Four cycles of docetaxel plus cisplatin as neoadjuvant chemotherapy followed by concurrent chemoradiotherapy in stage N2-3 nasopharyngeal carcinoma: phase 3 multicentre randomised controlled trial.
Xie WH, Xiao WW, Chang H, Xu MJ, Hu YH, Zhou TC, Zhong Q, Chen CY, Lu LX, Wang QX, Zhu YJ, Yang J, Shi XY, Kang HL, Wei JW, Huang R, Peng HH, Yuan Y, Wu SH, Jiang XH, Liu YJ, Wen BX, Gao YH
- DOI
- 10.1136/bmj-2024-081557
- Record issued
- 2026-08-15
- Engine
- 7.39.0
- Exported
- 2026-09-21
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/a922fcf3-fb1c-4671-b043-24ae8ea1a0cf is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×3−1.5★
- ReportingData & code availability partially met−0.25★
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
This is a well-conducted and well-reported phase 3 randomised controlled trial with a clear scientific premise, rigorous design, and transparent reporting. The main weakness is the vague data availability statement and lack of code sharing, which limits reproducibility.
Both reviewers agreed on all dimensions and study type (interventional). The statistics verification component recomputed 4 reported tests consistently, but coverage is limited to tests with test statistics/CI; other tests were not machine-verified. The citation check found no retracted or missing references. The integrity check raised only low-severity observations about toxicity percentages and CONSORT flow, which are not validity threats.
Numerical inconsistencies
1 finding · worst lowValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
- Internal contradictions in the reported numbersAssessed
Recomputed 4 tests: 4 consistent, 0 inconsistent; 2 recomputed directly from the reported test statistics, 2 via agent-written checks.
- CONSISTENTreported p = .020 · recomputed p = .022Recomputed hazard ratio 0.41 (95% CI 0.19–0.87), reported p=0.02
“hazard ratio 0.41 (95% CI 0.19 to 0.87); P=0.02”
Taken as given: 0.19–0.87 is a two-sided 95% confidence interval for the hazard ratio of 0.41, not a range, an IQR, or a different interval level; the hazard ratio is a RATIO measure, so the interval is symmetric on the log scale; p=0.02 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(0.41, 0.19, 0.87, 1) - CONSISTENTreported p = .020 · recomputed p = .025Recomputed hazard ratio 0.41 (95% CI 0.19–0.90), reported p=0.02
“hazard ratio 0.41 (95% CI 0.19 to 0.90); P=0.02”
Taken as given: 0.19–0.90 is a two-sided 95% confidence interval for the hazard ratio of 0.41, not a range, an IQR, or a different interval level; the hazard ratio is a RATIO measure, so the interval is symmetric on the log scale; p=0.02 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(0.41, 0.19, 0.9, 1) - CONSISTENTreported p = .010 · recomputed p = .012Reviewers 1, 2Hazard ratio for overall survival (0.38) with 95% CI (0.18-0.82) yields p=0.01.
“hazard ratio 0.38 (0.18 to 0.82); P=0.01”
Taken as given: The hazard ratio is 0.38.; The 95% confidence interval is 0.18 to 0.82.; The CI is two-sided at 95%.; The log hazard ratio is used.Method: Recomputed p-value from the hazard ratio and its 95% CI using the normal approximation for the log hazard ratio.How we recomputed it: pCI(0.38, 0.18, 0.82, 1) - CONSISTENTreported p = .005 · recomputed p = .007Reviewers 1, 2Hazard ratio for disease-free survival (0.46) with 95% CI (0.26-0.80) yields p=0.005.
“hazard ratio 0.46 (0.26 to 0.80); P=0.005”
Taken as given: The hazard ratio is 0.46.; The 95% confidence interval is 0.26 to 0.80.; The CI is two-sided at 95%.; The log hazard ratio is used.Method: Recomputed p-value from the hazard ratio and its 95% CI using the normal approximation for the log hazard ratio.How we recomputed it: pCI(0.46, 0.26, 0.80, 1)
- lowinternal contradictionThe text states 'The incidence of grade 3/4 toxicity during neoadjuvant chemotherapy was 38%' but then lists 27 (29%) leukopenia, 34 (37%) neutropenia, and 8 (9%) gastrointestinal toxicities, which sum to more than 38% if overlapping. This is likely due to overlapping toxicities, but the 38% is not clearly defined.
“The incidence of grade 3/4 toxicity during neoadjuvant chemotherapy was 38%, which included 27 (29%) grade 3/4 leukopenia, 34 (37%) grade 3/4 neutropenia, and eight (9%) grade 3 gastrointestinal toxicities.”
ResultsFind in source - lowinternal contradictionThe abstract states 'Grade 3/4 acute toxicities were observed in 60 (65%) and 46 (51%) patients' but the results section says 'We observed 60 (65%) and 46 (51%) grade 3/4 toxicities' - consistent. No contradiction found.
“Grade 3/4 acute toxicities were observed in 60 (65%) and 46 (51%) patients”
AbstractFind in source - lowinternal contradictionThe abstract reports 186 participants randomized, but the results section states 192 enrolled and 6 excluded, which is consistent. However, the CONSORT flow is not fully detailed in the text.
“We enrolled 192 patients with stage T1-4N2-3M0 nasopharyngeal carcinoma from 23 February 2016 to 18 February 2019. After exclusion of six patients who withdrew consent before randomisation, 186 patients were randomly assigned”
ResultsFind in source
Overstated conclusions
1 finding · worst lowConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
- Conclusions only partially backed by the presented evidenceAssessed
4 major claims checked against the paper's own evidence: 2 only partially supported (evidence backs part of the claim; gaps or caveats remain); the rest adequately supported.
- partialReviewers 1, 2The docetaxel plus cisplatin regimen is effective, safe, and economical for the treatment of locoregionally advanced nasopharyngeal carcinoma.Effectiveness and safety are supported by the trial results, but 'economical' is not directly assessed in the study.Evidence: Survival benefits and manageable toxicities are reported; no cost-effectiveness analysis is presented.
The docetaxel plus cisplatin neoadjuvant chemotherapy regimen is effective, safe, and economical for the treatment of locoregionally advanced nasopharyngeal carcinoma.
What this study addsreviewer’s wording - partialReviewers 1, 2The risk of distant metastasis in stage N2-3 nasopharyngeal carcinoma was primarily caused by preliminary micrometastasis and could be controlled by increasing neoadjuvant chemotherapy cycles.The improved DMFS supports the hypothesis, but the claim of causation is based on inference, not direct evidence of micrometastasis.Evidence: Improved DMFS in the neoadjuvant group; no direct detection of micrometastasis.
“Secondly, the risk of distant metastasis in stage N2-3 nasopharyngeal carcinoma was primarily caused by the preliminary micrometastasis and could be controlled by increasing neoadjuvant chemotherapy cycles, evidenced by the 13.1% (91.3% v 78.2%; P=0.02) improvement in five year distant metastasis-free survival.”
Discussion ¶1Find in source - supportedReviewers 1, 2Four cycles of docetaxel plus cisplatin neoadjuvant chemotherapy with concurrent chemoradiotherapy can effectively reduce distant metastasis and improve survival for patients with stage N2-3 nasopharyngeal carcinoma.The primary endpoints (distant metastasis-free survival and overall survival) were significantly improved in the neoadjuvant group, supporting the claim.Evidence: Five-year DMFS 91.3% vs 78.2%, HR 0.41, P=0.02; five-year OS 90.3% vs 82.6%, HR 0.38, P=0.01.
“Four cycles of docetaxel plus cisplatin neoadjuvant chemotherapy with concurrent chemoradiotherapy can effectively reduce distant metastasis and improve survival for patients with stage N2-3 nasopharyngeal carcinoma with manageable toxicities.”
ConclusionFind in source - supportedReviewers 1, 2Four cycles of neoadjuvant chemotherapy reduce the risks of distant metastasis and prolong survival for patients with stage N2-3 nasopharyngeal carcinoma.The trial directly shows reduced distant metastasis and improved survival in the four-cycle group.Evidence: Significant improvements in DMFS and OS with HRs <1 and p<0.05.
Four cycles of neoadjuvant chemotherapy reduce the risks of distant metastasis and prolong survival for patients with stage N2-3 nasopharyngeal carcinoma.
What this study addsreviewer’s wording
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary endpoints are hard clinical outcomes: five-year distant metastasis-free survival and overall survival, both measured directly from clinical events (distant metastasis and death). No surrogate biomarker is used as the basis for the efficacy claim.
“The primary endpoints were five year distant metastasis-free survival and overall survival, which were centrally assessed.”
- ADEQUATEEffect sizeThe effect sizes are clinically meaningful: five-year distant metastasis-free survival improved from 78.2% to 91.3% (absolute increase 13.1%) and overall survival from 82.6% to 90.3% (absolute increase 7.7%), with hazard ratios of 0.41 and 0.38, respectively, and statistically significant p-values. These are large, clinically material improvements in survival outcomes.
“the neoadjuvant chemotherapy plus concurrent chemoradiotherapy group had superior five year distant metastasis-free survival (91.3% ... versus 78.2% ...; hazard ratio 0.41 ...; P=0.02) and five year overall survival (90.3% ... versus 82.6% ...; hazard ratio 0.38 ...; P=0.01)”
Data authenticity concerns
None foundAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
Checked — nothing surfaced.
Reporting gaps
1 finding · worst mediumRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
- Data/code availability incompleteAssessed
The introduction cites multiple prior studies (Hui et al., Jin et al., and the authors' own 924-case analysis) and explains the rationale for using four cycles of docetaxel plus cisplatin. Limitations of prior research are implicitly addressed by noting the lack of consensus on neoadjuvant regimens and the need for a phase 3 trial. The hypothesis follows logically from the cited evidence.
“Hui and colleagues did a phase 2 trial comparing two cycles of docetaxel plus cisplatin neoadjuvant chemotherapy plus cisplatin based concurrent chemoradiotherapy with concurrent chemoradiotherapy alone in 2009 and showed a favourable impact on overall survival.”
“To further verify our hypothesis, we did a phase 3 multicentre randomised controlled trial.”
“However, before the results of the aforementioned randomised controlled trials were published, a consensus on neoadjuvant chemotherapy regimens was lacking.”
“Hui and colleagues did a phase 2 trial comparing two cycles of docetaxel plus cisplatin neoadjuvant chemotherapy plus cisplatin based concurrent chemoradiotherapy with concurrent chemoradiotherapy alone in 2009 and showed a favourable impact on overall survival.”
“To further verify our hypothesis, we did a phase 3 multicentre randomised controlled trial.”
“However, before the results of the aforementioned randomised controlled trials were published, a consensus on neoadjuvant chemotherapy regimens was lacking.”
Randomization method is described (computed random number code, sealed opaque envelopes, stratified by N stage). Blinding is described as open-label with blinded outcome assessment. Power analysis is reported with effect size assumptions and sample size calculation. Inclusion/exclusion criteria are pre-specified. Outlier handling is addressed through the definition of intention-to-treat and per-protocol populations. Controls are inherent in the concurrent chemoradiotherapy alone group. Independent replication is not applicable for a single pivotal trial.
“Random numbers were generated using a computed random number code and sealed in opaque envelopes marked with two colours to distinguish stage N2 (white) and N3 (brown) stratifications.”
“However, we applied blinding during the outcome assessment.”
“We used Power Analysis and Sample Size 11 software to analyse sample size, with a power of 80% and a two sided significance level of 0.05.”
“Random numbers were generated using a computed random number code and sealed in opaque envelopes marked with two colours to distinguish stage N2 (white) and N3 (brown) stratifications.”
“However, we applied blinding during the outcome assessment.”
“We used Power Analysis and Sample Size 11 software to analyse sample size, with a power of 80% and a two sided significance level of 0.05.”
Sex is reported in Table 1 (69 (74%) male in NACT+CCRT group, 67 (72%) in CCRT group). Age is reported as median (IQR). Health status is indicated by Karnofsky score. Demographics include sex, age, and clinical stage. Since both sexes are enrolled, sex_justified is not applicable. Species/strain and housing conditions are not applicable for a human trial.
“Male sex | 69 (74) | 67 (72)”
“Median (IQR) age, years | 48 (23-68) | 44 (24-69)”
“Male sex | 69 (74) | 67 (72)”
“Median (IQR) age, years | 48 (23-68) | 44 (24-69)”
The trial was approved by the ethics committee of Sun Yat-sen University Cancer Centre and each participating centre, with a protocol ID. Written informed consent was obtained from all patients. Regulatory compliance is implied by adherence to ethical standards, though not explicitly named; however, the approval statement is adequate.
“This trial was approved by the ethics committee of the Sun Yat-sen University Cancer Centre and each participating centre (ID: 5010-2015-03).”
“All patients gave written informed consent before enrolment.”
“This trial was approved by the ethics committee of the Sun Yat-sen University Cancer Centre and each participating centre (ID: 5010-2015-03).”
“All patients gave written informed consent before enrolment.”
The investigational products are named with doses and regimen (docetaxel 75 mg/m2, cisplatin 37.5 mg/m2). Statistical software (SPSS, R) and sample size software (PASS 11) are identified. No antibodies, cell lines, or mycoplasma testing are applicable. Reagents are not applicable beyond the drugs.
“The neoadjuvant chemotherapy regimen included docetaxel (75 mg/m 2 on day 1) and cisplatin (37.5 mg/m 2 on days 2-3).”
“We used SPSS Statistics software (version 22.0), and R (version 4.1.3) to analyse data collected in this trial.”
“The neoadjuvant chemotherapy regimen included docetaxel (75 mg/m 2 on day 1) and cisplatin (37.5 mg/m 2 on days 2-3).”
“We used SPSS Statistics software (version 22.0), and R (version 4.1.3) to analyse data collected in this trial.”
Tests named include log-rank test, Cox proportional hazards model, Wilcoxon rank-sum, chi-square, and t-tests. Exact p-values are reported (e.g., P=0.02). Effect sizes are reported as hazard ratios with 95% CIs. Software is identified. Data presentation includes Kaplan-Meier curves and per-group n. Mathematical plausibility is not applicable for large-N continuous outcomes.
“We described time-to-event data by using Kaplan-Meier curves and compared them with the log-rank test.”
“hazard ratio 0.41 (95% CI 0.19 to 0.87); P=0.02”
“hazard ratio 0.41 (95% CI 0.19 to 0.87)”
“We described time-to-event data by using Kaplan-Meier curves and compared them with the log-rank test.”
“hazard ratio 0.41 (95% CI 0.19 to 0.87); P=0.02”
“five year distant metastasis-free survival (91.3% (95% confidence interval (CI) 85.4% to 97.2%) versus 78.2% (69.8% to 86.6%); hazard ratio 0.41 (95% CI 0.19 to 0.87); P=0.02)”
The data availability statement says data can be requested from the corresponding author and will be deposited on a public platform, but the mechanism is not fully specified (no timeframe or conditions). No code sharing is mentioned, but the trial may not have bespoke code. Repository deposit is mentioned but not yet available.
Trial registration number is provided (NCT02512315). Methods are detailed enough for replication. Limitations are explicitly discussed. Conclusions are proportional to the evidence. Funding and competing interests are declared. Reporting guideline adherence is not explicitly mentioned, but the paper follows CONSORT-like structure.
“Trial registration ClinicalTrials.gov NCT02512315 .”
“This study had three main limitations.”
“Funding: This work was supported by grants from the Science and Technology Planning Project of Guangdong Province, China (grant number 2017A020215157).”
“Trial registration ClinicalTrials.gov NCT02512315 .”
“This study had three main limitations.”
“Funding: This work was supported by grants from the Science and Technology Planning Project of Guangdong Province, China (grant number 2017A020215157).”
Registered (2 IDs: ClinicalTrials.gov). No reporting guideline cited.
Broken references and links
None foundReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
Checked — nothing surfaced.
Checked 30 references by DOI: 30 verified.
Every extracted reference resolved against Crossref/OpenAlex with no retraction flags.
1 data/code link found; not probed for liveness in this run.
- datahttp://www.researchdata.org.cnUNVERIFIEDLiveness indeterminate — content not checked.
Copyediting
4 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 4 minor suggestions below.
4 copyedit issues flagged: mostly consistency, typo, clarity.
- MINORtypoMethods, Sample size and statistical methods“grouup”→ groupTypographical error.
- MINORconsistencyDiscussion, Comparison with other studies“fluororacil”→ fluorouracilInconsistent spelling of fluorouracil.
- MINORconsistencyDiscussion, Comparison with other studies“neoadvuvant”→ neoadjuvantTypographical error.
- MINORclarityData availability statement“The raw database will be deposited on the Research Data Deposit public platform ( www.researchdata.org.cn ).”→ Provide the full URL and specify the expected deposit date.The URL is present but the statement could be more specific.
The published work is robust and well-reported, with only minor reporting gaps. An informed reader should weigh the vague data availability statement and lack of code sharing as minor limitations, but these do not undermine the study's conclusions. No erratum or re-analysis appears warranted based on the checks performed.
- 1.HIGHdata codeIn the Data availability statement, specify the full URL of the Research Data Deposit platform (www.researchdata.org.cn) and provide the expected deposit date or accession number once available.The current statement is vague and the URL appears truncated, which hinders readers' ability to access the data.
- 2.HIGHdata codeAdd a statement about code availability, either providing a link to any custom analysis code or explicitly stating that no custom code was used.Lack of code sharing reduces reproducibility, and reviewers expect transparency about analysis code.
- 3.MEDIUMreportingAdd an explicit statement of adherence to the CONSORT reporting guideline in the Methods or Acknowledgements.Explicitly referencing CONSORT enhances transparency and aligns with journal expectations for RCT reporting.
- 4.MEDIUMcopyeditFix the typo 'grouup' to 'group' in Methods, Sample size and statistical methods.Typographical errors detract from manuscript quality.
- 5.MEDIUMcopyeditCorrect the inconsistent spelling 'fluororacil' to 'fluorouracil' in Discussion, Comparison with other studies.Consistent terminology is expected in scientific writing.
- 6.MEDIUMcopyeditFix the typo 'neoadvuvant' to 'neoadjuvant' in Discussion, Comparison with other studies.Typographical errors detract from manuscript quality.
- 7.LOWreportingClarify the definition of the 38% grade 3/4 toxicity incidence during neoadjuvant chemotherapy, as the listed individual toxicities sum to more than 38%.The current reporting is ambiguous and could confuse readers about overlapping toxicities.
- 8.LOWreportingProvide a link to the full trial protocol in the Data availability statement or as a supplementary file.Access to the full protocol enhances transparency and reproducibility.
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.