Search

Introduction: Recurrence after liver surgery or radiofrequency is a clinical and biological challenge because it worsens the colorectal cancer prognosis. To date, no biomarker is yet validated to predict the recurrence in order to intensify adjuvant therapy for patients with higher risk. Matrix metalloproteinases play a major role in the metastasis dissemination and tumoral microenvironment and could be a potential biomarker of interest., Methods: Forty-four patients with liver metastasis treated by surgery or radiofrequency were enrolled in this study. Serum levels of MMP-1, MMP-2, MMP-7, MMP-9 and TIMP-1 were monitored in Elisa after therapy and correlated to the recurrence from January 2004 to December 2007. After the curative treatment, patients were assessed for the recurence for two years by CT-scan and examination., Results: Post-operative serum level of MMP-9 was significantly higher between J0, J1 and J45 after liver surgery or radiofrequency (***P≤0.001). Level of MMP-2 was significantly increased at M3 and M6 (***P≤0.001) but does not appear to be a risk factor of liver recurrence. The level of TIMP-1 at J0 is a deleterious factor (HR=1.76, P=0.042*)., Conclusion: This is the first study wich correlates the post-operative level of 4 MMPs and TIMP-1 with the risk of liver recurrence after surgery or radiofrequency. Serum TIMP-1 level at J0 could be helpful to identify patients with higher risk but these results need to be confirmed in a large-scale study., (Copyright © 2018 Société Française du Cancer. Published by Elsevier Masson SAS. All rights reserved.)

Background: Although the number of colorectal liver metastases (CLM) is decreasingly considered as a contraindication to surgery, patients with 10 CLM or more are often denied liver surgery. This study aimed to evaluate the outcome after liver surgery and to identify prognostic factors of survival in such patients., Methods: The study population consisted of a multicentre cohort of patients with CLM (N=12 406) operated on, with intention to resect, from January 2005-June 2013 and whose data were prospectively collected in the LiverMetSurvey registry., Results: Overall, the group ⩾10 CLM (N=529, 4.3%) experienced a 5-year overall survival (OS) of 30%. A macroscopically complete (R0/R1) resection (72.8% of patients) was associated with a 3- and 5-year OS of 61% and 39% vs 29% and 5% for R2/no resection patients (P<0.0001). At multivariate analysis, R0/R1 resection emerged as the strongest favourable factor of OS (HR 0.35 (0.26-0.48)). Other independent favourable factors were as follows: maximal tumour size <40 mm (HR 0.67 (0.49-0.92)); age <60 years (HR 0.66 (0.50-0.88)); preoperative MRI (HR 0.65 (0.47-0.89)); and adjuvant chemotherapy (HR 0.73 (0.55-0.98)). The model showed that 5-year OS rates of 30% was possible provided R0/R1 resection associated with at least an additional favourable factor., Conclusions: Liver resection might provide long-term survival in patients with ⩾10 CLM staged with preoperative MRI, provided R0/R1 resection followed by adjuvant therapy. A validation of these results in another cohort is needed.

Dihydropyrimidine dehydrogenase is a crucial enzyme for the degradation of 5-fluorouracil (5FU). DPYD, which encodes dihydropyrimidine dehydrogenase, is prone to acquire genomic rearrangements because of the presence of an intragenic fragile site FRA1E. We evaluated DPYD copy number variations (CNVs) in a prospective series of 242 stage I-III colorectal tumours (including 87 patients receiving 5FU-based treatment). CNVs in one or more exons of DPYD were detected in 27% of tumours (deletions or amplifications of one or more DPYD exons observed in 17% and 10% of cases, respectively). A significant relationship was observed between the DPYD intragenic rearrangement status and dihydropyrimidine dehydrogenase (DPD) mRNA levels (both at the tumour level). The presence of somatic DPYD aberrations was not associated with known prognostic or predictive biomarkers, except for LOH of chromosome 8p. No association was observed between DPYD aberrations and patient survival, suggesting that assessment of somatic DPYD intragenic rearrangement status is not a powerful biomarker to predict the outcome of 5FU-based chemotherapy in patients with colorectal cancer.

Background: To test the prognostic value of tumour protein and genetic markers in colorectal cancer (CRC) and examine whether deficient mismatch repair (dMMR) tumours had a distinct profile relative to proficient mismatch repair (pMMR) tumours., Methods: This prospective multicentric study involved 251 stage I-III CRC patients. Analysed biomarkers were EGFR (binding assay), VEGFA, thymidylate synthase (TS), thymidine phosphorylase (TP) and dihydropyrimidine dehydrogenase (DPD) expressions, MMR status, mutations of KRAS (codons 12-13), BRAF (V600E), PIK3CA (exons 9 and 20), APC (exon 15) and P53 (exons 4-9), CpG island methylation phenotype status, ploidy, S-phase, LOH., Results: The only significant predictor of relapse-free survival (RFS) was tumour staging. Analyses restricted to stage III showed a trend towards a shorter RFS in KRAS-mutated (P=0.005), BRAF wt (P=0.009) and pMMR tumours (P=0.036). Deficient mismatch repair tumours significantly demonstrated higher TS (median 3.1 vs 1.4) and TP (median 5.8 vs 3.5) expression relative to pMMR (P<0.001) and show higher DPD expression (median 14.9 vs 7.9, P=0.027) and EGFR content (median 69 vs 38, P=0.037) relative to pMMR., Conclusions: Present data suggesting that both TS and DPD are overexpressed in dMMR tumours as compared with pMMR tumours provide a strong rationale that may explain the resistance of dMMR tumours to 5FU-based therapy.

Purpose: This study was designed to evaluate neoadjuvant intensified chemotherapy in potentially resectable or unresectable liver metastases (LM) from colorectal cancer (CRC)., Methods: Criteria for potentially resectable LM were complex hepatectomy and/or risky procedure, close contact with major vascular structures, and for unresectable LM, a future liver remnant predicted to be less than 25-30 % of total liver volume. Between October 2004 and August 2007, 125 patients were randomized to either standard (FOLFIRI/FOLFOX4) or intensified chemotherapy (FOLFIRI-HD/FOLFOX7/FOLFIRINOX). Primary endpoint was objective response rate (ORR) after 4 cycles of chemotherapy. Secondary endpoints included safety, R0 surgical resection, best ORR, progression-free survival (PFS), and overall survival (OS)., Results: A total of 122 patients were treated; 45 % of patients had less than 30 % of remaining liver tissue, 20 % had major vascular contact, and 35 % were potentially resectable. Grade 3/4 toxicities were neutropenia (24, 19, 10, 23 %) diarrhoea (0, 6, 3, 23 %), mucositis (0, 3, 0, 7 %), vomiting (7, 9, 0, 3 %), and neurotoxicity (0, 0, 10, 3 %) in arms (FOLFIRI + FOLFOX4)/FOLFIRI-HD/FOLFOX7/FOLFIRINOX, respectively. ORR was 33, 47, 43, and 57 % after the first 4 cycles in arms (FOLFIRI + FOLFOX4)/FOLFIRI-HD/FOLFOX7/FOLFIRINOX, respectively. FOLFIRINOX offered the best conversion rate to resectability (67 %). Disease-free status after chemotherapy and surgery (R0, R1, Rx) was achieved in 54 of 64 operated patients. Median PFS was 9.2 months in control arms versus 11.9 months in experimental arms (hazards ratio [HR] = 0.76, p = 0.115), and the median OS was 17.7 versus 33.4 months (HR = 0.73, p = 0.297), respectively., Conclusions: FOLFIRINOX showed promising activity in CRC patients with LM compared with standard or intensified bi-chemotherapy regimens.

Purpose: Tumor progression while receiving neoadjuvant chemotherapy (PD) has been associated with poor outcome and is commonly considered a contraindication to liver resection (LR). This study aims to clarify in a large multicenter setting whether PD is always a contraindication to LR., Methods: Data from the LiverMetSurvey international registry were analyzed. Patients undergoing LR for colorectal metastases without extrahepatic disease after neoadjuvant chemotherapy between 1990 and 2009 were reviewed., Results: Among 2143 patients, PD occurred in 176 (8.2 %). Risk of progression was increased after 5-FU or irinotecan (22.7 % vs. 6.8 % after other regimens, p < 0.0001; 14.9 % vs. 7.2 %, p < 0.0001), while it was reduced after oxaliplatin (5.6 % vs. 12.0 %, p < 0.0001) and still diminished among patients receiving targeted therapies (2.6 %). PD was an independent prognostic factor of survival at multivariate analysis (35 % vs. 49 %, p = 0.0006). In the PD group, 3 independent prognostic factors were identified: carcinoembryonic antigen (CEA) ≥ 200 ng/mL (p = 0.003), >3 metastases (p = 0.028), and tumor diameter ≥ 5 0 mm (p = 0.002). A survival predictive model showed that patients without any risk factors had 5-year survival rates of 53.3 %; good survival results were still observed if metastases were >3 or ≥ 50 mm (29.9 and 19.1 %, respectively). On the contrary, survival was less than 10 % at 3 years in the presence of >1 prognostic factor or CEA of ≥ 200 ng/mL., Conclusions: PD is a negative prognostic factor, but it is not an absolute contraindication to LR. Patients with PD could be scheduled for LR except for those with >3 metastases and ≥ 50 mm, or CEA ≥ 200 ng/mL in whom further chemotherapy is recommended.

Background: In advanced colorectal cancer, K-Ras somatic mutations predict resistance to mAbs targeting epidermal growth factor receptor (EGFR). Relationships between K-Ras mutations and EGFR status have not been examined so far. We analyzed relationships between K-Ras mutations and EGFR tumoral status based on EGFR germinal polymorphisms, gene copy number and expression., Methods: Eighty colorectal tumors (stage 0-IV) and 39 normal mucosas were analyzed. K-Ras mutations at codons 12 and 13 were detected by a sensitive enrichment double PCR-restriction fragment length polymorphism (RFLP) assay. EGFR gene polymorphisms at positions -216G>T, -191C>A and 497Arg>Lys were analyzed (PCR-RFLP), along with CA repeat polymorphism in intron 1 (fluorescent genotyping) and EGFR gene copy number (PCR amplification). EGFR expression was quantified by Scatchard binding assay., Results: The number of EGFR high-affinity sites, dissociation constant (Kd), gene copy number, intron 1, -216G>T, -191C>A or 497Lys>Arg genotypes was not different between K-Ras-mutated or K-Ras-non-mutated tumors. No relationship was observed between any of the analyzed EGFR genotypes and EGFR expression. EGFR expression was not related to gene copy number. EGFR gene copy number in tumor and normal tissue was not correlated. The mean value of the tumor/normal mucosa gene copy number ratio was 1.16., Conclusions: Present data clearly show that EGFR status is independent of K-Ras mutations in colorectal tumors.

Purpose: K-Ras mutations predict resistance to anti-epidermal growth factor receptor (EGFR) monoclonal antibodies. Because combinations of anti-EGFR with 5-fluorouracil (5-FU)-based chemotherapy are promising treatments, we analyzed the effect of K-Ras mutations in patients having received exclusive 5-FU therapy., Experimental Design: This study was conducted on 93 stage IV colorectal cancer patients with unresectable measurable liver metastasis receiving 5-FU-leucovorin (56 men and 37 women; 77 cancer deaths). Liver metastases (n = 93) along with primary tumors (n = 48) were analyzed for K-Ras mutations (codons 12 and 13), p53 mutations (exons 4-9), p53 polymorphism (codon 72), thymidylate synthase (TS) polymorphism (28-bp repeats including G>C mutation), methylenetetrahydrofolate reductase polymorphism (677C>T, 1298A>C), thymidylate synthase (TS) activity, dihydropyrimidine dehydrogenase activity, folylpolyglutamate synthase activity, and p53 protein expression., Results: Thirty-six of 93 (38.7%) metastases were K-Ras mutated (30 at codon 12 and 6 at codon 13). Mutated primary tumors (16 of 48) matched perfectly with mutated metastases. The additional analyzed tumor markers were not different between K-Ras mutated and wild-type tumors. The objective response rate was 37%: 44.4% in K-Ras mutated versus 32.1% in wild-type K-Ras metastasis (P = 0.27). Low TS activity in metastasis was the only significant predictor of tumor response (P = 0.047). K-Ras status did not influence specific survival., Conclusions: The present data indicate a perfect concordance of K-Ras mutations between primary and liver metastasis and suggest that any predictive and/or prognostic value of K-Ras mutations in treatments combining anti-EGFR monoclonal antibodies with 5-FU should be exclusively linked to the anti-EGFR agent.

Background: Advances in the understanding of tumor biology have led to the development of targeted therapies allowing progress in colorectal cancer treatment. One of the most promising targets is the epidermal growth factor receptor (EGFR)., Method: The presence and distribution of high- and low-affinity EGFR was investigated retrospectively in a group of 82 colorectal cancer samples (43 normal colon-colon cancer paired samples) using a specific ligand binding assay (Scatchard Analysis)., Findings: A large majority of tumor samples exhibited one class of high-affinity binding sites (78%). Eighteen cases (22%) exhibited both high- and low-affinity binding sites. A wide interpatient variability was observed for the site number, with physiologically-relevant high-affinity sites ranging from 7 to 310 fmol/mg protein in tumors and from 6 to 313 fmol/mg protein in normal mucosa. A significant positive correlation was demonstrated between tumor and normal mucosa for the high-affinity Kd values and for the number of high-affinity sites, suggesting a common regulation for both tumor and normal tissue., Interpretation: These observations (i) could explain recently-reported clinically-active EGFR targeting in colorectal tumors apparently negative for EGFR, and (ii) may offer a plausible explanation for the link observed between toxicity in normal tissue (cutaneous rash) and clinical outcome of patients treated with anti-EGFR drugs. Present data extends our understanding of EGFR identity in colorectal cancer which could be useful in reconsidering the predictive tools for the identification of tumors putatively responsive to EGFR targeted therapy.

Methylenetetrahydrofolate reductase (MTHFR) controls intracellular CH2FH4 concentrations (required for optimal fluoropyrimidine efficacy) by irreversibly converting CH2FH4 into 5-methyltetrahydrofolate. MTHFR 677C>T and 1298A>C polymorphisms are linked to altered enzyme activity. Thus, mutated MTHFR tumours should, in theory, be more sensitive to 5-fluorouracil (5FU) than wild-type tumours. MTHFR polymorphisms in position 677 and 1298 were analysed in 98 colorectal cancer patients with unresectable liver metastases (57 men, 41 women, mean age 64 years) receiving 5FU-folinic acid. 677C>T and 1298A>C genotypes were determined simultaneously by melting curve analyses on liver metastases. 677C>T genotype distribution was 46.9% wt/wt, 34.7% wt/mut and 18.4% mut/mut; that of 1298A>C was 52.0% wt/wt, 35.7% wt/mut and 12.3% mut/mut. The response rate was not related to 1298A>C genotype but was significantly linked to 677C>T genotype (response rate: 40%, 21% and 56% in wt/wt, wt/mut and mut/mut, respectively; P = 0.040), with an increased response rate in mut/mut tumours relative to wt/wt (odds ratio = 1.88). Thymidylate synthase activity measured in metastases was a significant predictor of 5FU responsiveness and the addition of the 677C>T genotype improved model prediction. MTHFR 1298A>C polymorphism was significantly linked to specific survival, with homozygous mutated patients having the worst prognosis (P = 0.009, relative risk = 2.48 in mut/mut versus wt/wt). MTHFR 1298A>C genotype remained a significant predictor in a multivariate analysis including metastasis characteristics. The results suggest that MTHFR genotypes are relevant and independent factors of patient outcome in 5FU-based treatment of advanced colorectal cancer.

Objective: We describe the rates and potential risk factors of complications of radiofrequency ablation of hepatic tumors. SUBJECTS AND METHODS. Over a 5-year period, 312 patients underwent 350 sessions of radiofrequency ablation (124 intraoperative and 226 percutaneous) for treatment of 582 liver tumors including 115 hepatocellular carcinomas and 467 metastatic tumors. The chi-square test was used for a group-to-group comparison of the occurrence of adverse events., Results: Thirty-seven (10.6%) adverse events and five (1.4%) deaths were related to radiofrequency treatment. The deaths were caused by liver insufficiency (n = 1), colon perforation (n = 1), and portal vein thrombosis (n = 3). Portal vein thrombosis was significantly (p < 0.00001) more frequent in cirrhotic livers (2/5) than in noncirrhotic livers (0/54) after intraoperative radiofrequency ablation performed during a Pringle maneuver. Liver abscess (n = 7) was the most common complication. Abscess occurred significantly (p < 0.00001) more frequently in patients bearing a bilioenteric anastomosis (3/3) than in other patients (4/223). We encountered five pleural effusions, five skin burns, four hypoxemias, three pneumothoraces, two small subcapsular hematomas, one acute renal insufficiency, one hemoperitoneum, and one needle-tract seeding. The 6.3% of minor complications did not require specific treatment or a prolonged hospital stay. Among the 5.7% major complications, 3.7% required less than 5 days of hospitalization for treatment or surveillance and 2% required more than 5 days for treatment., Conclusion: Radiofrequency ablation of liver tumors is a well-tolerated technique, but caution should be exercised when treating patients with a bilioenteric anastomosis, and radiofrequency ablation during vascular occlusion in cirrhotic livers should be avoided.

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.

In vitro and clinical studies have suggested that high-frequency microsatellite instability (MSI-H) phenotype, p53 and K-ras mutations might influence the response to chemotherapy in a variety of tumors, including primary colorectal cancers (CRC). Unresectable hepatic metastases from CRC are commonly treated with 5-fluorouracil (5FU) and folinic acid. Since several new active drugs are now used for treating CRC, molecular determinants predictive to response to 5FU would thus be crucial for optimizing indications of chemotherapy to those patients. MSI-H phenotype, p53 and K-ras status were characterized in a prospective study of 56 patients with CRC metastatic to the liver and treated with 5FU-based chemotherapy. The objective response rate after a 3-month treatment was 32.1%. The prevalence of p53 mutations, K-ras mutations and MSI-H phenotype was 62.5%, 30.3% and 1.8%, respectively. No significant association was found between response to chemotherapy and p53 mutations (78% mutated tumors in responders vs. 55% in nonresponders; p = 0.10) and K-ras mutations (39% mutated tumors in responders vs. 26% in nonresponders; p = 0.34). Survival was longer for patients with p53-mutated metastases than for patients with unresected wild-type p53 metastases (median survival 15 months vs. 17 months; p = 0.06). The determination of the MSI-H phenotype, p53 and K-ras status in hepatic metastases from CRC does not discriminate a group of patients that should preferentially benefit from 5FU-based chemotherapy. The prognosis of patients with treated liver metastases is better when p53 is mutated., (Copyright 2001 Wiley-Liss, Inc.)

Thymidylate synthase (TS) is the main target for fluorouracil (FU). Optimal cellular concentrations of reduced folates in polyglutamated forms [via folylpolyglutamate synthetase (FPGS)] are necessary for achieving maximal TS inhibition. The aim of this multicentric prospective study was to analyze the link between clinical response to FU therapy for liver metastases of colorectal carcinoma and tumoral TS and FPGS activities. Forty-four advanced colorectal cancer patients (15 women and 29 men; median age 63, range, 27-78 years) receiving a standard FU-folinic acid protocol were included. A single hepatic tumoral biopsy was obtained systematically at the time of diagnosis. For 24 patients, a biopsy in the primary colon tumor was available. TS and FPGS activities were measured by radioenzymatic assays. Clinical response on hepatic metastases was 1 complete response, 12 partial responses, 14 stabilizations, and 17 progressions. In hepatic biopsies, TS activity (median, 185; range, <10-3111 fmol/min/mg protein) and FPGS activity (median, 1270; range, <400-3730 fmol/min/mg protein) exhibited a wide variability. TS activity in primary tumors (median, 461; range, 35-2565 fmol/min/mg protein) was significantly higher than in hepatic metastases. No difference was observed between primaries and metastases for FPGS. FPGS activity expressed in liver metastases was significantly correlated to that expressed in primaries. The distribution of TS activity in liver metastases was not significantly different between responsive and nonresponsive patients. However, FPGS activity measured in liver metastases was significantly higher in responsive patients (median, 1550 fmol/min/mg protein) than in nonresponsive patients (median, 1100 fmol/min/mg protein). A discriminant analysis revealed that 24 of the 25 patients exhibiting a liver FPGS activity 320 fmol/min/mg protein were nonresponding patients. These data establish for the first time the potential importance of tumoral FPGS activity for assessing FU-folinic acid responsiveness in the clinical setting.