Influence of thymidylate synthase DNA polymorphisms and gender on the clinical evolution of patients with advanced colorectal cancer

Abstract

Abstract. Experimental evidence has revealed that several thymidylate synthase (TS) DNA polymorphisms modulate gene expression, which, in turn is known to be downregulated by oestrogen receptor subtypes. Consequently, this process might be influenced by female hormones. Based on these data, we investigated whether patient's gender and TS polymorphism exert an interactive effect on the clinical evolution of patients with advanced colorectal cancer (CRC) subjected to 5 fluorouracil (5FU)-based adjuvant chemotherapy. A retrospective study was carried out on paraffinembedded sections from 81 CRC patients. A variable tandem repeat (VNTR) of 28 bp, a G/C single nucleotide polymorphism (SNP), and a deletion of 6 bp (ins1494del 6 bp) were studied. Genotyping methods were polymerase chain reaction (PCR) for VNTR, and PCR followed by restriction length fragment polymorphism (PCR-RFLP) for SNP and ins1494del 6 bp. The effect of TS genotype and gender on overall and progression-free survival was assessed in univariate and multivariate (Cox regression model) tests. In male patients, the study of combined TS genotypes showed that G&6 + /6 + was an adverse marker for overall (P=0.04; median: not reached) and progression-free survival (P=0.03; median: 12 months, 95% CI: 0-32.4). In the multivariate analysis, the concurrence of G&6 + /6 + combination and male patients resulted in a 5.5-fold increased risk of relapse or disease progression (95% CI: 1-32.1; likelihood test P=0.004; interaction P=0.06). TS genotype did not affect survival among women. The present study supports that the effect of TS polymorphisms on the clinical evolution of advanced CRC patients is significantly influenced by gender.

Functional polymorphisms of folate metabolism and response to chemotherapy for colorectal cancer, a systematic review and meta-analysis

“…However, the utility of the assays could be compromised by the use of tumour tissue because somatic changes such as gene amplifications and loss of heterozygosity could arise as part of tumour evolution and this could affect genotype classification; both amplification and loss of heterozygosity have been described in studies of TYMS genotyping included in this review [14,55]. In our sensitivity analyses, we excluded studies that did not use predominantly peripheral blood or nonmalignant tissue for DNA extraction [21,37,38,[49][50][51], those studies in which the genotypes were not in HWE and the few studies that did not consider populations with European ancestry; we also considered the impact of combination versus fluoropyrimidine monotherapy because of the heterogeneity in the chemotherapeutic regimens described in the primary studies. The sensitivity analyses did not substantially alter the results or affect our overall conclusions (see Table 5).…”

“…Some studies used frozen or fixed and archived tissue samples (either tumour tissue or dissected normal tissue) from surgical resections [14,21,25,30,37,38,[49][50][51]55,56], whereas most extracted DNA was from peripheral blood samples. The analytical sensitivity and specificity for detecting SNPs are very high, even with the use of DNA-extracted fixed tissue.…”

“…The cut-off for exclusion was defined as P-value less than 0.05 using a twotailed w 2 -test. (2) Exclusion of studies that did not use predominantly peripheral blood or nonmalignant tissue for DNA extraction [21,37,38,[49][50][51]. (3) Exclusion of studies of populations that were not of European ancestry [5,24,36,48,52,53].…”

“…In addition, there were six studies in which the DNA extraction was from predominantly tumour tissue, rather than peripheral blood or normal tissue from the surgical resection [21,37,38,[49][50][51]. As the validity of study results may be affected in either of the above situations and because of variations in ethnicity and allele frequencies (see the population characteristics above), a sensitivity analysis, excluding studies as appropriate, is presented.…”

“…In addition, the following studies were excluded: Cho et al [52] (because risk ratios were determined using anonymized control samples from healthy individuals), Cui et al [53] (because outcome data were presented for a mixed cohort of patients with gastric or colorectal cancer), Fernandez-Contreras et al [49] (because only sex-specific outcome data were presented), Wisotzkey et al [56] (because of limited demographic data and in the survival analysis, no information was presented on the length of the follow-up period), Zhang et al [57] (because the format of the outcome data was not congruent with this research synthesis). Also Lurje et al [25], Kawakami and Watanabe [16], Prall et al [21], Sharma et al [26] and Terrazzino et al [31], because they presented haplotype data for TYMS 5 0 UTR and G > C SNP rather than data of individual genotypes.…”