Predictive Value of ERCC1 mRNA Level and Pretreatment EBV-DNA Virus Load in Stage II Nasopharyngeal Carcinoma Patients Receiving Intensity-Modulated Radiotherapy with Concurrent Cisplatin


 Background: The protein expression of ERCC1 in DNA repair genes was related to resistance platinum and predicting treatment outcomes in various malignant carcinoma ,the level of plasma Epstein-Barr virus（EBV）DNA concentrations is positively correlated with clinical stages of nasopharyngeal carcinoma(NPC), but the predictive value of ERCC1 mRNA and EBV-DNA level for stratified treatment with stage II NPC is unclear precisely. This study aimed to assess the predictive value of combined EBV-DNA and ERCC1 in stage II NPC patients treated with intensity-modulated radiotherapy (IMRT) with concurrent cisplatin and provide guidance for future stratified treatment.Methods: A total 78 stage II NPC patients who received IMRT and concurrent cisplatin-based chemotherapy had measurements of ERCC1 mRNA and pre-treatment EBV DNA levels by real-time PCR (RT-PCR) analysis were analyzed. Associations of ERCC1 mRNA and pre-treatment EBV DNA levels with clinical characteristics and survivals were evaluated.Results: Cut-off value of ERCC1 mRNA obtained from ROC curve was used and there were significant differences in progression-free survival (PFS) and overall survival (OS) between high expression group as compared to low expression group (P=0.021 and 0.030, respectively). Patients with pretreatment EBV-DNA<2000 copies/ml had significantly better PFS (P= 0.024) than those with pretreatment EBV-DNA≥2000 copies/ml, but there was no significant difference in OS (P= 0.062). Patients were divided into three groups by combination of ERCC1 mRNA and EBV-DNA level ERCC1 mRNA low expression/pre EBV-DNA<2000 copies/ml, ERCC1 mRNA low expression/pre EBV-DNA≥2000 copies/ml, ERCC1 mRNA high expression/pre EBV-DNA≥2000 copies/ml. In these groups, 1-year, 3-year, 5‐year OS were 100%, 100%, 100%; 100%, 94.1%, 90.9%; 100%, 85%, 72.9%, respectively (P=0.038); 1-year, 3-year, 5‐year PFS were 100%, 100%, 100%; 97.1%, 91.2%, 84.8%; 95%, 85%, 71.4%, respectively (P=0.028). Multivariate analysis showed combination of ERCC1 mRNA and EBV-DNA levels remained independent prognostic factor but not ERCC1 mRNA and EBV-DNA alone. Conclusion: Combined ERCC1 mRNA and pre EBV-DNA is a better prognostic factor in stage II NPC patients treated with concurrent chemoradiation. Patients with ERCC1 mRNA high expression/pre EBV-DNA≥2000 copies/ml should be treated with more aggressive regimen.

Conclusion: Combined ERCC1 mRNA and pre EBV-DNA is a better prognostic factor in stage II NPC patients treated with concurrent chemoradiation. Patients with ERCC1 mRNA high expression/pre EBV-DNA≥2000 copies/ml should be treated with more aggressive regimen. Background Nasopharyngeal carcinoma (NPC) is one of the most common malignancies in Southern China with an incidence rate of 20-30/10000 [1,2]. Non-keratinizing carcinoma (WHO 2005 classi cation), especially the undifferentiated type of Non-keratinizing carcinoma, is the major pathological type in regions with high incidence [3], which is more sensitive to chemoradiotherapy. Cisplatin is the standard concurrent chemotherapeutic agent in advanced NPC(stage III-IV) [4], due to its inhibitory effect on the tumor cells recovery from sub-lethal and potentially lethal radiation damage [5]. NCCN guidelines recommend concurrent chemoradiation for stage II NPC, however the role of chemotherapy in stage II NPC is still controversial. A phase III randomized controlled trial revealed improved 5-year overall survival (OS) rates in stage II NPC who received concurrent chemoradiotherapy compared to radiotherapy alone (94.5% vs 85.8%; P = 0.007) [6]. In contrast, data from the Real-World Study (RWD) showed that patients with stage II NPC did not bene t much with additional chemotherapy; those treated with radiotherapy alone had excellent long-term survival [7]. It is possible selection of patients for chemoradiation is important. The clinical stage NPC includes three subgroups: T2N0M0 T0 1N1M0, and T2N1M0. Long-term follow-up data showed that the survivals of these subgroups were different [8]. However, the current staging system based solely on anatomical information cannot fully re ect the risk factors strati cation of stage II NPC; there is biological heterogeneity among tumor individuals. Therefore, there are some limitations in guiding stage II NPC treatment through clinical staging system.
Repair of DNA damages is a key process after cancer cells exposed to radiotherapy and chemotherapy [9]. Mediators that are associated with DNA damage repair process may affect the outcomes of radiotherapy and chemotherapy [9]. Excision repair cross-complementation group 1 (ERCC1) act as a rate-limiting enzyme in the nucleotide excision repair (NER) pathway [10][11]. Recently, increasing experimental and clinical evidences demonstrated that high ERCC1 expression was associated with resistance to platinum-based chemotherapy in gastric cancer, ovarian cancer, and lung cancer, and low ERCC1 expression predicted good survival of these patients [12][13][14][15][16][17]. In locoregionally advanced NPC treated with chemoradiotherapy, there were several studies used immunohistochemical techniques (IHC) to determine the expression of ERCC1 and found that high tumor ERCC1 expression predicted poor treatment response and survival [18][19][20][21]. However, the relationship between ERCC1 mRNA and treatment outcomes in NPC has not been established.
Epstein-Barr virus(EBV)DNA concentrations correlate positively with disease stage as well as exhibit prognostic importance in NPC [22]. Many published studies have shown that EBV DNA is an important biomarker for monitoring survival for all stages of disease [23][24][25]. In stage II NPC, our previous study found that the plasma EBV-DNA level one week after radiotherapy completion could predict 3-year and 5year recurrence and metastasis rate [26], but the predictive value of pre-treatment plasma EBV-DNA level have not been studied.
Understanding the underlying intrinsic factors would help us to predict the prognosis and select optimal treatments. However, whether combined ERCC1 and EBV-DNA is associated with prognosis in the prediction of concurrent chemoradiotherapy (CCRT) treatment outcomes and provide clinical guidance remains unclear. The subject of this study was to retrospectively assess the prognostic value of combined ERCC1mRNA with the level of pre-treatment the plasma EBV-DNA and survival for stage II NPC.

Methods
Human NPC tissue samples All procedures performed in this study involving human participants were approved by Ethics Committee of the participating hospitals, and were in accordance with the 1964 Declaration of Helsinki. Informed consent was obtained from each patient. Human NPC tissues were collected from 78 patients who underwent CCRT between June 2012 and June 2015. All cases were biopsy-proven NPC and stage II disease according to the seventh edition TNM staging system. The median age was 46 years old (18-73 years old) and the gender ratio (male vs. female) was 3.1:1. The clinicopathological characteristics of the NPC patients were presented in Table 1. group. The response rate (RR) was calculated as number of CR plus PR cases divided by all cases. All patients were followed every 3 months in the rst year and every 6 months for the following years.

Real-time PCR analysis
RNA was extracted from the cancerous tissues using TRIzol reagent (Takara, Dalian, China) according to manufacturer's instructions, and cDNA was synthesized using PrimeScript 1st strand cDNA Synthesis Kit (Takara). ERCC1 expression level was detected by using SYBR Green real-time PCR in ABI7900 system test. The survival difference in OS and PFS between different groups was assessed by Kaplan-Meier plots. All statistical tests were two-sided, and P values less than 0.05 were considered as statistically signi cant.

ERCC1 mRNA expression in NPC tissues and treatment response
The mRNA expression levels of ERCC1 in NPC tissues determined by RT-PCR ranged from 0.456 to 1.412 with the median value of 1.119.
To determine the diagnostic value of the ERCC1 mRNA to response, the data from 78 cases were analyzed by ROC curve analysis. The results showed that the area under the curve (AUC) was signi cantly larger than the area expected by chance alone (P < 0.0001, 95% CI = 0.917-0.960; Fig. 1).
The optimal cut-off value was 1.295; with sensitivity of 87.5% and a speci city of 81.4%.
Using this cut-off value, ERCC1 mRNA expression was found signi cantly associated with N stage and TN stage but not signi cantly associated with gender, age or T stage (Table 2). Furthermore, using this cut-off value as dichotomy, 40/41 had complete response or partial response in the low expression group (< 1.295) and 30/37 had complete response or partial response in the high expression group (≥ 1.295).
Those with low expression had signi cant higher treatment response rate than those with high expression (97.65% vs 81.08%; P = 0.043). The relationship between ERCC1 mRNA expression levels and treatment outcomes The Kaplan-Meier survival curves for OS and PFS comparing high ERCC1 mRNA expression and low ERCC1 mRNA expression using the cut-off value were shown in Fig. 2A  The EBV-DNA viral load and treatment outcomes Using pretreatment EBV-DNA load 2000 copies/ml as a cut-off [22], EBV-DNA viral load was found to be signi cantly associated with T stage and TN stage but not signi cantly correlated with gender, age or N stage (Table 3).

Combination of ERCC1 mRNA expression and EBV DNA viral load improved prognostic strati cation
Patients were strati ed into three groups: low ERCC1 mRNA expression and low EBV-DNA load, low ERCC1 mRNA expression and high EBV-DNA load, high ERCC1 mRNA expression and high EBV-DNA load.

Discussion
In the present study, the ERCC1 mRNA expression level in NPC tissues determined by RT-PCR was examined rather than the ERCC1 protein level determined by the immunohistochemistry (IHC); the latter is more commonly used in most studies. Each method has its advantages and disadvantages. IHC procedure is easier to perform but with poorer reproducibility. In comparison, RT-PCR is more sensitive. In addition, ERCC1 has four protein subtypes 201, 202, 203 and 204. Only subtype 202 is related to the e cacy of platinum chemotherapy. However, the commonly used protein antibody detection method of ERCC1 cannot distinguish different subtypes [27,28]. A meta-analysis by Song et al suggested that RT-PCR method is superior to IHC in term of detecting ERCC1 expression level to predict the outcomes of platinum-based chemotherapy in gastric carcinoma [29]. However, this is still controversial due to the differences in experimental reagents and criteria for assessment.
Our study showed that ERCC1 expression level was strongly correlated with N stage and TN stage but not signi cantly associated with gender, age or T stage. This is different from the results by Olaussen et al that found ERCC1 expression signi cantly correlated with age, gender and histological type, but not associated with tumor size and TNM stage [30]. Our study also showed that ERCC1 mRNA level was signi cantly associated with treatment outcomes in stage II NPC. Using the cut-off value obtained from ROC curve, ERCC1 mRNA can be divided into high expression and low expression groups. The OS and PFS of high expression group were signi cantly worse than that of low expression group. It is possible that the cancer cells with high ERCC1 expression have better ability to repair DNA damages, which is associated with resistance to platinum radio-chemotherapy leading to poor response to initial radiotherapy and chemotherapy.
Several studies have shown that EBV-DNA levels in advanced stage NPC are signi cantly higher than those in stage I and II NPC [31][32][33]. Du et al observed that the level of EBV-DNA was correlated with age, T stage, N stage and total stage in stage II-IV NPC [34]. However, TWU et al reported that EBV-DNA was not correlated with clinical and pathological factors in stage III NPC [35]. In the present study with stage II diseases, we found that EBV-DNA was signi cantly associated with T stage and TN stage but not signi cantly correlated with gender, age or N stage.
Furthermore, in advanced stage NPC, patients with a high level of EBV-DNA have worse survival outcomes, and several prognostic subgroups can be classi ed based on their plasma EBV-DNA levels. However, in stage I and II disease, there was no study on the prognostic value of pre-therapy plasma EBV-DNA due to infrequent treatment failures. The optimal cut-off value of EBV-DNA has not been determined due to lack of harmonization of qPCR assay and different values were reported in different studies. Chai et al used 8000 copies/ml of pre-treatment plasma EBV DNA as a potential biomarker associated with OS for all stage NPC [36]. Lin et al, however reported that OS and relapse-free survival were signi cantly lower in patients with pretreatment plasma EBV DNA concentration ≥ 1500 copies/ml than those with < 1500 copies/ml [37]. Zhang et al in a meta-analysis of 14 prospective and retrospective control studies evaluated the relationship between EBV-DNA and half-life of EBV-DNA and survival and proposed a 4level risk strati cation model. Patients with stage II disease and pre-treatment plasma EBV-DNA < 4000 copies/ml could be down-staged to stage I and treated with radiation alone. However, patients with ≥ 4000copies/ml pre-treatment plasma EBV-DNA should be re-classi ed into stage III and treated more aggressively [24]. Guo and colleague analyzed 979 newly diagnosed nonmetastatic NPC patients with RPA model and proposed a new staging model based on pre-treatment EBV DNA level using 2000 copies/ml as cut off value and anatomical information [22]. Based on Guo's study, we also used 2000 copies/ml EBV-DNA as a cut-off value in our study and found that stage II NPC also has high EBV-DNA concentrations with 70% of the patients having more than 2000 copies / ml. Using this cut-off value, we show here that higher level of pre-treatment EBV DNA was associated with an increased risk of disease recurrence. There was a trend that higher level of pre-treatment EBV-DNA was associated with poor OS but did not reach statistical signi cance. Leung et al [38] also did not nd a signi cant association between pre-treatment EBV DNA levels and the probability of distant failure (P = 0.0967) and the probability locoregional failure (P = 0.719) in stage I-II NPC. The failure to detect signi cant association of pre-treatment EBV-DNA and OS in our study is possible due to small sample size and short follow up.
Pretreatment EBV-DNA has been shown to be an independent prognostic factor for NPC in many studies. But ERCC1 mRNA level was not an independent prognostic factor in lung cancer, gastric cancer and pancreatic cancer [39]. In our study with stage II NPC, neither pretreatment EBV-DNA nor ERCC1mRNA was independent prognostic factors. However, combination of pretreatment EBV-DNA and ERCC1 mRNA was shown to be a signi cant independent prognostic factor for OS and PFS.  Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
This study was approval by the Ethics Committee of Guangxi Medical University and the Forth A liated Hospital of Guangxi Medical University, written informed consent was obtained from each patient.

Consent for publication
Not applicable.