Journal of Pharmacogenomics & Pharmacoproteomics

5-Fluorouracil (5-FU) is a cytotoxic drug which inhibits the production of pyrimidine. Their indications are the treatment of several cancers such as head and neck tumor (squamous cell carcinoma), nasopharyngeal carcinoma, breast cancer, colon cancer, stomach cancer, hepatoma and cervical cancer. The severe side effects such as neutropenia, diarrhea and mucositis occurred in some patients even though they received normal dose of 5-FU. The cause of these severe adverse reactions is Dihydropyrimidine Dehydrogenase (DPD) deficiency [1].

Dihydropyrimidine Dehydrogenase (DPD) deficiency is a genetic disease caused by molecular defects in the Dihydropyrimidine Dehydrogenase (DPYD) gene. Because Dihydropyrimidine Dehydrogenase (DPD) is the rate-limiting enzyme in the catabolism of pyrimidines and pyrimidine base analogs such as 5-FU, cancer patients with DPD deficiency developed severe toxicities after receiving 5-FU based chemotherapy. Some toxic deaths were occasionally reported. To prevent this fatal toxicity, oncologists should know beforehand the genetic makeup of patient and 5-FU must be avoided. Appropriate pretreatment screening in every patient who plans to receive 5-FU could be very helpful. The prevalence of complete and partial DPD deficiency in Caucasian population varied between 0.1 and 3% [2] which these patients have 3.4 times higher risk of developing severe neutropenia compare with patients who have normal DPD [3].

There are more than thirty forms of DNA sequences of DPYD gene which have been identified in several researches. One percent of Caucasian population have DPYD*2A mutation (heterozygous mutation) [2] which is the most common mutation in DPD deficiency patients. 0.2 percent of Japanese population has homozygous mutation [4]. Furthermore in Turkish cancer patients, there are 1.5% (3/200) heterozygous mutation at 2q (1-q) and 0.000055% (1/18,043) homozygous mutation at 2q [5]. Th e IVS14+1G>A mutation is the most common mutation pattern. Eleven known mutations have been found in Asian population (IVS14+1G>A, 62G>A [6], 74A>G [4], 85T>C[DPYD*9A], 812delT [4], 1003G>T [7], 1156G>T [7], 1627A>G [DPYD*5] [8-12], 1714C>G [4], 1897delC [DPYD*3] [8-12] and 2194G>A [DPYD*6]. Four mutations are found in Taiwanese population [74A>G, 85T>C, 1627A>G, and 2194G>A] [12].

However the extent and mutation analysis of DPD deficiency in Thai population has never been investigated. The objective of this research is the determination of the DPYD gene polymorphisms in Thai patients treated with 5-FU based regimens at Ramathibodi Hospital. In addition, these patients may also have DPD deficiency. Furthermore, this study will determine the extent of other DPYD gene polymorphisms patterns of Thai patients with DPD deficiency by genetic analysis.

Study design and clinical specimens

This study was retrospective and descriptive translational study. One hundred and six teen patients who received 5-FU based regimen were enrolled in this study. Seventy six patients developed grade 3-4 toxicities such as ANC less than 1,000 cells/mm³, diarrhoea more than 7 times/day or life threatening conditions, encephalopathy, mucositis and anorexia after the first or second cycle of chemotherapy. Clinical patient documentation for all samples included hospital number, sex, age, diagnosis, staging, chemotherapy regimen, dose, body surface area, side effect, complete blood count before and two weeks after receiving chemotherapy, the affected cycle, and other prior treatments such as surgery and radiation. Patients were excluded if they refuse and withdraw from the study or patients received concurrent 5-FU based chemotherapy and radiotherapy. Patient taking other medicines that are known to inhibit DPYD gene activity are also excluded. The other forty patients had no serious side effects from 5-FU. Inform consent was performed.

Sequencing analysis

5 ml EDTA blood were obtained for genomic DNA isolation with In House Modified method and then stored at -20°c for DPYD gene mutations analysis. The first step is amplifying genomic DNA by PCR technique which use primers for eleven known mutations in Asian population on exon 1, 8, 10, 11, 13, 14 and 17 (Figure 1); IVS14+1G>A, 62G>A [7], 74A>G [4], 85T>C[DPYD*9A], 812delT [4], 1003G>T [7], 1156G>T [7], 1627A>G [DPYD*5] [8-12], 1714C>G [4], 1897delC [DPYD*3] [8-12] and 2194G>A [DPYD*6] then purifying PCR product by biogel-P100. DNA sequencing was performed by Big Dye Terminator Cycle Sequencing Kit and ABI 3100 Genetic Analyzer. Bioedit program was used for analysis the DNA sequencing result.

Figure 1: The agarose gel of amplifying genomic DNA by PCR technique which use primers for eleven mutations; IVS14+1G>A, 62G>A(7), 74A>G(4), 85T>C(DPYD*9A), 812delT(4), 1003G>T(7), 1156G>T(7), 1627A>G (DPYD*5), 1714C>G, 1897delC (DPYD*3) and 2194G>A (DPYD*6), on exon 1, 8, 10, 11, 13, 14 and 17.

Statistical analysis

Statistical analysis was performed by using SPSS version 15.0. The actual change of ANC, hematocrit, platelet and percentage of neutrophil were compared in each SNP using Mann Whitney U test. P value less than 0.05 was considered significant.

In this study, we enrolled 116 patients including 60 cases [51.7%] of breast cancer, 41 cases [35.3%] of gastrointestinal tract cancer, 14 cases [12.06%] of head and neck cancer patients and 1 case [0.94%] of squamous cell cancer at extremity. All of them received various regimens of chemotherapy as shown in Table 1 and Figure 2.

Table 1: Number of patients who developed grade 3, 4 toxicities or who had no serious adverse reaction from various 5-FU based regimens chemotherapy.

Figure 2: Diagram shows the incidence of toxicities and various regimens of chemotherapy with no statistical significant differences in toxicities among various regimens of chemotherapy.

Seventy six patients developed severe toxicities after receiving the first or second cycle of chemotherapy such as 52.58% of patients had ANC < 1,000 cells/mm³, 4.31% had diarrhea, 0.86% had encephalopathy, 2.59% had diarrhea with ANC < 1,000 cells/mm³, 0.86% had mucositis, 2.58% had anorexia and 1.72% had encephalopathy with ANC < 1,000 cells/mm³ as shown in Table 2. In addition, we found that there were no statistical significant differences in toxicities among various regimens of chemotherapy.

Table 2: Number and percentage of patients who developed grade 3-4 toxicities after the first or second cycle of 5-FU based regimens chemotherapy.

We found totally thirteen SNPs of DPYD gene on chromosome1 of which six SNPs were in exon 8, 10, 11, 13,14 (967G>A, 1011A>T, 1236G>A, 1774C>T, 1896T>C and 1627A>G) and seven SNPs were in intron (IVS14+1G>A, IVS8-118A>G, IVS8-17A>C, IVS11-15T>C, IVS14-68T>G, IVS14+19C>A, IVS14+134G>T). Furthermore, we also found ten novel SNPs from thirteen SNPs; 967G>A, 1011A>T, 1236G>A, 1774C>T, IVS8-118A>G, IVS8-17A>C, IVS11-15T>C, IVS14-68T>G, IVS14+19C>A, IVS14+134G>T of which the first four SNPs were found in exon. Here in, we correlated 2 different single nucleotide exchanges and 4 novel SNPs of the DPYD gene with the appearance of toxicity including; actual change of Absolute Neutrophil Count (ANC), hematocrit, platelet and percentage of neutrophil in each SNP. The DNA sequencing of all novel SNPs (967G>A, 1011A>T, 1236G>A, 1774C>T) has been done.

The 967G>A polymorphism which located in exon 10 and leading to amino acid change from alanine to threonine had in only one patient who carriers 967GA genotype with severe toxicity (ANC grade 3,897 cells/mm³). The 967A allele frequency was 0.43%. For SNP 1011A>T change in exon 10, which had a minor allele frequency (1011T) 0.43%. We found heterozygous AT genotype in only one patient without appearance of severe toxicities. One patient with SNP 1236G>A (1236GA genotype) had severe toxicity (ANC grade 4; 95.5 cells/mm³). The allele frequency of A was 0.43%. No novel amino acid change was found in this SNP. The SNP (1774C>T in exon 14), had only one patient who held 1774C>T genotype with severe toxicity (ANC grade 3; 586 cells/mm³). The minor allele frequency (1774T) was 0.43%. The base was changed from CGG to TGG that made the amino acid changed from arginine to tryptophan and also the group of amino acid was changed from basic amino acid group to aromatic amino acid group. The SNP 1896T>C, we found heterozygous TC in 30 patients (25.86%), homozygous CC in 3 patients (2.59%) and wild type TT in 83 patients (71.55%). There were 18 patients and one patient developed severe toxicities in heterozygous TC group and homozygous CC group respectively. The 1896C-allele was 15.52%and no novel amino acid change was observed. The last SNP (1627A>G in exon 13), the following frequencies were observed; 43 patients were AG heterozygous [37.07%], 4 patients were homozygous GG (3.45%) and 69 patients with wild type AA (59.48%). All of four patients in homozygous GG group developed severe toxicities. The base was changed from ATA to GTA that made the amino acid changed from isoleucine to valine. But these two amino acids are in aliphatic amino acid group. However, we found that the patients who had genotype homozygous GG in this SNP had the very low median nadir ANC (399.6 cells/mm³) after two weeks of the first or second cycle of 5-FU based regimen chemotherapy. Furthermore, we also found statistically significant difference in actual ANC change and percentage of neutrophil change in homozygous GG (P = .011 and .009) as shown in Table 3 and 4.

Table 3: The values of median nadir ANC, actual decreased change of ANC and actual decreased change of percentage of neutrophil in 1627A>G categorized by the genotype.

Table 4: The correlation of genotype and phenotype in seven SNPs which found in exon.

The variations in the DPYD gene have been shown to manipulate the breakdown of the common anticancer drug within 5-FU base regimen and to aggravate severe drug-adverse effects during systemic 5-FU-application in cancer patients. Thus, our study were approved to predict the patients who might have severe toxicities (grade3-4 toxicities) such as; neutropenia, diarrhea, mucositis, anorexia and encephalopathy after 5-FU based regimen of chemotherapy by finding the common DPYD gene polymorphisms in Thai patients including analysis the correlation between genotype and phenotype of each polymorphism. In addition, these patients might have DPD deficiency which is one of genetic disease caused by molecular defects in the dihydropyrimidine dehydrogenase gene.

This is the first study in which the mutant frequencies of these eleven mutations as reported in Asian population have been determined in Thai patients. We found totally thirteen SNPs on the first chromosome which six SNPs were in exon and seven SNPs were in intron. In addition, four novel SNPs were found in exon and six novel SNPs were found in intron. The polymorphism was not detected in exon 1 and 17 in this study.

Furthermore, 1627A>G, 967G>A, 1774C>T and IVS14+G>A might be the cause of DPD deficiency in Thai patients due to several reasons. The first reasons is; homozygous GG of 1627A>G, heterozygous GA of 967G>A, heterozygous CT of 1774C>T and IVS14+G>A consisted of seven patients who developed severe toxicities after 5-FU based regimen of chemotherapy. The second reason is; the first three SNPs made amino acid and group of amino acid change. Even though in 1627A>G, the amino acid group was not changed. But this SNP is the only one polymorphism in this study that homozygous GG group had the very low median nadir ANC (399.6 cells/mm³) and statistically significant difference in actual ANC change and percentage of neutrophil change (P = 0.011 and 0.009). In addition, IVS14+G>A is an intron splice site which this polymorphism can change the structure of normal protein to truncated protein that might be the cause of DPD deficiency as described in the previous literature [3,6].

Potential limitations of the present study should be considered as a result of the retrospective nature. The novel SNPs were not analyzed in tumor tissue and therefore a correlation between each polymorphism and mRNA expression levels and protein function could not be assessed. Therefore, these findings have to be interpreted with caution. Further study needs to establish the functional DPD protein in this population and the correlation between the patterns of mutation and phenotype maybe useful for determining DPD deficiency patients in the future before receiving 5-FU based regimen chemotherapy to prevent the severe toxicities and complication. Moreover, the assays used, sample size, and the treatment regimen used may have been confounding factors.

In conclusion, our results demonstrate a limited overall impact of DPYD polymorphisms on the risk for developing FU related toxicity. Thus, all information which we found in this study suggested that 1627A>G, 967G>A, 1774C>T and IVS14+G>A might be the cause of DPD deficiency in Thai patients. And two from these four SNPs are the novel polymorphisms. Nevertheless, prospective study with large-scale population screening is required before clinical application of DPYD as a predictive marker for toxicity in patients considered for 5-FU-based chemotherapy.

This study was supported financially by the Pharmacogenomics Project, The collaborative Project from the Thailand Center of Excellence for Life Science (TCELS) and Mahidol University (MU), The Thailand Research Fund Office of the Higher Education Commission.