Pharmacogenomic identification of novel determinants of response to chemotherapy in colon cancer. Cancer Res., March 1, 2006; 66(5): 2765 - 2777. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Tucker, J. T. Murphy, N. Kisiel, P. Diegelman, K. W. Barbour, C. Davis, M. Medda, L. Alhonen, J. Janne, D. L. Kramer, et al. Potent Modulation of Intestinal Tumorigenesis in Apcmin/+ Mice by the Polyamine Catabolic Enzyme Spermidine/Spermine N1-acetyltransferase Cancer Res., June 15, 2005; 65(12): 5390 - 5398. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. M. Ghadimi, M. Grade, M. J. Difilippantonio, S. Varma, R. Simon, C. Montagna, L. Fuzesi, C. Langer, H. Becker, T. Liersch, et al. Effectiveness of Gene Expression Profiling for Response Prediction of Rectal Adenocarcinomas to Preoperative Chemoradiotherapy J. Clin. Oncol., March 20, 2005; 23(9): 1826 - 1838. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. WORKMAN Drugging the Cancer Kinome: Progress and Challenges in Developing Personalized Molecular Cancer Therapeutics Cold Spring Harb Symp Quant Biol, January 1, 2005; 70(0): 499 - 515. [Abstract] [PDF]
|
|
|
|
 |
|
 X. Lu, W. E. Burgan, M. A. Cerra, E. Y. Chuang, M.-H. Tsai, P. J. Tofilon, and K. Camphausen Transcriptional signature of flavopiridol-induced tumor cell death Mol. Cancer Ther., July 1, 2004; 3(7): 861 - 872. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Workman The critical role of molecular biomarkers in clinical trials of novel cancer therapeutics. AACR Meeting Abstracts, March 1, 2004; 2004(1): 1313 - 1313. [Abstract]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
10%) pre- and during treatment samples. Node 1, genes expressed in endothelial cells; node 2, fibroblast and normal bowel genes decreased after MMC/5FU treatment; node 3, genes expressed in fibroblasts and normal bowel that were not altered after MMC/5FU treatment; node 4, lymphocyte associated genes; node 5, a cluster of 338 genes decreased during MMC/5FU treatment. Green, decreased relative to the median; red, increased relative to the median; black, unchanged relative to the median; gray, no confidence in signal. B, dendrogram of patient samples from A. Expression profiles of the pooled normal bowel sample, and tissue cultured T (Jurkat) and B (DOHH2) lymphocytes, endothelial cells (HUVEC) and fibroblasts were also included (pink text). C and D, TreeView output of K means clustering of the 247 genes identified by the SAM algorithm to be significantly altered >2-fold during treatment with 5FU/MMC, plotted as a ratio of treatment:pretreatment for all patients. CH., response of the tumor to the primary chemotherapy (CR, complete response; PR, partial response; SD, stable disease; and PD, progressive disease) and OV., overall response after completion of chemoradiotherapy (R, response; NR, nonresponse). Genes encompassed by the yellow box are listed (see Table 2