News

Article

Least Differentiation Analysis Reliably Measures LNM Risk in Submucosal Invasive CRC

Tumor grading via least differentiation analysis was found to be a way to measure the risk of lymph node metastasis (LNM) in patients with submucosal invasive colorectal cancer (CRC).

Tumor grading with least differentiation analysis was found to have higher sensitivity for lymph node metastasis (LNM) risk compared with predominant differentiation analysis for patients with submucosal invasive (T1) colorectal cancer (CRC), according to a study published in DEN Open. Tumor grading via least differentiation analysis was also found to be more reliable.

Patients with T1 CRC are expected to increase in number with more screening programs and development of endoscopic resection techniques. LNM occurs in 10% of cases of T1 CRC, which makes determining whether a patient needs surgical resection with lymph node dissection important. Although guidelines for determining this are in place, they have some issues limiting standardization, the authors wrote. This study aimed to evaluate which method to classify tumor grading is most accurate for assessing the risk of LNM in patients with T1 CRC.

Patients who were pathologically diagnosed with T1 CRC and had primary or secondary surgical resection between April 2001 and October 2021 were included in this study. All patients were from Showa University Northern Yokohama Hospital in Yokohama City, Japan. Patients were excluded if they only had an endoscopic resection, had synchronous advanced cancer, had Lynch syndrome, had familial adenomatous polyposis, had missing data, had inflammatory bowel disease, or had chemo- and/or radiotherapy before operation.

Cancer cells in colon | Image credit: Crystal light - stock.adobe.com

Cancer cells in colon | Image credit: Crystal light - stock.adobe.com

The goal of the study was to see which variable for predicting the risk of LNM was more accurate between the least differentiation analysis and the predominant differentiation analysis. Tumor location, age, sex, tumor morphology, tumor size, tumor grade, lymphovascular invasion, tumor budding, status of LNM, and depth of submucosal invasion were all analyzed in each patient.

There were 971 patients included in the study who had T1 CRC and had initial or additional surgical resection during the study period. The rate of LNM was 9.8%, and there were a mean (SD) of 20 (11) dissected lymph nodes.

The rate of high-grade disease in the least differentiation analysis was 17.0% (95% CI, 14.7%-19.5%), compared with 0.8% (95% CI, 0.45%-1.6%) when using the predominant differentiation analysis. The least differentiation analysis had a sensitivity of 27.4% (95% CI, 18.7%-37.5%) compared with 2.1% (95% CI, 0.3%-7.4%) in the predominant differentiation analysis; specificity was 84.1% (95% CI, 81.5%-86.5%) and 99.3% (95% CI, 98.5%-99.7%) respectively.

A multivariate logistic regression analysis found that tumor grade via the least differentiation analysis was an individual prognostic factor (OR, 1.68; 95% CI, 1.00-2.83) but it was not found to be an independent predictor when part of the predominant differentiation (OR, 2.07; 95% CI, 0.38-11.2).

There were some limitations to this study. There was a small sample size from a single center, which could have limited its generalizability. Patients with T1 CRC who were managed through endoscopic treatment alone were not considered in the analysis. Selection bias is also possible due to the exclusion of high-grade cases who were identified through least differentiation analysis.

The researchers concluded that tumor grading with least differentiation analysis was a risk factor for LNM in T1 CRC. The sensitivity was higher in the least differentiation analysis of assessing tumor grading for LNM in T1 CRC.

Reference

Shiina O, Kudo SE, Ichimasa K, et al. Differentiation grade as a risk factor for lymph node metastasis in T1 colorectal cancer. DEN Open. Published online December 28, 2023. doi:10.1002/deo2.324

Related Videos
Dr Ajay Goel
Dr Ajay Goel
Kristen K. Ciombor, MD, MSCI, associate professor, Vanderbilt University
Kristen K. Ciombor, MD, MSCI, associate professor, Vanderbilt University
Kristen K. Ciombor, MD, MSCI, associate professor of medicine, Vanderbilt University
Related Content
AJMC Managed Markets Network Logo
CH LogoCenter for Biosimilars Logo