The impact of omentectomy on cause-specific survival of Stage I-IIIA epithelial ovarian cancer: A PSM-IPTW analysis based on the SEER database

• Published in: Frontiers in surgery Vol. 9; p. 1052788
• Main Authors: Hao, Zhimin, Yu, Yangli, Yang, Sufen
• Format: Journal Article
• Language: English
• Published: Switzerland Frontiers Media S.A 29.12.2022
• Subjects: epithelial ovarian cancer; IPTW; lymphadenectomy; omentectomy; PSM; regression tree model; Surgery; IPTW; regression tree model; PSM; epithelial ovarian cancer; omentectomy; lymphadenectomy
• ISSN: 2296-875X; 2296-875X
• DOI: 10.3389/fsurg.2022.1052788

Routine omentectomy is generally performed during surgery for patients with epithelial ovarian cancer (EOC). The current study aims to evaluate the impact of omentectomy on cause-specific survival of Stage I-IIIA EOC patients. Patients who presented with clinical Stage I-IIIA serous, clear cell, endometrioid, and mucinous ovarian cancers were selected from the SEER Database for the period between 2004 and 2018. We extracted clinicopathological data and surgical information with the focus on the performance of omentectomy and lymphadenectomy. Binary logistic regression and recursive partitioning analyses were conducted to identify the significant factors for the performance of omentectomy during surgery. Propensity score matching (PSM) and inverse probability treatment weighting (IPTW) techniques were utilized to balance confounding factors. Multivariate, exploratory subgroup analyses and sensitivity analyses were conducted to evaluate the impact of omentectomy on cause-specific survival (CSS). A total of 13,302 patients with EOC were enrolled in the study. The cohort comprised 3,569 endometrioid, 4,915 serous, 2,407 clear cell, and 2,411 mucinous subtypes. A total of 48.62% (6,467/13,302) of patients underwent the procedure of omentectomy during primary surgery, and only 3% absolute improvement in CSS at the individual level was observed, without statistical significance based on multivariate analysis. According to the regression-tree model with recursive partitioning analysis, the procedure of lymphadenectomy was found to be the strongest factor to distinguish the performance of omentectomy, followed by the tumor stage. Patients who underwent omentectomy were more likely to be managed in Stage I than those who underwent lymphadenectomy. After PSM-IPTW adjustment, the inclusion of omentectomy in the initial surgical procedure did not demonstrate a beneficial impact on CSS compared with those who did not undergo the procedure. Exploratory subgroup analysis indicated that the performance of omentectomy improved 5-year CSS in Stage II-IIIA patients. In the sensitive analyses for various tumor stages, omentectomy appeared to benefit only Stage II patients. However, patients across various stages seemed to benefit from the performance of lymphadenectomy, irrespective of the performance of omentectomy on them. Routine omentectomy may not be associated with survival benefit for patients with a grossly normal-appearing omentum, especially for those with clinical Stage I epithelial ovarian cancers.