A Statistical Analysis of the Role of Various Co-variables in Determining the Number of Nodes to be Dissected in Endometrial Cancer Using the CART Model, AHGPIC

Authors

  • Jita Parija Department of Gynaecooncology Oncology, AHPGIC Cuttack Utkal University, India.
  • Bhagyalaxmi Nayak Department of Gynaecooncology Oncology, AHPGIC Cuttack Utkal University, India.
  • Ashok Padhi Department of Gynaecooncology Oncology, AHPGIC Cuttack Utkal University, India.
  • Smruti Sudha Pattnaik Department of Gynaecooncology Oncology, AHPGIC Cuttack Utkal University, India.
  • Manoranjan Mohapatra Department of Gynaecooncology Oncology, AHPGIC Cuttack Utkal University, India.
  • Janmejaya Mohaptra Department of Gynaecooncology Oncology, AHPGIC Cuttack Utkal University, India.
  • Rekha Das Department of Anesthesiology, AHPGIC Cuttack Utkal University, India.
  • Padmalaya Devi Department of Surgical Oncology, AHPGIC Cuttack Utkal University, India.
  • Sushil Kumar Giri Department of Gynaecooncology Oncology, AHPGIC Cuttack Utkal University, India.
  • Sagarika Samantray Department of Pathology, AHPGIC Cuttack Utkal University, India.
  • Lucy Pattnaik Department of Radiation Oncology, AHPGIC Cuttack Utkal University, India.
  • Sumita Mohanty Department of Oncopathology, AHPGIC Cuttack Utkal University, India.
  • Swadeep Mohanty Department of Surgical Oncology, AHPGIC Cuttack Utkal University, India.
  • Sasmita Panda Department of Radiation Oncology, AHPGIC Cuttack Utkal University, India.
  • Niharika Panda Department of Gynaecooncology Oncology, AHPGIC Cuttack Utkal University, India.
  • Surendra Nath Senapthi Department of Radiation Oncology, AHPGIC Cuttack Utkal University, India.

Keywords:

CART, Endometrial Cancer, ROC,, Lymph nodes, multiple regression.

Abstract

Background: The main objective of this paper is to predict the role of covariables in determining the number of nodes to be dissected in endometrial cancer, using the best regression model. Additionally, the study aims to compare the accuracy of the CART model with the traditional regression model, to accurately find and predict the co-variable in determining the number of nodes to be dissected.

Material and Methods: Data on 170 endometrial cancer patients, along with their covariates, were collected from the institute AHGPIC and used for the study. The data includes the dependent variable (total number of lymph nodes involved) and 10 co-variates (independent variables): age, postmenopausal bleeding, obstetrics history, nodal status, tumor size, histology, grade, myometrial invasion, lymphovascular space invasion, and cervical extension. The methods used include multiple regression and the CART model.

Results: The average number of lymph nodes dissected among patients with a tumor size less than 1.9 cm is 3.73 (approximately 4), while patients with a tumor size of 1.9 cm have an average of 12.4 nodes (approximately 13) dissected. Among patients with prior b/l pelvic lymphadenectomy, the average number of nodes dissected is 10.9 (approximately 11), while those with prior b/l para-aortic + b/l pelvic lymphadenectomy have an average of 14.1 nodes (approximately 14) dissected. The CART model predicts with an accuracy of 95.9%, which is higher than the multiple regression model’s accuracy of 88.3%, based on the selected covariates and validated by the receiver operating characteristics (ROC) curve.

Conclusion: The study concludes that if the tumor size is greater than 1.9 cm (approximately > 2 cm), then 12 nodes should be dissected, and if it is less than 1.9 cm (approximately < 2 cm), then approximately 4 nodes should be dissected. The classification and regression tree (CART) model is able to predict the role of the covariate, i.e., tumor size, in deciding the number of lymph nodes to be dissected for endometrial cancer patients with an accuracy of 95.9%, based on the selected covariates and validated by the ROC curve. The CART model predicts with more accuracy (95.9%) compared to the multiple regression model (88.3%).

Downloads

Published

2024-05-29

Issue

Vol. 9 No. 2 (2024)

Section

Systematic Review and Meta-analysis:

License

 West Asia Organization for Cabcer Prevention retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License 4 (This permits anyone to copy, distribute, transmit and adapt the published work, provided the original work and source are appropriately cited).