Implication of TYMP genetic variation on the prognosis of patients with colorectal cancer who received capecitabine-based adjuvant chemotherapy after surgical resection: a real-world exploratory study


 Background

Thymidine Phosphorylase (TYMP) gene was of crucial significance in the process of colorectal cancer (CRC) development and also played an important role in capecitabine metabolism. Present study was to identify the association between TYMP polymorphism and prognosis of postoperative patients with CRC who received capecitabine-based adjuvant chemotherapy.
Methods

A total of 218 patients with CRC who were treated with surgical resection and capecitabine-based adjuvant chemotherapy were included in this study retrospectively. Peripheral blood and peripheral blood mononuclear cell (PBMC) specimen of the patients with CRC were collected for the genotyping of TYMP polymorphism and TYMP mRNA expression, respectively. Univariate analysis of genotypes and prognosis was carried out by Kaplan-Meier survival analysis, and multivariate were adjusted by Cox regression analysis. Expression of TYMP according to genotype status was analyzed using non-parameter test.
Results

The median disease-free survival (DFS) of the 218 patients with CRC was 4.6 years, and the median overall survival (OS) was 5.8 years. Regarding the polymorphism analysis, only rs11479 was of clinical significance. The prevalence of rs11479 in TYMP among the 218 patients indicated that minor allele frequency was 0.20 (GG 141 cases, GA 68 cases and AA 9 cases), which accorded with Hardy-Weinberg Equilibrium (P=0.825). Prognostic analysis according to rs11479 genotype status suggested that the median DFS of patients with GG genotype and GA/AA genotype was 3.1 and 6.1 years, respectively (P=0.004). Furthermore, the median OS of patients with GG genotype and GA/AA genotype was 5.0 and 7.0 years, respectively (P=0.033). In order to adjust the confounding factors which might contribute to OS, a multivariate Cox regression analysis was introduced and the results exhibited that rs11479 polymorphism was an independent factor for DFS (HR=1.64, P=0.009). Additionally, of the 65 PBMC specimens, the mRNA expression results indicated that patients with GA/AA genotypes conferred significantly higher mRNA expression of TYMY than that of patients with GG genotype (P<0.001).
Conclusions

TYMP gene polymorphism rs11479 might involve in the prognosis of patients with CRC who received capecitabine based adjuvant chemotherapy through mediation of the mRNA expression of TYMP. Conclusion of present study should be confirmed in prospective clinical trial subsequently.

Implication of TYMP genetic variation on the prognosis of patients with colorectal cancer who received capecitabine-based adjuvant chemotherapy after surgical resection: a real-world exploratory study Thymidine Phosphorylase (TYMP) gene was of crucial signi cance in the process of colorectal cancer (CRC) development and also played an important role in capecitabine metabolism. Present study was to identify the association between TYMP polymorphism and prognosis of postoperative patients with CRC who received capecitabine-based adjuvant chemotherapy.

Methods
A total of 218 patients with CRC who were treated with surgical resection and capecitabine-based adjuvant chemotherapy were included in this study retrospectively. Peripheral blood and peripheral blood mononuclear cell (PBMC) specimen of the patients with CRC were collected for the genotyping of TYMP polymorphism and TYMP mRNA expression, respectively. Univariate analysis of genotypes and prognosis was carried out by Kaplan-Meier survival analysis, and multivariate were adjusted by Cox regression analysis. Expression of TYMP according to genotype status was analyzed using non-parameter test.

Results
The median disease-free survival (DFS) of the 218 patients with CRC was 4.6 years, and the median overall survival (OS) was 5.8 years. Regarding the polymorphism analysis, only rs11479 was of clinical signi cance. The prevalence of rs11479 in TYMP among the 218 patients indicated that minor allele frequency was 0.20 (GG 141 cases, GA 68 cases and AA 9 cases), which accorded with Hardy-Weinberg Equilibrium (P=0.825). Prognostic analysis according to rs11479 genotype status suggested that the median DFS of patients with GG genotype and GA/AA genotype was 3.1 and 6.1 years, respectively (P=0.004). Furthermore, the median OS of patients with GG genotype and GA/AA genotype was 5.0 and 7.0 years, respectively (P=0.033). In order to adjust the confounding factors which might contribute to OS, a multivariate Cox regression analysis was introduced and the results exhibited that rs11479 polymorphism was an independent factor for DFS (HR=1.64, P=0.009). Additionally, of the 65 PBMC specimens, the mRNA expression results indicated that patients with GA/AA genotypes conferred signi cantly higher mRNA expression of TYMY than that of patients with GG genotype (P<0.001).

Conclusions
TYMP gene polymorphism rs11479 might involve in the prognosis of patients with CRC who received capecitabine based adjuvant chemotherapy through mediation of the mRNA expression of TYMP. Conclusion of present study should be con rmed in prospective clinical trial subsequently.

Page 4/24
Background Colorectal cancer (CRC) is one of the most common gastrointestinal tumors worldwide. It is estimated that there are approximately 56,000 new cases and 29,000 new deaths of CRC in China currently according to the recent epidemiological data [1]. To our knowledge, surgical resection was established to be the only way to cure patients with CRC for decades [2]. Unfortunately, approximately 30% of the patients were diagnosed of distant metastases, thus missing the opportunity for surgical resection [3].
Consequently, almost half of the patients with CRC were able to receive surgical resection clinically. Recent years had witnessed in-depth development of molecular typing in the eld of colon cancer or rectal cancer with advanced stage, more and more targeted drugs with different targets of action were proved to provide considerable treatment options, thus bringing more survival bene ts for the patients consecutively [4]. Nevertheless, the meaningful research progress in early-stage and postoperative patients with CRC treatment was still limited comparatively. Only IDEA trial was proved to be the signi cant breakthrough with clinical signi cance recently, which suggested that 3 months rather than 6 months of adjuvant chemotherapy was recommended for low-risk stage colon cancer receiving capecitabine plus oxaliplatin regimen [5]. Conclusions in IDEA trial helped to attenuate the unnecessary side reactions of chemotherapy to some extent in view of the fact that 6 months adjuvant chemotherapy regimens had been established to be the standard of care for stage patients with CRC for decades [6]. Besides, with the adequate development of the more sensitive detection techniques, ctDNA had established to be the potential predictive value in the recurrence risk assessment for patients with CRC after surgical resection [7]. However, the predictive role regarding ctDNA for the intensity or duration of adjuvant chemotherapy among patients with CRC was still controversial and no standardized and uni ed criteria was available for the corresponding detection of ctDNA currently[8].
Capecitabine-based adjuvant chemotherapy had been established as the standard of care for the patients with CRC after surgical resection for decades, which improved 5-year survival rates of approximately 10% [9]. However, the bene ts of capecitabine-based adjuvant chemotherapy for patients with stage CRC regarding reducing recurrence and prolonging survival remained controversial currently [10]. Furthermore, large individual differences were still observed in the clinical application of capecitabine-based adjuvant chemotherapy, which suggested that many factors might compromise the prognosis of this regimen clinically. Capecitabine can be transformed into 5-FU only through the key metabolism enzyme of thymidine phosphorylase (TYMP), thus playing the cytotoxic effects to kill the tumor [11]. Furthermore, previous study indicated that TYMP gene was similar with the structure of platelet-derived endothelial cell growth factor in tumor cells, which played an important role in promoting tumor angiogenesis and metastasis [12]. Therefore, many studies had con rmed that TYMP was highly expressed in various kinds of tumors.
TYMP gene was located at chromosome 22q13.33 and contained 10 exons. As a member of the endothelial vascular growth factor family, TYMP gene was not conserved in Chinese population, which suggested that expression of TYMP also showed great individual differences among different individuals [13]. Relatively limited studies regarding the polymorphism study of this gene in Chinese population was observed. Especially in the eld of colorectal cancer, rare research had been performed on this gene polymorphism currently [14]. Previous work initiated by YB Du and colleagues implemented the exploration of TYMP genetic variation on clinical outcome and safety of CRC [15]. And they identi ed that polymorphism of TYMP could be of potential signi cance clinically. Besides, another study initiated by L Huang and colleagues investigated the clinical signi cance of TYMP gene polymorphism among patients with advanced gastrointestinal tumors [16]. Conclusion exhibited that the expression level of TYMP gene was higher among patients with T allele in the advanced gastrointestinal tumors. Nevertheless, the relevance of TYMP polymorphism to the prognosis of real-world patients with CRC who received capecitabine based adjuvant chemotherapy after surgical resection remained unknown.
Consequently, present study aimed to identify the association between TYMP polymorphism and prognosis of patients with CRC who received surgical resection and capecitabine-based adjuvant chemotherapy in real-world.

Patients
Given that capecitabine-based adjuvant chemotherapy was licensed in Chin over ten years, and considerable patients with CRC had received capecitabine-based adjuvant chemotherapy clinically.
Therefore, present study was designed as real-world retrospective research. Appropriate patients with CRC underwent surgical resection were collected in the department of Gastrointestinal surgery of A liated Hospital of Hebei University from January 2010 to March 2021. Baseline characteristics of the patients were obtained from the hospitals electronic medical record system. Patients who ful lled the eligibility criteria were included in the study retrospectively. Inclusion criteria manifested as: (1) aged ≥18 years; (2) ECOG performance status of 0-2 score; (3) adequate cardiac function, renal function and bone marrow function to receive adjuvant chemotherapy appropriately; (4) pathological diagnosis of colon cancer or rectal cancer; (5) received surgical resection and postoperative adjuvant chemotherapy; (6) pathological staging of or . The exclusion criteria included: (1) failed to receive capecitabine-based adjuvant chemotherapy (capecitabine monotherapy or capecitabine related regimens) after surgical resection; (2) peripheral blood specimen was not available for DNA extraction; (3) diagnosed of familial adenomatous polyposis or hereditary CRC; (4) concomitant with another tumor or serious diseases that might compromise the survival of the patients. Flow chart of present study was illustrated in Figure 1. Finally, a total of 218 patients with CRC patients met the eligibility criteria were enrolled in this study. The primary endpoint of present study was the association between prognosis and polymorphism genotype status. The study was approved by the ethics committee of the A liated Hospital of Hebei University. Informed consent was signed by each enrolled patient in accordance with the recommendation of the declaration of Helsinki. All the patients with CRC included in this study were treated with capecitabine-based adjuvant chemotherapy. The chemotherapy regimens included capecitabine monotherapy and CAPEOX combination therapy. And the usage and dosage of capecitabine monotherapy was as follows: 3-4 weeks after surgical treatment, capecitabine, 1000-1250 mg/m 2 , twice daily, day 1-14, every 21 days as one cycle. Additionally, the usage and dosage of CAPEOX regimen manifested as: capecitabine 1000mg/m 2 , twice daily, day 1 -14, every 21 days as one cycle, combined with oxaliplatin, 80-130 mg/m 2 , iv. infusion, day 1. The adjuvant chemotherapy period was planned for 8 cycles or depended on the actual situation of the patients.
Each patient was followed up at the onset of adjuvant chemotherapy administration. Initial follow-up was implemented when the patient received adjuvant chemotherapy in the hospital, where the baseline characteristics and the date of disease recurrence could be obtained through the electronic medical record system speci cally. Subsequent follow-up was performed mainly by telephone. Patients were followed up every three months for the date of recurrence and treatment after recurrence, and the death status were mainly inquired. Furthermore, when a patient was recurrence or death, it must be con rmed by at least two research colleagues before it could be con rmed as a recurrence events or death events.

Collection Of Peripheral Blood Specimens And Tymp Polymorphism Analysis
Peripheral blood specimen from each enrolled patient were collected when it was available. Genomic DNA was extracted using the traditional phenol chloroform method. Unfortunately, 19 patients failed to obtain the available peripheral blood specimen and 12 patients failed for the genomic DNA extraction. Consequently, a total of 218 patients with adjuvant chemotherapy had the suitable DNA specimens and included in the polymorphism analysis nally. The single nucleotide polymorphism included in this study were collected from the NCBI database with the minor allele frequency >10% among Chinese population or the previous study that identi ed the clinical signi cance of the polymorphisms. Consequently, polymorphisms included were rs11479, rs131804 and rs470119. In the preliminary analysis between the polymorphism status and prognosis, only rs11479 was signi cantly associated with prognosis. As a result, the subsequent analysis of present study was focused on rs11479 polymorphism accordingly.
The rs11479 polymorphism was genotyped using the method of polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). PCR product of rs11479 was ampli ed initially, the PCR product including this polymorphism was ampli ed, forward primer was 5'-TCTAACAGCCCCTCGCTCT-3', reverse primer was 5'-GGGTCACGTGTTCATCGAG-3'. Size of PCR product was 266bp. And a total of 2 µL PCR products were digested using the restriction enzyme NdeI (Thermo Fisher Scienti c, USA). Genotype status of rs11479 were distinguished through the size of PCR bands according to the previous study [17]. Additionally, Genotyping results for rs11479 were con rmed in some randomly selected samples using ionization-time-of-ight mass spectrometry (Sequenom, SanDiego, CA) method. And the two genotyping approaches reached a 100% consistency.
Peripheral blood Mononuclear Cell (PBMC) specimen collection and TYMP gene mRNA expression analysis As described in the ow chart of present study, PBMC specimens were collected in the 77 randomly selected subjects from the 218 patients with CRC initially. However, 7 patients' PBMC specimens were not available and 5 patients failed for RNA extraction experiment subsequently. Eventually, a total of 65 mRNA specimens were available for the subsequent analysis and preserved in Liquid nitrogen. Total RNA samples were extracted using TRizol reagents (Takara Biotechnology, China) according to the manufacture's instruction and stored at -80℃ for mRNA expression analysis. A total of 500ng RNA extracted from the PBMC specimens was used as the templates for reverse-transcription polymerase chain reaction. Relative quantitative analysis of TYMP gene mRNA expression was implemented using the LightCycler® 480 (Roche, Shanghai, China) with SYBR Premix EX Taq system. The forward primer of TYMP was 5'-ATGGCAGCCTTGATGACC-3', the reverse primer was 5'-TTATTGCTGCGGCGGCAG-3'.  As described in the method part, only TYMP rs11479 was of clinical signi cance preliminarily. The prevalence of rs11479 polymorphism among the 218 patients with CRC manifested as: GG genotype 141 cases (64.7%), GA genotype 68 cases (31.2%), AA genotype 9 cases (4.1%), minor allele frequency of rs11479 was 0.20. And the distribution of the three genotypes was in accordance with Hardy-Weinberg Equilibrium (P=0.825). Given that the frequency of AA genotype was relatively rare, patients with AA and GA genotypes were merged in the subsequent analysis. As presented in Table 2, patients with GG and GA/AA genotype were well balanced with similar baseline characteristics and no statistically signi cant difference was observed (P>0.05).

Implications of TYMP rs11479 polymorphism on the prognosis of the 218 patients with CRC
All the 218 patients with CRC included in this study were available for prognostic assessment. And the last follow-up date of present study was September 2021. The median follow-up duration from the patients included in this study to last follow-up date was 5.5 years (follow-up range: 0.25-10 years). Collectively, the prognostic data of the 218 patients with CRC was presented in Figure 3 and Figure 4. As shown in Figure 3, The median DFS of the 218 patients was 4.6 years [95% con dence interval (CI): 3.80-5.40], 3-year DFS rate was 59.04% (95%CI: 52.13%-65.29%) and 5-year DFS rate was 45.95% (95%CI: 38.89%-52.71%), respectively. Furthermore, as illustrated in Figure 4, the median OS of the 218-patient cohort was 5.8 years (95%CI: 5.12-6.48), the 3-year OS rate was 78.76% (95%CI: 72.60%-83.70%) and 5year OS rate was 54.60% (95%CI: 47.29%-61.32%).
In order to identify the clinical signi cance of TYMP rs11479 polymorphism, association analysis between genotype status of rs11479 and DFS was performed rstly. As exhibited in Figure 4, the median DFS of the patients with GG and GA/AA genotype of rs11479 polymorphism was 3.1 months (95%CI: 2.08-4.12) and 6.1 (95%CI: 4.65-7.55) months, respectively. And the 3-year DFS rate was 50.38% (95%CI: 41.78%-58.36%) and 74.97% (95%CI: 63.58%-83.25%), which was statistically signi cant difference (χ 2 = 8.164, P=0.004). Furthermore, regarding the association between rs11479 genotype status and OS, as illustrated in Figure 5, Furthermore, the median DFS according to different baseline characteristic subgroups in univariate analysis were performed simultaneously. As exhibited in Table 2, only ECOG PS score and pathological staging was signi cantly associated with DFS in univariate analysis according to baseline characteristics, which suggested that the median DFS of patients with ECOG 0 score was dramatically longer than that of patients with the 1-2 score (median DFS: 5.80 vs 3.89 months, P=0.009), the median DFS of patients with pathological staging of was signi cant longer than that of patients with (median DFS: 6.15 vs 3.89 months, P=0.002). Furthermore, a multivariate Cox regression model was introduced for DFS adjustment including the baseline characteristics which was signi cant in the univariate analysis as illustrated in Table 2. The multivariate analysis results were illustrated in Table 2, after the multivariate adjustment, the signi cant statistical difference was still observed regarding TYMP rs11479 polymorphism for DFS, which suggested that rs11479 polymorphism was an independent factor for DFS [hazard ratio (HR)=1.64, P=0.009). Besides, as presented in Table 2, after adjusted in the Cox regression analysis, both ECOG score (HR=0.65, P=0.015) and pathological staging (HR=0.57, P=0.007) were independent factors for DFS.

Relevance Of Rs11479 Polymorphism To Tymp Gene Mrna Expression
As mentioned in the methods part, a total of 65 PBMC specimens were available for TYMP gene mRNA expression analysis nally. The expression of TYMP gene mRNA was detected by extracting RNA from the 65 PBMC specimens and the relevance analysis between the genotype status of rs11479 polymorphism and TYMP mRNA expression was analyzed accordingly. Firstly, the genotype status of rs11479 polymorphism in the 65 PBMC specimens manifested as: GG genotype 42 cases (64.6%), GA genotype 20 cases (30.8%) and AA genotype 3 cases (4.6%). The MAF was 0.20 and distribution frequency of the three genotypes was in accordance with the Hardy-Weinberg equilibrium (P=0.756) as well, which were comparable to the genotype distribution among the 218 patients with CRC. Similarly, GA and AA genotypes were merged in the subsequent analysis. As illustrated in Figure 6, the relative expression of TYMP mRNA in PBMC of patients with GA/AA genotype was signi cantly higher than that of patients with GG genotype of rs11479 polymorphism (3.945±0.575 vs 3.191±0.726), which was statistically signi cant (t=4.294, P<0.001).

Discussion
Firstly, present study provided and highlighted the real-world evidence regarding the prognostic data of patients with CRC who were treated with surgical resection and capecitabine-based adjuvant chemotherapy in real-world. Simultaneously, the prognostic association analysis suggested that TYMP gene polymorphism rs11479 might involve in the prognosis of postoperative patients with CRC who received capecitabine based adjuvant chemotherapy through the mediation TYMP mRNA expression.
To the best of our knowledge, CRC was established to be a highly heterogeneous digestive system malignancy [19]. Comparatively limited research progresses that could strikingly improve the prognosis of the patients with CRC were observed in recent years [20]. At present, PD-1/PD-L1 blockades were reported to be effective for considerable patients with advanced stage of cancer which provided signi cant survival bene t to some extent [21]. Nevertheless, the availability of PD-1/PD-L1 blockades in advanced CRC was scanty, which indicated that only patients with dMMR might bene t from PD-1 blockade according to Keynote 177 clinical trial [22]. As a result, traditional chemotherapy still played an important role in the treatment for CRC. However, it should be noticed that ORR of chemotherapy was relatively limited, which exhibited the urgent exploration of potential biomarkers that could predict the e cacy of traditional chemotherapy was an important research direction [23]. At present, biomarkers that might predict the effectiveness to traditional chemotherapy were predominately focused on genetic polymorphisms, ctDNA and somatic gene mutations [24]. Speci cally, a recently reported polymorphism study initiated by JS Su and colleagues indicated that PD-L1 gene polymorphism 901T>C could be used as a potential biomarker to involve in the prognosis of patients with CRC receiving capecitabine-based adjuvant chemotherapy through mediation of the mRNA expression of PD-L1 [25]. Furthermore, another study initiated by JT and colleagues investigated the clinical signi cance of ctDNA for guiding the prognosis of patients with high-risk stage colon cancer, which indicated that the monitoring of ctDNA might be useful to predict the prognosis of patients with high-risk stage colon cancer [26]. Additionally, AJ Li and colleagues performed a study to explore the predictive association between PIK3CA and TP53 somatic mutations status and OS for patients with stage and CRC [27]. And the results suggested that patients with PIK3CA and TP53 double mutations were correlated with worse OS. Collectively, all the above ndings demonstrated that genome DNA status of colon cancer or rectal cancer might predict the prognosis of patients with CRC who were treated with conventional chemotherapy to some extent [28].
Although present study was designed as a retrospective analysis, we still carried out the prognostic analysis of the 218 patients with CRC who received capecitabine-based adjuvant chemotherapy. And the results exhibited that the median DFS of the 218 patients was 4.6 years (95%CI: 3.80-5.40) and the median OS of the 218-patient cohort was 5.8 years (95%CI: 5.12-6.48). The DFS and OS data in our study seemed to be lower than that in NO16968 clinical trial which was implemented to identify the feasibility of oxaliplatin combined with capecitabine as adjuvant therapy for stage CRC [29]. The discrepancy between the two study could be attributed to the following explanation: our study included more patients with an ECOG performance status of 2 score than that in the NO16968 study and the results of the Cox analysis in our study indicated that patients with ECOG of 2 score conferred a worse prognosis, which was consistent with the previous study [30]. Besides, it should be noted that this study was designed as a retrospective analysis. Management of the patients in retrospective study was not su cient and normative compared with well-designed phase clinical trial, which was also re ected in the fact that the actual completion of capecitabine monotherapy regimen in this study was only 5 cycles, and the CAPEOX regimen was only 4 cycles. The insu cient adjuvant chemotherapy might stand a good chance to compromise the survival of the patients [31].
Considerable previous studies indicated that genetic variation of drug metabolism gene might contribute to the effectiveness of a variety of drugs [32]. To our knowledge, our study might be the rst exploration disclosed that the carriers of the T allele at rs11479 of TYMP gene could be sensitive to capecitabine administration and bene t from capecitabine adjuvant chemotherapy in Chinese patients with CRC who received capecitabine based adjuvant chemotherapy by in uencing the mRNA expression of TYMP.
Interestingly, a previous study initiated by YB Du and colleagues was reported [15]. A total of 235 patients with CRC underwent surgical treatment were included in their study retrospectively and they investigated the in uence of TYMP genetic variation on clinical outcomes of patients with CRC. The conclusion in their study suggested that 5633C>T of TYMP might impact the prognosis of the patients, which was consistent with the design and results of our study. were included and the results suggested that T allele gene carriers were associated with higher TYMP gene expression, which was consistent with the mRNA expression results in our study. However, they failed to identify the positive association of the polymorphism with OS mainly own to the relatively limited sample size. Furthermore, a previous study initiated by BA Jennings and colleagues found that T allele gene carriers of rs11479 were correlated with higher incidence of adverse reaction among patients with CRC receiving capecitabine therapy, which was partly in line with the results of our study, highlighting the possibility that T allele carriers of rs11479 might be more sensitive to capecitabine administration [33].
Interestingly, the relationship between the expression level of TYMP gene and the prognosis of CRC remains controversial in recent study [34]. Our study preliminarily sugested that patients with higher mRNA expression of TYMP might be likely to bene t from capecitabine administration, thus conferring a superior prognosis, which was consistent with the results of the previous study initiated by M Lu and colleagues [35]. A total of 57 patients with advanced gastric cancer who received capecitabine-based regimen were included and identi ed that the mRNA expression of TYMP was positively associated with overall response and OS. Nevertheless, on the other hand, considerable studies had found that TYMP gene was an important factor promoting tumor angiogenesis, which demonstrated that higher expression of TYMP gene was usually accompanied by a higher tumor burden, which generated the tendency that the tumor was easy to recurrence and metastasis, thus contributing to the worse prognosis [36].
Collectively, TYMP gene played a dual role in vivo. On the one hand, it stimulated angiogenesis of tumor cells to promote tumor growth. On the other hand, TYMP was an important metabolic gene of capecitabine and a key gene for 5-FU chemotherapy drugs to play a cytotoxic role and kill tumor cells [37].
In a word, the conclusion in our study was needed to be con rmed in large-scale clinical trials subsequently.
The limitations of this study were as follows: rstly, the sample size of the study was relatively small, and the clinical signi cance of TYMP polymorphism in patient with CRC was still needed to be evaluated in a in a larger population. Secondly, present study was designed as a retrospective analysis and some bias could not be avoided. However, the clinical signi cance of TYMP rs11479 was fully evaluated, we thought our study was of clinical signi cance for the prognostic evaluation of patients with CRC who received surgical resection and capecitabine-based adjuvant chemotherapy.

Conclusions
Collectively, present study provided real-world evidence regarding the prognostic data of patients with CRC who were treated with surgical resection and capecitabine-based adjuvant chemotherapy in realworld. Simultaneously, the prognostic association analysis suggested that TYMP gene polymorphism rs11479 might involve in the prognosis of patients with CRC who received capecitabine based adjuvant chemotherapy through mediation of the mRNA expression of TYMP.

Funding
The authors received no speci c funding for this work.
Availability of data and materials The analyzed datasets generated during the study are possibly available from the corresponding author on reasonable request.
Ethics approval and consent to participate The study protocol was approved by the ethics committee of the A liated Hospital of Hebei University.

Consent for publication
Not applicable.  Figure 1 Flow chart of this retrospective study of in uence of TYMP polymorphism on the prognosis of patients with colorectal cancer receiving capecitabine-based adjuvant chemotherapy  Disease free survival of the 218 patients with colorectal cancer according to TYMP rs11479 genotype status Figure 5 Overall survival of the 218 patients with colorectal cancer according to TYMP rs11479 genotype status