Related Experiment Videos
A practical scoring system based upon ROC analysis for evaluating potential lymph nodes metastasis during gastric
Yulian Wu1, Jian Chen, Junxiu Yu
1Department of Surgery, the Second Affiliated Hospital, College of Medicine, Zhejiang University, P. R. China. wuyulian@medmail.com.cn
Journal of Surgical Oncology
|May 18, 2006
Summary
A new scoring system accurately predicts lymph node involvement in gastric cancer, aiding surgeons in determining the extent of lymphadenectomy needed. This system improves preoperative staging and minimizes unnecessary extensive lymph node dissection.
Area of Science:
- Oncology
- Surgical Oncology
- Diagnostic Accuracy
Background:
- Current preoperative staging for gastric cancer lymph node involvement lacks precision.
- Accurate staging is crucial for guiding appropriate lymphadenectomy strategies.
Purpose of the Study:
- To develop and validate a novel scoring system for predicting N stages in gastric cancer.
- To guide rational lymphadenectomy based on predicted lymph node involvement.
Main Methods:
- Multivariate logistic regression identified key predictors: tumor size, invasion depth, and histology.
- A scoring system was developed using odds ratios.
- Receiver operating characteristic (ROC) analysis defined scoring ranges for N0-N3 stages.
- Kappa statistics assessed agreement between predicted and actual pN classifications.
Main Results:
- The scoring system incorporates tumor size, invasion depth, and histological type.
- Scores 0-4, 5-7, 8-9, and 10-13 predict N0, N1, N2, and N3 stages, respectively.
- High negative predictive values (86.4-90.2%) and good overall accuracy (82-83.7%) were achieved.
- The system demonstrated good agreement with actual pN classifications.
Conclusions:
- The developed scoring system reliably predicts N stages in gastric cancer.
- Its strong negative predictive value supports avoiding unnecessary extensive lymph node dissection.
- This tool can optimize surgical planning and patient outcomes.