The NER pathway is one of the major repair mechanisms for platinum-DNA adducts and cross-links, playing a key role in the detection and repair process of the DNA damage exerted by the platinum compounds, particularly cisplatin and carboplatin (Rosell et al., 2002; García-Campelo et al., 2005). Several studies have described that genetic alterations in the SNPs of NER pathway genes and deregulation of NER pathway proteins may affect the DNA repair functions, influence the efficacy of platinum-CT and, eventually, impact in clinical outcomes of patients with lung cancer (Olaussen and Postel-Vinay, 2016; Pérez-Ramírez et al., 2017).
In our study we investigated a homogeneous cohort of 38 patients with ED-SCLC, assessing the role of three key genes encoding crucial factors of the NER pathway, ERCC1, ERCC2 and ERCC5. Our results showed a higher expression of these genes in SCLC-patients compared to the healthy controls; of note, patients with low ERCC1 and ERCC5 expression had a better mPFS and mOS, while an inverse trend emerged in survival outcomes for ERCC2 expression. The genotyping analysis of five polymorphic alterations in a panel of clinically relevant SNPs belonging to these genes showed that patients with the homozygous mutant genotype in the ERCC1 rs11615 SNP and in both the ERCC2 rs13181 and rs1799793 SNPs showed increased expression of all three genes, while a different gene expression pattern was found for the genetic variants of the two ERCC5 SNPs (rs1047768 and rs2296147). Finally, we investigated how genetic alterations in these SNPs may influence survival outcomes of SCLC patients. Overall, our data showed a longer PFS and OS, although not significant, in patients harboring the wild-type genotype or the heterozygous variant of the ERCC1 rs11615 SNP and of the rs13181 and rs1799793 ERCC2 SNPs compared to patients with the corresponding homozygous mutant genotype. On the other hand, the two polymorphisms of ERCC5 showed no association with patients' survival.
The higher gene expression of ERCC1, ERCC2 and ERCC5 in tumor samples compared to healthy control is consistent with previous studies that reported high tumor tissue levels of NER pathways genes, mainly ERCC1, in SCLC and in other solid tumors (Dabholkar et al., 1994; Metzger et al., 1998; Sereno Moyano et al., 2012; Sodja et al., 2012). The high expression of NER pathway genes could lead to an increased ability of tumor cells in repairing the DNA damage exerted by platinum-CT, conferring resistance to anti-cancer treatment. This may explain why, despite initial response to therapy, almost all ED-SCLC patients eventually experience disease progression and often early disease relapse. Consistently with this observation, ERCC1 has been extensively investigated as a predictive biomarker for efficacy of platinum agents and as a negative prognostic factor in various malignancies, including NSCLC and SCLC. Several studies have showed a significant correlation between low ERCC1 expression and both higher response rate to CT and better clinical outcomes in cancer patients (Metzger et al., 1998; Lord et al., 2002; Reed, 2005; Azuma et al., 2007; Chan et al., 2008; Hubner et al., 2011; de M Rêgo et al., 2017; Liu et al., 2018). As far as SCLC, protein or mRNA expression of ERCC1 was reported to be predictive of treatment efficacy and a prognostic factor for survival (Ceppi et al., 2008; Kim et al., 2009; Chiappori et al., 2010; Sereno Moyano et al., 2012). Moreover, in a large retrospective study of 184 SCLC patients, the low expression of ERCC1, as part of a favorable expression signature, was significantly correlated with better PFS and OS in both limited disease (LD) and ED-SCLC (Karachaliou et al., 2013). Interestingly, Lee et al. also reported that expression of ERCC1 in SCLC is lower compared to studies of tissue from NSCLC, suggesting that the greater biologically aggressiveness of SCLC could be related to the loss of ERCC1 gene, which leads to impaired DNA-repair functions of the NER pathway (Lee et al., 2008). According to previous data, our findings showed that patients with low ERCC1 expression had longer PFS and OS. Similarly, better outcomes were also found in patients with low ERCC5, for whom far less data is available to date (Walsh et al., 2008; Graf et al., 2011). Of note, Simon and colleagues reported that increased expression of ERCC1 is an independent predictor of improved survival in resected patients with NSCLC and that this may be secondary to a decreased accumulation of genomic aberrations as a result of efficient DNA-damage repair system (Simon et al., 2005). The favorable prognostic value of ERCC1 expression in resected-NSCLC patients was also confirmed by Olaussen et al (Olaussen et al., 2006). Consequently, upregulation of ERCC1 on the one hand seems to be a negative prognostic factor in advanced disease, as it reduces the benefit of platinum-CT, while on the other hand may act as a predictor of better survival in limited disease, as it reduces the risk of relapse after definitive treatment. Interestingly, it was reported that the ERCC1 gene generates different isoforms by alternative splicing and that only one isoform is implicated in the repair of platinum-DNA adducts (Friboulet et al., 2013). Consequently, the expression of different isoforms could be a major concern as it would lead to a tumor being considered ERCC1-positive, although the expressed protein might be non-functional.
In contrast with ERCC1 and ERCC5, our results showed that patients with low ERCC2 expression had an opposite survival trend with shorter PFS and OS. Although this finding should be taken with caution given the small size of ERCC2-low subgroup in our population, some previous data have reported a correlation between ERCC2 upregulation with a more aggressive cancer phenotype in head and neck tumors and in NSCLC cell lines, suggesting an inter-tissue variation in NER genes and chemoresistance (Weaver et al., 2005; Zafeer, Mahjabeen and Kayani, 2016; Wang et al., 2021). Furthermore, in our study we first reported a significant association between low ERCC2 expression and the presence of brain metastases at diagnosis. It is likely that the worst survival in these patients is associated with the presence of brain metastases at diagnosis. On the other hand, we may argue that the low or loss of expression of ERCC2 correlates with a more malignant SCLC-phenotype at diagnosis and with a higher propensity for metastatic dissemination, due to an altered DNA damage repair mechanism. To the best of our knowledge this finding has not been previously reported. Intriguingly, this might be investigated in future perspective studies in the limited-stage setting, where low ERCC2 expression could be a potential prognostic biomarker of encephalic dissemination, thus supporting the indication for PCI in these patients.
The genotyping analysis of SNPs on NER genes in our cohorts showed that the allele frequencies of the three genes of interest differed between SCLC patients and the general population, particularly for the SNP of ERCC5 rs2296147 (p<0.0001), suggesting a crucial role of the NER pathway in these patients. As wild-type and variant genotypes of the NER pathway are likely associated with differential activity of DNA repair functions (Lunn et al., 2000), polymorphic alterations in these genes may influence the variability on DNA damage repair activity. Consequently, we assessed whether specific genotypes of these genes may also impact the clinical outcomes of SCLC patients. Our data showed that homozygous mutant genotypes in the SNPs of ERCC1 and the ERCC2 are associated with decreased PFS and OS compared to wild-type or heterozygous variant genotypes. To date, very few studies have investigated the role of genetic polymorphisms in the NER pathway for SCLC. Nicos and colleagues evaluated the genotypes of the ERCC1 in a cohort of SCLC patients, reporting that patients harboring the heterozygous genotype in rs11615 had a significantly shorter OS compared to wild-type genotype, which instead was a favorable prognostic factor. Moreover, the different genotypes in two SNPs of the ERCC1 showed also a correlation with the hematological toxicity of the treatment (Nicoś et al., 2021). Conversely much more data about the role of gene polymorphisms in the NER pathway are available for NSCLC, where they have been extensively investigated, sometimes with contradictory results (Pérez-Ramírez et al., 2017). The SNP profiling of the ERCC1 mostly indicated clinical relevance for response to platinum-based CT and association with survival (Dong et al., 2012; Mazzoni et al., 2013; Mlak et al., 2013; Xu et al., 2013; Zhao et al., 2014) but its association with clinical outcomes remains unclear. Several studies also evaluated the role of ERCC2 SNPs rs13181 and rs1799793in prognosis and survival outcomes of advanced NSCLC. Generally, and in accordance to our results, variant genotypes were reported to be associated with decreased OS in these patients (Gurubhagavatula et al., 2004; Kalikaki et al., 2009; Dong et al., 2012; Wu et al., 2012; Zhang et al., 2020). Interestingly, two meta-analyses reported that SNPs in both ERCC1 and ERCC2 genes may play a significant role in lung cancer risk. In detail, Xu T-P et al indicated that individuals carrying at least one wild-type allele in ERCC1 rs11615 have a reduced risk of lung cancer development (Xu et al., 2013), while a large meta-analysis by Zhan P et al, including 22 studies about ERCC2 rs13181 and rs1799793 polymorphisms suggested that mutant genotypes of these SNPs are correlated with lung cancer development (Zhan et al., 2010). Taken together, these data highlight the dual role of DNA repair genes, both in carcinogenesis and in the restoration of platinum-induced DNA damage. For that reason DNA repair systems have been described as a double-edged sword (Wei, Frazier and Levin, 2000; Rosell et al., 2002) because, on the one hand, a deficiency in DNA repair functions may increase cancer susceptibility, while, on the other hand, it may improve survival in patients already diagnosed with cancer, when treated with platinum agents.
A novelty of the present study is that the integrated analysis of the expression of ERCC1 and ERCC2 and their SNPs variants was able to confirm a trend for a longer PFS and a significantly better OS for patients with a favorable signature (low gene expression and no homozygous mutant variants) than those with unfavorable signature (high gene expression or presence of homozygous mutant variants of the SNPs). This finding supports the prognostic role of the NER pathway genes and suggests that an integrated analysis of both mRNA expression and gene polymorphisms in the major components of the NER pathway may identify SCLC patients with different survival outcomes.
Limitations of the present study include its retrospective design and the relatively small sample size that greatly weakens the statistical power of survival analysis. Furthermore, it is known that the DNA repair process includes a complex set of different mechanisms, and thus several genetic polymorphisms in other DNA repair pathways could influence the clinical outcome of SCLC. Finally, our study population received platinum-based CT, as combination treatment with ICI had not yet been approved for patients with ED-SCLC at the time of the design of the study. However, we consider that our research may help to identify a subgroup of patients who achieve substantial benefit from standard CT and with better prognosis. Therefore, future large sample and prospective studies are warranted to validate the role of expression and polymorphisms in NER pathway genes on the prognosis of SCLC patients.
Finally, in the new era of chemo-immunotherapy and in a future therapeutic landscape in which new targeted therapies may also play a role in the therapeutic strategy of SCLC, the role of platinum-based CT will probably still remain relevant in maximizing the therapeutic benefit for these patients. Thus, future prospective trials may further investigate NER pathway genes and their role as prognostic factors for SCLC in a new and integrated therapeutic scenario.