PublishedNovember 6, 2026
JournalJAMA Oncology
TrialIATCCG retrospective cohort with SEER validation (not a registered trial)

Proposed Thyroid-Specific TNM System Reclassifies 36.6% of Poorly Differentiated Thyroid Cancers, With Level VII Nodes Upgraded to N1c and a Modest Gain in Discrimination

How this article was produced

How this article was produced

This article was written about a single named primary publication, with AI assistance for literature discovery, drafting and editing. AI tools are not the author and are not responsible for the conclusions.

How the draft was prepared

  • Discovery: PubMed is searched for oncology publications from the previous 14 days in a fixed list of major journals — JAMA Oncology, The Lancet Oncology, Journal of Clinical Oncology, Annals of Oncology, Nature Reviews Clinical Oncology, New England Journal of Medicine, Nature Medicine, Cancer Discovery and Clinical Cancer Research — falling back to a named set of open-access journals only if that list yields nothing. Candidates are ranked by how completely they report quantitative results, and studies already covered are excluded.
  • Source: the article is drafted from the open full text in PubMed Central where one exists. Where only the abstract is available, it is supplemented with cited external sources — the trial registry record, conference abstracts, sponsor releases — and facts drawn from those sources are labelled as such.
  • Numbers are reported exactly as the source gives them, with confidence intervals and p-values, and are not rounded. Data the source does not report is stated as not specified rather than estimated.
  • Subgroup statistics are reported only where the source states them explicitly, and quotations are reproduced verbatim.
  • Benchmarks from other studies are labelled as cross-study comparisons and never presented as results of the trial itself.
  • Validation: every factual claim in the draft is checked against the source, and key claims are spot-checked against independent sources such as the trial registry. Fixable discrepancies are corrected before the draft is handed on for publication.

These steps are instructions followed by an AI system, not checks enforced by software. The validation step is itself performed by AI rather than a person, and its verdict does not by itself stop an article from being published. This describes the current workflow; some earlier articles predate parts of it.

Checks completed before publication

  • The primary publication is recorded in structured form — journal, publication date, DOI, PubMed ID and PMC ID where they exist — rather than only as prose in a reference list.
  • Every cited DOI is confirmed to resolve.
  • Every cited PubMed ID is confirmed to exist in PubMed.
  • The cited DOI is cross-checked against the DOI that PubMed records for that PubMed ID, so a citation cannot point at a different paper from the one the article describes.
  • Trial registry identifiers are format-checked and linked to their ClinicalTrials.gov record.
  • Every reference must carry a resolvable identifier or link, unless it is explicitly labelled as a funding note, press release or background statistic.
  • The article must contain a limitations discussion specific to the study — what it could not establish — not a generic disclaimer.
  • Absolutist claims such as “proves”, “guarantees”, “will cure” or “completely effective” are rejected.
  • HTML markup and entities that would render as literal text — such as an unsupported tag or " " — are rejected outside the reference list.
  • Structural checks confirm the article has a unique URL and complete metadata.

These run automatically on every article. If any of them fails, the build fails and the article is held rather than published.

What these checks do not cover

These checks confirm that the sources are real and correctly cited. They do not confirm that the analysis drawn from them is sound. This article has not been read against its primary source by a human editor, and should be read as a signpost to the primary publication rather than a substitute for it.

Evidence note

This article summarizes research findings and does not provide medical advice or treatment recommendations. Investigational findings, preliminary results, and preclinical evidence should not be interpreted as established clinical care. Always consult the cited primary source, and a qualified clinician, before acting on anything described here.

Corrections

Report a factual error or a sourcing concern through our corrections policy. Substantive corrections are dated and described on the article page. Our editorial policy explains how these articles are sourced, verified, and corrected.

Article Summary
  • In an international cohort of 1010 patients with poorly differentiated thyroid carcinoma (PDTC) from 91 centers in 14 countries, the authors propose a PDTC-specific TNM staging system that removes the 55-year age cutoff and classifies level VII (upper mediastinal) nodal disease as a new N1c category. The proposed system reclassified 370 patients (36.6%) relative to TNM-8 (DTC criteria). Level VII lymph node metastasis carried the worst 5-year overall survival, 45.2% (95% CI, 33.4-61.2).
  • The C statistic was higher for the proposed system in the IATCCG derivation cohort (0.74 vs 0.71; difference, 0.03; 95% CI, 0.01-0.04) and numerically higher in the SEER validation cohort (0.76 vs 0.74; difference, 0.02; 95% CI, -0.01 to 0.06), where the interval crossed zero. The study is retrospective and drawn from high-volume referral centers, and the system is a proposal rather than an adopted standard.

Why This Study Matters

Poorly differentiated thyroid carcinoma (PDTC) accounts for 1% to 3% of all thyroid cancers, yet it is described as the predominant cause of mortality among nonanaplastic thyroid malignant tumors. The eighth AJCC/UICC TNM staging system (TNM-8) has no dedicated PDTC framework, so clinicians fall back on the differentiated thyroid cancer (DTC) criteria. The authors argue that this inaccurately stratifies PDTC risk and focus on two contentious features: the 55-year age cutoff and the classification of level VII (upper mediastinal) nodal disease as mild N1a.

1
Level VII lymph node metastasis carried the worst 5-year overall survival (OS): 45.2% (95% CI, 33.4-61.2).
2
The proposed PDTC system reclassified 370 patients (36.6%) of the IATCCG cohort relative to TNM-8 (DTC criteria).
3
C statistic was higher for the proposed system in IATCCG (0.74 vs 0.71; difference, 0.03; 95% CI, 0.01-0.04) and numerically higher in SEER (0.76 vs 0.74; difference, 0.02; 95% CI, -0.01 to 0.06).
Study at a glance

1010 patients in the IATCCG derivation cohort; 91 centers in 14 countries; 170 patients in the SEER validation cohort; 341 deaths (33.8%) in IATCCG; median follow-up 41 months (IQR, 15-85).

Study Design

This was a multicenter cohort study reported under STROBE. Informed consent was waived because of the retrospective design, use of deidentified data, and absence of interventions. The IATCCG cohort included consecutive patients with PDTC who underwent initial surgery at 91 centers in 14 countries from January 1984 to January 2026, and was used for derivation and internal evaluation. External validation used the SEER 17 Registries (November 2024 submission), selecting patients diagnosed from 2004 to 2015 with ICD-O-3 codes C73.9 and 8337/3. The primary end point was OS, defined as time from initial surgery to death from any cause. Analyses ran from December 2025 to June 2026. Randomization, blinding, and dosing do not apply to this design.

The stage groupings came from an exhaustive enumeration of 16 738 T-modified N combinations in 791 patients with M0 PDTC (201 deaths), each fitted into a multivariable Cox model. TNM-8 had an Akaike information criterion of 4020.22, higher than every enumerated configuration; the optimal model had 3969.81.

Patient Population

In the IATCCG cohort, 610 patients (60.4%) were female and 400 (39.6%) were male, with a median (IQR) age at diagnosis of 58 (46-68) years. At diagnosis, 219 patients (21.7%) had distant metastasis (M1) and 187 (18.5%) were classified as stage IV; 782 (77.4%) underwent total thyroidectomy. In the SEER cohort, 85 (50.0%) were female and 85 (50.0%) were male, median (IQR) age was 61 (49-72) years, 32 (18.8%) had M1 disease, 33 (19.4%) had TNM-8 stage IV disease, and 153 (90.0%) had total thyroidectomy. Median (IQR) follow-up was 92 (42-154) months, with 106 deaths (62.4%).

Primary Endpoint Results: Overall Survival by Stage

Under TNM-8 (DTC criteria), the IATCCG cohort showed inadequate separation between stages II and III and between stages IVA and IVB. The proposed system produced a more clearly delineated, progressive decline in 5-year OS. In SEER, TNM-8 again showed overlap, while the proposed system reproduced a more ordered gradient.

Crucially, level VII LNM portended the worst survival, with a 5-year OS of 45.2% (95% CI, 33.4-61.2).
— Shi et al, JAMA Oncology, Results
Stage (IATCCG)TNM-8 (DTC criteria), 5-year OS (95% CI)Proposed PDTC system, 5-year OS (95% CI)
I85.1% (80.7-88.6)89.2% (83.8-92.9)
II66.6% (59.9-72.5)78.8% (72.4-83.8)
III58.0% (44.6-69.2)68.3% (57.6-76.8)
IVA38.5% (12.1-64.9)53.8% (42.0-64.2)
IVB36.7% (28.1-45.3)38.2% (30.8-45.6)
Stage (SEER validation)TNM-8 (DTC criteria), 5-year OS (95% CI)Proposed PDTC system, 5-year OS (95% CI)
I88.5% (78.3-94.1)97.1% (80.9-99.6)
II60.3% (45.5-72.3)76.8% (64.5-85.3)
III61.5% (30.8-81.8)54.5% (32.1-72.4)
IVA22.2% (3.4-51.3)46.2% (19.2-69.6)
IVB45.8% (25.6-64.0)37.5% (21.3-53.7)

Under the proposed system, the proportion of patients with stage I disease fell from 44.4% (448 of 1010) to 24.6% (248 of 1010), and stage IV rose from 18.5% (187 of 1010) to 32.4% (327 of 1010).

Subgroup and Sensitivity Analyses

The nodal-site analysis drove the proposal for a new N1c category. Of 70 patients with level VII involvement, 47 (67.1%) had concurrent lateral neck metastases, whereas only 47 of 272 (17.3%) with lateral neck disease had spread to level VII. Among level VII patients, 92.9% (65 of 70) were T3-T4 and 40.0% (28 of 70) were M1.

Nodal group (IATCCG)5-year OS (95% CI)Log-rank comparisonR2 resection rate
None/no evidence74.4% (70.4-78.7)Reference4.6% (27 of 588)
Level VI73.2% (64.4-83.3)P < .001 vs level VII5.5% (7 of 127)
Lateral neck56.7% (49.4-65.0)P = .04 vs level VII12.0% (27 of 225)
Level VII45.2% (33.4-61.2)—22.9% (16 of 70)

N0 and NX had comparable 5-year OS (75.5% [95% CI, 67.2-82.0] vs 73.7% [95% CI, 68.3-78.4]; P = .81) and were combined. Among patients with M1 disease, there was no significant prognostic difference between those older and younger than 55 years, so the age threshold was discarded. Sensitivity analyses that excluded major centers, Asia and Europe, or China were reported as giving consistent results; numerical subgroup values are in the supplement and are Not specified in source (main text).

Prognostic Discrimination

CohortProposed system, C statistic (95% CI)TNM-8, C statistic (95% CI)Difference (95% CI)
IATCCG0.74 (0.71-0.76)0.71 (0.68-0.74)0.03 (0.01-0.04)
SEER0.76 (0.71-0.81)0.74 (0.68-0.79)0.02 (-0.01 to 0.06)

Safety Profile

Not applicable. This is a staging and prognostic study, and no treatment-related adverse event data were reported. The authors note that treatment was heterogeneous across institutions and that adjuvant therapies were omitted from the multivariable covariate set.

Interpretation and Broader Context

The authors propose a PDTC-specific system that removes the 55-year age cutoff and designates level VII nodal disease as N1c, intended as a reference for the upcoming AJCC version nine. The system is a proposal rather than an adopted standard.

Limitations

The authors report several caveats. The IATCCG cohort comes solely from high-volume referral centers, and the study is limited to surgically managed patients. SEER lacks pathology review, and molecular data (such as BRAF and TERT promoter status) were not fully available. Treatment was heterogeneous across institutions, and adjuvant therapies were omitted from the multivariable covariate set. The improvement in discrimination in SEER was modest, and its C statistic difference interval (-0.01 to 0.06) crossed zero. The design is retrospective, so this study could not establish that staging by the proposed system improves clinical decisions or outcomes.

Primary source: DOI 10.1001/jamaoncol.2026.4705 · PubMed 42994809 · PMC13744029

References

  1. Shi X, Liu Q, Scholfield D, et al. Global Proposal for the TNM Staging System for Poorly Differentiated Thyroid Cancer. JAMA Oncol. Published online November 6, 2026. doi: 10.1001/jamaoncol.2026.4705. PMID: 42994809. PMCID: PMC13744029.