Abstract
Aim: The aim of the study was to identify biomarkers capable of predicting response to preoperative chemoradiotherapy (CRT) including S-1 or UFT for rectal cancer using biopsy specimens obtained before CRT (Pre-samples) and 7 days after the start of CRT (Day-7 samples). Materials and Methods: Preoperative CRT including S-1 or UFT was performed in 82 patients with locally advanced rectal cancer. The expression levels of 18 genes related to 5-fluorouracil, folate, and radiation in the Pre-samples and the Day-7 samples were evaluated using reverse transcription polymerase chain reaction (RT-PCR) assay. Results: The gene expression levels of hypoxia inducible factor 1 alpha subunit (HIF1A), dihydropyrimidine dehydrogenase (DPYD) and thymidine phosphorylase (TYMP) were found significantly increased in Day-7 samples compared to Pre-samples in responders, but not in non-responders. Conclusion: Increases in gene expression levels of TYMP, DPYD, and HIF1A in tumor tissues at 7 days after the start of CRT may be useful for predicting the efficacy of CRT including S-1 or UFT.
- Rectal cancer
- chemoradiotherapy
- radiotherapy
- S-1
- UFT
- thymidine phosphorylase
- dihydropyrimidine dehydrogenase
- hypoxia inducible factor 1-alpha
Preoperative radiotherapy or chemoradiotherapy (CRT) has been recognized as the standard-of-care for locally advanced rectal cancer (1, 2). Reportedly, the histological response to preoperative CRT is closely related to oncological outcome. Disease-free survival and overall survival are significantly better in patients with histological complete regression or with tumor down-staging than in patients without such findings (3-5).
Approximately 20% of patients treated with neoadjuvant 5-fluorouracil (5-FU)-based CRT have been reported to show a pathological complete response (pCR) (6). Among patients with a pCR or a near-pCR, the amount of tumor cells in the resected tumor tissues tends to be insufficient. Therefore, to investigate changes in the expressions of genes related to drugs and radiation, tissues obtained before CRT (Pre-samples) and at an early time point after the start of CRT (before proceeding to tumor necrosis) should be compared.
Previous studies have confirmed the presence of irradiation-induced morphological changes in the nuclei of tumor cells at 7 days after the start of CRT (7, 8). We already reported that these histological changes were significantly related to histological regression and predicted response to CRT including S-1 or tegafur-uracil (UFT), when hematoxylin and eosin (H&E)-stained biopsy specimens obtained 7 days after the start of CRT (Day-7 samples) were examined (9, 10). These results suggested that the expression levels of genes related to CRT sensitivity might have changed in the Day-7 samples.
In this study, aiming to identify predictors of the response to CRT, we examined the association between the response to CRT and changes in the expression levels of 5-FU-, folate- and radiation-related genes in Day-7 samples, compared to expressions in Pre-samples.
Patients and Methods
Patients. Eighty-two consecutive patients with histologically confirmed adenocarcinoma of the middle or lower third of the rectum who were treated at our University Hospital between 2010 and 2013 were analyzed in this study. The preoperative diagnosis of these patients was clinical Stage II or Stage III, according to the tumor-node-metastasis (TNM) classification. The initial evaluation included digital examination of the rectum, colonoscopy, barium enema, computed tomography of the chest, abdomen, and pelvis, endorectal ultrasonography, and MRI of the pelvis. All patients received preoperative CRT followed by total mesorectal excision surgery 6 to 8 weeks after the completion of the irradiation. The patient characteristics are shown in Table I.
Patients' characteristics and pathological tumor responses.
This study was conducted with the approval of the Ethics Committees of Tokai University School of Medicine and Taiho Pharmaceutical Co., Ltd. All patients provided their written informed consent.
CRT and tissue sampling. Preoperative radiotherapy was performed with 18-MeV X-ray beams delivered by a linear accelerator (Clinac 2100C; Varian Medical Systems, Inc., Palo Alto, CA, USA) using the four-field technique. Irradiation was performed once (1.8 Gy) daily to a total dose of 45 Gy. For concomitant chemotherapy, 5 patients received oral uracil/tegafur (UFT, 400 mg/m2) and 77 received oral S-1 (80 mg/m2) or S-1+bevacizumab, starting at the same time as the radiotherapy. Oral UFT was given for 5 days, followed by a 2-day rest. This cycle was repeated. Oral S-1 (80 mg/m2) was given for 2 consecutive weeks, followed by a 1-week rest, and then was given for 2 more weeks (11, 12).
UFT and S-1 are the almost same oral 5-FU prodrugs that combine tegafur and DPYD inhibitor (uracil in UFT and 5-chloro-2,4-dihydroxypyridine, CDHP, in S-1) in order to keep high concentration of 5-FU in tumor tissues, and each drug was combined to potentiate the effects of radiation. Among 77 patients receiving S-1, 46 patients enrolled in the phase II study also received bevacizumab (11).
A colonoscopy was performed to obtain biopsy specimens before treatment (Pre-samples) and 7 days (range=4-9 days) after the start of CRT (Day-7 samples). We obtained 6 biopsy specimens from each patient both before and at 7 days after the start of CRT.
Comparison of mean gene expression levels between Pre- and Day-7 samples in all patients.
All biopsy samples were immediately immersed in RNAlater solution (Thermo Fisher Scientific, Waltham, MA USA) and incubated overnight at 4°C. Then, the tissues were removed from the RNAlater solution and stored at -80°C.
Histological evaluation. The antitumor effectiveness was evaluated based on the histological regression observed in the resected specimen. Histological regression was classified according to the tumor regression grade (TRG) (13) and the Japanese classification of colorectal carcinoma (JCCC) criteria (14). TRG was classified as Grade 1 (complete regression), Grade 2 (presence of rare residual cancer cells), Grade 3 (increased number of residual cancer cells), Grade 4 (residual cancer outgrowing fibrosis), or Grade 5 (absence of regression change). A patient with a TRG of 1 or 2 was defined as a responder. A patient with a JCCC grade of Grade 2 (necrosis or disappearance of the tumor is present in more than two thirds of the entire lesion, but viable tumor cells remain) or Grade 3 (complete regression) was also defined as a responder.
Gene expression analysis. The mRNA expressions of 5-FU-related enzymes (six genes: dihydropyrimidine dehydrogenase [DPYD], ribonucleotide reductase M1 [RRM1], thymidine kinase 1 soluble [TK1], thymidine phosphorylase [TYMP], thymidylate synthetase [TYMS], and uridine monophosphate synthetase [UMPS]), of reduced folate-related enzymes (eight genes: 5-aminoimidazole-4-carboxamide ribonucleotide formyltransferase/IMP cyclohydrolase [ATIC], dihydrofolate reductase [DHFR], folylpolyglutamate synthase [FPGS], phosphoribosylglycinamide formyltransferase [GART], gamma-glutamyl hydrolase [GGH], methylenetetrahydrofolate dehydrogenase 1 [MTHFD1], methylenetetrahydrofolate reductase [MTHFR], and 5,10-methenyltetrahydrofolate synthetase [MTHFS]), and of radiation-related enzymes (four genes: cyclin-dependent kinase inhibitor 1A [CDKN1A], hypoxia inducible factor 1 alpha subunit [HIF1A], tumor protein p53 [TP53], and vascular endothelial growth factor A [VEGFA]) were quantitatively evaluated using a RT-PCR assay as described below. Total RNA was isolated from the tissue using the RNeasy mini kit (Qiagen, Valencia, CA, USA) and reverse transcribed using a high-capacity cDNA reverse transcription kit (Thermo Fisher Scientific). Real-time reverse transcription polymerase chain reaction (RT-PCR) was performed using an ABI PRISM 7900HT sequence detection system (Thermo Fisher Scientific) and a real-time RT-PCR array (TaqMan Array; Thermo Fisher Scientific), which included duplicated wells of a reference gene (Assay ID: beta-actin [ACTB] Hs99999903_m1) and 18 target genes (Assay ID: DPYD Hs00559279_m1, RRM1 Hs01040698_m1, TK1 Hs00177406_m1, TYMP Hs00157317_m1, TYMS Hs00426586_m1, UMPS Hs00165978_m1, ATIC Hs00269671_m1, DHFR Hs00758822_s1, FPGS Hs00191956_m1, GART Hs00531926_m1, GGH Hs00608257_m1, MTHFD1 Hs01068263_m1, MTHFR Hs00195560_m1, MTHFS Hs00197574_ m1, CDKN1A Hs00355782_m1, HIF1A Hs00153153_m1, TP53 Hs01034249_m1, and VEGFA Hs00900055_m1). The gene expression levels were normalized to the reference gene, ACTB (15, 16). The relative gene expression levels were calculated using the delta threshold cycle (Ct) method according to the formula shown below. The expression levels of the target genes were expressed as 2−(delta Ct) × 1000 to simplify the calculation.
Comparison of gene expression levels in Pre-samples between responders and non-responders.

Statistical analysis. The differences in the gene expression levels between the Pre- and Day-7 samples were evaluated using a paired t-test. The differences in the gene expression levels in Pre-samples or the gene expression ratios (Day 7/Pre) between responders and non-responders were evaluated using the Student t-test. Log-transformed values were used for all the statistical analyses. JMP 9.0.2 statistical software (SAS Institute Inc., Cary, NC, USA) was used. Differences were considered significant when p<0.05.
Results
The patients' characteristics and pathological tumor response are shown in Table I. Pathological complete regression (pCR) was observed in 20.7% (17/82) of patients. The response rates according to the TRG criteria and JCCC criteria were 50.0% (41/82) and 72.0% (59/82), respectively. No significant relationships between either the pCR rate or the response rate and the clinical parameters (sex, age, tumor site, histological type, regimen, or interval to surgery) were observed (data not shown).
The mean expression levels of the genes evaluated in this study are shown in Table II. The gene expression levels of five genes, CDKN1A, GGH, DPYD, HIF1A, and TYMP, were significantly increased 7 days after the start of CRT. On the other hand, the gene expression levels of four genes, MTHFS, VEGFA, FPGS, and ATIC, were significantly decreased.
Comparison of Day 7/Pre ratio between responder and non-responders.
The relationship between the gene expression levels in the Pre-samples and the pathological tumor responses are shown in Table III. The gene expression level of TYMS was significantly higher in the non-responders than in the responders, as classified according to the JCCC criteria (p=0.028). On the other hand, no significant associations between the tumor response and the expression levels of the other genes were seen.
The relationships between the gene expression ratios of the Day-7 samples relative to the Pre-samples (Day 7/Pre ratios) and the pathological tumor responses are shown in Table IV. In the case of pCR criteria, the Day 7/Pre ratio of HIF1A was significantly higher in the responder group than in the non-responder group. In the case of TRG and JCCC criteria, the Day 7/Pre ratios of two genes (DPYD and HIF1A) and five genes (GART, HIF1A, MTHFR, TK1, and TYMP) were significantly higher in the responder group than in the non-responder group, respectively.
For these 6 genes with a significant difference in the Day 7/Pre ratio between the responder and non-responder groups, the mean gene expression levels were next compared between the Pre- and Day-7 samples for each group (Table V). The gene expression levels of HIF1A, DPYD, and TYMP were significantly elevated in the Day-7 samples in the responder group, but not in the non-responder group. On the other hand, the gene expression level of GART was significantly reduced in the Day-7 samples in the non-responder group, but not in the responder group.
Comparison of mean gene expression levels between Pre- and Day-7 samples in each group.
Significant negative correlations were observed between TRG and the Day 7/Pre ratios of three genes: HIF1A, DPYD, and TYMP (Figure 1). For the Day 7/Pre ratio, significant positive correlations were observed between HIF1A and DPYD (r=0.693, p<0.001), HIF1A and TYMP (r=0.650, p<0.001), and DPYD and TYMP (r=0.710, p<0.001) (data not shown).
Discussion
Previous studies have examined whether factors such as Ki-67, apoptosis, and apoptosis-related genes such as p53 and p21 in biopsy specimens obtained before treatment could be used to predict the response to CRT (17-20). The results of these studies were inconsistent, and reliable predictors of response remain to be established.
TYMS is one of the principle enzymes involved in DNA synthesis and is a molecular target of 5-FU (21). An inverse relationship has been reported between TYMS expression levels and response to 5-FU (22-24). In this study, we also showed a negative correlation between the gene expression level of TYMS in Pre-samples and the response to CRT including S-1 or UFT, as classified according to the JCCC criteria (Table III). However, no association was seen between the expression levels of TYMS and pCR or the response to CRT, as assessed according to the TRG. Moreover, Showalter et al. reported that only a slightly negative correlation existed between TYMS expression and the response to 5-FU based on the results of a meta-analysis, and TYMS expression alone could not be used as a predictor of the response to 5-FU (25). Therefore, we next attempted to investigate the changes in the expression levels of 5-FU-, folate-, and radiation-related genes after the start of CRT.
Kocakova et al. reported on changes in the gene expression levels of TYMS, DPYD, and TYMP during neoadjuvant CRT including capecitabine, an oral 5-FU prodrug, in patients with rectal cancer. Both TYMP and TYMS mRNA were induced using CRT two weeks after the start of CRT in both responders with TRG scores of 1 or 2 and in non-responders. However, in this study using S-1 or UFT for chemotherapy, the expression levels of TYMP and DPYD had increased at 7 days after the start of CRT, whereas the expression of TYMS had not increased.
TYMP is identical to platelet-derived endothelial cell growth factor and has angiogenic activity (26-29). The TYMP expression level in tumor tissues is reportedly higher than that in normal tissues (27, 30-32), and tumors with TYMP overexpression are reportedly more invasive (32) and associated with a relatively poor prognosis among colorectal cancers (29).
On the other hand, in vitro studies have shown that the overexpression of TYMP leads to the activation of 5-FU and the production of FdUMP, thereby enhancing the antitumor activity of 5-FU or tegafur (33-35). In an experiment using TYMP-overexpressing KB and AZ521 cell lines, TYMP reportedly played a principal role in the production of FdUMP and the enhanced responses to 5-FU induced by leucovorin (34). Sadahiro et al. previously reported that high gene expression levels of TYMP in tumor samples obtained before chemotherapy were associated with the response to UFT+leucovorin neoadjuvant chemotherapy in patients with colorectal cancer (36). Recently, a high TYMP gene expression level has been reported to be associated with the response to 5-FU-based chemotherapy in patients with colorectal cancer (37, 38). In this study, the gene expression level of TYMP was significantly increased at 7 days after the start of CRT in the responder group, but not in the non-responder group (Table V). Therefore, the increased gene expression of TYMP could lead to the activation of 5-FU and the production of FdUMP, thereby enhancing the antitumor activity of 5-FU-based chemotherapy (S-1 or UFT) after Day 7 in the present study.
Correlation between TRG and Day 7/Pre ratio of HIF1A, DPYD or TYMP.
The up-regulation of TYMP gene expression reportedly occurred after the administration of S-1 for 2 weeks in a gastric cancer xenograft model (39). A significant increase in TYMP gene expression was also reported in tumor tissues after radiotherapy in patients with rectal cancer (40). Therefore, the increase in the gene expression levels of TYMP in this study could be caused by both chemotherapy including S-1 and radiotherapy.
DPYD is the first and rate-limiting enzyme of 5-FU catabolism. A high intratumoral gene expression level of DPYD has been reported to be associated with resistance or non-response to 5-FU therapy in patients with colorectal cancer (41, 42). In this study, the gene expression level of DPYD was significantly increased at 7 days after the start of CRT in the responder group, but not in the non-responder group (Table V). The up-regulation of DPYD gene expression was reportedly found after the administration of S-1 for 2 weeks in a gastric cancer xenograft model (39), but the DPYD gene expression level had not changed at two weeks after the start of preoperative CRT including capecitabine in patients with rectal cancer (43). Therefore, an increase in DPYD gene expression may be related to the fact that S-1 contains 5-chloro-2,4-dihydroxypyridine (CDHP), an inhibitor of DPYD. Because S-1 contains CDHP, the antitumor effect of S-1 does not seem to be affected by the increase in DPYD gene expression.
HIF1 is an enzyme with a key role in the cellular response to hypoxia, and the alpha subunits of HIF (HIF1A) are rapidly degraded by proteasomes under normoxia, but are stabilized by hypoxia. HIF-1 affects many processes that have been shown to influence radioresponsiveness, including glycolysis, mitosis, apoptosis, and angiogenesis (44). Toiyama et al. reported that a low gene expression level of HIF1A in pre-treatment tumor biopsies was significantly associated with a high rate of tumor regression in patients with rectal cancer treated with neoadjuvant 5-FU based CRT (45). However, in this study, no association between the response to CRT and the gene expression levels of HIF1A was seen in the pre-treatment tumor biopsy samples (Table III).
Moeller et al. reported that radiation induced HIF-1 activation and that inhibiting postradiation HIF-1 activation increased tumor radiosensitivity as a result of enhanced vascular destruction (46, 47). Zeng et al. also reported that radiation induced the activation of HIF1 and that S-1 treatment suppressed the activation of HIF1 in a human lung cancer xenograft model (48). In the present study, the increase in gene expression level of HIF1A at 7 days after the start of CRT including S-1 or UFT was associated with the response to CRT (Table V). Although the reason why S-1 treatment did not suppress the activation of HIF1A in responders is unknown, differences in the radiotherapy schedules between the two studies (14 Gy only once vs. 1.8 Gy/day × 5) may have influenced the effect of S-1.
In this study, we demonstrated that the induction of the gene expressions of TYMP, DPYD, and HIF1A at an early time point (7 days) after the start of CRT could be regarded as predictive biomarkers of the response to 5-FU-based CRT. By determining the gene expression levels of TYMP, DPYD, or HIF1A at an early time point after the start of 5-FU-based CRT, future patients who are expected to have a poor response to CRT may additionally receive systemic neoadjuvant chemotherapy after completion of CRT, followed by surgery or another new individualized treatment strategy.
We selected and investigated only 18 genes that were related to the 5-FU and folate metabolism and radiation. The gene pathways and the number of genes chosen for the analysis in this study were extremely limited and the mechanism responsible for the correlations remains to be determined. In order to confirm these findings, we are now conducting a prospective phase II study on the prediction of response to CRT including S-1 and UFT/LV.
Footnotes
This article is freely accessible online.
Conflicts of Interest
S. Sadahiro, T. Suzuki, A. Tanaka, K. Okada, G Saito and A. Kamijo have no potential conflicts of interest to report. H. Nagase is an employee of Taiho Pharmaceutical Co., Ltd.
- Received March 8, 2016.
- Revision received April 12, 2016.
- Accepted April 13, 2016.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved