Head and Neck Cancer
Analysis of Plasma Epstein-Barr Virus DNA in Nasopharyngeal Cancer After Chemoradiation to Identify High-Risk Patients for Adjuvant Chemotherapy: A Randomized Controlled Trial
The contribution of adjuvant chemotherapy after chemoradiation therapy (CRT) in nasopharyngeal cancer (NPC) remains controversial. Plasma Epstein-Barr virus (EBV) DNA is a potential biomarker of subclinical residual disease in NPC. In this prospective, multicenter, randomized controlled trial, we used plasma EBV DNA to identify patients with NPC at a higher risk of relapse for adjuvant chemotherapy.
Eligible patients with histologically confirmed NPC of Union for International Cancer Control stage IIB to IVB, adequate organ function, and no locoregional disease or distant metastasis were screened by plasma EBV DNA at 6 to 8 weeks after radiotherapy (RT). Patients with undetectable plasma EBV DNA underwent standard surveillance. Patients with detectable plasma EBV DNA were randomly assigned to either adjuvant chemotherapy with cisplatin and gemcitabine for six cycles (arm 1) or observation (arm 2). Patients were stratified for primary treatment (RT v CRT) and stage (II/III v IV). The primary end point was relapse-free survival (RFS).
Seven hundred eighty-nine patients underwent EBV DNA screening. Plasma EBV DNA was undetectable in 573 (72.6%) and detectable in 216 (27.4%); 104 (13.2%) with detectable EBV DNA were randomly assigned to arms 1 (n = 52) and 2 (n = 52). After a median follow-up of 6.6 years, no significant difference was found in 5-year RFS rate between arms 1 and 2 (49.3% v 54.7%; P = .75; hazard ratio for relapse or death, 1.09; 95% CI, 0.63 to 1.89). The level of post-RT plasma EBV DNA correlated significantly with the hazards of locoregional failure, distant metastasis, and death.
Nasopharyngeal cancer (NPC) is endemic in southern China, including Hong Kong, where nearly all patients harbor the Epstein-Barr virus (EBV) in their tumor cells.1,2 Most patients with localized NPC will achieve complete clinical remission after curative-intent radiotherapy (RT) or chemoradiation therapy (CRT). Despite a high rate of initial control, disease subsequently will relapse in 30% to 40% of patients either at the local site or, more commonly, with distant metastasis, which prevents long-term cure. In the study conducted by the Prince of Wales and Queen Elizabeth Hospitals in Hong Kong, concurrent cisplatin and RT improved the 5-year overall survival (OS) rate from 58.6% to 70.3%, which confirms this treatment as a standard option in advanced-stage NPC.3 Although the addition of adjuvant plus concurrent chemotherapy to RT conferred superior survival over RT alone in the Intergroup 0099 study,4 the relative contribution of the concurrent and adjuvant components of chemotherapy has been inadequately assessed,5 and the extent to which adjuvant chemotherapy can affect OS in NPC remains uncertain.
Adjuvant chemotherapy after RT alone did not improve OS in two previous phase III trials.6,7 In the most recent phase III trial, 508 patients with advanced NPC were randomly assigned to adjuvant chemotherapy (cisplatin and fluorouracil) or observation after CRT, with no significant improvement in 5-year failure-free survival rate with adjuvant chemotherapy versus CRT alone (75% v 71%; hazard ratio [HR], 0.88; 95% CI, 0.64 to 1.22).8,9 In the updated Meta-Analysis of Chemotherapy in Nasopharynx Cancer of 4,806 patients, the addition of concurrent chemotherapy to RT significantly improved OS in locally advanced NPC.10 However, the interaction between treatment effect on OS and the timing of chemotherapy was significant in favor of both concurrent CRT with adjuvant chemotherapy (HR, 0.65; 95% CI, 0.56 to 0.76) and CRT without adjuvant chemotherapy (HR, 0.80; 95% CI, 0.70 to 0.93) but not adjuvant chemotherapy or induction chemotherapy alone. The observation of a small difference in the HRs in favor of CRT followed by adjuvant chemotherapy when indirectly compared with CRT without adjuvant chemotherapy in a network meta-analysis suggested that administration of more chemotherapy to patients with locally advanced NPC may achieve a better result.11 Not all patients likely will benefit from adjuvant chemotherapy after CRT, so the challenge is in patient selection to maximize the magnitude of benefit.
EBV DNA detected in cell-free plasma has higher specificity and sensitivity for EBV-positive disease compared with EBV DNA in peripheral blood mononuclear cells.12 We and others have shown that plasma EBV DNA level is an independent prognostic biomarker for NPC.13-15 Post-RT plasma EBV DNA has an even stronger correlation with prognosis and has been used to monitor recurrence during post-treatment surveillance.16,17 Elevated plasma EBV DNA level has been shown to predate clinical recurrence by 3 to 7 months.18 A detectable or high level of post-RT plasma EBV DNA can predict a poor progression-free survival or OS and represents a biomarker of subclinical residual disease.16
We hypothesized that by selecting high-risk patients with NPC with a significant likelihood of harboring occult distant metastasis (defined by detectable post-RT plasma EBV DNA) and using the more active chemotherapy regimen of cisplatin and gemcitabine with superior efficacy for metastatic disease19 administered earlier in the adjuvant setting, we can reduce distant metastasis and significantly improve relapse-free survival (RFS) by eradicating subclinical micrometastasis. Correlation of post-RT plasma EBV DNA with clinical outcome is a secondary end point of the study.
The Hong Kong NPC Study Group 0502 trial is a prospective, open-label, multicenter, randomized controlled trial conducted at all six oncology centers in Hong Kong. Eligible patients were age ≥ 18 years with a histologic diagnosis of locoregionally advanced NPC of Union for International Cancer Control (UICC; 6th edition) stage IIB, III, IVA, or IVB; no clinical evidence of persistent locoregional disease or distant metastasis after completion of primary RT or CRT; Eastern Collaborative Oncology Group performance status of 0 to 1; and adequate bone marrow and organ function. Exclusion criteria were second primary malignancy, > 12 weeks after completion of primary RT, and peripheral or ototoxicity greater than grade 2 (Appendix, online only).
Eligible patients were consented for plasma EBV DNA screening at 6 to 8 weeks post-RT. Patients with undetectable plasma EBV DNA underwent standard surveillance. Patients with detectable plasma EBV DNA were consented for restaging work-up and random assignment to adjuvant chemotherapy versus observation.
The study was conducted in accordance with the principles of the Declaration of Helsinki and International Conference on Harmonization of Good Clinical Practice and registered as a clinical trial. The study protocol (Data Supplement) was approved by the institutional review board of each study center. All patients provided written informed consent before any study procedure.
Patients who met all eligibility criteria after restaging were randomly assigned by telephone from the central office of the Comprehensive Cancer Trials Unit at The Chinese University of Hong Kong. They were randomly assigned in a 1:1 ratio, with stratification for the type of primary treatment (RT v CRT) and cancer stage (II/III v IV) to either adjuvant chemotherapy (arm 1) or clinical observation (arm 2). The masking of EBV DNA and positron emission tomography/computed tomography (PET/CT) results are described in the Appendix.
Ten milliliters of venous blood were collected into EDTA tubes; processed within 6 hours to obtain plasma; and transported to the Department of Chemical Pathology, The Chinese University of Hong Kong. Plasma EBV DNA was analyzed by real-time polymerase chain reaction assay that targeted the BamHI-W fragment of the EBV genome, as described previously.13,20,21 On the basis of the analysis of known EBV DNA standards, our assay was able to detect consistently 20 EBV genomes per milliliter of plasma.21 All samples were analyzed in duplicate. Amplification signals in any replicate were registered as a positive result, regardless of the level.
Patients randomly assigned to adjuvant chemotherapy would receive cisplatin 40 mg/m2 intravenously and gemcitabine 1,000 mg/m2 intravenously, with both administered on day 1 and day 8 every 3 weeks for a total of six cycles. Dose modification criteria were predefined in the study protocol (Appendix). Adverse events were graded according to Common Terminology Criteria for Adverse Events (version 3). Late toxicity of skin, subcutaneous tissue, and salivary glands were assessed by the Radiation Therapy Oncology Group and European Organisation for Research and Treatment of Cancer Late Radiation Morbidity Scoring Criteria.22
The primary end point was RFS, which was defined as the time from random assignment to the time of first recurrence of NPC (locoregional failure or distant metastasis) or death as a result of any cause, whichever occurred first, or censored at last follow-up. Secondary end points were OS, locoregional failure-free survival (LRFS), distant metastasis–free survival (DMFS), toxicity of adjuvant chemotherapy, and correlation of plasma EBV DNA with clinical outcome (Appendix).
Historically, RFS at 18 months is approximately 30% in patients with detectable post-RT plasma EBV DNA16,17; therefore, we estimated a magnitude of benefit from adjuvant chemotherapy to be 60% for the study treatment of major clinical value. With an estimated HR of 2 (or 0.5 if observation arm as reference), approximately 100 patients were required with a power of 0.8 and α of .05. At the closure of study accrual in July 2015 (when 104 patients were randomly assigned) and on the basis of the projected event rate of four per year, we planned to observe 48 events by the end of 2016, which can provide 80% power for a revised HR of 2.4 (or 0.42 if observation arm as reference). After a median follow-up of > 6 years, 98% of the events occurred in the first 3 years. The RFS curve was essentially flat after 3 years. The data and safety monitoring committee recommended to conduct the primary end point analysis. Kaplan-Meier method, log-rank test, and Cox proportional hazards regression modeling were used to analyze all time-to-event data (Appendix). We regarded P < .05 as significant. All statistical analyses were performed using SAS 9.3 software (SAS Institute, Cary, NC).
From September 2006 to July 2015, 789 patients with NPC who completed curative-intent RT or CRT consented for plasma EBV DNA screening (Fig 1); 573 patients (72.6%) had undetectable plasma EBV DNA, and 216 (27.4%) had detectable plasma EBV DNA (median, 61 copies/mL; interquartile range, 22.5 to 251 copies/mL). The baseline characteristics of the study population are listed in Table 1. Patients with detectable post-RT EBV DNA (cohort C in Table 1) had a higher T stage, N stage, and UICC overall stage than patients with undetectable plasma EBV DNA (cohort B in Table 1). Although RT technique was not a stratification factor in the randomization, intensity-modulated radiotherapy was adopted by all except one center during the study period. In one center where two-dimensional RT was still used in < 5% of patients before 2008, 11 patients recruited at that center received two-dimensional RT, including two who were subsequently randomly assigned (one to adjuvant chemotherapy and one to observation).
After work-up, 104 patients (13.2% of all screened) with detectable post-RT plasma EBV DNA were eligible and randomly assigned to adjuvant chemotherapy (n = 52) or observation (n = 52). The baseline clinical characteristics, including age, sex, Eastern Cooperative Oncology Group performance status, T stage, N stage, UICC overall stage, primary treatment modality, post- RT EBV DNA level, and baseline PET/CT positivity (blinded at random assignment), were evenly balanced in the two arms (cohort C1 v C2 in Table 1). The median follow-up was 6.6 years, and the relapse and death events in the study population are listed in Appendix Table A1 (online only). No significant difference was found in the 5-year RFS rate between adjuvant chemotherapy and observation (49.3% v 54.7%; HR for relapse or death, 1.09; 95% CI, 0.63 to 1.89; P = .75). Similarly, no significant difference was found in OS, LRFS, and DMFS between adjuvant chemotherapy and observation (Fig 2; Table 2). A sensitivity analysis that included patients who received at least one dose of adjuvant chemotherapy showed similar results (Appendix Table A2, online only). Forest plot analysis of RFS and OS (Figs 3A and 3B) in patient subgroups did not reveal heterogeneity in treatment effect except for female sex, which showed worse RFS in the adjuvant chemotherapy arm.
For patients randomly assigned to adjuvant chemotherapy, 84.6% started chemotherapy, and 50.0% completed six cycles (Appendix Table A3, online only). Three patients died within 90 days after the last dose of adjuvant chemotherapy. The cause of death was a result of a treatment-related complication in two (nasopharyngeal hemorrhage, hematemesis) and intercurrent illness in one. Other high-grade treatment-related toxicities were mainly hematologic (Appendix Table A4, online only). Late radiation toxicity was comparable in both arms, with the exception of subcutaneous toxicity, which was more common and severe in the adjuvant chemotherapy arm (Appendix Table A5, online only).
The absolute level of post-RT plasma EBV DNA correlated significantly with the risk of locoregional failure, distant metastasis, and death (Table 3; Fig 4). For each 10-log increase in plasma EBV DNA concentration, the HR was 1.96 (95% CI, 1.77 to 2.16) for locoregional failure, 2.14 (95% CI, 1.93 to 2.39) for distant metastasis, and 2.12 (95% CI, 1.89 to 2.37) for death. Post-RT plasma EBV DNA followed by UICC overall stage was the most significant prognostic factor in both univariable and multivariable analysis (Appendix Table A6, online only).
This study is the first biomarker-driven RCT in our knowledge to compare adjuvant chemotherapy versus observation in high-risk patients with NPC identified by detectable plasma EBV DNA after completion of RT and CRT. We used cisplatin and gemcitabine as the adjuvant chemotherapy, which has been shown to be more effective than cisplatin and fluorouracil in the metastatic setting.19 The current results show that adjuvant cisplatin and gemcitabine chemotherapy do not improve RFS or OS in this biomarker (plasma EBV DNA)–selected high-risk group of patients with NPC.
We observed several deaths without documented progression in the study arms (Appendix Table A1). To avoid the confounding effect of death as a competing event, we adopted a conservative composite end point definition in the analysis of RFS, LRFS, and DMFS by including either failure or death, whichever occurred first (Appendix). In a separate analysis with death censored, the results were similar (5-year RFS, 56.3% v 54.7%; P = .79; HR for relapse, 0.92; 95% CI, 0.51 to 1.66).
This study has several limitations. The compliance to adjuvant cisplatin and gemcitabine was suboptimal because only 84.6% of patients randomly assigned to adjuvant chemotherapy actually started treatment, and 65.4%, 57.7%, and 50.0% completed four, five, and six cycles of adjuvant cisplatin and gemcitabine, respectively (Appendix Table A3). The reasons for early discontinuation were patient refusal (73%) or treatment-related toxicity (27%). In the phase III trial that compared cisplatin and gemcitabine with cisplatin and fluorouracil in the first-line setting of recurrent or metastatic NPC, 82.9%, 66.9%, and 58% of patients were able to complete four, five, and six cycles of cisplatin and gemcitabine versus 75.7%, 56.4%, and 58.0% for cisplatin and fluorouracil, respectively.19 The lower compliance in the adjuvant setting than in the metastatic setting probably reflects a generally poor tolerance in the first 6 months after completion of RT and CRT. Another contributing factor could have been lower patient motivation to finish a regimen with significant toxicity but uncertain benefit in the adjuvant setting. Nevertheless, in a sensitivity analysis of survival outcome according to compliance of adjuvant chemotherapy, there was no suggestion of a different outcome in patients who received more than three cycles of adjuvant chemotherapy (Appendix Table A7, online only).
In the current study, the randomization was performed at a median of 78 days after completion of RT or CRT to accommodate the time to complete all prerandomization evaluations to assess patient disease status and eligibility criteria. As a result, of 216 patients with detectable post-RT plasma EBV DNA, 112 were excluded before random assignment (Fig 1). The reasons for exclusion were patient refusal (n = 39), presence of residual disease (n = 36) or distant metastasis (n = 25), inadequate renal (n = 6) or bone marrow function (n = 2), preexisting neuropathy (n = 2), poor general condition (n = 1), and diagnosis of second primary cancer (n = 1). We believe that the stringent prerandomization evaluation in the current study actually selected a more homogenous study population for adjuvant chemotherapy. Patients with clinically or radiologic-persistent locoregional disease, distant metastases, or inadequate organ function after completion of RT/CRT were excluded before random assignment. Therefore, the patients eventually randomly assigned had a much lower risk compared with the historical data.
We postulated that a time frame of 12 weeks after completing CRT would be necessary for patient recovery from treatment-related adverse effects but may be too late for adjuvant chemotherapy to be effective in minimal residual disease. In this study, the median time from the day of last RT to the day of first dose of adjuvant chemotherapy was 91 days (interquartile range, 81 to 99 days; Appendix Table A3). In comparison, the median time from last day of RT to first day of adjuvant chemotherapy was 33 days (interquartile range, 28 to 39 days) in another phase III study of adjuvant chemotherapy in NPC.8 In a sensitivity analysis on survival outcome according to time to initiation of adjuvant chemotherapy (Appendix Table A8, online only), there was no indication for a better outcome if the adjuvant chemotherapy was started earlier (less than the median) rather than later (greater than the median).
Because the companion PET/CT study was not reported until the second PET/CT scan was completed at 6 months after randomization, the blinding of baseline PET/CT could introduce imbalance into the two study arms. However, the number of baseline PET/CT scans retrospectively reported as positive were well balanced in the two study arms (19.1% in the adjuvant chemotherapy arm v 20.4% in the observation arm; Table 1). In a sensitivity analysis that excluded patients with positive baseline PET/CT scans, the survival outcome remained similar to that in the primary analysis (Appendix Table A9, online only).
The clinical utility and current state of polymerase chain reaction–based EBV DNA testing in NPC with recommendation for assay harmonization and validation have been recently discussed in a workshop meeting.15,23 Currently, no consensus exists on the optimal cutoff for post-RT plasma EBV DNA levels for risk classification.15 Previous studies have suggested that the quantification of a low level of plasma EBV DNA close to the detection limit of the assay system may be less accurate in patients with NPC.24,25 In the current study, we report any detectable signal as a positive result to maximize the sensitivity for detecting residual disease.
In the current RCT, 85% of patients had received prior cisplatin. We hypothesized that in patients with NPC with detectable plasma EBV DNA after cisplatin and RT, the residual subclinical disease is more likely platinum resistant. Early administration of cisplatin-based chemotherapy in the adjuvant setting not only is unable to eradicate the platinum-resistant clone but also adds toxicity, which further compromises the patient’s ability to receive subsequent chemotherapy. Indeed, we observed that most patients in the adjuvant chemotherapy arm responded poorly to subsequent chemotherapy at relapse. Retrospective data have suggested that the increased risk of relapse and death in patients with elevated plasma EBV DNA after CRT could be reduced by metronomic adjuvant chemotherapy with oral tegafur-uracil for 12 months.26 Recently, the use of immune checkpoint inhibitors has been demonstrated to improve progression-free survival in patients with stage III non–small-cell lung cancer after chemoradiation.27 The NRG Oncology Cooperative Group initiated NRG-HN001 (ClinicalTrials.gov identifier: NCT02135042) that used post-RT plasma EBV DNA to divide patients with NPC into low-risk and high-risk groups, with the hypothesis that the low-risk group would not need adjuvant chemotherapy and that the high-risk group would benefit from more-aggressive non–cross-resistant paclitaxel and gemcitabine chemotherapy.23
Current clinical guidelines in NPC generally recommend that the treatment decision be based on the anatomic classification of UICC TNM staging.1 The current study provides strong validation that post-RT plasma EBV DNA level is the major determinant of RFS and OS in patients with NPC at the end of RT or CRT. We propose to include post-RT EBV DNA level as a nonanatomic determinant of risk classification and treatment selection for adjuvant therapy in NPC. The optimal management of patients with NPC with detectable post-RT plasma EBV DNA remains undetermined. Patients should be encouraged to participate in clinical trials that investigate non–platinum-based systemic therapy or immune checkpoint inhibitors.28,29 Until more evidence from RCTs becomes available, concurrent cisplatin and RT should remain the standard of care in endemic locally advanced NPC.
Supported by research funding from The Chinese University of Hong Kong and Hong Kong Hospital Authority. Gemcitabine was provided by Eli Lilly. The Partner State Key Laboratory of Oncology in South China at The Chinese University of Hong Kong was supported by funding from the Innovative Technology Commission. Any opinions, findings, conclusions, or recommendations expressed in this article (or by members of the project team) do not reflect the views of the Government of the Hong Kong Special Administrative Region, the Innovation and Technology Commission, or the Panel of Assessors for the Innovation and Technology Support Programme of the Innovation and Technology Fund.
Presented at the 2017 Annual Meeting of American Society of Clinical Oncology, Chicago, IL, June 2-6, 2017, and the European Society of Medical Oncology Asia 2017 Congress, Singapore, November 17-19, 2017.
Clinical trial information: NCT00370890.
Conception and design: Anthony T.C. Chan, Edwin P. Hui, Roger K.C. Ngan, Brigette B.Y. Ma, Benny C.Y. Zee
Financial support: Y.M. Dennis Lo
Administrative support: Anthony T.C. Chan, Y.M. Dennis Lo
Provision of study materials or patients: Anthony T.C. Chan, Edwin P. Hui, Roger K.C. Ngan, Stewart Y. Tung, Ashley C.K. Cheng, Wai T. Ng, Victor H.F. Lee, Brigette B.Y. Ma, Hoi C. Cheng, Frank C.S. Wong, Herbert H.F. Loong, Macy Tong, Darren M.C. Poon, Anil T. Ahuja, Ann D. King, Ki Wang, K.C. Allen Chan
Collection and assembly of data: Edwin P. Hui, Roger K.C. Ngan, Stewart Y. Tung, Ashley C.K. Cheng, Wai T. Ng, Victor H.F. Lee, Brigette B.Y. Ma, Hoi C. Cheng, Frank C.S. Wong, Herbert H.F. Loong, Macy Tong, Darren M.C. Poon, Anil T. Ahuja, Ann D. King, Ki Wang, Frankie Mo, K.C. Allen Chan, Y.M. Dennis Lo
Data analysis and interpretation: Anthony T.C. Chan, Edwin P. Hui, Brigette B.Y. Ma, Benny C.Y. Zee
Manuscript writing: All authors
Final approval of manuscript: All authors
Accountable for all aspects of the work: All authors
The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/site/ifc.
Honoraria: Celgene, Merck Sharp & Dohme
Consulting or Advisory Role: Celgene
Research Funding: Pfizer, Boehringer Ingelheim, Bristol-Myers Squibb, Merck Sharp & Dohme
Travel, Accommodations, Expenses: Celgene, Pfizer, Merck Sharp & Dohme, Roche, Bristol-Myers Squibb, Novartis
Honoraria: Merck Sharp & Dohme, Merck
Consulting or Advisory Role: Bristol-Myers Squibb
Travel, Accommodations, Expenses: Eisai, Merck, Novartis, Pfizer, Roche, Celgene
No relationship to disclose
Research Funding: Eli Lilly (Inst)
No relationship to disclose
Honoraria: Eli Lilly, Pfizer, Roche
Honoraria: Novartis, Merck Sharp & Dohme, Merck Serono, Roche
Consulting or Advisory Role: Novartis, Merck Serono, Merck Sharp & Dohme
Research Funding: Novartis
Travel, Accommodations, Expenses: Amgen
No relationship to disclose
No relationship to disclose
Consulting or Advisory Role: GlaxoSmithKline, Roche, Genentech, Novartis
Speakers’ Bureau: AbbVie, Ignyta, Novartis
Research Funding: MSD (Inst)
Travel, Accommodations, Expenses: Roche, MSD, Bayer AG, Pfizer
No relationship to disclose
No relationship to disclose
No relationship to disclose
No relationship to disclose
No relationship to disclose
No relationship to disclose
Stock or Other Ownership: Health View Bioanalytic
Research Funding: Boston Therapeutics (Inst)
Patents, Royalties, Other Intellectual Property: Health View Bioanalytic
Leadership: Cirina, Xcelom
Stock or Other Ownership: Cirina, Xcelom, Grail
Honoraria: Bio-Rad Laboratories
Consulting or Advisory Role: Grail
Patents, Royalties, Other Intellectual Property: Grail, Illumina, Sequenom
Travel, Accommodations, Expenses: Bio-Rad Laboratories
Leadership: Cirina, Xcelom
Stock or Other Ownership: Grail, DRA Company Limited, Xcelom
Consulting or Advisory Role: Grail, Decheng Capital
Research Funding: Grail (Inst), Cirina (Inst)
Patents, Royalties, Other Intellectual Property: Co-inventor on patents or patent applications in the area of molecular diagnostics using circulating nucleic acids
Travel, Accommodations, Expenses: Grail, Cirina, Berry Genomics, Sequenom
The Hong Kong NPC Study Group 0502 trial is a prospective, open-label, multicenter, randomized controlled trial conducted at all six oncology centers in Hong Kong. The primary objective is to study the benefit of adjuvant chemotherapy with gemcitabine and cisplatin in patients with nasopharyngeal cancer (NPC) at higher risk of relapse defined by a detectable plasma Epstein-Barr virus (EBV) DNA level after completion of curative-intent radiotherapy (RT) or chemoradiation therapy (CRT).
Eligible patients were age ≥ 18 years with a histologic diagnosis of locoregionally advanced NPC of Union for International Cancer Control (6th edition) stage IIB, III, IVA, or IVB; no clinical evidence of persistent locoregional disease after completion of primary treatment; Eastern Collaborative Oncology Group performance status of 0 or 1; adequate bone marrow and organ function (defined as leukocytes ≥ 3,000/μL, absolute neutrophil count ≥ 1,500/μL; platelets ≥ 100,000/μL, total bilirubin ≤ 1.5 times the institutional upper limit of normal; transaminase ≤ 2.5 times the upper limit of normal; and creatinine clearance ≥ 50 mL/min as estimated by the Cockcroft-Gault equation. Exclusion criteria were hypercalcemia, second primary malignancy (except for in situ carcinoma of the cervix or adequately treated basal cell carcinoma of the skin), > 12 weeks after completion of primary RT, receipt of prior adjuvant chemotherapy, peripheral or ototoxicity with a grade > 2, serious active infection, pregnant or lactating female patients, and patients with reproductive potential who were not implementing adequate contraceptive measures.
Eligible patients were consented for plasma EBV DNA screening at 6 to 8 weeks post-RT. Those with undetectable plasma EBV DNA were followed up by standard surveillance. Those with detectable plasma EBV DNA were consented for work-up and random assignment to adjuvant chemotherapy versus clinical observation and surveillance. The protocol allowed primary RT to be delivered with either two- or three-dimensional techniques of ≥ 66 Gy in 33 fractions over 6.5 weeks (Lee et al: Int J Radiat Oncol Biol Phys 61:1107-1116, 2005). Concurrent cisplatin was administered per institutional practice either as a 100 mg/m2 bolus on days 1, 22, and 43 of RT,4 or 40 mg/m2 weekly on days 1, 8, 15, 22, 29, 36, and 43 of RT.3 Neoadjuvant chemotherapy was allowed. No patient received adjuvant chemotherapy, which was an exclusion criterion except for the patients randomly assigned to the adjuvant chemotherapy study arm.
The randomization procedure was performed through telephone from the central office of the Comprehensive Cancer Trials Unit at The Chinese University of Hong Kong. A computer program was used to generate the allocation list. Randomization was done 1:1, with stratification for the type of primary treatment (RT v CRT) and cancer stage (II/III v IV). Patients fulfilling all eligibility criteria were randomly assigned to either adjuvant chemotherapy followed by clinical observation and surveillance (arm 1) or clinical observation and surveillance only (arm 2).
All patients in both study arms had plasma EBV DNA checked at 6 to 8 weeks after completion of primary RT and again at 6 to 8 months after random assignment. Patients assigned to arm 1 would have the baseline positron emission tomography/computed tomography (PET/CT) scan at 6 to 12 weeks after completion of primary RT and then proceed to adjuvant chemotherapy. Patients assigned to arm 2 would have the baseline PET/CT scan at 6 to 12 weeks after completion of primary RT; they would then proceed to clinical observation and surveillance. When the study was first conceived in 2005, the value of performing plasma EBV DNA and PET/CT scan in patients with NPC who had already completed standard conventional staging procedures (as discussed next under Procedure) after completion of primary RT was undefined (King et al: Br J Radiol 81:291-298, 2008) and is currently not a routine investigation in the clinical care of asymptomatic patients with NPC in the study centers (Au et al: Oral Oncol 77:16-21, 2018). In the current study, the post-RT plasma EBV DNA was only reported as negative (undetectable) or positive (detectable) at the time of random assignment. The result of the baseline PET/CT was blinded in both arms to patients and investigators at the time of random assignment. This blinding was part of the study design and was included in the patients’ consent. When the patient had completed the second follow-up plasma EBV DNA and PET/CT scan at 6 to 8 months after random assignment, the result of both the baseline and the second follow-up PET/CT scan and the result of plasma EBV DNA would then be unblinded to both patient and investigator. Additional clinical management of the patient was at the discretion of the investigator or the attending clinician.
All pretreatment screening procedures had to be completed within 28 days before the day of randomization, which included demographic data, medical history (including concomitant medication), disease history (including date of initial diagnosis, histology, and Union for International Cancer Control/American Joint Committee on Cancer stage classification of NPC), complete physical examination (including performance status, body weight, height, and vital signs), baseline laboratory tests (hematology, biochemistry, and hepatitis B surface antigen), ECG, chest radiograph, ultrasound liver or CT scan of abdomen, and bone scan. All were the standard conventional staging procedures available at the study centers (unless the patient already had a staging PET or PET/CT scan as part of staging investigations [Au et al: Oral Oncol 77:16-21, 2018]).
Patients randomly assigned to adjuvant chemotherapy would receive cisplatin 40 mg/m2 intravenously and gemcitabine 1,000 mg/m2 intravenously, both administered on days 1 and 8 every 3 weeks for a total of six cycles. Carboplatin (area under the curve, 5) administered on day 1 only was used to replace cisplatin when significant (grade ≥ 2) peripheral or otoneurotoxicity existed. During adjuvant chemotherapy, laboratory hematology, biochemistry, and adverse events were evaluated on days 1 and 8 of each cycle. Adverse events were graded according to Common Terminology Criteria for Adverse Events (version 3). Chemotherapy was delayed by 1 week if the absolute neutrophil count was < 1.5 × 109/L or platelet count was < 75 × 109/L. Criteria of cisplatin dose modification for impaired renal function, neutropenic sepsis or grade 4 hematologic toxicity, and other nonhematologic toxicity were predefined in the study protocol. If adjuvant chemotherapy was delayed for > 4 weeks, the patient would be discontinued from subsequent doses of chemotherapy but would continue to complete the follow-up schedule.
All patients in both study arms were followed every 3 months in the first and second years, every 4 months in the third year, every 6 months in the fourth and fifth years, and then every year thereafter. Nasopharynx examination (with nasopharyngoscopy), physical examination (including body weight and performance status), and documentation of the nodal status were performed at each scheduled visit. Assessment of late toxicity (defined as toxicity that occurred at least 3 months after the last day of RT) of the skin, subcutaneous tissue, and salivary glands was conducted at each visit using the Radiation Therapy Oncology Group and European Organisation for Research and Treatment of Cancer late radiation morbidity scoring criteria.22
The primary end point of the trial is relapse-free survival (RFS), which is defined as the time from random assignment to the time of first recurrence of NPC (locoregional or distant metastasis) or death as a result of any cause, whichever was observed first, or until date of last follow-up. Secondary end points were overall survival, locoregional control, distant metastasis–free survival, toxicity of adjuvant chemotherapy, and correlation of plasma EBV DNA and PET/CT scan with clinical course and outcome. Overall survival is defined as the duration from the date of random assignment to the date of death as a result of any cause or censored at the date of last follow-up. Local recurrence-free survival is defined as the duration from the date of random assignment to the date of first local recurrence or death as a result of any cause or censored at the date of last follow-up. For the purpose of this study, local recurrence also includes regional (neck) nodal recurrence. Distant metastasis–free survival is defined as the duration from the date of random assignment to the date of first distant metastasis or death as a result of any cause or censored at the date of last follow-up. In the analysis of the whole patient population that underwent EBV DNA screening, the origin of all end points was calculated since the date of consent for post-RT plasma EBV DNA test.
When the current study was first conceptualized in 2005, we used the historical figure of approximately 30% for the RFS at 18 months in patients with detectable plasma EBV DNA post-RT.16,17 The magnitude of benefit from adjuvant chemotherapy is estimated to be approximately 60% for the study treatment to be of major clinical value. With an estimated hazard ratio (HR) of 2 (or 0.5 if observation arm as reference), approximately 100 patients were required with a power of 0.8 and an α of .05.
In the phase III trial that compared cisplatin and gemcitabine with cisplatin and fluorouracil for six cycles in the first-line recurrent or metastatic setting, the HR for RFS was 0.55 (95% CI, 0.44 to 0.68).19 Because the current study incorporated a follow-up PET/CT scan at 6 months after randomization, we estimated that we would be able to detect a much higher event rate in the observation arm on the basis of our previous historical data. Therefore, we anticipated that in a high-risk group defined by detectable plasma EBV DNA (a group with a very high relapse rate within 18 months) and as receiving a more active regimen of gemcitabine and cisplatin for six cycles, we could reasonably achieve an HR of 0.42 to 0.50 for RFS when the more active chemotherapy was compared with observation alone.
All analyses were performed by intention-to-treat principle except for toxicity of adjuvant chemotherapy, which was analyzed in the safety population (defined as patients who received at least one dose of chemotherapy). We performed a post hoc sensitivity analysis on patient groups of special interest: patients who received at least one dose of adjuvant chemotherapy, patients who complied with adjuvant chemotherapy, timing of initiation of adjuvant chemotherapy, and exclusion of patients with positive baseline PET/CT results. The linearity and proportional hazard assumptions were confirmed by model checking using plots of log-survival versus time, log-log curve, and martingale residuals (Altman et al: BMC Med 10:51, 2012). A Kaplan-Meier survival curve was used to analyze all time-to-event data. A log-rank test was used to assess the difference in survival between treatment groups. We regarded P < .05 as significant. All statistical analyses were performed using SAS 9.3 software (SAS Institute, Cary, NC).
|1.||Lee AW, Ma BB, Ng WT, et al: Management of nasopharyngeal carcinoma: Current practice and future perspective. J Clin Oncol 33:3356-3364, 2015 Link, Google Scholar|
|2.||Chua MLK, Wee JTS, Hui EP, et al: Nasopharyngeal carcinoma. Lancet 387:1012-1024, 2016 Crossref, Medline, Google Scholar|
|3.||Chan AT, Leung SF, Ngan RK, et al: Overall survival after concurrent cisplatin-radiotherapy compared with radiotherapy alone in locoregionally advanced nasopharyngeal carcinoma. J Natl Cancer Inst 97:536-539, 2005 Crossref, Medline, Google Scholar|
|4.||Al-Sarraf M, LeBlanc M, Giri PG, et al: Chemoradiotherapy versus radiotherapy in patients with advanced nasopharyngeal cancer: Phase III Randomized Intergroup Study 0099. J Clin Oncol 16:1310-1317, 1998 Link, Google Scholar|
|5.||Kwong DL, Sham JS, Au GK, et al: Concurrent and adjuvant chemotherapy for nasopharyngeal carcinoma: A factorial study. J Clin Oncol 22:2643-2653, 2004 Link, Google Scholar|
|6.||Rossi A, Molinari R, Boracchi P, et al: Adjuvant chemotherapy with vincristine, cyclophosphamide, and doxorubicin after radiotherapy in local-regional nasopharyngeal cancer: Results of a 4-year multicenter randomized study. J Clin Oncol 6:1401-1410, 1988 Link, Google Scholar|
|7.||Chi KH, Chang YC, Guo WY, et al: A phase III study of adjuvant chemotherapy in advanced nasopharyngeal carcinoma patients. Int J Radiat Oncol Biol Phys 52:1238-1244, 2002 Crossref, Medline, Google Scholar|
|8.||Chen L, Hu CS, Chen XZ, et al: Concurrent chemoradiotherapy plus adjuvant chemotherapy versus concurrent chemoradiotherapy alone in patients with locoregionally advanced nasopharyngeal carcinoma: A phase 3 multicentre randomised controlled trial. Lancet Oncol 13:163-171, 2012 Crossref, Medline, Google Scholar|
|9.||Chen L, Hu CS, Chen XZ, et al: Adjuvant chemotherapy in patients with locoregionally advanced nasopharyngeal carcinoma: Long-term results of a phase 3 multicentre randomised controlled trial. Eur J Cancer 75:150-158, 2017 Crossref, Medline, Google Scholar|
|10.||Blanchard P, Lee A, Marguet S, et al: Chemotherapy and radiotherapy in nasopharyngeal carcinoma: An update of the MAC-NPC meta-analysis. Lancet Oncol 16:645-655, 2015 Crossref, Medline, Google Scholar|
|11.||Ribassin-Majed L, Marguet S, Lee AWM, et al: What is the best treatment of locally advanced nasopharyngeal carcinoma? An individual patient data network meta-analysis. J Clin Oncol 35:498-505, 2017 Link, Google Scholar|
|12.||Kanakry JA, Hegde AM, Durand CM, et al: The clinical significance of EBV DNA in the plasma and peripheral blood mononuclear cells of patients with or without EBV diseases. Blood 127:2007-2017, 2016 Crossref, Medline, Google Scholar|
|13.||Lo YM, Chan LY, Lo KW, et al: Quantitative analysis of cell-free Epstein-Barr virus DNA in plasma of patients with nasopharyngeal carcinoma. Cancer Res 59:1188-1191, 1999 Medline, Google Scholar|
|14.||Lo YM, Chan AT, Chan LY, et al: Molecular prognostication of nasopharyngeal carcinoma by quantitative analysis of circulating Epstein-Barr virus DNA. Cancer Res 60:6878-6881, 2000 Medline, Google Scholar|
|15.||Kim KY, Le QT, Yom SS, et al: Current state of PCR-based Epstein-Barr virus DNA testing for nasopharyngeal cancer. J Natl Cancer Inst 109:djx007, 2017 Crossref, Google Scholar|
|16.||Chan AT, Lo YM, Zee B, et al: Plasma Epstein-Barr virus DNA and residual disease after radiotherapy for undifferentiated nasopharyngeal carcinoma. J Natl Cancer Inst 94:1614-1619, 2002 Crossref, Medline, Google Scholar|
|17.||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 350:2461-2470, 2004 Crossref, Medline, Google Scholar|
|18.||Chan AT, Ma BB, Lo YM, et al: Phase II study of neoadjuvant carboplatin and paclitaxel followed by radiotherapy and concurrent cisplatin in patients with locoregionally advanced nasopharyngeal carcinoma: Therapeutic monitoring with plasma Epstein-Barr virus DNA. J Clin Oncol 22:3053-3060, 2004 Link, Google Scholar|
|19.||Zhang L, Huang Y, Hong S, et al: Gemcitabine plus cisplatin versus fluorouracil plus cisplatin in recurrent or metastatic nasopharyngeal carcinoma: A multicentre, randomised, open-label, phase 3 trial. Lancet 388:1883-1892, 2016 Crossref, Medline, Google Scholar|
|20.||Le QT, Zhang Q, Cao H, et al: An international collaboration to harmonize the quantitative plasma Epstein-Barr virus DNA assay for future biomarker-guided trials in nasopharyngeal carcinoma. Clin Cancer Res 19:2208-2215, 2013 Crossref, Medline, Google Scholar|
|21.||Chan KCA, Woo JKS, King A, et al: Analysis of plasma Epstein-Barr virus DNA to screen for nasopharyngeal cancer. N Engl J Med 377:513-522, 2017 Crossref, Medline, Google Scholar|
|22.||Cox JD, Stetz J, Pajak TF: Toxicity criteria of the Radiation Therapy Oncology Group (RTOG) and the European Organization for Research and Treatment of Cancer (EORTC). Int J Radiat Oncol Biol Phys 31:1341-1346, 1995 Crossref, Medline, Google Scholar|
|23.||Kim KY, Le QT, Yom SS, et al: Clinical utility of Epstein-Barr virus DNA testing in the treatment of nasopharyngeal carcinoma patients. Int J Radiat Oncol Biol Phys 98:996-1001, 2017 Crossref, Medline, Google Scholar|
|24.||Hsu CL, Chan SC, Chang KP, et al: Clinical scenario of EBV DNA follow-up in patients of treated localized nasopharyngeal carcinoma. Oral Oncol 49:620-625, 2013 Crossref, Medline, Google Scholar|
|25.||Wang WY, Lin TY, Twu CW, et al: Long-term clinical outcome in nasopharyngeal carcinoma patients with post-radiation persistently detectable plasma EBV DNA. Oncotarget 7:42608-42616, 2016 Crossref, Medline, Google Scholar|
|26.||Twu CW, Wang WY, Chen CC, et al: Metronomic adjuvant chemotherapy improves treatment outcome in nasopharyngeal carcinoma patients with postradiation persistently detectable plasma Epstein-Barr virus deoxyribonucleic acid. Int J Radiat Oncol Biol Phys 89:21-29, 2014 Crossref, Medline, Google Scholar|
|27.||Antonia SJ, Villegas A, Daniel D, et al: Durvalumab after chemoradiotherapy in stage III non-small-cell lung cancer. N Engl J Med 377:1919-1929, 2017 Crossref, Medline, Google Scholar|
|28.||Hsu C, Lee SH, Ejadi S, et al: Safety and antitumor activity of pembrolizumab in patients with programmed death-ligand 1-positive nasopharyngeal carcinoma: Results of the KEYNOTE-028 study. J Clin Oncol 35:4050-4056, 2017 Link, Google Scholar|
|29.||Ma BBY, Lim WT, Goh BC, et al: Antitumor activity of nivolumab in recurrent and metastatic nasopharyngeal carcinoma: An international, multicenter study of the Mayo Clinic Phase 2 Consortium (NCI-9742). J Clin Oncol 36:1412-1418, 2018 Link, Google Scholar|