3. RESULTS
A total of 5,995 patients were identified with early-stage, low-grade EEC who received TLH between 2010 and 2020 from the NCR database. After linkage of the NCR database with the PALGA records, 86 patients did not meet the inclusion criteria. Moreover, 49 patients were excluded based on concurrent adnexal malignancy and 45 patients were excluded due to (partly) missing histopathology records. The online survey was distributed to 70 national hospitals. Amongst the hospitals, 9 underwent a change in hospital manipulator preference between 2010 and 2020. Therefore, these hospitals were seen as independent institutions before and after the date of change in manipulator preference, resulting in a total of 79 hospitals between 2010 and 2020. Of these hospitals, 5 were excluded from the analysis due to absence of response or inadequate data regarding hospital manipulator preference (N = 420 patients), and 4 because of simultaneous use of IU and non-IU manipulators (N = 190 patients). The remaining 70 (88.6%) hospitals, comprised of 5,205 patients (89.5%), were included in the analysis (Figure S1).
Of the total study population, 1,524 (29.3%) patients underwent surgery in hospitals that preferred non-IU manipulators (non-IU group) and 3,681 (70.7%) in hospitals that preferred IU manipulators (IU group). TLH with BSO was performed in 94.1% of patients in the non-IU group and 93.5% of patients in the IU group. The remaining patients underwent no or other types of (salpingo-)oophorectomy due to unknown reasons. In the non-IU group, 330 patients received adjuvant radiotherapy (21.7%), 1 patient received chemotherapy (0.1%), and 13 patients received (neo)adjuvant hormone therapy (0.8%). In the IU-group, 890 patients received adjuvant radiotherapy (24.2%), 1 patient underwent chemotherapy (0.0%), and 11 patients received adjuvant hormone therapy (0.3%). Mean age at diagnosis was 65.5 years (SD 10 years) in the non-IU group and 66.5 years (SD 9.7 years) in the IU group (p=0.001). In both non-IU and IU groups, most patients were diagnosed with FIGO IA disease (70.8% vs 67.8%, respectively, p=0.034) and without LVSI (89.5% vs 87.3%, p=0.036). No significant difference was observed between groups in maximum tumour diameter (p=0.485) (Table 1).
A total of 195 (3.7%) patients experienced recurrence of cancer during follow-up, involving 49/1524 (3.2%) patients in the non-IU group and 146/3681 (4.0%) patients in the IU group. There were no significant differences in site of recurrence (p=0.778). In both groups, the majority of the recurrences were distant (46.9% vs 41.8%, respectively), followed by local (32.7% vs 37.0%) and regional recurrences (20.4% vs 19.9%) (Table 2). There were 456 deaths during follow-up, including 142/1524 (9.3%) deaths in the non-IU group and 314/3681 (8.5%) in the IU group.
The median follow-up time was 64 months (interquartile range [IQR] 42.1–86.5 months) for the whole study population. Five-year DFS was 89.9% in the non-IU group and 89.5% in the IU group (Figure 1A). Five-year OS was 91.0% in the non-IU group and 91.5% in the IU group (Figure 1B). On univariable analysis, the risk of recurrence was comparable between the IU and non-IU groups (HR 1.04, 95% CI 0.87–1.23). After adjusting for age at onset, FIGO stage, type of hospital, and presence of LVSI, the risk of recurrence remained similar in both groups (HR 0.93, 95% CI 0.78–1.11). Similarly, manipulator preference did not affect the risk of death by any cause both at univariable (HR 1.02, 95% CI 0.85–1.22) and multivariable analyses (HR 0.90, 95% CI 0.75–1.09).
Of all patients, 1907 were treated in general hospitals, 2979 in teaching hospitals, and 319 in academic hospitals. The majority of patients seen in academic hospitals were treated with non-IU manipulators (66.8%), while patients in general and teaching hospitals were mainly operated on with IU manipulators (84.2% and 66.1%, respectively) (Table 3).