Plasma Epstein-Barr virus DNA level strongly predicts survival in metastatic/recurrent nasopharyngeal carcinoma treated with palliative chemotherapy

Authors

  • Xin An MD,

    1. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    2. State Key Laboratory of Oncology in South China, Guangzhou, China
    Search for more papers by this author
    • The first 3 authors contributed equally to this work.

  • Feng-Hua Wang MD, PhD,

    1. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    2. State Key Laboratory of Oncology in South China, Guangzhou, China
    Search for more papers by this author
    • The first 3 authors contributed equally to this work.

  • Pei-Rong Ding MD, PhD,

    1. State Key Laboratory of Oncology in South China, Guangzhou, China
    Search for more papers by this author
    • The first 3 authors contributed equally to this work.

  • Ling Deng MD,

    1. State Key Laboratory of Oncology in South China, Guangzhou, China
    2. Department of Pathology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • Wen-Qi Jiang MD,

    1. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    2. State Key Laboratory of Oncology in South China, Guangzhou, China
    Search for more papers by this author
  • Li Zhang MD,

    1. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    2. State Key Laboratory of Oncology in South China, Guangzhou, China
    Search for more papers by this author
  • Jian-Yong Shao MD, PhD,

    1. State Key Laboratory of Oncology in South China, Guangzhou, China
    2. Department of Pathology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • Yu-Hong Li MD, PhD

    Corresponding author
    1. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    2. Department of Pathology, Sun Yat-sen University Cancer Center, Guangzhou, China
    • 651 Dongfeng East Road, Guangzhou, China 510060
    Search for more papers by this author
    • Fax: (011) 86-20-87343535


Abstract

BACKGROUND:

Plasma Epstein-Barr virus (EBV) DNA is widely used in screening, monitoring, and prediction of relapse in nonmetastatic nasopharyngeal carcinoma (NPC). However, data regarding utility of plasma EBV DNA in metastatic NPC are rare. The current study was to test the prognostic implication of plasma EBV DNA level in metastatic/recurrent NPC patients treated with palliative chemotherapy.

METHODS:

Plasma EBV DNA level was measured at baseline and thereafter at the start of each treatment cycle in 127 histologically proven metastatic/recurrent NPC patients treated with palliative chemotherapy. Correlations of pre-treatment and post-treatment plasma EBV DNA levels to survival and response were analyzed.

RESULTS:

Patients with a low pre-treatment plasma EBV DNA level (<median) had significantly better survival than those with a high pre-treatment plasma EBV DNA level (≥median). Patients with a post-treatment plasma EBV DNA decline to an undetectable level had better survival and better tumor response compared with those with a sustained detectable post-treatment plasma EBV DNA level. The early decrease of post-treatment plasma EBV DNA to an undetectable level after 1 cycle of chemotherapy was associated with significantly increased survival. Patients with low pre-treatment plasma EBV DNA level and undetectable post-treatment plasma EBV DNA showed a favorable prognosis (5-year overall and progression-free survival of 50.6% and 21.7%, respectively).

CONCLUSIONS:

Plasma EBV DNA is of predictive value for prognosis in metastatic/recurrent NPC patients undergoing palliative chemotherapy. The pre-treatment plasma EBV DNA level as well as the early decrease of plasma EBV DNA after chemotherapy enabled easy and early discrimination between patients who will and those who will not benefit from continued treatment. Cancer 2011. © 2011 American Cancer Society.

Nasopharyngeal carcinoma (NPC) is an endemic carcinoma in Southern China and Southeast Asia, where about 50 to 80 cases per 100,000 of population per year are reported.1 Although NPC is sensitive to radiotherapy and chemotherapy, about 20% to 30% of patients will ultimately develop distant metastasis and/or recurrence after definitive treatment using conventional radiation therapy.2, 3 Once metastasis is detected, treatment is mostly palliative chemotherapy, with a median overall survival duration of 12 to 15 months.4, 5 However, patients with metastatic NPC do not behave uniformly. For selected patients, survival rates exceeding 4 years have been reported.6, 7 It is therefore critical to define reliable prognostic factors to identify patients with good or poor prognosis. Currently, evaluation of prognosis for metastatic NPC patients treated with palliative chemotherapy mainly relies on clinical characteristics, including performance status and disease-free interval.8 There is still no reliable laboratory prognostic factor. Recently, several studies including our own data have shown that quantification of plasma Epstein-Barr virus (EBV) DNA by the real-time quantitative polymerase chain reaction (PCR) is a useful marker in the detection, monitoring, and prognostic prediction for nondisseminated NPC treated by radiotherapy.9-13 However, data regarding utility of plasma EBV DNA in metastatic NPC are rare. To our knowledge, only 1 study has demonstrated the predictive and prognostic value of the clearance rates of plasma EBV DNA during the first month in metastatic/recurrent NPC.14 However, only 34 patients were included in that study, and a rather high complete response rate (41.2%) and long median survival (28 months) were achieved, although most patients were just treated with old chemotherapy agents such as mitomycin-C, epirubicin, cisplatin, or 5-fluorouracil, which might suggest a patient selection bias. Moreover, multiple plasma samples were needed during the first month of chemotherapy, which further limits the utility of plasma EBV DNA clearance rate in clinical practice. In our clinical experience, we have observed that a proportion of patients with metastatic NPC experience a dramatic decline of plasma EBV DNA to undetectable levels after palliative chemotherapy. Such a phenomenon has not previously been well described in metastatic NPC patients. Therefore, we initiated the current study to investigate the prognostic implication of pre-treatment and post-treatment plasma EBV DNA load in metastatic/recurrent NPC patients who were treated with palliative chemotherapy.

MATERIALS AND METHODS

Patients

The study included 127 patients with histologically proven metastatic NPC treated with first-line chemotherapy between May 2002 and April 2010 at Sun Yat-Sen University Cancer Center. Entry criteria consisted of: 1) an increased baseline plasma EBV DNA level; 2) radiologically measurable disease; 3) Karnofsky performance score (KPS) ≥60; 4) normal renal and liver function; and 5) informed written consent was obtained from all patients. The study was approved by the Research Ethics Committee of the Sun Yat-sen University Cancer Center.

Treatment

All eligible patients received 1 of the following chemotherapy regimens as the first-line treatment: cisplatin (25 mg/m2 intravenously [IV] on Days 1-3 of a 21-day cycle) plus 5-fluorouracil (500 mg/m2 IV on Days 1-5 of a 21-day cycle), cisplatin (25 mg/m2 IV on Days 1-3 of a 21-day cycle) plus capecitabine (1000 mg/m2 twice a day by mouth on Days 1-14 of a 21-day cycle), or paclitaxel (150 mg/m2 IV over 3 hours with standard premedication on Day 1 of a 21-day cycle) plus cisplatin (25 mg/m2 IV on Days 1-3 of a 21-day cycle) plus 5-fluorouracil (800 mg/m2, continuous IV infusion for 24 hours, on Days 1-5 of a 21-day cycle).

Evaluation of Tumor Response

Patients were evaluated for response every 2 treatment cycles during treatment and then every 2 months until disease progression. The response evaluation of the tumor to therapy was based on computed tomography (CT) or magnetic resonance imaging (MRI) scan. Patients were classified as having a complete response (CR), partial response (PR), stable disease (SD), or progressive disease according to Response Evaluation Criteria in Solid Tumors.15 Bone metastases were assessed radiologically (bone scan and x-ray or MRI) as either present or absent.

Real-Time Quantitative EBV DNA PCR

Measurement of plasma EBV DNA level was taken at baseline and thereafter at the start of each treatment cycle. Plasma DNA was extracted and subjected to a real-time quantitative PCR assay as described previously.13 The real-time quantitative PCR system was developed for plasma EBV DNA detection toward the BamHI-W region. The system consisted of the amplification primers W-44F (5′-AGT CTC TGC CTC AGG GCA-3′) and W-119R (5′-ACA GAG GGC CTG TCC ACCG-3′) and the dual-labeled fluorescent probe W-67T (5′-[FAM] CAC TGT CTG TAA AGT CCA GCC TCC [TAMRA]-3′).

Statistical Analyses

Statistical analysis was performed using the statistical package SPSS for Windows version 12.0 (SPSS Inc., Chicago, Ill). Progression-free survival (PFS) time was calculated from the day of first chemotherapy cycle to the date of radiologic progression or death. Overall survival (OS) was calculated from the first day of chemotherapy to the date of death or the last follow-up visit. Both PFS and OS were estimated by the Kaplan-Meier method and compared between the groups of patients by the log-rank test. Multivariate Cox regression model was used to determine whether plasma EBV DNA was a significant predictor of both OS and PFS. The relationship between plasma EBV DNA level and response rate was evaluated by the chi-square test. All analyses were 2-sided. The level of significance was set at P < .05.

RESULTS

Patient Characteristics and Outcome

The median age was 45 years (range, 24-70 years). One hundred two (80.3%) patients were men. Ninety-seven (76.3%) patients had good performance status (KPS ≥80). Ninety-two (72.4%) patients received prior radiotherapy at initial diagnosis for definitive treatment to head and neck area, among whom 51 (55.4%) patients received radiotherapy alone, and 41 (44.6%) patients received combined chemoradiotherapy. The adjuvant chemotherapy regimens were cisplatin or cisplatin plus 5-fluorouracil. At the time of recurrence, 3 (2.4%) patients had isolated locoregional recurrences, 117 (92.1%) patients had distant metastases only, and 7 (5.5%) had both. Among all patients with distant metastases only, 75 (59.1%) had lung metastases, 60 (47.2%) had liver metastases, 29 (22.8%) had lymph node metastases, and 23 (18.1%) had bone metastases. Patient characteristics are listed in Table 1.

Table 1. Demographic and Baseline Characteristics of Patients
CharacteristicsNo. of patients, N =127%
  • KPS indicates Karnofsky performance score; LDH, lactate dehydrogenase.

  • a

    Median (range).

Age, y45 (24-70)a
Sex  
 Women2519.7
 Men10280.3
KPS  
 ≤803023.7
 >809776.3
Tumor involved site  
 Nasopharyngeal107.9
 Lymph node2922.8
 Lung7559.1
 Liver6047.2
 Bone2318.1
 Other site53.9
Number of involved sites  
 17256.7
 23729.1
 ≥31814.2
Serum LDH  
 ≤245 U/L7559.1
 >245 U/L5240.9
Chemotherapy regimen  
 Cisplatin + fluorouracil/capecitabine6148.0
 Paclitaxel + cisplatin + fluorouracil6652.0

The median follow-up time was 21 months (range, 3-88 months). The Median PFS and OS were 7.2 and 17.0 months, respectively. The overall clinical response rate was 66.9% for all 127 patients (CR in 7 patients and PR in 78 patients).

Pre-treatment Plasma EBV DNA

Median baseline plasma EBV DNA level for all patients was 2.33 × 105 (range, 25-2.6 × 108) copies/mL. The responses (CR and PR) were not significantly different between patients with a baseline plasma EBV DNA level below the median (CR = 5 and PR = 42 in 64 assessable patients) and those with a baseline plasma EBV DNA level equal to or above the median (CR = 2 and PR = 36 in 63 assessable patients, P = .09) (Fig. 1, Top). The median PFS andOS in patients with a baseline plasma EBV DNA level below the median value were significantly longer than those with plasma EBV DNA level equal to or above the median value (median PFS was 8.2 vs 5.8 months, P < .001; median OS was 18.9 vs 15.3 months, respectively, P < .001) (Fig. 2 A, B).

Figure 1.

(Top) Pre-treatment plasma Epstein-Barr virus (EBV) DNA level and tumor response are shown (chi-square test, P = .09). (Bottom) Post-treatment plasma EBV DNA status and tumor response are shown (chi-square test, P = .001). CR indicates complete response; PR, partial response; SD, stable disease; PD, progressive disease.

Figure 2.

(A, B) Progression-free survival and overall survival, respectively, are shown according to the pre-treatment plasma Epstein-Barr virus DNA (pEBVDNA) level. (C, D) Progression-free survival and overall survival, respectively, are shown according to post-treatment plasma EBV DNA status after chemotherapy. (E, F) Progression-free survival and overall survival, respectively, are shown according to the timing of plasma EBV DNA decline after chemotherapy.

Post-Treatment Plasma EBV DNA

A drop of post-treatment plasma EBV DNA to an undetectable level at any time during treatment was noted in 82 (64.6 %) patients. The response rate (CR + PR) was significantly higher in patients with undetectable post-treatment plasma EBV DNA levels compared with those with sustained detectable post-treatment plasma EBV DNA levels. In detail, all 7 patients who achieved complete response showed a decrease in post-treatment plasma EBV DNA to an undetectable level; 57 (73.1%) of 78 PR and 18 (54.5%) of 33 SD patients showed a drastic fall of plasma EBV DNA to an undetectable level. However, none of the 9 patients who progressed on imaging showed a decrease in plasma EBV DNA to an undetectable level (Fig. 1, Bottom). With regard to survival, the median PFS and OS for patients with undetectable post-treatment plasma EBV DNA levels were significantly longer than those with sustained detectable post-treatment plasma EBV DNA (median PFS was 8.4 vs 3.9 months, P < .001; median OS was 18.9 vs 14.5 months, respectively, P < .001) (Fig. 2C, D).

Of note, among 82 patients whose post-treatment plasma EBV DNA dropped to undetectable levels, 47 (57.3%) patients demonstrated an early decline of plasma EBV DNA to an undetectable level after 1 cycle of chemotherapy (21 days), 22 (26.8%) more patients had a plasma EBV DNA decline to an undetectable level after 2 cycles of chemotherapy (42 days), and another 13 (15.9%) patients had plasma EBV DNA decline to an undetectable level after >2 cycles of chemotherapy. The median PFS and OS in patients who achieved an early drop of plasma EBV DNA to an undetectable level after 1 cycle of chemotherapy were 9.6 and 19.6 months, compared with 7.6 (P = .002) and 14.5 (P = .081) months in those who showed a late drop of plasma EBV DNA to an undetectable level after >1 cycle, and 3.9 (P < .001) and 14.5 months (P = .004) in those with persistently detectable post-treatment plasma EBV DNA (Fig. 2E, F).

Combination of the Pre-treatment and Post-Treatment Plasma EBV DNA Status

Patients were divided into 4 subgroups according to the baseline plasma EBV DNA level and the change of plasma EBV DNA after chemotherapy: 1) low pre-treatment plasma EBV DNA (<median) and undetectable post-treatment plasma EBV DNA (35 patients); 2) low pre-treatment plasma EBV DNA (<median) and detectable post-treatment plasma EBV DNA (28 patients); 3) high pre-treatment plasma EBV DNA (≥median) and undetectable post-treatment plasma EBV DNA (47 patients); and 4) high pre-treatment plasma EBV DNA (≥median) and detectable post-treatment plasma EBV DNA (17 patients). Median PFS for these 4 subgroups was 11.9, 5.5, 7.1, and 3.6 months, respectively. Median OS was 22.2, 14.5, 16.2, and 15.1 months, respectively. The 5-year PFS and OS rates in the low pre-treatment/undetectable post-treatment subgroup were 21.7% and 50.6%, respectively, whereas no patients in the other 3 subgroups of patients survived 5 years (Fig. 3).

Figure 3.

(Top) Progression-free survival is shown according to the combination of pre-treatment and post-treatment plasma Epstein-Barr virus (EBV) DNA status. (Bottom) Overall survival is shown according to the combination of the pre-treatment and post-treatment plasma EBV DNA status. Pre-L/Post-U indicates low pre-treatment and undetectable post-treatment pEBV DNA; Pre-L/Post-D, low pre-treatment and detectable post-treatment pEBV DNA; Pre-H/Post-U, high pre-treatment and undetectable post-treatment pEBV DNA; Pre-H/Post-D, high pre-treatment and detectable post-treatment pEBV DNA.

Multivariate Analysis

Various potential prognostic factors including pre-treatment and post-treatment plasma EBV DNA levels were univariately analyzed by the log-rank test (Table 2). Liver metastasis, pre-treatment plasma EBV DNA level, and post-treatment plasma EBV DNA status were predictors for PFS, whereas liver metastasis, lung metastasis, number of lesions, pre-treatment plasma EBV DNA levels, and post-treatment plasma EBV DNA status were predictors for OS. Multivariate analysis showed pre-treatment plasma EBV DNA level and post-treatment plasma EBV DNA status were the only 2 independent predictors for both PFS and OS. No statistically significant impact on survival time could be detected for the following factors evaluated: age, sex, baseline performance status, and chemotherapy regimens (Table 2and 3).

Table 2. Univariate and Multivariate Analysis of Variables Correlated With Progression-Free Survival
CharacteristicUnivariate AnalysisMultivariate Analysis
PHR (95% CI)PHR (95% CI)
  • HR indicates hazard ratio; CI, confidence interval; KPS = Karnofsky performance score; LDH, lactate dehydrogenase; PF, cisplatin+ fluorouracil/capecitabine regimen; TPF, paclitaxel + cisplatin + fluorouracil regimen; pEBVDNA, plasma Epstein-Barr virus DNA.

  • a

    Statistically significant.

Age, <50 y vs ≥50 y.7031.091 (0.697-1.709)  
Sex, men vs women.5000.824 (0.470-1.445)  
KPS, >80 vs ≤80.4161.227 (0.75-2.006)  
Liver metastasis, yes vs no.005a0.524 (0.334-0.824).3390.785 (0.477-1.290)
Lung metastasis, yes vs no.3571.226 (0.794-1.893)  
Number of involved site, 1 vs >1.2291.313 (0.841-2.050)  
Serum LDH, ≤245 U/L vs >245 U/L.1681.515 (0.839-2.736)  
Chemotherapy regimen, PF vs TPF.4160.835 (0.542-1.288)  
Pre-treatment pEBVDNA copies/mL, <median vs ≥median<.001a2.645 (1.671-4.171).008a1.985 (1.198-3.290)
Post-treatment pEBVDNA, detectable vs undetectable<.001a3.138 (1.970-4.997).001a2.319 (1.403-3.833)
Table 3. Univariate and Multivariate Analysis of Variables Correlated With Overall Survival
CharacteristicUnivariate AnalysisMultivariate Analysis
PHR (95% CI)PHR (95% CI)
  • HR indicates hazard ratio; CI, confidence interval; KPS = Karnofsky performance score; LDH, lactate dehydrogenase; PF, cisplatin+ fluorouracil/capecitabine regimen; TPF, paclitaxel + cisplatin + fluorouracil regimen; pEBVDNA, plasma Epstein-Barr virus DNA.

  • a

    Statistically significant.

Age, <50 y vs ≥50 y.1821.48 (0.832-2.632)  
Sex, men vs women.5250.783 (0.369-1.664)  
KPS, >80 vs ≤80.7530.895 (0.449-1.784)  
Liver metastasis, yes vs no.026a0.538 (0.312-0.927)0.4530.797 (0.440-1.443)
Lung metastasis, yes vs no.047a1.804 (1.007-3.232)0.2441.441 (0.779-2.665)
Number of involved sites,1 vs >1.041a1.191 (1.013-1.344)0.5630.815 (0.407-1.631)
Serum LDH, ≤245 U/L vs >245 U/L.221a1.497 (0.785-2.855)  
Chemotherapy regimen, PF vs TPF.9531.017 (0.581-1.780)  
Pre-treatment pEBVDNA, copies/mL, <median vs ≥median.001a2.609 (1.456-4.674)0.0051.893 (1.000-3.582)
Post-treatment pEBVDNA, detectable vs undetectable.001a2.710 (1.524-4.821)0.0321.980 (1.059-3.701)

DISCUSSION

The prognostic value of pre-treatment and post-treatment plasma EBV DNA levels has been validated in nonmetastatic NPC patients treated with radiotherapy16 and locally advanced NPC patients treated with concurrent chemoradiotherapy.17, 18 In our present study of a relatively large group of metastatic/recurrent NPC patients, we first demonstrated that both pre-treatment and post-treatment plasma EBV DNA were strong predictors of PFS and OS in metastatic/recurrent NPC treated with palliative chemotherapy.

Our results demonstrated that patients with a pre-treatment plasma EBV DNA level less than the median value had significantly better PFS and OS compared with those with a pre-treatment plasma EBV DNA level equal to or above the median value. These results were inconsistent with the most recently published study by Wang et al,14 which showed that the clearance rates of plasma EBV DNA during the first month of chemotherapy rather than baseline plasma EBV DNA level could predict tumor response and patient survival. However, as we mentioned above, a rather high CR rate (41.25%) and long median survival (28 months) were reported in his cohort, which suggested a patient selection bias and relatively small sample size, because most of previous studies in the same setting reported a CR rate of 3% to 6% and a median survival of 12 to 18 months.19–21 The current study showed that the CR rate was 5.5% (7 of 127), and median OS was 17.0 months, which were in accordance with most previous reports.19–21

Moreover, our study results showed that median PFS and OS were significantly longer in patients with undetectable post-treatment plasma EBV DNA levels compared with patients with a sustained detectable post-treatment plasma EBV DNA level. Of note, 47 (57.3%) of 82 patients showed an early decline of plasma EBV DNA within the first cycle of chemotherapy. The significance of an early decrease of plasma EBV DNA level was further evaluated, and was shown to correspond with a significant increase in median PFS and OS compared with patients who showed a late drop of plasma EBV DNA to an undetectable level later than the first cycle, or patients with persistently detectable post-treatment plasma EBV DNA. These findings suggested that undetectable post-treatment plasma EBV DNA levels after the first cycle of chemotherapy could serve as an early marker of response, allowing early discrimination between patients who will and those who will not benefit from continued treatment. Compared with calculation of clearance rates of plasma EBV DNA, which need multiple plasma samples during the first month of chemotherapy, this test is easier to perform in our clinical practices.

Meanwhile, our results demonstrated that the response rate was significantly higher in patients whose post-treatment plasma EBV DNA drops to undetectable levels compared with patients whose plasma EBV DNA did not. In particular, all 7 CR patients had a post-treatment plasma EBV DNA decline to an undetectable level, whereas none of the 9 patients who progressed on imaging showed a decrease in plasma EBV DNA to an undetectable level. These findings suggest that the decrease in plasma EBV DNA level could reflect the biosensitivity of the tumor cells to cytotoxic treatment during therapy. Although undetectable post-treatment plasma EBV DNA level was associated with better response, it could not further accurately discriminate PR from SD patients, as 73.1% of PR and 54.5% of SD patients showed a drastic fall of plasma EBV DNA to an undetectable level. Similar findings have been described by Ngan et al, who evaluated plasma EBV DNA before, during, and after salvage chemotherapy on 19 patients with metastatic/recurrent NPC and found that 4 of 9 PR and 2 of 4 SD patients showed a drastic drop of plasma EBV DNA level to background level.22 Ma et al also reported that 1 patient with local recurrence NPC who was treated by gefitinib experienced minor tumor shrinkage, but post-treatment plasma EBV DNA declined to 0.23 Why plasma EBV DNA does not reflect tumor burden after chemotherapy consistently has not been clear until now. Several studies have shown that circulating cell-free plasma EBV DNA originates from the primary tumor,10 and correlates very well with the tumor mass and tumor metabolic activity by positron emission tomography/CT scan before treatment.24, 25 Once released, plasma EBV DNA exists as short fragments in the circulation instead of within intact virions, which could be eliminated very quickly.9, 10 In addition to the direct cytotoxic effect, systemic chemotherapy might reduce plasma EBV DNA by interfering with other relevant steps. For example, chemotherapy could probably inhibit the metabolic activity of tumor cells, therefore inhibiting the secreting of plasma EBV DNA from primary tumor. Chemotherapy might also have an effect on the factors that control the release rate of plasma EBV DNA into the circulation, such as the degree of tumor invasion into the adjacent vascular plexus. Further studies are needed to elucidate the mechanism.

Furthermore, by combining the baseline plasma EBV DNA level and post-treatment plasma EBV DNA status, we identified a subgroup of patients with fairly favorable prognosis, with a 5-year survival of 50.4%. This might be of clinical importance. Several previous studies suggested that a small proportion of patients with metastatic NPC were potentially curable by aggressive chemotherapy.6, 7, 26 However, no specific indicator for this subgroup of patients was identified. In the current study, we found that 21.7% of patients with low pre-treatment plasma EBV DNA level (<median) and undetectable post-treatment plasma EBV DNA level were free of disease at 5-year follow-up, suggesting that plasma EBV DNA might be considered as an indicator for aggressive chemotherapy to achieve long-term survival.

In conclusion, the current study indicated that plasma EBV DNA is of predictive value for prognosis and determining the response in patients with metastatic NPC undergoing palliative chemotherapy. The pre-treatment plasma EBV DNA level and early decrease of plasma EBV DNA to an undetectable level after chemotherapy enable early determination of patients who will benefit from continued treatment.

CONFLICT OF INTEREST DISCLOSURES

Supported by the Chinese National High Technology Research and Development Project 863, Molecular Classification and personalized diagnose and therapy of Nasopharyngeal Carcinoma (Grant No.2006AA02A404).

There are no financial disclosures from any authors.

Ancillary