Alpha¹ by KaimenConfidential

Kaimen Rigor review

Pembrolizumab plus axitinib versus sunitinib for advanced clear cell renal cell carcinoma: 5-year survival and biomarker analyses of the phase 3 KEYNOTE-426 trial

Rini BI, Plimack ER, Stus V, Gafanov R, Waddell T, Nosov D, Pouliot F, Alekseev B, Soulières D, Melichar B, Vynnychenko I, de Azevedo SJ, Borchiellini D, McDermott RS, Bedke J, Tamada S, Wu S, Markensohn J, Zhang Y, Loboda A, Vajdi A, Perini RF, Burgents J, Powles T.

DOI: 10.1038/s41591-025-03867-5
Record issued: 2026-08-10
Engine: 7.29.0
Exported: 2026-09-22

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/e86b98f4-bafc-4552-8260-5abd0d3287fe is authoritative.

Kaimen Rigor review

Kaimen Rigor reviewed

3.75 / 5

Adequate — real issues to address

Interventional Study

Critical

High-priority

How this rating was calculated

Started at 5★ — deductions

ReportingStudy design partially met−0.25★
ReportingBiological variables partially met−0.25★
ReportingEthical approvals partially met−0.25★
ReportingKey resources partially met−0.25★
ReportingData & code availability partially met−0.25★

Model-reported confidence: highReviewer agreement: lowEngine: 7.29.0Source: full text

Findings by category

01ArithmeticNone found 02OverreachNone found 03AuthenticityNone found 504DisclosureReporting gaps 05AvailabilityNone found 06Copyedit9 minor suggestions

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 paper presents a robust long-term follow-up of a phase 3 clinical trial, with a strong scientific premise, well-described statistical analyses, and high reporting transparency. However, several methodological and reporting details, particularly regarding study design, biological variables, ethical approvals, and data/code availability, are inadequately specified.

This evaluation synthesized three independent reviewer runs and a copyedit report. Reviewers largely agreed on the strengths but diverged on the completeness of reporting in several dimensions. My synthesis weighed the specific evidence provided by each reviewer, prioritizing concrete omissions over general statements. The statistics component recomputed 4 reported tests consistently, but this covers only a subset of the reported statistics.

Numerical inconsistencies

None found

Values 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.

Checked — nothing surfaced.

Recomputed 4 tests: 4 consistent, 0 inconsistent; 4 via agent-written checks.

Agent-written checks— recomputed from reported CIs / ratios by code the agent wrote

CONSISTENTreported p = .040 · recomputed p = .040Reviewer 1
Check p-value for OS hazard ratio (0.84, 95% CI 0.71-0.99) using pCI function.
“The median OS was 47.2 months in the pembrolizumab plus axitinib arm and 40.8 months in the sunitinib arm (hazard ratio: 0.84; 95% confidence interval: 0.71–0.99)”
Find in source
Taken as given: The hazard ratio is 0.84.; The 95% confidence interval is 0.71 to 0.99.; The hazard ratio is on a log scale (log=1).
Method: Recomputed two-tailed p-value from the hazard ratio and its 95% confidence interval using the pCI function.
How we recomputed it: pCI(0.84, 0.71, 0.99, 1)
CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 1
Check p-value for PFS hazard ratio (0.69, 95% CI 0.59-0.81) using pCI function.
“Median PFS was 15.7 months in the pembrolizumab plus axitinib arm and 11.1 months in the sunitinib arm (hazard ratio: 0.69; 95% confidence interval: 0.59–0.81)”
Find in source
Taken as given: The hazard ratio is 0.69.; The 95% confidence interval is 0.59 to 0.81.; The hazard ratio is on a log scale (log=1).
Method: Recomputed two-tailed p-value from the hazard ratio and its 95% confidence interval using the pCI function.
How we recomputed it: pCI(0.69, 0.59, 0.81, 1)
CONSISTENTreported p = .043 · recomputed p = .040Reviewer 2
Recompute the two-tailed p-value for the overall survival hazard ratio from its 95% confidence interval.
“The median OS was 47.2 months in the pembrolizumab plus axitinib arm and 40.8 months in the sunitinib arm (hazard ratio: 0.84; 95% confidence interval: 0.71–0.99)”
Find in source
Taken as given: The confidence interval is two-sided at 95% level.; The hazard ratio is a ratio, so log=1.; The estimate is the point estimate of the HR.
Method: Using the pCI function which computes the p-value from the estimate and confidence interval under a normal approximation for the log hazard ratio.
How we recomputed it: pCI(0.84, 0.71, 0.99, 1)
CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 2
Recompute the two-tailed p-value for the progression-free survival hazard ratio from its 95% confidence interval.
“Median PFS was 15.7 months in the pembrolizumab plus axitinib arm and 11.1 months in the sunitinib arm (hazard ratio: 0.69; 95% confidence interval: 0.59–0.81)”
Find in source
Taken as given: The confidence interval is two-sided at 95% level.; The hazard ratio is a ratio, so log=1.; The estimate is the point estimate of the HR.
Method: Using the pCI function which computes the p-value from the estimate and confidence interval under a normal approximation for the log hazard ratio.
How we recomputed it: pCI(0.69, 0.59, 0.81, 1)

Overstated conclusions

None found

Conclusions 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.

12 major claims checked against the paper's own evidence: all adequately supported.

supportedReviewers 1, 2, 3
Pembrolizumab plus axitinib showed sustained benefits in OS, PFS, and ORR compared to sunitinib.
The reported hazard ratios and confidence intervals for OS and PFS, and the ORR percentages, directly support this claim.
Evidence: OS HR 0.84 (95% CI 0.71-0.99), PFS HR 0.69 (95% CI 0.59-0.81), ORR 60.6% vs 39.6%.
“Pembrolizumab plus axitinib showed sustained benefits in OS (hazard ratio: 0.84; 95% confidence interval: 0.71–0.99), PFS (hazard ratio: 0.69; 95% confidence interval: 0.59–0.81) and ORR (60.6% versus 39.6%) compared to sunitinib.”
AbstractFind in source
supportedReviewer 1
An 18-gene T-cell-inflamed gene expression profile (Tcell inf GEP) was positively associated with OS, PFS, and ORR within the pembrolizumab plus axitinib arm.
The reported p-values for Tcell inf GEP associations in the pembrolizumab plus axitinib arm support this claim.
Evidence: P = 0.002 for OS, P < 0.0001 for PFS, P < 0.0001 for ORR.
“An 18-gene T-cell-inflamed gene expression profile (Tcell inf GEP) was positively associated with OS ( P = 0.002), PFS ( P < 0.0001) and ORR ( P < 0.0001) within the pembrolizumab plus axitinib arm.”
AbstractFind in source
supportedReviewers 1, 2
An angiogenesis signature was positively associated with OS within the pembrolizumab plus axitinib arm and with OS, PFS, and ORR within the sunitinib arm.
The reported p-values for angiogenesis signature associations support this claim.
Evidence: P = 0.004 for OS in pembrolizumab+axitinib; P < 0.0001 for OS, P < 0.001 for PFS, P = 0.002 for ORR in sunitinib.
“An angiogenesis signature was positively associated with OS ( P = 0.004) within the pembrolizumab plus axitinib arm and with OS ( P < 0.0001), PFS ( P < 0.001) and ORR ( P = 0.002) within the sunitinib arm.”
AbstractFind in source
supportedReviewer 1
PD-L1 combined positive score was only associated (negatively) with OS within the sunitinib arm.
The reported p-value for PD-L1 CPS association with OS in the sunitinib arm supports this claim.
Evidence: P = 0.025 for OS in sunitinib arm.
“Across arms, programmed cell death ligand 1 combined positive score was only associated (negatively) with OS within the sunitinib arm ( P = 0.025).”
AbstractFind in source
supportedReviewers 1, 2
PBRM1 mutation had a positive association with ORR within the pembrolizumab plus axitinib arm.
The reported p-value and ORR rates support this claim.
Evidence: P = 0.002; ORR 71.4% vs 52.3%.
“Additionally, PBRM1 (polybromo-1) mutation had a positive association with ORR ( P = 0.002) within the pembrolizumab plus axitinib arm.”
AbstractFind in source
supportedReviewers 1, 2
Within the sunitinib arm, OS was positively associated with VHL and PBRM1 mutations and negatively associated with BAP1 mutation.
The reported p-values support this claim.
Evidence: P = 0.040 for VHL, P = 0.010 for PBRM1, P = 0.019 for BAP1.
“Within the sunitinib arm, OS was positively associated with VHL (von Hippel–Lindau tumor suppressor gene) ( P = 0.040) and PBRM1 ( P = 0.010) mutations and was negatively associated with BAP1 ( BRCA1- associated protein 1) mutation ( P = 0.019).”
AbstractFind in source
supportedReviewer 2
An 18-gene T-cell-inflamed GEP was positively associated with OS, PFS, and ORR within the pembrolizumab plus axitinib arm.
The paper provides p-values from logistic regression and Cox models showing significant associations.
Evidence: Table 2 with p-values: P = 0.002 for OS, P < 0.0001 for PFS and ORR.
In the pembrolizumab plus axitinib arm, higher Tcell inf GEP was associated with improved ORR (P < 0.0001), PFS (P < 0.0001) and OS (P = 0.002)
Resultsreviewer’s wording
supportedReviewer 2
Results showed a sustained clinical benefit with pembrolizumab plus axitinib over sunitinib and provide valuable information on biomarkers for immunotherapy-based treatment combinations.
The clinical benefit is demonstrated by the survival data; biomarker associations are exploratory and appropriately caveated.
Evidence: The entire Results and Discussion sections support this overarching claim.
“Results showed a sustained clinical benefit with pembrolizumab plus axitinib over sunitinib and provide valuable information on biomarkers for immunotherapy-based treatment combinations in advanced RCC.”
AbstractFind in source
supportedReviewer 3
Tcell inf GEP is positively associated with OS, PFS, and ORR within the pembrolizumab plus axitinib arm.
Table 2 shows statistically significant associations with p-values < 0.05.
Evidence: Table 2: Tcell inf GEP ORR P = 2.03e-6, PFS P = 1.41e-5, OS P = 0.002
Tcell inf GEP | 2.03 × 10^−6 (+) | 1.41 × 10^−5 (+) | 0.002 (+) | 0.741 | 0.464 | 0.547
Table 2reviewer’s wording
supportedReviewer 3
Angiogenesis signature is positively associated with OS in the pembrolizumab plus axitinib arm and with OS, PFS, and ORR in the sunitinib arm.
Table 2 shows significant associations for the angiogenesis signature.
Evidence: Table 2: Angiogenesis: pembro+axi OS P=0.004, sunitinib ORR P=0.002, PFS P=5.66e-4, OS P=1.69e-7
Angiogenesis | 0.202 | 0.244 | 0.004 (+) | 0.002 (+) | 5.66 × 10^−4 (+) | 1.69 × 10^−7 (+)
Table 2reviewer’s wording
supportedReviewer 3
PD-L1 CPS is negatively associated with OS only within the sunitinib arm.
Table 2 shows a significant negative association only for sunitinib OS (P=0.025).
Evidence: Table 2: PD-L1 CPS: sunitinib OS P=0.025 (-)
PD-L1 CPS | 0.053 | 0.168 | 0.544 | 0.558 | 0.331 | 0.025 (−)
Table 2reviewer’s wording
supportedReviewer 3
PBRM1 mutation is positively associated with ORR within the pembrolizumab plus axitinib arm.
Extended Data Table 3 shows a significant positive association for ORR (P=0.002).
Evidence: Extended Data Table 3: PBRM1 ORR P=0.002 (+) in pembro+axi
PBRM1 | ORR: 0.002 (+) | PFS: 0.758 | OS: 0.368
Table 3reviewer’s wording

Efficacy claim is anchored to an adequate endpoint and a meaningful effect.

ADEQUATESurrogate endpoint
The primary efficacy claim is based on overall survival (OS), a hard clinical endpoint, not a surrogate. No surrogate was used as the primary basis for the efficacy claim.
“Pembrolizumab plus axitinib showed sustained benefits in OS (hazard ratio: 0.84; 95% confidence interval: 0.71–0.99)”
Find in source
ADEQUATEEffect size
The effect size is a statistically significant improvement in OS (HR 0.84, 95% CI 0.71-0.99) with a median survival difference of 6.4 months, which is clinically meaningful in the context of advanced renal cell carcinoma.
“median OS was 47.2 months in the pembrolizumab plus axitinib arm and 40.8 months in the sunitinib arm (hazard ratio: 0.84; 95% confidence interval: 0.71–0.99)”
Find in source

Data authenticity concerns

None found

An 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

5 findings · worst medium

Required detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.

Biological variables underreported (sex, age, strain)Assessed
Data/code availability incompleteAssessed
Ethics/consent reporting incompleteAssessed
Key resources under-identified (antibodies, cell lines, RRIDs)Assessed
Study-design details incomplete (controls, blinding, power)Assessed

Eight Dimensions3 pass · 5 warn

The abstract and introduction cite the initial KEYNOTE-426 results, IMmotion151, JAVELIN Renal 101, CheckMate 9ER, among others. It acknowledges the need for durable outcomes and predictive biomarkers, addressing limitations of prior biomarker studies such as confounding from different drugs. The hypothesis follows from the cited evidence.

Sub-criteria

prior work citedADEQUATE

premise rationaleADEQUATE

limitations addressedADEQUATE

Per reviewer (3) — the votes behind the verdict

PASSReviewer 1· 90% conf

The paper establishes a clear scientific premise by citing prior research on PD-1/VEGFR-TKI combinations and biomarker studies, and provides a logical rationale for the long-term follow-up and biomarker analyses.

Evidence

paraphrase[Introduction, paragraph 1]

“Studies to evaluate predictive and prognostic biomarkers in metastatic RCC have been largely derived from phase 3 studies (for example, IMmotion151, JAVELIN Renal 101 and CheckMate 9ER) in which similar mechanisms (for example, PD-L1 inhibitor plus VEGF-TKI) but different drugs were evaluated as frontline therapy.”

“Because the treatment duration of pembrolizumab is limited to 2 years, it is important to assess the long-term durability of clinical outcomes.”

Introduction ¶2Find in source

“Thus, extrapolation of these data to define predictive biomarkers in frontline therapy for advanced RCC is potentially confounded.”

Introduction ¶1Find in source

PASSReviewer 2· 95% conf

The paper thoroughly cites prior work, establishes a clear rationale for long-term follow-up and biomarker analysis, and discusses limitations of previous studies.

Evidence

paraphrase[Introduction, paragraph 1]

“Because the treatment duration of pembrolizumab is limited to 2 years, it is important to assess the long-term durability of clinical outcomes. There is an unmet need for biomarkers that are predictive of patient outcomes after using available first-line treatment options in RCC.”

“Thus, extrapolation of these data to define predictive biomarkers in frontline therapy for advanced RCC is potentially confounded.”

Introduction ¶2Find in source

PASSReviewer 3· 95% conf

The paper clearly establishes the scientific foundation by citing prior work, providing a logical rationale, and addressing limitations of prior research.

Evidence

“At the first interim analysis of the phase 3 KEYNOTE-426 trial, first-line pembrolizumab plus axitinib showed superior overall survival (OS), progression-free survival (PFS) and objective response rate (ORR) over sunitinib for advanced renal cell carcinoma (RCC).”

Introduction ¶1Find in source

paraphrase[Introduction, paragraph 2]

“extrapolation of these data to define predictive biomarkers in frontline therapy for advanced RCC is potentially confounded.”

Introduction ¶2Find in source

Randomization method is not explicitly stated (only 'randomly assigned' with stratification). Blinding is described as 'open-label' but no rationale for the open-label design is provided. No a priori sample size calculation is reported in this paper. Inclusion/exclusion criteria and ITT analysis population are adequately defined.

Sub-criteria

controlsN/A

power analysisNOT REPORTED

blinding levelsINADEQUATE

outlier handlingADEQUATE

randomization unitADEQUATE

inclusion exclusionADEQUATE

randomization methodINADEQUATE

replicate distinctionN/A

independent replicationN/A

Per reviewer (3) — the votes behind the verdict

PASSReviewer 1· 90% conf

The trial design is rigorous: randomized, open-label, with stratification, prespecified endpoints, and a statistical analysis plan for the biomarker analyses.

Evidence

“Participants were randomly assigned in a 1:1 ratio to receive pembrolizumab 200 mg intravenously once every 3 weeks for up to 35 cycles (~2 years) plus axitinib 5 mg by mouth twice daily continuously or sunitinib 50 mg by mouth once daily for 4 weeks on and 2 weeks off, continuously. Randomization was stratified according to the IMDC risk group (favorable versus intermediate versus poor risk) and by geographic region (North America versus Western Europe versus rest of the world).”

MethodsFind in source

paraphrase[Methods, Inclusion and ethics]

“KEYNOTE-426 ( NCT02853331 ) is a randomized, open-label, phase 3 trial conducted across 129 medical centers globally.”

“The biomarker analysis followed a statistical analysis plan written before merging of the clinical data with biomarker assessment, specifying where statistical testing would be used and what biomarker cutoffs defined the subgroups for treatment arm comparisons.”

Statistical analysisFind in source

WARNReviewer 2· 85% conf

The study design is a randomized open-label trial, but several methodological details are insufficiently reported, such as the randomization method, blinding rationale, and a priori power analysis, leading to a warning.

Evidence

MethodsFind in source

paraphrase[Methods, Inclusion and ethics]

“KEYNOTE-426 (NCT02853331) is a randomized, open-label, phase 3 trial conducted across 129 medical centers globally.”

paraphrase[Methods, Statistical analysis]

“The paper does not mention a sample size calculation; it states that the present analysis is exploratory and no formal hypothesis testing was performed.”

WARNReviewer 3· 80% conf

The study design is reported but lacks details on randomization method, blinding rationale, and power analysis, while inclusion/exclusion criteria and analysis population are adequately defined.

Evidence

MethodsFind in source

paraphrase[Methods, 'Inclusion and ethics']

“KEYNOTE-426 (NCT02853331) is a randomized, open-label, phase 3 trial”

“eligible participants were adults with newly diagnosed stage IV or recurrent clear cell RCC who had not previously received systemic therapy for advanced disease. Participants had a Karnofsky Performance Scale score of 70% or higher at baseline, one or more measurable lesions per Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1) as assessed by the investigator and a tumor sample available for biomarker assessment.”

MethodsFind in source

Table 1 reports sex (male/female) and age (median and range). Health status is not reported in Table 1 (only eligibility criteria of KPS >=70). Weight is not reported. Race/ethnicity and comorbidities are not reported.

Sub-criteria

demographicsINADEQUATE

sex reportedADEQUATE

sex justifiedN/A

age weight healthINADEQUATE

housing conditionsN/A

species strain sourceN/A

Per reviewer (3) — the votes behind the verdict

PASSReviewer 1· 90% conf

Key biological variables (sex, age, disease characteristics) are reported in the baseline table, and the biomarker analysis includes relevant tumor molecular features.

Evidence

“Age, median (range), years | 62.0 (30–89) | 61.0 (26–90) | | <65 years | 260 (60.2) | 278 (64.8) | | Sex | | Male | 308 (71.3) | 320 (74.6) | | Female | 124 (28.7) | 109 (25.4) |”

Table 1Find in source

“Participants had a Karnofsky Performance Scale score of 70% or higher at baseline, one or more measurable lesions per Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1) as assessed by the investigator and a tumor sample available for biomarker assessment.”

MethodsFind in source

“Region of enrollment | | North America | 104 (24.1) | 103 (24.0) | | Western Europe | 106 (24.5) | 104 (24.2) | | Rest of the world | 222 (51.4) | 222 (51.7) |”

Table 1Find in source

WARNReviewer 2· 90% conf

Sex is reported, but age is reported without weight or health status, and demographics lack race/ethnicity and comorbidities, leading to inadequate reporting.

Evidence

paraphrase[Table 1]

“Sex: Male 308 (71.3%), Female 124 (28.7%)”

paraphrase[Table 1]

“Age, median (range), years: 62.0 (30–89)”

absence[Table 1 and Methods]

Race/ethnicity and comorbidities are not reported.

WARNReviewer 3· 70% conf

Sex is reported adequately, but age, weight, and health status are incompletely reported, and demographics lack race/ethnicity and comorbidities.

Evidence

paraphrase[Table 1]

“Male 308 (71.3%) | Female 124 (28.7%)”

paraphrase[Table 1]

“Age, median (range), years: 62.0 (30–89)”

paraphrase[Table 1]

“Table 1 includes age, sex, region, IMDC risk group, sarcomatoid features, PD-L1 CPS, number of organs with metastases, and most common sites of metastasis. No race/ethnicity or comorbidities.”

The paper states the trial was approved by 'appropriate institutional review boards' but does not name the specific body or provide a protocol number. Written informed consent and Good Clinical Practice compliance are stated.

Sub-criteria

iacuc statementN/A

informed consentADEQUATE

irb ethics statementINADEQUATE

regulatory complianceADEQUATE

Per reviewer (3) — the votes behind the verdict

PASSReviewer 1· 90% conf

The trial reports ethics approval by institutional review boards, written informed consent, and compliance with Good Clinical Practice.

Evidence

“The trial was conducted in accordance with the principles of Good Clinical Practice and was approved by the appropriate institutional review boards and regulatory agencies. Written informed consent was provided by all participants before enrollment.”

MethodsFind in source

WARNReviewer 2· 90% conf

Ethics approval is mentioned but not named specifically, and the regulatory compliance statement is adequate, but the vague IRB statement and lack of protocol number result in a warning.

Evidence

“The trial was conducted in accordance with the principles of Good Clinical Practice and was approved by the appropriate institutional review boards and regulatory agencies.”

MethodsFind in source

“Written informed consent was provided by all participants before enrollment.”

MethodsFind in source

“The trial was conducted in accordance with the principles of Good Clinical Practice”

MethodsFind in source

WARNReviewer 3· 95% conf

Informed consent and regulatory compliance are adequately reported, but the ethics approval statement is vague (no named IRB or protocol number).

Evidence

“The trial was conducted in accordance with the principles of Good Clinical Practice and was approved by the appropriate institutional review boards and regulatory agencies.”

MethodsFind in source

“Written informed consent was provided by all participants before enrollment.”

MethodsFind in source

“The trial was conducted in accordance with the principles of Good Clinical Practice”

MethodsFind in source

Pembrolizumab, axitinib, and sunitinib are named with dose and regimen but the manufacturer is not stated. The PD-L1 IHC assay is identified with vendor. Software tools (SAS, R, OmicSoft, etc.) are identified with versions. No custom code repository is provided.

Sub-criteria

mycoplasma testingN/A

reagents identifiedINADEQUATE

organisms identifiedN/A

antibodies identifiedN/A

cell line authenticationN/A

software tools identifiedADEQUATE

Per reviewer (3) — the votes behind the verdict

PASSReviewer 1· 90% conf

The investigational products (pembrolizumab, axitinib, sunitinib) are identified with doses and regimens, and the PD-L1 IHC assay and software tools are named.

Evidence

MethodsFind in source

“PD-L1 expression was centrally determined using the PD-L1 IHC 22C3 pharmDx (Agilent Technologies).”

MethodsFind in source

“Statistical analyses were performed in SAS version 9.4 and R version 4.2.1 software.”

Statistical analysisFind in source

WARNReviewer 2· 85% conf

The investigational drugs are identified with dose and regimen but not with manufacturer/source, and custom code is not shared, while software tools are adequately identified.

Evidence

“pembrolizumab 200 mg intravenously once every 3 weeks for up to 35 cycles (~2 years) plus axitinib 5 mg by mouth twice daily continuously or sunitinib 50 mg by mouth once daily for 4 weeks on and 2 weeks off, continuously.”

MethodsFind in source

“Statistical analyses were performed in SAS version 9.4 and R version 4.2.1 software.”

Statistical analysisFind in source

absence[Data availability section]

No custom code repository is provided.

WARNReviewer 3· 75% conf

Software tools are well identified, but the investigational drugs are not identified with manufacturer/source, and sequencing data are not deposited in a public repository.

Evidence

MethodsFind in source

“Statistical analyses were performed in SAS version 9.4 and R version 4.2.1 software.”

Statistical analysisFind in source

The paper uses Cox proportional hazards, logistic regression, Kaplan-Meier, and Miettinen-Nurminen methods. P-values are reported exactly (e.g., P = 0.002, P < 0.0001). Hazard ratios with 95% CIs are given. Software versions are stated. Data presentation includes KM curves, forest plots, and CONSORT diagram. No mathematical inconsistencies were detected.

Sub-criteria

tests namedADEQUATE

exact p valuesADEQUATE

effect sizes ciADEQUATE

data presentationADEQUATE

software identifiedADEQUATE

assumptions verifiedADEQUATE

mathematical plausibilityPASS

Per reviewer (3) — the votes behind the verdict

PASSReviewer 1· 85% conf

Statistical methods are well described, including tests, models, and software. P-values are reported exactly for many analyses, and effect sizes with confidence intervals are provided for primary outcomes.

Evidence

“To estimate the magnitude of the treatment difference (that is, hazard ratio) between the treatment arms, we used a stratified Cox proportional hazards model with the Efron method for handling ties.”

Statistical analysisFind in source

“Tcell inf GEP | 2.03 × 10 −6 (+) | 1.41 × 10 −5 (+) | 0.002 (+) | 0.741 | 0.464 | 0.547”

Table 2Find in source

“The median OS was 47.2 months in the pembrolizumab plus axitinib arm and 40.8 months in the sunitinib arm (hazard ratio: 0.84; 95% confidence interval: 0.71–0.99)”

ResultsFind in source

PASSReviewer 2· 90% conf

Statistical tests are named, exact p-values are reported, effect sizes with confidence intervals are provided, software is identified, and data presentation is appropriate for a clinical trial.

Evidence

Statistical analysisFind in source

“The median OS was 47.2 months in the pembrolizumab plus axitinib arm and 40.8 months in the sunitinib arm (hazard ratio: 0.84; 95% confidence interval: 0.71–0.99)”

ResultsFind in source

“Statistical analyses were performed in SAS version 9.4 and R version 4.2.1 software.”

Statistical analysisFind in source

PASSReviewer 3· 90% conf

Statistical tests are named, assumptions are handled by standard methods, effect sizes with CIs are reported, software is identified, and data presentation is appropriate for a clinical trial.

Evidence

paraphrase[Methods, 'Statistical analysis']

“We used the Kaplan–Meier method to estimate OS, PFS and DOR in each treatment arm. To estimate the magnitude of the treatment difference (that is, hazard ratio) between the treatment arms, we used a stratified Cox proportional hazards model with the Efron method for handling ties. The stratified Miettinen and Nurminen method, with weights proportional to the stratum size, was used for comparison of ORR between the treatment arms.”

“The median OS was 47.2 months in the pembrolizumab plus axitinib arm and 40.8 months in the sunitinib arm (hazard ratio: 0.84; 95% confidence interval: 0.71–0.99)”

ResultsFind in source

paraphrase[Table 2]

“Tcell inf GEP: ORR P = 2.03 × 10^−6 (+), PFS P = 1.41 × 10^−5 (+), OS P = 0.002 (+)”

The paper provides a detailed data availability statement via the MSD data-sharing website. However, no repository deposit or accession numbers for RNA-seq/WES data are mentioned, and custom analysis code is not shared.

Sub-criteria

code sharingNOT REPORTED

accession numbersNOT REPORTED

repository depositNOT REPORTED

data availability statementADEQUATE

Per reviewer (3) — the votes behind the verdict

PASSReviewer 1· 90% conf

A detailed data availability statement describes a managed-access process for clinical data, which is appropriate for patient-level data.

Evidence

“The MSD data-sharing website ( https://externaldatasharing-msd.com/ ) outlines the process and requirements for submitting a data request.”

Data availabilityFind in source

WARNReviewer 2· 85% conf

A data availability statement with a concrete platform (MSD data-sharing website) is provided, and a procedure for accessing genetic data is described, but custom code is not shared.

Evidence

paraphrase[Data availability section]

“Merck Sharp & Dohme (MSD) ... is committed to providing qualified scientific researchers access to anonymized data ... The MSD data-sharing website (https://externaldatasharing-msd.com/) outlines the process and requirements for submitting a data request.”

“Access to genetic or exploratory biomarker data requires a detailed, hypothesis-driven statistical analysis plan that is collaboratively developed by the requestor and MSD subject matter experts”

Data availabilityFind in source

absence[Data availability section and Methods]

No custom code is shared.

WARNReviewer 3· 90% conf

A data availability statement with a concrete mechanism for requesting data is provided, but sequencing data are not deposited in a public repository and code is not shared.

Evidence

paraphrase[Data availability section]

“Merck Sharp & Dohme (MSD) is committed to providing qualified scientific researchers access to anonymized data and clinical study reports from the company’s clinical trials for the purpose of conducting legitimate scientific research. The MSD data-sharing website (https://externaldatasharing-msd.com/) outlines the process and requirements for submitting a data request.”

absence[Data availability section]

No repository deposit for sequencing data.

absence[Data availability section]

No code sharing.

The paper includes a CONSORT diagram, trial registration number, detailed methods, and a limitations paragraph. All primary and secondary outcomes are reported. Funding source and competing interests are disclosed.

Sub-criteria

funding coiADEQUATE

trial registrationADEQUATE

reporting guidelineADEQUATE

methods completenessADEQUATE

all outcomes reportedADEQUATE

limitations discussedADEQUATE

conclusions proportionalADEQUATE

Per reviewer (3) — the votes behind the verdict

PASSReviewer 1· 90% conf

The manuscript is transparent: methods are detailed, trial registration is provided, limitations are discussed, conclusions are proportional, and funding/COI are disclosed.

Evidence

“ClinicalTrials.gov identifier: NCT02853331 (https://clinicaltrials.gov/study/NCT02853331)”

AbstractFind in source

“However, the prespecified exploratory biomarker analysis from the KEYNOTE-426 trial is limited by the small sample sizes of some of the subgroups and the lack of statistical power and (or) multiplicity adjustments for association analysis of some biomarkers, hindering definitive conclusions.”

Discussion ¶8Find in source

“B.I.R. reports receiving writing support for the present paper from Merck Sharp & Dohme (MSD); payment to his institution for clinical trial research from MSD; and consulting fees from MSD.”

Competing interestsFind in source

PASSReviewer 2· 90% conf

The paper is well-structured, includes trial registration, reports all outcomes, discusses limitations, and provides proportional conclusions, though it does not explicitly reference a reporting guideline.

Evidence

“ClinicalTrials.gov identifier: NCT02853331”

AbstractFind in source

“The prespecified exploratory biomarker analysis from the KEYNOTE-426 trial is limited by the small sample sizes of some of the subgroups and the lack of statistical power and (or) multiplicity adjustments for association analysis of some biomarkers, hindering definitive conclusions.”

Discussion ¶5Find in source

“Funding for this study was provided by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA”

FundingFind in source

PASSReviewer 3· 95% conf

Methods are complete, the trial is registered, a CONSORT diagram is included, all outcomes are reported, limitations are discussed, conclusions are proportional, and funding/COI are disclosed.

Evidence

“ClinicalTrials.gov identifier: NCT02853331”

AbstractFind in source

“Fig. 1 CONSORT diagram.”

Figure 1Find in source

“the prespecified exploratory biomarker analysis from the KEYNOTE-426 trial is limited by the small sample sizes of some of the subgroups and the lack of statistical power and (or) multiplicity adjustments for association analysis of some biomarkers, hindering definitive conclusions.”

DiscussionFind in source

Registered (1 ID: ClinicalTrials.gov). Reporting guideline cited: CONSORT.

ClinicalTrials.gov

NCT02853331

Reporting guidelines: CONSORT

Broken references and links

None found

References 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 49 references by DOI: 36 verified — 13 no DOI (shown, not verified).

NO DOI
Keytruda (pembrolizumab injection), for intravenous use
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
EAU guidelines on renal cell carcinoma
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
PBRM1 and KDM5C cooperate to define high-angiogenesis tumors and increased antiangiogenic response in renal cancer
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
Biomarker analysis from the phase 3 CheckMate 9ER trial of nivolumab + cabozantinib v sunitinib for advanced renal cell carcinoma (aRCC)
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
Novel serum glycoproteomic biomarkers predict response to nivolumab plus cabozantinib (NIVO+CABO) versus sunitinib (SUN) in advanced RCC (aRCC): analysis from CheckMate 9ER
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
KIM-1 as a circulating biomarker in metastatic RCC: post-hoc analysis of JAVELIN Renal 101
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
Evaluation of circulating kidney injury marker-1 (KIM-1) as a prognostic and predictive biomarker in advanced renal cell carcinoma (aRCC): post-hoc analysis of CheckMate 214
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
Association of gene expression with clinical outcomes in patients with renal cell carcinoma treated with pembrolizumab in KEYNOTE-427
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
Immunomodulatory effects of tyrosine kinase inhibitors (TKIs) in renal cell carcinoma (RCC) patients
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
Nivolumab plus cabozantinib (N+C) vs sunitinib (S) for previously untreated advanced renal cell carcinoma (aRCC): final follow-up results from the CheckMate 9ER trial
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
Biomarker-driven prospective clinical trial in renal cell carcinoma: developing machine learning models to allocate patients to treatment arms using RNA sequencing
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
SAMETA: a phase III study of savolitinib + durvalumab vs sunitinib and durvalumab monotherapy in patients with MET-driven, unresectable, locally advanced/metastatic papillary renal cell carcinoma
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
NO DOI
Guidance for industry: clinical trial endpoints for the approval of cancer drugs and biologics
No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).

3 data/code links checked; 3 live.

datahttps://clinicaltrials.gov/study/NCT02853331LIVEHTTP 200
Resolves, but the content could not be matched to the paper.
datahttps://clinicaltrials.gov/ct2/show/NCT02853331LIVEHTTP 200
Resolves, but the content could not be matched to the paper.
datahttps://externaldatasharing-msd.com/LIVEHTTP 200
Resolves, but the content could not be matched to the paper.

Copyediting

9 minor

Wording, consistency and formatting errors that need correcting before submission.

No major wording or formatting errors. 9 minor suggestions below.

9 copyedit issues flagged: mostly typo, consistency, clarity.

MINORtypoDiscussion, paragraph 6
“pressent dataset”
→ present dataset
Typographical error.
MINORtypoDiscussion, paragraph 6
“PRRM1 mutation”
→ PBRM1 mutation
Typographical error in gene name.
MINORconsistencyExtended Data Fig. 4-7
“von Lindau-Hippel tumor suppressor gene”
→ von Hippel-Lindau tumor suppressor gene
Inconsistent naming of VHL gene.
MINORtypoDiscussion, paragraph 4
“pressent”
→ present
Typo in 'present'.
MINORtypoDiscussion, paragraph 4
“PRRM1”
→ PBRM1
Incorrect abbreviation for polybromo-1.
MINORconsistencyExtended Data Table 3, footnote
“In the axitinib arm, non-zero associations (two-tailed test) were hypothesized for all biomarkers.”
→ In the sunitinib arm, non-zero associations (two-tailed test) were hypothesized for all biomarkers.
The footnote refers to the sunitinib arm but incorrectly says 'axitinib arm'.
MINORclarityDiscussion, paragraph 4
“The pressent dataset showed that PBRM1 mutation tended to be associated with improved outcomes, PRRM1 mutation does not seem to be a reliable biomarker”
→ The present dataset showed that PBRM1 mutation tended to be associated with improved outcomes; however, PBRM1 mutation does not seem to be a reliable biomarker
Run-on sentence with typo.
MINORtypoDiscussion, paragraph 8
“pressent”
→ present
Typo: 'pressent' should be 'present'.
MINORtypoDiscussion, paragraph 8
“PRRM1”
→ PBRM1
Typo: 'PRRM1' should be 'PBRM1'.

Kaimen Rigor Action ItemsFinal pass before journal submission

This published work demonstrates a generally robust methodology, but an informed reader should be aware of the reporting gaps in study design details, participant demographics, ethical approval specifics, and the lack of public deposition for sequencing data and analysis code. These issues, along with several minor copyedit errors, suggest that a formal erratum or a more detailed supplementary information document would enhance the paper's completeness and reproducibility for future researchers.

1.
HIGHreporting
In the Methods section, specify the exact randomization method used (e.g., computer-generated random sequence, block sizes) to enhance reproducibility.
The specific method of random sequence generation is a critical detail for assessing the rigor of a randomized controlled trial.
2.
HIGHreporting
In the Methods section, provide a rationale for the open-label design of the trial.
Explaining why blinding was not feasible or appropriate is important for transparency in clinical trial design.
3.
HIGHreporting
In the Methods section, include an a priori sample size or power calculation for the exploratory biomarker analyses, or explicitly state that these analyses were not formally powered.
Clarity on power analysis is crucial for interpreting the statistical significance and generalizability of exploratory findings.
4.
HIGHreporting
In Table 1 or a supplementary table, report additional demographic details such as weight, race/ethnicity, and key comorbidities.
Comprehensive demographic reporting is essential for assessing the generalizability and potential disparities in clinical trial results.
5.
HIGHethics
In the Methods section under 'Inclusion and ethics', name the specific institutional review board(s) that approved the trial and provide the protocol number.
Specific identification of ethics approvals enhances transparency and traceability of ethical oversight.
6.
HIGHrigor
In the Methods section, state the manufacturer/source for each investigational drug (pembrolizumab, axitinib, sunitinib).
Identifying the source of key reagents is fundamental for reproducibility and proper attribution.
7.
HIGHdata code
Deposit the RNA-seq and WES data in a public repository (e.g., GEO, dbGaP) and provide accession numbers in the Data Availability section.
Public deposition of sequencing data is crucial for data transparency, re-analysis, and reproducibility of biomarker findings.
8.
HIGHdata code
Share any custom analysis code in a public repository (e.g., GitHub, Zenodo) with a permanent identifier and include the link in the Data Availability section.
Sharing custom code is essential for computational reproducibility and allows other researchers to verify or build upon the analyses.
9.
MEDIUMreporting
In the Methods section, add a statement on how outliers were handled in the biomarker analyses, or clarify that no outliers were excluded.
Explicitly stating outlier handling procedures improves transparency and allows for critical evaluation of data processing.
10.
MEDIUMreporting
In the Methods section, add a brief note on the verification of statistical assumptions (e.g., proportional hazards) for the Cox models.
Reporting on assumption checks for statistical models enhances the credibility of the analysis.
11.
LOWcopyedit
Correct all instances of 'pressent' to 'present' in the Discussion section.
Typographical errors detract from the professionalism and readability of the manuscript.
12.
LOWcopyedit
Correct all instances of 'PRRM1' to 'PBRM1' in the Discussion section.
Accurate gene nomenclature is essential for scientific precision and clarity.
13.
LOWcopyedit
Ensure consistent naming of the 'von Hippel-Lindau tumor suppressor gene' throughout the manuscript, especially in Extended Data Fig. 4-7.
Consistency in terminology improves readability and avoids confusion.
14.
LOWcopyedit
In Extended Data Table 3, footnote, correct 'In the axitinib arm' to 'In the sunitinib arm'.
Accuracy in table footnotes is important for correct interpretation of data.
15.
LOWcopyedit
Revise the sentence in Discussion, paragraph 4, to improve clarity and correct the run-on structure.
Clear and grammatically correct sentences enhance the overall quality of writing.

Full text (70,578 chars)Analyzed: Aug 10, 2026

RigorReviewer 1

RigorReviewer 2

RigorReviewer 3

RigorPre-Submission Reviewer

CopyeditAgent Scribe

CitationsCitation Integrity

StatisticsStatistics Consistency

ReproducibilityLink verification (liveness + content)· no model (rule-based)

PreregistrationDeterministic check· no model (rule-based)

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.”

How stars are deducted

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.

The eight reporting dimensions

Scientific PremisePrior work, rationale, and how it addresses gaps.

Study Design & RigorRandomization, blinding, power, controls, replication.

Biological VariablesSex, age, species/strain, housing, demographics.

Ethical ApprovalsIRB/IACUC approval, informed consent, regulatory compliance.

Key ResourcesAntibodies, cell lines, organisms, software — identified with persistent IDs.

Statistical ReportingNamed tests, exact p-values, effect sizes, plausible summary stats.

Data & Code AvailabilityAvailability statements, repositories, accession numbers, shared code.

Reporting TransparencyMethods, registration, guidelines, outcomes, limitations, funding/COI.

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.

Pembrolizumab plus axitinib versus sunitinib for advanced clear cell renal cell carcinoma: 5-year survival and biomarker analyses of the phase 3 KEYNOTE-426 trial

DOI

10.1038/s41591-025-03867-5

Record issued

2026-08-10

Engine

7.29.0

Exported

2026-09-22