Review Reports - Impact of Facility Volume on Therapy and Survival for Endometrial Cancer: A Retrospective Multicenter Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Editor,

Thanks for inviting me to review the manuscript entitled “Impact of Facility Volume on Therapy and Survival for Endometrial Cancer: A Retrospective Multicenter Study.” The topic is clinically relevant and addresses an important healthcare policy issue regarding centralization of oncologic care. However, the manuscript requires substantial language editing and clarification of several methodological and conceptual aspects. My detailed comments are provided below.

General Comments

Overall, the language and grammar need to be strongly improved. Numerous grammatical errors, unclear phrasing, and typographical mistakes impair readability and scientific clarity. I strongly recommend professional language editing prior to publication.

Specific Comments

Introduction and Study Objective

Line 47: Please add a clear study objective, for example: “The objective of our study was to test the impact of facility volume on treatment patterns and survival outcomes in endometrial cancer.”
Line 48: Please correct “multicenters study comparing” to “multicenter study comparing.”

Typographical and Formatting Errors

Line 90: The sentence “Early articles suggested ≥ 10 cases per year 10 and more recent articles ≥ 80 cases per year [11]” contains a duplicated number (“10”), likely left unintentionally. Please correct this.
Line 101–102 and Line 165: The number “17” appears to be intended as a citation but is not formatted correctly. Please insert the appropriate reference or remove the number.
Citation 16: Please clarify this reference, as it is currently unclear how it supports the statement in the text.

Role of Imaging and High-Volume Centers

In the Introduction, stronger emphasis should be placed on the critical role of imaging in preoperative staging of endometrial cancer, particularly for assessing myometrial invasion and lymph node involvement.
This staging should ideally be performed in high-volume centers with specialized expertise, which is directly relevant to the centralization hypothesis.
Please cite:

PMID: 40335530
PMID: 38473269

Definition of Facility Volume Thresholds

Line 103: The manuscript defines facility volume thresholds, but it is unclear which studies support these cut-offs. Please cite at least two articles that justify the chosen volume categories and explain the rationale for these thresholds.

Data Source and Study Population

Please clearly specify which hospitals were included, which database was used, and the exact study timeframe (years). This information is essential for transparency and reproducibility.

Centralization and Waiting Times

The manuscript argues for centralization of care but does not address potential unintended consequences.
Centralization could lead to longer waiting lists and treatment delays, which may negatively impact survival outcomes.
This important issue should be discussed in the Discussion section, with reference to existing literature on healthcare system capacity and treatment delays.

Summary

This study addresses an important question regarding the relationship between facility volume and outcomes in endometrial cancer. However, major language revision and clarification of methodological details are required. Strengthening the Introduction, clarifying references and definitions, and discussing potential drawbacks of centralization will significantly improve the manuscript’s scientific rigor and clinical relevance.

Author Response

Dear Editor,

General Comments

Specific Comments

Introduction and Study Objective

Line 47: Please add a clear study objective, for example: “The objective of our study was to test the impact of facility volume on treatment patterns and survival outcomes in endometrial cancer.”

Reply

We revised as requested:

“This study aimed to evaluate the influence of facility volume on treatment strategies and survival outcomes among EC patients.”

Line 48: Please correct “multicenters study comparing” to “multicenter study comparing.”

Reply

We revised as requested:

Typographical and Formatting Errors

Line 90: The sentence “Early articles suggested ≥ 10 cases per year 10 and more recent articles ≥ 80 cases per year [11]” contains a duplicated number (“10”), likely left unintentionally. Please correct this.

Reply

We revised as requested.

Line 101–102 and Line 165: The number “17” appears to be intended as a citation but is not formatted correctly. Please insert the appropriate reference or remove the number.

Reply

We revised as requested:

Reference [17] refers to our previous work, which provides a detailed description of the study population analyzed in the current manuscript.

Citation 16: Please clarify this reference, as it is currently unclear how it supports the statement in the text.

We have revised the connection and it now works correctly. You can consult the Sixth National Survey on the state of implementation of the Regional oncology networks. Emilia Romagna is a region with high centralization (94%):

https://www.agenas.gov.it/images/Rapporto_ROR_versione_finale_29052025_vers_rettifica.pdf

Role of Imaging and High-Volume Centers

In the Introduction, stronger emphasis should be placed on the critical role of imaging in preoperative staging of endometrial cancer, particularly for assessing myometrial invasion and lymph node involvement.

This staging should ideally be performed in high-volume centers with specialized expertise, which is directly relevant to the centralization hypothesis.
Please cite: Arezzo F, Fanizzi A, Mancari R, Cocco E, Bove S, Comes MC, Gianciotta M, Lanza G, Lopez S, Cazzato G, Silvestris E, Vitale E, Vizza E, Cormio G, Massafra R, Loizzi V. A Radiomic-based model to predict the depth of myometrial invasion in endometrial cancer on ultrasound images. Sci Rep. 2025 May 7;15(1):15901. doi: 10.1038/s41598-025-00906-6. PMID: 40335530. Madár I, Szabó A, Vleskó G, Hegyi P, Ács N, Fehérvári P, Kói T, Kálovics E, Szabó G. Diagnostic Accuracy of Transvaginal Ultrasound and Magnetic Resonance Imaging for the Detection of Myometrial Infiltration in Endometrial Cancer: A Systematic Review and Meta-Analysis. Cancers (Basel). 2024 Feb 23;16(5):907. doi: 10.3390/cancers16050907. PMID: 38473269; PMCID: PMC10931203.

Reply

We revised as requested:

“A high-volume center should ensure both the diagnostic phase and the treatment of pa-tients with endometrial cancer. Accurate preoperative diagnosis could enable the identifi-cation of low-risk endometrial cancer cases that may be managed in lower-volume centers [9]. Preoperative staging relies on expert-led transvaginal or transrectal ultrasonography (US) and pelvic Magnetic Resonance Imaging (MRI) to evaluate myometrial invasion, cer-vical stromal involvement, and adnexal status. Furthermore, chest-abdominal-pelvic Computed Tomography (CT) or Positron Emission Tomography (PET) scans are recom-mended to exclude nodal metastases, peritoneal carcinomatosis, and distant secondary le-sions [6]. Currently, US and MRI are the primary modalities employed for the preoperative assessment of myometrial invasion depth. While both techniques demonstrate comparable diagnostic efficacy, US remains the preferred first-line approach due to its cost-effectiveness and shorter acquisition times [9,10, 17, 18]. Nevertheless, the accuracy of US-based preoperative staging can be significantly compromised in complex clinical sce-narios. For instance, the presence of concomitant uterine pathologies, such as leiomyomas or adenomyosis, can obscure the tumor-myometrial interface [19]. Notably, in low-risk EC patients, the specificity of MRI is significantly superior [20]. Furthermore, unlike the standardized protocols of MRI, US is inherently operator-dependent; its diagnostic relia-bility is highly contingent on the clinician's expertise, which may lead to substantial in-ter-observer variability in staging results. Today, the integration of the US with radiomic models has made it possible to standardize subjective evaluations, overcoming critical is-sues and significantly improving accuracy in identifying myometrial infiltration [21].“

Definition of Facility Volume Thresholds

Line 103: The manuscript defines facility volume thresholds, but it is unclear which studies support these cut-offs. Please cite at least two articles that justify the chosen volume categories and explain the rationale for these thresholds.

Reply

We added these information as requested:

Due to the considerable variability in thresholds used to define “high volume” facilities, in contrast to a broader consensus on “low volume” definitions in the literature, we established our classification based on a synthesis of international data. Specifically, low-volume hospitals were defined as those treating up to 10 cases per year [10, 14]. To define medium-volume hospitals, we harmonized European [13,15] and non-European studies [12, 14, 24], the latter often using higher thresholds. Consequently, we defined medium-volume centers (MVCs) as those treating 11 to 29 cases/year and high-volume centers (HVCs) as those treating ≥30 cases/year
Data Source and Study Population

Please clearly specify which hospitals were included, which database was used, and the exact study timeframe (years). This information is essential for transparency and reproducibility.

Reply

We added information as requested:

“This multi-center study included all patients who underwent hysterectomy for endometrial cancer (EC) between 2000 and 2019 across seven hospitals in the Emilia-Romagna region of Northern Italy: the University Hospitals of Bologna, Modena and Reggio Emilia, Parma, and Ferrara, as well as the USL-IRCCS of Reggio Emilia, Ospedale di Forlì, and Ospedale degli Infermi (Rimini) [17]”.

Centralization and Waiting Times

The manuscript argues for centralization of care but does not address potential unintended consequences.
Centralization could lead to longer waiting lists and treatment delays, which may negatively impact survival outcomes.
This important issue should be discussed in the Discussion section, with reference to existing literature on healthcare system capacity and treatment delays.

Reply

This issue was discussed as requested

“Centralization and treatment by gynecologic oncologists have been associated with improved quality of care, with lower postoperative morbidity and adequate surgical staging. Consequently, improved patient selection for adjuvant treatment, particularly chemotherapy, has been observed, especially in high-risk patients [28-30]. Moreover, treatment in HVC may reduce the survival disparity between black and white women with EC [14]. However, the centralization of care for EC is not without criticism. While it is widely associated with better surgical outcomes and adherence to guidelines, it presents several logistical and socio-ethical challenges. Centralization inevitably leads to longer travel distances for patients living in rural or peripheral areas. This can be physically exhausting for patients who are often elderly or have comorbidities, travel costs, lost wages for caregivers, and the logistical difficulty of arranging transportation can create significant barriers, potentially exacerbating existing socioeconomic inequalities in healthcare access. Centralization may favor patients living in urban centers near tertiary hospitals, while those in peripheral regions may suffer from delayed diagnosis or suboptimal initial management. If the number of designated reference centers is too low, these institutions may become overwhelmed. This can lead to increased waiting times for surgery or consultations, which may negatively affect oncological outcomes. Several papers investigated the impact of the delay of surgery on EC oncological outcomes [31-38]. The experience of the COVID-19 pandemic has demonstrated how diagnostic delay can lead to a diagnostic de-lay with a consequent increase in the stage and aggressiveness of EC at diagnosis [39]. Another study reported that delaying surgery by more than eight weeks had no impact on patient mortality, but increased the need for adjuvant pelvic radiation therapy and worsened recurrence rates. [31]. A previous large population-based study demonstrated that patients experiencing wait times exceeding 12 weeks exhibited significantly poorer survival outcomes compared to those treated within the 2.1–12 -week interval [32]. However, particularly in low-risk patient populations, treatment delays showed little or no correla-tion with stage progression, likely due to a greater interaction with tumor biology [33, 34]. On the contrary, a previous paper showed that treatment delays were identified as a significant risk factor for mortality exclusively in low-risk EC, likely reflecting restricted access to specialized care. However, authors concluded that referring to an experienced surgical team and thorough preoperative optimization should be prioritized over expedited surgery [35]. A new paper reported that surgical delays beyond four weeks were associated with reduced PFS and increased recurrence [36]. Recent reviews showed that the most common cut-off for the time to surgery interval was six weeks, and for the time to adjuvant treatment, nine weeks. Only 24- 74 % of EC patients were treated according to this indication [37,38]. These reviews suggested performing surgery between two and eight weeks from diagnosis [37,38]. While the centralization of EC patients within HVCs may result in surgical delays and a dilution of clinical proficiency among gynecologists in LVCs—potentially compromising their readiness to manage unexpected complication—this organizational model remains preferable and should be vigorously supported. However, HVCs must extend their role beyond the surgical procedure to lead multidisciplinary management, ranging from the selection of adjuvant therapies to the coordination of follow-up. Simultaneously, they must provide consistent support and continuous training to LVCs to ensure their operational resilience.”

Summary

Reviewer 2 Report

Comments and Suggestions for Authors

The paper compares HVC and MVC regarding the outcomes of treatment and therapy for EC patients. It is a large multi-centre study involving 2402 endometrial cancer patients treated between 2000 and 2019 either at HVC ( 1431 patients) or MVC (971 patients). The patients differed in mean age, ASA score, additional comorbidities, surgical approach, adjuvant treatment.

The paper focuses on which type of centre achieved best results for EC patients, concluding that HVC patients presented better PFS and OS. Authors propose that centralized treatment is of higher quality regarding recovery post-surgery and delivers better outcomes for patients with more aggressive cancers.

26 references are appropriately cited within the paper, tables and figures are clear and accessible for the reader.

This paper is of high value from a clinical perspective.

Author Response

Comments and Suggestions for Authors

26 references are appropriately cited within the paper, tables and figures are clear and accessible for the reader.

This paper is of high value from a clinical perspective.

Reply

We would like to thank the reviewer for their time and valuable expertise in evaluating our manuscript

Reviewer 3 Report

Comments and Suggestions for Authors

This is a retrospective, non‑randomized comparison between MVC and HVC across almost 20 years (2000–2019), with clear imbalances at baseline: HVC patients are younger, have fewer comorbidities, better ASA, and more high‑risk disease.: These variables are strong confounders of both treatment decisions and outcomes.

Several key variables have substantial missingness and it differs significantly between groups (e.g. total lymph nodes: missing in 158 MVC vs 643 HVC; comorbidities: 53 vs 446 missing; hemoglobin variation and others likewise) The manuscript does not explain how missing data were handled (complete‑case analysis, exclusion from models, imputation). Given the imbalance in missingness between MVC and HVC, complete‑case analyses may introduce bias.

Justify more explicitly why only MVC vs HVC (no LVC) are analysed here and how this might bias towards smaller differences (region already highly centralized).

PFS and OS are defined, but there is no description of how recurrence dates were ascertained, whether competing risks (death without recurrence) were considered in PFS, or how loss to follow‑up was handled.

The conclusion should distinguish between: (a) strong, data‑supported statements (e.g. lower perioperative morbidity and better PFS in selected risk groups) and (b) more speculative policy recommendations.

HVC patients received adjuvant therapy less frequently overall but more often chemotherapy when treated, and this difference is particularly marked in high‑risk cases. In the Cox model, adjuvant therapy is associated with worse PFS (HR >1), which is expected because of confounding by indication (higher‑risk patients receive more adjuvant therapy). This is not adequately discussed.

The observation that HVC patients show more loco‑regional and fewer extra‑abdominal relapses is interesting and plausible (better staging, systemic therapy) but the analysis is descriptive and not adjusted. Small numbers in some subgroups and lack of molecular data may limit interpretation, especially for patterns of distant recurrence in non‑endometrioid and p53‑abn‑enriched populations. Present numbers (not only percentages) of recurrence types by center in the main text or supplementary table and acknowledge that these exploratory findings could be influenced by stage/histotype distributions and follow‑up intensity.

The narrative sometimes over‑emphasizes the benefit of centralization relative to the heterogeneousliterature, which includes neutral or modest effects and differences by stage, histotype and geography. Provide a more structured summary, e.g.:studies showing clear benefit for high‑risk or advanced disease,studies showing benefit mainly in early‑stage or type 1 EC., Studies with null or modest effects.

Author Response

REVIEWER 3

Comments and Suggestions for Authors

Overall, the paper addresses a relevant question with a large, real‑world cohort, but there are important issues in the statistical methods, interpretation of results, and completeness/positioning of the discussion and literature that should be addressed This is a retrospective, non‑randomized comparison between MVC and HVC across almost 20 years (2000–2019), with clear imbalances at baseline: HVC patients are younger, have fewer comorbidities, better ASA, and more high‑risk disease.: These variables are strong confounders of both treatment decisions and outcomes.

Reply

“We thank the Reviewer for this insightful observation. In our study, we initially adopted a variable-by-variable (pairwise) approach, utilizing all available data for each specific analysis to maximize the use of the collected information. However, the point raised by the Reviewer prompted us to conduct a thorough audit of our database, which allowed us to identify a significant classification error in our initial report. Regarding variables such as total lymph node count, percentage of positive lymph nodes, and Grading, we realized that we incorrectly labeled "Not Applicable" cases as "Missing." Specifically, for total lymph node count and the percentage of positive lymph nodes, we wrongly classified data from patients who did not undergo lymphadenectomy as "missing." In the same way, we incorrectly classified as missing the grading for patients with non-endometrioid tumors, for whom standard grading is not clinically applicable. In the revised manuscript, we have corrected these classifications in Table 1, which significantly reduces the perceived imbalance in missingness between the MVC and HVC groups for these surgical and pathological variables.

Regarding the variables where a substantial amount of missing data remains (e.g., comorbidities and hemoglobin variation), we acknowledge this as a limitation of our study. This missingness is primarily due to the retrospective nature of the data collection across different centers, where some secondary clinical details were not consistently recorded. We have now addressed this point more explicitly in the Limitations section of the Discussion, highlighting that while we utilized all available data, the imbalance in reporting between groups should be taken into account when interpreting these specific secondary variables.”

Justify more explicitly why only MVC vs HVC (no LVC) are analysed here and how this might bias towards smaller differences (region already highly centralized).

Reply

“In the Emilia-Romagna region, surgical cases are highly centralized; consequently, all study centers were medium- or high-volume. Although comparing low-volume to high-volume centers might have yielded more significant disparities, it would not have been representative of our current practice. Moreover, analyzing medium- versus high-volume centers helps identify the optimal volume targets necessary to achieve substantial improvements in patient care”

Reply

“We thank the Reviewer for the opportunity to clarify these methodological aspects.

Overall Survival (OS) was defined as the time from surgery to death from any cause. For Progression-Free Survival (PFS), the date of recurrence was ascertained by the onset of symptoms or during routine follow-up imaging, such as transvaginal ultrasound and computed tomography.

In our analysis, PFS was specifically focused on tumor progression as the primary event. Consequently, patients who died without documented evidence of disease progression were censored at the date of death, as the retrospective nature of the study did not always allow us to reliably distinguish between cancer-related and non-cancer-related deaths. Furthermore, all patients who were lost to follow-up or reached the end of the study period without an event (recurrence or death) were censored at the date of their last known clinical contact. We have clarified this methodological point in the Statistical Methods section; however, we understand the Reviewer’s point regarding the distinction between PFS and TTP. While we have maintained the term PFS in this version of the manuscript, we are fully prepared to redefine this endpoint as Time to Progression (TTP) should the Reviewer deem it necessary.

For OS, the multivariable model shows that the center effect disappears when adjusting for age, ASA and ESMO‑ESGO risk, which is appropriate, but this contradicts the rather strong language in the conclusions about survival. The conclusion should distinguish between: (a) strong data‑supported statements (e.g. lower perioperative morbidity and better PFS in selected risk groups) and (b) more speculative policy recommendations

Reply.

“Multivariate analysis demonstrated that HVC was an independent predictor of improved PFS; however, this protective effect diminished when accounting for heterogeneous recurrence risk groups, a trend even more pronounced in the OS analysis”

Reply

“In the Cox proportional hazards model, adjuvant therapy was associated with worse progression-free survival (PFS) (HR > 1), particularly within the total cohort. This finding was anticipated and likely reflects confounding by indication, as patients with higher-risk clinical features were more frequently selected for adjuvant treatment. Consequently, when the analysis was restricted to homogeneous risk classes, the negative impact of adjuvant therapy on PFS was substantially attenuated.”

Reply

We have added observations regarding the interpretation of the data into the main text:

“However, the interpretability of our findings may be constrained by the limited sample size within specific subgroups and the absence of comprehensive molecular data. This limitation is par-ticularly relevant when evaluating distant recurrence patterns in non-endometrioid and p53-abnormal (p53-abn) enriched populations.”

The narrative sometimes over‑emphasizes the benefit of centralization relative to the heterogeneous literature, which includes neutral or modest effects and differences by stage, histotype and geography. Provide a more structured summary, e.g.:studies showing clear benefit for high‑risk or advanced disease, studies showing benefit mainly in early‑stage or type 1 EC., Studies with null or modest effects.

Reply

A structured summary was reported in Conclusion section

“Centralization of EC patients should be promoted because it is associated with lower perioperative morbidity and better quality of care. Patients treated at HVC, especially those at high risk of recurrence, underwent more frequent LPS and achieved more accurate sur-gical staging (higher counts of biopsies, resected nodes, and positive nodes) than those at MVC. Patients treated at HVC showed longer PFS, in particular in intermediate-high- and high-risk patients. All EC patients could be centralized to receive higher quality treatment to improve recovery from surgery and improve oncologic outcome, particularly for patients with more aggressive cancers. New studies are needed to establish a globally uniform minimum annual case volume of treated EC patients required to designate a hospital as a referral center.”

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

After the completed corrections I support the acceptance of the manuscript.