ERCC1mRNA Expression Is Associated With Clinical Outcome Of Non-small Cell Lung Cancer Treated With Platinum-based Chemotherapy

Non-small cell lung cancer (NSCLC) is the most common type oflung cancer, ac-counting for70%of all lung cancer cases (IARC,2008).NSCLC is the leading cause of cancer-related deaths worldwide (IARC,2008). A recent report from the World Health Orga-nization (WHO)estimated that the age-standardized incidence of lung cancer is33.5casesper100,000people for the entire Chinese population. The overallincidence was higher in male patients (45.9cases per100,000) thanfemale patients (21.3cases per100,000). NSCLC has a poor prognosis,and it has been reported that more than50%of NSCLC patients presentwith stage III disease or metastatic disease at the time of diagnosis.Recently, platinum-based doublet chemotherapy has become the mostfrequently used chemotherapy regimen for advanced NSCLC.Interestingly, patients with cancer of a similar clinical stage andhistological type experience different clinical outcomes; therefore,genetic variants may be associated with platinum-based chemotherapyefficacy.Platinum is cytotoxic because it disrupts double-stranded DNA incells by forming platinum-DNA adducts. These adducts cause interstrandcrosslinking, induce bulky distortion of DNA, and inhibit DNAreplication. These DNA lesions can be mended by DNA repairmechanisms; therefore, the activities of repair molecules can play a role in clinical outcomes, response to chemotherapy, and survival of patientstreated with platinum-based chemotherapy. Nucleotide excision repair(NER) is a central DNA repair pathway, and it is involved in eliminatingboth cisplatin-induced DNA adducts as well as nucleotides damaged byultraviolet irradiation. The NER pathway involves the cooperation ofaround20enzymes to restore a segment of DNA containing a bulkyadduct. The excision repair cross-complementing group1(ERCC1) gene,together with the xeroderma pigmentosum group F (XPF) gene, isresponsible for making a5-prime incision in DNA and is a rate-limitingstep in the NER DNA repair pathway. The ribonucleotide reductasesubunit M1gene (RRM1), located at the chromosomal locus11p15.5, isinvolved in DNA synthesis. Specifically, RRM1catalyzesdeoxyribonucleotide biosynthesis from corresponding ribonucleotides. Previous studies have suggested that ERCC1and RRM1may play a rolein chemosensitivity and the clinical outcomes of NSCLC patients treatedwith platinum-based chemotherapy. However, few studies haveinvestigated the effect of ERCC1and RRM1on the clinical outcome ofNSCLC or the association between ERCC1and RRM1expression andclinical characteristics. Therefore, we conducted a prospective study toanalyze the expression of ERCC1and RRM1in297Chinese patientswith advanced NSCLC and examined the potential of these genes asbiomarkers for predicting tumor response and clinical outcome.This study included323patients with unresectable, locally advanced,or metastatic NSCLC who were enrolled between September2007andSeptember2009at the Second Hos-pital of Jilin University. All patientshad histologically confirmed inoperable stage IIIB or IV NSCLC. Theinclusion criteria for this study specified that patients had to have beentreated with at least2cycles of first-line platinum-based chemotherapy, had histologically confirmed inoperable stage IIIB or IV NSCLC, had aWHO performance status of0to2, and had measurable tumor response.Clinical and demographic characteristics were obtained from medicalrecords. The histological classification of NSCLC was determinedaccording to WHO criteria (WHO,1981). Patients were excluded fromthe study if they had a second primary tumor, were pregnant, had severecardiopulmonary insufficiency, or were severely malnourished. Allpatients signed written informed consent forms. The ethics committees ofthe Second Hospital of Jilin University approved our study protocol.All patients received platinum-based doublet chemotherapytreatment. The regimens were75mg/m2cisplatin or carboplatin on day1and1000mg/m2gemcitabine on days1and8, or25mg/m2vinorelbineon days1and8, or175mg/m2paclitaxel on day1. Chemotherapytreatment was stopped if individuals experienced disease progression orunacceptable toxicity.Patients’ responses to platinum-based doublet chemotherapy wereevaluated, and patients were classified as either good or poor responders.Patients who had a complete response or partial response tochemotherapy were considered good responders, and patients with stabledisease or progressive disease were defined as poor responders. ChestX-rays and computed tomography scans were used to examine patientsfor progressive disease. Overall survival (OS) was calculated from thedate of chemotherapy initiation to the date of death or the date of lastfollow-up.All patients were asked to provide5mL whole blood, and the wholeblood samples were kept at-70oC until use. Total RNA was extractedfrom blood with an EZNA Blood RNA Mini Kit (Omega, Berkeley, CA,USA). A fluorescence-based real-time detection method was used to quantify relative cDNA levels of ERCC1and RRM1. The relativeamounts of ERCC1and RRM1cDNA were calculated by comparing tothe amount of β-actin. Polymerase chain reaction (PCR) was performedin a20μL reaction volume containing50ng genomic DNA,200μMdNTP,2.5U Taq DNA polymerase (Promega Corporation, Madison, WI,USA), and200μM primers. PCR conditions were as follows:94°C for2min,35cycles of94°C for30s, reduction to the annealing temperature(64°C) for30s, and then72°C for1min. For quality control, differentinvestigators performed genotyping on a random sample of10%of all thecases and control subjects; the reproducibility was100%.The SPSS software version16.0(SPSS Inc., Chicago, IL, USA) wasused for statistical analysis. Continuous variables are expressed as themeans±SD and categorical variables are expressed as the frequency andpercentage. The associations between ERCC1and RRM1expression andchemotherapy response were assessed by ORs and their95%CIs. Theassociations between ERCC1and RRM1expression and OS of NSCLCpatients were assessed by Cox regression models with hazards ratios and95%CIs. The OS times of all patients were plotted using theKaplan-Meier method. All tests were2-sided and P <0.05was consideredto be statistically significant.The clinical characteristics of all patients are shown in Table1. Ofthe323patients,297patients agreed to participate in our study, for a91.95%participation rate. The median age of the patients included was61.5years (range,34.6-74.1years), and69.1%of the patients were male.In this study,57.5%of the patients were TNM stage IV. Adenocarcinomaand squamous cell carcinoma were the histological types of56.7and37.4%of the NSCLC patients, respectively. There were54.5%of thepatients with an Eastern Cooperative Oncology Group performance status of0or1.During the follow-up period,5patients were dropped from the studybecause of loss of contact. All patients were followed up until September2012. At the end of the follow-up period,132patients had died and165patients had experienced progressive disease. The me-dian OS time was18.7months (range,1-60months). The median levels of ERCC1andRRM1were2.46×10-2and0.97×10-2, respectively. Expression levels ofERCC1and RRM1mRNA were classified as either high or lowexpression by comparing to the median level. Of the297patients in thestudy,146patients had a complete response to chemotherapy and151patients had a partial response to chemotherapy. Patients with lowERCC1expression had a significantly higher rate of complete response tochemotherapy, with an OR (95%CI) of1.56(1.03-2.47). The associationsbetween ERCC1and RRM1mRNA expression and OS were analyzedusing a Cox regression analysis. We found that low levels of ERCC1expression were associated with longer OS times compared to highexpression, with an adjusted HR (95%CI) of0.57(0.35-0.93).Individualized chemotherapy based on molecular prognostic markerscan improve the prognosis of cancer patients. Our study showed thatERCC1expression was predictive of tumor response to chemotherapyand identified ERCC1expression as an independent prognostic factor forChinese patients with NSCLC treated with platinum-based chemotherapy.Patients with high expression of ERCC1achieved significantly longersurvival times than patients with low ERCC1expression. The results ofour study are in agreement with previous findings in a small cohort ofNSCLC patients treated with platinum-based chemotherapy. However,another study found no association between ERCC1expression andtreatment response or survival. Our study suggests that ERCC1 expression may predict the prognosis of advanced NSCLC patients.ERCC1plays a crucial role in NER and has been reported toinfluence the effective-ness of platinum-based chemotherapy in variouscancers including gastric cancer, colorectal cancer, and bladder cancer, aswell as NSCLC. Huang et al. reported that overexpression of ERCC1isan important predictor of early treatment failure in patients with stage IIIcolorectal cancer treated with FOLFOX adjuvant chemotherapy(leucovorin,5-fluorouracil, and oxaliplatin), and ERCC1expressioncould be used to identify patients who would benefit from intensivefollow-up and enhanced therapeutic programs. Li et al. reported a studyconducted in an American population that indicated that combinedERCC1and ERCC2functional single nucleotide polymorphisms mightinfluence the OS of gastric cancer patients. For NSCLC, the associationbetween ERCC1expression and prognosis is inconsistent among studies.Tantraworasin et al. conducted a retrospective cohort study of247patients with completely resected advanced NSCLC and found thatERCC1and RRM1expression were not prognostic factors for tumorrecurrence and overall survival. Another study found no associationbetween ERCC1expression and survival or treatment response. However,a study conducted in Greece indicated that ERCC1could be used torefine prognosis and thus individualize chemotherapy for advanced-stageNSCLC patients. Moreover, a recent meta-analysis found a significantlybetter therapy response and longer OS in patients with low or negativeERCC1expression compared to patients with high or positive ERCC1expression. The discrepancies among the reports may be explained bydifferences in ethnicities of the patients studied, the source of controlsubjects, the sample sizes, and also by chance. Future studies are neededto confirm the association between ERCC1polymorphisms and NSCLC prognosis.The expression of RRM1protein by immunostaining has beenreported to predict the clinical outcome of human cancers includingbreast cancer, pancreatic cancer, and gastric cancer. For NSCLC, somestudies have reported that there is no association between RRM1expression and clinical outcomes. In our study, we found that RRM1expression was not a prognostic factor for advanced NSCLC patients whodid or did not receive adjuvant chemotherapy.In conclusion, our results indicate that low ERCC1mRNAexpression is associated with better treatment response in advancedNSCLC patients who receive platinum-based chemotherapy and thatERCC1expression correlates with longer survival. Therefore, we suggestthat ERCC1mRNA expression could be used as a surrogate marker tohelp individualize NSCLC treatment strategies.