Correlative analysis of putative molecular predictive factors in patients with curatively resected stage III colon cancer, treated with adjuvant oxaliplatin-based chemotherapy.

Colon cancer is the second cause of death for neoplasm worldwide. In most cases it is diagnosed when still localized to the intestinal wall or in regional lymph nodes. Adjuvant therapy with 5-Fluorouracil (5FU) and folinic acid (FA), in combination with oxaliplatin (FOLFOX) are the standard options for patients with radically resected stage III disease. However, a proportion of patients will develop recurrence due to drug resistance and oxaliplatin-based chemotherapy is a regimen that may cause potentially disabling sensory neuropathy. Therefore there is an increase needing for a better selection of patients to be addressed to the most appropriate chemotherapeutic treatment, also in the adjuvant setting. Several proteins and genetic markers have been described in an attempt to refine prognostic information and predict the benefit derived from systemic treatment. In particular TS protein expression, MSI, p53 expression, BRAF and TP53 mutations, have been described in several reports in relation to 5FU treatment, whereas ERCC1 polymorphism, ERCC1 expression and KRAS mutations, seem to be related to oxaliplatin efficacy in advanced colon cancer patients. At this purpose we enrolled 230 patients from Argentina and Switzerland who underwent surgical resection, followed by 6-months adjuvant treatment: 106 were treated with 5FU alone and 124 with FOLFOX. In all the cases we investigated the MSI status by fragment analysis, we analyzed BRAF, KRAS, TP53 mutations and ERCC1 codon 118 polymorphism by direct sequencing and we performed ERCC1 expression analysis at protein and mRNA levels by immunohistochemistry and real-time PCR, respectively. Finally, we correlated the molecular and immunohistochemical results with the clinical data. Above all, a little advantage in survival was observed for patients treated with FOLFOX regimen if compared to those treated with 5FU (51.3 and 41.6 months, respectively, for DFS; 55.4 and 49.3 months, respectively, for OS), although the difference was not statistically significant, probably due to the low number of analyzed cases. We found MSI in 12% of cases, BRAF mutations in 9% of cases, KRAS mutations in 28% of cases and ERCC1 resulted over-expressed in 40% of cases detected by IHC and in 49% of cases detected by real-time PCR. These percentages, as well as the types of alterations, are in line with those published in the literature. Concerning the correlations among markers, we observed a significant association between MSI and BRAF mutations (in agreement with the literature) and absence of association between KRAS mutations and ERCC1 expression (at odds with the hypothesis proposed in a recent preclinical study). When we matched the clinical data of the whole patients cohort with molecular alterations, we found a trend towards a better prognosis for patients with MSI than for those with a MSS status (p=0.17); we observed that KRAS mutations confer a worse prognosis to advanced colon cancer patients, borderline for the DFS (p=0.07) and statistically significant for the OS (p=0.004); finally we found a trend towards a better DFS (p=0.11) for patients showing low levels of ERCC1 mRNA expression. When we subdivided the patients on the basis of the received treatment (5FU versus oxaliplatin-based chemotherapy), we observed similar percentages of alterations of all the markers between the two groups. By correlating the molecular alterations with clinical data, we found a trend towards a better survival for MSI patients treated 5FU (p=0.16 and p=0.37 for DFS and OS, respectively), while for FOLFOX patients no clinical differences were found between MSI and MSS cases. As for KRAS mutations, in 5FU group we observed a statistical significant worse DFS (p=0.04) and a trend towards a worse OS (p=0.07) in KRAS mutated patients if compared to wild-type patients. In FOLFOX group, no statistical differences were identified between KRAS mutated and wild-type cases. Stratifying the population on the basis of KRAS mutational status, we noticed that in wild-type patients there was no difference in the clinical outcome in the two treatment modalities. On the contrary, in mutated cases a trend towards a better DFS (p=0.28) and OS (p=0.20) was observed in FOLFOX treated patients if compared to 5FU group. As regards ERCC1 expression, we found only a trend toward a better DFS (p=0.17) in patients characterized by low ERCC1 mRNA levels when treated with FOLFOX. As for the last markers, ERCC1 codon 118 polymorphism (AAT/AAC) and TP53 mutations, we found percentages of alterations in line with the literature (for ERCC1 polymorphism: TT genotype in 31% of cases, CC genotype in 21% of patients; for TP53: 44% of cases showed at least one mutation). The correlations between these two markers and the clinical outcome are now under evaluation. In conclusion, looking at the whole cohort, we can confirm a better clinical outcome for adjuvant colon cancer patients treated with FOLFOX regimen with respect to 5FU treatment. MSI could be a useful tool indicating a better prognosis also for advanced colon cancer but its role in predicting 5FU or FOLFOX efficacy remains controversial. In addition, we propose to assess ERCC1 mRNA expression analysis before the administration of oxaliplatin-based chemotherapy, in order to early identify the patients who may benefit the most from this treatment. Finally, we suggest that KRAS mutational status could help clinicians in selecting the best chemotherapeutic treatment in the adjuvant setting: only KRAS mutant patients should be treated with a platinum-based chemotherapy, while patients whose tumour is KRAS wild-type can be treated with 5FU alone, thus preventing adverse side effect in a consistent number of cases. Our results, of course, deserve confirmations.