Advanced Search

Home > Journals > Minerva Chirurgica > Past Issues > Minerva Chirurgica 2006 October;61(5) > Minerva Chirurgica 2006 October;61(5):385-91



A Journal on Surgery

Indexed/Abstracted in: EMBASE, PubMed/MEDLINE, Science Citation Index Expanded (SciSearch), Scopus
Impact Factor 0,877

Frequency: Bi-Monthly

ISSN 0026-4733

Online ISSN 1827-1626


Minerva Chirurgica 2006 October;61(5):385-91


Variance of surgeons versus pathologists in staging of colorectal cancer

Jha M. K. 1, Corbett W. A. 1, Wilson R. G. 1, Koreli A. 2, Papagrigoriadis S. 2

1 Coloproctology Unit James Cook University Hospital Middlesbrough, UK
2 Department of Colorectal Surgery King’s College Hospital, London, UK

Aim. Accurate staging of colorectal cancer depends on adequate retrieval and reporting of lymph nodes in the specimen. The presence of positive lymph nodes is an indication for adjuvant therapy. Both surgeons and pathologists influence the number of lymph nodes that are retrieved and reported in specimens. Although several recommendations exist in the literature regarding the minimum number of lymph nodes required for reliable staging, the relationship of examined to infiltrated lymph nodes has not been clarified. The aims of this study were to examine variance among surgeons and pathologists in the retrieval and reporting of lymph nodes in colorectal cancer specimens; to examine the relationship between retrieved/examined lymph nodes and infiltrated lymph nodes; to identify in our own series the minimum number of retrieved lymph nodes required to secure accurate staging.
Methods. Cross-sectional study of 284 patients with colorectal cancer followed in our hospital and retrospective analysis of histopathology reports. Correlation analysis, ANOVA, and survival analysis were performed on the data.
Results. There were 127 patients with cancer of the rectum and 157 patients with cancer of the colon under follow-up. The median number of lymph nodes per specimen was 8 (range 0-29). There was no difference in the number of retrieved lymph nodes among 9 surgeons. There were 2 outliers among pathologists, with one reporting a mean of 11.4 (9.8-12.9) 95% CI nodes per specimen and another reporting a mean 4.9 (3.6-6.2) 95% CI nodes per specimen. Dukes and T stage did not affect the number of nodes. Correlation analysis revealed a linear correlation between the total number of reported lymph nodes and the existence of positive lymph nodes. From the correlation equation we calculated that, in order to have one positive node, a minimum of 8.4 nodes was required in the specimen. Therefore, in our group of patients, a minimum of 8.4 nodes was required for accurate Dukes staging. However, survival analysis did not show any difference between patients with more and patients with less than 9 reported lymph nodes.
Conclusions. Variance among pathologists exists and may be at least as important as variance among surgeons. Specialisation of pathologists similar to that of surgeons as well as employment of new techniques may be required . There is a linear correlation between the number of examined lymph nodes and the presence of positive nodes in a colorectal cancer specimen. This linear correlation makes the calculation of the minimum number of lymph nodes possible. In our series a minimum of nine nodes must be examined. However, we have not demonstrated an effect of inadequate nodes numbers on survival, possibly because survival in colorectal cancer is multifactorial.


top of page