Thymidylate synthase expression and genotype have no major impact on the clinical outcome of colorectal cancer patients treated with 5-fluorouracil

Abstract

Background and objectives

Thymidylate synthase (TS) expression levels appear to be related to response to 5-fluorouracil-(5-FU)-based chemotherapy in colorectal cancer (CRC) patients. Three polymorphisms have been proposed as modulators of TS expression: a tandemly repeated sequence (2R/3R) in the 5′ UTR, a SNP (G > C) within the 3R allele and a 6 bp deletion in the 3′ UTR.

To evaluate the influence of TS expression and polymorphisms on clinical outcome of 5-FU-treated patients we performed a comprehensive genetic analysis on 63 CRC patients.

Methods

TS expression levels were analyzed in normal and tumor tissues. TS coding sequence and UTR polymorphisms were investigated on DNA from normal tissue. LOH analysis was performed to determine tumor genotype.

Results

A difference in disease-free survival (DFS), although not statistically significant, was observed between high and low mRNA expression levels: patients with low levels showed longer DFS. The 2R2R genotype showed significantly lower expression than the 3R3R and 2R3R genotypes in normal tissue. No other TS polymorphism was associated with mRNA expression or clinical outcome.

Conclusions

The results obtained in this pilot study indicate that the number of 5′ UTR repeats is the major genetic determinant of TS expression. The lack of association with other polymorphisms might be partially explained by the existence of linkage disequilibrium in the TS gene. Our data support the growing evidence that TS control may require multiple mechanisms acting in close coordination with one another and suggest that TS genotyping alone in tumor samples is not sufficient to accurately predict response to 5-FU.

Introduction

More than 50 years after its introduction into clinical practice, 5-fluorouracil (5-FU) is still a fundamental drug in the treatment of colorectal cancer (CRC) and many other tumors, either alone or in combination with other drugs [1], [2]. Several variables associated with genes involved in the 5-FU metabolic pathway have been studied for their potential relationship with clinical outcome and response to chemotherapy. To date the most widely studied molecular marker is thymidylate synthase (TS), the biological target of 5-FU and related drugs. TS is the key enzyme of the de novo synthesis of deoxythymidine monophosphate (dTMP), that catalyzes the methylation of deoxyuridine monophosphate to dTMP [3], [4], an essential step in DNA synthesis.

The expression level of the TS gene appears to be related to clinical outcome and response to 5-FU chemotherapy and has been suggested as a potential prognostic and/or predictive marker. Johnston et al. [5] first demonstrated a correlation between low TS levels and improved 5-year disease-free survival (DFS) and overall survival (OS) in rectal cancer patients receiving 5-FU adjuvant chemotherapy. A meta-analysis by Popat et al. [6] showed that CRC patients with advanced disease treated with TS inhibitors had a significantly better OS if they had low TS expression in primary tumors or metastases, whereas a predictive role of TS expression was not established for the adjuvant setting.

Three different polymorphisms in the TS untranslated regions (UTRs) have been proposed as modulators of TS mRNA transcriptional and translational efficiency. The 5′ UTR contains a variable number of 28 bp tandem repeats (VNTR) [7], [8]. Although up to nine repeats have been described, the vast majority of TS alleles harbor either 2 or 3 repeats, creating genotypes defined as 2R/2R, 2R/3R and 3R/3R, respectively. The 3R alleles present a G > C single nucleotide polymorphism (SNP) at the 12th position of the second repeat [9]. The two alleles of this SNP are defined as 3RG and 3RC, respectively. The third polymorphism is a 6 bp insertion/deletion at nucleotide 1494 within the 3′ UTR [10]. In addition, a further SNP in the VNTR region, consisting of a C > G substitution in the first repeat of the 2R allele, has recently been described [11].

Many studies have been conducted to investigate whether TS genotypes might explain differences in mRNA expression levels, but the results are heterogeneous and even controversial.

In vitro experiments have shown that the 3R/3R genotype is associated with higher levels of TS gene expression than the 2R/2R genotype [8]. Subsequent experiments on DNA samples from CRC patients provided support to these results, indicating that the 3R sequence has greater transcriptional efficiency than the 2R allele [12], [13]. The influence of the VNTR on TS expression has been ascribed to the presence of a USF family E-box consensus element in repeat units containing the G nucleotide at position 12. The G > C substitution eliminates the USF-1 binding site, thus abolishing the translation enhancer effect of the 3R allele [9]. Each of the first two repeats of the 3RG allele has a G at this position; therefore, 3RG alleles contain two potential USF-1 binding sites, whereas the G is present only in the first repeat of 2R and 3RC alleles, that consequently have a single USF-1 site.

The ins/del polymorphism within the 3′ UTR seems to modulate TS expression by affecting mRNA stability; the 6bpdel allele has been associated with decreased mRNA stability in vitro and lower intratumoral TS expression in vivo [14], [15].

By contrast, other studies did not detect any correlation between TS polymorphisms and mRNA or protein expression levels [16], [17], [18], or even a significantly decreased TS mRNA expression in samples from patients with the 3R/3R genotype [19].

Discrepancies in results among different studies may be due to methodological differences or incomplete analysis leading to partial results. Some groups analyzed TS expression levels by immunohistochemistry (IHC) [14], [15], [20] while others used real time quantitative PCR [12], [21], [22]. In addition, loss of heterozygosity (LOH), that modifies TS genotype in tumor cells, has been reported to affect tumor response and survival and TS expression [21], [23].

In order to take into account all major intrinsic factors potentially involved in the relationship between TS genotype, expression and clinical response to 5-FU, we performed a comprehensive pilot study; this involved investigation of the whole TS coding sequence, 5′ and 3′ UTR polymorphisms and mRNA expression in normal and matching tumor tissues of a series of CRC patients receiving treatment with 5-FU only. We analyzed possible associations between TS genotype and expression as well as those between these experimental data and survival parameters (DFS and OS). Additional associations between TS genotype, expression and clinical/pathological characteristics were explored.