Keywords chemotherapy - dose - induction chemotherapy - nasopharyngeal cancer - survival -
toxicity
Introduction
Nasopharyngeal carcinoma (NPC) is one of the most common aggressive malignancies of
the head and neck, which is highly endemic in Southeast Asia, North Africa, and Southern
China. According to the Global Cancer Statistics, there were 133,354 new cases of
NPC diagnosed and 80,008 deaths occurred worldwide in 2020.[1 ] The initial presentation of NPC is atypical and unspecific, which can be manifested
as pain in the nose, ears, neck, or head. As a result, up to 80% of NPC were diagnosed
at locally advanced stage.[2 ]
Due to its anatomical location near critical structures, surgical treatment is not
the main strategy for NPC. NPC is sensitive to chemoradiation; therefore, the mainstay
of NPC treatment is the combination of radiotherapy and chemotherapy (concurrent chemoradiotherapy
[CCRT]). Several studies had reported that induction chemotherapy (IC) followed by
CCRT was superior to CCRT alone for locally advanced nasopharyngeal cancer (LA-NPC),
manifesting as higher overall survival (OS), progression-free survival (PFS), and
distant metastasis-free survival (DMFS).[3 ]
[4 ]
[5 ]
[6 ]
The combination of IC + CCRT is recommended in the latest National Comprehensive Cancer
Network Guidelines, with not only survival benefits but also early alleviation of
symptoms, better radiotherapy compliance with reduced targets, and elimination of
small metastatic lesions which is the main failure pattern of LA-NPC.[7 ]
[8 ]
TPF is one of the IC regimens which consists of docetaxel, cisplatin, and 5-fluorouracil.
A randomized controlled trial (RCT) by Sun et al reported that TPF IC followed by
CCRT significantly improved OS, DMFS, and failure-free survival compared with CCRT
alone.[9 ] Previous retrospective studies in Chinese patients reported low-dose TPF (L-TPF),
consists of docetaxel (60 mg/m2 ), cisplatin (65 mg/m2 ), and then 5-fluorouracil (550 mg/m2 /d; 5 days), showed better tolerance and compliance rates, and similar efficacy to
high-dose TPF (H-TPF) (docetaxel [75 mg/m2 ; 1 hour infusion], cisplatin [75 mg/m2 ; 0.5–3 hours], and then 5-fluorouracil [600 mg/m2 /d; 4 days]). NPC patients receiving H-TPF had more frequent treatment delays due
to chemotoxicity, such as grades 3 and 4 anemia, thrombocytopenia, and neutropenia.[10 ]
The optimal dose of TPF in IC + CCRT for NPC patients is still unclear. Thus, we aim
to evaluate the effect of L-TPF + CCRT and H-TPF + CCRT on survival and chemotoxicity
in LA-NPC patients.
Materials and Methods
Search Strategy
This systematic review has been registered in PROSPERO (international database of
prospectively registered systematic reviews) (CRD42023483635). A literature search
for this systematic review was conducted in online resources, including PubMed, PubMed
Central (PMC), and Science Direct. The literature search was done from November 13–16,
2023, using keywords related to NPC, TPF regimen, survival, and chemotoxicity (listed
in [Table 1 ]). Systematic analysis approaches were used in this study, including PICO analysis
(P: LA-NPC patients of any age; I: L-TPF regimen; C: H-TPF regimen; O: OS, PFS, and
chemotoxicity).
Table 1
Literature search strategy
Database
Keywords
Results
PubMed
(“TPF”[All Fields] AND “dose”[All Fields] AND (“nasopharyngeal carcinoma”[MeSH Terms]
OR (“nasopharyngeal”[All Fields] AND “carcinoma”[All Fields]) OR “nasopharyngeal carcinoma”[All
Fields])) AND (2013:2023[pdat])
12
PubMed Central
(“TPF”[All Fields] AND “dose”[All Fields] AND (“nasopharyngeal carcinoma”[MeSH Terms]
OR (“nasopharyngeal”[All Fields] AND “carcinoma”[All Fields]) OR “nasopharyngeal carcinoma”[All
Fields])) AND (2013:2023[pdat])
462
Science Direct
TPF” AND “dose” AND “nasopharyngeal carcinoma”*Filter: 10 years
166
The studies that we included are cohort studies or RCTs with participants of any age,
who were diagnosed with LA-NPC. The participants were treated with systemic chemotherapeutic
agents with or without radiotherapy. The studies must report data of survival, including
OS and PFS, and chemotherapy toxicity. The studies must be in English language and
published by the end of October 2023. Unpublished articles, abstracts or lectures,
dissertations, books and book chapters, editorials, online articles, letters to the
editor or opinion pieces, poster or conference presentations, case reports, case series,
meta-analyses, animal studies, cross-sectional, case–control studies, systematic reviews,
and literature reviews were excluded. We also excluded studies with reported opinions
or outcomes through nonstandardized questionnaires or studies with other main outcomes,
such as psychological outcomes. Studies that include patients with pregnancy or lactating
females or other types of cancer as primary exposure were also excluded. Studies that
include patients with cardiac arrhythmia, coronary heart disease, peripheral neuropathy,
or any psychiatric disorders that may affect treatment compliance were also excluded.
Data selection, extraction, and analysis were conducted by two independent reviewers.
Any dissimilarities were resolved through discussing with the third author and reaching
a general agreement between the reviewers. The reviewers evaluated the titles and
abstracts for all selected studies using the PRISMA search strategy. Relevant titles
and abstracts from each database were chosen. If there is limited information for
the reviewers to determine the inclusion and exclusion criteria, the full texts will
be evaluated. References that are found in the included and excluded articles were
reviewed to discover studies that failed to be included through the primary search.
Data Extraction
Data regarding authors' name; year of publication; study design; country of the study;
inclusion and exclusion criteria; mean age of study population; population size and
NPC stage; TPF regimen dose; and main outcomes (OS, PFS, and chemotoxicity) were extracted
by at least two reviewers.
Collective article reviews were performed on three databases. Full-text articles of
the appropriate studies were read and assessed. Quality assessment is conducted according
to the Newcastle–Ottawa scale (NOS) for cohort studies and the JADAD scale for RCT
with minimal two reviewers.[11 ] The study selection process is illustrated in a flow diagram ([Fig. 1 ]).[12 ]
Fig. 1 PRISMA diagram of the detailed process of study selection to be included in the review.
Outcome Definitions
Primary outcome includes OS, PFS, and chemotoxicity. OS is defined as the length of
time from either the date of diagnosis or the start of treatment for a disease, such
as cancer, that patients diagnosed with the disease are still alive.[13 ] PFS is defined as the length of time during and after the treatment of a disease,
such as cancer, that a patient lives with the disease, but it does not get worse.[14 ] Chemotoxicity is toxicity due to chemical effects, especially the effects of chemotherapy.[15 ]
Quality Assessment
The quality of the included cohort studies was assessed using the NOS. The three domains
assessed included: (1) selection of study groups (four points), (2) comparability
of the groups (two points), and (3) assessment of the outcome (three points). A score
of 5 or less was considered poor, 6 or 7 was considered moderate, and 8 or 9 was considered
good quality.[11 ] JADAD scale was used for assessing RCTs. The JADAD scale is a five-point scale for
evaluating the quality of randomized trials in which three points or more indicate
superior quality. The scale contains two questions each for randomization and masking,
and one question evaluating reporting of withdrawals and dropouts. Studies with poor
quality will be excluded.[16 ]
Two independent researchers assessed the quality of methods and standard of outcome
reporting in the included studies. Disagreements between the researchers were resolved
through consensus or the opinion of a third reviewer. The quality of evidence assessment
using the GRADE (Cochrane Group) analysis of findings was not done.
Statistical Analysis
RevMan version 5.4 software (Cochrane Collaboration) was used to conduct the meta-analysis.
Hazard ratio (HR) and its 95% confidence interval (CI) were used to evaluate OS and
PFS. We estimated the HR and 95% CI based on the method by Tierney et al if there
is no available data on HRs and 95% CI from the study.[17 ] Inconsistency index (I
2 ) test, ranging from 0 to 100%, was used to assess heterogeneity of the studies. If
the value of I
2 >50% or p < 0.10, it indicates that the heterogeneity is statistically significant, and therefore,
random-effect model was used.[18 ] All p -values in this meta-analysis were two-tailed, and the statistical significance was
set at ≤0.05. Subgroup analyses were performed by classifying studies according to
regions. If meta-analysis could not be conducted, study results were presented descriptively.
Results
Literature Search
The initial search through online databases identified a total of 657 studies, with
12 studies from PubMed, 462 studies from PMC, 166 studies from Science Direct, and
17 studies identified through literature or other sources. The search process was
followed by the screening of titles and abstracts, and the remaining studies were
further assessed for eligibility. After a series of selections according to inclusion
criteria and exclusion of duplicated studies, we acquired six studies for this systematic
review and meta-analysis. The selection process is shown in [Fig. 1 ].
Quality Assessment
The result of the quality assessment is presented in [Table 2 ]. According to the NOS quality assessment, there was one study with good quality
and three studies with moderate quality. Two RCT studies were assessed using the JADAD
scale and both were high quality. There was no study with poor quality; therefore,
all the studies were included in this review.
Table 2
Newcastle–Ottawa quality assessment of cohort studies
Author (year), country
Study design
Selection
Comparability
Exposure/outcome
Total score
Result
Sun et al (2016), China
RCT
★★★
★
★★★
7
Moderate
Ou et al (2016), China
Retrospective cohort
★★★
★
★★★
7
Moderate
Kawahira et al (2017), Japan
Retrospective cohort
★★★
★
★★
7
Moderate
Frikha et al (2018), France and Tunisia
RCT
★★★
★
★★★
7
Moderate
Zhu et al (2019), China
Retrospective cohort
★★★
★
★★★
7
Moderate
Mohamad et al (2022), Jordan
Retrospective cohort
★★★★
★
★★★
8
Good
Abbreviation: RCT, randomized controlled trial.
Star symbol Indicates points.
Characteristics of the Included Studies
The characteristics of the included studies are presented in [Table 3 ]. A total of six studies were included in this quantitative analysis. There were
two RCTs and four cohort studies. The studies were published between the year 2016
and 2022. The countries of the studies are China, Japan, Jordan, France, and Tunisia.
All the included studies had evaluated OS and PFS.
Table 3
Summary of studies on OS, PFS, and chemotoxicity of NPC patients
Author (year)
Country
Type of study
TNM stage
Population (median age)
NAC TPF dose and cycles
Median follow-up time (mo)
Outcome
Sun et al (2016)
China
RCT
Stages III–IVb
241 (42)
L-TPF: docetaxel 60 mg/m2 on day 1, cisplatin 60 mg/m2 on day 1, and 5-fluorouracil 600 mg/m2 per day as a continuous 120-h infusion on days 1–5
45
OS: HR 0.59 (95% CI, 0.36–0.96)
PFS: HR 0.68 (95% CI, 0.48–0.97)
The most common grade 3 or 4 adverse events during treatment in the NAC + CCRT group
vs. CCRT alone group were neutropenia (101 [42%] vs. 17 [7%]), leukopenia (98 [41%]
vs. 41 [17%]), and stomatitis (98 [41%] vs. 84 [35%])
Kawahira et al (2017)
Japan
Retrospective cohort
Stages III–IVb
12 (61)
L-TPF: docetaxel (60–70 mg/m2 ) and cisplatin (60–70 mg/m2 ) on day 1, and a continuous infusion of 5-fluorouracil (750 mg/m2 /d) on days 1–5
36.4
OS: HR 0.32 (95% CI, 0.01–11.47)
PFS: HR 0.97 (95% CI 0.08–12.19)
The common grades 3–4 acute toxicities were anemia, anorexia, febrile neutropenia,
and stomatitis in the TPF group
Zhu et al (2019)
China
Retrospective Cohort
Stages III–IV
87 (45)
L-TPF: docetaxel (60 mg/m2 ) on day 1, cisplatin (20–25 mg/m2 ) per day on days 1–3, and fluorouracil (600 mg/m2 ) per day as a continuous 120-hour infusion
40
OS: HR 0.32 (95% CI, 0.14–0.75)
PFS: HR 0.18 (95% CI, 0.06–0.59)
Treatment failure was 26/87 (29.9%) in the TPF group
The most common grades 3–4 adverse effect was leukopenia, which occurred significantly
more often in the TPF group (25/87 [28.7%], p = 0.007)
There was also significantly more neutropenia in the TPF group (18/87 [20.7%], p = 0.033)
Ou et al (2016)
China
Retrospective cohort
Stages II–IV
58 (49)
H-TPF: docetaxel (75 mg/m2 ) on day 1, cisplatin (75 mg/m2 ) on day 1, and 5-fluorouracil (750 mg/m2 /d) on days 1–5
78
OS: HR 0.40 (95% CI, 0.06–2.75)
PFS: HR 0.56 (95% CI, 0.13–2.38)
There was less total cumulative incidence of grades 3–4 toxicities in the NAC + CCRT
group (44.8 vs. 70.8%, p = 0.01)
Frikha et al (2018)
France and Tunisia
RCT
Stages T2b, T3, T4 and/or N1-N3, M0
40 (46)
H-TPF: docetaxel (75 mg/m2 ) on day 1, cisplatin (75 mg/m2 ) on day 1, and 5-fluorouracil 750 mg/m2 /d on days 1–5
43.1
OS: HR 0.40 (95% CI, 0.15–1.05)
PFS: HR 0.44 (95% CI, 0.20–0.97)
At least one toxicity grade 3 or 4 was seen in 20 patients (50%), especially neutropenia
(27.5%), febrile neutropenia (7.5%), mucositis (12.5%), alopecia (15%), and asthenia
(10%)
Mohamad et al (2022)
Jordan
Retrospective cohort
Stages II–IVa
65 (37)
H-TPF: docetaxel (75 mg/m2 ) and cisplatin (100 mg/m2 ) on day 1 and 5-fluorouracil (1,000 mg/m2 ) on days 1–5
51
OS: HR 1.04 (95% CI, 0.22–4.86)
PFS: HR 2.25 (95% CI, 0.80–6.33)
TPF group had more frequent acute grade (G) II anemia (13, p < 0.01), late G II brain toxicity (4, p < 0.01), and late G II dysphagia (32, p < 0.01)
Abbreviations: CCRT, concurrent chemoradiotherapy; CI, confidence interval; HR, hazard
ratio; H-TPF, high-TPF; L-TPF, low-TPF; NAC, neoadjuvant chemotherapy; OS, overall
survival; PFS, progression-free survival; RCT, randomized controlled trial; TNM, Tumor
Node Metastasis.
Characteristics of the Studied Population
The studies consisted of 503 nasopharyngeal cancer patients, with TNM staging of II
to IVb. All patients underwent IC with a TPF regimen plus CCRT. Three of the studies
used L-TPF with a similar range of dose regimens. Dose regimens of L-TPF by Sun et
al were docetaxel (60 mg/m2 ) on day 1, cisplatin (60 mg/m2 ) on day 1, and 5-fluorouracil (600 mg/m2 ) per day as a continuous 120-hour infusion on days 1 to 5.[9 ] Dose regimens by Kawahira et al were docetaxel (60–70 mg/m2 ) on day 1, cisplatin (60–70 mg/m2 ) on day 1, and 5-fluorouracil (750 mg/m2 ) per day as a continuous 120-hour infusion on days 1 to 5.[19 ] Meanwhile, Zhu et al used docetaxel (60 mg/m2 ) on day 1, cisplatin (20–25 mg/m2 ) on day 1, and 5-fluorouracil (600 mg/m2 ) per day as a continuous 120-hour infusion on days 1 to 5.[20 ] Three other studies used H-TPF which consisted of docetaxel (75 mg/m2 ) on day 1, cisplatin (75 mg/m2 ) on day 1, and 5-fluorouracil (750–1,000 mg/m2 /d) on days 1 to 5.[21 ]
[22 ]
[23 ]
Primary Outcomes
Overall Survival
A total of six studies evaluated OS in NPC patients receiving L-TPF or H-TPF.[9 ]
[19 ]
[20 ]
[21 ]
[22 ]
[23 ] The population of patients receiving L-TPF and H-TPF were 340 and 163 patients,
respectively. The synthesized result of OS is presented in [Fig. 2 ]. There was no significant heterogeneity found (I
2 = 0% and p > 0.10); therefore, the fixed-effect model was used. We found that L-TPF was significantly
associated with improved OS, with a pooled HR of 0.50 (95% CI, 0.33–0.76; p = 0.001). H-TPF was not significantly associated with improved OS in NPC patients
(p = 0.08).
Fig. 2 Forest plot of the association between L-TPF or H-TPF + CCRT and overall survival
in NPC patients. CCRT, concurrent chemoradiotherapy; CI, confidence interval; H-TPF,
high-dose TPF; L-TPF, low-dose TPF; NPC, nasopharyngeal cancer; SE, standard error.
Progression-Free Survival
All the included studies evaluated PFS in NPC patients receiving L-TPF or H-TPF.[9 ]
[19 ]
[20 ]
[21 ]
[22 ]
[23 ] The synthesized result of PFS is presented in [Fig. 3 ]. Random-effect model was used due to the presence of significant heterogeneity found
among studies (I
2 >50% and p < 0.10). We found that both L-TPF and H-TPF were not associated with improved PFS
in NPC patients (p = 0.13 and p = 0.72, respectively).
Fig. 3 Forest plot of the association between L-TPF or H-TPF + CCRT and progression-free
survival in NPC patients. CCRT, concurrent chemoradiotherapy; CI, confidence interval;
H-TPF, high-dose TPF; L-TPF, low-dose TPF; NPC, nasopharyngeal cancer; SE, standard
error.
Chemotherapy Toxicity
Chemotherapy toxicity was reported in all recruited studies. However, due to limited
data available from the studies, analysis was not done. Studies with L-TPF had reported
similar incidence of grades 3 and 4 chemotoxicity. Among these, neutropenia, febrile
neutropenia, leukopenia, stomatitis, and anorexia were the most prevalent.[9 ]
[19 ]
[20 ] Conversely, H-TPF + CCRT, as studied by Ou et al, demonstrated less total cumulative
incidence of grades 3 and 4 toxicities than CCRT only.[21 ] However, studies by Frikha et al and Mohamad et al had contrasting results. Frikha
et al reported a similar rate of grades 3 and 4 toxicity in the H-TPF + CCRT group
and CCRT-only group, mainly neutropenia, febrile neutropenia, mucositis, alopecia,
and asthenia.[22 ] Study by Mohamad et al reported more frequent grade 2 anemia, late grade 2 brain
toxicity, and late grade 2 dysphagia in the H-TPF group.[23 ]
Discussion
This systematic review examined survival and chemotherapy toxicity in NPC patients
undergoing different TPF dose regimens. The results indicated that both L-TPF and
H-TPF were not associated with improved PFS in NPC patients. Conversely, L-TPF significantly
improved OS in NPC patients. However, H-TPF did not show significant improvement in
terms of OS. Patients in the L-TPF group had better tolerance and compliance rates
than H-TPF, suggesting L-TPF may be a more effective treatment option for improving
survival outcomes. This may be because of the increased toxicity in H-TPF regimens
that delayed further chemotherapy schedule and this may impact the survival. However,
further investigation into factors contributing to the reduced efficacy of H-TPF compared
with L-TPF may be warranted.
In 2017, Jin et al compared the efficacy and toxicity between L-TPF and H-TPF in NPC
patients. The study reported that L-TPF was superior to H-TPF in terms of survival
and chemotoxicities. The result could be attributed to multiple factors. Treatment
delays were more common in the H-TPF group than in the L-TPF group (33.3 vs. 19.4%,
p = 0.034), which could counteract any possible survival benefits of H-TPF by allowing
tumor cell proliferation.[10 ] Previous research by Lee et al and Loong et al showed that OS and locoregional control
were significantly impacted by the total dose of cisplatin given during CCRT.[24 ]
[25 ] According to Jin et al, there were fewer patients in the H-TPF group that finished
two cycles of concurrent cisplatin (p < 0.001). This was due to hematologic and nonhematologic adverse events in the H-TPF
group that restrict the completion of treatment.[10 ]
Chemotoxicity and adverse events were the major problems causing incomplete treatment
cycles, particularly in H-TPF patients. However, our included studies had inconsistent
results. The included studies demonstrated that both hematological toxicities, such
as anemia and leukopenia, and nonhematological toxicities, such as stomatitis, mucositis,
and anorexia, were common in L-TPF patients. One study had reported that H-TPF was
associated with less total cumulative incidence of grades 3 and 4 adverse events.
Another study had shown a similar rate of grades 3 and 4 toxicity in the H-TPF + CCRT
group and CCRT-only group. Meanwhile, Mohamad et al reported more frequent adverse
events, such as grade 2 anemia, late grade 2 brain toxicity, and late grade 2 dysphagia
in the H-TPF group. Further investigations regarding the chemotherapy toxicity effect
of different TPF dose regimens are required.
There were several limitations in this analysis. The limited number of samples and
included studies may cause a limitation of the power of pooled results for survival
analysis. There were also differences in the dose regimen of cisplatin and 5-fluorouracil
in both L-TPF and H-TPF groups which may cause bias in the result. Analysis of chemotoxicity
was not done due to limited data from the included studies.
One of the strengths of our study is that all studies with poor quality were excluded
from this review. Thus, all the included studies were of moderate and good qualities.
Our review focused on LA-NPC patients only; therefore, it is more specific.
Understanding the literature support for these findings further strengthens the implications
for clinical practice and emphasizes the need for continued research to elucidate
the factors underlying the differential outcomes between L-TPF and H-TPF in NPC patients.
By building upon the existing evidence, future studies can contribute to a more comprehensive
understanding of the efficacy and mechanisms of these treatment regimens.
Implications for Clinical Practice
The findings that support the superiority of L-TPF over H-TPF in terms of survival
in NPC patients have significant implications for clinical practice. The consideration
of the mechanistic dissimilarities, toxicity profiles, and patient-specific factors
can guide oncologists in making informed decisions regarding the selection of treatment
regimens for NPC.
Beyond the comparison of L-TPF and H-TPF, the challenges in treatment selection and
patient care in NPC are multifaceted. The complexity of NPC, which may involve tumor
staging, histological subtypes, and individual variations in disease progression,
necessitates a patient-specific approach to treatment decision-making. Factors such
as the presence of distant metastases, involvement of critical structures in the head
and neck region, and the potential for chemotherapy toxicities further emphasize the
need for personalized care pathways for NPC patients.
The management of NPC often involves a multimodal treatment approach, which may include
a combination of chemotherapy, radiotherapy, and in some cases, targeted therapies,
or immunotherapies. Understanding the interplay between these modalities and their
potential additive or synergistic effects is crucial for optimizing treatment outcomes.
Additionally, the sequencing of these modalities and their timing in relation to disease
progression represents a dynamic area of research and clinical decision-making.
Conclusion
The comparison of L-TPF and H-TPF in NPC patients has revealed important considerations
for clinical practice and patient care. The superior OS and better tolerance associated
with L-TPF exhibit its potential as a preferred treatment regimen for NPC patients.
Further, larger multicenter RCT studies are needed to focus on evaluating the optimal
TPF regimen dose in LA-NPC. However, the diverse nature of NPC and the complexities
involved in treatment selection and patient care necessitate a tailored and multidisciplinary
approach. The insights gained from this analysis highlight the need for personalized
treatment strategies that consider not only the efficacy of a given regimen but also
its tolerability, impact on patient quality of life, and potential toxicities.
