Dihydropyrimidine dehydrogenase (DPYD) gene c.1627A>G A/G and G/G genotypes are risk factors for lymph node metastasis and distant metastasis of colorectal cancer.

BACKGROUND: Dihydropyrimidine dehydrogenase (DPD) acts as the key enzyme catabolizing pyrimidines, and may affect the tumor progression. DPYD gene mutations affect DPD activity. The relationship between DPYD IVS14+1G>A, c.1627A>G, c.85T>C and lymph node metastasis (LNM) and distant metastasis (DM) of colorectal cancer (CRC) was investigated.
METHODS: A total of 537 CRC patients were enrolled in this study. DPYD polymorphisms were analyzed by polymerase chain reaction (PCR)-Sanger sequencing. The relationship between DPYD genotypes and clinical features of patients, metastasis of CRC was analyzed.
RESULTS: About DPYD c.1627A>G, A/A (57.7%) was the most common genotype, followed by A/G (35.6%), G/G (6.7%) genotypes. In c.85T>C, T/T, T/C, and C/C genotypes are accounted for 83.6%, 16.0%, and 0.4%, respectively. Logistic regression analysis revealed that DPYD c.1627A>G A/G and G/G genotypes in the dominant model (A/G + G/G vs. A/A) were significant risk factors for the LNM (p = 0.029, OR 1.506, 95% CI = 1.048-2.165) and DM (p = 0.039, OR 1.588, 95% CI = 1.041-2.423) of CRC. In addition, DPYD c.1627A>G polymorphism was more common in patients with abnormal serum carcinoembryonic antigen (CEA) (>5 ng/ml) (p = 0.003) or carbohydrate antigen 24-2 (CA24-2) (>20 U/ml) level (p = 0.015).
CONCLUSIONS: The results suggested that DPYD c.1627A>G A/G, G/G genotypes are associated with increased risk of LNM and DM of CRC.

INTRODUCTION

With the burden of cancer morbidity and mortality rapidly growing worldwide, cancer is a major barrier to increasing life expectancy worldwide. Colorectal cancer (CRC) is one of the most common gastrointestinal malignancies. According to the Global Cancer Statistics in 2020 by International Agency for Research on Cancer (IARC), CRC is the third most prevalent cancer and the second leading cause of cancer death in the world. In clinical treatment, CRC can be treated with endoscopic treatment, surgical resection, chemotherapy drugs, targeted drugs, immunotherapy, and radiation. , The multiple disciplinary team (MDT) model also improved the treatment level of CRC. However, the recurrence and metastasis of CRC are the major problems affecting the survival of the patients. Metastasis is the process by which cancer cells spread from the primary lesion to the distal organs and is the leading cause of cancer mortality. Metastasis of CRC includes lymph nodes metastasis (LNM) and distant metastasis (DM).

Capecitabine is an oral prodrug of 5‐fluorouracil (5‐FU) and has been approved for the treatment of various malignancies. There has been reports that the curative effect and toxic effects of 5‐FU exist noticeable individual differences. After fluorouracil administration, 5‐FU can be transformed into 5‐fluoro‐2'‐deoxyuridine 5’ monophosphate (FdUMP), 5‐fluoro‐2'‐deoxyuridine 5'‐triphosphate (FdUTP), and 5‐fluorouridine 5'‐triphosphate (FUTP) in cells, which are three cytotoxic metabolites. FdUMP inhibits the thymine ceoxyribonucleotide synthetase, the enzyme is necessary for DNA replication and repair, while FdUTP and FUTP disrupt the processing and function of DNA and RNA. Dihydropyrimidine dehydrogenase (DPD) is a rate‐limiting enzyme in the catabolic pathway of fluorouracil. DPD can inactivate up to 85% of 5‐Fu into 5, 6‐dihydro‐5‐fluorouracil, and the intermediate is further metabolized to β‐alanine or β‐aminoisobutyric acid. These processes will increase nucleotide synthesis, which is conducive to DNA synthesis and cell growth. While DPD enzyme activity is decreased, fluorouracil clearance rate in vivo is decreased, the half‐life is prolonged and cytotoxicity is enhanced. DPD enzyme activity is affected by DPYD gene polymorphisms. In addition, DPD is associated with epithelial‐to‐mesenchymal transition (EMT). EMT has been implicated in carcinogenesis and tumor metastasis by enhancing mobility, invasion, and resistance to apoptotic stimuli. DPYD gene polymorphisms may affect the process of EMT by changing the activity of DPD, thus participating in the metastasis of tumor cells.

The human DPYD gene is located on chromosome 1p21.3, it is 850 kb in length encompassing 23 exons. Genetic variations of DPYD lead to changes in DPD enzyme activity, which could result in some adverse side effects. The DPYD gene has more than 1700 different genetic variants, and more than 600 are missense variants impacting on the DPD protein sequence, according to the report in the GnomAD database (https://gnomad.broadinstitute.org/). So far, the variants or polymorphisms of DPYD gene attracted more attention including: DPYD IVS14+1 G>A (rs3918290, DPYD *2A), DPYD c. 1627 A>G (rs1801159, DPYD *5A), DPYD c. 85 T>C (rs1801265, DPYD *9A). ,

Studies have shown that the clinical outcome, the survival of CRC is associated with gene polymorphisms and gene expression level. One study showed that polymorphisms of DPYD have a significant effect on toxicity and clinical outcome in colorectal or gastroesophageal cancer patients receiving capecitabine‐based chemotherapy. Another study showed that the mRNA expression of DPYD is associated with clinicopathological characteristics and may be useful for predicting survival in CRC patients. The relationship between DPYD gene polymorphisms and metastasis of CRC has not been studied. In the present study, the relationship between DPYD gene polymorphisms and the clinical features of CRC patients, metastasis of CRC (including LNM and DM) was analyzed. It is expected to provide a valuable marker for the prognosis of CRC and a valuable target for the clinical treatment of metastatic CRC. This study may provide a valuable reference for the relationship between gene polymorphism and pathological features and metastasis of CRC.

MATERIALS AND METHODS

Subjects

A total of 537 CRC patients were recruited from Meizhou People's Hospital, from January 2016 to May 2019. Inclusion criteria: (1) Imaging diagnosis and histologically confirmed diagnosis met the diagnostic criteria for CRC. (2) Patients without serious cardiovascular and cerebrovascular diseases and infectious diseases. Exclusion criteria: (1) Patients without colorectal cancer. (2) Patients with dysfunction of vital organs. (3) Patients who also have other tumors. This study was supported by the Ethics Committee of the Meizhou People's Hospital. The flow chart of the present study is shown in Figure 1.

FIGURE 1

The flow chart of the present study

Genotyping of DPYD gene

Two milliliters of venous blood sample were obtained from each subject. Genomic DNA was extracted using a QIAamp DNA Kit (Qiagen GmbH). DPYD IVS14+1 G>A variant and polymorphisms of DPYD c. 1627 A>G and DPYD c. 85 T>C were analyzed. DPYD Genotyping Test Kit (SINOMD Gene Detection Technology Co. Ltd.) based on Sanger sequencing was used for testing. Polymerase chain reaction (PCR) was performed according to the following procedure: Initial denaturation at 95℃ for 3 min, followed by 45 cycles of denaturation at 94℃ for 15 s, annealing at 63℃ for 1 min, and extension at 72℃ for 1 min. PCR products were purified with ExoSap‐It (ABI PCR Product Cleanup Reagent). DNA sequences determination was detected using ABI Terminator v3.1 Cycle Sequencing kit and performed on ABI 3500 Dx Genetic Analyzer, analyzed with Sequencing Analysis v5.4 (Life Technologies).

Data collection and statistical analysis

Relevant information and medical records of these participants were collected. Clinical information, including age, gender, histopathological type, degree of tumor differentiation, TNM stage, and tumor grade, was collected. SPSS statistical software version 21.0 (IBM Inc.) was used for the data analysis. The Hardy–Weinberg equilibrium (HWE) of DPYD genotypes was assessed using the χ2 test. Association between DPYD variants status with the clinical features of patients and metastasis of CRC were evaluated by Fisher's exact test. A p value <0.05 was set as statistically significant.

RESULTS

Population characteristics

A total of 537 CRC patients were enrolled in this study, including 349 (65.0%) men and 188 (35.0%) women. The average age of the patients was 59.34 ± 10.14 years (26–85 years), 273 (50.8%) patients with ≤60 years old, and 264 (49.2%) patients with >60 years old. According to the pathological degree of tumor differentiation, 8 (1.5%) samples were well‐differentiated tumors, 497 (92.5%) samples were moderately differentiated tumors, 26 (5.0%) samples were poorly differentiated tumors, and 6 samples were unknown. According to the tumor stage, 3 (0.6%), 27 (5.0%), 364 (67.8%), and 142 (26.4%) cases were pT1, pT2, pT3, and pT4 stage, respectively. The proportion of higher stage tumors (pT3+ pT4 categories) was 94.2%. According to the lymph nodes status, 192 (35.8%), 196 (36.5%), 145 (27.0%), and 4 (0.7%) cases were N0, N1, N2, and N3 stage, respectively. In addition, 428 (79.7%) and 109 (20.3%) cases were M0 and M1 stage, respectively (Table 1).

TABLE 1

Baseline characteristics of study objects

	Colorectal cancer patients (n = 537)
Gender
Male, n (%)	349 (65.0)
Female, n (%)	188 (35.0)
Age, mean ± SD (years)	59.34 ± 10.14 (26–85)
≤60, n (%)	273 (50.8)
>60, n (%)	264 (49.2)
Differentiation
Well, n (%)	8 (1.5)
Moderate, n (%)	497 (92.5)
Poor, n (%)	26 (5.0)
Unknown, n (%)	6 (1.0)
T stages
pT1, n (%)	3 (0.6)
pT2, n (%)	27 (5.0)
pT3, n (%)	364 (67.8)
pT4, n (%)	142 (26.4)
Unknown, n (%)	1 (0.2)
N stages
N0, n (%)	192 (35.8)
N1, n (%)	196 (36.5)
N2, n (%)	145 (27.0)
N3, n (%)	4 (0.7)
M stages
M0, n (%)	428 (79.7)
M1, n (%)	109 (20.3)

The frequency of DPYD gene polymorphisms in the patients

In this study, the DPYD IVS14+1 G>A, DPYD c. 1627 A>G, DPYD c. 85 T>C genotypes in the patients were identified. About the DPYD IVS14+1G>A variant, there were 537 (100%) cases with G/G genotype (wild type), 0 (0%) cases with G/A heterozygous, and 0 (0%) cases with A/A homozygous. That is to say, no DPYD IVS14+1G>A mutation was found in the patients in this study. In the DPYD c.1627A>G, there were 310 (57.7%) cases with A/A genotype (wild type), 191 (35.6%) cases with A/G heterozygous, and 36 (6.7%) cases with G/G homozygous. Among DPYD c.85T>C, there were 449 (83.6%) cases with T/T genotype (wild type), 86 (16.0%) cases with T/C heterozygotes, and 2 (0.4%) cases with C/C homozygous. The genotype distributions of DPYD c.1627A>G, and DPYD c.85T>C in the CRC patients were consistent with Hardy–Weinberg equilibrium (χ2 = 0.425, p = 0.802 and χ2 = 0.715, p = 0.750, respectively).

Association of DPYD polymorphisms with metastasis of CRC

Logistic regression analysis of the relationship between the genotype of DPYD polymorphisms and the LNM status of CRC was studied. The frequency of DPYD c.1627A>G A/G genotype (39.4%) in the LNM group was obviously higher than that (28.6%) in the non‐LNM CRC patients. It was demonstrated that the A/G genotype of DPYD c.1627A>G might increase the risk of LNM in CRC patients (p = 0.016, OR = 1.626, 95% CI = 1.104–2.395). The variants were analyzed under different genetic models. It was showed that DPYD c.1627A>G A/G and G/G genotypes in the dominant model (DPYD c.1627A>G A/G + G/G vs. DPYD c.1627A>G A/A) were the significant risk factors (p = 0.029, OR 1.506, 95% CI = 1.048–2.165) for the LNM of CRC (Table 2).

TABLE 2

Association of DPYD polymorphisms with metastasis of CRC patients

Genotype	LNM n (%)	Non‐LNM n (%)	OR (95% CI)	p value	DM n (%)	Non‐DM n (%)	OR (95% CI)	p value
DPYD c. 1627 A>G
A/A	187 (54.2)	123 (64.1)	1.000 (reference)		53 (48.6)	257 (60.0)	1.000 (reference)
A/G	136 (39.4)	55 (28.6)	1.626 (1.104–2.395)	0.016	49 (45.0)	142 (33.2)	1.673 (1.079–2.596)	0.023
G/G	22 (6.4)	14 (7.3)	1.034 (0.509–2.097)	1.000	7 (6.4)	29 (6.8)	1.170 (0.487–2.813)	0.816
Dominant model (A/G+G/G vs. A/A)			1.506 (1.048–2.165)	0.029			1.588 (1.041–2.423)	0.039
Recessive model (G/G vs. A/A+A/G)			0.866 (0.432–1.735)	0.720			0.944 (0.402–2.217)	1.000
Allele frequency			1.280 (0.952–1.722)	0.104			1.333 (0.956–1.860)	0.094
A allele	510 (73.9)	301 (78.4)			155 (71.1)	656 (76.6)
G allele	180 (26.1)	83 (21.6)			63 (28.9)	200 (23.4)
DPYD c. 85 T>C
T/T	290 (84.1)	159 (82.8)	1.000 (reference)		90 (82.6)	359 (83.9)	1.000 (reference)
T/C	54 (15.7)	32 (16.7)	0.925 (0.574–1.492)	0.806	19 (17.4)	67 (15.7)	1.131 (0.647–1.979)	0.770
C/C	1 (0.3)	1 (0.5)	0.548 (0.034–8.825)	1.000	0 (0)	2 (0.5)	—	1.000
Dominant model (T/C+C/C vs. T/T)			0.914 (0.569–1.466)	0.716			1.098 (0.629–1.919)	0.772
Recessive model (C/C vs. T/T+T/C)			0.555 (0.035–8.927)	1.000			—	1.000
Allele frequency			0.909 (0.582–1.420)	0.731			1.056 (0.622–1.793)	0.891
T allele	634 (91.9)	350 (91.1)			199 (91.3)	785 (91.7)
C allele	56 (8.1)	34 (8.9)			19 (8.7)	71 (8.3)

Abbreviations: CRC, colorectal cancer; DM, distant metastasis; LNM, lymph node metastasis.

Bold numbers indicate significant values (p < 0.05).

Logistic regression analysis of the relationship between the genotype of DPYD polymorphisms and DM status of CRC was studied. The frequency of DPYD c.1627A>G A/G genotype (45.0%) in the DM group was obviously higher than that (33.2%) in the non‐DM group. It was demonstrated that the A/G genotype of DPYD c.1627A>G might increase the risk of DM in CRC patients (p = 0.023, OR = 1.673, 95% CI = 1.079–2.596). In addition, DPYD c.1627A>G A/G and G/G genotypes in the dominant model (DPYD c.1627A>G A/G + G/G vs. DPYD c.1627A>G A/A) were the significant risk factors (p = 0.039, OR = 1.588, 95% CI = 1.041–2.423) for the DM of CRC (Table 2).

Association of DPYD polymorphisms with clinicopathological parameters in the CRC patients

The association between DPYD c.1627A>G, c.85T>C polymorphisms, and clinicopathological features of CRC patients have been evaluated. The clinical features including gender, age, degree of differentiation of the tumor sample, serum tumor marker levels (carcinoembryonic antigen (CEA), carbohydrate antigen 24–2 (CA24‐2), carbohydrate antigen 19–9 (CA19‐9)), tumor stage, lymph nodes status, and distant metastasis status was collected. There was no relationship between the DPYD c.1627A>G, c.85T>C polymorphisms and gender, degree of differentiation of the tumor sample, serum CA19‐9 level, and tumor stage (T stage) of CRC patients. However, the frequency of DPYD c.1627A>G A/G+G/G genotypes in older patients (>60 years old) was significantly higher than that in the younger patients (≤60 years old) (p = 0.036). The frequency of DPYD c.1627A>G A/G+G/G genotypes in patients with abnormal serum CEA level (>5 ng/ml) and abnormal serum CA24‐2 level (>20 U/ml) was significantly higher than that in the patients with normal serum CEA level (≤5 ng/ml) (p = 0.003) and normal serum CA24‐2 level (≤20 U/ml) (p = 0.015), respectively (Table 3).

TABLE 3

Association of DPYD polymorphisms with clinicopathological parameters in the CRC patients

Parameters	DPYD c. 1627 A>G						DPYD c. 85 T>C
	Dominant model			Recessive model			Dominant model			Recessive model
	A/A	A/G+G/G	p value	A/A+A/G	G/G	p value	T/T	T/C+C/C	p value	T/T+T/C	C/C	p value
Gender
Male	200	149	0.855	325	24	0.859	289	60	0.542	348	1	1.000
Female	110	78		176	12		160	28		187	1
Age, years
≤60	170	103	0.036	259	14	0.168	228	45	1.000	271	2	0.499
>60	140	124		242	22		221	43		264	0
Differentiation
Well	4	4	0.242	6	2	0.069	8	0	0.329	8	0	1.000
Moderate	291	206		463	34		417	80		495	2
Poor	11	15		26	0		20	6		26	0
Serum CEA
≤5 ng/ml	244	152	0.003	370	26	0.845	332	64	0.895	394	2	1.000
>5 ng/ml	66	75		131	10		117	24		141	0
Serum CA24‐2
≤20 U/ml	290	198	0.015	459	29	0.036	407	81	0.697	486	2	1.000
>20 U/ml	20	29		42	7		42	7		49	0
Serum CA19‐9
≤37 U/ml	272	187	0.084	430	29	0.460	378	81	0.068	457	2	1.000
>37 U/ml	38	40		71	7		71	7		78	0
T stages
pT1‐2	15	15	0.447	28	2	1.000	27	3	0.450	30	0	1.000
pT3‐4	295	211		472	34		421	85		504	2
N stages
N0	123	69	0.029	178	14	0.720	159	33	0.716	191	1	1.000
N1‐3	187	158		323	22		290	55		344	1
M stages
M0	257	171	0.039	399	29	1.000	359	69	0.772	426	2	1.000
M1	53	56		102	7		90	19		109	0

Abbreviations: CA19‐9, carbohydrate antigen 19–9; CA24‐2, carbohydrate antigen 24–2; CEA, carcinoembryonic antigen.

Bold numbers indicate significant values (p < 0.05).

DISCUSSION

CRC is one of the common malignant tumors in human digestive tracts. , Metastasis is a biological phenotype of malignant tumors and an important factor affecting the prognosis of malignant tumors. Tumor metastasis is a dynamic process in which multiple factors are involved in multiple stages of development, including the biology of tumor cells and the interaction between tumor and microenvironment. , At present, the research on tumor metastasis mainly focuses on tumor metastasis genes and tumor metastasis suppressor genes, tumor angiogenesis, extracellular matrix, cell adhesion, tumor microenvironment, and so on. ,

Studies have shown that some gene polymorphisms were associated with the metastasis of cancer. It is a lower risk of LNM in oral cancer patients carrying A/A genotype of the single nucleotide polymorphism (SNP) rs10399805 or rs6691378 in chitinase‐3‐like protein 1 (CHI3L1) gene. Polymorphisms in the promoter regions of matrix metalloproteinase (MMP)1, 3, 7, and 9 genes are associated with metastasis of head/neck and breast cancer. Luminal A and luminal B breast cancer patients with the A/G genotype of C‐C motif chemokine ligand 4 (CCL4) gene SNP rs10491121 were less likely to develop LNM. The SNPs rs1143630, rs1143633, and rs1143643 of interleukin‐1 beta (IL‐1B) gene showed a relationship with LNM of papillary thyroid carcinoma (PTC). SNP rs1989839 C/T genotype of Ras‐association domain family 1 isoform A (RASSF1A) gene increases the risk of lung metastasis of osteosarcoma. Transforming growth factor‐β1 (TGFB1) gene promoter −509C/T polymorphism affected the metastasis of CRC. Granzyme B (GZMB) gene polymorphisms were not associated with the metastasis of CRC. Studies have shown that DPYD gene polymorphisms were associated with the susceptibility to CRC and the toxicity of chemotherapy drugs. However, the relationship between DPYD gene polymorphisms and metastasis of CRC has not been studied.

DPYD IVS14+1G>A variant was not found in this study, and this result was similar to those reported in other populations, such as Caucasians, African‐Americans, Egyptians, Turks, and Taiwanese. Many studies have reported that CRC patients with DPYD IVS14+1G>A variant might suffer from severe toxicity and even death after the 5‐FU administration. , However, DPYD IVS14+1G>A variant is rare in most populations. In this study, DPYD c.1627A>G, A/A, A/G, and G/G genotypes accounted for 57.7%, 35.6%, and 6.7%, respectively. The result is in line with those of another Chinese population study. DPYD c.85T>C T/T, T/C, and C/C genotypes accounted for 83.6%, 16.0%, and 0.4%, respectively. The result in this study was consistent with that in the previous study. A study of a population of a mixed racial background showed that DPYD c.85T>C T/C and C/C genotypes were 41% and 10%, respectively. The frequencies of DPYD c.85T>C variants in patients were higher than that in this study.

In this study, DPYD c.1627A>G A/G and G/G genotypes in the dominant model (A/G + G/G vs. A/A) were significant risk factors for the LNM and DM of CRC. DPD activity is in association with the epithelial‐to‐mesenchymal transition (EMT). EMT is a process during which the epithelial features of cancer cells are lost, the cytoskeletal architecture is re‐organized, the cell shape is changed, and some genes are activated, which leads to increased cell motility and dissemination of tumor to distant metastatic sites. EMT results in decreased adhesion and enhanced migration or invasion. Studies have shown that dihydrothymine and dihydrouracil, the metabolites catabolized by DPD, play an important role in tumor EMT. , DPD is necessary for cells to acquire mesenchymal characteristics in vitro and tumorigenic cells overflow. It is a metabolic process essential associated with the acquisition of metastatic and aggressive cancer cell traits for the EMT. Mechanistically, DPD may act as a regulator of EMT by targeting the p38/NF‐κB/Snail1 pathway.

In the present study, the frequency of DPYD c.1627A>G A/G+G/G genotypes in patients with abnormal serum CEA levels was significantly higher than that in patients with normal serum CEA levels. Serum CEA levels can be used as biomarkers for diagnosis, postoperative recurrence, or efficacy monitoring of colorectal cancer. The CEA gene family belongs to the immunoglobulin (Ig) superfamily and codes for a vast number of glycoproteins that differ greatly both in amino acid composition and function. The CEA family is divided into two groups, the carcinoembryonic antigen‐related cell adhesion molecules (CEA‐CAMs) and the pregnancy‐specific glycoproteins. CEA expression on epithelial cells may directly influence tumor development by CEA‐CEA bridges between tumor cells or tumor‐stromal cells. That is to say, DPYD gene mutations may affect the process of EMT by changing the activity of DPD, thus participating in the metastasis of tumor cells. Elevated CEA expression level and DPYD gene mutations may be associated with CRC metastasis.

CA24‐2 is a serum tumor marker, which is one of the indicators reflecting the number and activity of tumor cells. A study has shown that the CA24‐2 level was higher in gastric cancer patients with distant metastasis than in patients without distant metastasis. Increased serum CA24‐2 concentrations were significantly associated with the risk of invasiveness of intraductal papillary mucinous neoplasm (IPMN). CEA, CA19‐9, CA24‐2, and CA72‐4, examined postoperatively during follow‐up, were useful to find early tumor recurrence and metastasis, and evaluate prognosis. Tumorigenesis is dependent on the reprogramming of cellular metabolism. A common feature of metabolism in the cancer cells is the ability to acquire necessary nutrients from a frequently nutrient‐poor environment and utilize these nutrients to both maintain viability and build new biomass. Some studies have shown that Pantothenate and CoA biosynthesis signaling pathway was significantly altered in tumor cells. , , DPD is a key enzyme in the Pantothenate and CoA biosynthesis signaling pathway (https://www.genome.jp/pathway/ko00770+K00207). So, DPYD c.1627A>G A/G+G/G genotypes may affect the activity of DPD, and regulate tumor cells tumorigenesis through signaling pathway regulation in the reprogramming of cellular metabolism, which is manifested as changes in serum tumor markers.

Tumor invasion and metastasis is a dynamic and complex process, including multiple simultaneous steps. The persistent emergence of populations of cells with different invasion and metastasis capabilities is a barrier to tumor therapy. In order to prevent the invasion and metastasis of tumor, it is a hot spot of research to design modulatory blocking methods specifically aiming at some key links in tumor invasion and metastasis. With the deepening understanding of the occurrence and mechanism of tumor invasion and metastasis, it can promote the design and search for effective anti‐tumor drugs, provide new ideas for the treatment of tumors, and have a positive significance to reduce the mortality of tumor patients.

This is the first study about the relationship of DPYD gene variants/polymorphisms and lymph node metastasis, distant metastasis of CRC. There are some limitations to this study that should be noted. First of all, the number of cases included in this study is not large, which may lead to some deviations in the results. Second, the number of gene polymorphisms included in this study was relatively single. Tumor cell metastasis is affected by tumor metastasis‐related genes and tumor metastasis‐suppressor genes, tumor angiogenesis, extracellular matrix degradation, cell adhesion, tumor microenvironment, and other factors. It may be more meaningful to include some related genes for comprehensive analysis. In addition, a tumor is a kind of multifactorial disease caused by genetic and environmental factors. As a retrospective analysis, the limitations of the original data included in this study constrained assessment of potential gene‐environment interactions.

CONCLUSION

DPYD c.1627A>G A/G and G/G genotypes are associated with the increased risk of lymph node metastasis and distant metastasis of CRC. Future studies need to include more relevant genes for analysis and to assess potential gene‐environment interactions. This study may provide a valuable reference for the relationship between gene polymorphism and pathological features and metastasis of CRC.

CONFLICT OF INTEREST

The authors declare that they have no competing interests.

AUTHORS’ CONTRIBUTIONS

Zhixiong Zhong, Heming Wu, and Juanzi Zeng designed the study. Juanzi Zeng, Qingyan Huang, and Zhikang Yu performed the experiments. Juanzi Zeng and Jiaquan Li collected the clinical data. Heming Wu and Juanzi Zeng analyzed the data. Heming Wu and Juanzi Zeng prepared the manuscript. All authors were responsible for critical revisions, and all authors read and approved the final version of this work.