4.3 Interpretation
The oncological safety of IU manipulators in endometrial cancer remains a subject of debate. Although previous studies did not specifically compare IU with non-IU manipulators, our findings support earlier research in a homogenous population with low risk EEC patients12–14,23–25. A recent meta-analysis by Scutiero et al. demonstrated that the use of IU manipulators did not impact the recurrence rate compared to when no manipulators were used in TLH for clinically early-stage endometrial cancer (risk ratio [RR] 1.11, 95% CI 0.71–1.74)13. Furthermore, Uccella et al. found no association between different IU manipulators used and the risk of recurrence. Additionally, no difference in recurrence pattern, DFS, and OS between the use and non-use of manipulators during TLH were observed24. In line with this, Alletti et al. illustrated similar DFS and OS after TLH with and without IU manipulator in a multicentric randomized controlled trial14.
On the contrary, several research groups have indicated that the use of IU manipulators negatively affects oncological outcome15,16. Padilla-Iserte et al. showed that the recurrence rate (HR 2.31, 95% CI 1.27–4.20) and survival were worse after TLH with IU manipulator than without manipulator, but no difference in recurrence pattern was found. Interestingly, the decrease in DFS and OS was only observed in patients with FIGO I-II endometrial cancer (HR 0.74, 95% CI 0.57–0.97 vs HR 1.74, 95% CI 1.07–2.83, respectively) and not in those with FIGO III endometrial cancer15.
In our cohort, the positive LVSI rate was significantly higher in the IU group (12.7%) than in the non-IU group (10.5%), but no correlation was observed with worse survival. LVSI and PPC have both been considered poor prognostic factors for recurrence and survival in endometrial cancer. However, it remains disputable whether these factors are associated with the use of uterine manipulators7–11. Scutiero et al. reported that there were no differences in LVSI and PPC rate between the use and non-use of IU manipulators during TLH. Moreover, the incidence of PPC before and after insertion of the IU manipulator was similar, which suggested that the use of IU manipulators was not associated with PPC conversion13. In disagreement with these findings, Siegenthaler et al. showed that laparoscopy with IU manipulation was followed by PPC conversion in 8.1% of patients, which was significantly associated with higher recurrence rate and lower DFS and OS. In their study, peritoneal washings were taken at three time points: at the beginning of surgery, after manipulator insertion, and after vaginal vault closure. They found that 80% of cytology conversions occurred at the third washing, implying that the presence of the manipulator in the uterine cavity during the whole procedure is the main issue, rather than the insertion of the manipulator itself16. Interestingly, hysteroscopy has been associated with higher PPC rates, but without worse oncological outcome26–29. This discrepancy might result from different ways of handling the IU device. During hysterectomy, the IU manipulator is in theory more likely to induce trauma, which might lead to cancer recurrence by disrupting the containment barrier, whereas hysteroscopy involves passively rinsing out tumour cells16. Our observation of no significant differences in site of recurrence argues against this theory.
One explanation for the contrasting results observed across studies is the small sample size of the majority of research, combined with the low recurrence rate in endometrial cancer14,23,25. Although studies with small sample sizes should not be disregarded, their results are limited in statistical power. In addition, the variation in follow-up duration between studies, ranging from 1925 to 120 months16, contributes to the inconsistency in earlier findings, as time is important when evaluating oncological outcomes. Furthermore, most studies included a heterogeneous population of patients in terms of tumour stage, grade, and histotype15,16,23–25. Alletti et al. was the only study that specifically focused on clinically early-stage, low-grade EEC patients, but only included 154 patients14. Another contributing factor could be the different manipulators used during surgery12,13. Since favourable and adverse effects of tumour manipulation may be affected by the type of manipulator, potential manipulator-specific differences should be considered.
Moreover, the overall recurrence in our study was 3.7%, which is lower than the 9.7% reported by Reijntjes et al. in the same low-risk population30. This suggests a relative underreporting of recurrence in our data. One explanation is that we defined recurrence as histologically confirmed recurrence according to PALGA, leading to some patients with not-histologically confirmed recurrences being missed in our study. Although NCR has not systematically recorded cancer recurrence, NCR has documented recurrence data between 2015 and 2017 in a pilot study. By comparing the NCR pilot data to the PALGA data, we established that the number of missed recurrences was similar between the IU and non-IU group in this period.