Association between DNA mismatch repair gene polymorphisms and platinum-based chemotherapy toxicity in non-small cell lung cancer patients

Background Chemotherapy toxicity is a serious problem from which non-small cell lung cancer (NSCLC) patients suffer. The mismatch repair (MMR) system is associated with platinum-based chemotherapy toxicity in NSCLC patients. In this study, we aimed to investigate the relationship between genetic polymorphisms in the MMR pathway and platinum-based chemotherapy toxicity in NSCLC patients. Methods A total of 220 Chinese lung cancer patients who received at least two cycles of platinum-based chemotherapy were recruited for this study. Toxicity was evaluated in each patient after two cycles of chemotherapy. A total of 44 single nucleotide polymorphisms were selected to investigate their associations with platinum-based chemotherapy toxicity. Results MutS homolog 2 (MSH2) rs6544991 [odds ratio (OR) 2.98, 95% confidence interval (CI) 1.20–7.40, P = 0.019] was associated with gastrointestinal toxicity in the dominant model; MSH3 rs6151627 (OR 2.38, 95% CI 1.23–4.60, P = 0.010), rs6151670 (OR 2.05, 95% CI 1.07–3.93, P = 0.031), and rs7709909 (OR 2.38, 95% CI 1.23–4.64, P = 0.010) were associated with hematologic toxicity in the dominant model. Additionally, MSH5 rs805304 was significantly associated with overall toxicity (OR 2.21, 95% CI 1.19–4.09, P = 0.012), and MSH5 rs707939 was significantly associated with both overall toxicity (OR 0.42, 95% CI 0.23–0.76, P = 0.004) and gastrointestinal toxicity (OR 0.44, 95% CI 0.20–0.96, P = 0.038) in the dominant model. Conclusion Genetic polymorphisms in the MMR pathway are potential clinical markers for predicting chemotherapy toxicity in NSCLC patients. Electronic supplementary material The online version of this article (doi:10.1186/s40880-016-0175-2) contains supplementary material, which is available to authorized users.


Background
Non-small cell lung cancer (NSCLC) is a common cancer and the main cause of cancer-related mortality worldwide [1][2][3][4][5]. Although many new target drugs, such as gefitinib and erlotinib, have been used to treat NSCLC, cytotoxic drugs, such as platinum, are still used as firstline agents in the treatment of NSCLC [6,7]. However, adverse drug reactions, such as nephrotoxicity, hepatotoxicity, hematologic toxicity, and gastrointestinal toxicity, are major obstacles to successful treatment [8][9][10]. Thus, it is important to identify biomarkers that can be used to predict platinum-based chemotherapeutic toxicity [11].
There are several partially overlapping DNA repair pathways, including base excision repair (BER), nucleotide excision repair (NER), double-strand break repair (DSBR), and MMR [22]. A previous study showed that mutations in DNA repair genes affect the effectiveness and toxicity of platinum-based chemotherapy in NSCLC patients [23]. MMR plays a key role in maintaining genomic stability through the highly conserved biological pathway. It is an important determinant of platinum cellular toxicity. The formation of platinum/DNA adducts blocks replication and transcription of DNA, and the MMR system plays an important role in removing these adducts. MMR defects lead to replication and recombination errors and cause 6-thioguanine (6-TG)-and O6-methylguanine-induced toxicity in DNA glycosylasedeficient cells [24]. It has been reported that mutations in MMR pathway genes may be associated with platinumbased chemotherapy toxicity in NSCLC patients [23,25]. In addition, mutations in MMR genes, particularly those in MSH3 and MSH5, may be associated with the risk of lung cancer and even lead to increased alkylation tolerance [26].

Study subjects
All patients met the following inclusion criteria were selected: (1) patients between 18 and 80 years old; (2) patients newly diagnosed with NSCLC, including adenocarcinoma or squamous cell carcinoma, with histological or cytological examination at the Affiliated Cancer Hospital or Xiangya Hospital of Central South University (Changsha, Hunan, China) between December 2012 and December 2015; (3) patients who received at least two cycles of platinum-based chemotherapy, e.g., cisplatin or carboplatin chemotherapy; (4) patients with no history of chemotherapy or radiotherapy; and (5) patients with no history of surgery before or during chemotherapy. Patients with active infections or other concomitant malignancies were excluded.
All patients provided written informed consent before they participated in this study. The study protocol was approved by the Ethics Committee of Xiangya School of Medicine, Central South University (approval number: CTXY-110008-1). This clinical research project was approved by the Chinese Clinical Trial Registry under the following registration number: ChiCTR-RNC-12002892 (http://www.chictr.org/cn/).

SNP selection, DNA extraction, and genotyping
All single nucleotide polymorphisms (SNPs) were selected by Haploview (Broad Institute, Cambridge, MA, USA) using pair-wise tagging with default settings (pairwise r 2 threshold = 0.8). The following SNPs were eligible for further study: SNPs with a minor allele frequency (MAF) ≥ 5% in the Han Chinese population and SNPs in Hardy-Weinberg equilibrium (HWE) (P > 0.05).
All blood samples were collected in the morning and stored at −20°C for 4 h. Genomic DNA was isolated using a Genomic DNA Purification Kit (Promega, Madison, WI, USA) and stored at −20°C before use. Genotyping was conducted using a Sequenom MassARRAY Genotyping Platform (Sequenom, San Diego, CA, USA).

Toxicity evaluation criteria
Platinum-based chemotherapy-induced toxicity was estimated according to the National Cancer Institute Common Toxicity Criteria, Version 3.0. The toxicity intensity was graded on a scale of 1-5 as follows: grade 1, mild adverse events; grade 2, moderate adverse events; grade 3, severe adverse events; grade 4, life-threatening or disabling adverse events; and grade 5, death related to adverse events. We recruited patients experiencing grade 0 to grade 4 toxicity, and they were divided into two categories. Patients experiencing grade 0-2 adverse events were classified into the low-toxicity category, whereas patients with grades 3 and 4 adverse events were classified into the severe toxicity category.

Statistical analysis
The genotype frequencies observed among all patients were compared with their expected frequencies under Hardy-Weinberg equilibrium using a χ 2 test (P > 0.05). Sex, age, smoking status, tumor histology, clinical stage, and Eastern Cooperative Oncology Group (ECOG) performance status were considered potential covariates for logistic regression. All analyses were performed using PLINK (version 1.07, http://pngu.mgh.harvard.edu/purcell/plink/) and SPSS 13.0 software (SPSS Inc, Chicago, IL, USA). Odds ratios (OR) and their 95% confidence intervals (95% CI) were used to assess the association between treatment outcomes and gene polymorphisms. P < 0.05 was considered statistically significant.

Patient characteristics
A total of 220 patients who received first-line platinumbased chemotherapy were recruited for this study. A total of 44 SNPs were genotyped in these patients, and 37 of them were in HWE (P > 0.05) and exhibited an MAF ≥ 5%. The basic information of these SNPs and the clinical characteristics of these lung cancer patients are summarized in Tables 1 and 2, respectively. Hematologic,

Association between MMR gene polymorphisms and toxicity
The genotypes of the 37 SNPs in 6 DNA MMR genes were determined in the 220 patients. The results are summarized in Additional file 1: Table S1. Six SNPs exhibited significant associations with toxicity (

Stratification analyses
Stratification analyses were performed to investigate the associations between all SNPs that were significantly associated with overall toxicity. Patients were stratified by cancer type (squamous cell carcinoma or adenocarcinoma), age (≤55 years or >55 years), smoking status (non-smoker or smoker), and sex (male or female). As shown in Fig. 1

Discussion
In this study, we investigated whether polymorphisms of MMR genes (MLH1, MSH2, MSH3, MSH4, MSH5, and MSH6) were associated with platinum-based chemotherapy toxicity in 220 NSCLC patients. We evaluated the associations between these gene polymorphisms and gastrointestinal, hematologic, and overall toxicities. Our results showed that MSH2 rs6544991 was associated with gastrointestinal toxicity, MSH3 rs6151627, rs6151670, and rs7709909 were associated with hematologic toxicity, and MSH5 rs707939 and rs805304 were associated with gastrointestinal toxicity and overall toxicity. A previous study showed that MSH2 was a key protein that influenced 6-thioguanine (6-TG)-and O 6 -methylguanine-induced toxicity in DNA glycosylase-deficient cells, indicating that MSH2 plays an important role in attenuating oxidative DNA damage [24]. In our study, C allele carriers of rs6544991, which features an A/C single-nucleotide variation located in the intron area of MSH2, exhibited poor gastrointestinal toxicity tolerance after being treated with platinum-based chemotherapy. We speculated that this SNP may affect the ability of MSH2 to remove platinum adducts. In addition, our results showed that MSH3 rs6151627 G allele carriers, rs6151670 G allele carriers, and rs7709909 T allele carriers exhibited poor hematologic toxicity tolerance after being treated with platinum-based chemotherapy. MSH3 rs6151627 is an A/G single-nucleotide variation, rs6151670 is a C/G single-nucleotide variation, and rs7709909 is a C/T single-nucleotide variation. All of these polymorphisms are intron variants of MSH3. Methylation of the MSH3 promoter is involved in esophageal tumorigenesis, suggesting that it plays an important role in modulating cell chemosensitivity. As a DNA MMR gene, MSH3 forms the MutSβ heteroduplex with MSH2. The MSH2/MSH3 heterodimer is an ATPase that plays a critical role in mismatch recognition and repair initiation. It binds to DNA mismatches by recognizing 2-to 13-bp insertion-deletion loops [30]. The three SNPs are speculated to affect the function of MSH3; however, the underlying mechanism is still unclear.
Another important finding of our study was that MSH5 polymorphisms were significantly related to overall toxicity. MSH5 rs707939 T allele carriers exhibited better gastrointestinal and overall toxicity tolerance after being treated with platinum-based chemotherapy. Moreover, among MSH5 rs707939 T allele carriers, stratification analysis showed that male patients, patients ≤55 years old, smokers, and patients diagnosed with squamous cell carcinoma faced a lower risk of overall severe toxicity than their counterparts. All these results indicated that MSH5 rs707939 was associated with reduced cisplatin-induced gastrointestinal and overall toxicities in NSCLC patients. In addition, MSH5 rs805304 was also significantly associated with gastrointestinal and overall toxicities. It is noteworthy that rs707939 is a G/T single-nucleotide variation in the intron of MSH5. Previous studies suggested that mutations in MSH5 result in alkylation tolerance in mammalian cells, which is associated with lung cancer risk [26,31]. Thus, rs707939 and rs805304 may also affect MSH5 activity.
As far as we know, DNA MMR gene defects lead to MMR function loss, which increases the spontaneous Many error messages across the entire genome eventually affect the effectiveness and toxicity of chemotherapy. However, the detailed mechanisms underlying the effects of these SNPs on gene function need to be studied further.

Conclusions
Our findings indicate that carriers of the MSH5 rs707939 T allele, the MSH2 rs6544991 C allele, the MSH3 rs6151627 and rs6151670 G alleles, and the MSH3 rs7709909 T allele have poor toxicity tolerance. Therefore, these polymorphisms are potential clinical markers for predicting platinum-based chemotherapy toxicity in Chinese NSCLC patients. However, a study with a larger sample size is needed to validate these findings in the future. Fig. 1 Stratification analyses of the associations between the MutS homolog 5 (MSH5) rs707939 polymorphism and overall platinumbased chemotherapy toxicity in the 220 non-small cell lung cancer patients using the additive, dominant, and recessive models. Each box and horizontal line represents an odds ratio (OR) and a 95% confidence interval (CI)