Atezolizumab plus FOLFOX for Stage III Mismatch Repair-Deficient Colon Cancer.
Sinicrope FA, Ou FS, Arnold D, Peters WR, Behrens RJ, Lieu CH, Matin K, Cohen DJ, Potter SL, Nixon AB, Kottschade LA, Kathol E, Frankel WL, Shergill A, Hsu D, Reinacher-Schick A, Mehan P, Gold PJ, Khalil MF, Zemla T, Gatten C, O'Reilly EM, Meyerhardt JA
- DOI
- 10.1056/NEJMoa2507874
- 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/f2ddd234-f13f-4f07-8f23-21b7d26909e9 is authoritative.
How this rating was calculated
- ReportingData & code availability not met−0.5★
- ReportingBiological variables partially met−0.25★
- ReportingKey resources partially met−0.25★
- 01Data and code not shared
The paper does not include a data availability statement, which is a significant omission for a clinical trial. No code sharing is mentioned, but this is not required for a standard clinical trial analysis.
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 methodologically strong phase 3 RCT with a well-founded scientific premise, rigorous randomization and powering, complete ethics reporting, and sound statistical reporting; the verification components found no statistical inconsistencies, no retracted/non-existent references, and no dead links. The principal weaknesses are the complete absence of a data availability statement (a critical trial-reporting gap) and partial reporting of biological variables (weight, race/ethnicity) and of the drug manufacturer.
Synthesis of two independent reviewer runs that converged on seven of eight dimensions; the sole divergence was biological variables (warn vs pass), resolved to warn on the strength of the jointly confirmed absence of weight and race/ethnicity. Statistics verification recomputed 2/2 reported tests consistently, but coverage is limited to tests reported with a test statistic + df or an effect estimate + CI; threshold-only and resampling-based p-values were not machine-checked and should be treated as unverified rather than confirmed. No retracted or not-found references and no dead reproducibility links were identified.
Numerical inconsistencies
None foundValues 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 2 tests: 2 consistent, 0 inconsistent; 2 via agent-written checks.
- CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 2Primary DFS hazard ratio: 0.50, 95% CI 0.35-0.73, p=0.0002
“Disease-free survival at 3 years was 86.3% (95% CI, 81.8 to 89.8) for atezolizumab plus mFOLFOX6 compared to 76.2% (95% CI, 70.9 to 80.6) for mFOLFOX6 (stratified HR, 0.50; 95% confidence interval [CI] , 0.35 to 0.73). The stratified log-rank p-value was 0.0002.”
Taken as given: The HR is from a stratified Cox model, and the CI is a 95% Wald interval.; The p-value is from the stratified log-rank test, which should be consistent with the CI-based p.; The log-rank p-value is two-sided (as stated in the paper: 'two-sided P values are reported').Method: The pCI function computes a two-sided p-value from the point estimate and 95% CI on the log scale (log=1 for HR). The expected p-value is approximately 0.0002, consistent with the reported p.How we recomputed it: pCI(0.50, 0.35, 0.73, 1) - CONSISTENTreported p = .679 · recomputed p = .674Reviewer 2Overall survival: HR=0.90, 95% CI 0.55-1.47, p=0.6787
“The 5-year overall survival rate was 89.7% (95% CI, 85.2 to 92.9) for atezolizumab plus mFOLFOX6 versus 87.9% (95% CI, 83.1 to 91.4) for mFOLFOX6 alone (stratified HR, 0.90; 95% CI, 0.55 to 1.47). Overall survival showed no statistically significant difference by study arm (stratified log-rank p value of 0.6787).”
Taken as given: The HR is from a stratified Cox model, and the CI is a 95% Wald interval.; The p-value is from the stratified log-rank test.; The p-value is two-sided.Method: The pCI function computes a two-sided p-value from the point estimate and 95% CI on the log scale. The expected p-value is approximately 0.68, consistent with the reported p.How we recomputed it: pCI(0.90, 0.55, 1.47, 1)
Overstated conclusions
1 finding · worst lowConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
- Conclusions only partially backed by the presented evidenceAssessed
6 major claims checked against the paper's own evidence: all adequately supported.
- partialReviewers 1, 2These findings support universal MMR testing in all patients with newly diagnosed colon cancer.The study demonstrates benefit of immunotherapy in dMMR, but the claim about universal MMR testing is an extrapolation beyond the study's direct evidence.Evidence: Discussion, last paragraph: 'These findings support universal MMR testing in all patients with newly diagnosed colon cancer, both to identify the Lynch syndrome and to determine eligibility for immunotherapy.'
“These findings support universal MMR testing in all patients with newly diagnosed colon cancer, both to identify the Lynch syndrome and to determine eligibility for immunotherapy.”
DiscussionFind in source - partialReviewer 1The therapeutic benefit of adjuvant atezolizumab may depend on adequate chemotherapy exposure.This is based on a pre-specified subgroup analysis, which is exploratory and not definitive.Evidence: Results, Efficacy: 'In a pre-specified subgroup analysis, we examined the duration of adjuvant chemotherapy in relationship to patient disease-free survival. When stratified by duration of mFOLFOX6 treatment, an apparent disease-free survival benefit for atezolizumab was observed among patients who received more than six cycles of mFOLFOX6 (HR 0.41, 95% CI 0.27–0.64), whereas no advantage was seen in those who received ≤6 cycles (HR 0.97, 95% CI 0.44–2.11).'
“These exploratory findings suggest, but do not prove, that the therapeutic benefit of adjuvant atezolizumab may depend on adequate chemotherapy exposure.”
Discussion ¶3Find in source - supportedReviewers 1, 2The addition of atezolizumab to mFOLFOX6 significantly improved disease-free survival in stage III dMMR colon cancer.The primary endpoint analysis shows a statistically significant improvement in DFS with HR 0.50 (95% CI 0.35-0.73, p<0.0001).Evidence: Results, Efficacy: 'Disease-free survival at 3 years was 86.3% (95% CI, 81.8 to 89.8) for atezolizumab plus mFOLFOX6 compared to 76.2% (95% CI, 70.9 to 80.6) for mFOLFOX6 (stratified HR, 0.50; 95% CI, 0.35 to 0.73). The stratified log-rank p-value was 0.0002.'
“The addition of atezolizumab to mFOLFOX6 significantly improved disease-free survival in stage III dMMR colon cancer.”
AbstractFind in source - supportedReviewers 1, 2The addition of atezolizumab to mFOLFOX6 was associated with a ten percentage point absolute reduction and a 50% reduction in the hazard of disease recurrence or death.The absolute reduction is 86.3% - 76.2% = 10.1%, and the HR is 0.50, consistent with the claim.Evidence: Results, Efficacy: '3-year disease-free survival was 86.3% (95% CI, 81.8 to 89.9) in the atezolizumab plus mFOLFOX6 arm and 76.2% (95% CI, 70.9 to 80.6) in the mFOLFOX6 arm [hazard ratio (HR) for recurrence or death, 0.50; 95% CI, 0.35 to 0.73; P<0.0001].'
“The addition of atezolizumab to mFOLFOX6 was associated with a ten percentage point absolute reduction and a 50% reduction in the hazard of disease recurrence or death compared to mFOLFOX6 alone.”
Discussion ¶1Find in source - supportedReviewers 1, 2No difference in overall survival has been observed between the study arms at this first analysis.The stratified log-rank p-value is 0.6787, indicating no statistically significant difference.Evidence: Results, Efficacy: 'Overall survival showed no statistically significant difference by study arm (stratified log-rank p value of 0.6787).'
“No difference in overall survival has been observed between the study arms at this first analysis with a median follow-up of 40.9 months, and longer follow-up is required for mature estimates.”
Discussion ¶1Find in source - supportedReviewer 1The addition of atezolizumab to mFOLFOX6 resulted in a significant reduction in recurrence or death indicating it is an effective adjuvant treatment.The primary endpoint supports this claim, as discussed.Evidence: Results, Efficacy; Discussion, paragraph 1
“In conclusion, the addition of atezolizumab to mFOLFOX6 resulted in a significant reduction in recurrence or death indicating it is an effective adjuvant treatment of patients with resected dMMR stage III colon cancer.”
DiscussionFind in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary endpoint is disease-free survival, defined as time from randomization to disease recurrence or death from any cause. This is a hard clinical outcome (a clinical event), not a surrogate biomarker. The trial directly measured this endpoint and found a significant improvement.
“The primary endpoint was disease-free survival, defined as the time from randomization to disease recurrence or death from any cause.”
- ADEQUATEEffect sizeThe treatment effect is large and clinically material: a 50% reduction in the hazard of recurrence or death (HR 0.50, 95% CI 0.35–0.73), with a 10.1 percentage-point absolute improvement in 3-year disease-free survival (86.3% vs 76.2%). This is well beyond typical minimal clinically important differences for adjuvant colon cancer and is statistically significant.
“The 3-year disease-free survival was 86.3% (95% CI, 81.8 to 89.9) in the atezolizumab plus mFOLFOX6 arm and 76.2% (95% CI, 70.9 to 80.6) in the mFOLFOX6 arm [hazard ratio (HR) for recurrence or death, 0.50; 95% CI, 0.35 to 0.73; P<0.0001].”
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
3 findings · worst highRequired 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 and code not sharedAssessed
- Biological variables underreported (sex, age, strain)Assessed
- Key resources under-identified (antibodies, cell lines, RRIDs)Assessed
The introduction cites prior evidence on dMMR colon cancer, resistance to fluoropyrimidines, and the success of immunotherapy in metastatic disease. It provides a clear rationale for testing atezolizumab in the adjuvant setting. The paper acknowledges limitations such as the study design predating the IDEA collaboration and the immaturity of overall survival data.
“Evidence indicates that 10-15% of localized colon cancers show deficient DNA mismatch repair (MMR) that results in microsatellite instability-high (MSI-H) characterized by hypermutation immune checkpoint up-regulation, and abundant predicted neoantigens.”
“The phase 3 ATOMIC study was designed to investigate whether the addition of atezolizumab, an anti-PD-L1 antibody, to standard FOLFOX chemotherapy can improve disease-free survival in patients with curatively resected, dMMR stage III colon cancer.”
“In ATOMIC, patients received 6 months of adjuvant mFOLFOX6 without the possibility of 3 months of treatment, as the study was designed before results from the IDEA collaboration, and accrual was far along at the time of the IDEA publication.”
“Evidence indicates that 10-15% of localized colon cancers show deficient DNA mismatch repair (MMR) that results in microsatellite instability-high (MSI-H) characterized by hypermutation immune checkpoint up-regulation, and abundant predicted neoantigens.”
“The phase 3 ATOMIC study was designed to investigate whether the addition of atezolizumab, an anti-PD-L1 antibody, to standard FOLFOX chemotherapy can improve disease-free survival in patients with curatively resected, dMMR stage III colon cancer.”
“In ATOMIC, patients received 6 months of adjuvant mFOLFOX6 without the possibility of 3 months of treatment, as the study was designed before results from the IDEA collaboration, and accrual was far along at the time of the IDEA publication.”
The trial uses permuted block randomization, a priori power analysis, clear eligibility criteria, and ITT analysis. The missing blinding statement is a minor reporting gap, but the trial is likely open-label due to the nature of the intervention.
“The trial was designed to randomize 700 patients, with 165 events providing an overall 90% power to detect a hazard ratio (HR) of 0.6 favoring the atezolizumab plus mFOLFOX6 arm at the one-sided significance level of 0.025.”
“The trial was designed to randomize 700 patients, with 165 events providing an overall 90% power to detect a hazard ratio (HR) of 0.6 favoring the atezolizumab plus mFOLFOX6 arm at the one-sided significance level of 0.025.”
The paper reports sex (55.1% female) and median age (64 years), but does not report weight or race/ethnicity. Health status is reported via ECOG performance status. Demographics are incomplete.
The paper states that the protocol was approved by the NCI central institutional review board, that all patients provided written informed consent, and that the trial was conducted in accordance with GCP and the Declaration of Helsinki.
“The protocol was approved by the NCI central institutional review board (IRB).”
“All the patients provided written informed consent before enrollment.”
“The trial was conducted in accordance with the Good Clinical Practice guidelines of the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use and the principles of the Declaration of Helsinki.”
“The protocol was approved by the NCI central institutional review board (IRB).”
“All the patients provided written informed consent before enrollment.”
“The trial was conducted in accordance with the Good Clinical Practice guidelines of the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use and the principles of the Declaration of Helsinki.”
For a drug trial, the applicable sub-criteria are reagents_identified (investigational products) and software_tools_identified. The SAS software is identified with version (9.04) and vendor (SAS Institute). The investigational products are described with dose and regimen, but the manufacturer/source is not explicitly stated in the methods. The funding section acknowledges Genentech for atezolizumab, but the main text should ideally include the source. This is a minor inadequacy.
“Statistical analyses were performed using SAS software, version 9.04 (SAS Institute).”
“Statistical analyses were performed using SAS software, version 9.04 (SAS Institute).”
The paper uses stratified log-rank tests and Cox models, reports HR with 95% CIs, and provides exact p-values. The data presentation includes Kaplan-Meier curves and forest plots. No arithmetic errors were detected.
“Disease-free and overall survival were compared between treatment groups using stratified log-rank tests; their hazard ratios were estimated using a stratified Cox proportional-hazards model without covariates adjustment.”
“Disease-free and overall survival were compared between treatment groups using stratified log-rank tests; their hazard ratios were estimated using a stratified Cox proportional-hazards model without covariates adjustment.”
“The stratified log-rank p-value was 0.0002.”
The only applicable sub-criterion is data_availability_statement, as repository_deposit and accession_numbers are not applicable for patient-level data, and code_sharing is not applicable because standard SAS software was used. The paper does not include any statement about data availability, not even a vague 'available on request' line. The ICMJE requires a data sharing statement for clinical trials. The absence of this statement is a clear reporting gap.
The paper provides a CONSORT diagram but does not explicitly reference the CONSORT statement. All other transparency criteria are met, including trial registration, funding disclosure, and balanced discussion of limitations.
“NCT02912559 (https://clinicaltrials.gov/ct2/show/NCT02912559)”
“ATOMIC was not designed to test the benefit of atezolizumab in the absence of standard chemotherapy.”
“NCT02912559 (https://clinicaltrials.gov/ct2/show/NCT02912559)”
“The secondary endpoints of patient-reported outcome-CTCAE (PRO-CTCAE) and quality of life will be reported separately.”
Registered (1 ID: ClinicalTrials.gov). Reporting guideline cited: CONSORT.
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 24 references by DOI: 24 verified.
Every extracted reference resolved against Crossref/OpenAlex with no retraction flags.
3 data/code links checked; 1 live.
- datahttps://clinicaltrials.gov/ct2/show/NCT02912559LIVEHTTP 200Resolves, but the content could not be matched to the paper.
- datahttps://NEJM.orgUNVERIFIEDHTTP 403Liveness indeterminate — content not checked.
- datahttps://acknowledgments.alliancefound.orgUNVERIFIEDHTTP 403Liveness indeterminate — content not checked.
Copyediting
4 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 4 minor suggestions below.
4 copyedit issues flagged: mostly consistency, typo, punctuation.
- MINORconsistencyResults, Table 2 footnote“mFOLFOX arm”→ mFOLFOX6 armInconsistent abbreviation: 'mFOLFOX' used instead of 'mFOLFOX6'.
- MINORtypoResults, Table 1 footnote“Lansky-Play performance status”→ Lansky play performance statusHyphen missing or incorrect capitalization.
- MINORconsistencyAuthor list“Sinicrope Frank A M.D.”→ Commas and periods: Sinicrope, Frank A., M.D.Author name formatting is inconsistent with standard medical journal style.
- MINORpunctuationAbstract, Results“Atezolizumab Plus FOLFOX for Stage III Mismatch Repair Deficient Colon Cancer”→ Consider adding a colon after 'FOLFOX' for clarity: 'Atezolizumab Plus FOLFOX for Stage III Mismatch Repair Deficient Colon Cancer' is fine.No change needed; title is acceptable.
The published trial report is robust on its primary efficacy analysis and, absent any detected statistical, citation, or link problems, an informed reader should not doubt the headline DFS result on the basis of this review. The most consequential gaps to weigh are the missing data availability statement (an ICMJE requirement for trials that would warrant a journal correction/amendment), the unstated blinding status, and the omissions of weight, race/ethnicity, and drug manufacturer — none of which undermines the primary efficacy finding but all of which limit reproducibility and generalizability assessment.
- 1.HIGHdata codePublish a correction/amendment adding an explicit data availability statement to the paper (e.g., data available from the Alliance for Clinical Trials in Oncology upon request via a data access committee).The paper contains no data availability statement of any kind, an ICMJE requirement for clinical trials that an informed reader cannot resolve without a stated access route.
- 2.HIGHrigorState the trial's open-label (unblinded) design explicitly in the Methods 'Trial Design and Treatment' section.Blinding status is not reported anywhere, and the differing infusion schedules make open-label the likely design, which readers must know to weigh bias risk.
- 3.HIGHrigorAdd race/ethnicity and weight (or BSA) to the baseline characteristics table (Table 1) via a correction or supplementary appendix.Both are absent; race/ethnicity limits generalizability assessment and weight is biologically relevant given BSA-based chemotherapy dosing.
- 4.HIGHrigorAdd the manufacturer/source of atezolizumab and the mFOLFOX6 components to the Methods 'Trial Design and Treatment' section (Genentech is currently acknowledged only in the funding section).Investigational products should be identified with source for reproducibility, and the current omission is a reporting gap.
- 5.MEDIUMstatisticsEither verify the proportional-hazards assumption for the stratified Cox model or explicitly justify its use given the non-parametric primary log-rank analysis.Assumption verification is not reported, an incompleteness readers may weigh when interpreting the reported HR.
- 6.MEDIUMreportingReference the CONSORT reporting guideline explicitly in the Methods (a CONSORT diagram is present but not cited).Explicitly citing the guideline strengthens transparency and aligns with trial reporting standards.
- 7.MEDIUMreportingClarify in the primary report that the PRO-CTCAE and quality-of-life secondary outcomes are deferred to separate reports (currently a single sentence in Statistical Analysis).A reader could otherwise infer incomplete outcome reporting; making the deferral explicit preempts that concern.
- 8.LOWcopyeditCorrect 'mFOLFOX arm' to 'mFOLFOX6 arm' in the Results, Table 2 footnote.The abbreviation is inconsistent with the rest of the manuscript, which consistently uses 'mFOLFOX6'.
- 9.LOWcopyeditCorrect 'Lansky-Play performance status' to 'Lansky play performance status' in the Results, Table 1 footnote.Incorrect hyphenation/capitalization of the standard Lansky play-performance term.
- 10.LOWcopyeditReformat the author name 'Sinicrope Frank A M.D.' to standard journal style (e.g., 'Sinicrope, Frank A., M.D.').Author name formatting is inconsistent with standard medical journal conventions.
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.