Article Text
PDF
Abstract
Objective This retrospective, multicenter, observational study aimed to refine patient selection criteria for secondary cytoreductive surgery in recurrent endometrial cancer. The objective was to identify preoperative predictors of complete cytoreduction, assess surgical complexity, and propose a preoperative predictive scoring system to identify suitable candidates for secondary cytoreductive surgery.
Methods Data from 331 women with recurrent endometrial cancer were analyzed across three Italian centers from January 2010 to December 2021. Patients were categorized based on treatment received (medical treatment, diagnostic laparoscopy/examination under anesthesia, or secondary cytoreductive surgery). Preoperative predictors, surgical complexity, complications, and a predictive scoring system were assessed. Logistic regression and receiver operating characteristic analysis were used for statistical evaluation.
Results Of the cohort, 56.2% underwent debulking surgery, 17.2% had diagnostic laparoscopy, and 26.6% received medical treatment. Patients undergoing secondary cytoreductive surgery were younger, with a lower body mass index, better performance status, and fewer comorbidities. Single site locoregional relapse was common in secondary cytoreductive surgery patients. Age <65 years, single site relapse, lymph node, and hematogenous relapse were independent predictors of complete cytoreduction. A predictive scoring system demonstrated a clear relationship between the score and the likelihood of complete cytoreduction.
Conclusion This study identified age <65 years, single site recurrence, as well as nodal and hematogenous recurrence, as predictive factors for achieving optimal cytoreduction. A predictive scoring system incorporating these factors has been proposed to identify optimal candidates for secondary cytoreductive surgery in recurrent endometrial cancer. The scoring system showed promising predictive accuracy and could aid in refining the decision making process, ensuring appropriate patient selection for secondary cytoreductive surgery. Further prospective studies are warranted to validate and enhance the predictive model.
- Endometrial Neoplasms
- Neoplasm Recurrence, Local
- Cytoreduction surgical procedures
Data availability statement
Data are available upon reasonable request.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
The landscape of treatment options for recurrent endometrial cancer varies widely.
Secondary cytoreductive surgery, particularly when achieving complete gross resection, has the potential to offer significant survival benefits, but the absence of well defined criteria presents a challenge in accurately selecting patients who would benefit from this approach.
WHAT THIS STUDY ADDS
Our study identified age <65 years, single site recurrence, and nodal or hematogenous recurrence as predictive factors for optimal cytoreduction.
The proposed scoring system derived from these factors could facilitate the identification of patients suitable for surgery.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The study’s findings could influence the decision making process in treatment selection, offering valuable insights for future prospective studies and potential impacts on research and clinical practice.
Introduction
Endometrial cancer is a highly heterogeneous disease, with distinct histological and molecular characteristics that significantly influence its natural history.1–3 Most of the patients are diagnosed at an early stage, resulting in a 5 year survival rate of up to 95%.4 However, a significant subgroup of patients, approximately 18%, experience disease recurrence,5 6 leading to unfavorable survival outcomes.7 8 Relapses can occur as single or multiple nodules and can follow different pathways, including locoregional, lymph node, hematogenous/parenchymal, or peritoneal, either individually or in combination.5–7 9
Recurrence management is strongly affected by type and number of relapses, prior treatments, and patient clinical characteristics. Treatment options include surgery, chemotherapy, radiation therapy, hormone therapy, and immunotherapy, which has recently gained a leading role in the treatment of relapses.10–12 In this broad scenario, studies indicate that patients with endometrial cancer recurrence who benefit from secondary cytoreductive surgery experience significantly improved survival outcomes compared with those who do not, but only a minority is deemed eligible (16–56%).5 13–15
In this study, we performed a retrospective analysis of clinical and histological variables to identify preoperative predictors of complete cytoreduction. In addition, we assessed surgical complexity, documenting rates of intraoperative and postoperative complications, and proposed a preoperative predictive scoring system to identify suitable candidates for surgery.
Methods
This was a retrospective, multicenter, observational study involving three centers in Italy: Fondazione Policlinico Universitario A Gemelli IRCCS, Rome, Gemelli Molise Hospital, Campobasso, and University of Parma. Following approval of the institutional review board of Fondazione Policlinico Universitario A Gemelli, data from patients with first endometrial cancer recurrence from January 2010 to December 2021 were retrospectively retrieved. All patients signed an informed consent for their data to be collected and analyzed for scientific purpose.
Clinical and Surgical Data
Only patients with histological and/or a radiological diagnosis of first endometrial cancer recurrence were included. Patients who did not undergo surgery as the primary treatment, and those with concomitant primary cancers and lacking relevant clinical or pathological data were excluded. Non-epithelial tumors were also excluded. All demographic, surgical, and postoperative variables were retrieved from RedCap electronic database.
Relapses were categorized based on the number of metastatic nodules identified through radiological assessments, including CT, MRI, or positron emission tomography (single vs multiple nodules). Relapses were also classified according to the specific pathway of recurrence, which could be locoregional, lymphatic, hematogenous/parenchymal (lung, liver, bone), or peritoneal. If multiple spreading pathways were revealed at radiological evaluations, relapses were classified as mixed.
Following a thorough evaluation of each case by the multidisciplinary tumor board, patients were directed toward different treatment options. The decision between radiation therapy and surgery was evaluated for all patients who had not previously undergone radiation treatment. The selection between these two options was based on various factors, including the number and location of the recurrence, its proximity to critical tissues such as the bowel and bladder, the presence of comorbidities, the patient’s performance status, and the complexity of the surgical procedure. All patients deemed eligible for secondary cytoreductive surgery underwent clinical examination under general anesthesia to assess the extension of pelvic resection and evaluate parametrial involvement, particularly in cases with central or lateral pelvic recurrence, and/or diagnostic laparoscopy to evaluate distant nodules and the presence of miliary carcinomatosis. The decision to proceed with debulking surgery was based on the combined expertise of at least two experienced gynecologic oncology surgeons, who evaluated the surgical complexity and the likelihood of achieving a complete gross resection.
The population was then divided into three groups: group 1 patients were directly assigned to medical treatment; group 2 patients underwent diagnostic laparoscopy/examination under anesthesia and subsequent medical treatment; and group 3 included patients who underwent diagnostic laparoscopy/examination under anesthesia and secondary cytoreductive surgery. Complete cytoreduction was defined as the absence of macroscopic residual disease (complete gross resection), whereas patients with residual tumor >0 were defined as suboptimally debulked.
Intraoperative complications were assessed using the Common Terminology Criteria for Adverse Events (CTCAE) V.5, while the extended Clavien–Dindo classification of surgical complications was used for early postoperative complication (within 30 days from surgery).16 The primary endpoint was to identify preoperative factors that could predict the achievement of complete cytoreduction during surgery. Secondary endpoints included describing surgical complexity, documenting rates of intraoperative and postoperative complication, assessing early mortality rates and proposing a preoperative predictive scoring system that would identify patients who are suitable candidates for surgery.
In accordance with the journal’s guidelines, we will provide our data for independent analysis by a selected team by the editorial ream for the purposes of additional data analysis or for the reproducibility of this study in other centers if such is requested.
Statistical Analysis
According to the number of patients undergoing debulking surgery or diagnostic laparoscopy, the lower frequency was considered and, based on this, several variables in a ratio 1:10 were included in the logistic model to respect the rule of at least 10 events per variable. Descriptive statistics were used to describe the clinical, surgical, and pathological characteristics of patients. Absolute frequency and percentages were adopted for qualitative variables. Groups were compared using the Pearson χ2 test for categorical variables. Binomial logistic regression was performed to evaluate factors that could predict complete gross resection. This analysis only included patients directed to surgery (groups 2 and 3), excluding those directly assigned to medical treatment (group-1).
Only variables with a p value <0.100 were included in the multivariable model. Significant variables were assigned a predictive score based on their beta coefficient, calculated considering the highest beta value. A score of 0 was assigned to items with a ratio to the highest value ≤0.50, and a score of 1 to items with a value >0.50. Subsequently, the total predictive score for each patient was calculated by summing individual scores. Finally, the complete gross resection rate corresponding to each total score was determined.
To assess the predictive accuracy of this scoring system, a receiver operating characteristic curve was constructed. The area under the curve (AUC), along with its 95% CI, was calculated. To validate the scoring system’s effectiveness and reliability, it was applied to a separate, smaller internal population of patients who had endometrial cancer recurrence but were not part of the initial analysis. A p value <0.050 was considered statistically significant (two tailed test). Statistical analyses were performed using SPSS V.29.0 (IBM, Armonk, New York, USA) software for Macintosh.
Results
Data for 331 women with recurrent endometrial cancer were retrieved. Patient clinical and histopathological characteristics are provided in online supplemental Table S1. Among the cohort, 243 patients (73.4%) were assessed for potential surgical intervention (groups 2 and 3). Of these, 186 (56.2% of the whole population) received debulking surgery, while 57 (17.2%) underwent diagnostic laparoscopy. The remaining 88 (26.6%) patients were directly assigned to medical treatment.
Supplemental material
Significant differences were observed among the three groups in terms of age, body mass index (BMI), Eastern Cooperative Oncology Group performance status (ECOG-PS), and Charlson Comorbidity Index (online supplemental Table S1). Patients undergoing secondary cytoreductive surgery were significantly younger (age <65 years: 34.1% vs 29.8% vs 51.1%, p=0.003, in groups 1, 2, and 3, respectively), with lower BMI (BMI <30: 52.3% vs 61.4% vs 69.4%, p=0.022), better ECOG-PS (score 0–1: 63.6% vs 78.9% vs 89.2%, p<0.001), and lower Charlson Comorbidity Index (score 0–2: 12.5% vs 15.8% vs 32.3%, p<0.001) (online supplemental Table S1).
Online supplemental Table S2 reports the histopathological and surgical characteristics of the study population at the initial diagnosis. A higher rate of grade 3 tumors, lymph-vascular space invasion, and pelvic lymph node metastases at first diagnosis was reported in group 1 (grade 3: 56.8% vs 40.4% vs 39.2%, p=0.019; lymph-vascular space invasion: 38.6% vs 23.3% vs 31.2%, p=0.041; positive pelvic nodes: 41.3% vs 32.4%, 22.2%, p=0.023). The three groups were comparable in terms of International Federation of Gynecology and Obstetrics (FIGO) stage (2009) and prognostic risk groups17 (p=0.18 and p=0.27, respectively) (online supplemental Table S2). The number and pattern of recurrence significantly differed among the groups (Table 1).
Supplemental material
Recurrences sites and management
Patients who underwent secondary cytoreductive surgery had a significantly higher rate of single site metastases (group 1 vs group 2 vs group 3: 29.5% vs 54.4% vs 78.0%, p<0.001) and negative Ca125 (47.5% vs 51.2% vs 72.9%, p=0.001). Among patients in group 3, 45.2% had locoregional relapse (central or lateral), 26.3% had lymph node relapse, 17.2% hematogenous, and 11.3% had peritoneal or mixed relapse (Table 1). Analyzing the type of treatment based on the radiological pattern of recurrence, lymph node recurrences were assigned to secondary cytoreductive surgery in 72.1% of cases, locoregional and hematogenous in 63.6% and 54.2%, respectively, and peritoneal and mixed in 29.2% of cases. Details regarding the characteristics of non-surgical relapses (group 1) are reported in online supplemental Table S3.
Supplemental material
Surgical characteristics, postoperative course, and adjuvant treatment of patient undergoing secondary cytoreductive surgery are reported in Table 2. Complete gross resection was reached in 95.7% of cases, median operative time was 204 min, with a median postoperative hospital stay of 4 days. Total intraoperative and postoperative complication rates were 14.5% and 21.5%, respectively. No severe intraoperative complications were reported, and among postoperative complications, only 10 (5.4%) were recorded as severe. No perioperative deaths were reported.
Surgical characteristics, postoperative course, and adjuvant treatment of patients undergoing secondary cytoreductive surgery (group 3)
Of patients undergoing secondary cytoreductive surgery, 72.6% received adjuvant chemotherapy, 7.0% both chemotherapy and radiotherapy, 10.8% radiotherapy, and 3.2% were given hormonal therapy. Twelve patients (6.4%) did not undergo adjuvant treatment due to poor postoperative conditions, severe postoperative complication, or refused chemotherapy. We then evaluated possible predictors of complete gross resection in patients undergoing surgery (groups 2 and 3) (Table 3). Age <65 years, Charlson Comorbidity Index 0–2, single site relapse, and the pattern of recurrence showed statistical significance at the univariate analysis (age <65 years: OR 2.407, 95% CI 1.32 to 4.40, p=0.004; Charlson Comorbidity Index 0–2: OR 2.37, 95% CI 1.16 to 4.88, p=0.019; single site relapse: OR 3.60, 95% CI 1.96 to 6.61, p<0.001; locoregional relapse OR 3.09, 95% CI 1.47 to 6.50, p=0.003; lymph node relapse: OR 5.56, 95% CI 2.22 to 13.90, p<0.001; hematogenous relapse: OR 11.96, 95% CI 3.15 to 45.45, p<0.001). For the multivariate logistic regression analysis, only the following variables were confirmed as independent positive predictors for complete gross resection: age <65 years (OR 2.53, 95% CI 1.12 to 5.70, p=0.025), single site relapse (OR 3.14, 95% CI 1.38 to 7.12, p=0.006), lymph node relapse (OR 4.36, 95% CI 1.61 to 11.80, p=0.004), and hematogenous relapse (OR 5.69, 95% CI 1.30 to 24.88, p=0.021).
Predictors of complete gross resection in patients undergoing surgery (groups 2 and 3): univariate and multivariate analysis
As reported in the materials and methods, each significant variable was assigned a value according to their beta coefficient. A value of 1 was assigned to age <65 years, single site relapse, and lymph node and hematogenous relapse (Figure 1). These values were then summed for each variable to create an overall predictive risk score, which ranged from 0 to 3. The predictive score performance is presented in online supplemental Figure S1, with an AUC of 0.74 (95% CI 0.67 to 0.81).
Supplemental material
Predictive score based on multivariate analysis (A), and its application in the internal cohort (B) and in the validation group (C). CGR, complete gross resection; LPS, laparoscopy; RT, radiation therapy.
An increasing rate in complete gross resection was recorded, from score 0 to 3 (score 0 vs 3: 33.3% to 93.3%) (Figure 1). This trend was consistent in the validation population (Figure 1). A cut-off of 2 (distinguishing between scores 0–1 and scores 2–3) was identified according to the Youden Index with a sensitivity value of 64.6% (57.6–71.6), specificity 75.4% (64.9–85.9), accuracy 67.5% (61.6–73.4), positive predictive value 87.8% (82.2–93.4), and negative predictive value 43.8% (34.6–52.9).
Discussion
Summary of Main Results
Patients who underwent debulking surgery tended to be younger, with a lower BMI, better performance status, and fewer comorbidities. In this subgroup, the most common type of recurrence was single site locoregional relapse. However, when considering the radiological pattern, nodal relapses were the most frequently treated with secondary cytoreductive surgery (72.1% of cases), followed by locoregional and hematogenic relapse (63.6% and 54.2% of cases). Analyzing the potential predictors of complete gross resection, it was found that age <65 years, single site relapse, lymph node relapse, and hematogenous relapse were significantly and independently associated with a higher likelihood of achieving complete gross resection.
The rate of complete gross resection consistently increased with higher predictive scores, ranging from 33.3% for a score of 0 to 93.3% for a score of 3. This underscores the correlation between the predictive score and the likelihood of achieving complete gross resection, with elevated scores indicating a higher probability of surgical success.
Results in the Context of the Literature
The recently published results from the DESKTOP-III trial18 demonstrated a significantly longer survival outcomes with the combination of surgical and systemic modalities compared with systemic treatment alone in carefully selected patients with recurrent ovarian cancer. However, this benefit was seen only in patients with complete gross resection, highlighting the importance of both optimal selection and surgical expertise.
In the context of endometrial cancer recurrence, several studies5 13 19 demonstrated the survival advantage conferred by secondary cytoreductive surgery, and our recently published research aligns with these findings, confirming a significant statistical survival benefit for patients who achieve complete gross resection.14 19–23 Nevertheless, secondary cytoreductive surgery is a demanding procedure, marked by technical intricacies and a non-negligible rate of intraoperative and postoperative complications (14%–42%)24 and thus only offered to a minority of patients, with complete gross resection rates ranging between 57.3% and 95.7%.14 15 20–23 25 26
In line with the existing literature,14 20 21 23 26 27 our study reported 14.5% of intraoperative complications with 11.3% of patients requiring intraoperative blood transfusion. Additionally, we observed an overall postoperative complications rate of 21.5%, and 5.4% of severe complications. In light of these considerations, secondary cytoreductive surgery should be reserved for patients with a high likelihood of complete cytoreduction, and it is therefore imperative to identify optimal surgical candidates to avoid harmful and unnecessary surgery.
Moukarzel et al13 proposed offering surgery to patients aged ≤70 years, with a long progression free survival, grade 1/2 endometrioid or clear cell histology, no more than two grade 3 complications at first surgery, patients who received adjuvant radiation therapy, and with single site disease at the time of recurrence. However, only a few studies specifically investigated predictive factors for optimal debulking, giving conflicting results. A first study by Campagnutta25 involving 75 patients identified the presence of a solitary nodule as the sole independent factor positively influencing a cytoreduction outcome, a finding later confirmed by Awtrey and colleagues in 200627 in a cohort of 27 patients.
A more recent study by Ren et al26 found that age at recurrence, relapse size, and multiplicity of recurrence were significantly associated with optimal debulking (defined as residual disease ≤1 cm). In contrast, Papadia et al23 could only identify performance status as a positive predictor. However, several other studies failed to identify any predictors of optimal surgery,14 21 22 limiting the ability to correctly identify the subset of patients who would benefit from it.
In our analysis, we identified four variables with independent and significant impact on surgical outcome, from which we derived a predictive score. Factors were age <65 years, single site relapse, and lymph node and hematogenous relapse. Assigning a value of 1 to each positive factor, we observed an increasing rate of complete gross resection as the score increased from 0 to 3. Patients with scores of 2 and 3 exhibited rates of 86.1% and 93.3%, respectively, indicating their suitability as candidates for surgery. By setting a cut-off value of 2, a positive predictive value of 87.8% was identified, but a relatively large number of false negative cases were recorded, resulting in a negative predictive value of 43.8%.
Strengths and Weakness
This study had several strengths, particularly its large sample size and multicenter design, which enhanced the reliability and generalizability of the findings. Comprehensive data collection enabled the identification of four key preoperative predictors of successful cytoreductive surgery, and the development and internal validation of a predictive scoring system based on these variables provides a practical tool for clinical decision making. Additionally, the study’s detailed statistical analysis, focus on surgical complexity and complications, and proposal of a decision making algorithm have significant practical implications. The retrospective methodology of the study clearly represents its main bias, since we analyzed predictors of resectability in patients selected for surgery after discussion at the multidisciplinary tumor board. Intraoperative findings were excluded from our analysis since our primary objective was to construct a predictive model of resectability relying on preoperative criteria. However, it is worth noting that diagnostic laparoscopy could offer valuable insights to refine the decision making process for debulking surgery.
Implication for Practice and Future Research
According to our findings and predictive score, we propose a decision making algorithm (Figure 2) for guiding the management of recurrent endometrial cancer. Based on preoperative criteria, patients with scores of 2 and 3 should be considered for surgery. Given the relatively low negative predictive value, and to individualize the decision for a cytoreductive attempt on a case-by-case basis, a two-step process incorporating diagnostic laparoscopy and an examination under anesthesia could be evaluated for patients with a score of 1.
Decision making algorithm.
Careful consideration should be given to the clinical characteristics of the patient, including comorbidities and performance status.28 A minimally invasive approach, whether robotic or laparoscopic, could be evaluated in highly selected patients with single parenchymal or lymph node recurrence.29–31
The importance of patient selection is amplified in the current era of personalized therapy, characterized by significant insights into tumor biology that have ushered in a new era of targeted therapies and immunotherapy modalities that are showing encouraging efficacy. In this dynamic landscape, it is imperative to carefully reassess and integrate the role of surgery into the comprehensive spectrum of therapeutic interventions to optimize patient care and outcomes.
Conclusions
Determining the optimal treatment approach for recurrent endometrial cancer poses a significant challenge, as it is strongly influenced by the patient’s clinical condition, previous treatments, and number and location of metastases. Surgery represents a viable option only for those patients who can be optimally debulked, with acceptable intraoperative and postoperative complication rates.
We identified four predictors of optimal debulking: age <65 years, recurrence at a single site, lymphatic recurrence pattern, and hematogenous pattern, and hypothesized a predictive score that can be applied preoperatively to identify patients with a high likelihood of achieving optimal surgery. This score may represent a valuable assistance in the decision making algorithm of the treatment of endometrial cancer recurrence and could form the basis for larger prospective studies in the future.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by the institutional review board of Fondazione Policlinico Universitario A Gemelli (DIPUSVSP-PD-03-228). Participants gave informed consent to participate in the study before taking part.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
X @frafanfani
Contributors VV: conceptualization, data curation, formal analysis, methodology, writing-original draft and guarantor. AR: conceptualization, data curation, formal analysis, methodology, and writing-original draft. LT: conceptualization and writing-original draft. DG: formal analysis and methodology. VAC: data curation and methodology. VG: writing-review and editing and supervision. AG, EDS, and MC: data curation. RB and FC: writing-review and editing, supervision, and validation. GS: project administration, supervision, and validation. FF: project administration, writing-review and editing, supervision, and validation.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial, or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.