This is the first study to examine the effect of germline genetic polymorphisms on the clinical response of patients with esophageal cancer to preoperative DCF chemotherapy.
In our previous study, we reported that genetic polymorphisms involved in docetaxel pharmacokinetics affect the development of grade 3 or higher neutropenia. In addition, a study of preoperative chemotherapy for esophageal cancer reported that grade 3 or higher neutropenia was correlated with a high histological response . The incidence of docetaxel-induced neutropenia is particularly high in Asians , and neutropenia is considered a major concern because it is a dose-limiting toxicity of docetaxel. Therefore, we performed this pharmacogenomic study with the hypothesis that genetic polymorphisms that affect docetaxel pharmacokinetics may influence the response to DCF chemotherapy. However, contrary to expectations, the two ABC transporter gene polymorphisms (ABCB1 rs1045642 and ABCC2 rs12762549), which were shown to be associated with grade 3 or higher neutropenia in a previous study, did not affect the clinical response to DCF. No pharmacokinetic-related genetic polymorphisms, including those in other drug transporters and metabolic enzymes, had any impact on the clinical efficacy of DCF therapy.
Polymorphisms in genes related to DNA repair processes, such as NER, are also known to affect the response to chemotherapy. Some reports suggest that polymorphisms in the ERCC and XRCC genes, which are involved in DNA repair mechanisms, may affect the response to platinum-based chemotherapy and chemoradiotherapy [23,24].
In this study, no association was observed between polymorphisms located in NER and BER pathway genes and DCF response rate, which indicates that the XRCC3 rs17997944 polymorphism in the XRCC3 gene, which is involved in HR, might be the only genetic factor that affects tumor objective response. XRCC3 facilitates the binding of RAD51 to DNA, which has been implicated in the recognition and repair of DNA interstrand crosslinks . Functional analysis has also shown that XRCC3 overexpression is associated with increased Rad51C protein levels  and is strongly correlated with DNA damage resistance . XRCC3 rs1799794, which was shown to be associated with tumor reduction in this study, is located in the 5′-UTR region, 316 bp upstream of the ATG start codon of the XRCC3 gene. According to an in silico analysis of regulatory RNA motifs, this genetic polymorphism is assumed to cause the loss of upstream open reading frames (uORFs), and because it is involved in translational regulation by controlling uORF expression, this polymorphism may affect gene expression .
To the best of our knowledge, no studies have reported an association between XRCC3 gene polymorphisms and tumor shrinkage in esophageal cancer patients treated with DCF chemotherapy. Although the clinical importance of XRCC3 rs17997944, which was shown to be significantly associated with the response to DCF chemotherapy, is still unclear, several clinical studies support the results of this study. A study of patients with localized prostate adenocarcinoma who were treated with radiation therapy revealed no relationship between XRCC3 rs1799794 and the response to radiation therapy, but an association between this genotype and the development of common late radiation injury was found . Furthermore, in a recent study of patients with glioblastoma treated with chemoradiotherapy consisting of temozolomide, a statistically significant association was found between progression-free survival and overall survival and XRCC3 rs1799794 .
The development of clinically useful biomarkers that can predict sensitivity to preoperative chemotherapy is an important issue. The tumor response to chemotherapy is believed to be affected by both somatic and germline gene mutations. The results of a systematic review and meta-analysis of studies that evaluated the impact of somatic and germline polymorphisms on tumor response to preoperative treatment in rectal and esophageal cancers were reported in 2015 . In the present study, a germline XRCC3 polymorphism was shown to be a predictive factor of a good clinical response to preoperative DCF chemotherapy, but in the abovementioned meta-analysis, which focused on esophageal cancer, only wild-type TP53 predicted better efficacy of preoperative treatment. On the contrary, few articles have reported germline mutations suitable for incorporation into a meta-analysis, and thus, a meta-analysis could not be conducted because of lack of data. Therefore, it is desirable to develop pharmacogenetic biomarkers for clinical application.
The current study has some limitations. First, the sample size is relatively small. Although all patients who received DCF chemotherapy were surveyed, only 56 evaluable patients with measurable lesions were ultimately included in the analysis. This study was two-sided without adjustment for multiple testing. Therefore, it is suggested that the germline XRCC3 rs1799794 genetic polymorphism might not be significant. Second, no patients had the homozygous variant of the minor allele in some genetic polymorphisms, which may be related to the small sample size. Third, the timing of the tumor response evaluation was limited. This study was based on a single evaluation after the first cycle, but an evaluation after three cycles of chemotherapy may be necessary. However, if the evaluation had been performed after three cycles, it would have been impossible to accurately evaluate the relationship with tumor shrinkage because of dose reduction and prolonged dosing intervals caused by adverse reactions. Therefore, tumor response was evaluated only after the first cycle. That is, we believe that this limitation is actually a strength and that the relationship between tumor response and gene polymorphisms was evaluated correctly in this study.
In conclusion, this study suggests that the germline XRCC3 rs1799794 genetic polymorphism in the DNA repair pathway might influence the clinical response of patients treated with preoperative DCF therapy. Because this study was performed retrospectively with a small number of patients whose DNA samples in the NCC Biobank Registry were used, external validation in another cohort with a large sample is necessary to clarify the clinical significance of this XRCC3 polymorphism in DCF therapy.