This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Etienne, M.-C. Articles by Milano, G. Search for Related Content PubMed PubMed Citation Articles by Etienne, M.-C. Articles by Milano, G. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Medline Plus Health Information Colorectal Cancer Hazardous Substances DB FLUOROURACIL LEUCOVORIN Social Bookmarking                            What's this?

Prognostic Value of Tumoral Thymidylate Synthase and p53 in Metastatic Colorectal Cancer Patients Receiving Fluorouracil-Based Chemotherapy: Phenotypic and Genotypic Analyses

From the Centre Antoine Lacassagne and Centre Hospitalier Universitaire, Nice; L’Institut National de la Santé et de la Recherche Médicale U490, Paris; Institut Paoli Calmette, Marseille; Centre Hospitalier Universitaire, Grenoble; Centre Hospitalier Universitaire, Clermont-Ferrand, France.

Address reprint requests to Gerard Milano, PhD, Laboratoire d’Oncopharmacologie, Centre Antoine Lacassagne, 33 Av de Valombrose, 06189 Nice Cedex 2, France; email: gerard.milano{at}nice.fnclcc.fr

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: The aim of this multicenter prospective study was to evaluate the role of intratumoral parameters related to fluorouracil (FU) sensitivity in 103 metastatic colorectal cancer patients receiving FU–folinic acid.

PATIENTS AND METHODS: Liver metastatic biopsy specimens were obtained for all patients and primary tumor biopsy specimens for 54 patients. Thymidylate synthase (TS), folylpolyglutamate synthetase, and dihydropyrimidine dehydrogenase were measured by radioenzymatic assays; TS promoter polymorphism (2R/2R v 2R/3R v 3R/3R) was determined by polymerase chain reaction; and p53 protein and mutations were analyzed by immunoluminometric assay and denaturing gradient gel electrophoresis, respectively.

RESULTS: p53 mutations were observed in 56.7% of metastases. TS activity was significantly higher in 2R/3R tumors as compared with 2R/2R or 3R/3R. TS activity in metastasis was the only parameter linked to clinical responsiveness (responders exhibited the lower TS, P = .047). Univariate Cox analyses demonstrated that TS activity in primary tumor (the greater the TS, the poorer the survival; P = .040), TS promoter polymorphism in primary tumor (risk of death of 2R/3R v 2R/2R, 2.68; P = .035), and p53 stop mutation in metastasis (risk of death of stop mutations v wild type, 3.14; P = .018) were the only significant biologic predictors of specific survival. Stepwise analysis did not discriminate between TS activity and TS polymorphism.

CONCLUSION: Present results confirm the value of tumoral TS activity for predicting FU responsiveness, point out the importance of detailed p53 mutation analysis for predicting survival, and suggest that TS genotype in primary tumor carries a prognostic value similar to that of TS activity.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
THE BASIS OF THE chemotherapeutic strategy for the treatment of advanced colorectal cancer is currently broadening with the clinical emergence, among a plethora of potential candidates, of new active drugs such as irinotecan¹ and oxaliplatin.² It follows that fluorouracil (FU) is no longer the sole agent with significant antitumor activity in colorectal cancer. This new therapeutic context heightens the need for objective arguments when choosing the most active drug for a given colorectal tumor. Concerning FU itself³ or new promising FU prodrugs such as capecitabine,⁴ several tumoral candidates able to predict FU efficacy have been identified.

Thymidylate synthase (TS), one of the key enzymes controlling DNA replication, is a target enzyme for numerous anticancer drugs, of which FU is the main one. Importantly, the TS promoter enhancer region is polymorphic, and this polymorphism influences the translation efficiency of TS mRNA.^5-7 TS promoter comprises 28-base pair sequences, usually presented as a double-tandem repeat (2R) or a triple-tandem repeat (3R), and it was demonstrated that homozygous 3R/3R cells overexpressed TS mRNA as compared with homozygous 2R/2R cells.^6,7 Numerous investigators have demonstrated that a low tumoral TS expression in colorectal cancer patients receiving FU-based chemotherapy was related to clinical responsiveness as well as to longer survival.^8-14 Relevance of TS activity^15,16 or TS protein¹⁷ has been less extensively studied. In contrast to TS expression or activity, few data have been reported on the clinical relevance of TS polymorphism.¹⁸

The potential role of tumoral dihydropyrimidine dehydrogenase (DPD) is interesting in terms of controlling FU responsiveness¹⁹ because DPD is not only relevant for FU itself but also for new FU prodrugs such as capecitabine or tegafur, which produce FU at the target site. A previous in vitro analysis conducted on 19 human cancer cell lines revealed that DPD activity in tumor cells was significantly related to FU sensitivity, the lower the DPD activity, the greater the FU cytotoxicity.²⁰ This experimental result was confirmed on a human cancer xenograft model demonstrating significant relationships between FU efficacy and tumoral DPD activity or expression.²¹ Clinical confirmation of the predictive role of DPD was provided by Salonga et al,⁸ who reported that DPD gene expression in colorectal tumors was associated with FU responsiveness, with responding patients having lower DPD expression than nonresponding patients.

Because FU requires elevated cellular concentrations of reduced folates in polyglutamate form in order to achieve maximal TS inhibition,²² important for FU cytotoxic activity is the enzyme folylpolyglutamate synthetase (FPGS), which controls the intracellular concentration of polyglutamated reduced folates.²³ We previously demonstrated in vitro²⁴ as well as in a preliminary study on 44 liver metastatic colorectal cancer patients receiving FU-based therapy²⁵ that tumoral FPGS activity was a potential predictor of FU efficacy.

Experimental studies have demonstrated that FU-induced TS inhibition mediates apoptosis mainly through the Fas system.^26,27 Data from Maecker et al²⁸ have demonstrated that p53 status may regulate the Fas expression level. Thus, the central role of p53 at the crossroads between apoptosis and cell proliferation, along with the high proportion of p53-mutated colorectal tumors, highlights p53 as a potential candidate for predicting FU efficacy in colon cancer patients.²⁹

With TS, p53, DPD, and FPGS, one thus disposes of a set of complementary parameters, which could faithfully reflect different aspects of FU tumoral resistance. So far, multifactorial pharmacologic investigations designed to predict FU antitumor efficacy at the clinical level have been rare. This lack of multifactorial studies is accounted for by the fact that such investigations, in order to achieve sufficient statistical power, necessitate large numbers of patients. We thus conducted a prospective multicentric study on a homogenous group of Dukes’ D colorectal cancer patients with unresectable hepatic metastasis treated by FU–folinic acid chemotherapy. Liver metastatic biopsy specimens were available for all patients. Additionally, biopsy specimens of the primary tumor were obtained in 54 patients.

The aim of the present study was to evaluate the predictive value on clinical responsiveness and the prognostic role on specific survival of TS, DPD, and FPGS activities, TS promoter polymorphism, and p53 (evaluated both as intracellular protein concentration and mutation status). This is one of the first studies to include both phenotypic and genotypic investigations of TS and p53.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patients
From January 1994 to November 1998, 103 patients (60 men, 43 women; mean age, 63.7 years, range, 27 to 82 years) were included. This French prospective multicenter study was conducted in four medical centers: Hopital Pasteur, Centre Hospitalier Universitaire (CHU) de Nice (n = 35); Hopital Albert Michallon, CHU de Grenoble (n = 39); Hotel Dieu, CHU de Clermont Ferrand (n = 19), and Institut Paoli Calmettes, Cancer Center of Marseille (n = 10).

All patients exhibited Dukes’ D colorectal cancer with isolated synchronous (n = 90) or metachronous (n = 13) liver metastases assessable by computed tomography scan. Primary tumor location was left colon and sigmoid for 49 patients, right colon for 14 patients, transverse colon for two patients, and rectum for 38 patients. Liver metastases were single in 11 patients and multifocal in 92 patients. Inclusion criteria excluded patients who had received previous FU-based chemotherapy. All patients had a laparotomy, either for primary tumor resection or for an unsuccessful attempted removal of liver metastasis. In both situations, a liver metastatic biopsy specimen was taken at that time for diagnostic and pharmacologic purposes. In addition, for 54 patients, we obtained a tissue sample of the primary tumor. For each biopsy specimen, a slice was cut off for histologic control.

After laparotomy, all patients received chemotherapy combining FU and folinic acid (Fol) (racemic form) given intravenously: 74 patients received a 5-day infusion associating FU (350 mg/m²/d) and Fol (200 mg/m²/d), three cycles at 4-week intervals; nine patients received 12 cycles of weekly 24-hour infusion of FU (1,300 mg/m²) plus Fol (200 mg/m²), and 20 received a 2-hour infusion of Fol (200 mg/m²/d), followed by an FU bolus (400 mg/m²/d) and 22-hour FU infusion (600 mg/m²/d) for 2 days every 14 days (six cycles administered).

Three months after the start of chemotherapy, a new abdominal computed tomography scan was performed to assess the variation in size of hepatic lesions compared with the first tomography, without knowledge of the molecular results. Assessment of objective response was made according to the criteria of the World Health Organization.³⁰ Complete response (CR) was defined as disappearance of all lesions, and partial response (PR) was defined as regression of at least 50% of all lesions. Stabilization (St) corresponded to tumor regression less than 50% or tumor progression less than 25%. Progression (Pg) corresponded to appearance of new lesion or tumor size increase greater than 25%. The same chemotherapy regimen was maintained until disease progression. None of the patients underwent liver metastasis resection (for patients with PR, resection was impracticable as a result of multifocal liver metastases or poor general status). Second-line treatments were as follows: raltitrexed alone (n = 3), irinotecan alone (n = 6), oxaliplatin alone (n = 4), and oxaliplatin-FU combination (n = 5).

Duration of survival was calculated from the start of chemotherapy. For specific survival, the end point was cancer-related death, excluding chemotherapy-related death. Only one patient was lost to follow-up. The median follow-up duration was 52.8 months for patients who did not die. At the time of analysis, 84 patients had died from their disease, and four had died from other causes. Median overall survival was 16.0 months (95% confidence interval, 13.6 to 18.5 months) and median specific survival was 16.4 months (95% confidence interval, 14.0 to 18.9 months).

Investigation of Tumor
Liver metastatic tissue samples (mean, 220 mg; range, 50 to 630 mg) and samples of the primary tumors (mean, 640 mg; range, 120 to 1980 mg) were immediately frozen and stored in liquid nitrogen until processed. Then samples were mechanically pulverized in liquid nitrogen. The resulting powder was homogenized with a Polytron PT-1020 in Tris-HCl buffer, pH 7.4, containing EDTA 1 mmol/L, dithiothreitol 0.5 mmol/L, and sodium molybdate 10 mmol/L (50 to 400 mg in 500 to 1,000 µL) and centrifuged at 105,000 x g for 1 hour to obtain a cytosol stored at -80°C for analysis of p53 protein content, TS, and DPD activity. FPGS activity was measured immediately after cytosols were obtained. The pellet resulting from the 105,000 x g centrifugation was stored at -20°C and used for DNA extraction. Cytosolic proteins were determined by the Bradford colorimetric assay (Protein Assay Reagent, Bio-Rad Laboratories, Munich, Germany) with serum albumin as standard.

TS Activity Assay
TS activity was measured according to the tritium-release assay described by Spears and Gustavsson.³¹ The assay consisted of incubating 25 µL of cytosol with ³H-deoxyuridine monophosphate (1 µmol/L final concentration) and CH₂FH₄ (0.62 mmol/L final concentration) in a total volume of 55 µL. After 0, 10, 20, and 30 minutes of incubation at 37°C, the reaction was stopped on ice. The excess of ³H-deoxyuridine monophosphate was removed by adding 300 µL of activated charcoal (15%) containing 4% trichloroacetic acid (5 minutes’ centrifugation at 14,000 x g, room temperature). The ³H₂O formed during the incubation was then counted in an aliquot of the above supernatant. Results were expressed as femtomoles of ³H₂O formed per minute per milligram of protein, on the basis of the linear regression obtained from the incubation times. The sensitivity limit was 10 fmol/min/mg protein. The coefficient of variation for interassay reproducibility (pooled cytosol, n = 12) was 21.7%.

FPGS Activity Assay
FPGS activity was measured according to the method described by Montero and Llorente³² and consisted of incorporating an additional ¹⁴C-glutamic acid residue into the glutamate chain of aminopterin. Each cytosol was assayed in duplicate. The assay consisted of incubating 100 µL of cytosol with ¹⁴C-glutamic acid (isotopic dilution, 250 µmol/L final concentration) and aminopterin (250 µmol/L final concentration) in a total volume of 250 µL. After 2-hour incubation at 37°C, the reaction was stopped by addition of 50 µL of 40% trichloroacetic acid. The tubes were then centrifuged for 10 minutes at 3,000 x g. The supernatant was analyzed for the presence of aminopterin polyglutamates by high-performance liquid chromatography with an RP18 5-µm Lichrospher column and a radioactive flow monitor (LD 506; Berthold, Wildbad, Germany). Results were expressed as femtomoles per minute per milligram of protein. The limit of sensitivity was 400 fmol/min/mg protein. The coefficient of variation for interassay reproducibility (pooled cytosol, n = 16) was 9.69%.

DPD Activity Assay
DPD activity was measured according to the method described by Harris et al.³³ The assay consisted of incubating 50 µL of cytosol with ¹⁴C-FU (20 µmol/L final), beta-nicotinamide adenine dinucleotide phosphate (250 µmol/L final), and magnesium chloride (2.5 mmol/L final). Total volume was 125 µL (in 35 mmol/L sodium phosphate buffer, pH 7.5, containing sodium azide). The duration of incubation was 30 minutes at 37°C. The reaction was stopped by addition of 125 µL of ice-cold ethanol, followed by 30 minutes’ storage at -20°C. The samples were centrifuged (5 minutes, 400 x g) to remove proteins, and the supernatant was analyzed for the presence of ¹⁴C-dihydrofluorouracil, ¹⁴C- fluoro ß alanine, and ¹⁴C- fluoro ß ureidopropionate with a high-pressure liquid chromatographic method. Detection was performed via a radioactive flow monitor (LD 506; Berthold). DPD activity was calculated by taking into account the sum of dihydrofluorouracil, fluoro ß alanine, and fluoro ß ureidopropionate peaks. DPD activity was expressed as picomoles of ¹⁴C-FU catabolized per minute per milligram of protein. Each sample was assayed in duplicate. The sensitivity limit was 20 pmol/min/mg protein. The interassay reproducibility (pooled cytosol, n = 14) gave a coefficient of variation of 13.8%.

TS Promoter Polymorphism
Polymerase chain reaction (PCR) amplifications were run on a GeneAmp PCR system 9700 (Applied Biosystems, Courtabuf, France) in a 25-µL final volume with 50 ng of DNA as template, 1 mmol/L MgCl₂, 2.5 µL of 10x buffer, 1.25 mmol/L of dNTPs, 0.15 µmol/L of each forward GTGGCTCCTGCGTTTCCCCC and reverse GCTCCGAGCCGGCCACAGGCA primers, and 0.05 U/µL of Taq polymerase Cetus (Perkin-Elmer, Courtabuf, France). Primers were synthesized by Genset (Paris, France). After 30 cycles of amplification (denaturation at 94°C for 30 seconds, annealing at 62°C for 60 seconds, and extension at 72°C for 90 seconds), amplification products were electrophoresed on acrylamide gel at 8%. Products of 248 base pairs (2R/2R), 270 base pairs (3R/3R), or both of these products (2R/3R), depending on the TS genotype, were observed. For heterozygous genotype (2R/3R), allelic imbalance was assessed by two independent observers. Correlations with biologic parameters and clinical data were performed after correction for allelic loss in primary tumors and in metastases (ie, a 2R/3R tumor exhibiting loss of the 2R allele was classified as a 3R/3R tumor, and a 2R/3R tumor exhibiting loss of the 3R allele was classified as a 2R/2R tumor).

p53 Protein Assay
The p53 protein (wild type and mutated forms) was analyzed on 200 µL of cytosol via a monoclonal two-site single incubation immunoluminometric assay (LIA-Mat; Sangtec, Bromma, Sweden). The assay uses two monoclonal antibodies against different denaturing-resistant epitopes of the N-terminal part of the p53 protein. Polystyrene tubes were coated with PAb 1801 monoclonal anti-p53 antibody, and the tracer consisted of isoluminol-conjugated D01 monoclonal antibody. The sensitivity limit was 0.002 ng/mg protein. The coefficient of variation for interassay reproducibility (pooled cytosol, n = 10) was 7.66%.

p53 Mutation Analyses
DNA was extracted by proteinase K digestion and phenol-chloroform purification. Denaturing gradient gel electrophoresis, with primers covering exons 4 through 9, was performed to screen mutations at the p53 gene, as previously described.³⁴ Samples displaying an abnormal electrophoretic migration pattern were sequenced by standard dideoxy sequencing reaction and loading on an ABI 377 sequencer (Applied Biosystems). Mutations leading to a putative truncated protein (stop mutations) and mutations affecting the zinc-binding domain (L2/L3 loop) were specifically recorded.

Statistical Analysis
Statistics were performed on log 10–transformed data for TS, FPGS, and p53, which exhibited asymmetric distribution. Mean comparisons were performed by Student’s t test (for two groups) or analysis of variance (for more than two groups). Relationships between continuous variables were analyzed by means of Pearson correlations, and links between categorical variables were assessed by means of ² test. The potential influence of intratumoral parameters on clinical responsiveness (CR + PR v St + Pg) was analyzed according to logistic regression. Overall and specific survival were computed according to the Kaplan-Meier method, and potential prognostic factors were analyzed according to the Cox proportional hazard regression. Statistical analysis was performed using SPSS software (Paris, France).

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Description of Tumoral Parameters
Table 1 lists the wide variability observed for all parameters, both in primary tumors and liver metastases. TS activity was approximately two-fold higher in primary tumor as compared with metastasis (geometric mean, 325 v 150 fmol/min/mg protein, P = .037) and DPD activity was significantly lower in primary tumor as compared with liver metastasis (mean 100 v 166 pmol/min/mg protein, P < .001). Interestingly, with the exception of DPD, significant positive correlations were observed between primary tumor and metastasis for all parameters, the strongest relationship being observed for p53 protein (Table 1).

View larger version (20K): [in this window] [in a new window]   Fig 1. Distribution of p53 protein concentrations according to p53 mutation status, both measured in metastases. Boxes delimit the first and third quartiles (median inside); bars represent the range of values falling within 1.5-fold the interquartile range. Triangles represent outliers (> 1.5-fold the interquartile range).

TS promoter polymorphism was investigated in 88 liver metastases and 50 primary tumors. Distribution of TS genotype was 20.0% 2R/2R, 40.0% 2R/3R, and 40.0% 3R/3R in primary tumors and 23.9% 2R/2R, 43.2% 2R/3R, and 33.0% 3R/3R in metastases. The above distributions were in close agreement with those predicted by the Hardy-Weinberg equilibrium. In patients with TS polymorphism analyzed both in primary tumor and metastasis, TS genotype was superimposable between primary tumor and derived metastasis. Allelic loss was observed in 50% of heterozygous primary tumor biopsy specimens and in 47% of heterozygous liver metastases. Each allele was lost at a similar rate. Analysis of allelic imbalances revealed discordance between primary tumor and metastasis for three patients. TS genotype was finally considered after correction for allelic loss (ie, 2R/3R were classified as 2R/2R or 3R/3R, depending on the allele lost). Distribution of TS genotype after correction for the allelic loss is listed in Table 3.

View larger version (17K): [in this window] [in a new window]   Fig 2. Distribution of TS enzymatic activity according to TS promoter polymorphism in primary tumors (A) and metastases (B). Boxes delimit the first and third quartiles (median inside); bars represent the range of values falling within 1.5-fold the interquartile range. Triangles represent outliers (> 1.5-fold the interquartile range).

Analysis of Survival
Table 6 lists univariate Cox analyses for potential predictors of specific survival. Characteristics of metastases: multifocal versus single (relative risk, 2.79; P = .017) and synchronous versus metachronous (relative risk, 2.23; P = .033) were significantly related to survival, as well as patient age. TS activity measured in primary tumor was a significant survival predictor: the greater the TS activity, the poorer the survival (P = .040). Interestingly, TS promoter polymorphism in primary tumor carries a similar prognostic value (P = .035, Table 6), with heterozygous primary tumors having the poorer prognosis (median survivals were 10.3 months for 2R/3R, 14.0 months for 3R/3R, and 19.0 months for 2R/2R, log rank P = .025, Fig 3). In fact, a bivariate stepwise Cox analysis revealed that TS polymorphism was selected from a forward analysis, whereas TS activity was the selected variable in a backward analysis. When considered as categorical (lower v greater than median value), none of the quantitative variables were significant predictors of survival.

View larger version (16K): [in this window] [in a new window]   Fig 3. Plot of cumulative specific survival according to TS genotype in primary tumors. Solid line, 2R/3R (n = 10, 10 events); dashed line, 2R/2R (n = 16, 16 events); dotted line, 3R/3R (n = 24, 20 events). P = .025, log-rank test.

View larger version (16K): [in this window] [in a new window]   Fig 4. Plot of cumulative specific survival according to p53 mutation status in metastases. Solid line, stop mutations (n = 6, 6 events); dashed line, p53 wild type (n = 42, 34 events); dotted line, nonstop mutations (n = 49, 40 events). P = .011, log rank test.

Last, because TS activity in primary tumor was linked to clinical responsiveness, it was relevant to know whether TS prognostic value was independent from clinical response or not. In fact, survival of colorectal cancer patients with liver metastases is known to be closely related to clinical response on metastasis, as presently confirmed (univariate Cox regression for St + Pg v CR + PR gave P < .001, with a relative risk at 2.47). A multivariate analysis taking into account clinical parameters related to survival (metastasis characteristics and patient age), along with clinical response, confirmed the prognostic value of TS activity (P = .045) independently of clinical responsiveness.

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
This prospective study was performed on the largest homogenous population of liver metastatic colorectal cancer patients (n = 103) published to date. All patients received first line FU–folinic acid chemotherapy, and none of them underwent liver metastasis resection. Importantly, the pharmacologic parameters potentially related to FU sensitivity (TS, DPD, FPGS, p53) were investigated in the treatment-targeted tissue (liver metastasis) for all patients, as well as in primary tumor when possible (metastasis synchronous with primary cancer). Based on the observed variability of tumoral parameters, calculation of the statistical power to detect differences between responders and nonresponders indicated that adequate statistical power was reached for DPD, FPGS, and TS activities, whereas a weaker power was detected for p53 protein (data not shown).

Contrary to a previous study by our group on head and neck cancer patients receiving FU-based therapy,³⁵ the present study does not demonstrate a significant role of tumoral DPD activity on clinical response (Tables 4 and 5) or on survival (Table 6). This result contrasts with that of a study⁸ that reported that DPD gene expression in colorectal cancer was a significant predictor of FU responsiveness. Also, present data (Tables 4 and 5) do not confirm the previously reported value of FPGS activity for predicting FU responsiveness in colorectal cancer patients.²⁵ This latter observation could be accounted for by the fact that patients received folinic acid, thus lessening the role of FPGS. From the present study, DPD and FPGS evaluated as enzymatic activities do not play a major role for predicting FU responsiveness or survival of colorectal cancer patients.

Among the pharmacologic factors related to fluoropyrimidine sensitivity, TS expression or activity has been the most thoroughly investigated parameter over the last decade.^8-17 Few data have been reported on the clinical interest of TS promoter polymorphism.^18,36 TS activity was presently investigated, along with TS polymorphism, in primary tumor and metastasis tissue samples. TS promoter genotypes were concordant between primary tumors and derived liver metastases, with the exception of different allelic imbalances in three patients: two patients exhibited an allelic loss in primary tumor not present in the corresponding metastasis, and one patient demonstrated different allelic losses in primary tumor and metastasis. These discrepancies could be explained by different contamination rates from normal cells in primary tumors and metastases. The latter case (allelic loss different in primary tumor and metastasis) could be explained by early metastatic dissemination occurring before allelic loss in the primary tumor. Present data clearly demonstrate that TS activity was higher in heterozygous as compared with homozygous patients, both in primary tumor and metastases (Table 3 and Fig 2). This result contrasts with the data of Horie et al⁶ and Kawakami et al⁷ and suggests the existence of additional posttranscriptional regulatory pathways.

TS activity measured in metastasis, and not TS promoter polymorphism, was the only significant predictor of treatment efficacy, with responding patients exhibiting lower TS activity than nonresponding patients (Tables 4 and 5). This result contrasts with the findings by Villafranca et al¹⁸ and McLeod et al³⁶ suggesting that TS genotype may be related to FU-based responsiveness. Numerous experimental and clinical studies have demonstrated that elevated tumoral TS is related to FU resistance.^8-16 Even though TS promoter polymorphism may modulate the basal level of TS protein,^5-7 present results clearly demonstrate that TS genotype cannot serve as a substitute for TS activity in order to predict FU responsiveness. Besides, although TS activity in metastasis is correlated to TS activity in primary tumor, the latter is not predictive of FU efficacy on liver metastasis, in line with data from Aschele et al.³⁷ Considering the tumoral evolution from initial tumor to distant metastasis, it is not surprising that measurement at the target level is the most relevant approach.

Regarding the influence of TS on specific survival, both TS activity and TS genotype measured in primary tumor were significant survival predictors (Table 6). Heterozygous patients, exhibiting the highest TS activities, had the poorest prognosis (Fig 3). In order to select the strongest factor between TS activity and TS genotype, we performed a bivariate stepwise analysis (backward and forward), which did not allow one of these two factors to be discriminated, suggesting that TS activity and TS genotype carried similar prognostic values. Interestingly, TS activity (P = .061) and TS genotype (P = .070) remained close to significance after stratification on metastasis characteristics. Of note is the fact that TS activity (P = .045) and not TS genotype (P = .14, data not shown) remained a significant survival predictor when combined with clinical responsiveness. The independence of TS activity from clinical responsiveness for predicting survival suggests a dual role of TS, having a prognostic value linked to its intrinsic tumor aggressiveness potential, as well as a role toward clinical responsiveness likely linked to the pharmacologic interaction between TS and the FU anabolite fluorodeoxyuridine monophosphate.

By contrast, when measured in metastasis, TS activity and TS genotype were not related to patient survival (Table 6). Regarding TS genotype, this discrepancy is explained by discordant allelic imbalances in the three patients with a different allele loss between primary tumor and metastasis: analyses performed in the subset of 44 patients with TS genotype evaluated both in primary tumor and metastasis (data not shown) confirmed that TS genotype remained significant in primary tumor (P = .051) and nonsignificant in metastasis (P = .99).

TS activity was two-fold lower in p53 wild-type metastases than in p53 mutated metastases, corroborating clinical data from Lenz et al¹² that was based on reverse transcriptase–PCR mRNA quantification. Such observations are in line with a study by Lee et al³⁸ suggesting in a mouse model that p53 wild-type protein could downregulate TS. Conversely, Chu et al³⁹ reported a specific interaction between the TS protein and the protein-coding region of p53 mRNA and drew the conclusion that TS could downregulate the expression of p53 at the translational level. In this context, we looked at a possible negative correlation between TS activity and p53 protein in the subgroup of p53 wild-type patients and did not observe any significant relationship (data not shown).

p53 mutations were analyzed in metastases and the concentration of the p53 protein was determined both in metastases and primary tumor. p53 protein concentrations, reflecting the protein cellular retention, were significantly higher in p53-mutated metastases as compared with wild type (approximately seven-fold, Fig 1), in line with the literature.⁴⁰ In concordance with the data by Belluco et al,¹⁰ we observed a significant positive correlation between p53 protein measured in primary tumors and derived metastases (r² = .28, P < .001, Table 1). p53 mutations were observed in 56.7% of metastatic tissue samples (Table 2). p53 protein concentrations (measured in metastasis or in primary tumor) were not significantly different in responding and nonresponding patients (Table 4). Accordingly, response rates were similar between p53 mutated and p53 wild-type tumors (36.4% v 36.6%). The value of p53 for predicting FU responsiveness is conflicting, as emphasized in the review by Ferreira et al.²⁹

In addition, neither p53 protein nor p53 mutations considered globally were significant survival predictors (Table 6). A new finding arising from the present study is that patients with mutations leading to a putative truncated p53 protein (stop mutations) exhibited a dramatically reduced survival, with a relative risk of death of 3.14 as compared with p53 wild-type patients (Table 6 and Fig 4). Moreover, survival of patients with nonstop mutations was comparable to that of p53 wild-type patients (Fig 4). A multivariate analysis taking into account the clinical parameters related to survival demonstrated that p53 stop mutation status remained close to significance (P = .059). However, these results must be considered carefully because stop mutations accounted for a small number of patients (14.5% of mutated samples). The clinical relevance of p53 stop mutations for predicting patient outcome thus merits further confirmation on a larger set of patients. The present observation concords well with data⁴¹ demonstrating that a large number of p53 mutations do not promote disease progression in colorectal cancer and demonstrating that some mutations, particularly in exons 5 to 9, may even counteract negative effects of other p53 structural alterations.

In conclusion, the present study clearly confirms the value of TS activity measured in the treatment-targeted tissue, and not TS genotype, for predicting FU responsiveness in patients with advanced colorectal cancer. Present data suggest that TS activity as well as TS genotype, both measured in primary tumor, are prognostic factors close to significance, independent of classical prognostic factors (metastasis characteristics and patient age). In addition, this study suggests that p53 stop mutation analysis might represent a promising approach for predicting colorectal cancer patient outcome. Considering the small number of patients with p53 stop mutations, along with the fact that a limited subset of patients (18 of 103) received second-line chemotherapy, the above results on prognostic factors must be given careful consideration and deserve further confirmation.

	ACKNOWLEDGMENTS

 
Supported in party by grants from the French Research Ministry (Programme Hospitalier de Recherche Clinique 1996).

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. Saltz L: Irinotecan-based combinations for the adjuvant treatment of stage III colon cancer. Oncology 14: 47-50, 2000 (suppl 14)

2. Pelley RJ: Oxaliplatin: A new agent for colorectal cancer. Curr Oncol Rep 3: 147-155, 2001[Medline]

3. Grem JL: 5-Fluorouracil: Forty-plus and still ticking—A review of its preclinical and clinical development. Invest New Drugs 18: 299-313, 2000[CrossRef][Medline]

4. Budman DR: Capecitabine. Invest New Drugs 18: 355-363, 2000[CrossRef][Medline]

5. Kanada S, Takeishi K, Ayusawa D, et al: Role in translation of a triple tandemly repeated sequence in the 5'-untranslated region of human thymidylate synthase mRNA. Nucleic Acids Res 15: 1259-1270, 1987[Abstract/Free Full Text]

6. Horie N, Aiba H, Oguro K, et al: Functional analysis and DNA polymorphism of the tandemly repeated sequences in the 5'-terminal regulatory region of the human gene for thymidylate synthase. Cell Struct Funct 20: 191-197, 1995[Medline]

7. Kawakami K, Salonga D, Park JM, et al: Different lengths of a polymorphic repeat sequence in the thymidylate synthase gene affect translational efficiency but not its gene expression. Clin Cancer Res 7: 4096-4101, 2001[Abstract/Free Full Text]

8. Salonga D, Danenberg KD, Jonhson M, et al: Colorectal tumors responding to 5-fluorouracil have low gene expression levels of dihydropyrimidine dehydrogenase, thymidylate synthase and thymidine phosphorylase. Clin Cancer Res 6: 1322-1327, 2000[Abstract/Free Full Text]

9. Johnston PG, Lenz HJ, Leichman CG, et al: Thymidylate synthase gene and protein expression correlate and are associated with response to 5-fluorouracil in human colorectal and gastric tumors. Cancer Res 55: 1407-1412, 1995[Abstract/Free Full Text]

10. Belluco C, Guillem JG, Kemeny N, et al: p53 nuclear protein overexpression in colorectal cancer: A dominant predictor of survival in patients with advanced hepatic metastases. J Clin Oncol 14: 2696-2701, 1996[Abstract/Free Full Text]

11. Leichman CG, Lenz HJ, Leichman L, et al: Quantification of intratumoral thymidylate synthase expression predicts for disseminated colorectal cancer response and resistance to protracted-infusion fluorouracil and weekly leucovorin. J Clin Oncol 15: 3223-3229, 1997[Abstract]

12. Lenz HJ, Hayashi K, Salonga D, et al: p53 mutations and thymidylate synthase messenger RNA levels in disseminated colorectal cancer: An analysis of response and survival. Clin Cancer Res 4: 1243-1250, 1998[Abstract]

13. Paradiso A, Simone G, Petroni S, et al: Thymidylate synthase and p53 primary tumor expression as predictive factors for advanced colorectal cancer patients. Br J Cancer 82: 560-567, 2000[CrossRef][Medline]

14. Shirota Y, Stoehlmacher J, Brabender J, et al: ERCC1 and thymidylate synthase mRNA levels predict survival for colorectal cancer patients receiving combination oxaliplatin and fluorouracil chemotherapy. J Clin Oncol 19: 4298-4304, 2001[Abstract/Free Full Text]

15. Peters GJ, Van der Wilt CL, Van Triest B, et al: Thymidylate synthase and drug resistance. Eur J Cancer 31: 1299-1305, 1995

16. Peters GJ, Van der Wilt CL, Van Groeninger CJ, et al: Thymidylate synthase inhibitor after administration of fluorouracil with or without leucovorin in colon cancer patients: Implications for treatment with fluorouracil. J Clin Oncol 12: 2035-2042, 1994[Abstract/Free Full Text]

17. Findlay MPN, Cunningham D, Morgan G, et al: Lack of correlation between thymidylate synthase levels in primary colorectal tumors and subsequent response to chemotherapy. Br J Cancer 75: 903-909, 1997[Medline]

18. Villafranca E, Okrushnov Y, Dominguez MA, et al: Polymorphisms of the repeated sequences in the enhancer region of the thymidylate synthase gene promoter may predict downstaging after preoperative chemoradiation in rectal cancer. J Clin Oncol 19: 1779-1786, 2001[Abstract/Free Full Text]

19. Milano G, McLeod HL: Can dihydropyrimidine dehydrogenase impact 5-fluorouracil–based treatment? Eur J Cancer 36: 37-42, 2000[CrossRef][Medline]

20. Beck A, Etienne MC, Cheradame S, et al: A role for dihydropyrimidine dehydrogenase and thymidylate synthase in tumor sensitivity to fluorouracil. Eur J Cancer 30: 1517-1522, 1994

21. Ishikawa Y, Kubota T, Otami Y, et al: Dihydropyrimidine dehydrogenase activity and messenger RNA level may be related to the antitumor effect of 5-fluorouracil on human tumor xenografts in nude mice. Clin Cancer Res 5: 883-889, 1999[Abstract/Free Full Text]

22. Keyomarsi K, Moran RG: Mechanism of the cytotoxic synergism of fluoropyrimidines and folinic acid in mouse leukemic cells. J Biol Chem 28: 14402-14409, 1998

23. Boarman DM, Allegra CJ: Intracellular metabolism of 5-formyltetrahydrofolate in human breast and colon cell lines. Cancer Res 52: 36-44, 1992[Abstract/Free Full Text]

24. Cheradame S, Etienne MC, Chazal M, et al: Relevance of tumoral folylpolyglutamate synthetase and reduced folates for optimal 5-fluorouracil efficacy: Experimental data. Eur J Cancer 33: 950-959, 1997[CrossRef][Medline]

25. Chazal M, Cheradame S, Formento JL, et al: Decreased folylpolyglutamate synthetase activity in tumors resistant to fluorouracil–folinic acid treatment: Clinical data. Clin Cancer Res 3: 553-557, 1997[Abstract]

26. Tillman DM, Petak I, Houghton JA: A Fas-dependent component in 5-fluorouracil/leucovorin–induced cytotoxicity in colon carcinoma cells. Clin Cancer Res 5: 425-430, 1999[Abstract/Free Full Text]

27. Fichhorst ST, Muerkoster S, Weigand MA, et al: The chemotherapeutic drug 5-fluorouracil induces apoptosis in mouse thymocytes in vivo via activation of the CD95 (APO-1/Fas) system. Cancer Res 61: 243-248, 2001[Abstract/Free Full Text]

28. Maecker HL, Koumenis C, Giaccia AJ: p53 promotes selection for Fas-medicated apoptotic resistance. Cancer Res 60: 4638-4644, 2000[Abstract/Free Full Text]

29. Ferreira CG, Tolis G, Giaccone G: p53 and chemosensitivity. Ann Oncol 10: 1011-1021, 1999[Abstract/Free Full Text]

30. World Health Organization: Handbook for Reporting Results of Cancer Treatment. Publication no. 48. Geneva, Switzerland, World Health Organization, 1979

31. Spears CP, Gustavsson BG: Methods for thymidylate synthase pharmacodynamics: Serial biopsy, free and total TS, FdUMP and dUMP, and H4Pteglu and CH2-H4Pteglu assays, in Rustum Y, McGuire JJ (eds): The Expanding Role of Folates and Fluoropyrimidines in Cancer Chemotherapy. New York, NY, Plenum Press, 1988, pp 97-104

32. Montero C, Llorente P: Single step LC method for determination of folylpolyglutamate synthetase activity and separation of polyglutamates. Chromatography 37: 73-78, 1993[CrossRef]

33. Harris BE, Song R, Soong S et al: Relationship between dihydropyrimidine dehydrogenase activity and plasma 5-fluorouracil levels with evidence for circadian variation of enzyme activity and plasma drug levels in cancer patients receiving 5-fluorouracil by protracted continuous infusion. Cancer Res 50: 197-201, 1990[Abstract/Free Full Text]

34. Rosty C, Chazal M, Etienne MC, et al: Determination of microsatellite instability, p53 and K-RAS mutations in hepatic metastases from patients with colorectal cancer: Relationship with response to 5-fluorouracil and survival. Int J Cancer 95: 162-167, 2001[CrossRef][Medline]

35. Etienne MC, Cheradame S, Fischel JL, et al: Response to fluorouracil therapy in cancer patients: The role of tumoral dihydropyrimidine dehydrogenase activity. J Clin Oncol 13: 1663-1670, 1995[Abstract/Free Full Text]

36. McLeod HL, Marsh S, McKay JA, et al: Thymidylate synthase pharmacogenetics in colorectal cancer. Proc Am Soc Clin Oncol 20: 129a, 2001 (abstr 512)

37. Aschele C, Debernardis D, Tunesi G, et al: Thymidylate synthase protein expression in primary colorectal cancer compared with the corresponding distant metastases and relationship with the clinical response to 5-fluorouracil. Clin Cancer Res 6: 4797-4802, 2000[Abstract/Free Full Text]

38. Lee Y, Chen Y, Chang LS, et al: Inhibition of mouse thymidylate synthase promoter activity by the wild type p53 tumor suppressor protein. Exp Cell Res 234: 270-276, 1997[CrossRef][Medline]

39. Chu E, Copur SM, Ju J, et al: Thymidylate synthase protein and p53 mRNA form an in vivo ribonucleoprotein complex. Mol Cell Biol 19: 1582-1594, 1999[Abstract/Free Full Text]

40. Raybaud-Diogène H, Tetu B, Morenay R, et al: p53 overexpression in head and neck squamous cell carcinoma: Review of the literature. Oral Oncol 32B: 143-149, 1996

41. Forslund A, Lönnroth C, Andersson M, et al: Mutations and allelic loss of p53 in primary tumor DNA from potentially cured patients with colorectal carcinoma. J Clin Oncol 19: 2829-2836, 2001[Abstract/Free Full Text]

Submitted September 18, 2001; accepted March 21, 2002.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				G. Lurje, P. C. Manegold, Y. Ning, A. Pohl, W. Zhang, and H.-J. Lenz Thymidylate synthase gene variations: predictive and prognostic markers Mol. Cancer Ther., May 1, 2009; 8(5): 1000 - 1007. [Abstract] [Full Text] [PDF]

				M.-C. Etienne-Grimaldi, J.-L. Formento, M. Francoual, E. Francois, P. Formento, N. Renee, P. Laurent-Puig, M. Chazal, D. Benchimol, J.-R. Delpero, et al. K-Ras Mutations and Treatment Outcome in Colorectal Cancer Patients Receiving Exclusive Fluoropyrimidine Therapy Clin. Cancer Res., August 1, 2008; 14(15): 4830 - 4835. [Abstract] [Full Text] [PDF]

				R. Soong, N. Shah, M. Salto-Tellez, B. C. Tai, R. A. Soo, H. C. Han, S. S. Ng, W. L. Tan, N. Zeps, D. Joseph, et al. Prognostic significance of thymidylate synthase, dihydropyrimidine dehydrogenase and thymidine phosphorylase protein expression in colorectal cancer patients treated with or without 5-fluorouracil-based chemotherapy Ann. Onc., May 1, 2008; 19(5): 915 - 919. [Abstract] [Full Text] [PDF]

				D. G. Mollevi, T. Serrano, M. M. Ginesta, J. Valls, J. Torras, M. Navarro, E. Ramos, J. R. Germa, E. Jaurrieta, V. Moreno, et al. Mutations in TP53 are a prognostic factor in colorectal hepatic metastases undergoing surgical resection Carcinogenesis, June 1, 2007; 28(6): 1241 - 1246. [Abstract] [Full Text] [PDF]

				J. R. Brody, T. Hucl, E. Gallmeier, J. M. Winter, S. E. Kern, and K. M. Murphy Genomic Copy Number Changes Affecting the Thymidylate Synthase (TYMS) Gene in Cancer: A Model for Patient Classification to Aid Fluoropyrimidine Therapy Cancer Res., October 1, 2006; 66(19): 9369 - 9373. [Abstract] [Full Text] [PDF]

				R. Largillier, M.-C. Etienne-Grimaldi, J.-L. Formento, J. Ciccolini, J.-F. Nebbia, A. Ginot, M. Francoual, N. Renee, J.-M. Ferrero, C. Foa, et al. Pharmacogenetics of capecitabine in advanced breast cancer patients. Clin. Cancer Res., September 15, 2006; 12(18): 5496 - 5502. [Abstract] [Full Text] [PDF]

				P. A. Wood, J. Du-Quiton, S. You, and W. J.M. Hrushesky Circadian clock coordinates cancer cell cycle progression, thymidylate synthase, and 5-fluorouracil therapeutic index. Mol. Cancer Ther., August 1, 2006; 5(8): 2023 - 2033. [Abstract] [Full Text] [PDF]

				E. Dotor, M. Cuatrecases, M. Martinez-Iniesta, M. Navarro, F. Vilardell, E. Guino, L. Pareja, A. Figueras, D. G. Mollevi, T. Serrano, et al. Tumor Thymidylate Synthase 1494del6 Genotype As a Prognostic Factor in Colorectal Cancer Patients Receiving Fluorouracil-Based Adjuvant Treatment J. Clin. Oncol., April 1, 2006; 24(10): 1603 - 1611. [Abstract] [Full Text] [PDF]

				K. P. de Jong, A. S.H. Gouw, P. M.J.G. Peeters, M. Bulthuis, L. Menkema, R. J. Porte, M. J.H. Slooff, H. van Goor, and A. van den Berg P53 Mutation Analysis of Colorectal Liver Metastases: Relation to Actual Survival, Angiogenic Status, and p53 Overexpression Clin. Cancer Res., June 1, 2005; 11(11): 4067 - 4073. [Abstract] [Full Text] [PDF]

				M. A. Wright, G. Morrison, P. Lin, G. D. Leonard, D. Nguyen, X. Guo, E. Szabo, J. L. Hopkins, J. P. Leguizamo, N. Harold, et al. A Phase I Pharmacologic and Pharmacogenetic Trial of Sequential 24-Hour Infusion of Irinotecan Followed by Leucovorin and a 48-Hour Infusion of Fluorouracil in Adult Patients with Solid Tumors Clin. Cancer Res., June 1, 2005; 11(11): 4144 - 4150. [Abstract] [Full Text] [PDF]

				S. Popat, R. Wort, and R. S. Houlston Relationship Between Thymidylate Synthase (TS) Genotype and TS Expression: A Tissue Microarray Analysis of Colorectal Cancers International Journal of Surgical Pathology, April 1, 2005; 13(2): 127 - 133. [Abstract] [PDF]

				D. J. Sargent, B. A. Conley, C. Allegra, and L. Collette Clinical Trial Designs for Predictive Marker Validation in Cancer Treatment Trials J. Clin. Oncol., March 20, 2005; 23(9): 2020 - 2027. [Abstract] [Full Text] [PDF]

				A. Jakobsen, J. N. Nielsen, N. Gyldenkerne, and J. Lindeberg Thymidylate Synthase and Methylenetetrahydrofolate Reductase Gene Polymorphism in Normal Tissue As Predictors of Fluorouracil Sensitivity J. Clin. Oncol., March 1, 2005; 23(7): 1365 - 1369. [Abstract] [Full Text] [PDF]

				J. I. Geller, K. Szekely-Szucs, I. Petak, B. Doyle, and J. A. Houghton P21Cip1 Is a Critical Mediator of the Cytotoxic Action of Thymidylate Synthase Inhibitors in Colorectal Carcinoma Cells Cancer Res., September 1, 2004; 64(17): 6296 - 6303. [Abstract] [Full Text] [PDF]

				T. Lecomte, J.-M. Ferraz, F. Zinzindohoue, M.-A. Loriot, D.-A. Tregouet, B. Landi, A. Berger, P.-H. Cugnenc, R. Jian, P. Beaune, et al. Thymidylate Synthase Gene Polymorphism Predicts Toxicity in Colorectal Cancer Patients Receiving 5-Fluorouracil-based Chemotherapy Clin. Cancer Res., September 1, 2004; 10(17): 5880 - 5888. [Abstract] [Full Text] [PDF]

				R. Chan and D. J. Kerr Can we individualise chemotherapy for colorectal cancer? Ann. Onc., July 1, 2004; 15(7): 996 - 999. [Full Text] [PDF]

				L. Sharp and J. Little Polymorphisms in Genes Involved in Folate Metabolism and Colorectal Neoplasia: A HuGE Review Am. J. Epidemiol., March 1, 2004; 159(5): 423 - 443. [Abstract] [Full Text] [PDF]

				S. Popat, A. Matakidou, and R. S. Houlston Thymidylate Synthase Expression and Prognosis in Colorectal Cancer: A Systematic Review and Meta-Analysis J. Clin. Oncol., February 1, 2004; 22(3): 529 - 536. [Abstract] [Full Text] [PDF]

				K. Uchida, K. Hayashi, K. Kawakami, S. Schneider, J. M. Yochim, H. Kuramochi, K. Takasaki, K. D. Danenberg, and P. V. Danenberg Loss of Heterozygosity at the Thymidylate Synthase (TS) Locus on Chromosome 18 Affects Tumor Response and Survival in Individuals Heterozygous for a 28-bp Polymorphism in the TS Gene Clin. Cancer Res., January 15, 2004; 10(2): 433 - 439. [Abstract] [Full Text] [PDF]

				J. Little, L. Sharp, S. Duthie, and S. Narayanan Colon Cancer and Genetic Variation in Folate Metabolism: The Clinical Bottom Line J. Nutr., November 1, 2003; 133(11): 3758S - 3766. [Abstract] [Full Text] [PDF]

				K. Kawakami and G. Watanabe Identification and Functional Analysis of Single Nucleotide Polymorphism in the Tandem Repeat Sequence of Thymidylate Synthase Gene Cancer Res., September 15, 2003; 63(18): 6004 - 6007. [Abstract] [Full Text] [PDF]

				J. R. Bertino and D. Banerjee Is the Measurement of Thymidylate Synthase to Determine Suitability for Treatment with 5-Fluoropyrimidines Ready for Prime Time? Clin. Cancer Res., April 1, 2003; 9(4): 1235 - 1239. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Etienne, M.-C. Articles by Milano, G. Search for Related Content PubMed PubMed Citation Articles by Etienne, M.-C. Articles by Milano, G. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Medline Plus Health Information Colorectal Cancer Hazardous Substances DB FLUOROURACIL LEUCOVORIN Social Bookmarking                            What's this?