Lung cancer is the leading cause of cancer-related death worldwide (1). Radical resection is the standard treatment for patients with clinical stage I to III non-small cell lung cancer (NSCLC). In Japan, adjuvant chemotherapy is recommended for patients with NSCLC having a pathological maximum tumor size with diameter of ≥2 cm, even with complete tumor resection. Elderly individuals tend to have poor treatment compliance because of physical and cognitive declines. The European Organization for Research and Treatment of Cancer and the International Society of Geriatric Oncology recommend adjuvant chemotherapy treatment for elderly individuals with NSCLC as postoperative adjuvant chemotherapy treatment can increase survival and should not be denied to patients based on age (2). Furthermore, prolonged exposure to some external factors and a decrease in DNA repair functions in elderly individuals may lead to various genetic abnormalities, suggesting that molecular mechanisms underlying carcinogenesis may differ by age group and may affect the clinical outcome of systemic therapy in geriatric patients with NSCLC.

S-1 (Taiho Pharmaceutical Co., Ltd, Tokyo, Japan) is an oral fluoropyrimidine agent, consisting of tegafur [a 5-fluorouracil (5-FU)prodrug] gimeracil [a dihydropyrimidine dehydrogenase (DPD) inhibitor that degrades 5-FU], and oteracil, a phosphorylation inhibitor. S-1 reduces the toxic effects of 5-FU in the gastrointestinal tract (3). S-1 monotherapy is effective in elderly patients with advanced NSCLC, and is an alternative treatment for platinum-doublet chemotherapy, has been indicated (4-6). Furthermore, the phase 2 clinical trial conducted by Setouchi Lung Cancer Group (SLCG1201) showed that the alternate-day and the daily oral administrations of S-1 for 14 consecutive days followed by 7-day rest were feasible in elderly patients with completely resected NSCLC (7).

5-FU sensitivity is influenced by its biomarkers, such as the target molecule, thymidylate synthase (TS) (8), fluoropyrimidine metabolizing enzymes (orotate phosphoribosyl transferase [OPRT] and thymidine phosphatase [TP] (9,10), and the 5-FU-degrading enzyme, DPD (11). Activating mutations in the epidermal growth factor receptor (EGFR) gene affects the efficacy of EGFR tyrosine kinase inhibitors (EGFR-TKIs). In addition, clinical and experimental studies have shown that EGFR mutations reduce the efficacy of adjuvant chemotherapy with oral uracil-tegafur (UFT) (Taiho Pharmaceutical Co., Ltd) (12,13). Excision repair cross-complementation group 1 (ERCC1) is implicated in DNA repair pathways. Further, ERCC1 expression is associated with poor prognosis in patients with NSCLC treated with cisplatin-based chemotherapy, and low ERCC1 expression correlated with better prognosis in patients with gastric and colon cancer treated with 5-FU (14). However, the relationship between the expression levels of ERCC1 and the therapeutic efficacy of 5-FU chemotherapy in patients with NSCLC remains unclear.

This study investigated the effect of aging on the expression levels of 5-FU biomarkers by analyzing RNA-seq data from The Cancer Genome Atlas (TCGA) database and assessed the ability of these biomarkers to predict the clinical outcomes of elderly patients with NSCLC who underwent complete resection in the SCLG1201 followed by adjuvant chemotherapy with S-1.

Aging affects gene expression via mRNA splicing and genetic regulation (18,19). The expression of several genes is affected by age in multiple tissues (20,21). To our knowledge, the six genes investigated in this study have not been previously linked with aging. However, the analysis of TCGA RNA-Seq data suggests that TS expression is affected by age. Aging exacerbates the effects of folate deficiency, including its downstream pathway (22). TS is a key downstream molecular target of folate and plays a crucial role in DNA replication and repair. TS expression is tightly regulated by epigenetic modifications (e.g., DNA methylation), transcription factors (E2F, p53, c-Myc), microRNAs, folate availability, hypoxia, and drug interactions. Genome-wide DNA methylation declines with age (22), which may contribute to the downregulation of TS. Additionally, the age-related downregulation of TS expression may be due to epigenetic repression, reduced cell proliferation, metabolic changes, chronic inflammation, and oxidative stress. These findings support the notion that TS expression is influenced by age and that the expression levels of 5-FU biomarkers changes with age, potentially impacting the clinical outcomes of adjuvant treatment in elderly patients with NSCLC compared with younger patients.

This study examined the mRNA expression of six biomarker genes associated with 5-FU chemotherapy in patients aged ≥75 years with pStage IA (2 cm) to IIIA NSCLC. We investigated the association of these markers with clinicopathological characteristics and prognosis. The upregulation of EGFR and DPD genes and downregulation of ERCC1 and TS genes were significantly associated with the presence of EGFR mutation. The univariate analysis for RFS and OS revealed that the downregulation of EGFR was an unfavorable factor for RFS, and the upregulation of TS was an unfavorable factor for OS. Multivariate analysis showed that pStage II/III was identified as an independent unfavorable factor for both RFS and OS.

EGFR mutations were significantly associated with the upregulation of EGFR and DPD and the downregulation of ERCC1 and TS. Consistent with these findings, previous studies have shown that mutant alleles specific imbalance is common in mutant EGFR cells and correlates with increased mutant allele transcription (23-25). The upregulation of DPD is associated with EGFR mutations in clinical samples and cell lines (26). Moreover, EGFR signaling regulates DPD expression via Sp-1 in EGFR-mutant cells (27). These findings support the notion that EGFR mutations contribute to resistance to 5-FU-based therapies by upregulating DPD. Regarding ERCC1, an experimental study showed that increased DNA damage and reduced damage repair (ERCC1 and RAD51 foci formation) were more common in EGFR exon 19 deletion mutant cells than in EGFR wild-type cells (28). Furthermore, ERCC1 expression was increased by inhibiting EGFR exon 19 deletion signals and decreased by blocking wild-type EGFR signals. NSCLC specimens with EGFR activating mutations tend to have low ERCC1 mRNA levels (28-31). ERCC1 and TS are essential for DNA synthesis and repair, and TS expression was decreased in EGFR-mutant lung cancer specimens, consistent with our findings (29,31).

Univariate analysis showed that EGFR downregulation and TS upregulation were significantly associated with unfavorable RFS and OS, respectively. To our knowledge, the prognostic impact of EGFR mRNA expression has not been well-investigated in patients who received 5-FU adjuvant chemotherapy after complete resection. However, we have previously showed that adjuvant UFT improved prognosis in patients without EGFR mutations but not in patients with EGFR mutation (12), in line with a previous study (13). In contrast, a meta-analysis of 18 studies involving 2972 patients conducted before EGFR-TKIs were developed showed that EGFR overexpression was not associated with favorable prognosis (combined HR of 1.14 with 95%CI of 0.97 to 1.34; p=0.103) (32). Although our univariate analysis indicated that EGFR upregulation was associated with favorable RFS, its impact on RFS in patients receiving 5-FU-based therapies remains controversial. Several studies and meta-analyses showed that low TS protein expression was significantly associated with favorable prognosis in patients with lung cancer who received S-1-based chemotherapy (33,34) and pemetrexed-based chemotherapy (35). Furthermore, TS overexpression was associated with poor prognosis in patients treated with 5-FU and UFT, including those with NSCLC (33,36,37) and gastrointestinal cancers (38). Moreover, in vitro studies have suggested that EGFR-TKIs induce TS downregulation in TKI-resistant NSCLC cells with MET amplification but not in cells harboring the T790M EGFR mutation (39-41). These findings suggest a link between TS expression and EGFR mutations.

Regarding EGFR mutation subtypes, a pooled analysis of 12 clinical trials involving patients with advanced NSCLC and EGFR mutations indicated that exon 19 deletion was significantly associated with better clinical outcomes than the L858R mutation after EGFR-TKI treatment (42). Additionally, preclinical studies have demonstrated that the molecular differences in these two subtypes influence the efficacy of EGFR-TKIs (43-46). Therefore, we compared the expression levels of six genes corresponding to different EGFR mutation subtypes. There were no significant differences in expression because of the small number of cases with specific EGFR mutation subtypes (12 cases with exon 19 deletions, 9 cases with L858R mutation, and 3 cases with rare mutations) (data not shown). Exon 19 deletion was significantly more prevalent in younger patients than the L858R mutation (42), supporting age affects the mechanisms of lung cancer. Although S-1 monotherapy was effective and feasible as a subsequent-line treatment in a small cohort of elderly patients treated with anti-cancer therapies, including EGFR-TKIs (47), large-scale studies are needed to evaluate the efficacy of S-1 chemotherapy, including its role as adjuvant therapy, in elderly patients with EGFR-mutated NSCLC, with a focus on EGFR mutation subtypes.

This study has some limitations. First, the small sample size may limit the generalizability of our findings to larger populations. Thus, studies with larger, independent and diverse cohorts are needed to confirm the robustness and reproducibility of our results. Notwithstanding, all patients were monitored until death or for at least 5 years from the registration. Second, as a preliminary study, we performed receiver operating characteristic (ROC) analysis using survival outcomes (dead or alive). However, except for DPD [which had an area under the curve (AUC) of 0.631 and a P-value of 0.0332], the AUC values for the other markers were below 0.543, with P-values exceeding 0.10. We also conducted ROC analysis based on recurrence events; however, this approach similarly failed to yield suitable cutoff values (data not shown). Additionally, we observed a substantial imbalance in the number of subjects classified according to the ROC-derived cutoffs. Given these limitations, we determined the cutoff values using median values instead of ROC analysis, as the small number of patients did not allow for the determination of suitable cutoff values for mRNA expression. Third, although mRNA was detected using a commercial real-time PCR assay using primers for a specific lesion of the target gene, we did not assess the correlation between mRNA expression level and protein expression levels. Nevertheless, as a preliminary study, we evaluated the correlation between mRNA and protein expression levels for TP, DPD, and EGFR (data not shown). Experimental and clinical studies have shown that the non-synonymous SNP 538G>A in MRP8/ABCC11, an ABC transporters for which 5-FU, methotrexate, and pemetrexed are substrates, is a potential biomarker for S-1 treatment (48,49). Clarifying the clinical implications of these candidate biomarkers is warranted. Fourth, we found that pStage was an independent prognostic factor but did not account for other potential confounding factors, such as comorbidities. Thus, future studies with more patient data and rigorous adjustments for confounders are needed to elucidate the prognostic value of 5-FU biomarkers.

In conclusion, the analysis of TCGA data showed that TS expression was significantly decreased with age. Univariate analysis indicated that, among 5-FU derivatives, TS downregulation and EGFR upregulation were favorable prognostic markers in elderly patients with NSCLC who underwent radical resection followed by adjuvant chemotherapy with S-1. Although pStage was an independent prognostic factor in the multivariable analysis, the findings suggested that elderly patients with NSCLC exhibited low TS expression, which might have improved the clinical outcomes of S-1 adjuvant chemotherapy in this population.