- Email: [email protected]
Comparison of the prognostic impact of serum anti-EBV antibody and plasma EBV DNA assays in nasopharyngeal carcinoma
Int. J. Radiation Oncology Biol. Phys., Vol. 67, No. 1, pp. 130 –137, 2007 Copyright © 2007 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/07/$–see front matter
doi:10.1016/j.ijrobp.2006.07.012
CLINICAL INVESTIGATION
Head and Neck
COMPARISON OF THE PROGNOSTIC IMPACT OF SERUM ANTI-EBV ANTIBODY AND PLASMA EBV DNA ASSAYS IN NASOPHARYNGEAL CARCINOMA CHIH-WEN TWU, M.D.,*† WEN-YI WANG, PH.D.,‡ WEN-MIIN LIANG, PH.D.,§ JIAN-SHENG JAN, M.D.,储 RONG-SAN JIANG, M.D.,† JEFFREY CHAO, M.D.,储 YING-TAI JIN, D.D.S., M.S.,¶ 储# AND JIN-CHING LIN, M.D., PH.D.* *Institute of Clinical Medicine, School of Medicine, National Yang-Ming University, Taipei; †Department of Otorhinolaryngology, Taichung Veterans General Hospital, Taichung; ‡Section of Basic Medicine, Department of Nursing, Hung Kuang University, Taichung; §Department of Public Health, China Medical University, Taichung; 储Department of Radiation Oncology, Taichung Veterans General Hospital, Taichung; ¶Department of Pathology, National Cheng Kung University Hospital, Tainan; # Department of Medicine, School of Medicine, China Medical University, Taichung, Taiwan Purpose: Nasopharyngeal carcinoma (NPC) has been proven as an Epstein-Barr virus (EBV)-associated cancer. Serum anti-EBV antibodies and plasma EBV DNA have been investigated as surrogate markers for NPC. A comparison of the prognostic impacts of both assays has never been reported. Methods and Materials: Paired serum and plasma samples from 114 previously untreated NPC patients were collected and subjected to an immunofluorescence assay for immunoglobulin (Ig)A and IgG antibodies against the viral capsid antigen (VCA) and a real-time quantitative polymerase chain reaction assay for EBV DNA measurement. The effects of both assays on patient prognosis were thoroughly investigated. Results: Relapsed patients had significantly higher pretreatment EBV DNA concentration than patients without relapse ( p ⴝ 0.0006). No associations of VCA-IgA ( p ⴝ 0.9669) or VCA-IgG ( p ⴝ 0.6125) were observed between patients with and without relapse. The 4-year overall survival (60.3% vs. 93.1%, p < 0.0001) and relapse-free survival rates (54.4% vs. 77.9%, p ⴝ 0.0009) were significantly lower in patients with higher pretreatment EBV DNA load than in those with lower EBV DNA load. Patients with persistently detectable EBV DNA after treatment had significantly worse 4-year overall (30.8% vs. 84.6%, p < 0.0001) and relapse-free survival rates (15.4% vs. 74.0%, p < 0.0001) than those with undetectable EBV DNA. The VCA-IgA and VCA-IgG titer could not predict survivals (all p > 0.1). Cox multivariate analyses also showed the same results. Conclusion: Plasma EBV DNA is superior to serum EBV VCA antibodies in prognostic predictions for NPC. © 2007 Elsevier Inc. Nasopharyngeal carcinoma, Epstein-Barr virus, Immunofluorescence, Quantitative polymerase chain reaction.
Nasopharyngeal carcinoma (NPC) is distinguished from other cancers of the head and neck by its epidemiology, histopathology, clinical characteristics, and methods of therapy. It is a geographically endemic, Epstein-Barr virus (EBV)-associated carcinoma of epidermoid origin. It has poorly differentiated or undifferentiated pathology with a higher incidence of neck lymph node metastasis, and greater radiosensitivity and chemosensitivity. Radiotherapy is the primary treatment of NPC because of the anatomic constraints and a high degree of radiosensitivity. Recent trials
support the use of combined chemotherapy and radiotherapy for patients with advanced NPC (1– 4). Nasopharyngeal carcinoma has been proven as an EBVassociated cancer for a long time by several lines of evidence: the presence of deoxyribonucleic acid (DNA), ribonucleic acid, and proteins of EBV in almost all cancer cells of nearly all tissue samples (primary and various metastatic sites) obtained from patients with NPC (5–7); tumor cells proved to be derived from a single EBV-infected cell (8); and high levels of EBV protein antibodies in newly found NPC patients (9 –11) and in healthy individuals in whom NPC later developed (12, 13). Furthermore, EBV has been
Reprint requests to: Jin-Ching Lin, M.D., Ph.D., Department of Radiation Oncology, Taichung Veterans General Hospital, Taiwan, No. 160, Sec. 3, Taichung-Kang Rd., Taichung 407, Taiwan. Tel: (⫹886) 4-23592525 ext. 5613; Fax: (⫹886) 4-23741316; E-mail: [email protected] This study was supported by grants from the National Science
Council (NSC89-2314-B-075A-021-M08, NSC90-2314-B-075A007, and NSC91-2314-B-075A-004-M08) and Taichung Veterans General Hospital (TCVGH-937104C), Taiwan. Conflict of interest: none. Received May 3, 2006, and in revised form July 7, 2006. Accepted for publication July 7, 2006.
INTRODUCTION
130
EBV antibody and DNA in nasopharyngeal carcinoma
detected in premalignant (preinvasive) nasopharyngeal lesions, including carcinoma in situ and dysplasia (14, 15). Thus, EBV has been investigated as a surrogate marker for NPC in past decades. Multiple EBV-specific antibodies were detectable in patients with NPC (9 –11, 16 –18). Among these markers, the immunoglobulin A (IgA) antibody against viral capsid antigen (VCA-IgA), measured by indirect immunofluorescence (9 –11, 18) or enzyme-linked immunosorbent assays (17), is one of the most widely used antibody markers for assisting in diagnosis (9 –11, 16, 18), and for screening (12, 13) of NPC. Recently, the quantification of plasma EBV DNA by the real-time quantitative polymerase chain reaction (RTQ-PCR) has been demonstrated as a useful marker in the detection, monitoring, and prognostic prediction for NPC (19 –21). Comparing the prognostic impact of these two different assays in NPC has never been reported. Thus, we investigated the prognostic effects between serum antiVCA antibodies and plasma EBV DNA load on NPC patients.
METHODS AND MATERIALS Study subjects and blood sampling One hundred fourteen patients with previously untreated, biopsyproven NPC and no evidence of distant metastasis were included. The routine staging workup included clinical examinations of the head and neck region, fiberoptic nasopharyngoscopy, computed tomographic or magnetic resonance imaging scan from the skull base to the whole neck, chest radiography, whole-body bone scan, abdominal sonography, complete blood count with cell differentials of leukocytes, and biochemical profiling. Chest computed tomography and bone marrow biopsy were performed when lung metastasis was suspected on the basis of chest X-ray films, and abnormal blood routines were noted. The cancer stage was defined according to the 1997 American Joint Committee on Cancer TNM staging system. All patients received five cycles of neoadjuvant chemotherapy (cisplatin 60 mg/m2 2-h infusion on Day 1, 5-fluorouracil 2500 mg/m2 mixed with leucovorin 250 mg/m2 24-h continuous intravenous infusion on Day 8, repeated every 2 weeks) followed by radiotherapy 70 –74 Gy/7– 8 weeks. Table 1 lists the pretreatment patient and tumor characteristics of the 114 patients. The TNM stage distribution indicated that most (72.8%) patients had Stage IV disease. After a median follow-up of 46 (range 22 to 67) months, there were 33 failures—5 in nasopharynx, 1 in neck, 23 in distant sites, 3 in nasopharynx plus distant sites, and 1 in neck plus distant site. Paired serum and plasma samples from NPC patients were collected and subjected to immunofluorescence assays for IgA and IgG antibodies against the viral capsid antigen (VCA) and a RTQ-PCR assay for plasma EBV DNA measurement. The timing of blood samplings was before treatment and 1 week after completion of chemoradiotherapy. Fifty healthy adults without histories of any malignancy and 63 patients with squamous cell carcinoma of the head and neck other than nasopharynx (SCCHN) were enrolled as controls. The primary sites of these 63 SCCHN patients were as follows: buccal mucosa (21), tongue (15), gingiva (2), lip (1), larynx (6), hypopharynx (5), tonsil (5), palate (4), oropharynx (3), and middle ear (1).
●
C.-W. TWU et al.
131
Table 1. Patient characteristics (n ⫽ 114) Characteristics Age (years) Range, 24–74 Median, 46 Mean, 46 Sex Male Female Karnofsky scale ⱖ90% ⬍90% Pathology (WHO classification) Type I Type II Type III T stage (1997 AJCC) T1 T2 T3 T4 N stage (1997 AJCC) N0 N1 N2 N3 Overall stage II III IV Failure site(s) Nasopharynx Neck Distant Nasopharynx ⫹ distant Neck ⫹ distant
No. of cases
%
87 27
76.3 23.7
29 85
25.4 74.6
4 90 20
3.5 78.9 17.6
7 31 21 55
6.1 27.2 18.4 48.3
2 3 70 39
1.8 2.6 61.4 34.2
1 30 83
0.9 26.3 72.8
5 1 23 3 1
4.4 0.9 20.2 2.6 0.9
Abbreviations: WHO ⫽ World Health Organization; AJCC ⫽ American Joint Committee on Cancer.
The Institutional Review Board of our hospital approved this study, and informed consent was obtained from each study subject.
Immunofluorescence assay for serum VCA antibodies titers Immunofluoresence tests of IgA and IgG antibodies against EBV VCA were performed routinely using commercially available antigenic slides. Titers were considered positive if VCA-IgA ⱖ1:10 and VCA-IgG ⱖ1:40.
Real-time quantitative polymerase chain reaction for plasma EBV DNA measurement Plasma DNA was extracted using a QIAamp DNA Blood MiniKit (Qiagen, Hilden, Germany). EBV DNA concentrations of plasma were measured using a RTQ-PCR assay toward the BamHI-W region of the EBV genome. The sequences of forward and reverse primers were 5=-CCCAACACTCCACCACACC-3= and 5=-TCTTAGGAGCTGTCCGAGGG-3=. A dual fluorescencelabeled oligomer, 5=-(FAM)CACACACTACACACACCCACCCGTCTC(TAMRA)-3=, served as a probe. The principles of the RTQ-PCR and detailed reaction setup procedures were as described previously (19, 21).
132
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 67, Number 1, 2007
Table 2. Pretreatment viral load and antibody titers versus relapse or not EBV DNA Median Relapse No relapse Wilcoxon rank-sum test
4669 681
Interquartile range 779–13990 207–3225 p ⫽ 0.0006
VCA-IgA Median 1:40 1:40
VCA-IgG
Interquartile range 1:10–1:80 1:10–1:80 p ⫽ 0.9669
Median 1:2560 1:2560
Interquartile range 1:1280–1:5120 1:1280–1:5120 p ⫽ 0.6125
Abbreviations: EBV ⫽ Epstein-Barr virus; VCA ⫽ viral capsid antigen; Ig ⫽ immunoglobulin.
Statistical analysis The relapse-free survival was calculated from the first day of chemoradiotherapy until the date of disease relapse or until the date of the last follow-up visit. The overall survival was calculated from the first day of chemoradiotherapy until the date of death or until the date of the last follow-up visit. Life-table estimation was done according to the method of Kaplan-Meier. Univariate comparisons of survival curves were performed by use of the log–rank test. All statistical tests were two-sided, and p values of less than 0.05 were considered statistically significant. The multivariate Cox proportional hazards model was used to estimate the hazard ratios and 95% confidence intervals. Variables in the model included age, sex, performance status, pathologic type, response to neoadjuvant chemotherapy, T-stage, N-stage, overall stage, pretreatment and posttreatment plasma EBV DNA levels, pretreatment and posttreatment serum VCA-IgA titers, and pretreatment and posttreatment serum VCA-IgG titers. The relationship between the
plasma EBV DNA level or serum antibodies titers and the relapse rate was evaluated by using the chi-square test. Analyses were performed by use of SAS (Version 8.0; SAS Institute, Inc., Cary, NC).
RESULTS Sensitivity and specificity One hundred six of 114 NPC patients had detectable plasma EBV DNA by RTQ-PCR before treatment, whereas no detectable EBV DNA was found in 63 SCCHN patients and 50 healthy people. The sensitivity and specificity by the RTQ-PCR assay are 93.0% and 100%, respectively. The VCA-IgA antibody titers were positive in 82.5% (94/114) NPC patients, 46.0% (29/63) SCCHN, and 54.5% (27/50) healthy people. The VCA-IgG antibody titers were 100%
Fig. 1. Overall survival (A, C) and relapse-free survival (B, D) by plasma Epstein-Barr virus (EBV) deoxyribonucleic acid levels.
EBV antibody and DNA in nasopharyngeal carcinoma
positive for all tested samples, including 114 NPC patients, 65 SCCHN patients, and 50 healthy people. Thus, plasma EBV DNA proved to be superior to serum VCA antibodies in sensitivity and specificity. Effect of pretreatment EBV DNA levels or VCA antibody titers on subsequent relapse To study the prognostic impacts of the plasma EBV DNA level at initial presentation on subsequent relapses, we compared the pretreatment plasma EBV DNA concentrations between patients with and without tumor relapses. The median plasma EBV DNA levels in patients with (33 cases) and without (81 cases) relapse were 4669 copies/mL (interquartile range, 779 to 13,990) and 681 copies/mL (interquartile range, 207 to 3,225), respectively. The difference between these two groups was statistically significant ( p ⫽ 0.0006). There were no significant differences for pretreatment VCA-IgA ( p ⫽ 0.9669) or VCA-IgG ( p ⫽ 0.6125) titers in patients with or without relapse (Table 2). Relapse rates according to EBV DNA level or EBV antibody titers Twenty-two of 51 (43.1%) patients with pretreatment plasma EBV DNA ⱖ1500 copies/mL later developed tumor relapse, whereas only 17.5% (11/63) patients with pretreatment plasma EBV DNA ⬍1,500 copies/mL experienced
●
C.-W. TWU et al.
133
tumor relapse ( p ⫽ 0.0027). The pretreatment serum VCAIgA and VCA-IgG titers had limited value in predicting subsequent tumor relapse. The relapse rates between patients with VCA-IgA ⱖ1:80 and those with VCA-IgA ⬍ 1:80 were 27.0% (10/37) and 29.9% (23/77), respectively (p ⫽ 0.7540). The corresponding relapse rates for VCAIgG ⱖ and ⬍1:2560 were 27.8% (22/79) and 31.8% (11/ 35), respectively ( p ⫽ 0.6974). The presence or absence of plasma EBV DNA 1 week after chemoradiotherapy showed significant association to subsequent tumor relapse. The relapse rates between patients with plasma EBV DNA ⬎ and ⫽ 0 after chemoradiotherapy were 84.6% (11/13) and 21.8% (22/101), respectively ( p ⫽ 0.0002). However, the posttreatment serum VCA-IgA and VCA-IgG titers did not show any significant impact on subsequent tumor relapse. The relapse rates between patients with posttreatment VCA-IgA ⱖ1:80 and those with VCA-IgA ⬍1:80 were 23.9% (11/46) and 32.4% (22/68), respectively ( p ⫽ 0.3297). The corresponding relapse rates for VCA-IgG ⱖ and ⬍1:2560 were 25.4% (17/67) and 34.0% (16/47), respectively ( p ⫽ 0.3151). Correlation of the plasma EBV DNA levels and sites of first failure The pretreatment plasma EBV DNA levels in patients with distant failure (median, 5085; interquartile range,
Fig. 2. Overall survival (A, C) and relapse-free survival (B, D) by serum viral capsid antigen–immunoglobulin A (VCA-IgA) titers.
134
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 67, Number 1, 2007
Fig. 3. Overall survival (A, C) and relapse-free survival (B, D) by serum viral capsid antigen–immunoglobulin G (VCA-IgG) titers.
1,126 to 26,381 copies/mL) were higher than those with locoregional failure (median, 603; interquartile range, 113 to 10,799 copies/mL). The difference reached statistically significant ( p ⫽ 0.0443) levels. Thus, the patients with higher EBV DNA levels have a higher chance of developing microscopic metastases at presentation. Regardless of local therapy with chemoradiotherapy, those patients relapsed more frequently at distant than at local or regional sites. Survival analysis Kaplan-Meier estimates showed that both pretreatment EBV DNA levels and the presence/absence of EBV DNA in plasma samples after treatment significantly affected overall survival and relapse-free survival rates. The overall survival at 4 years was 93.1% in patients with pretreatment plasma EBV DNA ⬍1,500 copies/mL, and 60.3% in those with pretreatment plasma EBV DNA ⱖ1500 copies/mL ( p ⬍ 0.0001, Fig. 1A). The 4-year relapse-free survival rates of the patients with pretreatment plasma EBV DNA ⬍ and ⱖ1,500 copies/mL were 77.9% and 54.4%, respectively (p ⫽ 0.0009, Fig. 1B). The 4-year overall survival rates between patients with persistently detectable plasma EBV DNA and those with undetectable EBV DNA after chemoradiotherapy were 30.8% and 84.6%, respectively ( p ⬍ 0.0001, Fig. 1C). The relapse-free survival rate at 4 years
was significantly worse in the patients with persistently detectable plasma EBV DNA than in those with undetectable EBV DNA after radiotherapy (15.4% vs. 74.0%, p ⬍ 0.0001, Fig. 1D). Figures 2 and 3 illustrate detailed survival curves according to antibody titers of VCA-IgA and VCAIgG, measured at either pretreatment or posttreatment. The curves clearly showed no significant relationship between antibody titers and survival rates. Cox multivariate analysis When EBV data (serum EBV antibody titers and plasma EBV DNA levels) along with clinical parameters (including age, sex, performance status, pathologic type, response to neoadjuvant chemotherapy, T-stage, N-stage, and overall stage) were entered for Cox multivariate analysis, both pretreatment and posttreatment plasma EBV DNA levels were shown to be the most important prognostic factors for both overall survival (pretreatment, p ⫽ 0.001; hazard ratio ⫽ 9.16; 95% confidence interval [CI], 2.49 –33.64; posttreatment, p ⬍ 0.001; hazard ratio ⫽ 13.3; 95% CI, 3.92– 45.11) and relapse-free survival (pretreatment, p ⫽ 0.021; hazard ratio ⫽ 2.68; 95% CI, 1.16 – 6.18; posttreatment, p ⬍ 0.001; hazard ratio ⫽ 15.7; 95% CI, 6.02– 40.79) after adjustment for other variables (Table 3). The pretreatment and posttreatment serum VCA-IgA and VCA-IgG titers had no appreciable relationship to either overall or relapse-free
EBV antibody and DNA in nasopharyngeal carcinoma
●
C.-W. TWU et al.
135
Table 3. Summary of survival analyses by multivariate Cox proportional hazards model Overall survival Variables Age, years ⱕ45 vs. ⬎45 Sex Female vs. male Karnofsky performance status ⬍90% vs. ⱖ90% Pathology (WHO type) I vs. II/III Response to NeoCT CR vs. PR⫹SD T stage T3–4 vs. T1–2 N stage N3 vs. N0–2 Overall stage IV vs. II⫹III Pretreatment plasma EBV DNA ⱖ1,500 vs. ⬍1,500 Posttreatment plasma EBV DNA ⫹ vs. ⫺ Pretreatment serum VCA-IgA ⱖ1:80 vs. ⬍1:80 Posttreatment serum VCA-IgA ⱖ1:80 vs. ⬍1:80 Pretreatment serum VCA-IgG ⱖ1:2560 vs. ⬍1:2560 Posttreatment serum VCA-IgG ⱖ1:2560 vs. ⬍1:2560
p value
HR (95% CI)
Relapse-free survival p value
HR (95% CI)
0.401
0.66 (0.26–1.73)
0.235
0.60 (0.26–1.39)
0.412
1.64 (0.50–5.34)
0.979
1.01 (0.42–2.47)
0.027
4.99 (1.20–20.74)
0.999
1.00 (0.32–3.15)
0.794
1.37 (0.13–14.27)
0.026
5.32 (1.22–23.26)
0.968
1.02 (0.41–2.57)
0.712
1.16 (0.53–2.54)
0.411
1.63 (0.51–5.18)
0.658
1.26 (0.45–3.54)
0.012
5.32 (1.44–19.68)
0.082
2.46 (0.89–6.79)
0.299
0.41 (0.08–2.20)
0.332
2.17 (0.45–10.35)
0.001
9.16 (2.49–33.64)
0.021
2.68 (1.16–6.18)
⬍0.001
13.3 (3.92–45.11)
⬍0.001
15.7 (6.02–40.79)
0.533
0.70 (0.23–2.13)
0.568
0.76 (0.30–1.93)
0.319
0.60 (0.22–1.63)
0.365
0.67 (0.28–1.60)
0.078
0.39 (0.13–1.12)
0.834
0.91 (0.37–2.21)
0.800
0.88 (0.31–2.45)
0.777
1.14 (0.46–2.84)
Abbreviations: HR ⫽ hazard ratio; CI ⫽ confidence interval; WHO ⫽ World Health Organization; NeoCT ⫽ neoadjuvant chemotherapy; PR ⫽ partial response; CR ⫽ complete response; SD ⫽ stable disease; EBV ⫽ Epstein-Barr virus; VCA ⫽ viral capsid antigen; Ig ⫽ immunoglobulin.
survival rates. Table 3 also demonstrated that plasma EBV DNA level had greater effects than any other clinical factors on survivals. DISCUSSION Previously, NPC had no reliable tumor markers. Because NPC has a close association with the EBV, various antiEBV antibodies have been studied as surrogate markers of NPC since 1970 (9 –13, 16 –18). However, most studies used the same design: the collection of blood samples from NPC patients with active disease (previously untreated or relapsed cases) in addition to various controls (healthy volunteer or non-NPC tumors) and then comparing the sensitivity and specificity of different assay systems targeting various EBV antigens (9 –11, 18). All studies demonstrated higher positive detection rates of various anti-EBV antibodies for patients with active disease than those of control subjects. However, these data prove to be of little help in the diagnosis of newly found NPC or relapsed cases. The definite diagnosis of a newly found NPC or a relapsed case depends on suspected symptoms/signs, nasopharyngoscopic and imaging findings, and confirmatory biopsy results, irrespective of the titer of various antibodies. Further-
more, the antibody titer remains persistently high in most patients in remission after treatment (22, 23). There is no reliable cutoff value in the differential diagnosis of recurrence or remission. A clinically useful tumor marker should not only have higher sensitivity and specificity in the diagnosis setting but also have significant impact on the prognosis prediction. The latter is more important than sensitivity and specificity calculations alone in clinical practice. Only few studies have previously addressed the prognostic effects of various antiEBV antibodies (24 –27). By the antibody-dependent cellular cytotoxicity assay, which measures antibodies to an EBV-induced membrane antigen component, Neel et al. demonstrated that patients with low antibody-dependent cellular cytotoxicity titers at diagnosis had significantly worse progression-free survival ( p ⫽ 0.0001) and overall survival ( p ⫽ 0.0008) for World Health Organization types II and III NPC patients (24). Conventional antibody (VCAIgG, early antigen (EA)-IgA, and EA-IgG) assays before treatment could not reflect the prognosis (24). After longterm follow-up, they also found that the posttreatment sequential measurements did not accurately predict the outcome (25). A multicenter follow-up study on 319 NPC patients showed no prognostic values of initial serology,
136
I. J. Radiation Oncology
●
Biology
●
Physics
including IgG and IgA antibodies to VCA, EA or EBV nuclear antigen (EBNA) (26). However, increasing titers of EA-IgA and EA-IgG 1 year after completion of radiotherapy became highly significant for prediction of relapse, regardless of the initial titers (26). Yip et al. investigated the prognostic role of the IgG antibody against EBV BZLF-1 replication activators and found no significant difference in patients’ survival between the low and high titers for sera taken before treatment ( p ⫽ 0.331) (27). In contrast, for sera taken 10 months after radiotherapy, the high titer group predicted significantly worse overall survival rates ( p ⫽ 0.0019) (27). Based on the foregoing discussion, the definite role of anti-EBV antibodies on prognosis prediction is still controversial and the precise point of blood sampling has not been well established. On the contrary, circulating EBV DNA loads measured before or after treatment are consistently correlated with the survival and prognosis of NPC patients (19 –21). This may relate to the essential difference between a marker directly released from tumor cells (EBV DNA) and a marker that essentially is a host response to viral tumor antigens. The latter marker inevitably lags behind with regard to kinetics upon treatment. Also, the dynamic range of antibody responses is fairly limited in general, with only weak correlations observed between the extent of exposure and the level (titer) of the response. On the other hand, unlike conventional protein markers such as carcinoembryonic antigen or prostate-specific antigen, released mostly from tumor cells, EBV can be found in patients with EBV-related diseases or even in normal populations in high concentration, especially after an infection with EBV. These situations should be considered in the clinical interpretation of EBV DNA as a tumor marker for NPC.
Volume 67, Number 1, 2007
To the best of our knowledge, there are only five papers comparing the EBV-specific antibody-based and DNAbased assays in NPC (22, 23, 28 –30). They have usually focused on sensitivity, specificity, or correlation with staging. The major findings of these studies are that the EBV DNA assay is superior to the anti-EBV antibody assay in both sensitivity/specificity calculation (23, 28 –30) and differential diagnosis of active disease or remission state (22, 23). In addition, plasma EBV DNA levels have better correlation with clinical staging than VCA-IgA titers (22). So far, no other report has been shown to compare the prognostic values between anti-EBV antibody and EBV DNA assays. In this study, except for routine sensitivity/specificity calculations, we first report a detailed comparison of the prognostic value between serum anti-VCA antibody titers and plasma EBV DNA copy numbers. Our results clearly demonstrate that plasma EBV DNA levels are superior to serum VCA-IgA or VCA-IgG titers for all analyzed items, including parameters of sensitivity, specificity, prediction of subsequent relapse, survival curves, and prognostic factors analyses. Results of this study will have a significant impact in the management of NPC. Patients with high risk of relapse (persistently detectable EBV DNA after treatment or higher EBV DNA level before treatment) should be further studied for a more aggressive therapy, and routine adjuvant therapy should be prohibited for patients with low risk of relapse to avoid unnecessary morbidity and waste of medical resources. Of note, one limitation of our study is that only antibody against the VCA was compared. Comparisons of the prognostic impact between the EBV DNA assay and many different antibody assays against other antigens are still lacking and need to be investigated further.
REFERENCES 1. Al-Sarraf M, LeBlanc M, Giri PGS, et al. Chemoradiotherapy versus radiotherapy in patients with advanced nasopharyngeal cancer: Phase III randomized Intergroup study 0099. J Clin Oncol 1998;16:1310 –1317. 2. Lin JC, Jan JS, Hsu CY, et al. Phase III study of concurrent chemoradiotherapy versus radiotherapy alone for advanced nasopharyngeal carcinoma: Positive effect on overall and progression-free survival. J Clin Oncol 2003;21:631– 637. 3. Lee AWM, Lau WH, Tung SY, et al. Preliminary results of a randomized study on therapeutic gain by concurrent chemotherapy for regionally-advanced nasopharyngeal carcinoma: NPC-9901 trial by the Hong Kong Nasopharyngeal Cancer Study Group. J Clin Oncol 2005;23:6966 – 6975. 4. Wee J, Tan EH, Tai BC, et al. Randomized trial of radiotherapy versus concurrent chemoradiotherapy followed by adjuvant chemotherapy in patients with American Joint Committee on Cancer/International Union Against Cancer stage III and IV nasopharyngeal cancer of the endemic variety. J Clin Oncol 2005;23:6730 – 6738. 5. Chang YS, Tyan YS, Liu ST, et al. Detection of Epstein-Barr virus DNA sequences in nasopharyngeal carcinoma cells by enzymatic DNA amplification. J Clin Microbiol 1990;28:2398 – 2402. 6. Chen CL, Wen WN, Chen JY, et al. Detection of Epstein-Barr
7.
8. 9. 10. 11. 12. 13. 14.
virus genome in nasopharyngeal carcinoma by in situ DNA hybridization. Intervirology 1993;36:91–98. Tsai ST, Jin YT, Su IJ. Expression of EBER1 in primary and metastatic nasopharyngeal carcinoma tissues using in situ hybridization: A correlation with WHO histologic subtypes. Cancer 1996;77:231–236. Raab-Traub N, Flynn K. The structure of the termini of the Epstein-Barr virus as a marker of clonal cellular proliferation. Cell 1986;47:883– 889. Henle W, Henle G, Ho HC, et al. Antibodies to Epstein-Barr virus in nasopharyngeal carcinoma, other head and neck neoplasms and control groups. J Natl Cancer Inst 1970;44:225–231. Henle G, Henle W. Epstein-Barr virus-specific IgA serum antibodies as an outstanding feature of nasopharyngeal carcinoma. Int J Cancer 1976;17:1–7. Ho HC, Ng MH, Kwan HC, et al. Epstein-Barr virus-specific IgA and IgG serum antibodies in nasopharyngeal carcinoma. Br J Cancer 1976;34:655– 660. Chien YC, Chen JY, Liu MY, et al. Serologic markers of Epstein-Barr virus infection and nasopharyngeal carcinoma in Taiwanese men. N Engl J Med 2001;345:1877–1882. Zeng Y, Zhong JM, Li LY, et al. Follow-up studies on Epstein-Barr virus IgA/VCA antibody-positive persons in Zangwu County, China. Intervirology 1983;20:190 –194. Pathmanathan R, Prasad U, Sadler R, et al. Clonal prolifera-
EBV antibody and DNA in nasopharyngeal carcinoma
15. 16.
17.
18.
19. 20.
21. 22.
tions of cells infected with Epstein-Barr virus in preinvasive lesions related to nasopharyngeal carcinoma. N Engl J Med 1995;333:693– 698. Yeung WM, Zong YS, Chiu CT, et al. Epstein-Barr virus carriage by nasopharyngeal carcinoma in situ. Int J Cancer 1993;53:746 –750. Liu MY, Chang YL, Ma J, et al. Evaluation of multiple antibodies to Epstein-Barr virus as markers for detecting patients with nasopharyngeal carcinoma. J Med Virol 1997;52: 262–269. Uen WC, Luka J, Pearson GR. Development of an enzymelinked immunosorbent assay (ELISA) for detecting IgA antibodies to the Epstein-Barr virus. Int J Cancer 1988;41:479 – 482. Neel HB 3rd, Pearson GR, Taylor WF. Antibodies to EpsteinBarr virus in patients with nasopharyngeal carcinoma and in comparison groups. Ann Otol Rhinol Laryngol 1984;93:477– 482. Lo YMD, Chan LYS, Lo KW, et al. Quantitative analysis of cell-free Epstein-Barr virus DNA in plasma of patients with nasopharyngeal carcinoma. Cancer Res 1999;59:1188 –1191. Chan ATC, Lo YMD, Zee B, et al. Plasma Epstein-Barr virus DNA and residual disease after radiotherapy for undifferentiated nasopharyngeal carcinoma. J Natl Cancer Inst 2002;94: 1614 –1619. Lin JC, Wang WY, Chen KY, et al. Quantification of plasma Epstein-Barr virus DNA in patients with advanced nasopharyngeal carcinoma. N Engl J Med 2004;350:2461–2470. Shao JY, Li YH, Gao HY, et al. Comparison of plasma Epstein-Barr virus (EBV) DNA levels and serum EBV immunoglobulin A/viral capsid antigen antibody titers in patients with nasopharyngeal carcinoma. Cancer 2004;100:1162–1170.
●
C.-W. TWU et al.
137
23. Fan H, Nicholls J, Chua D, et al. Laboratory markers of tumor burden in nasopharyngeal carcinoma: A comparison of viral load and serologic tests for Epstein-Barr virus. Int J Cancer 2004;112:1036 –1041. 24. Neel HB 3rd, Pearson GR, Taylor WF. Antibody-dependent cellular cytotoxicity: Relation to stage and disease course in North American patients with nasopharyngeal carcinoma. Arch Otolaryngol 1984;110:742–747. 25. Neel HB 3rd, Taylor WF. Epstein-Barr virus-related antibody. Change in titers after therapy for nasopharyngeal carcinoma. Arch Otolaryngol Head Neck Surg 1990;116:1287–1290. 26. de Vathaire F, Sancho-Garnier H, de-The H, et al. Prognostic value of EBV markers in the clinical management of nasopharyngeal carcinoma (NPC): A multicenter follow-up study. Int J Cancer 1988;42:176 –181. 27. Yip TTC, Ngan RKC, Lau WH, et al. A possible prognostic role of immunoglobulin-G antibody against recombinant Epstein-Barr virus BZLF-1 transactivator protein ZEBRA in patients with nasopharyngeal carcinoma. Cancer 1994;74: 2414 –2424. 28. Chan KH, Gu YL, Ng F, et al. EBV specific antibody-based and DNA-based assays in serologic diagnosis of nasopharyngeal carcinoma. Int J Cancer 2003;105:706 –709. 29. Leung SF, Tam JS, Chan ATC, et al. Improved accuracy of detection of nasopharyngeal carcinoma by combined application of circulating Epstein-Barr virus DNA and anti-EpsteinBarr viral capsid antigen IgA antibody. Clin Chem 2004;50: 339 –345. 30. Kondo S, Horikawa T, Takeshita H, et al. Diagnostic value of serum EBV-DNA quantification and antibody to viral capsid antigen in nasopharyngeal carcinoma patients. Cancer Sci 2004;95:508 –513.
doi:10.1016/j.ijrobp.2006.07.012
CLINICAL INVESTIGATION
Head and Neck
COMPARISON OF THE PROGNOSTIC IMPACT OF SERUM ANTI-EBV ANTIBODY AND PLASMA EBV DNA ASSAYS IN NASOPHARYNGEAL CARCINOMA CHIH-WEN TWU, M.D.,*† WEN-YI WANG, PH.D.,‡ WEN-MIIN LIANG, PH.D.,§ JIAN-SHENG JAN, M.D.,储 RONG-SAN JIANG, M.D.,† JEFFREY CHAO, M.D.,储 YING-TAI JIN, D.D.S., M.S.,¶ 储# AND JIN-CHING LIN, M.D., PH.D.* *Institute of Clinical Medicine, School of Medicine, National Yang-Ming University, Taipei; †Department of Otorhinolaryngology, Taichung Veterans General Hospital, Taichung; ‡Section of Basic Medicine, Department of Nursing, Hung Kuang University, Taichung; §Department of Public Health, China Medical University, Taichung; 储Department of Radiation Oncology, Taichung Veterans General Hospital, Taichung; ¶Department of Pathology, National Cheng Kung University Hospital, Tainan; # Department of Medicine, School of Medicine, China Medical University, Taichung, Taiwan Purpose: Nasopharyngeal carcinoma (NPC) has been proven as an Epstein-Barr virus (EBV)-associated cancer. Serum anti-EBV antibodies and plasma EBV DNA have been investigated as surrogate markers for NPC. A comparison of the prognostic impacts of both assays has never been reported. Methods and Materials: Paired serum and plasma samples from 114 previously untreated NPC patients were collected and subjected to an immunofluorescence assay for immunoglobulin (Ig)A and IgG antibodies against the viral capsid antigen (VCA) and a real-time quantitative polymerase chain reaction assay for EBV DNA measurement. The effects of both assays on patient prognosis were thoroughly investigated. Results: Relapsed patients had significantly higher pretreatment EBV DNA concentration than patients without relapse ( p ⴝ 0.0006). No associations of VCA-IgA ( p ⴝ 0.9669) or VCA-IgG ( p ⴝ 0.6125) were observed between patients with and without relapse. The 4-year overall survival (60.3% vs. 93.1%, p < 0.0001) and relapse-free survival rates (54.4% vs. 77.9%, p ⴝ 0.0009) were significantly lower in patients with higher pretreatment EBV DNA load than in those with lower EBV DNA load. Patients with persistently detectable EBV DNA after treatment had significantly worse 4-year overall (30.8% vs. 84.6%, p < 0.0001) and relapse-free survival rates (15.4% vs. 74.0%, p < 0.0001) than those with undetectable EBV DNA. The VCA-IgA and VCA-IgG titer could not predict survivals (all p > 0.1). Cox multivariate analyses also showed the same results. Conclusion: Plasma EBV DNA is superior to serum EBV VCA antibodies in prognostic predictions for NPC. © 2007 Elsevier Inc. Nasopharyngeal carcinoma, Epstein-Barr virus, Immunofluorescence, Quantitative polymerase chain reaction.
Nasopharyngeal carcinoma (NPC) is distinguished from other cancers of the head and neck by its epidemiology, histopathology, clinical characteristics, and methods of therapy. It is a geographically endemic, Epstein-Barr virus (EBV)-associated carcinoma of epidermoid origin. It has poorly differentiated or undifferentiated pathology with a higher incidence of neck lymph node metastasis, and greater radiosensitivity and chemosensitivity. Radiotherapy is the primary treatment of NPC because of the anatomic constraints and a high degree of radiosensitivity. Recent trials
support the use of combined chemotherapy and radiotherapy for patients with advanced NPC (1– 4). Nasopharyngeal carcinoma has been proven as an EBVassociated cancer for a long time by several lines of evidence: the presence of deoxyribonucleic acid (DNA), ribonucleic acid, and proteins of EBV in almost all cancer cells of nearly all tissue samples (primary and various metastatic sites) obtained from patients with NPC (5–7); tumor cells proved to be derived from a single EBV-infected cell (8); and high levels of EBV protein antibodies in newly found NPC patients (9 –11) and in healthy individuals in whom NPC later developed (12, 13). Furthermore, EBV has been
Reprint requests to: Jin-Ching Lin, M.D., Ph.D., Department of Radiation Oncology, Taichung Veterans General Hospital, Taiwan, No. 160, Sec. 3, Taichung-Kang Rd., Taichung 407, Taiwan. Tel: (⫹886) 4-23592525 ext. 5613; Fax: (⫹886) 4-23741316; E-mail: [email protected] This study was supported by grants from the National Science
Council (NSC89-2314-B-075A-021-M08, NSC90-2314-B-075A007, and NSC91-2314-B-075A-004-M08) and Taichung Veterans General Hospital (TCVGH-937104C), Taiwan. Conflict of interest: none. Received May 3, 2006, and in revised form July 7, 2006. Accepted for publication July 7, 2006.
INTRODUCTION
130
EBV antibody and DNA in nasopharyngeal carcinoma
detected in premalignant (preinvasive) nasopharyngeal lesions, including carcinoma in situ and dysplasia (14, 15). Thus, EBV has been investigated as a surrogate marker for NPC in past decades. Multiple EBV-specific antibodies were detectable in patients with NPC (9 –11, 16 –18). Among these markers, the immunoglobulin A (IgA) antibody against viral capsid antigen (VCA-IgA), measured by indirect immunofluorescence (9 –11, 18) or enzyme-linked immunosorbent assays (17), is one of the most widely used antibody markers for assisting in diagnosis (9 –11, 16, 18), and for screening (12, 13) of NPC. Recently, the quantification of plasma EBV DNA by the real-time quantitative polymerase chain reaction (RTQ-PCR) has been demonstrated as a useful marker in the detection, monitoring, and prognostic prediction for NPC (19 –21). Comparing the prognostic impact of these two different assays in NPC has never been reported. Thus, we investigated the prognostic effects between serum antiVCA antibodies and plasma EBV DNA load on NPC patients.
METHODS AND MATERIALS Study subjects and blood sampling One hundred fourteen patients with previously untreated, biopsyproven NPC and no evidence of distant metastasis were included. The routine staging workup included clinical examinations of the head and neck region, fiberoptic nasopharyngoscopy, computed tomographic or magnetic resonance imaging scan from the skull base to the whole neck, chest radiography, whole-body bone scan, abdominal sonography, complete blood count with cell differentials of leukocytes, and biochemical profiling. Chest computed tomography and bone marrow biopsy were performed when lung metastasis was suspected on the basis of chest X-ray films, and abnormal blood routines were noted. The cancer stage was defined according to the 1997 American Joint Committee on Cancer TNM staging system. All patients received five cycles of neoadjuvant chemotherapy (cisplatin 60 mg/m2 2-h infusion on Day 1, 5-fluorouracil 2500 mg/m2 mixed with leucovorin 250 mg/m2 24-h continuous intravenous infusion on Day 8, repeated every 2 weeks) followed by radiotherapy 70 –74 Gy/7– 8 weeks. Table 1 lists the pretreatment patient and tumor characteristics of the 114 patients. The TNM stage distribution indicated that most (72.8%) patients had Stage IV disease. After a median follow-up of 46 (range 22 to 67) months, there were 33 failures—5 in nasopharynx, 1 in neck, 23 in distant sites, 3 in nasopharynx plus distant sites, and 1 in neck plus distant site. Paired serum and plasma samples from NPC patients were collected and subjected to immunofluorescence assays for IgA and IgG antibodies against the viral capsid antigen (VCA) and a RTQ-PCR assay for plasma EBV DNA measurement. The timing of blood samplings was before treatment and 1 week after completion of chemoradiotherapy. Fifty healthy adults without histories of any malignancy and 63 patients with squamous cell carcinoma of the head and neck other than nasopharynx (SCCHN) were enrolled as controls. The primary sites of these 63 SCCHN patients were as follows: buccal mucosa (21), tongue (15), gingiva (2), lip (1), larynx (6), hypopharynx (5), tonsil (5), palate (4), oropharynx (3), and middle ear (1).
●
C.-W. TWU et al.
131
Table 1. Patient characteristics (n ⫽ 114) Characteristics Age (years) Range, 24–74 Median, 46 Mean, 46 Sex Male Female Karnofsky scale ⱖ90% ⬍90% Pathology (WHO classification) Type I Type II Type III T stage (1997 AJCC) T1 T2 T3 T4 N stage (1997 AJCC) N0 N1 N2 N3 Overall stage II III IV Failure site(s) Nasopharynx Neck Distant Nasopharynx ⫹ distant Neck ⫹ distant
No. of cases
%
87 27
76.3 23.7
29 85
25.4 74.6
4 90 20
3.5 78.9 17.6
7 31 21 55
6.1 27.2 18.4 48.3
2 3 70 39
1.8 2.6 61.4 34.2
1 30 83
0.9 26.3 72.8
5 1 23 3 1
4.4 0.9 20.2 2.6 0.9
Abbreviations: WHO ⫽ World Health Organization; AJCC ⫽ American Joint Committee on Cancer.
The Institutional Review Board of our hospital approved this study, and informed consent was obtained from each study subject.
Immunofluorescence assay for serum VCA antibodies titers Immunofluoresence tests of IgA and IgG antibodies against EBV VCA were performed routinely using commercially available antigenic slides. Titers were considered positive if VCA-IgA ⱖ1:10 and VCA-IgG ⱖ1:40.
Real-time quantitative polymerase chain reaction for plasma EBV DNA measurement Plasma DNA was extracted using a QIAamp DNA Blood MiniKit (Qiagen, Hilden, Germany). EBV DNA concentrations of plasma were measured using a RTQ-PCR assay toward the BamHI-W region of the EBV genome. The sequences of forward and reverse primers were 5=-CCCAACACTCCACCACACC-3= and 5=-TCTTAGGAGCTGTCCGAGGG-3=. A dual fluorescencelabeled oligomer, 5=-(FAM)CACACACTACACACACCCACCCGTCTC(TAMRA)-3=, served as a probe. The principles of the RTQ-PCR and detailed reaction setup procedures were as described previously (19, 21).
132
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 67, Number 1, 2007
Table 2. Pretreatment viral load and antibody titers versus relapse or not EBV DNA Median Relapse No relapse Wilcoxon rank-sum test
4669 681
Interquartile range 779–13990 207–3225 p ⫽ 0.0006
VCA-IgA Median 1:40 1:40
VCA-IgG
Interquartile range 1:10–1:80 1:10–1:80 p ⫽ 0.9669
Median 1:2560 1:2560
Interquartile range 1:1280–1:5120 1:1280–1:5120 p ⫽ 0.6125
Abbreviations: EBV ⫽ Epstein-Barr virus; VCA ⫽ viral capsid antigen; Ig ⫽ immunoglobulin.
Statistical analysis The relapse-free survival was calculated from the first day of chemoradiotherapy until the date of disease relapse or until the date of the last follow-up visit. The overall survival was calculated from the first day of chemoradiotherapy until the date of death or until the date of the last follow-up visit. Life-table estimation was done according to the method of Kaplan-Meier. Univariate comparisons of survival curves were performed by use of the log–rank test. All statistical tests were two-sided, and p values of less than 0.05 were considered statistically significant. The multivariate Cox proportional hazards model was used to estimate the hazard ratios and 95% confidence intervals. Variables in the model included age, sex, performance status, pathologic type, response to neoadjuvant chemotherapy, T-stage, N-stage, overall stage, pretreatment and posttreatment plasma EBV DNA levels, pretreatment and posttreatment serum VCA-IgA titers, and pretreatment and posttreatment serum VCA-IgG titers. The relationship between the
plasma EBV DNA level or serum antibodies titers and the relapse rate was evaluated by using the chi-square test. Analyses were performed by use of SAS (Version 8.0; SAS Institute, Inc., Cary, NC).
RESULTS Sensitivity and specificity One hundred six of 114 NPC patients had detectable plasma EBV DNA by RTQ-PCR before treatment, whereas no detectable EBV DNA was found in 63 SCCHN patients and 50 healthy people. The sensitivity and specificity by the RTQ-PCR assay are 93.0% and 100%, respectively. The VCA-IgA antibody titers were positive in 82.5% (94/114) NPC patients, 46.0% (29/63) SCCHN, and 54.5% (27/50) healthy people. The VCA-IgG antibody titers were 100%
Fig. 1. Overall survival (A, C) and relapse-free survival (B, D) by plasma Epstein-Barr virus (EBV) deoxyribonucleic acid levels.
EBV antibody and DNA in nasopharyngeal carcinoma
positive for all tested samples, including 114 NPC patients, 65 SCCHN patients, and 50 healthy people. Thus, plasma EBV DNA proved to be superior to serum VCA antibodies in sensitivity and specificity. Effect of pretreatment EBV DNA levels or VCA antibody titers on subsequent relapse To study the prognostic impacts of the plasma EBV DNA level at initial presentation on subsequent relapses, we compared the pretreatment plasma EBV DNA concentrations between patients with and without tumor relapses. The median plasma EBV DNA levels in patients with (33 cases) and without (81 cases) relapse were 4669 copies/mL (interquartile range, 779 to 13,990) and 681 copies/mL (interquartile range, 207 to 3,225), respectively. The difference between these two groups was statistically significant ( p ⫽ 0.0006). There were no significant differences for pretreatment VCA-IgA ( p ⫽ 0.9669) or VCA-IgG ( p ⫽ 0.6125) titers in patients with or without relapse (Table 2). Relapse rates according to EBV DNA level or EBV antibody titers Twenty-two of 51 (43.1%) patients with pretreatment plasma EBV DNA ⱖ1500 copies/mL later developed tumor relapse, whereas only 17.5% (11/63) patients with pretreatment plasma EBV DNA ⬍1,500 copies/mL experienced
●
C.-W. TWU et al.
133
tumor relapse ( p ⫽ 0.0027). The pretreatment serum VCAIgA and VCA-IgG titers had limited value in predicting subsequent tumor relapse. The relapse rates between patients with VCA-IgA ⱖ1:80 and those with VCA-IgA ⬍ 1:80 were 27.0% (10/37) and 29.9% (23/77), respectively (p ⫽ 0.7540). The corresponding relapse rates for VCAIgG ⱖ and ⬍1:2560 were 27.8% (22/79) and 31.8% (11/ 35), respectively ( p ⫽ 0.6974). The presence or absence of plasma EBV DNA 1 week after chemoradiotherapy showed significant association to subsequent tumor relapse. The relapse rates between patients with plasma EBV DNA ⬎ and ⫽ 0 after chemoradiotherapy were 84.6% (11/13) and 21.8% (22/101), respectively ( p ⫽ 0.0002). However, the posttreatment serum VCA-IgA and VCA-IgG titers did not show any significant impact on subsequent tumor relapse. The relapse rates between patients with posttreatment VCA-IgA ⱖ1:80 and those with VCA-IgA ⬍1:80 were 23.9% (11/46) and 32.4% (22/68), respectively ( p ⫽ 0.3297). The corresponding relapse rates for VCA-IgG ⱖ and ⬍1:2560 were 25.4% (17/67) and 34.0% (16/47), respectively ( p ⫽ 0.3151). Correlation of the plasma EBV DNA levels and sites of first failure The pretreatment plasma EBV DNA levels in patients with distant failure (median, 5085; interquartile range,
Fig. 2. Overall survival (A, C) and relapse-free survival (B, D) by serum viral capsid antigen–immunoglobulin A (VCA-IgA) titers.
134
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 67, Number 1, 2007
Fig. 3. Overall survival (A, C) and relapse-free survival (B, D) by serum viral capsid antigen–immunoglobulin G (VCA-IgG) titers.
1,126 to 26,381 copies/mL) were higher than those with locoregional failure (median, 603; interquartile range, 113 to 10,799 copies/mL). The difference reached statistically significant ( p ⫽ 0.0443) levels. Thus, the patients with higher EBV DNA levels have a higher chance of developing microscopic metastases at presentation. Regardless of local therapy with chemoradiotherapy, those patients relapsed more frequently at distant than at local or regional sites. Survival analysis Kaplan-Meier estimates showed that both pretreatment EBV DNA levels and the presence/absence of EBV DNA in plasma samples after treatment significantly affected overall survival and relapse-free survival rates. The overall survival at 4 years was 93.1% in patients with pretreatment plasma EBV DNA ⬍1,500 copies/mL, and 60.3% in those with pretreatment plasma EBV DNA ⱖ1500 copies/mL ( p ⬍ 0.0001, Fig. 1A). The 4-year relapse-free survival rates of the patients with pretreatment plasma EBV DNA ⬍ and ⱖ1,500 copies/mL were 77.9% and 54.4%, respectively (p ⫽ 0.0009, Fig. 1B). The 4-year overall survival rates between patients with persistently detectable plasma EBV DNA and those with undetectable EBV DNA after chemoradiotherapy were 30.8% and 84.6%, respectively ( p ⬍ 0.0001, Fig. 1C). The relapse-free survival rate at 4 years
was significantly worse in the patients with persistently detectable plasma EBV DNA than in those with undetectable EBV DNA after radiotherapy (15.4% vs. 74.0%, p ⬍ 0.0001, Fig. 1D). Figures 2 and 3 illustrate detailed survival curves according to antibody titers of VCA-IgA and VCAIgG, measured at either pretreatment or posttreatment. The curves clearly showed no significant relationship between antibody titers and survival rates. Cox multivariate analysis When EBV data (serum EBV antibody titers and plasma EBV DNA levels) along with clinical parameters (including age, sex, performance status, pathologic type, response to neoadjuvant chemotherapy, T-stage, N-stage, and overall stage) were entered for Cox multivariate analysis, both pretreatment and posttreatment plasma EBV DNA levels were shown to be the most important prognostic factors for both overall survival (pretreatment, p ⫽ 0.001; hazard ratio ⫽ 9.16; 95% confidence interval [CI], 2.49 –33.64; posttreatment, p ⬍ 0.001; hazard ratio ⫽ 13.3; 95% CI, 3.92– 45.11) and relapse-free survival (pretreatment, p ⫽ 0.021; hazard ratio ⫽ 2.68; 95% CI, 1.16 – 6.18; posttreatment, p ⬍ 0.001; hazard ratio ⫽ 15.7; 95% CI, 6.02– 40.79) after adjustment for other variables (Table 3). The pretreatment and posttreatment serum VCA-IgA and VCA-IgG titers had no appreciable relationship to either overall or relapse-free
EBV antibody and DNA in nasopharyngeal carcinoma
●
C.-W. TWU et al.
135
Table 3. Summary of survival analyses by multivariate Cox proportional hazards model Overall survival Variables Age, years ⱕ45 vs. ⬎45 Sex Female vs. male Karnofsky performance status ⬍90% vs. ⱖ90% Pathology (WHO type) I vs. II/III Response to NeoCT CR vs. PR⫹SD T stage T3–4 vs. T1–2 N stage N3 vs. N0–2 Overall stage IV vs. II⫹III Pretreatment plasma EBV DNA ⱖ1,500 vs. ⬍1,500 Posttreatment plasma EBV DNA ⫹ vs. ⫺ Pretreatment serum VCA-IgA ⱖ1:80 vs. ⬍1:80 Posttreatment serum VCA-IgA ⱖ1:80 vs. ⬍1:80 Pretreatment serum VCA-IgG ⱖ1:2560 vs. ⬍1:2560 Posttreatment serum VCA-IgG ⱖ1:2560 vs. ⬍1:2560
p value
HR (95% CI)
Relapse-free survival p value
HR (95% CI)
0.401
0.66 (0.26–1.73)
0.235
0.60 (0.26–1.39)
0.412
1.64 (0.50–5.34)
0.979
1.01 (0.42–2.47)
0.027
4.99 (1.20–20.74)
0.999
1.00 (0.32–3.15)
0.794
1.37 (0.13–14.27)
0.026
5.32 (1.22–23.26)
0.968
1.02 (0.41–2.57)
0.712
1.16 (0.53–2.54)
0.411
1.63 (0.51–5.18)
0.658
1.26 (0.45–3.54)
0.012
5.32 (1.44–19.68)
0.082
2.46 (0.89–6.79)
0.299
0.41 (0.08–2.20)
0.332
2.17 (0.45–10.35)
0.001
9.16 (2.49–33.64)
0.021
2.68 (1.16–6.18)
⬍0.001
13.3 (3.92–45.11)
⬍0.001
15.7 (6.02–40.79)
0.533
0.70 (0.23–2.13)
0.568
0.76 (0.30–1.93)
0.319
0.60 (0.22–1.63)
0.365
0.67 (0.28–1.60)
0.078
0.39 (0.13–1.12)
0.834
0.91 (0.37–2.21)
0.800
0.88 (0.31–2.45)
0.777
1.14 (0.46–2.84)
Abbreviations: HR ⫽ hazard ratio; CI ⫽ confidence interval; WHO ⫽ World Health Organization; NeoCT ⫽ neoadjuvant chemotherapy; PR ⫽ partial response; CR ⫽ complete response; SD ⫽ stable disease; EBV ⫽ Epstein-Barr virus; VCA ⫽ viral capsid antigen; Ig ⫽ immunoglobulin.
survival rates. Table 3 also demonstrated that plasma EBV DNA level had greater effects than any other clinical factors on survivals. DISCUSSION Previously, NPC had no reliable tumor markers. Because NPC has a close association with the EBV, various antiEBV antibodies have been studied as surrogate markers of NPC since 1970 (9 –13, 16 –18). However, most studies used the same design: the collection of blood samples from NPC patients with active disease (previously untreated or relapsed cases) in addition to various controls (healthy volunteer or non-NPC tumors) and then comparing the sensitivity and specificity of different assay systems targeting various EBV antigens (9 –11, 18). All studies demonstrated higher positive detection rates of various anti-EBV antibodies for patients with active disease than those of control subjects. However, these data prove to be of little help in the diagnosis of newly found NPC or relapsed cases. The definite diagnosis of a newly found NPC or a relapsed case depends on suspected symptoms/signs, nasopharyngoscopic and imaging findings, and confirmatory biopsy results, irrespective of the titer of various antibodies. Further-
more, the antibody titer remains persistently high in most patients in remission after treatment (22, 23). There is no reliable cutoff value in the differential diagnosis of recurrence or remission. A clinically useful tumor marker should not only have higher sensitivity and specificity in the diagnosis setting but also have significant impact on the prognosis prediction. The latter is more important than sensitivity and specificity calculations alone in clinical practice. Only few studies have previously addressed the prognostic effects of various antiEBV antibodies (24 –27). By the antibody-dependent cellular cytotoxicity assay, which measures antibodies to an EBV-induced membrane antigen component, Neel et al. demonstrated that patients with low antibody-dependent cellular cytotoxicity titers at diagnosis had significantly worse progression-free survival ( p ⫽ 0.0001) and overall survival ( p ⫽ 0.0008) for World Health Organization types II and III NPC patients (24). Conventional antibody (VCAIgG, early antigen (EA)-IgA, and EA-IgG) assays before treatment could not reflect the prognosis (24). After longterm follow-up, they also found that the posttreatment sequential measurements did not accurately predict the outcome (25). A multicenter follow-up study on 319 NPC patients showed no prognostic values of initial serology,
136
I. J. Radiation Oncology
●
Biology
●
Physics
including IgG and IgA antibodies to VCA, EA or EBV nuclear antigen (EBNA) (26). However, increasing titers of EA-IgA and EA-IgG 1 year after completion of radiotherapy became highly significant for prediction of relapse, regardless of the initial titers (26). Yip et al. investigated the prognostic role of the IgG antibody against EBV BZLF-1 replication activators and found no significant difference in patients’ survival between the low and high titers for sera taken before treatment ( p ⫽ 0.331) (27). In contrast, for sera taken 10 months after radiotherapy, the high titer group predicted significantly worse overall survival rates ( p ⫽ 0.0019) (27). Based on the foregoing discussion, the definite role of anti-EBV antibodies on prognosis prediction is still controversial and the precise point of blood sampling has not been well established. On the contrary, circulating EBV DNA loads measured before or after treatment are consistently correlated with the survival and prognosis of NPC patients (19 –21). This may relate to the essential difference between a marker directly released from tumor cells (EBV DNA) and a marker that essentially is a host response to viral tumor antigens. The latter marker inevitably lags behind with regard to kinetics upon treatment. Also, the dynamic range of antibody responses is fairly limited in general, with only weak correlations observed between the extent of exposure and the level (titer) of the response. On the other hand, unlike conventional protein markers such as carcinoembryonic antigen or prostate-specific antigen, released mostly from tumor cells, EBV can be found in patients with EBV-related diseases or even in normal populations in high concentration, especially after an infection with EBV. These situations should be considered in the clinical interpretation of EBV DNA as a tumor marker for NPC.
Volume 67, Number 1, 2007
To the best of our knowledge, there are only five papers comparing the EBV-specific antibody-based and DNAbased assays in NPC (22, 23, 28 –30). They have usually focused on sensitivity, specificity, or correlation with staging. The major findings of these studies are that the EBV DNA assay is superior to the anti-EBV antibody assay in both sensitivity/specificity calculation (23, 28 –30) and differential diagnosis of active disease or remission state (22, 23). In addition, plasma EBV DNA levels have better correlation with clinical staging than VCA-IgA titers (22). So far, no other report has been shown to compare the prognostic values between anti-EBV antibody and EBV DNA assays. In this study, except for routine sensitivity/specificity calculations, we first report a detailed comparison of the prognostic value between serum anti-VCA antibody titers and plasma EBV DNA copy numbers. Our results clearly demonstrate that plasma EBV DNA levels are superior to serum VCA-IgA or VCA-IgG titers for all analyzed items, including parameters of sensitivity, specificity, prediction of subsequent relapse, survival curves, and prognostic factors analyses. Results of this study will have a significant impact in the management of NPC. Patients with high risk of relapse (persistently detectable EBV DNA after treatment or higher EBV DNA level before treatment) should be further studied for a more aggressive therapy, and routine adjuvant therapy should be prohibited for patients with low risk of relapse to avoid unnecessary morbidity and waste of medical resources. Of note, one limitation of our study is that only antibody against the VCA was compared. Comparisons of the prognostic impact between the EBV DNA assay and many different antibody assays against other antigens are still lacking and need to be investigated further.
REFERENCES 1. Al-Sarraf M, LeBlanc M, Giri PGS, et al. Chemoradiotherapy versus radiotherapy in patients with advanced nasopharyngeal cancer: Phase III randomized Intergroup study 0099. J Clin Oncol 1998;16:1310 –1317. 2. Lin JC, Jan JS, Hsu CY, et al. Phase III study of concurrent chemoradiotherapy versus radiotherapy alone for advanced nasopharyngeal carcinoma: Positive effect on overall and progression-free survival. J Clin Oncol 2003;21:631– 637. 3. Lee AWM, Lau WH, Tung SY, et al. Preliminary results of a randomized study on therapeutic gain by concurrent chemotherapy for regionally-advanced nasopharyngeal carcinoma: NPC-9901 trial by the Hong Kong Nasopharyngeal Cancer Study Group. J Clin Oncol 2005;23:6966 – 6975. 4. Wee J, Tan EH, Tai BC, et al. Randomized trial of radiotherapy versus concurrent chemoradiotherapy followed by adjuvant chemotherapy in patients with American Joint Committee on Cancer/International Union Against Cancer stage III and IV nasopharyngeal cancer of the endemic variety. J Clin Oncol 2005;23:6730 – 6738. 5. Chang YS, Tyan YS, Liu ST, et al. Detection of Epstein-Barr virus DNA sequences in nasopharyngeal carcinoma cells by enzymatic DNA amplification. J Clin Microbiol 1990;28:2398 – 2402. 6. Chen CL, Wen WN, Chen JY, et al. Detection of Epstein-Barr
7.
8. 9. 10. 11. 12. 13. 14.
virus genome in nasopharyngeal carcinoma by in situ DNA hybridization. Intervirology 1993;36:91–98. Tsai ST, Jin YT, Su IJ. Expression of EBER1 in primary and metastatic nasopharyngeal carcinoma tissues using in situ hybridization: A correlation with WHO histologic subtypes. Cancer 1996;77:231–236. Raab-Traub N, Flynn K. The structure of the termini of the Epstein-Barr virus as a marker of clonal cellular proliferation. Cell 1986;47:883– 889. Henle W, Henle G, Ho HC, et al. Antibodies to Epstein-Barr virus in nasopharyngeal carcinoma, other head and neck neoplasms and control groups. J Natl Cancer Inst 1970;44:225–231. Henle G, Henle W. Epstein-Barr virus-specific IgA serum antibodies as an outstanding feature of nasopharyngeal carcinoma. Int J Cancer 1976;17:1–7. Ho HC, Ng MH, Kwan HC, et al. Epstein-Barr virus-specific IgA and IgG serum antibodies in nasopharyngeal carcinoma. Br J Cancer 1976;34:655– 660. Chien YC, Chen JY, Liu MY, et al. Serologic markers of Epstein-Barr virus infection and nasopharyngeal carcinoma in Taiwanese men. N Engl J Med 2001;345:1877–1882. Zeng Y, Zhong JM, Li LY, et al. Follow-up studies on Epstein-Barr virus IgA/VCA antibody-positive persons in Zangwu County, China. Intervirology 1983;20:190 –194. Pathmanathan R, Prasad U, Sadler R, et al. Clonal prolifera-
EBV antibody and DNA in nasopharyngeal carcinoma
15. 16.
17.
18.
19. 20.
21. 22.
tions of cells infected with Epstein-Barr virus in preinvasive lesions related to nasopharyngeal carcinoma. N Engl J Med 1995;333:693– 698. Yeung WM, Zong YS, Chiu CT, et al. Epstein-Barr virus carriage by nasopharyngeal carcinoma in situ. Int J Cancer 1993;53:746 –750. Liu MY, Chang YL, Ma J, et al. Evaluation of multiple antibodies to Epstein-Barr virus as markers for detecting patients with nasopharyngeal carcinoma. J Med Virol 1997;52: 262–269. Uen WC, Luka J, Pearson GR. Development of an enzymelinked immunosorbent assay (ELISA) for detecting IgA antibodies to the Epstein-Barr virus. Int J Cancer 1988;41:479 – 482. Neel HB 3rd, Pearson GR, Taylor WF. Antibodies to EpsteinBarr virus in patients with nasopharyngeal carcinoma and in comparison groups. Ann Otol Rhinol Laryngol 1984;93:477– 482. Lo YMD, Chan LYS, Lo KW, et al. Quantitative analysis of cell-free Epstein-Barr virus DNA in plasma of patients with nasopharyngeal carcinoma. Cancer Res 1999;59:1188 –1191. Chan ATC, Lo YMD, Zee B, et al. Plasma Epstein-Barr virus DNA and residual disease after radiotherapy for undifferentiated nasopharyngeal carcinoma. J Natl Cancer Inst 2002;94: 1614 –1619. Lin JC, Wang WY, Chen KY, et al. Quantification of plasma Epstein-Barr virus DNA in patients with advanced nasopharyngeal carcinoma. N Engl J Med 2004;350:2461–2470. Shao JY, Li YH, Gao HY, et al. Comparison of plasma Epstein-Barr virus (EBV) DNA levels and serum EBV immunoglobulin A/viral capsid antigen antibody titers in patients with nasopharyngeal carcinoma. Cancer 2004;100:1162–1170.
●
C.-W. TWU et al.
137
23. Fan H, Nicholls J, Chua D, et al. Laboratory markers of tumor burden in nasopharyngeal carcinoma: A comparison of viral load and serologic tests for Epstein-Barr virus. Int J Cancer 2004;112:1036 –1041. 24. Neel HB 3rd, Pearson GR, Taylor WF. Antibody-dependent cellular cytotoxicity: Relation to stage and disease course in North American patients with nasopharyngeal carcinoma. Arch Otolaryngol 1984;110:742–747. 25. Neel HB 3rd, Taylor WF. Epstein-Barr virus-related antibody. Change in titers after therapy for nasopharyngeal carcinoma. Arch Otolaryngol Head Neck Surg 1990;116:1287–1290. 26. de Vathaire F, Sancho-Garnier H, de-The H, et al. Prognostic value of EBV markers in the clinical management of nasopharyngeal carcinoma (NPC): A multicenter follow-up study. Int J Cancer 1988;42:176 –181. 27. Yip TTC, Ngan RKC, Lau WH, et al. A possible prognostic role of immunoglobulin-G antibody against recombinant Epstein-Barr virus BZLF-1 transactivator protein ZEBRA in patients with nasopharyngeal carcinoma. Cancer 1994;74: 2414 –2424. 28. Chan KH, Gu YL, Ng F, et al. EBV specific antibody-based and DNA-based assays in serologic diagnosis of nasopharyngeal carcinoma. Int J Cancer 2003;105:706 –709. 29. Leung SF, Tam JS, Chan ATC, et al. Improved accuracy of detection of nasopharyngeal carcinoma by combined application of circulating Epstein-Barr virus DNA and anti-EpsteinBarr viral capsid antigen IgA antibody. Clin Chem 2004;50: 339 –345. 30. Kondo S, Horikawa T, Takeshita H, et al. Diagnostic value of serum EBV-DNA quantification and antibody to viral capsid antigen in nasopharyngeal carcinoma patients. Cancer Sci 2004;95:508 –513.