Background: A relationship between dietary folate intake and efficacy of fluorouracil (FU) is supported by preclinical data. Furthermore, there are several reports that evaluated genetic polymorphisms of MTHFR (methylenetetrahydrofolate reductase) or TYMS (thymidylate synthase) and efficacy of FU. However, to our knowledge, there are no reports that evaluate simultaneously the effects of folate intake and genetic polymorphisms on clinical outcome of gastric cancer patients.

Methods: We retrospectively analyzed the survival impact of estimated folate intake by a food frequency questionnaire and MTHFR and TYMS polymorphisms in 132 patients with advanced gastric cancer who were treated with first-line FU-based chemotherapy.

Results: Median overall survival was 11.3 months (95% confidence interval, 9.4-13.4 mo) and median progression-free survival was 5.2 months (95% confidence interval, 4.1-6.3 mo). Patients with folate intake of >260 μg/day (n = 88) showed longer overall survival compared with low folate intake (n = 44; overall survival, 12.2 versus 8.4 mo). In a multivariate Cox model, patients who had folate intake of >260 μg/day, MTHFR 677 TT polymorphism, and TYMS-3′ untranslated region 6-bp insertion were associated with better survival. Similar tendency was observed in progression-free survival. No interaction was observed between folate intake and favorable genotypes.

Conclusion: Folate intake and genetic polymorphisms of MTHFR and TYMS were associated with better clinical outcome by FU-based chemotherapy in advanced gastric cancer.

Impact: Our results suggested folate intake and folate-related genetic polymorphisms may play an important role in efficacy of FU-based chemotherapy in advanced gastric cancer. Cancer Epidemiol Biomarkers Prev; 19(5); 1311–9. ©2010 AACR.

Fluorouracil (FU) is the most widely used drug for advanced gastric cancer. Oral fluoropyrimidines, such as capecitabine or S-1, which contain the prodrug of FU, show similar efficacy to FU (1-3). FU is converted to 5-fluoro-dUMP, which forms a ternary complex with thymidylate synthase (TYMS) and 5-10-methylene tetrahydrofolate (4). Formation of this ternary complex results in sustained inhibition of TYMS and further DNA synthesis, which is thought to be the predominant mechanism of the antitumor effect of FU or fluoropyrimidines (4).

Folate metabolism is an important pathway for the antitumor effect of FU because antitumor activity is dependent on an interaction with folate metabolism (4). Increased 5-10-methylene tetrahydrofolate may produce tighter ternary complexes and improved antitumor efficacy of FU. 5,10-Methylenetetrahydrofolate reductase (MTHFR) is a key enzyme in folate metabolism, which catalyzes the irreversible conversion of 5-10-methylene tetrahydrofolate to 5-methyltetrahydrofolate (5). Because decreased activity of MTHFR may result in accumulation of 5-10-methylene tetrahydrofolate and improve the antitumor efficacy of FU, several studies evaluated genetic polymorphisms of MTHFR, with or without genetic polymorphisms of TYMS in patients with advanced gastric cancer, although the clinical data are still controversial (6-10).

In preclinical data, a substantial effect of dietary folate intake on the efficacy and safety of FU was suggested (11, 12), although only two clinical studies in colorectal cancer evaluated this relationship (13, 14). In addition, combined analysis between folate intake and genetic polymorphisms of MTHFR or TYMS is reported to be important when investigating gastric cancer risk (15-17). However, there are no reports that evaluate the effect of folate intake and genetic polymorphisms simultaneously on clinical outcome of advanced gastric cancer.

To address this issue, we did a retrospective cohort study using data from the Hospital-Based Epidemiologic Research Program at Aichi Cancer Center (HERPACC) combined with clinical data from Aichi Cancer Center Hospital, Japan.

Patients

Cases were selected from the database of the HERPACC conducted at Aichi Cancer Center Hospital. Details of the HERPACC have been described elsewhere (18, 19). In brief, 23,408 HERPACC-enrolled, first-visit outpatients treated between January 2001 and November 2005 were asked to provide blood samples in addition to information on lifestyle factors. Of those who participated, 22,727 (97.1%) subjects completed the questionnaire satisfactorily and were enrolled in the HERPACC. The study was approved by the Institutional Ethical Committee of Aichi Cancer Center Hospital.

In the present study, cases of newly diagnosed advanced gastric cancer who participated in the HERPACC with the following criteria were included: (a) presence of histologically or cytologically proven, inoperable gastric cancer, (b) treated with first-line chemotherapy with FU, (c) performance status according to the Eastern Cooperative Oncology Group criteria of 0 to 2, (d) sufficient oral intake possible (no use of i.v. hyperalimentation, (e) written informed consent before chemotherapy, and (f) blood samples available for the analysis. During the same period of HERPACC, overall 267 patients with advanced gastric cancer were treated by chemotherapy in our hospital. Among them, 212 patients completed the questionnaire of HERPACC (79.4%) and 132 patients met the criteria and were included as subjects in this study.

Estimation of folate intake

Folate intake was estimated through responses in the HERPACC questionnaire. The HERPACC questionnaire included items on demographic characteristics, family and individual medical history, height and weight, exercise, smoking and drinking habits, vitamin use, and consumption of selected foods and beverages. All dietary exposures were determined by a food frequency questionnaire (20, 21), a self-administered questionnaire given to patients at their first visit to Aichi Cancer Center Hospital before any diagnostic procedures were conducted. Briefly, the food frequency questionnaire consisted of 47 single food items with frequencies in eight categories. We estimated the average daily intake of nutrients by multiplying the food intake (in grams) or serving size by the nutrient content per 100 g of food as listed in the standard tables of food composition. Consumption of nutrients from supplements was not considered in total vitamin consumption because the questionnaire for multivitamins was not quantitative (22). However, this might affect the results of folate intake; therefore, we included the variable use of vitamin supplement (yes, 1; no, 2) in the multivariate analysis. Energy-adjusted intakes of food groups and nutrients were calculated using the residual method (23). The food frequency questionnaire was validated by referring to a 3-day weighted dietary record as a standard, which showed validity (20) and reproducibility to be satisfactory. The deattenuated r's for energy-adjusted intakes of folate were 0.36 [95% confidence interval (95% CI), 0.12-0.58] in men and 0.38 (95% CI, 0.25-0.62) in women, respectively.

Evaluation of genetic polymorphisms

DNA of each subject was extracted from the buffy coat fraction with the DNA Blood mini Kit (Qiagen K.K.). Genotyping for the MTHFR C677T [a database of Single Nucleotide Polymorphism (dbSNP) ID, rs1801133] was based upon Taqman Assays (Applied Biosystems). The TYMS 28-bp variable number of tandem repeat polymorphism (dbSNP ID, rs45445694) was defined by PCR using 5′-CGTGGCTCCTGCGTTTCC-3′ and 5′-GAGCCGGCCACAGGCAT-3′ primers. The TYMS 6-bp insertion/deletion (6/6) in the 3' untranslated region (UTR) polymorphism (dbSNP ID, rs16430) was determined by PCR using 5′-CAAATCTGAGGGAGCTGAGT-3′ and 5′-CAGATAAGTGGCAGTACAGA-3′ primers followed by digestion with the restriction enzyme DraI (New England BioLabs). Five percent of the samples were examined in duplicate for consistency, and 100% agreement was observed.

Evaluation of treatment and statistical methods

The primary purpose of this study was to evaluate the association between estimated folate intake, genetic polymorphisms, and overall survival, which was defined as the interval between the date of initial chemotherapy to the date of death or last follow-up using the Kaplan-Meier method. Progression-free survival associated with first-line chemotherapy was also measured from the beginning of treatment to the date of disease progression, which was evaluated by each physician. Vital status or disease status was confirmed by checking medical record at the last date of follow-up visit. In the case of lost to follow-up, vital status was confirmed by census registration conducted annually. Association between genetic polymorphisms, folate intake, and progression-free survival was also evaluated. To evaluate the effect of genetic polymorphisms and folate intake on overall survival and progression-free survival, univariate and multivariate Cox proportional hazards modeling was applied. Therefore, a measure of association in this study was the hazard ratio along with a 95% CI. Forward and backward stepwise methods were used for model building using threshold P values 0.10 for inclusion and 0.20 for exclusion. Toxicity during first-line chemotherapy was also evaluated and graded according to the National Cancer Institute Common Toxicity Criteria version 3.0. Distribution of subject characteristics was assessed by the χ2 test or the Fisher's exact test as appropriate. Statistical analyses were done using STATA ver. 10 (StataCorp LP). All tests were two-sided, and P < 0.05 was considered statistically significant.

Patient characteristics and survival

Detailed characteristics of 132 patients are shown in Table 1. Although vitamin supplement was not reported in detail, any vitamin supplement was used in 30 (22.7%) of 132 patients. First-line chemotherapy was administered as follows: S-1 alone (n = 92), S-1 combination (n = 19; S-1 + cisplatin in 16 patients, S-1 + docetaxel in two patients, S-1 + irinotecan in one patient), and other FU combinations (n = 21; FU + cisplatin in 13 patients, FU + methotrexate in eight patients), indicating fewer patients received combination chemotherapy (n = 40; 30%) compared to monotherapy (n = 92; 70%). Detailed schedule of each chemotherapy was shown in Supplementary Table S1 and in reference (3, 24-28). Median duration of first-line chemotherapy was 4.8 months. One hundred twenty-three patients experienced disease progression with a median progression-free survival of 5.2 months (95% CI, 4.1-6.3 mo). Among the patients who experienced disease progression, second-line chemotherapy was applied in 90 patients (73%). At the time of analysis, 118 (89%) patients had died, with a median follow-up of 61 months since initiation of first-line chemotherapy. Median overall survival for all patients was 11.3 months (95% CI, 9.4-13.4 mo), which was almost similar to the patients with advanced gastric cancer who were treated in our hospital and were not included in this analysis (n = 135; 11.5 mo; 95% CI, 9.8-12.6 mo).

Table 1.

Patient characteristics and genetic polymorphisms

Characteristicsn (%; N = 132)
Age Median (range) 58 (30-80) 
Gender Male 91 (69) 
Female 41 (31) 
ECOG PS 0-1 111 (84) 
21 (16) 
Disease status Advanced 98 (74) 
Recurrent 34 (26) 
Pathologic type Diffuse 103 (78) 
Intestinal 29 (22) 
Previous gastrectomy Yes 59 (45) 
No 73 (55) 
Adjuvant Yes 8 (6) 
No 124 (94) 
Metastatic place 73 (55) 
≥2 59 (45) 
Ascites Yes 28 (21) 
No 104 (79) 
Folate intake* Low 44 (33) 
Medium 44 (33) 
High 44 (33) 
MTHFR 677 C/C 53 (40) 
C/T 59 (45) 
T/T 20 (15) 
TYMS-5′UTR 2R/2R 3 (2) 
2R/3R 41 (31) 
3R/3R or 3R/other 88 (67) 
TYMS-3′UTR +6 bp/+6 bp 19 (14) 
+6 bp/−6 bp 68 (52) 
−6 bp/−6 bp 45 (34) 
Characteristicsn (%; N = 132)
Age Median (range) 58 (30-80) 
Gender Male 91 (69) 
Female 41 (31) 
ECOG PS 0-1 111 (84) 
21 (16) 
Disease status Advanced 98 (74) 
Recurrent 34 (26) 
Pathologic type Diffuse 103 (78) 
Intestinal 29 (22) 
Previous gastrectomy Yes 59 (45) 
No 73 (55) 
Adjuvant Yes 8 (6) 
No 124 (94) 
Metastatic place 73 (55) 
≥2 59 (45) 
Ascites Yes 28 (21) 
No 104 (79) 
Folate intake* Low 44 (33) 
Medium 44 (33) 
High 44 (33) 
MTHFR 677 C/C 53 (40) 
C/T 59 (45) 
T/T 20 (15) 
TYMS-5′UTR 2R/2R 3 (2) 
2R/3R 41 (31) 
3R/3R or 3R/other 88 (67) 
TYMS-3′UTR +6 bp/+6 bp 19 (14) 
+6 bp/−6 bp 68 (52) 
−6 bp/−6 bp 45 (34) 

Abbreviations: ECOG, Eastern Cooperative Oncology Group; PS, performance status.

*Folate intake was divided in three groups: low (≤260 μg/d), medium (>260 and <340 μg/d), and high (≥340 μg/d).

Results of folate intake and genetic polymorphisms of MTHFR and TYMS

Estimated folate intake was divided into three groups: low (≤260 μg/d; lowest tertile), medium (>260 and ≤340 μg/d; middle tertile), and high (>340 μg/d; highest tertile), and the number of patients in these groups was 44, 44, and 44, respectively. The estimated median total calorie intakes in each folate group were as follows: 1,580 kcal (range, 1,131-2,498), 1,600 kcal (range, 910-2,667), and 1,703 kcal (range, 965-2,467).

The frequencies and types of the MTHFR C677T polymorphisms, polymorphic variable number of tandem repeat in the TYMS-5′UTR region, and the 6+/6− polymorphism in the TYMS-3′UTR were shown in Table 1. Genotype distributions of all the polymorphisms were in accordance with Hardy-Weinberg equilibrium. We also confirmed genotyping of insertion/deletion for by sequencing of 10% of samples with completed accordance.

Overall survival according to folate intake and genetic polymorphisms

Table 2 shows univariate and multivariate analyses of folate intake and genetic polymorphisms as prognostic factors for a better overall survival. In the multivariate analysis, patients with intermediate or high folate intake (n = 88) were significantly associated with better survival than patients with low folate intake (n = 44; overall survival, 12.2 versus 8.2 mo; hazard ratio, 0.65; 95% CI, 0.44-0.96; P = 0.03; Fig. 1A). Hazard ratios for medium and high groups were 0.65 (95% CI, 0.41-1.03) and 0.62 (95% CI, 0.38-1.01), respectively. Therefore, we decided to conduct analyses by dichotomization by medium/high (>260 μg/d) versus low (≤260 μg/d) folate intake. In addition, when we stratified other clinical factors, vitamin use, or estimated total calorie, medium/high folate intake tended to be associated with improved prognosis in almost all subgroups (Fig. 1B).

Table 2.

Univariate and multivariate analysis of overall survival

VariantGenotype/classificationnUnivariate analysisMultivariate analysis
HR (95% CI)PHR (95% CI)P
Folate intake Low (≤260 μg/d) 44 1.00  1.00  
Medium/high (>260 μg/d) 88 0.60 (0.41-0.89) 0.013 0.65 (0.44-0.96) 0.030 
MTHFR 677 C/C or C/T 112 1.00  1.00  
T/T 20 0.78 (0.48-1.2) 0.300 0.57 (0.33-0.97) 0.039 
TYMS-5′UTR 2R/2R or 2R/3R 44 1.00  1.00  
3R/3R or 3R/other 88 0.87 (0.59-1.28) 0.508 0.78 (0.51-1.18) 0.220 
TYMS-3′UTR −6 bp/−6 bp or +6 bp/−6 bp 113 1.00  1.00  
+6 bp/+6 bp 19 0.45 (0.25-0.81) 0.008 0.41 (0.22-0.76) 0.005 
VariantGenotype/classificationnUnivariate analysisMultivariate analysis
HR (95% CI)PHR (95% CI)P
Folate intake Low (≤260 μg/d) 44 1.00  1.00  
Medium/high (>260 μg/d) 88 0.60 (0.41-0.89) 0.013 0.65 (0.44-0.96) 0.030 
MTHFR 677 C/C or C/T 112 1.00  1.00  
T/T 20 0.78 (0.48-1.2) 0.300 0.57 (0.33-0.97) 0.039 
TYMS-5′UTR 2R/2R or 2R/3R 44 1.00  1.00  
3R/3R or 3R/other 88 0.87 (0.59-1.28) 0.508 0.78 (0.51-1.18) 0.220 
TYMS-3′UTR −6 bp/−6 bp or +6 bp/−6 bp 113 1.00  1.00  
+6 bp/+6 bp 19 0.45 (0.25-0.81) 0.008 0.41 (0.22-0.76) 0.005 

NOTE: Adjusted by age, performance status, pathologic type, disease status, previous gastrectomy, adjuvant, ascites, metastatic location, regimens, vitamin use, and calorie intake (less than median or more than or equal to median).

Abbreviation: HR, hazard ratio.

Figure 1.

A, Kaplan-Meier survival curves of overall survival. Patients with intermediate or high folate intake (n = 88) were significantly associated with better survival than patients with low folate intake (n = 44; overall survival, 12.2 versus 8.2 mo; hazard ratio, 0.65; 95% CI, 0.44-0.96; P = 0.03). B, Kaplan-Meier survival curves of progression-free survival. Patients with intermediate or high folate tended to be associated with longer progression-free survival without statistical significance (progression-free survival, 6.3 versus 4.0 mo; hazard ratio, 0.74; 95% CI, 0.50-1.07; P = 0.094; Fig. 2).

Figure 1.

A, Kaplan-Meier survival curves of overall survival. Patients with intermediate or high folate intake (n = 88) were significantly associated with better survival than patients with low folate intake (n = 44; overall survival, 12.2 versus 8.2 mo; hazard ratio, 0.65; 95% CI, 0.44-0.96; P = 0.03). B, Kaplan-Meier survival curves of progression-free survival. Patients with intermediate or high folate tended to be associated with longer progression-free survival without statistical significance (progression-free survival, 6.3 versus 4.0 mo; hazard ratio, 0.74; 95% CI, 0.50-1.07; P = 0.094; Fig. 2).

Close modal

Patients who had MTHFR TT also had significantly better survival compared with patients with CT or CC (hazard ratio, 0.57; 95% CI, 0.33-0.97; P = 0.039). Hazard ratios for CT and TT compared with CC were 1.05 (95% CI, 0.70-1.65) and 0.59 (95% CI, 0.32-1.06), respectively. Therefore, we used the recessive model (CC/CT versus TT).

Patients who had +6/+6 in TYMS-3′UTR had also significantly better survival compared with −6/−6 or −6/+6 (hazard ratio, 0.41; 95% CI, 0.22-0.76; P = 0.01). In contrast, no association was seen between presence of TYMS-5′UTR and survival (hazard ratio, 0.78; 95% CI, 0.51-1.18; P = 0.220). We did not conduct the haplotype analysis because linkage disequilibrium between repeat polymorphism and insertion/deletion polymorphism was very low (R2 = 0.03; D′ = 0.3).

Progression-free survival according to folate intake and genetic polymorphisms

According to the multivariate analysis for progression-free survival, patients with intermediate or high folate tended to be associated with longer progression-free survival with marginal statistical significance (progression-free survival, 6.3 versus 4.0 mo; hazard ratio, 0.74; 95% CI, 0.50-1.07; P = 0.094; Table 3; Fig. 2). Presence of +6/+6 in TYMS-3′UTR or MTHFR TT was significantly associated with longer progression-free survival (Table 3), similar to what was seen in overall survival.

Table 3.

Univariate and multivariate analysis of progression-free survival

VariantGenotype/classificationnUnivariate analysisMultivariate analysis
HR (95% CI)PHR (95% CI)P
Folate intake Low (≤260 μg/d) 44 1.00  1.00  
Medium/high (>260 μg/d) 88 0.81 (0.55-1.18) 0.280 0.74 (0.50-1.07) 0.094 
MTHFR 677 C/C or C/T 112 1.00  1.00  
T/T 20 0.66 (0.38-1.15) 0.140 0.53 (0.29-0.99) 0.046 
TYMS-5′UTR 2R/2R or 2R/3R 44 1.00  1.00  
3R/3R or other 88 1.10 (0.77-1.69) 0.500 0.94 (0.62-1.43) 0.780 
TYMS-3′UTR −6 bp/−6 bp or +6 bp/−6 bp 113 1.00  1.00  
+6 bp/+6 bp 19 0.56 (0.31-0.99) 0.049 0.49 (0.25-0.9) 0.029 
VariantGenotype/classificationnUnivariate analysisMultivariate analysis
HR (95% CI)PHR (95% CI)P
Folate intake Low (≤260 μg/d) 44 1.00  1.00  
Medium/high (>260 μg/d) 88 0.81 (0.55-1.18) 0.280 0.74 (0.50-1.07) 0.094 
MTHFR 677 C/C or C/T 112 1.00  1.00  
T/T 20 0.66 (0.38-1.15) 0.140 0.53 (0.29-0.99) 0.046 
TYMS-5′UTR 2R/2R or 2R/3R 44 1.00  1.00  
3R/3R or other 88 1.10 (0.77-1.69) 0.500 0.94 (0.62-1.43) 0.780 
TYMS-3′UTR −6 bp/−6 bp or +6 bp/−6 bp 113 1.00  1.00  
+6 bp/+6 bp 19 0.56 (0.31-0.99) 0.049 0.49 (0.25-0.9) 0.029 

NOTE: Adjusted by age, performance status, pathologic type, disease status, previous gastrectomy, adjuvant, ascites, metastatic location, regimens, vitamin use, and calorie intake (less than median or more than or equal to median).

Figure 2.

Hazard ratios for death and 95% CIs. HR, hazard ratio. In subgroup analyses, medium/high folate intake tended to be associated with improved prognosis in almost all subgroups. PS, performance status.

Figure 2.

Hazard ratios for death and 95% CIs. HR, hazard ratio. In subgroup analyses, medium/high folate intake tended to be associated with improved prognosis in almost all subgroups. PS, performance status.

Close modal

Interaction of genetic polymorphisms and folate intake

The interaction of favorable genotypes and folate intake is shown in Table 4. When we stratified patients according to genotypes, medium or high folate intake was associated with better survival regardless of any genotypes, and no significant interaction was observed between folate intake and genetic polymorphisms on either overall survival or progression-free survival.

Table 4.

Association of folate intake and survival stratification by genotype or other clinical factors

Folate intake*nMultivariate analysis for OSMultivariate analysis for PFS
HR (95% CI)PHR (95% CI)P
MTHFR 677 C/C or C/T Low 36 1.00 0.55 1.00 0.83 
Medium/high 76 0.59 (0.37-0.92) 0.73 (0.48-1.11) 
T/T Low 1.00 1.00 
Medium/high 12 1.19 (0.22-6.3) 3.1 (0.52-19) 
TYMS-5′UTR 2R/2R or 2R/3R Low 17 1.00 0.45 1.00 0.12 
Medium/high 27 0.44 (0.21-0.92) 0.51 (0.24-1.07) 
3R/3R or other Low 27 1.00 1.00 
Medium/high 61 0.69 (0.43-1.15) 1.02 (0.61-1.71) 
TYMS-3′UTR −6 bp/−6 bp or +6 bp/−6 bp Low 37 1.00 0.27 1.00 0.98 
Medium/high 76 0.66 (0.43-1.01) 0.71 (0.47-1.08) 
+6 bp/+6 bp Low 1.00 1.00 
Medium/high 12 0.10 (0.01-0.46) 0.33 (0.03-3.9) 
Folate intake*nMultivariate analysis for OSMultivariate analysis for PFS
HR (95% CI)PHR (95% CI)P
MTHFR 677 C/C or C/T Low 36 1.00 0.55 1.00 0.83 
Medium/high 76 0.59 (0.37-0.92) 0.73 (0.48-1.11) 
T/T Low 1.00 1.00 
Medium/high 12 1.19 (0.22-6.3) 3.1 (0.52-19) 
TYMS-5′UTR 2R/2R or 2R/3R Low 17 1.00 0.45 1.00 0.12 
Medium/high 27 0.44 (0.21-0.92) 0.51 (0.24-1.07) 
3R/3R or other Low 27 1.00 1.00 
Medium/high 61 0.69 (0.43-1.15) 1.02 (0.61-1.71) 
TYMS-3′UTR −6 bp/−6 bp or +6 bp/−6 bp Low 37 1.00 0.27 1.00 0.98 
Medium/high 76 0.66 (0.43-1.01) 0.71 (0.47-1.08) 
+6 bp/+6 bp Low 1.00 1.00 
Medium/high 12 0.10 (0.01-0.46) 0.33 (0.03-3.9) 

NOTE: Adjusted by age, performance status, pathologic type, disease status, previous gastrectomy, adjuvant, metastatic location, regimens, vitamin supplement use, and calorie intake (less than median or more than or equal to median).

Abbreviations: OS, overall survival; PFS, progression-free survival.

*Folate intake was divided as follows: low (≤260 μg/d) and medium/high (>260 μg/d).

For interaction.

Relationships between toxicity and folate intake or genetic polymorphisms

Hematologic toxicity (grade 3-4) was observed in 34 patients (20.3%), and nonhematologic toxicity (grade 3-4) was seen in 21 patients (15.9%). The frequency of hematologic and nonhematologic grade 3 to 4 toxicity was significantly higher in patients with medium/high folate intake than low folate intake after it was adjusted by age, performance status, gender, and regimens (Table 5). Although favorable genotypes did not correlate to grade3 to 4 toxicity, MTHFR TT tend to be associated with higher frequency of hematologic toxicity with borderline significance (P = 0.072; Table 5).

Table 5.

The frequency of toxicity according to genetic polymorphisms or folate intake

VariantToxicity
Hematologic (grade 3-4)Nonhematologic (grade 3-4)
n (%)OR (95% CI)Pn (%)OR (95% CI)P
Folate intake* Low (n = 44) 4 (9) 1.00  5 (11) 1.00  
Medium/high (n = 88) 23 (26) 3.91 (1.21-12.6) 0.022 16 (18) 1.81 (0.60-5.34) 0.29 
MTHFR 677 C/C or C/T (n = 112) 20 (17) 1.00  16 (14) 1.00  
T/T (n = 20) 7 (35) 2.87 (0.90-8.91) 0.072 5 (25) 2.00 (0.59-6.75) 0.24 
TYMS-5′UTR 2R/2R or 2R/3R (n = 44) 9 (20) 1.00  7 (15) 1.00  
3R/3R or 3R/other (n = 88) 18 (20) 1.14 (0.42-3.12) 0.81 14 (15) 1.23 (0.42-3.59) 0.71 
TYMS-3′UTR −6 bp/−6 bp or +6 bp/−6 bp (n = 113) 23 (20) 1.00  17 (19) 1.00  
+6 bp/+6 bp (n = 19) 4 (21) 1.07 (0.28-4.11) 0.92 4 (21) 1.66 (0.42-6.41) 0.47 
VariantToxicity
Hematologic (grade 3-4)Nonhematologic (grade 3-4)
n (%)OR (95% CI)Pn (%)OR (95% CI)P
Folate intake* Low (n = 44) 4 (9) 1.00  5 (11) 1.00  
Medium/high (n = 88) 23 (26) 3.91 (1.21-12.6) 0.022 16 (18) 1.81 (0.60-5.34) 0.29 
MTHFR 677 C/C or C/T (n = 112) 20 (17) 1.00  16 (14) 1.00  
T/T (n = 20) 7 (35) 2.87 (0.90-8.91) 0.072 5 (25) 2.00 (0.59-6.75) 0.24 
TYMS-5′UTR 2R/2R or 2R/3R (n = 44) 9 (20) 1.00  7 (15) 1.00  
3R/3R or 3R/other (n = 88) 18 (20) 1.14 (0.42-3.12) 0.81 14 (15) 1.23 (0.42-3.59) 0.71 
TYMS-3′UTR −6 bp/−6 bp or +6 bp/−6 bp (n = 113) 23 (20) 1.00  17 (19) 1.00  
+6 bp/+6 bp (n = 19) 4 (21) 1.07 (0.28-4.11) 0.92 4 (21) 1.66 (0.42-6.41) 0.47 

NOTE: Adjusted by age, performance status, gender, regimens, vitamin supplement use, and calorie intake (less than median or more than or equal to median).

Abbreviation: OR, odds ratio.

*Folate intake was divided as follows: low (≤260 μg/d) and medium/high (>260 μg/d).

In this study, we found that medium or high amount of folate intake and genetic polymorphisms in TYMS and MTHFR were associated with significantly better survival in advanced gastric cancer patients treated with FU-based chemotherapy. Similar tendencies were observed in progression-free survival, thus inferring that folate intake and these two polymorphisms have predictive values for FU-based chemotherapy in advanced gastric cancer. However, folate intake and genetic polymorphisms were independent, and no significant interaction was observed. In addition, folate intake was associated with increased frequency of toxicity.

To our knowledge, there are only two studies that directly evaluated treatment outcome of FU-based chemotherapy and folate (13, 14). In one study, Canadian patients receiving adjuvant FU and leucovorin were prospectively assessed for biomarkers of folate metabolism (13). Multivariate analyses identified baseline serum folate as an independent positive predictor of grade 3 and/or 4 toxic effect. Similar results were found in a study on capecitabine monotherapy in Australia (14), in which patients with higher baseline levels of serum folate had a significantly increased incidence of toxic events. These two reports suggested importance of folate in patients with treated with FU, although they did not evaluate the efficacy of treatment or the genetic polymorphisms. In contrast, we evaluated efficacy (overall survival and progression-free survival) and toxicity in this study. In addition, we evaluate folate and genetic polymorphisms simultaneously. As a result, increased folate intake was associated with better survival of advanced gastric cancer in this study. Although the cause of this association is unclear, one may suggest that sufficient folate intake might be important for FU to exert its antitumor effect. Supporting this hypothesis, folinic acid, which increases the 5-10-methylene tetrahydrofolate levels in cancer cells, is reported to result in a tighter ternary complex of TS, 5-10-methylene tetrahydrofolate, and 5-fluoro-dUMP (4) and showed increased antitumor effects of FU in gastric cancer (29). Therefore, in addition to folinic acid, daily supplementation of folate may be an important factor having an impact on the efficacy of FU. Toxicity is also higher in patients with medium/or high folate intake patients, which suggests increased cytotoxicity by folate intake not only in cancer cells but also in normal tissue. Although methodology was different in our study (estimated folate intake) and previous studies (serum folate), these results might suggest importance of folate in FU-based chemotherapy. Because the impact of folate intake was more intense in overall survival than progression-free survival in our study, other mechanisms than through FU may exist that explain the better survival of advanced gastric cancer patients with higher folate intake.

The MTHFR 677TT genotype was strongly associated with better clinical outcome according to our multivariate survival analysis. MTHFR 677C/T results in an alanine-to-valine substitution that induces a thermolabile variant of the enzyme with reduced activity (5). This may result in accumulation of 5-10-methylene tetrahydrofolate and improved efficacy of FU. Not only better survival but also relatively higher toxicity was seen in TT type in this study, which may support increased cytotoxicity in TT type. In addition, no gene-folate interaction was seen in this study; MTHFR 677TT and folate intake may complementarily increase levels of 5-fluoro-dUMP and therefore increase the effect of FU.

In our study, patients homozygous for the insertion (+6 bp/+6 bp) polymorphism TYMS-3UTR had significantly better survival than those homozygous for the deletion or those that were heterozygous (+6 bp/−6 bp). Given the controversies on the biological significance of the TYMS-3UTR insertion/deletion polymorphism (30, 31), further studies are necessary to evaluate several genetic polymorphisms of TYMS in terms of not only clinical outcome but also resulting TYMS activity in blood and tumor tissue simultaneously.

An additional TYMS polymorphism consisting of two or three 28-bp repeated sequences in the TYMS 5′UTR locus did not show any impact on clinical outcome in our study. Several studies on TYMS expression and this polymorphism were reported especially in colorectal cancer, with some clinical studies showing favorable results in 2R/2R (32-34). However, two recent studies in advanced gastric cancer did not show any significant impact of this polymorphism as seen with our results (7, 9). Although further studies may be necessary, the clinical significance of this polymorphism may be limited.

Our study had several methodologic strengths. First, exposure of interest, folate intake, and polymorphisms were measured before the treatment; therefore, chronological relation between exposure and outcome was in order. Moreover, because clinicians associated with cases in this study did not know the exposure status until the study, it is less likely to introduce the view of researchers as a bias. Secondly, potential confounders such as performance status and disease status were considered in the analyses; therefore, associations that we observed were theoretically independent of confounders, although we cannot completely rule out effect of residual confounding by unevaluated factors. Lastly, given that our allele frequencies were comparable to those previously reported in public databases, bias in the distribution of selected polymorphisms was negligible.

There are several methodologic issues in this study. The food frequency questionnaire was quite short, and validity and reliability of folate intake were modest. There is a possibility of misclassification of exposure and high folate intake being a marker for other behaviors. However, we tried to include other possible factor such as performance status or disease site to exclude this bias. In addition, when we evaluated other estimated nutrients (Supplementary Table S2), none was considered to be significant other than folate intake. In addition, we did not conduct validation between estimated nutrients and biomarkers such as serum folate concentration. In addition, we did not evaluate behavior change after the diagnosis. These points are also limitations in this study.

A small portion of patients received combination chemotherapy because monotherapy of FU or S-1 was the standard chemotherapy regimen at the time of this study. Current standard treatment for advanced gastric cancer in Japan and other countries is FU plus platinum if tolerable, so our patient population does not directly reflect current clinical practice. However, combined agents may also make the impact of genetic polymorphism on the effect of FU itself more obscure. Therefore, this study may be more suggestive in evaluating the impact of polymorphisms on FU itself compared with other studies that evaluate patients receiving combined chemotherapy. In our study, when we limited the cohort to patients who received S-1 alone as first-line chemotherapy (n = 91), almost similar results were obtained (Supplementary Table S3). FU or fluoropyrimidine alone is considered to be optimal for frail patients or those receiving adjuvant chemotherapy, making our results useful to predict patients who may benefit from FU-based chemotherapy. As for genetic polymorphisms, we did not genotype the SNP located within the promoter repeat of TS, which might be confounding the results. In addition, the small sample size used may be a study limitation, which may contribute to lack of statistical power to show the interaction between folate intake and genetic polymorphisms. Therefore, further study is required to duplicate this work in a larger cohort.

In conclusion, this is the first report that simultaneously evaluated the effects of folate intake and genetic polymorphisms on clinical outcome of FU-based chemotherapy in advanced gastric cancer. Our findings indicate that folate intake and genetic polymorphisms of MTHFR or TYMS may play an important role in FU-based chemotherapy in advanced gastric cancer. Further prospective evaluation is warranted.

No potential conflicts of interest were disclosed.

We thank the many doctors, nurses, and technical and administration staff of Aichi Cancer Center Hospital for the daily administration of the HERPACC study.

Grant Support: Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports, Culture and Technology of Japan.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Cunningham
D
,
Starling
N
,
Rao
S
, et al
. 
Capecitabine and oxaliplatin for advanced esophagogastric cancer
.
N Engl J Med
2008
;
358
:
36
46
.
2
Kang
YK
,
Kang
WK
,
Shin
DB
, et al
. 
Capecitabine/cisplatin versus 5-fluorouracil/cisplatin as first-line therapy in patients with advanced gastric cancer: a randomised phase III noninferiority trial
.
Ann Oncol
2009
;
20
:
666
73
.
3
Boku
N
,
Yamamoto
S
,
Fukuda
H
, et al
. 
Fluorouracil versus combination of irinotecan plus cisplatin versus S-1 in metastatic gastric cancer: a randomised phase 3 study
.
Lancet Oncol
2009
;
10
:
1063
9
.
4
Longley
DB
,
Harkin
DP
,
Johnston
PG
. 
5-fluorouracil: mechanisms of action and clinical strategies
.
Nat Rev Cancer
2003
;
3
:
330
8
.
5
De Mattia
E
,
Toffoli
G
. 
C677T and A1298C MTHFR polymorphisms, a challenge for antifolate and fluoropyrimidine-based therapy personalisation
.
Eur J Cancer
2009
;
45
:
1333
51
.
6
Lee
J
,
Jeong
CK
,
Hong
SP
, et al
. 
Clinical significance of thymidylate synthase and methylenetetrahydrofolate reductase gene polymorphism in Korean patients with gastric cancer
.
Korean J Gastroenterol
2005
;
46
:
32
8
.
7
Ruzzo
A
,
Graziano
F
,
Kawakami
K
, et al
. 
Pharmacogenetic profiling and clinical outcome of patients with advanced gastric cancer treated with palliative chemotherapy
.
J Clin Oncol
2006
;
24
:
1883
9
.
8
Lu
JW
,
Gao
CM
,
Wu
JZ
,
Cao
HX
,
Tajima
K
,
Feng
JF
. 
Polymorphism in the 3-untranslated region of the thymidylate synthase gene and sensitivity of stomach cancer to fluoropyrimidine based chemotherapy
.
J Hum Genet
2006
;
51
:
155
60
.
9
Keam
B
,
Im
SA
,
Han
SW
, et al
. 
Modified FOLFOX-6 chemotherapy in advanced gastric cancer: results of phase II study and comprehensive analysis of polymorphisms as a predictive and prognostic marker
.
BMC Cancer
2008
;
8
:
148
.
10
Lu
JW
,
Gao
CM
,
Wu
JZ
,
Sun
XF
,
Wang
L
,
Feng
JF
. 
Relationship of methylenetetrahydrofolate reductase C677T polymorphism and chemosensitivity to 5-fluorouracil in gastric carcinoma
.
Ai Zheng
2004
;
23
:
958
62
.
11
Branda
RF
,
Nigels
E
,
Lafayette
AR
,
Hacker
M
. 
Nutritional folate status influences the efficacy and toxicity of chemotherapy in rats
.
Blood
1998
;
92
:
2471
6
.
12
Branda
RF
,
Chen
Z
,
Brooks
EM
,
Naud
SJ
,
Trainer
TD
,
McCormack
JJ
. 
Diet modulates the toxicity of cancer chemotherapy in rats
.
J Lab Clin Med
2002
;
140
:
358
68
.
13
Ho
C
,
Ng
K
,
O'Reilly
S
,
Gill
S
. 
Outcomes in elderly patients with advanced colorectal cancer treated with capecitabine: a population-based analysis
.
Clin Colorectal Cancer
2005
;
5
:
279
82
.
14
Sharma
R
,
Rivory
L
,
Beale
P
,
Ong
S
,
Horvath
L
,
Clarke
SJ
. 
Phase II study of fixed-dose capecitabine and assessment of predictors of toxicity in patients with advanced/metastatic colorectal cancer
.
Br J Cancer
2006
;
94
:
964
8
.
15
Lin
D
,
Li
H
,
Tan
W
,
Miao
X
,
Wang
L
. 
Genetic polymorphisms in folate-metabolizing enzymes and risk of gastroesophageal cancers: a potential nutrient-gene interaction in cancer development
.
Forum Nutr
2007
;
60
:
140
5
.
16
Tan
W
,
Miao
X
,
Wang
L
, et al
. 
Significant increase in risk of gastroesophageal cancer is associated with interaction between promoter polymorphisms in thymidylate synthase and serum folate status
.
Carcinogenesis
2005
;
26
:
1430
5
.
17
Galván-Portillo
MV
,
Cantoral
A
,
Oñate-Ocaña
LF
, et al
. 
Gastric cancer in relation to the intake of nutrients involved in one-carbon metabolism among MTHFR 677 TT carriers
.
Eur J Nutr
2009
;
48
:
269
76
.
18
Tajima
K
,
Hirose
K
,
Inoue
M
,
Takezaki
T
,
Hamajima
N
,
Kuroishi
T
. 
A model of practical cancer prevention for out-patients visiting a hospital: the Hospital-based Epidemiologic Research Program at Aichi Cancer Center (HERPACC)
.
Asian Pac J Cancer Prev
2000
;
1
:
35
47
.
19
Hamajima
N
,
Matsuo
K
,
Saito
T
, et al
. 
Gene-environment interactions and polymorphism studies of cancer risk in the Hospital-based Epidemiologic Research Program at Aichi Cancer Center II (HERPACCII)
.
Asian Pac J Cancer Prev
2001
;
2
:
99
107
.
20
Tokudome
Y
,
Goto
C
,
Imaeda
N
, et al
. 
Relative validity of a short food frequency questionnaire for assessing nutrient intake versus three-day weighed diet records in middle-aged Japanese
.
J Epidemiol
2005
;
15
:
135
45
.
21
Imaeda
N
,
Goto
C
,
Tokudome
Y
,
Hirose
K
,
Tajima
K
,
Tokudome
S
. 
Reproducibility of a short food frequency questionnaire for Japanese general population
.
J Epidemiol
2007
;
17
:
100
7
.
22
Imaeda
N
,
Goto
C
,
Tokudome
Y
,
Hirose
K
,
Tajima
K
,
Tokudome
S
. 
Dietary supplement use by community-living population in Japan: data from the National Institute for Longevity Sciences Longitudinal Study of Aging (NILS-LSA)
.
J Epidemiol
2006
;
16
:
249
60
.
23
Willett
W
,
Stampfer
MJ
. 
Total energy intake: implications for epidemiologic analyses
.
Am J Epidemiol
1986
;
124
:
17
27
.
24
Koizumi
W
,
Narahara
H
,
Hara
T
, et al
. 
S-1 plus cisplatin versus S-1 alone for first-line treatment of advanced gastric cancer (SPIRITS trial): a phase III trial
.
Lancet Oncol
2008
;
9
:
215
21
.
25
Yoshida
K
,
Ninomiya
M
,
Takakura
N
, et al
. 
Phase II study of docetaxel and S-1 combination therapy for advanced or recurrent gastric cancer
.
Clin Cancer Res
2006
;
12
:
3402
7
.
26
Imamura
H
,
Iishi
H
,
Tsuburaya
A
, et al
. 
Randomized phase III study of irinotecan plus S-1 (IRIS) versus S-1 alone as first-line treatment for advanced gastric cancer(GC0301/TOP-002)
.
2008 Gastrointestinal Cancer Symposium; Abstract #5
,
Orland, USA
. 
2008
.
27
Ohtsu
A
,
Shimada
Y
,
Shirao
K
, et al
. 
Randomized phase III trial of fluorouracil alone versus fluorouracil plus cisplatin versus uracil and tegafur plus mitomycin in patients with unresectable, advanced gastric cancer: the Japan Clinical Oncology Group Study (JCOG9205)
.
J Clin Oncol
2003
;
21
:
54
9
.
28
Yamao
T
,
Shimada
Y
,
Shirao
K
, et al
. 
Phase II study of sequential methotrexate and 5-fluorouracil chemotherapy against peritoneally disseminated gastric cancer with malignant ascites: a report from the Gastrointestinal Oncology Study Group of the Japan Clinical Oncology Group, JCOG 9603 Trial
.
Jpn J Clin Oncol
2004
;
34
:
316
22
.
29
Lutz
MP
,
Wilke
H
,
Wagener
DJ
, et al
. 
Weekly infusional high-dose fluorouracil (HD-FU), HD-FU plus folinic acid (HD-FU/FA), or HD-FU/FA plus biweekly cisplatin in advanced gastric cancer: randomized phase II trial 40953 of the European Organisation for Research and Treatment of Cancer Gastrointestinal Group and the Arbeitsgemeinschaft Internistische Onkologie
.
J Clin Oncol
2007
;
25
:
2580
5
.
30
Kawakami
K
,
Watanabe
G
. 
The association of thymidylate synthase mRNA expression with its three gene polymorphisms in colorectal cancer
.
Proc Am Assoc Cancer Res
2004
;
45
:
2104
.
31
Mandola
MV
,
Stoehlmacher
J
,
Zhang
W
, et al
. 
A 6 bp polymorphism in the thymidylate synthase gene causes message instability and is associated with decreased intratumoral TS mRNA levels
.
Pharmacogenetics
2004
;
15
:
319
27
.
32
Marsh
S
,
McKay
JA
,
Cassidy
J
,
McLeod
HL
. 
Polymorphism in the thymidylate synthase promoter enhancer region in colorectal cancer
.
Int J Oncol
2001
;
19
:
383
6
.
33
Pullarkat
ST
,
Stoehlmacher
J
,
Ghaderi
V
, et al
. 
Thymidylate synthase gene polymorphism determines response and toxicity of 5-FU chemotherapy
.
Pharmacogenomics J
2001
;
1
:
65
70
.
34
Park
DJ
,
Stoehlmacher
J
,
Zhang
W
,
Tsao-Wei
D
,
Groshen
S
,
Lenz
HJ
. 
Thymidylate synthase gene polymorphism predicts response to capecitabine in advanced colorectal cancer
.
Int J Colorectal Dis
2002
;
17
:
46
9
.

Supplementary data