Interobserver variation in the classification of tumor deposits in rectal cancer-is the use of histopathological characteristics the way to go?
Virchows Arch
; 479(6): 1111-1118, 2021 Dec.
Article
em En
| MEDLINE
| ID: mdl-34480612
ABSTRACT
The focus on lymph node metastases (LNM) as the most important prognostic marker in colorectal cancer (CRC) has been challenged by the finding that other types of locoregional spread, including tumor deposits (TDs), extramural venous invasion (EMVI), and perineural invasion (PNI), also have significant impact. However, there are concerns about interobserver variation when differentiating between these features. Therefore, this study analyzed interobserver agreement between pathologists when assessing routine tumor nodules based on TNM 8. Electronic slides of 50 tumor nodules that were not treated with neoadjuvant therapy were reviewed by 8 gastrointestinal pathologists. They were asked to classify each nodule as TD, LNM, EMVI, or PNI, and to list which histological discriminatory features were present. There was overall agreement of 73.5% (κ 0.38, 95%-CI 0.33-0.43) if a nodal versus non-nodal classification was used, and 52.2% (κ 0.27, 95%-CI 0.23-0.31) if EMVI and PNI were classified separately. The interobserver agreement varied significantly between discriminatory features from κ 0.64 (95%-CI 0.58-0.70) for roundness to κ 0.26 (95%-CI 0.12-0.41) for a lone arteriole sign, and the presence of discriminatory features did not always correlate with the final classification. Since extranodal pathways of spread are prognostically relevant, classification of tumor nodules is important. There is currently no evidence for the prognostic relevance of the origin of TD, and although some histopathological characteristics showed good interobserver agreement, these are often non-specific. To optimize interobserver agreement, we recommend a binary classification of nodal versus extranodal tumor nodules which is based on prognostic evidence and yields good overall agreement.
Palavras-chave
Texto completo:
1
Coleções:
01-internacional
Base de dados:
MEDLINE
Tipo de estudo:
Observational_studies
/
Prognostic_studies
/
Risk_factors_studies
Limite:
Humans
País/Região como assunto:
Europa
Idioma:
En
Ano de publicação:
2021
Tipo de documento:
Article