The survival impact of adjuvant radiotherapy and chemotherapy in patients with non-endometrioid endometrial carcinomas: a PSM-IPTW analysis based on SEER database

Abstract

PurposeTo investigate outcomes of adjuvant treatments for non-endometrioid endometrial carcinomas (NEEC), as previous studies are limited by its rarity and heterogeneity.Patients and methodsPatients with endometrial serous carcinoma (SC), clear cell carcinoma (CCC) and carcinosarcoma were identified between 2004 and 2018 from SEER database. Propensity score matching (PSM) along with inverse probability treatment weighting (IPTW) technique were employed to balance confounding factors. Multivariate, exploratory subgroup and sensitivity analyses were conducted to evaluate the impact of adjuvant treatment on overall survival (OS) and cause-specific survival (CSS).ResultsThe cohort comprised 5577 serous, 977 clear cell, and 959 carcinosarcomas. Combined chemotherapy and radiotherapy (CRT), chemotherapy alone, and radiotherapy alone were respectively administered in 42.21%, 47.27% and 10.58% of the whole cohort. Prior to adjusting, chemotherapy plus brachytherapy yielded the most beneficial effect among various strategies. After PSM-IPTW adjustment, CRT still demonstrated beneficial effect on OS and CSS. Subgroup analysis indicated CRT improved survival among various TNM stages, particularly with uterine carcinosarcoma. In the sensitivity analyses for serous histology, brachytherapy with or without chemotherapy appeared to benefit stage I-II patients. In stage III-IV SC patients, chemotherapy plus brachytherapy was still associated with improved survival outcomes. When nodal metastases were identified, additional external beam radiotherapy (EBRT) to CT was more utilized with survival improvement.ConclusionIn NEEC patients, combined CRT yielded beneficial effects than any single mode. Both chemotherapy and brachytherapy promoted survival in early stage SC patients. Late stage SC patients may benefit from chemotherapy plus either EBRT or brachytherapy.