The aim of this systematic review and meta-analysis was to determine the survival benefit of CRS in patients with primary stage IV EC. Through a comprehensive search across five databases, we included twelve studies. Ten studies showed a favourable effect on OS of complete/optimal CRS compared to incomplete CRS. Meta-analysis showed a pooled HR of 0.38 (95% CI 0.28–0.69, p = 0.0016) in favour of complete/optimal CRS. Three studies provided OS data for complete CRS vs. optimal vs. incomplete CRS and one for complete CRS vs. incomplete CRS. Three of them reported an OS of more than 40 months after complete CRS compared to 13–23 months after optimal and 7–14 months after incomplete CRS. Our results highlight the value of complete/optimal CRS, applying the same paradigm of high-stage ovarian cancer for EC.7 Current guidelines cautiously recommend CRS in women with advanced EC if complete resection is deemed feasible.4–6 However, reluctance among clinicians remains, due to the challenges of CRS in patients with stage IV EC which are also opposed in our study.
Primary stage IV EC comprises a heterogeneous patient population due to several factors, firstly due to the presence of extra-abdominal metastases in some patients with stage IV EC. In our review, the majority of the patients had intra-abdominal metastasis only, however details on intra-abdominal localization were scarcely provided by the included studies. The significance of differentiation between pelvic, abdominal, and extra-abdominal metastases is demonstrated in the study by Height et al.32 These authors applied the updated 2023 FIGO stage for endometrial cancer33, which classifies pelvic metastases as stage III and extra-abdominal metastases as stage IVC. Their findings revealed a significant prolonged OS among patients with stage III under the updated classification, while these patients were classified as stage IV in the FIGO 2009 system. Furthermore, current guidelines do not specify whether CRS should be attempted in the presence of extra-abdominal disease. Within our review, three of the seven studies that included patients with extra-abdominal disease provided information regarding the treatment of these metastases. These studies documented the surgical removal in some patients with extra-abdominal metastases located in the lungs, liver, supraclavicular and inguinal lymph nodes, umbilicus/skin, and breast. On the contrary, surgical removal of bone and brain metastases was not performed. Additionally, in one study (Eto et al.24) radiotherapy was administered for extra-abdominal metastases. The majority of this cohort with extra-abdominal metastases had metastasis localized in a single anatomical region. These authors, together with Ueda et al.31 reported a significant longer OS in patients who underwent complete or optimal CRS, regardless of the presence of extra-abdominal metastases. Guo et al.34 have demonstrated a survival benefit after the surgical removal of lung, bone, or multiple organ metastasis. In contrast, surgical and/or radio therapeutic treatment of brain metastases does not seem to lead to a survival benefit.34,35 Consequently, considering the findings of our and the abovementioned studies, we oppose the notion that complete/optimal CRS provides a benefit even if extra-abdominal disease is present, particularly if is the extra-abdominal metastasis is localized in a single anatomical region other than the brain.
Additionally, the histologic subtype of EC introduces heterogeneity in patients with primary stage IV EC. Consequently, certain studies have focused on patients with a specific histological subtype, including only patients with endometroid EC or only patients with serous EC. Serous EC, characterized by its high-grade nature and aggressiveness, is more frequently detected in advanced stages of the disease.36,37 This is also reflected in the number of patients with serous EC in our population (42%), despite the overall incidence of serous EC across all stages of EC being only 10%.4,38 It has been suggested that less aggressive tumours are better suited for complete resection, implying that observed survival benefit may also be attributed to the histological characteristics of the tumour rather than solely the surgical intervention itself.20,23,31 In our review five studies included patients with serous EC only, and four reported a significant advantage in OS after complete or optimal CRS compared to incomplete CRS. These findings demonstrate that CRS can be advantageous for both endometroid and serous stage IV EC patients.
Lastly, due to patient heterogeneity it is not feasible to establish a uniform recommendation. Patients with EC typically exhibit characteristics such as advanced age, frailty, and potential comorbidities. These factors contribute to possible prolonged post-surgical recovery and an increased risk of peri- and postoperative complications and mortality26, particularly when extensive surgical procedures, including bowel resection, are needed for complete/optimal CRS. Notably, in our review, a considerable proportion of patients (40–47%) had macroscopic residual disease larger than 1 or 2 cm following CRS. Unfortunately, the studies included in our review provided limited information on peri- and postoperative complications. However, when reported, a notable percentage of patients experienced complications, with rates ranging from 23–51.6%. Furthermore, there are two reported patients with fatal surgery-related complications. Objective tools such as frailty assessment questionnaires like the G8 questionnaire39 and WHO performance status can help predict post-surgical morbidity rates. Of the studies in our review, Bristow et al.22 identified age and a good performance status, in addition to complete/optimal CRS, as significant predictors of survival. Eto et al.24 further supported this notion, finding that patients with a good performance score and limited metastases were more likely suitable candidates for complete CRS. In conclusion, the treatment of elderly, frail women with primary stage IV EC poses additional challenges, requiring a delicate balance between the risk of complications and the potential survival benefits, where the focus should not be solely on age but also on physical health and patients’ preferences.
We observed variations in the definition of optimal CRS among the studies included in our systematic review and meta-analysis. While most studies defined optimal CRS as ≤ 1cm residual disease, some studies used a less strict criterion of ≤ 2cm residual disease. Despite the difference in cut-off values, both optimal CRS definitions were associated with improved survival outcomes. Importantly, the survival benefit was superior in patients who underwent complete CRS compared to patients who underwent with optimal CRS. In our study cohort, a significant proportion (40 to 47%) still had residual macroscopic disease exceeding 1 or 2 cm following surgery. This emphasizes the importance of accurate pre-operative imaging and the need for innovative tools and techniques to maximize the chances of achieving complete CRS. Based on our findings, we recommend a thorough pre-operative assessment to determine the feasibility of complete CRS and only perform surgery if complete CRS appears achievable. In patients in whom complete CRS is not feasible during surgery, efforts should be made to achieve optimal CRS. Residual disease ≤ 1cm is more effectively managed through adjuvant radiotherapy or systemic therapy, emphasizing the importance of striving for the best possible surgical outcome.
This study and two previously published systematic review and meta-analyses14–16 have investigated the benefit of complete and/or optimal CRS in patients with advanced stage EC. The most recent study, a systematic review by Capozzi et al.16 has investigated the optimal management for stage IVB EC and stated that determining the optimal treatment for these patients remains complex due to the lack of comprehensive evidence. Regarding the role of CRS in management for these patients the authors, similar to our perspective, conclude after the systematic review that optimal CRS remains crucial, however this effect is not quantified in a meta-analysis. The systematic review and meta-analysis by Albright et al.14, included studies on patients with both stage III and stage IV EC. Thirty-four studies were selected, which included a total of 1329 patients. Eleven of these studies were also included in our meta-analysis. The studies not included in our analysis focused on stage III EC, or did not report the survival data of patients with stage IV EC separately. To impute the HR of studies that were only reporting on median survival, Albright et al.14 calculated the HR by dividing the median survival in the two groups, rather than taking the whole survival curves into account as recommended by Tierney et al.19. The estimated effect of complete/optimal CRS was similar to our meta-analysis. This does, however, not imply that the results presented in the meta-analysis of Albright et al.18 can be extrapolated to patients with primary stage IV EC because patients with stage III EC are more often treated with emphasis on curative treatment, in contrast to patients with stage IV EC. In 2010, Barlin et al.15 conducted a meta-analysis to determine prognostic factors for OS in patients with stage III and IV and recurrent EC. The authors selected 14 studies, including a total of 672 patients, of which 157 had recurrent EC. Their analysis also showed that complete CRS was significantly associated with a longer OS (9.3 months increase, range 0.1–18.5, p = 0.04). In their analysis, optimal CRS also positively impacted on OS (16 months increase, range − 0.7–32.7) compared to incomplete CRS, though this was on the verge of significance (p = 0.05). Hence, both meta-analyses concluded, in line with our meta-analysis, that complete CRS and optimal CRS are associated with improved survival outcomes.
Future studies should prospectively investigate the effect of CRS in patients with primary stage IV EC on survival, morbidity and quality of life. Recording details on pre-operative imaging, surgical decision-making, and the extent of surgery might shed light on the predictability of achieving complete CRS, for example with the frailty assessment questionnaires.39 Moreover, the heterogeneity resulting from the intra- or extra-abdominal localization of metastases can be more effectively classified through the incorporation of the updated 2023 FIGO stage for endometrial cancer.33 Additionally, the role of the molecular classification40, both prognostic impact and predictive for completeness of CRS should be investigated. In stage I-III EC, the different molecular EC classes are known to have a distinct prognosis and respond differently to adjuvant chemotherapy and radiotherapy.41–43 Besides one study in recurrent EC, which showed prognostic significance44, the value of molecularly classifying advanced EC has not been explored. Given the rarity of primary stage IV EC, a multi-center, international prospective study is most likely needed.
To our knowledge, this study stands out as the sole systematic review and meta-analysis that specifically focuses on CRS in patients diagnosed with primary stage IV EC. The comprehensive analysis of the available literature presented here is particularly valuable, as the heterogeneity within the stage IV EC patient group is substantial, which often leads to reluctance in considering CRS. We conducted a comprehensive and systematic search across five databases to ensure the inclusion of all relevant studies that reported on survival outcomes after CRS in patients with primary stage IV EC. There was optimal use of the available data by advanced imputation techniques and performed state-of-the-art statistical analyses. The aim was to investigate the role of complete CRS in primary stage IV EC. However, due to limitations in the available data from the included studies, we could only investigate the effect of complete/optimal CRS versus incomplete CRS.
The quality of every meta-analysis depends on the quality of the original studies that are pooled. Here, only relatively small, retrospective studies were available, which are hampered by selection bias, registration bias, and confounding by indication. In addition, our analysis showed that there is significant publication bias in this field, in favour of small studies showing a positive impact of complete/optimal CRS on survival. Hence, the estimated benefit of complete/optimal CRS might be overestimated. Another limitation is the substantial heterogeneity between the included studies. This causes additional uncertainty in the pooled estimate of the HR that is not fully addressed by the 95% confidence interval.