Non-detection in exfoliative airway cytology of lung cancer: associations with sampling context and reported lepidic morphology
Abstract Introduction: We examined whether unequivocally negative airway cytology was associated with sampling context and reported histopathological features in histologically confirmed lung cancers. Methods: This retrospective cohort included selected linked accessions with histologically confirmed lung cancer and definitive exfoliative-airway cytology interpretations. Accessions containing intrapulmonary fine-needle aspiration were excluded. The full-cohort model excluded suspicious-only and unknown records and used patient-clustered robust standard errors. A co-primary analysis was restricted to adenocarcinomas without the institution-specific processing prefix because the opportunity to report lepidic growth varied with the histological material available. Three pathologists audited 300 report-text records. Results: The selected linked cohort comprised 5,376 pathology accessions from 4,846 patient clusters; 1,388 (25.8%) were unequivocally negative. Detection proportions were 78.7% with brushing plus another airway sample, 67.9% with brushing alone, 45.1% with lavage/washing alone, and 18.5% with sputum alone; these are accession-level proportions within the selected cohort and are not overall diagnostic sensitivity estimates. A lepidic feature reported in the paired histopathology was associated with non-detection in the full cohort (odds ratio, 3.52; 95% confidence interval, 2.24–5.53) and in the ascertainment-restricted adenocarcinoma analysis (odds ratio, 3.05; 95% confidence interval, 1.82–5.14). When suspicious interpretations were counted as non-detection, the odds ratio was 3.44 (95% confidence interval, 2.25–5.26). Conclusion: Sampling context and a reported lepidic feature were associated with non-detection in this selected cohort. The data cannot distinguish tumour biology from sampling, visibility, specimen adequacy, interpretive factors, or outcome-dependent availability of histological material, and do not establish comparative sampling effectiveness. They support cytology–histology audits stratified by specimen type, diagnostic category, and histological material.
Authors
- Kaiwen Fu
- Enle Zhou
- Fang Yuan
- lifeng wang
Publication Details
- Journal
- Acta Cytologica
- Published
- 2026-11-13
- DOI
- https://doi.org/10.1159/acy/abjag020
- Primary Topic
- Lung Cancer Diagnosis and Treatment
- Type
- article
- Field-Weighted Citation Impact
- 0.00