Work

Incomplete nodal verification limits risk estimation for intermediate size rectal neuroendocrine tumors in SEER

Year: 2026
Type: article
Study design: Observational Study
Abstract: Management of 10–20 mm rectal neuroendocrine tumors (NETs) remains uncertain. Registry analyses are vulnerable to verification bias because regional nodal status is selectively observed, particularly after local excision. We quantified the nodal-verification deficit and the resulting limits on nodal-risk inference in SEER. We analyzed adults with pathologically confirmed, non-distant, 10–20 mm rectal NETs in SEER 17 registries diagnosed during 2004–2023 who underwent strict local excision or radical resection. Surgery codes were harmonized across the 2023 coding transition. Node-evaluable status included exact examined-node counts and SEER special codes documenting aspiration, sampling, dissection, or node removal. The primary analysis reported verification rates, an NX = N0 lower estimate, complete-case estimates, Manski-type bounds, and illustrative fixed-risk scenarios. Descriptive survival analyses were restricted to 2004–2021. The primary cohort included 1,157 patients, of whom 1,008 underwent local excision and 149 radical resection. Ninety-four patients (8.1%) were node-evaluable: 13/1,008 (1.3%) after local excision and 81/149 (54.4%) after radical resection. Pathologically verified nodal positivity was present in 62 patients. Treating all unverified patients as N0 yielded a lower estimate of 5.4%, whereas the complete-case estimate was 62/94 (66.0%; 95% CI 55.9–74.7%). The conservative partial-identification interval was 5.4%-97.2%. Under purely illustrative assumptions that 5%, 7.5%, or 10% of unverified patients were node-positive, overall scenario prevalences were 10.0%, 12.2%, and 14.5%, respectively. The 2004–2021 survival cohort included 1,006 patients and 39 rectal NET-specific deaths. SEER empirically demonstrates a profound nodal-verification deficit after local excision. The lower bound obtained by coding NX as N0 and the high complete-case estimate answer different, highly selected questions and neither identifies population nodal risk. The principal contribution of this analysis is a quantified, citable demonstration of this structural limitation rather than a corrected surgical threshold.
Language: English
Subfield: Oncology
Field: Medicine
Open Access status: gold
Estimated APC paid: $2,690
OpenAlex ID: W7225905450
FWCI:
Cites:
Cited by:
Related to: 10