When Standards Meet Reality: An Inverted PORTEC-3 Protocol for High-Risk Endometrial Cancer in Resource-Limited Settings^†

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

 

1. To be statistical correct - whenever mentioned the word “significant” – need to also report the corresponding p-value to support such statement.   Please change the wording if no p-value to support the word of “significant”. 

 

2. “uni/multivariate” need to change to “uni/multivariable”. In statistics, multivariable means one dependent variable (Y) with multiple independent variables (X's), while multivariate means multiple dependent variables (Y's) with potentially multiple independent variables, analyzing outcomes as a group. The key distinction is the number of outcomes: multivariable focuses on one outcome, adjusting for several predictors; multivariate looks at several related outcomes simultaneously.

 

3. When reports ”median age”, need to also report Interquartile Range (IQR) values.

 

4. Gender information needs to be reported.

 

5. There are several study disadvantages: (1) small sample size (n=52) causing lack of statistical power, (2) lack of large enough sample size, 10 subjects per variable, to conduct true multi-variable analysis and (3) lack of important prior comorbidity index of patients (Charlson comorbidity index or Elixhauser comorbidity index) (4)

   The other disadvantage of the study is lacking a comparison group with different therapy sequence - this could be added into the limitation section.

   – suggesting adding “Limitations” section.

 

6. For a good Receiver operating characteristic (ROC) curve fit, looking for a high Area Under the Curve (AUC), generally above 0.8 (excellent), with values 0.75-0.8 considered acceptable and >0.9 outstanding. This part should be addressed when discussing ROC comparison section or limitation section.

Author Response

Comments 1: [To be statistically correct – whenever mentioned the word ‘significant’ – need to also report the corresponding p-value to support such statement. Please change the wording if no p-value to support the word of ‘significant.]

 

Response 1:
We thank the reviewers for this comment. We agree with the suggestion and have carefully revised the manuscript to ensure that all statements using the word “significant” are now supported by the corresponding p-values. Where no p-value was available, the wording has been modified to remove “significant” and use a more descriptive term instead.

These changes have been made throughout :

-The Results section, paragraph “Treatment related outcomes, (pages 11-12, lines 347-349 and 433).

-The Discussion section, page 18, line 622.

-The Discussion section, page 19, line 713 ans 729.

Comments 2: [“uni/multivariate” need to change to “uni/multivariable”. In statistics, multivariable means one dependent variable (Y) with multiple independent variables (X's), while multivariate means multiple dependent variables (Y's) with potentially multiple independent variables, analyzing outcomes as a group. The key distinction is the number of outcomes: multivariable focuses on one outcome, adjusting for several predictors; multivariate looks at several related outcomes simultaneously.]

Response 2:

Thank you for this helpful clarification. We agree with the reviewer’s comment.

All instances of the terms “univariate” and “multivariate” have been corrected to “univariable” and “multivariable”, respectively, to accurately reflect the statistical approach used.

The Results and Statistical Analysis sections were revised accordingly.

These changes can be found in the revised manuscript at the following locations:

• Methods section, paragraph 4 (Statistical Analysis), page 5, line 241-248

• Results section, paragraph 3 (Univariable and multivariable analyses of prognostic factors), page 10, line 257

• Results section, paragraph 3 (Univariable and multivariable analyses of prognostic factors), page 12, line 446

• Table 3 title, page 11, line 339.

• Table 4 title, page 12, line 442.

• Table 5 title, page 12, line 452

The updated text now reads:

“Univariable and multivariable analyses of prognostic factors.”

Comments 3: [When reports ”median age”, need to also report Interquartile Range (IQR) values..]

Response 3:
We thank the reviewers for this comment. We agree with the suggestion and have accordingly revised the manuscript to report the median age together with the interquartile range (IQR). The change has been made in the Results section, paragraph 1 (Patient Characteristics), page 6, line 271.

 

“Median age was 63 years (IQR 37–76).”

Comments 4: [Gender information needs to be reported]

Response 4:
We thank the reviewer for this important comment. As this study focuses on endometrial cancer, all patients are female. For clarity, gender information has been explicitly stated in the Results section under “Patient Characteristics, page 6, line 258.

Comments 5: [There are several study disadvantages: (1) small sample size (n=52) causing lack of statistical power, (2) lack of large enough sample size, 10 subjects per variable, to conduct true multi-variable analysis and (3) lack of important prior comorbidity index of patients (Charlson comorbidity index or Elixhauser comorbidity index) (4) The other disadvantage of the study is lacking a comparison group with different therapy sequence - this could be added into the limitation section.]

 

Response 5:
We thank the reviewer for this thorough and constructive comment. We agree that these are important considerations. Accordingly, we have expanded the Limitations paragraph in the Discussion section to explicitly acknowledge these points. The revised text now includes mention of the small sample size, the inability to perform a full multivariable analysis due to sample size limitations, the lack of systematic comorbidity indices, and the absence of a comparison group with different therapy sequences. These changes can be found in the Discussion, last paragraph under “Strengths, Limitations, and Future Directions” (page 20, Line 759-761).

Updated text in the manuscript:
" Limitations include its retrospective design, relatively small sample size, which limits statistical power and precludes a full multivariable analysis, and the lack of systematic comorbidity assessment. Comprehensive molecular classification (e.g., p53, MMR, POLE status) was unavailable for all patients, which may impact risk stratification and therapeutic tailoring [3,31].

Prospective trials are required to compare chemotherapy-first versus conventional sequencing definitively, ideally stratified by molecular subtype. Efforts should also be made to minimize delays to radiotherapy, particularly in resource-constrained settings, to optimize outcomes [3,4,16,20]."

 

Comments 6:

[For a good Receiver operating characteristic (ROC) curve fit, looking for a high Area Under the Curve (AUC), generally above 0.8 (excellent), with values 0.75-0.8 considered acceptable and >0.9 outstanding. This part should be addressed when discussing ROC comparison section or limitation section.]

Response 6:

We thank the reviewer for this insightful comment. We agree that commenting on the discriminative ability of our ROC models improves transparency. In the revised manuscript, we have clarified that the area under the curve (AUC) values for the identified bone marrow dose–volume thresholds ranged from 0.6 to 0.88, indicating acceptable to excellent predictive performance. These values support the robustness of our proposed constraints while acknowledging the limitations inherent to retrospective dosimetric analyses. This discussion has been added to the ROC comparison section of the Discussion (page 20, lines 741-747).

During the revision process, we noted that the AUC column in Table 8 was inadvertently reversed during transcription, although the correct results had previously been presented at the ESTRO 2025 and SFRO 2025 conferences. We have corrected this error in the revised manuscript and have clarified the interpretation of AUC values in the ROC comparison section.

The changes can be found in Table 8 (page 17)

4. Response to Comments on the Quality of English Language

Point 1:

The English is fine and does not require any improvement.

Response:

We thank the reviewer for this positive comment and are pleased to note that the English language and style were found to be satisfactory.

 

Additional clarifications:

We would like to clarify that line numbers may differ between the revised manuscript without tracked changes and the tracked-changes version of the manuscript. This discrepancy is due to the inclusion of insertions, deletions, and formatting adjustments visible in the tracked-changes version. For accuracy, all line number references provided in our responses to the reviewers should be interpreted with reference to the tracked-changes version of the revised manuscript.

 

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript addresses a clinically relevant and timely topic, focusing on the feasibility and outcomes of a “chemo-first” adjuvant approach for high-risk endometrial cancer in resource-limited settings. The study provides valuable real-world data and explores an important practical question that is often underrepresented in randomized clinical trials.

The overall study design is appropriate for the research question, and the results are generally coherent and clinically meaningful. Nevertheless, several aspects of the manuscript would benefit from clarification and refinement to strengthen its scientific rigor, transparency, and readability.

1. Introduction
The introduction provides an adequate background; however, it would benefit from a more explicit and structured rationale for the “chemo-first” strategy. In particular, the authors may wish to better differentiate between logistical constraints and potential biological or clinical advantages of this sequencing. In addition, citing more recent international guidelines and real-world studies addressing adjuvant treatment sequencing in high-risk endometrial cancer would further strengthen the context.

2. Methods
The Methods section would benefit from greater precision in several areas:

The definitions of the time intervals analyzed (surgery-to-chemotherapy, chemotherapy-to-radiotherapy, and surgery-to-radiotherapy) should be clearly stated and consistently used throughout the manuscript.

Additional justification for the statistical approaches employed, particularly the use of logistic regression models and ROC curve analyses, would improve transparency. Clarifying how variables were selected for inclusion and whether potential confounders were considered would strengthen the robustness of the analysis.

The description of “pelvic bones” as a surrogate for bone marrow deserves further clarification, including a brief rationale and discussion of its limitations.

3. Results
The results are generally clear; however, their presentation could be improved. Some tables, especially those in the supplementary material, would benefit from a more consistent presentation of hazard ratios, 95% confidence intervals, and p-values. In a few instances, the narrative of the Results section appears to overlap with interpretative statements that might be more appropriate for the Discussion.

4. Discussion
The Discussion appropriately addresses the main findings, but it could be strengthened by a more balanced comparison with existing literature, particularly regarding survival outcomes and toxicity profiles. Expanding on the clinical implications of the observed associations between hematological toxicity and survival outcomes, while clearly acknowledging the limitations of a retrospective and unicentric design, would enhance the interpretative value of this section.

5. Figures and Tables
All figures and tables are relevant; however, improvements in formatting, consistency, and explanatory captions would facilitate interpretation for the reader. Ensuring uniform terminology and abbreviations across all visual elements is recommended.

In summary, this is a well-conducted and clinically relevant study that addresses an important real-world issue. With the suggested clarifications and minor revisions, the manuscript would be significantly strengthened and suitable for publication.

Comments on the Quality of English Language

Overall, the manuscript is understandable and the scientific content can be followed. However, the quality of the English would benefit from a thorough and careful revision to improve clarity, precision, and readability. In several parts of the manuscript, particularly in the Introduction and Discussion sections, sentences are excessively long and contain multiple subordinate clauses, which makes them difficult to follow on first reading. Splitting these sentences into shorter, more direct statements would help convey the key messages more clearly.

There are also recurrent minor issues related to grammar and syntax, including article usage (“the” vs. no article), prepositions, verb tenses, and subject–verb agreement. While none of these errors is critical on its own, their accumulation affects the overall fluency of the text.

In addition, some expressions are imprecise or overly complex and could be replaced by more standard scientific phrasing. Improving consistency in terminology and simplifying certain constructions would enhance readability for an international audience.

A careful language revision by a fluent English speaker with experience in scientific writing is strongly recommended, as it would substantially improve the clarity, flow, and overall quality of the manuscript without altering its scientific content.

Author Response

Comments 1: [The introduction provides an adequate background; however, it would benefit from a more explicit and structured rationale for the “chemo-first” strategy. In particular, the authors may wish to better differentiate between logistical constraints and potential biological or clinical advantages of this sequencing. In addition, citing more recent international guidelines and real-world studies addressing adjuvant treatment sequencing in high-risk endometrial cancer would further strengthen the context.]

 

Response 1: Thank you for this constructive and insightful comment. We agree with the reviewer that the rationale for the “chemo-first” strategy required further clarification and structuring.

The Introduction has been revised to clearly differentiate between logistical constraints related to limited access to radiotherapy in resource-limited settings and the potential biological and clinical considerations supporting early initiation of systemic chemotherapy in high-risk endometrial cancer. We now explicitly discuss the risk of prolonged treatment-free intervals due to radiotherapy delays, as well as the theoretical advantage of early chemotherapy in addressing micrometastatic disease.

In addition, we have strengthened the context by referencing and acknowledging real-world sequencing practices outside randomized clinical trials.

These revisions can be found in the Introduction section, page 3, paragraph 4, lines 126–129.

Updated text in the manuscript:

“This “chemo-first” approach is primarily driven by logistical constraints, but it is also supported by a biological rationale: high-risk EC has a high propensity for early distant relapse, and early chemotherapy may reduce micro-metastatic burden, potentially lowering the risk of distant failure [15].”

 

Comments 2:

[The Methods section would benefit from greater precision in several areas:

The definitions of the time intervals analyzed (surgery-to-chemotherapy, chemotherapy-to-radiotherapy, and surgery-to-radiotherapy) should be clearly stated and consistently used throughout the manuscript.

Additional justification for the statistical approaches employed, particularly the use of logistic regression models and ROC curve analyses, would improve transparency. Clarifying how variables were selected for inclusion and whether potential confounders were considered would strengthen the robustness of the analysis.

The description of “pelvic bones” as a surrogate for bone marrow deserves further clarification, including a brief rationale and discussion of its limitations.]

 

Response 2:

 

Thank you for this detailed and valuable comment. We agree that additional precision and clarification in the Methods section would enhance the transparency and scientific rigor of the manuscript.

First, we have explicitly defined all time intervals analyzed in the study, namely surgery-to-chemotherapy, chemotherapy-to-radiotherapy, and surgery-to-radiotherapy. These definitions are now clearly stated in the Methods section and have been used consistently throughout the manuscript.

These revisions have been incorporated in the Methods section, page 5, paragraph “Follow-Up and Outcome Definitions”, lines 222–227.

 

Proposed revised text for the Methods section

“Time intervals were defined as follows. The surgery-to-chemotherapy interval was calculated from the date of definitive surgery to the start of adjuvant chemotherapy. The chemotherapy-to-radiotherapy interval was calculated from the end of the last chemotherapy cycle to the initiation of pelvic radiotherapy. The surgery-to-radiotherapy interval was defined as the time from surgery to the start of pelvic radiotherapy. These definitions were applied consistently throughout the analysis.”

 

 

Second, we have expanded the description of the statistical methodology to better justify the use of logistic regression models and ROC curve analyses. Logistic regression was employed to identify factors associated with binary clinical outcomes, while ROC curve analyses were used to evaluate the discriminatory performance of the selected models. We also clarified the process of variable selection, indicating that clinically relevant variables and those showing an association in univariable analyses were considered for inclusion in multivariable models, with attention to potential confounding factors.

These revisions have been incorporated in the Methods section, page 5, paragraph “Statistical analysis, lines 243–252.

 

Proposed revised text for the Methods section

Statistical analysis

“Receiver operating characteristic (ROC) curve analysis was performed to assess the association between bone marrow dose–volume parameters (V10 Gy, V20 Gy, V30 Gy, and V40 Gy) and grade ≥ 2 hematologic toxicities. Logistic regression models were used to evaluate factors associated with binary clinical and dosimetric outcomes. Variables included in multivariable analyses were selected based on clinical relevance and results of univariable analyses, and potential confounders were considered to ensure model robustness. Model performance was assessed using ROC curves, with the area under the curve (AUC) used to evaluate discriminative ability. Optimal cut-off values were determined, and statistical significance was defined as p < 0.05.”

 

Finally, we have clarified the rationale for using pelvic bones as a surrogate for pelvic bone marrow in dosimetric analyses. A brief explanation and acknowledgment of the limitations of this surrogate approach have been added, noting that it represents an indirect estimation and may not fully capture functional bone marrow distribution.

These revisions have been incorporated in the Methods section, page 4, paragraph “Study procedures”, lines 194–199.

 

Proposed revised text for the Methods section

[“Pelvic bones were contoured as a surrogate for bone marrow, extending from the inferior border of the ischial tuberosities to 25 mm above the planning target volume (PTV). This method is commonly used in radiotherapy studies due to the practical limitations of directly delineating active bone marrow on planning CT scans. It provides an indirect estimate of bone marrow exposure, although it does not capture inter-patient variability in functional bone marrow distribution.”]

 

Comments 3:

[The results are generally clear; however, their presentation could be improved. Some tables, especially those in the supplementary material, would benefit from a more consistent presentation of hazard ratios, 95% confidence intervals, and p-values. In a few instances, the narrative of the Results section appears to overlap with interpretative statements that might be more appropriate for the Discussion.]

Response 3:

Thank you for this helpful comment. We agree that the clarity and consistency of the Results section could be further improved.

Accordingly, we have revised the presentation of the results to ensure a more consistent reporting of hazard ratios, 95% confidence intervals, and p-values, particularly in the supplementary tables. All tables have been harmonized to follow a uniform format, facilitating interpretation and comparison across analyses.

 

These revisions have been implemented in the Results section , paragraph 3 , page 10, line 332 and Tables 4–6.

 

 

In addition, we carefully reviewed the Results section to remove or rephrase any statements with interpretative content. Interpretative and explanatory comments have been relocated to the Discussion section where appropriate, ensuring a clearer distinction between the presentation of results and their interpretation.

These revisions have been implemented in the Results section:

-Paragraph “Univariable and multivariable analyses of prognostic factors” and  “Treatment-Related Outcomes”, page 10-12, line 317-332 and 347-441.

-Paragraph “treatment related toxicities” page 13-14, line 497-505.

 

Comments 4:

The Discussion appropriately addresses the main findings, but it could be strengthened by a more balanced comparison with existing literature, particularly regarding survival outcomes and toxicity profiles. Expanding on the clinical implications of the observed associations between hematological toxicity and survival outcomes, while clearly acknowledging the limitations of a retrospective and unicentric design, would enhance the interpretative value of this section.

 

Response 4:

Thank you for this thoughtful and constructive comment. We agree that the Discussion section could be further strengthened by a more balanced and nuanced comparison with existing literature and by a clearer acknowledgment of the study’s limitations.

Accordingly, we have revised the Discussion to include a broader comparison of our survival outcomes and toxicity profiles with those reported in pivotal randomized trials and recent real-world studies. This comparison highlights both similarities and differences, thereby placing our findings in a more comprehensive clinical context.

In addition, we expanded the discussion of the clinical implications of the observed associations between hematological toxicity and survival outcomes, with particular attention to their potential impact on treatment tolerance, treatment intensity, and overall outcomes in high-risk endometrial cancer patients. These interpretations are presented cautiously, given the exploratory nature of these analyses.

Finally, we have explicitly reinforced the limitations of our study, emphasizing its retrospective and unicentric design, potential selection bias, and the inherent constraints of real-world data. These limitations are now clearly stated to ensure appropriate interpretation of the results.

These revisions have been incorporated in the Discussion section, page 20 paragraphs “Strengths, Limitations, and Future Directions“, lines 760-764.

 

Comments 5:

[All figures and tables are relevant; however, improvements in formatting, consistency, and explanatory captions would facilitate interpretation for the reader. Ensuring uniform terminology and abbreviations across all visual elements is recommended. In summary, this is a well-conducted and clinically relevant study that addresses an important real-world issue. With the suggested clarifications and minor revisions, the manuscript would be significantly strengthened and suitable for publication.]

 

Response 5:

Thank you for this constructive comment and for the overall positive assessment of our work. We agree that improvements in formatting and consistency would enhance the clarity and interpretability of the figures and tables.

Accordingly, all figures and tables, including those in the supplementary material, have been revised to ensure a more uniform presentation in terms of formatting, terminology, and abbreviations. Captions have been expanded to provide clearer and more self-explanatory descriptions, including definitions of abbreviations and relevant analytical details where appropriate.

These revisions have been implemented throughout the manuscript and supplementary files (Figures 2 and 3; Tables 3,4,5-8).

 

We thank the Reviewer for this positive assessment and believe that the revisions have further strengthened the manuscript.

 

4. Response to Comments on the Quality of English Language

Point 1:

[Overall, the manuscript is understandable and the scientific content can be followed. However, the quality of the English would benefit from a thorough and careful revision to improve clarity, precision, and readability. In several parts of the manuscript, particularly in the Introduction and Discussion sections, sentences are excessively long and contain multiple subordinate clauses, which makes them difficult to follow on first reading. Splitting these sentences into shorter, more direct statements would help convey the key messages more clearly.]

 

Response 1:

Thank you for this careful and constructive assessment of the language quality of our manuscript. We fully agree with the reviewer that the clarity and readability of the English could be further improved.

Accordingly, the manuscript has undergone a thorough and comprehensive language revision. Particular attention was paid to the Introduction and Discussion sections, where long and complex sentences were restructured into shorter and clearer statements to better convey the key messages.

 

Point 2:

[There are also recurrent minor issues related to grammar and syntax, including article usage (“the” vs. no article), prepositions, verb tenses, and subject–verb agreement. While none of these errors is critical on its own, their accumulation affects the overall fluency of the text.]

 

Reponse 2 :

 

We thank the reviewer for this careful assessment of the linguistic quality of the manuscript. Recurrent grammatical and syntactic issues, including article usage, prepositions, verb tenses, and subject–verb agreement, were carefully corrected throughout the text.

 

Point 3:

[In addition, some expressions are imprecise or overly complex and could be replaced by more standard scientific phrasing. Improving consistency in terminology and simplifying certain constructions would enhance readability for an international audience.

A careful language revision by a fluent English speaker with experience in scientific writing is strongly recommended, as it would substantially improve the clarity, flow, and overall quality of the manuscript without altering its scientific content.]

 

 

Response 3:

We fully agree with the reviewer’s comment. Accordingly, imprecise or overly complex expressions were revised and replaced with more standard scientific phrasing, and terminology was harmonized throughout the manuscript to improve consistency and readability for an international audience. This language revision was performed without altering the scientific content of the manuscript.

The manuscript was revised by a fluent English speaker with experience in scientific writing. All changes have been highlighted in the revised version using track changes.

 

 

 

 

Additional clarifications

We would like to clarify that line numbers may differ between the revised manuscript without tracked changes and the tracked-changes version of the manuscript. This discrepancy is due to the inclusion of insertions, deletions, and formatting adjustments visible in the tracked-changes version. For accuracy, all line number references provided in our responses to the reviewers should be interpreted with reference to the tracked-changes version of the revised manuscript.

 

 

 

 

 

 

Author Response File: Author Response.pdf