Broad versus limited gene panels to guide treatment in patients with advanced solid tumors: a randomized controlled trial.
Trédan O, Pouessel D, Penel N, Chabaud S, Gomez-Roca C, Delord JP, Pannier D, Brahmi M, Fabbro M, Garcia ME, Larrieu-Ciron D, Ray-Coquard I, Viala M, Italiano A, Tosi D, Cassier P, Dufresne A, Attignon V, Boyault S, Treilleux I, Viari A, Pérol D, Blay JY
- DOI
- 10.1038/s41591-025-03613-x
- Record issued
- 2026-08-15
- Engine
- 7.39.0
- Exported
- 2026-09-20
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/5352b57d-5e7d-4d0b-bad5-9705045d3b63 is authoritative.
How this rating was calculated
- IntegrityIntegrity concern−0.5★
- CitationsUnresolved reference ×2−0.5★
- StatisticsPrinted percentage does not match its own count (capped) ×2−0.25★
- Declared data/code links were not checked for liveness or content.
- 01Printed percentage does not match its own count
63.3% is unattainable for n=339 (nearest: 63.1, 63.4%)
“214 (63.3%)”
Table 2B - 02Printed percentage does not match its own count
4.2% is unattainable for n=171 (nearest: 4.1, 4.7%)
“7 (4.2%)”
Table 1
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.
The paper is a well-conducted multicenter randomized controlled trial with rigorous design, clear reporting of ethics, demographics, and statistical methods, and transparent data availability. Minor reporting gaps include the lack of an explicit CONSORT statement and a few copyedit issues such as a duplicate row in Table 1 and unspecified software versions.
Both reviewers classified the study as interventional, which is consistent with the randomized controlled trial design. The evaluation covered all eight dimensions; several sub-criteria were marked not applicable (e.g., animal housing, cell line authentication) due to the human clinical trial context. The statistics verification component checked only a subset of tests, and the two inconsistent findings were not specified, so they were not treated as demonstrable errors.
Numerical inconsistencies
2 findings · worst mediumValues 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.
- Printed percentage does not match its own countRecomputed
- Internal contradictions in the reported numbersAssessed
Recomputed 2 tests: 2 consistent, 0 inconsistent; 2 via agent-written checks. 2 printed percentages that do not match their own count.
- PERCENT63.3% is unattainable for n=339 (nearest: 63.1, 63.4%)
“214 (63.3%)”
Table 2B - PERCENT4.2% is unattainable for n=171 (nearest: 4.1, 4.7%)
“7 (4.2%)”
Table 1
- CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Primary endpoint McNemar test: F1CDX+CTL- vs F1CDX-CTL+ discordant pairs
“MBRTs were exclusively identified with one panel in 84 (24.8%) patients (F1CDX+/CTL−: 67, 19.8%; F1CDX−/CTL+: 17, 5.0%)”
Taken as given: The discordant pairs are 67 (F1CDX+CTL-) and 17 (F1CDX-CTL+).; The McNemar test is equivalent to a chi-square test on discordant pairs with 1 degree of freedom.; The p-value reported is two-sided.Method: McNemar test computed as chi-square on discordant pairs (67 vs 17) using pChi2x2 with cell counts (67,108,17,147) to approximate the paired test.How we recomputed it: pChi2x2(67, 108, 17, 147) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 2Secondary endpoint McNemar test for MBRT initiated: discordant pairs F1CDX+CTL- vs F1CDX-CTL+.
“A total of 27 (8.0%) patients had MBRTs initiated based on both panels, 3 (0.9%) additional patients had MBRTs initiated exclusively with CTL and 21 (6.2%) additional patients had MBRTs initiated exclusively with F1CDX”
Taken as given: Discordant pairs are 21 (F1CDX+CTL-) and 3 (F1CDX-CTL+).; McNemar test is approximated by chi-square on discordant pairs.; p-value is two-sided.Method: McNemar test computed as chi-square on discordant pairs (21 vs 3) using pChi2x2 with cell counts (21,27,3,288).How we recomputed it: pChi2x2(21, 27, 3, 288)
- lowinternal contradictionTable 1 lists 'Pancreatic' twice with different counts, which may be a typographical error.
Pancreatic | 14 (4.1%) | 7 (4.1%) | 6 (3.6%) ... Pancreatic | 7 (4.2%) | 0 (0%) | 1 (0.6%)
Table 1reviewer’s wording
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.
3 major claims checked against the paper's own evidence: all adequately supported.
- supportedReviewers 1, 2The F1CDX panel identified MBRTs in a significantly higher proportion of patients than the CTL panel.The primary endpoint analysis with paired data shows a 14.8 percentage point increase (P<0.001), directly supporting the claim.Evidence: Primary endpoint result: 51.6% vs 36.9%, McNemar P<0.001.
“MBRTs were identified with F1CDX in 175 (51.6%) patients and with CTL in 125 (36.9%) patients, translating to a significant increase of 14.8 percentage points ( P < 0.001) with the more comprehensive gene panel versus the more limited panel, meeting the primary endpoint.”
AbstractFind in source - supportedReviewers 1, 2No differences in clinical outcomes were observed between the two panels.The paper reports no significant differences in PFS or other clinical outcomes, consistent with the claim.Evidence: PFS analysis showed no significant difference (P=0.3409) and no differences in BOR or duration of response.
“However, no differences in clinical outcomes were observed in these patients with advanced and/or metastatic cancer in need of treatment beyond standard genomic alterations.”
AbstractFind in source - supportedReviewers 1, 2Larger gene panels increase the number of molecularly matched therapies.The increase in MBRT identification and initiation supports this claim, though clinical benefit remains unproven.Evidence: F1CDX led to more MBRTs identified and initiated (48 vs 30 patients).
“These findings illustrate the potential for larger gene panels to increase the number of molecularly matched therapies.”
AbstractFind in source
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
None foundRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
Checked — nothing surfaced.
The introduction cites prior work (ProfiLER-01, SHIVA, MPACT, MATCH, 100,000 Genomes Project) and acknowledges limitations such as low MBRT initiation rates and unclear clinical impact of large panels. The rationale for comparing F1CDX (324 genes) versus CTL (87 genes) is clearly stated. The study addresses prior limitations by using a randomized design and paired data from both panels.
“In an attempt to assess this impact, the multicenter prospective randomized study ProfiLER-02 aimed to evaluate the added value of a larger number of screened genomic biomarkers from the gene panel Foundation OneCDX (F1CDX panel, 324 genes) compared to our home-based molecular profiling panel (CTL panel with limited biomarker detection, 87 genes) to identify more MBRTs in patients with advanced cancer.”
“In an attempt to assess this impact, the multicenter prospective randomized study ProfiLER-02 aimed to evaluate the added value of a larger number of screened genomic biomarkers from the gene panel Foundation OneCDX (F1CDX panel, 324 genes) compared to our home-based molecular profiling panel (CTL panel with limited biomarker detection, 87 genes) to identify more MBRTs in patients with advanced cancer.”
Randomization used an IWRS with permuted block design, and the randomization list was generated by a statistician. Blinding was maintained for study collaborators except biologists. A sample size calculation was performed with 98% power. Inclusion/exclusion criteria are detailed. The analysis population (ITT and per-protocol) is defined. Outlier handling is addressed through the ITT principle and sensitivity analysis.
“Registration and randomization used an interactive web response system (IWRS, Ennov) with a permuted block design. A statistician was in charge of generating the randomization list.”
“At randomization request, the IWRS automatically sent an email with the NGS panel assigned at MTB#1 to biologists only. Other study collaborators remained blinded.”
“Registration and randomization used an interactive web response system (IWRS, Ennov) with a permuted block design. A statistician was in charge of generating the randomization list.”
“At randomization request, the IWRS automatically sent an email with the NGS panel assigned at MTB#1 to biologists only. Other study collaborators remained blinded.”
Sex is reported (54.9% female), age (median 57 years), metastatic status, and tumor types are detailed in Table 1. Since both sexes are enrolled, sex_justified is not applicable. Demographics are adequately reported. Species/strain and housing conditions are not applicable for a human trial.
“In the ITT population, female patients ( n = 186) accounted for 54.9% of the population. At inclusion, the median age was 57 (19–85) years and 222 (65.5%) patients had metastasis.”
“In the ITT population, female patients ( n = 186) accounted for 54.9% of the population. At inclusion, the median age was 57 (19–85) years and 222 (65.5%) patients had metastasis.”
The study obtained approval from the Ethics Committee of Lyon Sud-Est IV on 23 May 2017 and authorization from the national competent authority. All patients provided written informed consent. The study was conducted in accordance with Good Clinical Practice guidelines and the Declaration of Helsinki. Regulatory compliance is explicitly stated.
“This multicenter randomized prospective study, ProfiLER-02, was performed in 12 institutions after obtaining approval from the Ethics Committee of Lyon Sud-Est IV on 23 May 2017 and after receiving authorization from the national competent authority on 15 May 2017.”
“All patients provided written informed consent.”
“This study was conducted in accordance with the Good Clinical Practice guidelines of the International Conference on Harmonization and the Declaration of Helsinki, as well as relevant French and European laws and directives.”
“This multicenter randomized prospective study, ProfiLER-02, was performed in 12 institutions after obtaining approval from the Ethics Committee of Lyon Sud-Est IV on 23 May 2017 and after receiving authorization from the national competent authority on 15 May 2017.”
“All patients provided written informed consent.”
“This study was conducted in accordance with the Good Clinical Practice guidelines of the International Conference on Harmonization and the Declaration of Helsinki, as well as relevant French and European laws and directives.”
The F1CDX panel is identified as provided by Foundation Medicine. The CTL panel is described with its gene list and the kit used (Forma Pure RNA kit, Beckman Coulter, # C19158). Sequencing platforms and software versions are specified (e.g., FastQC v0.11.5, SAMtools v1.3.1, R v3.5.3, Annovar v02-2016, VEP release 92). Antibodies, cell lines, and mycoplasma testing are not applicable for this clinical trial.
“Statistical analyses were performed using R v3.5.3 (R Foundation for Statistical Computing). Mutations were annotated with Annovar (v02-2016) and VEP (release 92).”
“The quality of sequencing data was checked using FastQC (v0.11.5) and SAMtools (v1.3.1) for DNA and FastQC and RSeQC for RNA. Statistical analyses were performed using R v3.5.3 (R Foundation for Statistical Computing).”
The primary analysis used McNemar test for paired proportions, and p-values are reported exactly (P < 0.001). Effect sizes are reported as percentage point differences with 95% CIs for PFS. Statistical software is identified (SAS v9.4). Data presentation includes Kaplan-Meier curves and per-group n. Mathematical plausibility checks were not applicable due to large N and continuous outcomes.
“corresponding to a significant increase in MBRT identification of 14.8 percentage points using F1CDX versus CTL (McNemar test: P < 0.001)”
“Statistical analyses were performed using SAS software v9.4 or later.”
“corresponding to a significant increase in MBRT identification of 14.8 percentage points using F1CDX versus CTL (McNemar test: P < 0.001)”
“Statistical analyses were performed using SAS software v9.4 or later.”
The data availability statement provides a specific repository (Zenodo) and DOI (10.5281/zenodo.14794347). Since the data are anonymized individual-level data, repository_deposit and accession_numbers are applicable and adequate. Code sharing is not applicable as no bespoke code is mentioned.
“Anonymized individual data, including selected actionable gene alterations, detailed alterations, ESCAT class, pharmaceutical class proposed by the MTB and initiated, and MBRTs initiated, are available via Zenodo at 10.5281/zenodo.14794347 (ref. ).”
“Anonymized individual data, including selected actionable gene alterations, detailed alterations, ESCAT class, pharmaceutical class proposed by the MTB and initiated, and MBRTs initiated, are available via Zenodo at 10.5281/zenodo.14794347 (ref. ).”
The trial is registered at ClinicalTrials.gov (NCT03163732). Methods are comprehensive. Limitations are explicitly discussed, including the small number of MBRT initiations and the overrepresentation of certain tumor types. Conclusions appropriately state that no differences in clinical outcomes were observed. Funding and COI statements are provided.
“ClinicalTrials.gov registration: NCT03163732”
“Some limitations have to be underlined. These results should be interpreted with caution considering the small number of patients in whom an MBRT was initiated (the main limitation resulting from the critical lack of accessibility of MBRTs) and the aggressiveness of most of the included cancers (gliomas and sarcomas).”
“However, no differences in clinical outcomes were observed in this population of patients with advanced and/or metastatic cancer in need of treatment beyond standard genomic alterations.”
“ClinicalTrials.gov registration: NCT03163732”
“Some limitations have to be underlined. These results should be interpreted with caution considering the small number of patients in whom an MBRT was initiated (the main limitation resulting from the critical lack of accessibility of MBRTs) and the aggressiveness of most of the included cancers (gliomas and sarcomas).”
“This work received partial funding from Roche Pharma AG. The F1CDX panel was provided for free by Foundation Medicine.”
Registered (3 IDs: ClinicalTrials.gov). No reporting guideline cited.
Broken references and links
1 finding · worst lowReferences 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.
- References not resolvable to a published paperRecomputed
Checked 26 references by DOI: 24 verified — 2 DOI unresolved.
- UNRESOLVED10.1038/s41588-024-01885-2Analysis of 10,478 cancer genomes identifies candidate driver genes and opportunities for precision oncologyCited DOI does not resolve to any Crossref record.
- UNRESOLVED10.5281/zenodo.14794347PROFILER02Cited DOI does not resolve to any Crossref record.
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, clarity.
- MINORconsistencyTable 1“Pancreatic | 14 (4.1%) | 7 (4.1%) | 6 (3.6%)”→ Check if the row 'Pancreatic' is duplicated; there is another 'Pancreatic' row later with different values.Potential duplicate row in tumor type table.
- MINORconsistencyTable 1“Pancreatic | 7 (4.2%) | 0 (0%) | 1 (0.6%)”→ Verify the correct count for pancreatic cancer; the two rows may be inconsistent.Duplicate pancreatic row with conflicting numbers.
- MINORclarityMethods, Procedures“The quality of sequencing data was checked using FastQC (v0.11.5) and SAMtools (v1.3.1) for DNA and FastQC and RSeQC for RNA.”→ Specify the version of RSeQC for completeness.RSeQC version not provided.
- MINORclarityResults, Description of the cohort“A total of 233 deaths occurred during the study period.”→ Clarify whether this includes the 17 deaths during screening and 58 after randomization.Ambiguity in death counts.
The published work is robust and well-reported; an informed reader should weigh the minor reporting gaps (missing CONSORT statement, duplicate Table 1 row, unspecified RSeQC version) and the unverified subset of statistical tests. No erratum is warranted for the core findings, but the authors should consider issuing a correction for the Table 1 duplication and clarifying the death counts.
- 1.HIGHcopyeditIn Table 1, remove the duplicate 'Pancreatic' row and verify the correct counts for pancreatic cancer, ensuring consistency with the text.The duplicate row with conflicting numbers is an internal contradiction that could confuse readers and undermine data integrity.
- 2.HIGHreportingIn the Methods or Reporting Summary, explicitly state that the study followed the CONSORT reporting guideline and provide the checklist as supplementary material.The absence of an explicit CONSORT statement is a reporting gap that reviewers and readers expect for a randomized trial.
- 3.HIGHstatisticsIn the Results, report the exact p-values for secondary endpoints (e.g., PFS) rather than only thresholds, and clarify the handling of missing data for the primary endpoint.Exact p-values and missing-data handling improve statistical transparency and reproducibility.
- 4.MEDIUMcopyeditIn Methods, Procedures, specify the version of RSeQC used for RNA quality checking.Providing software versions for all tools enhances reproducibility.
- 5.MEDIUMcopyeditIn Results, Description of the cohort, clarify whether the 233 deaths include the 17 during screening and 58 after randomization.The ambiguity in death counts could lead to misinterpretation of the survival analysis.
- 6.MEDIUMdata codeIn the Data Availability section, consider also depositing the analysis code (e.g., SAS scripts) used for statistical analyses in a public repository.Sharing analysis code enhances reproducibility, even though it is not strictly required.
- 7.MEDIUMreportingIn the Discussion, add a brief note on the generalizability of the findings given the overrepresentation of gliomas and sarcomas and the underrepresentation of common cancers.This would help readers interpret the applicability of the results to broader patient populations.
- 8.LOWotherVerify the two references flagged as not found in the registry (the cancer genomes analysis and the Zenodo dataset) and correct or replace them if they are erroneous.Unresolved references may indicate fabrication or citation errors that need correction.
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.