ESCC is a highly malignant neoplasm generally associated with poor patient prognosis [2]. NAC followed by surgery is the standard treatment for the patients with surgically treatable advanced ESCC [5]. Therefore, it is important to assess clinical outcomes prior to administering NAC in order to intervene with suitable treatment at the appropriate time during the clinical course. However, prediction of therapeutic response to NAC, including the histological response, has not yet been established. CF therapy is used in NAC for esophageal cancer [5]. The mechanism underlying the therapeutic effects of cisplatin is production of ROS which causes DNA damage in cancer cells [17]. However, 5-FU alone does not necessarily induce apoptosis in cancer cells. When 5-FU is combined with cisplatin, marked stress is inflicted on cancer cells, leading to increased ROS production and DNA damage [18]. Cisplatin, generally used in NAC for ESCC, induces ROS formation and apoptosis [10, 11]. In contrast, 5-FU alone does not necessarily induce apoptosis but in combination with cisplatin, ROS are generated, which eventually results in more marked anti-tumor effects [18]. When DNA is damaged by ROS, 8-OHdG is produced and could serve as a biomarker of oxidative stress in tumors [9]. HO-1 is a target of the Nrf2 pathway and regulates systemic oxidative stress response, participates in cancer cell proliferation, and increases therapeutic resistance [8]. Therefore, in this study, we examined the status of HO-1 and 8-OHdG in patients with ESCC who underwent NAC followed by surgery to evaluate the possible correlation between these findings and the therapeutic response to NAC.
Tumor cells often experience hypoxia during progression, activating the Nrf2 pathway [19]. Therefore, for tumor cells to survive under hypoxic conditions, they must adapt through neoangiogenesis and other biological reactions [19]. Increased HO-1 expression in tumor tissues exerts potentially detrimental effects on the patients because HO-1 promotes tumor neovascularization and provides a selective advantage to tumor cells to overcome oxidative stress[8, 20]. In this study, Kaplan–Meier curve analysis demonstrated that the high HO-1 group had significantly worse OS compared to the low group. These findings indicated that the HO-1 status in the tumor cells of resected ESCC specimens following NAC could serve as a prognostic marker for patients.
The Kaplan–Meier curve in this study also demonstrated that both HO-1 and 8-OHdG tumor H-scores were significantly lower in patients with ESCC without recurrence than in those with recurrence. ESCC with high HO-1 and 8-OHdG H scores had a significantly higher recurrence rate, which also indicated that HO-1 and 8-OHdG status in primary tumors could predict tumor recurrence following NAC and surgery in patients with ESCC. HO-1 is a target of Nrf2 and acts as a potent reducing agent in heme degradation to eliminate ROS [20]. Nrf2 increases under hypoxia [19], with a concomitant increase of HO-1 expression, in conjunction with neoangiogenesis [8, 20]. This is advantageous for the survival of tumor cells by formulating a favorable tissue microenvironment for adaptation to hypoxia. The release of free iron, a metabolite of heme, also results in in situ reduction of ROS formation via the Fenton reaction [20]. ROS cause oxidative stress and therefore 8-OHdG has been widely used as a marker of oxidative stress [9]. 8-OHdG is also related to DNA oxidative damage, characterized by hydroxylation at the eighth position of deoxyguanosine, one of the bases constituting DNA. In particular, 8-OHdG occurring in chromosomal DNA induces G ⇒ T mutation during replication, and the increased 8-OHdG in chromosomes is associated with an increased risk of carcinogenesis [21]. In addition, cancer tissues harbor relatively low levels of antioxidant enzymes and high levels of 8-OHdG [22]. Higher 8-OHdG levels are associated with worse OS in preoperative untreated esophageal cancers [13]. However, the roles and mechanism of action of 8-OHdG in human malignancies are largely unknown. In patients with ESCC who underwent NAC in this study, high levels of 8-OHdG were associated with significantly (p < 0.05) worse DFS and OS. Multivariate analysis also demonstrated that higher 8-OHdG levels significantly shortened DFS, independent of other factors.
In this study, univariate analysis demonstrated a significant prolongation of OS in the group with higher histologically proven therapeutic effects of NAC, and multivariate analysis also revealed a tendency for prolonged OS in patients with ESCC with high NAC effects. In contrast, both univariate and multivariate analyses demonstrated a significant prolongation of DFS in the high NAC group. The significant association between HO-1 and 8-OHdG in the primary tumor and the histological response also indicated the important roles of oxidative stress in the development of NAC resistance in ESCC patients. It has been shown that the addition of cisplatin induced stress in cancer cells, leading to increased production of ROS and the induction of apoptosis [11]. High 8-OHdG levels were significantly (p < 0.05) associated with shortened OS in patients with ESCC who did not receive neoadjuvant chemotherapy [13]. Only post-NAC specimens were evaluated in this study but high 8-OHdG expression tended to shorten OS and DFS [13]. High levels of 8-OHdG indicated an increased ROS production and subsequently apoptosis. Nevertheless, worse prognosis was detected in patients with ESCC with high 8-OHdG levels in our present study. Therefore, it is difficult to determine whether high 8-OHdG expression could be a prerequisite for the progression of carcinogenesis or merely represent secondary changes in the carcinogenic process.
LN metastasis has been considered the most important prognostic factor in ESCC [6, 12]. In this study, the HO-1 H-score of LN metastatic lesions on the Kaplan–Meier curve was significantly (p < 0.05) correlated with OS. However, only pN was significantly (p < 0.05) associated with both OS and DFS in the multivariate analysis. Therefore, the status of pN could be more significantly correlated with prognosis than HO-1 or 8-OHdG in LN metastatic lesions in the evaluation of surgically resected specimens. Therefore, further investigations, including the analysis of HO-1 or 8-OHdG status in endoscopic biopsy specimens before NAC, are required to clarify the response to NAC in patients with ESCC.