Association of variations in platinum resistance-related genes and prognosis in lung cancer patients

Purpose: We aimed to investigate the association of single-nucleotide polymorphisms (SNPs) in HMGB1, REV3L, and NFE2L2 with prognosis in lung cancer patients with platinum-based chemotherapy. Methods: We have recruited 348 lung cancer patients treated with platinum. Log-rank test and Cox regression analysis were used to assess overall survival (OS) and progression-free survival (PFS) among SNP genotypes. Results: The results revealed that patients carrying TC or CC genotype in REV3L rs462779 (HR=0.67, 95% CI=0.51-0.90, P=0.007) and AG or GG genotype in HMGB1 rs1045411 (HR=0.61, 95% CI=0.38-0.99, P=0.046) had a better overall survival. Additionally, carrying TC or TT genotype in rs462779 had a lower risk (OR=0.38, 95% CI=0.17-0.89, P=0.025) of lymph node metastasis, carrying AG or AA genotype in rs1045411 was significantly related to early T stage (OR=0.47, 95% CI=0.29-0.76, P=0.002). In stratified analysis, patients with TC or CC genotype in rs462779 were significantly associated with overall survival in male patients, never-smokers, patients with younger age (≤56), no family history of cancer, adenocarcinoma, advanced stage (stage III or IV), or ECOG PS 0-1. While patients with AG or GG genotype in rs1045411 were significantly associated with overall survival in patients with advanced stage (stage III or IV) or ECOG PS 0-1. Conclusion: Our results indicate that the TC or CC genotype in rs462779 and AG or GG genotype in rs1045411 are contributed to better overall survival. The REV3L rs462779 and HMGB1 rs1045411 may serve as prognosis markers in lung cancer patients with platinum-based chemotherapy.


Introduction
According to recent studies, lung cancer is still the leading cause of cancer mortality both in males and females, which accounts for 18.4% of total cancer death [1]. A majority of lung cancer patients, approximately 85%, are diagnosed with non-small cell lung cancer (NSCLC) [2]. To the best of our knowledge, tobacco smoking is the main risk factor for lung cancer, and other factors are environmentaland occupational-related. Though the incidence rate has slightly decreased in recent years, the 5-year Ivyspring International Publisher survival of lung cancer is merely 19% [3]. Most of the patients have been diagnosed with an advanced stage, the prognosis for patients with lung cancer is still depressing. Obviously, the prognosis of lung cancer patients with different stage varies greatly, it is necessary to improve survival by finding biomarkers for prognosis prediction. In addition to traditional prognostic factors including tumor microenvironment state and clinical stage, diverse outcomes of cancer patients may also attribute to genetic alterations [4].
In spite of other treatments, like surgery, targeted therapy, radiation, and immunity therapy, platinum-based chemotherapy is still the first-line treatment for lung cancer patients [5]. Platinum-based chemotherapy drugs mainly function by forming DNA adducts intrastrand and interstrand, cells recognize the adducts as DNA damage then activate DNA repair mechanisms, like nucleotide excision repair, mismatch repair, base-excision repair, non-homologous end-joining, and homologous recombination [6,7]. The initial response of platinum in most lung cancer patients is generally good, but intrinsic or acquired resistance would influence the sensitivity of platinum. The underlying mechanisms are described as following: increased damage, decreased accumulation, detoxification, epigenetic changes, decreased apoptosis, membrane trafficking changes, and genetic variants [8][9][10].
Platinum resistance-related genes affect the efficacy of platinum-based chemotherapy, which may act on the prognosis of patients. There are many genes associated with platinum resistance, according to database searching and literature references, we chose HMGB1, REV3L, and NFE2L2 to study the association between genetic variations and clinical outcome. HMGB1 (high mobility group box 1) is situated in the nucleus to maintain nuclear homeostasis, which plays an indispensable role in many diseases and cellular processes, like tumor, inflammation, cell differentiation, and migration [11]. HMGB1 directly binds to lesions of DNA then pursuits to DNA repair and tends to relate with platinum toxicity and liver damage [12,13]. There are many reports about variants of HMGB1. The rs1045411 is related to the overall survival of gastric cancer and the risk of developing lung cancer, hepatocellular carcinoma, and lymph node metastasis of breast cancer [14][15][16][17]. The rs1412125 and rs2249825 appear to be significantly related to the risk of lung cancer and platinum response and seem to relate with the progression of breast cancer [15,[17][18][19]. REV3L (the zeta catalytic subunit of DNA polymerase) is the most important DNA polymerase, it plays a role in maintaining genome stability in the advent of DNA damage and cell proliferation after damage, also may inhibit tumorigenesis [20,21]. Inhibition of REV3L sensitizes lung cancer and gliomas to cisplatin chemotherapy [22,23]. The rs462779 and rs465646 in REV3L are associated with overall survival in platinum-treated malignant mesothelioma, rs462779 is also correlated with event-free survival and risk of colorectal cancer [24,25]. Transcription factor NFE2L2 (nuclear factor, erythroid 2 like 2), also named NRF2, functions as an oncogene or tumor suppressor, it regulates the cellular antioxidant response and promotes cancer chemotherapy resistance, metastasis, and progression [26][27][28]. The rs6706649 and rs6721961 in NFE2L2 may influence breast cancer prognosis, rs6721961 and rs35652124 are associated with susceptibility to colorectal cancer [29,30].

Materials and methods
Subjects 351 lung cancer patients were recruited in the study between November 2011 and May 2013 in Xiangya Hospital of Central South University and Hunan Provincial Tumour Hospital in Changsha, Hunan Province, and clinical data were collected in the same place. No patients had received surgery, targeted therapy, radiotherapy, and/or biological therapy before first-line chemotherapy. In demographic characteristics study and outcome analysis, we excluded 3 patients who had incomplete clinical information. In tumor progression analysis, we excluded 46 patients who had incomplete TNM stage information. All subjects were Han people of Chinese and successfully genotyped after donated 5 mL blood, all patients were treated with platinum-based therapy, and each had signed informed consent. A standard follow-up was carried out to collect characteristic data. Clinical data including age, gender, smoking history, family history of cancer, histology classification, TNM stage, and Eastern Cooperative Oncology Group Performance Status. The last follow-up date was in July 2019.

SNPs selection and genotyping
HMGB1 SNPs (rs1045411, rs1412125, and rs2249825), REV3L SNPs (rs462779 and rs465646), and NFE2L2 SNPs (rs6706649, rs6721961, and rs35652124) were selected for genotyping in the Han Chinese patients. They were chosen according to the following criteria: (1) based on our previous research, HMGB1 SNPs were related with lung cancer platinum-based chemotherapy response [18], (2) association with the outcome of cancers and involvement in other cancer types in other research, (3) functional relevance of gene transcription or protein expression.
EDTA tube was used to hold 5 mL of venous blood donated from each patient. DNeasy Blood & Tissue Kit (Qiagen, Shanghai, China) or Genomic DNA Purification Kit (Promega, Madison, WI, USA) was used to extracting genomic DNA based on the instructions. Sequenom Mass Array Genotype Platform (Sequenom, San Diego, CA, USA) was used to genotype SNPs of each gene. Primers were designed using Primer-Premier 6 software (Premier Biosoft Interpairs, Palo Alto, CA). The Sequenom Mass Array Genotype Platform was used to design polymerase chain reaction (PCR), the shrimp alkaline phosphatase enzyme (SAP) reaction and associated extension reactions. The PCR system was heating for 15 minutes at 94℃, thermocycling for 45 times (94℃ for 20 seconds, 56℃ for 30 seconds, then 72℃ for 1 minute) and extension at 72℃ for 3 minutes. The SAP reaction was performed at 37℃ for 40 minutes then 85℃ for 5 minutes. The extension reaction was conducted at 94℃ for 30 seconds, 94℃ for 5 seconds, 40 cycles for 5 seconds at 52℃, five cycles for 5 seconds at 80℃, finally 72℃ for 3 minutes. The resin was used to purify the reaction product, and the Mass Array system (Sequenom) was used to resolve the data.

Statistical analysis
The SPSS version 25.0 (SPSS Inc., Chicago, IL, USA) and GraphPad Prism (version 8, GraphPad Software Inc., San Diego, CA) were used for statistical analysis. Three genetic models (additive, dominant, and recessive model) were applied to evaluate the association between SNPs and prognosis of lung cancer patients. Overall survival (OS) and progression-free survival (PFS) were analyzed as endpoints and were defined as the time from diagnosis to death for any reason, the time from diagnosis to the first time when patients progressed, respectively. The association between clinical or SNPs data and OS or PFS was assessed using Cox proportional hazards regression analysis, computed as HRs with corresponding 95% CIs. Binary logistic regression was used to test the difference in disease progression between groups. The log-rank test was used to examine the difference in overall survival or progression-free survival between groups. Kaplan-Meier plot was used to visualize the results. All the P-values were two-sided, P<0.05 were supposed to be significant.

Demographic characteristics of patients and prognosis analysis
A total of 348 subjects were included in this investigation, all had received platinum-based chemotherapy. As shown in Table 1, the median age was 56 years (a range of 21 to 75 years). Among them, 78 (22.4%) were females, 270 (77.6%) were males. There were 205 (58.9%) patients who had ever smoked, and the rest 143 (41.1%) patients never smoked. Most of the patients (95.4%) did not have a family history of cancer. For histology, 179 (51.4%) patients were diagnosed with adenocarcinoma, 142 (40.8%) with squamous cell carcinoma, the remainings with other types. 338 (97.1%) of the patients were in an advanced stage (stage III or IV). Likewise, most patients (94.0%) were with Eastern Cooperative Oncology Group Performance Status (ECOG PS) 0 or 1. In the univariate Cox regression analysis, histology was significantly associated with progression-free survival, the risk of progression in patients with squamous cell carcinoma (HR=0.69, 95% CI=0.54-0.89, P=0.004) was lower than that in patients with adenocarcinoma (Table 1).

Relationship between the eight SNPs and clinical outcome of platinum-based chemotherapy
Log-rank test and multivariate Cox regression analysis with adjustment for age, gender, and histology were conducted to analyze the association between the eight SNPs and OS or PFS. The rs462779 in REV3L was significantly correlated with overall survival of lung cancer patients in the additive (P for log-rank=0.018) and recessive (P for log-rank=0.005) models ( Table 2). Patients carrying TC or CC genotype in rs462779 showed a markedly lower death risk (HR=0.67, 95% CI=0.51-0.90, P=0.007) when compared with TT genotype in the recessive model ( Table 2). There was also a trend difference in progression risk that patients with TC or CC genotype had better progression-free survival (P for log-rank=0.034), while the P-value for Cox regression (HR=0.77, 95% CI=0.59-1.02, P=0.073) was not significant ( Table 2). Additionally, patients carrying AG or GG genotype in rs1045411 were significantly related to better overall survival (HR=0.61, 95% CI=0.38-0.99, P=0.046) ( Table 2). The other SNPs were not significantly associated with either OS or PFS.

Association between the eight SNPs and lung cancer progression
To investigate whether the eight single nucleotide polymorphisms were related to tumor progression, binary logistic regression was used to analyze the association. As shown in Table 3, patients carrying REV3L rs462779 TC or TT genotypes had a lower hazard (OR=0.38, 95% CI=0.17-0.89, P=0.025) of lymph node metastasis in the dominant model when compared with patients carrying CC genotype. The rs1045411 and rs2249825 in HMGB1 gene were significantly associated with T stage. Patients with AG or AA genotype in rs1045411 were significantly related to early T stage (T1 or T2) when compared with GG genotype (OR=0.47, 95% CI=0.29-0.76, P=0.002), and patients with GC or GG genotype in rs2249825 were also associated with early T stage when compared with CC genotype (OR=0.56, 95% CI=0.33-0.94, P=0.028) ( Table 3).

Subgroup analysis of association between REV3L/HMGB1 polymorphisms and prognosis
To further study the associations of rs462779 and rs1045411 with the prognosis of platinum-based chemotherapy, multivariate Cox proportional hazard analysis was performed in the recessive model stratified by clinical characteristics. In the patients with younger age (≤56 years old), we observed that the TC or CC genotype in rs462779 had significantly lower death risk (HR=0.61, 95% CI=0. 42 Figure 1A-1G).

Discussion
Platinum remains the commonly used drug for lung cancer patients, it is generally used in combination with other antitumor drugs, but the appearance of drug resistance leads to unsatisfied efficacy. Stable DNA adducts formed by cisplatin lead to DNA damage then mainly pursuit to nucleotide excision repair, finally result in cell apoptosis [6,7]. Enhanced DNA repair contributes to platinum resistance, which compromises the chemotherapy efficacy and therefore influences clinical outcome [10].  Polymorphism is one of the factors that may affect prognosis, and variants of genes of drug transporters, metabolic enzymes, DNA repair system, apoptosis pathway, and folate metabolism pathway are the most studied biomarkers for predicting platinum-based chemotherapy response, while more meaningful markers still need to be discovered [10].
In the present study, we focused on the genetic alterations of HMGB1, REV3L, and NFE2L2, a total of eight SNPs, to study the prognostic effect of platinum-based chemotherapy. Among them, REV3L rs462779 and HMGB1 rs1045411 were significantly associated with overall survival; none of them was significantly associated with progression-free survival. In male patients, patients who never smoked, or patients with younger age (≤56), no family history of cancer, adenocarcinoma, advanced tumor (stage III or IV), or ECOG PS 0-1, patients carrying TC or CC genotype in rs462779 had better overall survival when compared with TT genotype. Patients carrying AG or GG genotype in rs1045411 had better overall survival in lung cancer patients with advanced stage (stage III or IV) or ECOG PS 0-1. In terms of tumor progression, carrying TC or TT genotype in REV3L rs462779 had a lower risk of lymph node metastasis when compared with CC genotype, while carrying AG or AA genotype in HMGB1 rs1045411 or carrying GC or GG genotype in HMGB1 rs2249825 were significantly related to early T stage (T1 or T2).
REV3L is an important DNA polymerase involved in DNA replication, repair, recombination, and has an increased mRNA expression in non-small cell lung cancer tissue, the C-terminal portion encompasses conserved DNA polymerase domain, and the N-terminal domain contains residues direct contacting to DNA [20,[31][32][33]. Pol zeta plays a major extender role in translesion DNA synthesis (TLS) after DNA lesions due to its capability to extend mismatched or distorted primer templates of sorts, the lesions were directly bypassed by specialized DNA polymerases, such as DNA polymerases zeta [34,35]. Loss expression of REV3L increases the frequency of chromosome translocation and break, thus results in genomic instability, and it acts on as a tumor suppressor because of inhibition of spontaneous tumor formation [31,35,36]. REV3L rs462779 is a nonsynonymous SNP (p.Thr1224Ile) that may influence protein function, REV3L rs465646 alters the microRNA binding site in the 3´-UTR, thus may affect REV3L expression [24].
Genetic variations in REV3L have been reported that it was associated with tumor risk or survival in multiple kinds of tumors. Patients carrying the TC or TT genotype in rs462779 had significantly increased colorectal cancer risk, and REV3L rs462779 and RAD18 rs373572 seemed to have a strong cumulative relation with CRC risk, while carrying the TC or CC genotype in rs465646 had significantly decreased lung cancer risk [25,37]. Carrying the TC or CC genotype in rs462779 or carrying the AG or GG genotype in rs465646 was associated with good overall survival in platinum-treated malignant mesothelioma patients (n=139), the TC or CC genotype in rs462779 had poor overall survival in osteosarcoma patients (n=66), the TC or CC genotype in rs462779 had poor event-free survival in aggressive breast cancer patients (n=738) [24,38,39].  Furthermore, REV3L rs462779 showed a slight association with severe toxicity (thrombocytopenia) in NSCLC patients treated with platinum-based chemotherapy [40]. Nonetheless, there is no report about the association between OS or PFS and polymorphisms in the platinum resistance-related REV3L gene. In our research, carrying TC or CC genotype in rs462779 had good overall survival in lung cancer patients (n=348), but no SNP was significantly associated with progression-free survival. It perhaps due to the differences in sample size and tumor type compared with previous studies. Besides, patients carrying TC or TT genotype had a lower risk of lymph node metastasis. However, the mechanisms of how rs462779 influences overall survival and lymph node metastasis still need further investigation.
HMGB1 plays nuclear factor role or extracellular signaling molecule role during cell migration and tumor metastasis. Overexpression of HMGB1 induced by chemotherapy or radiotherapy was associated with all hallmarks of cancer, which leads to tumor microenvironment disorder [12,13,41]. Recent studies also show that HMGB1 is involved in positive regulation and maintenance of ferroptosis in acute myeloid leukemia (AML) cells and autophagy in thyroid cancer cells, also prevents necroptosis in AML cells [42][43][44]. On the aspect of cancer risk, rs1045411 was significantly associated with susceptibility to hepatocellular carcinoma, oral squamous cell carcinoma, lung cancer, breast cancer, and urothelial cell carcinoma [15-17, 45, 46]. In terms of association of genetic variants of HMGB1 and prognosis, previous reports showed that carrying AG or AA genotype of rs1045411 in HMGB1 gene was significantly associated with good overall survival in gastric cancer patients [14]. HMGB1 rs1045411 is located in the 3'UTR, and was reported to decrease HMGB1 expression through hsa-miR-505-5p [45,47]. HMGB1 rs1045411 C/T heterozygous polymorphism was associated with a significantly lower ratio of developing EGFR mutation in the smoking population; it may be a protective factor in lung adenocarcinoma [47]. In our study, carrying AG or GG genotype in rs1045411 was significantly associated with good overall survival. Moreover, rs1045411 and rs2249825 were significantly related to the early T stage.
There are also some limitations in our study. The recruitment hospitals are located in the same region, and the multi-central clinical studies may be needed to overcome results bias. It is a retrospective study with the sample size not large enough, therefore prospective or other analogous studies are warranted to validate our results. Thus we are enrolling other patients and independent validation will be done in our next studies. The underlying mechanism of rs462779 relating with cancer prognosis still needs further investigation.
In conclusion, our study suggested that REV3L rs462779 was significantly associated with overall survival and lymph node metastasis in lung cancer patients with platinum-based chemotherapy, and HMGB1 rs1045411 was related to overall survival and T stage. Genotypes of REV3L rs462779 and HMGB1 rs1045411 may be biomarkers for predicting platinum-based chemotherapy prognosis in lung cancer patients.