CoChat
Sign Up
Cited by 852Lung Cancer Treatments and MutationsCancer Genomics and DiagnosticsLung Cancer Diagnosis and Treatment

Association Between Plasma Genotyping and Outcomes of Treatment With Osimertinib (AZD9291) in Advanced Non–Small-Cell Lung Cancer

Open access
Oxnard, Geoffrey R., Thress, Kenneth S., Alden, Ryan S., Lawrance, Rachael, Paweletz, Cloud P., Cantarini, Mireille, Yang, James Chih-Hsin, Barrett, J. Carl +1 more·2016·Journal of Clinical Oncology
DOI 10.1200/jco.2016.66.7162PMC PMC5035123
Open this paper in CoChat — free account

Abstract

Purpose: Third-generation epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) have demonstrated potent activity against TKI resistance mediated by EGFR T790M. We studied whether noninvasive genotyping of cell-free plasma DNA (cfDNA) is a useful biomarker for prediction of outcome from a third-generation EGFR-TKI, osimertinib. Methods: Plasma was collected from all patients in the first-in-man study of osimertinib. Patients who were included had acquired EGFR-TKI resistance and evidence of a common EGFR-sensitizing mutation. Genotyping of cell-free plasma DNA was performed by using BEAMing. Plasma genotyping accuracy was assessed by using tumor genotyping from a central laboratory as reference. Objective response rate (ORR) and progression-free survival (PFS) were analyzed in all T790M-positive or T790M-negative patients. Results: Sensitivity of plasma genotyping for detection of T790M was 70%. Of 58 patients with T790M-negative tumors, T790M was detected in plasma of 18 (31%). ORR and median PFS were similar in patients with T790M-positive plasma (ORR, 63%; PFS, 9.7 months) or T790M-positive tumor (ORR, 62%; PFS, 9.7 months) results. Although patients with T790M-negative plasma had overall favorable outcomes (ORR, 46%; median PFS, 8.2 months), tumor genotyping distinguished a subset of patients positive for T790M who had better outcomes (ORR, 69%; PFS, 16.5 months) as well as a subset of patients negative for T790M with poor outcomes (ORR, 25%; PFS, 2.8 months). Conclusion: In this retrospective analysis, patients positive for T790M in plasma have outcomes with osimertinib that are equivalent to patients positive by a tissue-based assay. This study suggests that, upon availability of validated plasma T790M assays, some patients could avoid a tumor biopsy for T790M genotyping. As a result of the 30% false-negative rate of plasma genotyping, those with T790M-negative plasma results still need a tumor biopsy to determine presence or absence of T790M.

Verified 2026-10-06 · 10 references not checked yet

AI snapshot

Written from the paper’s full text.

TL;DR

In a retrospective analysis of the phase I AURA trial, patients with T790M detected in plasma by BEAMing had osimertinib outcomes similar to those with T790M detected in tumor tissue. Because plasma genotyping had a 30% false-negative rate for T790M, patients with T790M-negative plasma still needed tumor biopsy to determine T790M status.

Key findings

  • Plasma genotyping sensitivity was 70% for T790M (111 of 158; 95% CI, 63% to 77%), 82% for exon 19 deletion (112 of 136; 95% CI, 75% to 88%), and 86% for L858R (63 of 73; 95% CI, 76% to 93%).
  • Of 58 patients with T790M-negative tumors, T790M was detected in plasma in 18 (31%); repeat testing with an alternative plasma assay confirmed T790M in 14 of 18 (78%).
  • ORR and median PFS were similar for T790M-positive plasma (ORR, 63%; PFS, 9.7 months) and T790M-positive tumor (ORR, 62%; PFS, 9.7 months).
  • Among patients with T790M-negative plasma, tumor genotyping distinguished T790M-positive tumors (ORR, 69%; PFS, 16.5 months) from T790M-negative tumors (ORR, 25%; PFS, 2.8 months).

Methods

  • Post hoc/exploratory analysis of the phase I AURA study of osimertinib in advanced EGFR-mutant NSCLC; 402 patients enrolled, 60 first-line patients excluded, 34 without a common EGFR-sensitizing mutation excluded, leaving 308 patients.
  • Plasma cfDNA was collected before osimertinib and genotyped by BEAMing for T790M, L858R, and exon 19 deletion variants; positivity thresholds were ≥ 0.04% allelic fraction for exon 19 deletion or L858R and ≥ 0.06% for T790M; discordant plasma/tumor T790M results were retested with cobas plasma EGFR assay or ddPCR.
  • Outcomes included confirmed ORR, maximum change in target lesion size, and PFS by RECIST version 1.1 per investigator; analyses used Clopper-Pearson exact CIs, Fisher’s exact or χ2 tests, and Kaplan-Meier estimates with log-rank tests.

Limitations

  • The paper states the analysis was not preplanned, the study population was not fully representative of all acquired EGFR-TKI resistance because of intentional enrichment for T790M-positive patients, BEAMing was not performed under Clinical Laboratory Improvement Amendments conditions, and prospective validation is needed.
  • Evident from the text: this was a retrospective post hoc analysis, and some key subgroups were small, such as the 18 patients with T790M-positive plasma but T790M-negative tumor genotyping.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1200/jco.2016.66.7162

Cites

13 of 23 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 18 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § RESULTS

    T790M was detected in 80% of patients with a detectable EGFR-sensitizing mutation in plasma versus only 5% of patients without a detectable EGFR-sensitizing mutation in plasma (P<.001).

    “T790M was detected in 80% of patients (110 of 137) with versus only 5% of patients (1 of 21) without a detectable EGFR-sensitizing mutation in plasma (P<.001, Fig A2B)”
    Quote in the paperSubstantial claim
  • Findingown data
    § RESULTS

    Specificity of plasma genotyping for L858R was 97%.

    “Plasma genotyping resulted in rare false positives for exon 19 deletions and L858R (specificity, 98% and 97%, respectively; Table 1, Appendix Fig A2C, online only)”
    Quote in the paperNot a substantial claim
  • Findingown data
    § RESULTS

    Sensitivity of plasma genotyping for exon 19 deletion was 82% (112 of 136; 95% CI, 75% to 88%).

    “sensitivity of plasma genotyping for the known sensitizing mutation was 82% for exon 19 deletions (112 of 136; 95% CI, 75% to 88%)”
    Quote in the paperSubstantial claim
  • Findingown data
    § RESULTS

    Sensitivity of plasma genotyping for EGFR-sensitizing mutations was significantly higher in patients with liver metastases (94% v 79%; P = .008).

    “sensitivity of plasma genotyping for EGFR-sensitizing mutations was significantly higher in patients with liver metastases (94% v 79%; P = .008)”
    Quote in the paperNot a substantial claim
  • Findingown data
    § DISCUSSION

    Those 40 patients with T790M-negative plasma and T790M-negative tumor genotyping had a median PFS of 2.8 months.

    “Of 102 patients from this study with T790M-negative plasma, 45 had tumor genotyping that was positive for T790M, and these patients had a median PFS of 16.2 months, whereas 40 patients had tumor genotyping that was negative for T790M and a median PFS of 2.8 months.”
    Quote in the paperSubstantial claim
  • May combine two claims. 21
  • Classified as probably not a substantial claim. 8

  • May combine two claims.

    Findingown data
    § DISCUSSION

    In patients treated with osimertinib on the basis of T790M-positive tumor genotyping, the ORR was 62% and median PFS was 9.7 months.

    “If plasma genotype is positive for T790M, this may obviate the need for a biopsy—this predicts for excellent outcomes on osimertinib (ORR, 63%; median PFS, 9.7 months), similar to that observed when treating with osimertinib on the basis of tumor genotyping results (ORR, 62%; median PFS, 9.7 months).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § DISCUSSION

    In patients with T790M-positive plasma genotyping treated with osimertinib, the ORR was 63% and median PFS was 9.7 months.

    “If plasma genotype is positive for T790M, this may obviate the need for a biopsy—this predicts for excellent outcomes on osimertinib (ORR, 63%; median PFS, 9.7 months), similar to that observed when treating with osimertinib on the basis of tumor genotyping results (ORR, 62%; median PFS, 9.7 months).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § RESULTS

    Median PFS was 8.2 months (95% CI, 5.3 to 10.9 months) in patients with T790M-negative plasma genotyping, not significantly lower than in patients with T790M-positive plasma (P = .188).

    “median PFS was not significantly lower in those with T790M-negative plasma genotyping (8.2 months; 95% CI, 5.3 to 10.9 months; P = .188; Fig 3B).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § RESULTS

    ORR was 64% (21 of 33; 95% CI, 45% to 79%) in patients with T790M-negative plasma and no detectable EGFR mutations, higher than in those with detectable TKI-sensitive EGFR mutation (P = .019).

    “In contrast, in 33 patients with no EGFR mutations detected (T790M negative/sens negative), ORR was higher at 64% (21 of 33; 95% CI, 45% to 79%; P = .019)”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § RESULTS

    Median PFS was 9.7 months (95% CI, 8.3 to 11.1 months) in patients positive for T790M in plasma, similar to outcomes in patients positive for T790M on tumor biopsy.

    “a 9.7-month median PFS (95% CI, 8.3 to 11.1 months; Fig 3B), which is similar to outcomes in patients positive for T790M on tumor biopsy.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § RESULTS

    ORR was 46% (47 of 102; 95% CI, 36% to 56%) in patients with T790M-negative plasma genotyping, lower than in patients with T790M-positive plasma (P = .011).

    “ORR was surprisingly high at 46% (47 of 102; 95% CI, 36% to 56%) in patients with T790M-negative plasma genotyping; however, this was still lower than in patients with T790M-positive plasma (P = .011; Fig 2D)”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § RESULTS

    Median PFS was 15.2 months (95% CI, 11.0 to 17.9 months) in patients with T790M-negative plasma and no detectable EGFR mutations, longer than in those with detectable TKI-sensitive EGFR mutation (P = .002).

    “and median PFS was longer at 15.2 months (95% CI, 11.0 to 17.9 months; P = .002; Appendix Fig A3)”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § RESULTS

    ORR was 62% (108 of 173; 95% CI, 54% to 70%) in patients with T790M-positive tumors, higher than the 26% ORR (15 of 58; 95% CI, 15% to 39%; P < .001) in patients with T790M-negative tumors.

    “ORR was 62% (108 of 173; 95% CI, 54% to 70%) in patients with T790M-positive tumors, higher than the 26% ORR (15 of 58; 95% CI, 15% to 39%; P < .001) in patients with T790M-negative tumors”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § DISCUSSION

    In patients with T790M-negative/sens-positive plasma genotyping, ORR to osimertinib was 38% and median PFS was 4.4 months.

    “In patients with T790M-negative/sens-positive plasma genotyping, ORR to osimertinib was 38% and median PFS was 4.4 months.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § DISCUSSION

    Those 18 patients with T790M-positive plasma and T790M-negative tumor genotyping had a response rate of 28% (5 of 18) and median PFS of 4.2 months.

    “Of 164 patients from this study with T790M-positive plasma genotyping, 18 had T790M-negative tumor genotyping, presumably as a result of heterogeneous presence of resistance mutation as a minor clone, and these patients had a response rate of 28% (5 of 18) and median PFS of 4.2 months.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § DISCUSSION

    In patients with T790M-negative/sens-negative plasma results, high ORR and prolonged PFS was observed.

    “In patients with T790M-negative/sens-negative plasma results, high ORR and prolonged PFS was observed.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation
    § INTRODUCTION

    Third-generation mutant-selective EGFR-TKIs have activity against cells that harbor TKI-sensitive EGFR mutations, and this activity is maintained in the presence of the T790M resistance mutation.

    “Standard of care is rapidly changing with the development of third-generation, mutant-selective EGFR-TKIs that have activity against cells that harbor TKI-sensitive EGFR mutations. Unlike conventional EGFR-TKIs, this activity is maintained in the presence of T790M resistance mutation.8,9”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [8] source

      Not checked in the free run.

    • cites [9] source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § DISCUSSION

    Plasma and tumor genotyping can have complementary roles for T790M testing, with plasma genotyping as the initial step and tumor biopsy as supplementary.

    “Our data suggest that plasma and tumor genotyping can have complementary roles for T790M testing, where plasma genotyping could be the initial step and a biopsy for tumor genotyping could be supplementary (Fig 5B).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation
    § INTRODUCTION

    Tumor genotyping for EGFR T790M requires a biopsy after resistance develops, which carries risks, can delay subsequent therapy, and may not always be feasible.

    “This represents a clinical challenge because tumor genotyping for EGFR T790M requires a biopsy to be performed after resistance develops, which is a procedure that carries risks, can delay subsequent therapy, and may not always be feasible.18”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [18] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation
    § INTRODUCTION

    At least five third-generation EGFR-TKIs are in clinical development, and all have reported objective response rates exceeding 45% in patients with EGFR-mutant NSCLC and T790M-positive resistance.

    “At least five third-generation EGFR-TKIs are in clinical development (rociletinib, osimertinib, HM61713/BI1482694, ASP8273, and EGF816) and all have reported objective response rates (ORRs) that exceed 45% in patients with EGFR-mutant NSCLC and T790M-positive resistance.12-16”
    Quote in the paperNot a substantial claim0 of 5 sources checked
    • cites [12] source

      Not checked in the free run.

    • cites [13] source

      Not checked in the free run.

    • cites [14]

      Not checked in the free run.

    • cites [15]

      Not checked in the free run.

    • cites [16]

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § RESULTS

    Heterogeneous presence of resistance mutations across disease sites, and not false positives of the BEAMing assay, account for the majority of T790M tissue-plasma discordance.

    “which suggests that heterogeneous presence of resistance mutations across disease sites, and not false-positives of the BEAMing assay, account for the majority of T790M tissue–plasma discordance.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § INTRODUCTION

    Plasma genotyping for EGFR-sensitizing mutations is an established alternative in Europe when tumor genotyping is not feasible.

    “Noninvasive genotyping of cell-free plasma DNA (cfDNA) represents an attractive alternative for detection of EGFR T790M, noting that plasma genotyping for EGFR-sensitizing mutations is an established alternative in Europe when tumor genotyping is not feasible.19”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [19] source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation
    § INTRODUCTION

    Acquired resistance to conventional EGFR-TKIs develops after a median of 10 to 12 months and is most commonly mediated by an acquired T790M mutation.

    “Resistance to conventional EGFR-TKIs develops after a median of 10 to 12 months and is most commonly mediated by an acquired mutation, T790M1-5.”
    Quote in the paperSubstantial claim0 of 5 sources checked
    • cites [1] source

      Not checked in the free run.

    • cites [2] source

      Not checked in the free run.

    • cites [3] source

      Not checked in the free run.

    • cites [4] source

      Not checked in the free run.

    • cites [5] source

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning
    § DISCUSSION

    Given the ease and reduced risk of plasma analysis compared with an invasive biopsy procedure, data support a new paradigm for resistance management, with rapid plasma genotyping as a diagnostic option before undergoing a tumor biopsy.

    “Given the ease and reduced risk of plasma analysis compared with an invasive biopsy procedure, data support a new paradigm for resistance management, with rapid plasma genotyping as a diagnostic option before undergoing a tumor biopsy.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning
    § DISCUSSION

    This analysis was not a preplanned analysis.

    “Whereas it is a large analysis of a prospective dataset, it was not a preplanned analysis, in part because of the rapid evolution of plasma genotyping technologies.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data
    § DISCUSSION

    Despite discordance, outcomes in patients with T790M-positive plasma are similar to outcomes of patients with T790M-positive tumors, which highlights that a single tumor biopsy may not be representative of the spatial heterogeneity of resistant lung cancer.

    “Despite this discordance, outcomes in patients with T790M-positive plasma are similar to outcomes of patients with T790M-positive tumors, which highlights that a single tumor biopsy may not be representative of the spatial heterogeneity of resistant lung cancer.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Resistant cancers are inherently more heterogeneous than treatment-naïve cancers.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given
    § DISCUSSION

    BEAMing was not performed under Clinical Laboratory Improvement Amendments conditions; this analysis used an investigational assay identical to the commercially available BEAMing assay.

    “Lastly, BEAMing was not performed under Clinical Laboratory Improvement Amendments conditions; this analysis used an investigational assay that is identical to the commercially available BEAMing assay.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning
    § DISCUSSION

    In a setting in which it is unclear whether tumor genotyping is representative of the entirety of the disease, clinical outcome will be the best reference standard for development of a noninvasive assay.

    “In a setting in which it is unclear whether tumor genotyping is representative of the entirety of the disease, clinical outcome will be the best reference standard for development of a noninvasive assay.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Resistant cancers are inherently more heterogeneous than treatment-naïve cancers.

  • Classified as probably not a substantial claim.

    Interpretationreasoning
    § DISCUSSION

    Patients negative for T790M in plasma should undergo a biopsy to determine T790M status because of the risks of false-negative plasma results.

    “Patients negative for T790M in plasma, however, should undergo a biopsy to determine T790M status because of the risks of false-negative plasma results.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Interpretationreasoning
    § DISCUSSION

    Biopsy to further investigate T790M-positive tumor tissue is warranted in patients with T790M-negative plasma.

    “As the plasma T790M-negative population is a mixture of true and false negatives, biopsy to further investigate the presence of T790M-positive tumor tissue is warranted (Fig 5B).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning
    § DISCUSSION

    In this example, approximately 54% of resistant patients would be T790M positive in plasma and could potentially avoid a biopsy.

    “In this example, approximately 54% of resistant patients would be T790M positive in plasma and, if treated on the basis of that result, could potentially avoid a biopsy.”
    Quote in the paperNot a substantial claim
26 of 26

  • The paper relies on this without support.

    Resistant cancers are inherently more heterogeneous than treatment-naïve cancers.

    “Resistant cancers are inherently more heterogeneous than treatment-naïve cancers and, therefore, a single tumor biopsy may not be representative of the entire resistant cancer.”

  • The paper relies on this without support.

    The BEAMing assay has negligible false positives for T790M.

    “We found that 31% of patients who are negative for T790M on central tumor genotyping have detectable T790M in plasma with a BEAMing assay that otherwise has negligible false positives, though the relative AF of plasma T790M is lower in these patients than in patients with T790M-positive plasma and tumors.”
2 of 2

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Different median PFS values for the same subgroup

    Minor

    For patients with T790M-negative plasma but T790M-positive tumors, the Results report a median PFS of 16.5 months, while the Discussion reports 16.2 months.

  • Response-evaluable count presented as the plasma-positive cohort size

    Minor

    The Results say that 164 of the 271 plasma-genotyped patients were T790M positive, but Table A1 lists 167. The 164 figure is the denominator used for ORR, not the tabulated size of the plasma-positive cohort.

  • Incorrect percentage for the 80 mg tumor-genotyping subgroup

    Minor

    Table A1 reports 90 patients at 80 mg among 237 patients with central tumor genotyping. That is 38.0%, not 40.0%.

    Reported 40 · recomputed 37.974684

3 of 3

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Oxnard, G. R., Thress, K. S., Alden, R. S., Lawrance, R., Paweletz, C. P., Cantarini, M., Yang, J. C.-H., Barrett, J. C., & Jänne, P. A.
Affiliations
  • National Taiwan University (TW)
  • Geoffrey R. Oxnard, Ryan S. Alden, Cloud P. Paweletz, and Pasi A. Jänne, Dana-Farber Cancer Institute, Boston; Kenneth S. Thress and J. Carl Barrett, AstraZeneca, Waltham, MA; Rachael Lawrance and Mireille Cantarini, AstraZeneca, Macclesfield, United Kingdom; and James Chih-Hsin Yang, National Taiwan University and National Taiwan University Hospital, Taipei, Taiwan.
ORCID
Venue
Journal of Clinical Oncology
Publisher
American Society of Clinical Oncology (ASCO)
Published
2016-10-01
Volume
34
Issue
28
Pages
3375-3382
ISSN
  • 0732-183X (ISSN-L)
  • 1527-7755
DOI
10.1200/jco.2016.66.7162
PMC
PMC5035123
OpenAlex
W2468983064
Semantic Scholar
120643ae96c264f871673844d9f7ad68e25ce532
Cited by
850
References
23
Citations by year
2026: 23 · 2025: 21 · 2024: 35 · 2023: 61 · 2022: 66 · 2021: 105 · 2020: 99 · 2019: 120 · 2018: 151 · 2017: 149 · 2016: 18 · 2014: 1
FWCI
130.49
Type
journal article
Language
en
Open access
Locations
  • Journal of Clinical Oncology (publishedVersion)
  • PubMed Central (publishedVersion)
  • PubMed (publishedVersion)
Funders
  • AstraZeneca
  • National Institutes of Health
  • National Cancer Institute (R01CA135257, U01CA210171)
Topics
  • Health Sciences › Medicine › Pulmonary and Respiratory Medicine › Lung Cancer Treatments and Mutations
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cancer Research › Cancer Genomics and Diagnostics
  • Health Sciences › Medicine › Pulmonary and Respiratory Medicine › Lung Cancer Diagnosis and Treatment
Keywords
osimertinib, T790M mutation, liquid biopsy, plasma cell-free DNA, beaming, EGFR tyrosine kinase inhibitors, non-small cell lung cancer, treatment response prediction, false negative rate, tumor biopsy, progression-free survival, objective response rate, first-in-man trial
MeSH terms
Acrylamides, Aniline Compounds, Antineoplastic Agents, Carcinoma, Non-Small-Cell Lung, DNA, Neoplasm, Genotype, Humans, Indoles, Lung Neoplasms, Piperazines, Pyrimidines, Retrospective Studies, Biomarkers, Tumor, Treatment Outcome, Disease-Free Survival, Protein Kinase Inhibitors, Kaplan-Meier Estimate, ErbB Receptors

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-06.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-06.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review