Benmelstobart, anlotinib and chemotherapy in extensive-stage small-cell lung cancer: a randomized phase 3 trial.
Cheng Y, Chen J, Zhang W, Xie C, Hu Q, Zhou N, Huang C, Wei S, Sun H, Li X, Yu Y, Lai J, Yang H, Fang H, Chen H, Zhang P, Gu K, Wang Q, Shi J, Yi T, Xu X, Ye X, Wang D, Xie C, Liu C, Zheng Y, Lin D, Zhuang W, Lu P, Yu G, Li J, Gu Y, Li B, Wu R, Jiang O, Wang Z, Wu G, Lin H, Zhong D, Xu Y, Shu Y, Wu D, Chen X, Wang J, Wang M, Yang R
- DOI
- 10.1038/s41591-024-03132-1
- Record issued
- 2026-08-16
- 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/47a07c99-0b93-4bcb-b427-4eed1cad1b8f is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×4−2★
- ReportingData & code availability partially met−0.25★
- 01Other integrity concern
Trial NCT04234607 was first submitted to ClinicalTrials.gov on 2020-01-16, after the registered study start date of 2020-01. Retrospective registration means the protocol and outcomes were not on the public record before the study ran, which is what prospective registration exists to establish.
NCT04234607
reviewer’s wording
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 rigorously designed and reported phase 3 randomized controlled trial with strong methodology: comprehensive blinding, adequate power, detailed statistical methods, and transparent reporting of outcomes and limitations. The main weakness is the vague data availability statement, which lacks a concrete access mechanism for de-identified data, and several minor copyedit issues including an incomplete sentence and a percentage typo.
Both reviewers independently scored all eight dimensions and agreed on every status; no divergence required reconciliation. The statistics verification covered only 4 recomputable tests (all consistent), so the broader statistical claims remain unverified. The integrity check flagged retrospective registration (submitted 2020-01-16 after study start 2020-01) as a medium concern, which is a transparency caveat rather than a rigor failure.
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; 4 via agent-written checks.
- CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Verify the p-value for the OS hazard ratio (0.61) comparing benmelstobart+anlotinib+EC vs EC alone.
“hazard ratio (HR), 0.61 (95% CI 0.47–0.79); P = 0.0002”
Taken as given: The HR is 0.61 with 95% CI 0.47-0.79.; The CI is two-sided at 95%.; The p-value is two-sided.Method: Recomputed p-value from the hazard ratio and its 95% confidence interval using the normal approximation for the log hazard ratio.How we recomputed it: pCI(0.61, 0.47, 0.79, 1) - CONSISTENTreported p = .172 · recomputed p = .233Reviewers 1, 2Verify the p-value for the OS hazard ratio (0.86) comparing anlotinib+EC vs EC alone.
“HR, 0.86 (95% CI 0.67–1.10); P = 0.1723”
Taken as given: The HR is 0.86 with 95% CI 0.67-1.10.; The CI is two-sided at 95%.; The p-value is two-sided.Method: Recomputed p-value from the hazard ratio and its 95% confidence interval using the normal approximation for the log hazard ratio.How we recomputed it: pCI(0.86, 0.67, 1.10, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Verify the p-value for the PFS hazard ratio (0.32) comparing benmelstobart+anlotinib+EC vs EC alone.
“HR 0.32 (95% CI 0.26–0.41); P < 0.0001”
Taken as given: The HR is 0.32 with 95% CI 0.26-0.41.; The CI is two-sided at 95%.; The p-value is two-sided.Method: Recomputed p-value from the hazard ratio and its 95% confidence interval using the normal approximation for the log hazard ratio.How we recomputed it: pCI(0.32, 0.26, 0.41, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Verify the p-value for the PFS hazard ratio (0.44) comparing anlotinib+EC vs EC alone.
“HR 0.44 (95% CI 0.36–0.55); P < 0.0001”
Taken as given: The HR is 0.44 with 95% CI 0.36-0.55.; The CI is two-sided at 95%.; The p-value is two-sided.Method: Recomputed p-value from the hazard ratio and its 95% confidence interval using the normal approximation for the log hazard ratio.How we recomputed it: pCI(0.44, 0.36, 0.55, 1)
- lowinternal contradictionThe abstract reports grade ≥3 TRAE rates as 93.1%, 94.3%, 87.0%, while the Results section reports grade ≥3 TEAE rates as 94.3%, 95.9%, 89.0%. These are different metrics (TRAE vs TEAE) and not contradictory, but the close numbers might cause confusion.
The incidence of grade 3 or higher treatment-related adverse events was 93.1%, 94.3% and 87.0% ... (Abstract) ... Treatment-emergent adverse events (TEAEs) of any grade (100.0% in each group) and of grade 3 or higher were similar (94.3%, 95.9% and 89.0%) (Results)
Abstractreviewer’s wording - lowinternal contradictionTable 1 reports a stage II patient percentage as 1 (0.04%) which is inconsistent with the N=247 group size; should be 0.4%.
“Stage II | – | – | 1 (0.04%) c”
Table 1Find in source
Overstated conclusions
None foundConclusions 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.
Checked — nothing surfaced.
4 major claims checked against the paper's own evidence: all adequately supported.
- supportedReviewers 1, 2Benmelstobart and anlotinib plus EC significantly improved OS compared with EC alone.The presented OS analysis shows a statistically significant improvement with HR 0.61 and P=0.0002, meeting the prespecified criteria.Evidence: Median OS 19.3 vs 11.9 months, HR 0.61 (95% CI 0.47-0.79), P=0.0002.
“Compared with EC alone, median OS was prolonged with benmelstobart and anlotinib plus EC (19.3 versus 11.9 months; hazard ratio 0.61; P = 0.0002)”
AbstractFind in source - supportedReviewers 1, 2Anlotinib plus EC did not significantly improve OS compared with EC alone.The OS analysis shows no statistically significant difference, with HR 0.86 and P=0.1723.Evidence: Median OS 13.3 vs 11.9 months, HR 0.86 (95% CI 0.67-1.10), P=0.1723.
“improvement of OS was not statistically significant with anlotinib plus EC (13.3 versus 11.9 months; hazard ratio 0.86; P = 0.1723)”
AbstractFind in source - supportedReviewers 1, 2Benmelstobart and anlotinib plus EC significantly improved PFS compared with EC alone.The PFS analysis shows a significant improvement with HR 0.32 and P<0.0001.Evidence: Median PFS 6.9 vs 4.2 months, HR 0.32 (95% CI 0.26-0.41), P<0.0001.
“median IRC-assessed PFS was significantly longer, compared with the EC alone group, in both the benmelstobart and anlotinib plus EC group (6.9 months (95% CI 6.2–8.3) versus 4.2 months (95% CI 4.17–4.24); HR 0.32 (95% CI 0.26–0.41); P < 0.0001)”
ResultsFind in source - supportedReviewers 1, 2The addition of anti-angiogenesis therapy to immunochemotherapy may represent an efficacious and safe approach to the management of ES-SCLC.The efficacy and safety data support this conclusion, though the word 'may' appropriately reflects the interim nature of the OS analysis.Evidence: Significant OS and PFS improvements with manageable safety profile.
“Our findings suggest that the addition of anti-angiogenesis therapy to immunochemotherapy may represent an efficacious and safe approach to the management of ES-SCLC.”
DiscussionFind in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary efficacy endpoints are overall survival (OS) and progression-free survival (PFS), which are hard clinical outcomes. OS is a direct measure of mortality, and PFS is a validated clinical endpoint in oncology. The trial reports a statistically significant improvement in OS (median 19.3 vs 11.9 months, HR 0.61, P=0.0002) and PFS (median 6.9 vs 4.2 months, HR 0.32, P<0.0001) for the benmelstobart+anlotinib+EC arm compared to EC alone. These are not surrogate biomarkers but direct clinical outcomes.
“Compared with EC alone, median OS was prolonged with benmelstobart and anlotinib plus EC (19.3 versus 11.9 months; hazard ratio 0.61; P = 0.0002)... median IRC-assessed PFS was significantly longer... (6.9 months versus 4.2 months; HR 0.32; P < 0.0001)”
- ADEQUATEEffect sizeThe effect size for OS is a 7.4-month improvement in median OS (19.3 vs 11.9 months) with a hazard ratio of 0.61, which is clinically meaningful in extensive-stage small-cell lung cancer, where historical median OS is around 10-12 months. The 12-month OS rate improved from 49.0% to 64.1%. These differences are statistically significant and exceed the minimal clinically important difference for survival in this setting.
“median OS was prolonged with benmelstobart and anlotinib plus EC (19.3 versus 11.9 months; hazard ratio 0.61; P = 0.0002)... The estimated OS rate at 12 months was 64.1% ... and 49.0% ... in the benmelstobart and anlotinib plus EC and EC alone groups, respectively.”
Data authenticity concerns
1 finding · worst mediumAn 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.
- Other integrity concernAssessed
4 integrity concerns flagged (0 high).
- mediumotherTrial NCT04234607 was first submitted to ClinicalTrials.gov on 2020-01-16, after the registered study start date of 2020-01. Retrospective registration means the protocol and outcomes were not on the public record before the study ran, which is what prospective registration exists to establish.
NCT04234607
reviewer’s wording - lowotherThe data availability statement says 'All data required to interpret, verify or build new research on the published claims are included in the article or uploaded in .' The phrase 'uploaded in .' is incomplete, likely missing a repository name.
“All data required to interpret, verify or build new research on the published claims are included in the article or uploaded in .”
Data availabilityFind in source
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 four prior phase 3 trials (IMpower133, CASPIAN, CAPSTONE-1, ASTRUM-005) and discusses the failure of TIGIT inhibitors, establishing the premise. The rationale for combining anlotinib with PD-L1 inhibition is supported by preclinical and phase 1b data. Limitations of prior work are implicitly addressed by the trial design, though not explicitly framed as such.
“four randomized, controlled phase 3 studies (IMpower133, CASPIAN, CAPSTONE-1 and ASTRUM-005) have reported significant improvements in median OS on the order of 2–4 months”
“Anti-angiogenesis is known to augment the efficacy of ICIs in multiple cancer types”
“four randomized, controlled phase 3 studies (IMpower133, CASPIAN, CAPSTONE-1 and ASTRUM-005) have reported significant improvements in median OS on the order of 2–4 months”
“Anti-angiogenesis is known to augment the efficacy of ICIs in multiple cancer types”
Randomization method (central stratified randomization) and unit (patient) are described. Blinding is comprehensive, covering patients, investigators, reviewers, statisticians, and sponsor. Power analysis is provided (85% power, HR detection). Inclusion/exclusion criteria are detailed. Outlier handling is addressed through ITT and per-protocol analyses. Controls are appropriate (placebo and active comparator). Independent replication is not applicable for a single pivotal trial.
“Randomization was conducted by the central stratified randomization method, and stratified by ECOG PS (0 versus 1), brain metastases (‘yes’ or ‘no’) and liver metastases (‘yes’ or ‘no’).”
“Patients and their families or guardians, investigators, local and central radiological reviewers, the study statistician, data management personnel who were involved in data cleaning and analysis of the data, and the study sponsor were masked to treatment allocation until the final database was locked.”
“approximately 738 patients (combined sample size for PFS and OS) were needed for 1:1:1 randomization and to provide a power of 85% to detect an HR across treatment groups for disease progression or death at a two-sided significance level of 0.05.”
“Randomization was conducted by the central stratified randomization method, and stratified by ECOG PS (0 versus 1), brain metastases (‘yes’ or ‘no’) and liver metastases (‘yes’ or ‘no’).”
“Patients and their families or guardians, investigators, local and central radiological reviewers, the study statistician, data management personnel who were involved in data cleaning and analysis of the data, and the study sponsor were masked to treatment allocation until the final database was locked.”
“approximately 738 patients (combined sample size for PFS and OS) were needed for 1:1:1 randomization and to provide a power of 85% to detect an HR across treatment groups for disease progression or death at a two-sided significance level of 0.05.”
Sex is reported for all groups. Age (median and range) is reported. Demographics include ECOG performance status, smoking status, and disease stage. Species/strain and housing conditions are not applicable. Sex justification is not applicable as both sexes are enrolled.
“Male | 209 (85.0) | 204 (83.3) | 207 (83.8)”
“Median (range) | 62.0 (36–75) | 60.0 (32–75) | 63.0 (30–75)”
“Male | 209 (85.0) | 204 (83.3) | 207 (83.8)”
“Median (range) | 62.0 (36–75) | 60.0 (32–75) | 63.0 (30–75)”
The ethics statement names the Jinlin Province Cancer Hospital Institutional Review Board with an approval number. Written informed consent is stated. Compliance with Good Clinical Practice and Declaration of Helsinki is declared. This is a human trial, so IACUC is not applicable.
“central approval was obtained from the Jinlin Province Cancer Hospital Institutional Review Board (ethics approval no. 201909-053-01).”
“All patients provided written informed consent for participation before enrollment.”
“The trial was conducted per the principles of Good Clinical Practice guidelines and the Declaration of Helsinki.”
“central approval was obtained from the Jinlin Province Cancer Hospital Institutional Review Board (ethics approval no. 201909-053-01).”
“All patients provided written informed consent for participation before enrollment.”
“The trial was conducted per the principles of Good Clinical Practice guidelines and the Declaration of Helsinki.”
The drugs are named with doses and regimens. The statistical software (SAS 9.4) is identified. Antibodies, cell lines, mycoplasma, and organisms are not applicable as this is a clinical trial without wet-lab assays. Reagents are scored against the investigational products, which are adequately described.
“Benmelstobart was given intravenously at a dose of 1,200 mg once daily on day 1 of each cycle. Anlotinib was given orally, once daily, at a dose of 12 mg for 2 weeks followed by 1 week off.”
“The randomization list was generated by an independent statistician using a SAS statistical package (version 9.4, SAS Institute)”
“Benmelstobart was given intravenously at a dose of 1,200 mg once daily on day 1 of each cycle. Anlotinib was given orally, once daily, at a dose of 12 mg for 2 weeks followed by 1 week off.”
“The randomization list was generated by an independent statistician using a SAS statistical package (version 9.4, SAS Institute)”
Tests are named (stratified log-rank, Cox model, Cochran–Mantel–Haenszel). Assumptions are handled by design (stratified analyses). Exact p-values are reported (e.g., P = 0.0002). Effect sizes with 95% CIs are provided. Software is identified. Data presentation includes Kaplan-Meier curves and forest plots. Mathematical plausibility checks were not performed due to large N and continuous outcomes.
“Survival data were estimated with the Kaplan–Meier method and compared using the stratified log-rank test. HRs and associated 95% CIs were assessed using a stratified Cox proportional-hazards model”
“median OS was greater in the benmelstobart and anlotinib plus EC group compared with the EC alone group (19.3 months (95% CI 14.2 to not estimable) versus 11.9 months (95% CI 10.7–13.4); hazard ratio (HR), 0.61 (95% CI 0.47–0.79); P = 0.0002)”
“Survival data were estimated with the Kaplan–Meier method and compared using the stratified log-rank test. HRs and associated 95% CIs were assessed using a stratified Cox proportional-hazards model”
“hazard ratio 0.61; P = 0.0002”
The data availability statement says all data required to interpret the claims are included in the article or uploaded, but does not specify a repository or access mechanism. Individual deidentified participant data are not shared due to reidentification risk. This is a controlled-access situation, but the statement lacks a concrete mechanism or platform, making it inadequate. Repository deposit and accession numbers are not applicable for patient-level data. Code sharing is not applicable as no bespoke code is mentioned.
“All data required to interpret, verify or build new research on the published claims are included in the article or uploaded in . We cannot share individual deidentified participant data owing to the risk of reidentification and loss of patient confidentiality.”
“All data required to interpret, verify or build new research on the published claims are included in the article or uploaded in . We cannot share individual deidentified participant data owing to the risk of reidentification and loss of patient confidentiality.”
The trial is registered (NCT04234607). Methods are detailed. All primary and secondary endpoints are reported, including negative results (anlotinib plus EC OS). Limitations are discussed. Conclusions are proportional. Funding and competing interests are declared. Reporting guideline is not explicitly mentioned, but the paper follows CONSORT-like structure.
“ClinicalTrials.gov identifier: NCT04234607”
“Several limitations should be acknowledged. This study provided no direct comparison with the current standard-of-care immunochemotherapy.”
“ClinicalTrials.gov identifier: NCT04234607”
“Several limitations should be acknowledged. This study provided no direct comparison with the current standard-of-care immunochemotherapy.”
Registered (1 ID: 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 29 references by DOI: 28 verified — 1 no DOI (shown, not verified).
- NO DOIKeynote address on biostatistics and data retrievalNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
2 data/code links checked; 2 live.
- datahttps://clinicaltrials.gov/study/NCT04234607?term=NCT04234607&rank=1LIVEHTTP 200Resolves, but the content could not be matched to the paper.
- datahttps://clinicaltrials.gov/ct2/show/NCT04234607LIVEHTTP 200Resolves, but the content could not be matched to the paper.
Copyediting
6 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 6 minor suggestions below.
6 copyedit issues flagged: mostly consistency, typo, clarity.
- MINORconsistencyAbstract“93.1%, 94.3% and 87.0%”→ Ensure these percentages match the safety table (Table 3) where the same figures appear.The abstract reports grade ≥3 TRAE rates as 93.1%, 94.3%, 87.0%, which match Table 3, but the text in Results says '94.3%, 95.9% and 89.0%' for TEAEs; ensure clarity between TRAE and TEAE.
- MINORtypoResults, Patients and treatment“1,005 patients at 72 sites were assessed for eligibility. Overall, 738 were randomized”→ Consider adding a CONSORT flow diagram reference to clarify the numbers.The numbers are consistent, but a flow diagram is referenced as Fig. 1.
- MINORclarityDiscussion“The median OS in the ETER701 trial was 19.3 months.”→ Clarify that this is the median OS for the benmelstobart+anlotinib+EC arm, not the overall trial.The sentence could be misinterpreted as the overall trial median OS.
- MINORtypoAbstract“hereafter to referred to as”→ hereafter referred to asMissing 'be'
- MINORconsistencyTable 1“1 (0.04%) c”→ 1 (0.4%) cPercentage appears inconsistent with other values; likely a typo.
- MINORclarityData availability“uploaded in .”→ uploaded in the supplementary information.Incomplete sentence.
The published work is methodologically robust and the reported results are internally consistent for the tests that could be recomputed. An informed reader should weigh the retrospective registration timing and the lack of a concrete data-sharing mechanism as transparency limitations, but neither undermines the core findings. The minor copyedit issues (incomplete data-availability sentence, percentage typo) are cosmetic and do not warrant a correction on their own, though the authors may consider a correction for the data-availability statement.
- 1.HIGHdata codeIn the Data availability section, replace the incomplete sentence 'uploaded in .' with a concrete statement naming the repository or supplementary file, and specify a managed-access mechanism (e.g., Vivli, a data access committee with contact details and conditions) for any de-identified data that could be shared.The current statement is vague and incomplete, failing to provide readers any route to access or verify the data, which is a transparency gap in a data-driven trial.
- 2.HIGHreportingIn the Methods or a reporting summary, explicitly state adherence to CONSORT guidelines for this randomized trial.Reporting guideline adherence is not stated in the manuscript, and explicit mention would strengthen transparency and reproducibility.
- 3.MEDIUMcopyeditIn Table 1, correct the stage II percentage from '1 (0.04%)' to '1 (0.4%)' to match the group size.The percentage is internally inconsistent with the N=247 group size and is a clear typo that could mislead readers.
- 4.MEDIUMcopyeditIn the Abstract, fix the typo 'hereafter to referred to as' to 'hereafter referred to as'.This is a grammatical error that detracts from the manuscript's professionalism.
- 5.MEDIUMcopyeditIn the Discussion, clarify that the median OS of 19.3 months refers specifically to the benmelstobart+anlotinib+EC arm, not the overall trial.The current wording could be misinterpreted as the overall trial median OS, which would be misleading.
- 6.MEDIUMreportingIn the Data availability section, add a statement clarifying that no custom code was used, or provide access to any analysis code if applicable.This addresses the code-sharing sub-criterion and removes ambiguity about reproducibility.
- 7.MEDIUMreportingIn the Acknowledgements or Competing interests, clarify the specific role of the funder (Chia Tai Tianqing Pharmaceutical Group) in study design, data collection, analysis, and manuscript preparation.Transparency about funder involvement is important for assessing potential conflicts of interest.
- 8.LOWcopyeditIn the Results, add a reference to the CONSORT flow diagram (Fig. 1) when describing the 1,005 assessed and 738 randomized patients.This clarifies the patient flow and aligns with reporting guidelines.
- 9.LOWcopyeditIn the Abstract and Results, ensure the distinction between TRAE (93.1%, 94.3%, 87.0%) and TEAE (94.3%, 95.9%, 89.0%) rates is clearly labeled to avoid confusion.The close numbers for different metrics could be misread as a contradiction; explicit labeling prevents misinterpretation.
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.