The n e w e ng l a n d j o u r na l of m e dic i n e

Original Article

Gemcitabine and Cisplatin Induction

Chemotherapy in Nasopharyngeal Carcinoma
Y. Zhang, L. Chen, G.-Q. Hu, N. Zhang, X.-D. Zhu, K.-Y. Yang, F. Jin, M. Shi,
Y.-P. Chen, W.-H. Hu, Z.-B. Cheng, S.-Y. Wang, Y. Tian, X.-C. Wang, Yan Sun,
J.-G. Li, W.-F. Li, Y.-H. Li, L.-L. Tang, Y.-P. Mao, G.-Q. Zhou, R. Sun, X. Liu,
R. Guo, G.-X. Long, S.-Q. Liang, L. Li, J. Huang, J.-H. Long, J. Zang, Q.-D. Liu,
L. Zou, Q.-F. Su, B.-M. Zheng, Y. Xiao, Y. Guo, F. Han, H.-Y. Mo, J.-W. Lv,
X.-J. Du, C. Xu, N. Liu, Y.-Q. Li, M.L.K. Chua, F.-Y. Xie, Ying Sun, and J. Ma​​

A BS T R AC T

BACKGROUND
The authors’ full names, academic de- Platinum-based concurrent chemoradiotherapy is the standard of care for patients
grees, and affiliations are listed in the Ap- with locoregionally advanced nasopharyngeal carcinoma. Additional gemcitabine and
pendix. Address reprint requests to Dr.
Ma at the Department of Radiation On- cisplatin induction chemotherapy has shown promising efficacy in phase 2 trials.
cology, Sun Yat-sen University Cancer METHODS
Center, No. 651 Dongfeng Rd. E., Guang-
zhou 510060, China, or at m ­ ajun2@​­mail​
In a parallel-group, multicenter, randomized, controlled, phase 3 trial, we compared
.­sysu​.­edu​.­cn. gemcitabine and cisplatin as induction chemotherapy plus concurrent chemoradio-
Drs. Y. Zhang, L. Chen, G.-Q. Hu, N.
therapy with concurrent chemoradiotherapy alone. Patients with locoregionally ad-
Zhang, Zhu, Yang, Jin, Shi, and Y.-P. Chen vanced nasopharyngeal carcinoma were randomly assigned in a 1:1 ratio to receive
and Drs. Chua, Xie, Ying Sun, and Ma gemcitabine (at a dose of 1 g per square meter of body-surface area on days 1 and 8)
contributed equally to this article.
plus cisplatin (80 mg per square meter on day 1), administered every 3 weeks for
This article was published on May 31, 2019, three cycles, plus chemoradiotherapy (concurrent cisplatin at a dose of 100 mg per
at NEJM.org.
square meter every 3 weeks for three cycles plus intensity-modulated radiotherapy)
N Engl J Med 2019;381:1124-35. or chemoradiotherapy alone. The primary end point was recurrence-free survival
DOI: 10.1056/NEJMoa1905287
Copyright © 2019 Massachusetts Medical Society.
(i.e., freedom from disease recurrence [distant metastasis or locoregional recurrence]
or death from any cause) in the intention-to-treat population. Secondary end points
included overall survival, treatment adherence, and safety.
RESULTS
A total of 480 patients were included in the trial (242 patients in the induction
chemotherapy group and 238 in the standard-therapy group). At a median follow-
up of 42.7 months, the 3-year recurrence-free survival was 85.3% in the induction
chemotherapy group and 76.5% in the standard-therapy group (stratified hazard ratio
for recurrence or death, 0.51; 95% confidence interval [CI], 0.34 to 0.77; P = 0.001).
Overall survival at 3 years was 94.6% and 90.3%, respectively (stratified hazard ratio
for death, 0.43; 95% CI, 0.24 to 0.77). A total of 96.7% of the patients completed
three cycles of induction chemotherapy. The incidence of acute adverse events of
grade 3 or 4 was 75.7% in the induction chemotherapy group and 55.7% in the
standard-therapy group, with a higher incidence of neutropenia, thrombocytope-
nia, anemia, nausea, and vomiting in the induction chemotherapy group. The in-
cidence of grade 3 or 4 late toxic effects was 9.2% in the induction chemotherapy
group and 11.4% in the standard-therapy group.
CONCLUSIONS
Induction chemotherapy added to chemoradiotherapy significantly improved recur-
rence-free survival and overall survival, as compared with chemoradiotherapy alone,
among patients with locoregionally advanced nasopharyngeal carcinoma. (Funded
by the Innovation Team Development Plan of the Ministry of Education and others;
ClinicalTrials.gov number, NCT01872962.)
1124 n engl j med 381;12 nejm.org  September 19, 2019

The New England Journal of Medicine

Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 Gemcitabine and Cisplatin in Nasopharyngeal Carcinoma

N
asopharyngeal carcinoma is a ticipating center approved the trial protocol, avail-
head and neck cancer with a specific geo- able at NEJM.org. The trial was performed accord-
graphic distribution. It affected an esti- ing to the principles of the Declaration of Helsinki
mated 130,000 patients worldwide in 2018, with and Good Clinical Practice guidelines as defined
the highest rates occurring in regions in South by the International Conference on Harmonisa-
China, Southeastern Asia, and North Africa.1 More tion. Written informed consent was obtained from
than 70% of patients receive a diagnosis of lo- all the patients before enrollment. Patients could
coregionally advanced disease at presentation,2 withdraw consent at any time after enrollment
and in this subgroup of patients with an unfa- and could discontinue the trial if disease progres-
vorable prognosis, concurrent chemoradiothera- sion or severe coexisting conditions occurred dur-
py with a platinum-based agent constitutes the ing treatment.
backbone of treatment, with the chemotherapy This was an investigator-initiated trial. The
sensitizing the tumor to the toxic effects of the last author designed, wrote, and implemented the
radiotherapy. Distant metastasis predominates as trial protocol and managed the trial. Lead inves-
the pattern of disease relapse, and it accounts for tigators from each center gathered the data and
cancer-specific mortality among approximately ensured its accuracy and completeness. One of the
70% of patients.3,4 authors conducted the statistical analyses. No one
The addition of chemotherapy as an induction who is not an author contributed to the writing of
or adjuvant regimen to chemoradiotherapy has the manuscript. The first author wrote the first
been investigated with mixed results.5-9 The tox- draft of manuscript, which was reviewed by all
icity of systemic therapy after chemoradiotherapy the authors. All the authors approved the final
remains a pertinent issue.6,10 The use of induction content of the manuscript. The trial sponsors had
chemotherapy is supported by the long-term re- no access to the data and were not involved in the
sults of a randomized, controlled trial in which data interpretation or the manuscript preparation
docetaxel, cisplatin, and fluorouracil were added or review. Qilu Pharmaceutical provided gem-
to chemoradiotherapy in patients with locoregion- citabine and cisplatin free of charge and was not
ally advanced nasopharyngeal carcinoma; patients involved in the trial design, data collection or
had prolonged overall survival with this regimen.7,9 analysis, or manuscript preparation or review.
Previous phase 2 trials have shown that gem- The last author vouches for the completeness and
citabine plus cisplatin is an effective chemothera- accuracy of the data and for the adherence of the
py in patients with nasopharyngeal carcinoma11-13 trial to the protocol.
and has been established as the first-line treat- Eligibility criteria included the following: an age
ment of choice over cisplatin plus fluorouracil in between 18 and 64 years; histologic confirmation
patients with recurrent or metastatic disease.14 of nonkeratinizing nasopharyngeal carcinoma; no
However, in the context of newly diagnosed, non- previous treatment for cancer; nondistant meta-
metastatic, locoregionally advanced disease, the static, newly diagnosed stage III to IVB disease
efficacy and safety profile of induction therapy (excluding subgroups of patients with low risk of
with gemcitabine plus cisplatin to chemoradio- metastasis; i.e., those with bulky primary tumor
therapy is unclear. We therefore conducted a mul- with no nodal involvement) that was staged accord-
ticenter, randomized, controlled, phase 3 clinical ing to the American Joint Committee on Cancer–
trial to investigate the efficacy and safety of add- Union for International Cancer Control 7th edition
ing gemcitabine plus cisplatin to chemoradio- stage-classification system15; a Karnofsky perfor-
therapy in patients with locoregionally advanced mance-status score of at least 70 (on a scale from
nasopharyngeal carcinoma. 0 to 100, with lower scores indicating greater dis-
ability); and adequate hematologic, renal, and he-
patic function. Key exclusion criteria were the fol-
Me thods
lowing: receipt of treatment with palliative intent;
Trial Design and Participants a history of cancer; receipt of previous treatment
This open-label, parallel-group, randomized, phase (radiotherapy, chemotherapy, or surgery [except di-
3 trial enrolled patients from 12 hospitals in China agnostic procedures]) to the nasopharynx or neck;
(Table S1 in the Supplementary Appendix, available lactation or pregnancy; or severe coexisting illness.
with the full text of this article at NEJM.org). For this trial, essential pretreatment evalua-
The institutional ethics review board at each par- tions included the following: complete history;

n engl j med 381;12 nejm.org  September 19, 2019 1125

The New England Journal of Medicine
Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

physical examination; hematologic and biochemi- tion of induction chemotherapy and 16 weeks
cal analyses; flexible nasopharyngoscopy; histo- after chemoradiotherapy. We used the Common
pathological diagnosis; and magnetic resonance Terminology Criteria for Adverse Events, version
imaging (MRI) or enhanced computed tomogra- 4.0, to grade acute toxic effects during treatment,
phy (CT) (if patients had contraindications to and late toxic effects that were associated with
MRI) of the nasopharynx and neck for primary radiotherapy were graded according to the Late
tumor staging. Distant metastasis staging was Radiation Morbidity Scoring Scheme of the Radia-
completed with CT examination of the chest and tion Therapy Oncology Group.18
abdomen and with skeletal scintigraphy. The use In the first 3 years of follow-up, all the patients
of 18F-fluorodeoxyglucose–positron-emission to- underwent assessment every 3 months and then
mography was recommended for patients with every 6 months thereafter until death. All end
advanced node stage or if there was a clinical points were assessed or confirmed by the physi-
suspicion of distant metastases.16 cian in charge. Fine-needle aspiration or biopsy of
suspected lesions was performed if deemed nec-
Randomization and Masking essary in order to confirm locoregional or distant
The randomization procedure was carried out by disease progression.
telephone from the Clinical Trials Center of the
Sun Yat-sen University Cancer Center. A computer End Points
program was used to generate the assignment list. The primary end point was recurrence-free sur-
Randomization was stratified according to treat- vival, which was defined as the time from ran-
ment center and tumor–node–metastasis (TNM) domization to documented disease recurrence
stage (III or IV), and patients were randomly as- (either distant metastasis or locoregional disease
signed in a 1:1 ratio in blocks of four to receive recurrence) or death from any cause, whichever
either three cycles of induction chemotherapy occurred first. Secondary end points included over-
plus chemoradiotherapy (induction chemothera- all survival, distant recurrence–free survival (free-
py group) or chemoradiotherapy alone (standard- dom from documented distant metastasis or death
therapy group). Treatment group assignment was from any cause), locoregional recurrence–free sur-
not masked. vival (freedom from documented locoregional
recurrence or death from any cause), treatment
Procedures response, treatment adherence, and safety. (Def-
Gemcitabine at a dose of 1 g per square meter of initions of the end points are provided in the
body-surface area on days 1 and 8 and cisplatin Supplementary Appendix.) Patients who were lost
at a dose of 80 mg per square meter on day 1 were to follow-up or were still alive without distant
administered intravenously once every 3 weeks metastasis or locoregional recurrence at the end
for three cycles.14 Cisplatin that was concurrent of the trial had their data censored at the date of
with radiotherapy was then administered intra- last follow-up.
venously at a dose of 100 mg per square meter
every 3 weeks on days 1, 22, and 43. Details of the Statistical Analysis
chemotherapy dose modifications and supportive This trial aimed to evaluate whether the addition
measures are provided in the Supplementary Ap- of induction chemotherapy to chemoradiotherapy
pendix. improved recurrence-free survival as compared
For radiotherapy, an intensity-modulated tech- with chemoradiotherapy alone. We estimated that
nique was mandatory in both groups. The guide- approximately 452 patients would need to un-
lines regarding radiotherapy3,17 are provided in the dergo randomization in a 1:1 ratio (226 patients
Supplementary Appendix. It was recommended per group) in order for 77 events to be observed
that patients in the induction chemotherapy group for the primary analysis of recurrence-free survival.
commence chemoradiotherapy within 21 to 28 We estimated that the trial would have 80% power
days after the first day of the last cycle of induc- to detect a hazard ratio for disease recurrence or
tion chemotherapy. death of 0.52 using a log-rank test with a two-sided
Tumors were assessed with the use of flexible significance level of 0.05. We further assumed
nasopharyngoscopy and MRI of the nasopharyn- that 5% of the patients would be lost to follow-
geal and neck areas at 1 week after the comple- up or would prematurely discontinue the trial.

1126 n engl j med 381;12 nejm.org  September 19, 2019

The New England Journal of Medicine

Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 Gemcitabine and Cisplatin in Nasopharyngeal Carcinoma

This yielded a final sample size of 476 (238 pa- R e sult s

tients per group).19
Efficacy analyses were performed in both the Patients and Treatment
intention-to-treat and per-protocol trial popula- From December 2013 through September 2016,
tions (see the Supplementary Appendix). The we enrolled 480 patients across 12 sites. The in-
safety population comprised patients who start- duction chemotherapy group comprised 242 pa-
ed treatment in each group. Kaplan–Meier curves tients, and the standard-therapy group comprised
were used to present time-to-event data, and the 238 patients (Fig. 1). The characteristics of the
two treatment groups were compared by means patients at baseline were well balanced between
of log-rank tests that were stratified according to the two groups (Table 1).
trial center and disease stage (primary analysis).20 Of the 242 patients who had been randomly
The stratified Cox proportional-hazards model, assigned to undergo induction chemotherapy plus
with treatment as a single covariate, was used to chemoradiotherapy, 239 (98.8%) started protocol-
calculate the hazard ratios and 95% confidence defined induction chemotherapy and were in-
intervals, and the proportional-hazards assump- cluded in the safety population (Fig. 1). A total of
tion was tested with Schoenfeld residuals.21 3 patients withdrew from the trial before the
We further performed an interaction analysis initiation of trial treatment (1 patient received
to explore whether the effect of the experimental chemoradiotherapy alone, and 2 received differ-
treatment varied in the subgroups defined accord- ent induction chemotherapy plus chemoradiother-
ing to sex, age, Karnofsky performance-status apy). A total of 231 of the 239 patients (96.7%)
score, tumor category, node category, and TNM completed three cycles of induction chemotherapy;
stage (see the Supplementary Appendix). The in- 8 patients (3.3%) did not complete all three cy-
teraction analysis was conducted by means of a cles of induction chemotherapy (3 patients re-
test of treatment-by-covariate interaction on the ceived two cycles and 5 received one cycle; the
basis of the Cox proportional-hazards model.22 reasons included a case of the patient declining
For each chemotherapy drug, we calculated the the treatment, a case of disease progression, and
relative dose intensity as the total dose actually discontinuation due to adverse events in the re-
received during treatment, divided by the dose maining six cases) (Table S2 in the Supplemen-
defined by the protocol. Dose intensity in the two tary Appendix). A total of 31 of the 239 patients
groups was compared with the use of the Wil- (13.0%) had dose reductions of gemcitabine or
coxon rank-sum test. cisplatin, mainly because of hematologic toxic
An independent data monitoring committee effects (in 21 patients). Overall, the median rela-
monitored the trial and made decisions regarding tive dose intensity was 100% (interquartile range,
possible early trial stoppage. An interim analysis 100 to 100) for gemcitabine and 100% (interquar-
was performed on October 30, 2016, when the tile range, 100 to 100) for cisplatin.
trial reached approximately half the expected In the induction chemotherapy group, 234 of
number of disease recurrences or deaths (i.e., 38 239 patients (97.9%) received concurrent chemo-
events). To preserve an overall type I error rate of radiotherapy after induction chemotherapy, and
0.05 for the entire trial, the O’Brien–Fleming type 5 patients (2.1%) did not (Fig. 1). A total of 93 pa-
boundary (alpha of 0.003) was used for early tients (38.9%) completed three cycles of concurrent
trial stoppage. The data monitoring committee cisplatin, 127 (53.1%) received two cycles, and 14
reviewed the analyses and approved the continu- (5.9%) received one cycle.
ation of the trial. Separately, we implemented a In the standard-therapy group, of the 238
central safety monitoring committee to monitor patients who had undergone randomization, 237
unexpected adverse events and treatment-related received the protocol-defined concurrent chemo-
death. The database was locked on April 15, 2019, radiotherapy and were included in the safety popu-
and here we report the survival and toxicity lation; 1 patient who deviated from the protocol
analyses. Analyses were conducted with the use of received weekly cisplatin (Table S3 in the Supple-
SPSS software, version 22.0 (IBM), and Stata soft- mentary Appendix). A total of 177 of 237 patients
ware, version 14.2 (StataCorp). The type I error (74.7%) completed three cycles of concurrent cis-
rate was set to 0.05 for the primary end point, and platin, 56 (23.6%) received two cycles, and 4 (1.7%)
all tests were two-sided. received one cycle.

n engl j med 381;12 nejm.org  September 19, 2019 1127

The New England Journal of Medicine
Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

480 Patients underwent randomization

242 Were assigned to receive induction 238 Were assigned to receive concurrent
chemotherapy with gemcitabine and cisplatin chemoradiotherapy alone
plus concurrent chemoradiotherapy (intention-to-treat population)
(intention-to-treat population)

3 Withdrew consent and did not receive

induction chemotherapy with gem-
citabine and cisplatin
2 Received other induction chemo-
therapy and concurrent chemo-
radiotherapy
1 Received concurrent chemoradio-
therapy alone

239 Started induction chemotherapy with

gemcitabine and cisplatin (safety
population) 1 Did not receive protocol-defined
8 Discontinued induction chemotherapy concurrent cisplatin (withdrew
5 Received 1 cycle consent and received radiotherapy
5 Had adverse event and weekly cisplatin)
3 Received 2 cycles
1 Had adverse event
1 Had disease progression
1 Declined to participate

5 Did not receive concurrent cisplatin

4 Received radiotherapy alone
2 Had adverse event
1 Had disease progression
1 Withdrew consent
1 Received radiotherapy and carboplatin
1 Had adverse event

234 Started concurrent chemoradiotherapy 237 Started concurrent chemoradiotherapy

141 Discontinued concurrent cisplatin (safety population)
14 Completed 1 cycle 2 Discontinued radiotherapy
7 Had adverse event 2 Declined to participate
7 Declined to participate 60 Discontinued concurrent cisplatin
127 Completed 2 cycles 4 Completed 1 cycle
91 Declined to participate 3 Declined to participate
13 Had adverse event 1 Had adverse event
23 Had other reason 56 Completed 2 cycles
44 Declined to participate
6 Had adverse event
6 Had other reason

218 Were included in the per-protocol 177 Were included in the per-protocol
population (completed 3 cycles of induction population (completed 3 cycles of concurrent
chemotherapy with gemcitabine and cisplatin cisplatin and radiotherapy)
and 2 to 3 cycles of concurrent cisplatin and
radiotherapy)

Figure 1. Enrollment, Randomization, and Follow-up.

The induction chemotherapy group received gemcitabine and cisplatin plus concurrent chemoradiotherapy, and the standard-therapy
group received chemoradiotherapy alone. Other induction chemotherapy regimens included docetaxel, cisplatin plus fluorouracil, and
docetaxel plus cisplatin.

1128 n engl j med 381;12 nejm.org  September 19, 2019

The New England Journal of Medicine

Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 Gemcitabine and Cisplatin in Nasopharyngeal Carcinoma

Overall, 191 of 239 patients (79.9%) in the in- Table 1. Characteristics of the Patients at Baseline.*
duction chemotherapy group and 227 of 237 pa-
tients (95.8%) in the standard-therapy group Induction Chemotherapy Standard Therapy
Characteristic (N = 242) (N = 238)
received at least 200 mg per square meter of
concurrent cisplatin. The median dose intensity Median age (range) — yr 46 (18–64) 45 (20–64)
for concurrent cisplatin, which had been admin- Sex — no. (%)
istered every 3 weeks, was 200 mg per square Male 182 (75.2) 164 (68.9)
meter (interquartile range, 200 to 300) in the Female 60 (24.8) 74 (31.1)
induction chemotherapy group and 300 mg per
Karnofsky performance-
square meter (interquartile range, 200 to 300) in status score†
the standard-therapy group (P<0.001). The medi-
100 10 (4.1) 10 (4.2)
an cumulative dose of cisplatin was 440 mg per
90 189 (78.1) 198 (83.2)
square meter (interquartile range, 440 to 540) in
the induction chemotherapy group; only 63 of 80 36 (14.9) 21 (8.8)
239 patients (26.4%) received the full cumulative 70 7 (2.9) 9 (3.8)
dose of 540 mg per square meter. Tumor category — no. (%)‡
Regarding radiotherapy, all 239 patients in the T1 2 (0.8) 3 (1.3)
induction chemotherapy group completed proto- T2 16 (6.6) 16 (6.7)
col-defined intensity-modulated radiotherapy. The
T3 115 (47.5) 116 (48.7)
median time from the start of the last induction
T4 109 (45.0) 103 (43.3)
chemotherapy cycle to the commencement of
chemoradiotherapy was 25 days (interquartile Node category — no. (%)‡
range, 22 to 28). In the standard-therapy group, N1 114 (47.1) 106 (44.5)
radiotherapy was discontinued in 2 of 237 pa- N2 101 (41.7) 108 (45.4)
tients (0.8%) because the patients declined the N3A 12 (5.0) 8 (3.4)
treatment. The time to the completion of radio-
N3B 15 (6.2) 16 (6.7)
therapy and the radiotherapy doses received were
Disease stage — no. (%)‡
similar in the two groups (Table S3 in the Sup-
plementary Appendix). III 111 (45.9) 120 (50.4)
IVA 104 (43.0) 94 (39.5)
Efficacy IVB 27 (11.2) 24 (10.1)
Overall, 94.6% of the patients (226 of 239) had a
response after induction chemotherapy before the * Patients in the induction chemotherapy group received gemcitabine plus cis-
platin as well as concurrent chemoradiotherapy, and patients in the standard-
start of chemoradiotherapy; 24 patients (10.0%) therapy group received concurrent chemoradiotherapy alone. There were no
had a complete response, and 202 (84.5%) had a significant differences between the treatment groups in the characteristics at
partial response. At 16 weeks after radiotherapy, baseline. Percentages may not total 100 because of rounding.
† Karnofsky performance-status scores are assessed on a scale from 0 to 100,
97.1% of the patients (235 of 242) in the induc- with lower scores indicating greater disability.
tion chemotherapy group had a complete response, ‡ Tumor and node categories and disease stage were assessed according to the
as did 96.6% of the patients (230 of 238) in the 7th edition of American Joint Committee on Cancer–Union for International
Cancer Control stage classification system.15
standard-therapy group (Table 2).
At the last follow-up on April 15, 2019, the
median follow-up was 42.7 months (range, 3.5 to
65.0). A total of 296 of the 427 patients who were standard-therapy group. Details regarding the
alive (69.3%) as of this date had follow-up re- patterns of relapse and subsequent therapies
cords of at least 36 months, and the last patient after relapse are provided in Tables S5 and S6 in
who had enrolled in the trial had a follow-up of the Supplementary Appendix. The 3-year recur-
31.2 months. We recorded a total of 100 events rence-free survival was 85.3% (95% confidence
of recurrence or death (20.8% of the patients in interval [CI], 80.0 to 89.3) in the induction che-
the overall trial population), including events in motherapy group, as compared with 76.5% (95%
37 of 242 patients (15.3%) in the induction che- CI, 70.4 to 81.5) in the standard-therapy group
motherapy group and in 63 of 238 (26.5%) in the (stratified hazard ratio for recurrence or death,

n engl j med 381;12 nejm.org  September 19, 2019 1129

The New England Journal of Medicine
Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 2. Survival and Response to Treatment.*

Induction Chemotherapy Standard Therapy Hazard Ratio

Variable (N = 242) (N = 238) (95% CI)
Recurrence-free survival
Recurrence or death — no. (%) 37 (15.3) 63 (26.5)
Percentage of patients alive and without recurrence 85.3 (80.0–89.3) 76.5 (70.4–81.5) 0.51 (0.34–0.77)
at 3 yr (95% CI)
Overall survival
Death — no. (%) 18 (7.4) 35 (14.7)
Percentage of patients alive at 3 yr (95% CI) 94.6 (90.6–96.9) 90.3 (85.6–93.5) 0.43 (0.24–0.77)
Distant recurrence–free survival
Distant metastasis or death — no. (%) 23 (9.5) 40 (16.8)
Percentage of patients alive and without distant 91.1 (86.4–94.2) 84.4 (79.1–88.5) 0.43 (0.25–0.73)
metastasis at 3 yr (95% CI)
Locoregional recurrence–free survival
Locoregional recurrence or death — no. (%) 17 (7.0) 22 (9.2)
Percentage of patients alive and without locoregional 91.8 (87.3–94.7) 91.0 (86.2–94.0) 0.77 (0.42–1.41)
recurrence at 3 yr (95% CI)
Response to induction chemotherapy†
Complete response — no./total no. (%) 24/239 (10.0) —
Partial response — no./total no. (%) 202/239 (84.5) —
Stable disease — no./total no. (%) 10/239 (4.2) —
Progressive disease — no./total no. (%) 3/239 (1.3) —
Response to whole treatment — no. (%)
Complete response 235 (97.1) 230 (96.6)
Partial response 2 (0.8) 5 (2.1)
Progressive disease 1 (0.4) 1 (0.4)
Could not be assessed 4 (1.7) 2 (0.8)

* Hazard ratios were calculated by a stratified Cox proportional-hazards model. CI denotes confidence interval.
† Response to induction chemotherapy was assessed only in the 239 patients who began receiving induction chemotherapy.

0.51; 95% CI, 0.34 to 0.77; P = 0.001) (Table 2 and the standard-therapy group (91.1% [95% CI, 86.4
Fig. 2A). to 94.2] vs. 84.4% [95% CI, 79.1 to 88.5]; strati-
At the time of analysis, 18 of 242 patients fied hazard ratio for distant recurrence or death,
(7.4%) in the induction chemotherapy group and 0.43; 95% CI, 0.25 to 0.73) (Table 2 and Fig. 2C).
35 of 238 patients (14.7%) in the standard-therapy However, the 3-year locoregional recurrence–free
group had died. Details regarding the cause of survival was similar in the induction chemother-
death are provided in Table S5 in the Supplemen- apy group and the standard-therapy group
tary Appendix. Patients in the induction chemo- (91.8% [95% CI, 87.3 to 94.7] and 91.0% [95%
therapy group had better 3-year overall survival CI, 86.2 to 94.0], respectively; stratified hazard
than those in the standard-therapy group (94.6% ratio for locoregional recurrence or death, 0.77;
[95% CI, 90.6 to 96.9] vs. 90.3% [95% CI, 85.6 95% CI, 0.42 to 1.41) (Table 3 and Fig. 2D).
to 93.5]; stratified hazard ratio for death, 0.43;
95% CI, 0.24 to 0.77) (Table 2 and Fig. 2B). The Adverse Events
3-year distant recurrence–free survival was bet- During induction chemotherapy, 93 of 239 pa-
ter in the induction chemotherapy group than in tients (38.9%) had acute adverse events of grade

1130 n engl j med 381;12 nejm.org  September 19, 2019

The New England Journal of Medicine

Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 Gemcitabine and Cisplatin in Nasopharyngeal Carcinoma

A Recurrence-free Survival B Overall Survival

100 100 Induction chemotherapy
Induction chemotherapy
90 90

Percentage of Patients

Percentage of Patients
80 80
70 70 Standard therapy
60 Standard therapy 60
50 3-Yr recurrence-free survival: 50
40 Induction chemotherapy, 85.3% 40 3-Yr overall survival:
30 vs. standard therapy, 76.5% 30 Induction chemotherapy, 94.6%
Hazard ratio for recurrence or death, 0.51 vs. standard therapy, 90.3%
20 20
(95% CI, 0.34–0.77) Hazard ratio for death, 0.43
10 10
P=0.002 (95% CI, 0.24–0.77)
0 0
0 12 24 36 48 60 0 12 24 36 48 60
Months since Randomization Months since Randomization
No. at Risk No. at Risk
Induction chemotherapy 242 234 215 146 93 35 Induction chemotherapy 242 241 236 162 100 36
Standard therapy 238 217 194 130 73 26 Standard therapy 238 232 219 152 87 29

C Distant Recurrence–free Survival D Locoregional Recurrence–free Survival

100 Induction chemotherapy 100 Induction chemotherapy
90 90
Percentage of Patients

Percentage of Patients
80 80 Standard therapy
70 Standard therapy 70
60 60
50 50
40 3-Yr distant recurrence–free survival: 40 3-Yr locoregional recurrence–free survival:
Induction chemotherapy, 91.1% Induction chemotherapy, 91.8%
30 30
vs. standard therapy, 84.4% vs. standard therapy, 91.0%
20 20
Hazard ratio for recurrence or death, 0.43 Hazard ratio for recurrence or death, 0.77
10 (95% CI, 0.25–0.73) 10 (95% CI, 0.42–1.41)
0 0
0 12 24 36 48 60 0 12 24 36 48 60
Months since Randomization Months since Randomization
No. at Risk No. at Risk
Induction chemotherapy 242 238 226 154 96 35 Induction chemotherapy 242 237 225 152 97 36
Standard therapy 238 217 204 140 80 28 Standard therapy 238 230 206 141 81 27

Figure 2. Kaplan–Meier Analysis of Recurrence-free Survival, Overall Survival, Distant Recurrence–free Survival, and Locoregional
­Recurrence–free Survival (Intention-to-Treat Population).
Hazard ratios and 95% confidence intervals were calculated by a stratified Cox proportional-hazards model. The primary end point was
recurrence-free survival, defined as the time from randomization to documented disease recurrence (either distant metastasis or loco­
regional disease recurrence) or death from any cause, whichever occurred first. Secondary end points included overall survival, distant
recurrence–free survival, and locoregional recurrence–free survival. Tick marks indicate censored data.

3 or 4 (Table S4 in the Supplementary Appendix). [11.3%] vs. 3 [1.3%]), anemia (23 [9.6%] vs. 2
Neutropenia was the most common event (in 49 [0.8%]), nausea (55 [23.0%] vs. 33 [13.9%]), and
patients [20.5%]), followed by leukopenia (in 26 vomiting (54 [22.6%] vs. 33 [13.9%]). The induc-
[10.9%]) and vomiting (in 26 [10.9%]). Over the tion chemotherapy group also had a higher inci-
entire treatment course, 181 patients (75.7%) in dence than the standard-therapy group of grade
the induction chemotherapy group and 132 (55.7%) 1 or 2 nephrotoxic effects (46 patients [19.2%] vs.
in the standard-therapy group reported adverse 27 [11.4%]) but not of ototoxic effects such as
events of grade 3 or 4 (Table 3). Mucositis was deafness or otitis (172 [72.0%] and 178 [75.1%],
the most common adverse event of grade 3 or 4 respectively).
(in 69 patients [28.9%] in the induction chemo- The incidence of all late toxic effects of grade
therapy group and in 76 [32.1%] in the standard- 1 or 2 was 84.9% (203 of 239 patients) in the
therapy group). The induction chemotherapy group induction chemotherapy group and 87.8% (208
had a higher incidence than the standard-therapy of 237) in the standard-therapy group. A total of
group of grade 3 or 4 neutropenia (67 patients 9.2% of the patients (22 of 239) in the induction
[28.0%] vs. 25 [10.5%]), thrombocytopenia (27 chemotherapy group and 11.4% of the patients

n engl j med 381;12 nejm.org  September 19, 2019 1131

The New England Journal of Medicine
Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 3. Adverse Events, According to Trial Group and Grade.*

Induction Chemotherapy Standard Therapy

Event (N = 239) (N = 237)

Grade 1 or 2 Grade 3 or 4 Grade 1 or 2 Grade 3 or 4

number of patients with event (percent)

Any acute adverse event 58 (24.3) 181 (75.7) 105 (44.3) 132 (55.7)
Leukopenia 168 (70.3) 63 (26.4) 178 (75.1) 48 (20.3)
Neutropenia 135 (56.5) 67 (28.0) 147 (62.0) 25 (10.5)
Febrile neutropenia 0 1 (0.4) 0 0
Neutropenic infection 0 0 0 0
Anemia 178 (74.5) 23 (9.6) 157 (66.2) 2 (0.8)
Thrombocytopenia 91 (38.1) 27 (11.3) 54 (22.8) 3 (1.3)
Mucositis 139 (58.2) 69 (28.9) 154 (65.0) 76 (32.1)
Vomiting 85 (35.6) 54 (22.6) 52 (21.9) 33 (13.9)
Nausea 176 (73.6) 55 (23.0) 188 (79.3) 33 (13.9)
Dry mouth 168 (70.3) 12 (5.0) 166 (70.0) 6 (2.5)
Diarrhea 18 (7.5) 6 (2.5) 15 (6.3) 4 (1.7)
Dermatitis 141 (59.0) 5 (2.1) 152 (64.1) 9 (3.8)
Weight loss 148 (61.9) 5 (2.1) 145 (61.2) 4 (1.7)
Deafness or otitis 172 (72.0) 0 178 (75.1) 0
Nephrotoxic event 46 (19.2) 6 (2.5) 27 (11.4) 1 (0.4)
Hepatotoxic event 68 (28.5) 6 (2.5) 53 (22.4) 0
Any late adverse event 203 (84.9) 22 (9.2) 208 (87.8) 27 (11.4)
Symptomatic temporal-lobe necrosis 14 (5.9) 0 19 (8.0) 2 (0.8)
Cranial neuropathy 6 (2.5) 2 (0.8) 8 (3.4) 2 (0.8)
Peripheral neuropathy 21 (8.8) 3 (1.3) 4 (1.7) 0
Eye damage 3 (1.3) 0 2 (0.8) 0
Deafness or otitis 60 (25.1) 13 (5.4) 65 (27.4) 16 (6.8)
Dry mouth 179 (74.9) 7 (2.9) 190 (80.2) 5 (2.1)
Neck tissue damage 62 (25.9) 1 (0.4) 74 (31.2) 3 (1.3)
Bone necrosis 4 (1.7) 0 6 (2.5) 2 (0.8)
Trismus 7 (2.9) 0 9 (3.8) 0
Nephrotoxic event 7 (2.9) 0 5 (2.1) 0

* This analysis was conducted in the safety population, which included only patients who began receiving the trial treatment.

(27 of 237) in the standard-therapy group had Discussion

one or more late toxic effects of grade 3 or 4
(Table 3). The incidence of late toxic effects was We report the results of a randomized, controlled,
similar in the treatment groups, with the excep- phase 3 trial that showed the superior tumor
tion of grade 1 or 2 peripheral neuropathy, the control and survival with the addition of induc-
incidence of which was higher in the induction tion chemotherapy to chemoradiotherapy in a
chemotherapy group than in the standard-therapy selected cohort of patients with high-risk loco­
­
group (8.8% vs. 1.7%). regionally advanced nasopharyngeal carcinoma.

1132 n engl j med 381;12 nejm.org  September 19, 2019

The New England Journal of Medicine

Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 Gemcitabine and Cisplatin in Nasopharyngeal Carcinoma

The majority of patients had unfavorable prog- docetaxel plus cisplatin and fluorouracil (differ-
nostic features of N2 or N3 disease or bulky ence vs. standard therapy, 8.0 percentage points).7
primary tumors (T3 or T4), all of which are sur- It remains unclear whether these regimens differ
rogates for occult metastasis.23 The efficacy of significantly with regard to efficacy or toxicity.
induction chemotherapy was due to the lower We noted a higher overall incidence of acute
incidence of distant metastatic recurrences in adverse events among patients treated with in-
the induction chemotherapy group than in the duction chemotherapy than among those treated
standard-therapy group, which probably explained with chemoradiotherapy alone; in particular, the
the early overall survival advantage in the pa- incidence of severe neutropenia, thrombocytope-
tients treated in the induction chemotherapy nia, anemia, nausea, and vomiting was higher in
group. The 3-year recurrence-free survival was the induction chemotherapy group. The incidence
85.3% in the induction chemotherapy group and of acute grade 1 or 2 nephrotoxic effects was also
76.5% in the standard-therapy group, which cor- greater with the higher cumulative dose of cis­
responded to overall survival at 3 years that was platin in the induction chemotherapy group, al-
4.3 percentage points higher with induction though the frequencies of late nephrotoxic effects
chemotherapy than with standard therapy. This and ototoxic effects were similar in the treatment
clinical advantage was evident when the out- groups, with the exception of grade 1 or 2 periph-
comes of the two treatment groups were ana- eral neuropathy. The incidence of severe late com-
lyzed in patients who completed the planned plications was low in both groups, and we did
course of treatment (three cycles of chemother­ not find any treatment-related deaths. In our
apy plus two or three cycles of concurrent cis- previous chemotherapy trial (which investigated
platin and radiotherapy vs. three cycles of con- docetaxel plus cisplatin plus fluorouracil),7 a high
current cisplatin and radiotherapy alone) (see the incidence of grade 3 or 4 acute adverse events
Supplementary Appendix). such as neutropenia (35.5%), leukopenia (27.2%),
Among patients with recurrent or metastatic and diarrhea (8.0%) was observed despite the
disease, objective response rates of 64% and 91% modified doses (a 20% reduction of each drug as
(including complete and partial responses) have compared with the other trials26,27); in that trial,
been observed with gemcitabine plus cisplatin four patients (2%) had leukopenic fever, and one
alone and together with camrelizumab (an anti– patient died from neutropenic sepsis (Table S7
programmed death 1 antibody), respectively.24 in the Supplementary Appendix). Given the pau-
Likewise, in patients with locoregionally advanced city of comparative data, the choice of either a
disease, Yau and colleagues found that three gemcitabine-based or taxane-based induction che-
cycles of induction therapy with gemcitabine plus motherapy regimen could be made on the basis of
cisplatin yielded high percentages (>90%) of pa- the expected toxic effects matched against the pa-
tients with a clinical response.12 Our results con- tient’s status with regard to coexisting conditions.
trast with those from a similar trial conducted by In conclusion, the addition of induction thera-
Tan et al., in which a combination of gemcitabine, py with gemcitabine plus cisplatin to a backbone
carboplatin, and paclitaxel did not prolong pro- of chemoradiotherapy with cisplatin, adminis-
gression-free survival or overall survival.25 Sev- tered every 3 weeks, improved recurrence-free
eral explanations may be offered for the con- survival among patients with high-risk locoregion-
flicting results, including a cohort with more ally advanced nasopharyngeal cancer. This result
favorable characteristics in the trial conducted translated into a 4.3-percentage-point advantage
by Tan et al. (fewer patients with N2 or N3 dis- in survival over standard therapy at 3 years, at the
ease than in the present trial) and the use of cost of a higher incidence of acute adverse events.
low-dose carboplatin (area under the curve, 2.0) Supported by grants from the Innovation Team Development
that could have compromised the synergy with Plan of the Ministry of Education (IRT_17R110), the Natural Sci-
ence Foundation of Guangdong Province (2017A030312003), the
gemcitabine. The improvements in recurrence- Planned Science and Technology Project of Guangdong Province
free survival at 3 years are similar among patients (2019B020230002), the Health and Medical Collaborative Inno-
who received induction therapy with gemcitabine vation Project of Guangzhou City, China (201803040003), the
Sun Yat-Sen University Clinical Research 5010 Program
plus cisplatin (difference vs. standard therapy, (2014009), and the Overseas Expertise Introduction Project for
8.8 percentage points) and those who received Discipline Innovation (111 Project, B14035) and by a National

n engl j med 381;12 nejm.org  September 19, 2019 1133

The New England Journal of Medicine
Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 The n e w e ng l a n d j o u r na l of m e dic i n e

Medical Research Council Singapore Clinician-Scientist Award study centers; Xiao-Qing Hu, Xiao-Dan Jiang, Xiao-Fen Xiao,
(NMRC/CSA/0027/2018) and the Duke–NUS Oncology Academic Qiu-Hui Zheng, and Hui-Xia Feng (Department of Radiation On-
Program Proton Research Program (EX/42-A92) (both to Dr. Chua). cology, Sun Yat-sen University Cancer Center) for assistance with
Disclosure forms provided by the authors are available with data management and logistic support; the staff of the National
the full text of this article at NEJM.org. Clinical Study Center for Anticancer Drugs, Sun Yat-sen Univer-
A data sharing statement provided by the authors is available sity Cancer Center, for trial monitoring, data management, and
with the full text of this article at NEJM.org. statistical analysis; and Tan Sze Huey (Division of Clinical Trials
We thank the patients who participated in this study, their and Epidemiologic Sciences, National Cancer Center Singapore)
families, and the medical, nursing, and research staff at the for assistance with statistics.

Appendix
The authors’ full names and academic degrees are as follows: Yuan Zhang, M.D., Ph.D., Lei Chen, M.D., Ph.D., Guo‑Qing Hu, M.D.,
Ning Zhang, M.D., Xiao‑Dong Zhu, M.D., Ph.D., Kun‑Yu Yang, M.D., Feng Jin, M.D., Mei Shi, M.D., Ph.D., Yu‑Pei Chen, M.D.,
Wei‑Han Hu, M.D., Zhi‑Bin Cheng, M.D., Si‑Yang Wang, M.D., Ye Tian, M.D., Xi‑Cheng Wang, M.D., Yan Sun, M.D., Ph.D., Jin‑Gao
Li, M.D., Wen‑Fei Li, M.D., Ph.D., Yu‑Hong Li, M.D., Ling‑Long Tang, M.D., Ph.D., Yan‑Ping Mao, M.D., Ph.D., Guan‑Qun Zhou,
M.D., Ph.D., Rui Sun, M.D., Xu Liu, M.D., Ph.D., Rui Guo, M.D., Ph.D., Guo‑Xian Long, M.D., Ph.D., Shao‑Qiang Liang, M.D., Ling
Li, M.D., Ph.D., Jing Huang, M.D., Ph.D., Jin‑Hua Long, M.D., Jian Zang, M.D., Qiao‑Dan Liu, M.D., Ph.D., Li Zou, M.D., Ph.D.,
Qiong‑Fei Su, M.D., Bao‑Min Zheng, M.D., Ph.D., Yun Xiao, M.D., Ying Guo, Ph.D., Fei Han, M.D., Ph.D., Hao‑Yuan Mo, M.D., Jia‑Wei
Lv, M.D., Xiao‑Jing Du, M.D., Ph.D., Cheng Xu, M.D., Ph.D., Na Liu, M.D., Ph.D., Ying‑Qin Li, M.D., Ph.D., Melvin L. K. Chua, M.D.,
Ph.D., Fang‑Yun Xie, M.D., Ying Sun, M.D., Ph.D., and Jun Ma, M.D.
The authors’ affiliations are as follows: the Departments of Radiation Oncology (Y.Z., L.C., Y.-P.C., W.-H.H., W.-F.L., L.-L.T.,
Y.-P.M., G.-Q.Z., R.S., X.L., R.G., F.H., J.-W.L., X.-J.D., C.X., N.L., Y.-Q.L., F.-Y.X., Ying Sun, J.M.), Medical Oncology (Y.-H.L.), and
Nasopharyngeal Carcinoma (H.-Y.M.) and the Clinical Trials Center (Y.G.), Sun Yat-sen University Cancer Center, the State Key Labora-
tory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangdong Key Laboratory of Nasopharyngeal
Carcinoma Diagnosis and Therapy (Y.Z., L.C., Y.-P.C., W.-H.H., W.-F.L., L.-L.T., Y.-P.M., G.-Q.Z., R.S., X.L., R.G., F.H., J.-W.L., X.-J.D.,
C.X., N.L., Y.-Q.L., F.-Y.X., Ying Sun, J.M.), and the Department of Radiation Oncology, First Affiliated Hospital of Guangdong Phar-
maceutical University (X.-C.W., Q.-F.S.), Guangzhou, the Cancer Center, Tongji Hospital Affiliated to Tongji Medical College (G.-Q.H.,
G.-X.L.), and the Cancer Center, Union Hospital, Tongji Medical College (K.-Y.Y., J.H.), Huazhong University of Science and Technol-
ogy, Wuhan, the Department of Radiation Oncology, First People’s Hospital of Foshan, Foshan (N.Z., S.-Q.L.), the Department of Ra-
diation Oncology, Affiliated Cancer Hospital of Guangxi Medical University, Nanning (X.-D.Z., L.L.), the Department of Head and Neck
Oncology, Affiliated Hospital of Guizhou Medical University, Guizhou Cancer Hospital, Guiyang (F.J., J.-H.L.), the Department of Ra-
diation Oncology, XiJing Hospital of Fourth Military Medical University, Xi’an (M.S., J.Z.), the Cancer Center (Z.-B.C.), and the Depart-
ment of Head and Neck Oncology (S.-Y.W., Q.-D.L.), Fifth Affiliated Hospital of Sun Yat-sen University, Zhuhai, the Department of
Radiation Oncology, Second Affiliated Hospital of Soochow University, Suzhou (Y.T., L.Z.), the Department of Radiation Oncology,
Peking University Cancer Hospital, Beijing (Yan Sun, B.-M.Z.), and the Department of Radiation Oncology, Jiangxi Cancer Hospital,
Nanchang (J.-G.L., Y.X.) — all in China; and the Divisions of Radiation Oncology and Medical Sciences, National Cancer Center Singa-
pore, and the Oncology Academic Program, Duke–National University of Singapore Medical School — both in Singapore (M.L.K.C.).

References
1. Bray F, Ferlay J, Soerjomataram I, Sie- geal cancer: phase III randomized Inter- Oncology Group (TCOG) 1303 Study. Ann
gel RL, Torre LA, Jemal A. Global cancer group study 0099. J Clin Oncol 1998;​16:​ Oncol 2018;​29:​1972-9.
statistics 2018: GLOBOCAN estimates of 1310-7. 9. Li WF, Chen NY, Zhang N, et al. Con-
incidence and mortality worldwide for 36 6. Chen L, Hu CS, Chen XZ, et al. Con- current chemoradiotherapy with/without
cancers in 185 countries. CA Cancer J Clin current chemoradiotherapy plus adjuvant induction chemotherapy in locoregional­
2018;​68:​394-424. chemotherapy versus concurrent chemo- ly advanced nasopharyngeal carcinoma:
2. Mao YP, Xie FY, Liu LZ, et al. Re-eval- radiotherapy alone in patients with loco­ long-term results of phase 3 randomized
uation of 6th edition of AJCC staging sys- regionally advanced nasopharyngeal carci- controlled trial. Int J Cancer 2019;​145:​
tem for nasopharyngeal carcinoma and noma: a phase 3 multicentre randomised 295-305.
proposed improvement based on magnetic controlled trial. Lancet Oncol 2012;​13:​163- 10. Chan ATC, Hui EP, Ngan RKC, et al.
resonance imaging. Int J Radiat Oncol 71. Analysis of plasma Epstein-Barr virus
Biol Phys 2009;​73:​1326-34. 7. Sun Y, Li WF, Chen NY, et al. Induc- DNA in nasopharyngeal cancer after
3. Lai SZ, Li WF, Chen L, et al. How does tion chemotherapy plus concurrent chemo- chemoradiation to identify high-risk pa-
intensity-modulated radiotherapy versus radiotherapy versus concurrent chemo- tients for adjuvant chemotherapy: a ran-
conventional two-dimensional radiother- radiotherapy alone in locoregionally domized controlled trial. J Clin Oncol
apy influence the treatment results in advanced nasopharyngeal carcinoma: 2018 July 10 (Epub ahead of print).
nasopharyngeal carcinoma patients? Int a phase 3, multicentre, randomised con- 11. He X, Ou D, Ying H, Zhu G, Hu C, Liu T.
J Radiat Oncol Biol Phys 2011;​80:​661-8. trolled trial. Lancet Oncol 2016;​17:​1509- Experience with combination of cisplatin
4. Lee AW, Ng WT, Chan LL, et al. Evolu- 20. plus gemcitabine chemotherapy and inten-
tion of treatment for nasopharyngeal can- 8. Hong RL, Hsiao CF, Ting LL, et al. sity-modulated radiotherapy for locoregion-
cer — success and setback in the intensity- ­Final results of a randomized phase III ally advanced nasopharyngeal carcinoma.
modulated radiotherapy era. Radiother trial of induction chemotherapy followed Eur Arch Otorhinolaryngol 2012;​269:​1027-
Oncol 2014;​110:​377-84. by concurrent chemoradiotherapy versus 33.
5. Al-Sarraf M, LeBlanc M, Giri PG, et al. concurrent chemoradiotherapy alone in 12. Yau TK, Lee AW, Wong DH, et al. In-
Chemoradiotherapy versus radiotherapy patients with stage IVA and IVB nasopha- duction chemotherapy with cisplatin and
in patients with advanced nasopharyn- ryngeal carcinoma — Taiwan Cooperative gemcitabine followed by accelerated ra-

1134 n engl j med 381;12 nejm.org  September 19, 2019

The New England Journal of Medicine

Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.
 Gemcitabine and Cisplatin in Nasopharyngeal Carcinoma

diotherapy and concurrent cisplatin in cisplatin concurrent chemotherapy: long- dation of the 8th edition of the UICC/
patients with stage IV(A-B) nasopharyn- term results of a phase 2 study. Cancer AJCC Staging System for Nasopharyngeal
geal carcinoma. Head Neck 2006;​28:​880- 2011;​117:​1874-83. Carcinoma from Endemic Areas in the
7. 18. Cox JD, Stetz J, Pajak TF. Toxicity cri- Intensity-Modulated Radiotherapy Era.
13. Ngan RK, Yiu HH, Lau WH, et al. teria of the Radiation Therapy Oncology J Natl Compr Canc Netw 2017;​15:​913-9.
Combination gemcitabine and cisplatin Group (RTOG) and the European Organi- 24. Fang W, Yang Y, Ma Y, et al. Camreliz­
chemotherapy for metastatic or recurrent zation for Research and Treatment of umab (SHR-1210) alone or in combination
nasopharyngeal carcinoma: report of a Cancer (EORTC). Int J Radiat Oncol Biol with gemcitabine plus cisplatin for naso-
phase II study. Ann Oncol 2002;​13:​1252-8. Phys 1995;​31:​1341-6. pharyngeal carcinoma: results from two
14. Zhang L, Huang Y, Hong S, et al. 19. Chow S-C, Shao J, Wang H, eds. Sam- single-arm, phase 1 trials. Lancet Oncol
Gemcitabine plus cisplatin versus fluoro- ple size calculation in clinical research. 2018;​19:​1338-50.
uracil plus cisplatin in recurrent or meta- New York:​Marcel Dekker, 2003. 25. Tan T, Lim WT, Fong KW, et al. Con-
static nasopharyngeal carcinoma: a multi- 20. Pocock SJ, Clayton TC, Altman DG. current chemo-radiation with or without
centre, randomised, open-label, phase 3 Survival plots of time-to-event outcomes induction gemcitabine, carboplatin, and
trial. Lancet 2016;​388:​1883-92. in clinical trials: good practice and pit- paclitaxel: a randomized, phase 2/3 trial in
15. Edge SB, Compton CC. The American falls. Lancet 2002;​359:​1686-9. locally advanced nasopharyngeal carcino-
Joint Committee on Cancer: the 7th edi- 21. Schoenfeld D. Partial residuals for the ma. Int J Radiat Oncol Biol Phys 2015;​91:​
tion of the AJCC cancer staging manual proportional hazards regression model. 952-60.
and the future of TNM. Ann Surg Oncol Biometrika 1982;​69:​239-41. 26. Posner MR, Hershock DM, Blajman
2010;​17:​1471-4. 22. Fisher DJ, Copas AJ, Tierney JF, Par- CR, et al. Cisplatin and fluorouracil alone
16. Chan AT. Nasopharyngeal carcinoma. mar MK. A critical review of methods for or with docetaxel in head and neck can-
Ann Oncol 2010;​21:​Suppl 7:​vii308-vii312. the assessment of patient-level interac- cer. N Engl J Med 2007;​357:​1705-15.
17. Xiao WW, Huang SM, Han F, et al. Lo- tions in individual participant data meta- 27. Vermorken JB, Remenar E, van Her-
cal control, survival, and late toxicities of analysis of randomized trials, and guid- pen C, et al. Cisplatin, fluorouracil, and
locally advanced nasopharyngeal carci- ance for practitioners. J Clin Epidemiol docetaxel in unresectable head and neck
noma treated by simultaneous modulated 2011;​64:​949-67. cancer. N Engl J Med 2007;​357:​1695-704.
accelerated radiotherapy combined with 23. Tang LL, Chen YP, Mao YP, et al. Vali- Copyright © 2019 Massachusetts Medical Society.

receive immediate notification when an article

is published online first

To be notified by email when Journal articles

are published online first, sign up at NEJM.org.

n engl j med 381;12 nejm.org  September 19, 2019 1135

The New England Journal of Medicine
Downloaded from nejm.org on August 1, 2020. For personal use only. No other uses without permission.
Copyright © 2019 Massachusetts Medical Society. All rights reserved.