Review Reports - The Comprehensive Repair–Inflammation Index (CRII) Predicts Tooth Extraction After Chemoradiotherapy: A Continuous and Nonlinear Modeling Analysis

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript entitled “The Comprehensive Repair–Inflammation Index (CRII) Predicts Tooth Extraction After Chemoradiotherapy” presents a well-conducted retrospective analysis addressing a clinically relevant and insufficiently explored dimension of post-chemoradiotherapy dental outcomes. The study is methodologically robust, clearly structured, and statistically sophisticated, particularly in its use of continuous modeling and formal assessment of nonlinearity.

The introduction provides a coherent and up-to-date rationale, appropriately framing the limitations of dose-centric approaches and justifying the exploration of host-related factors. The research objective is clearly articulated and aligns well with the analytical strategy implemented.

The methodological section is comprehensive and transparent. The inclusion and exclusion criteria are well defined, and the clinical and radiographic assessment protocols are described with sufficient detail to ensure reproducibility. The definition and operationalization of the CRII are clearly presented, and the decision to model the predictor as a continuous variable—supplemented by spline-based nonlinear analysis—represents a significant methodological strength. The statistical approach is appropriate for the study design, and the efforts to address overfitting, collinearity, and model calibration are commendable .

The results are clearly reported and logically organized. The consistency of the association between CRII and tooth extraction across multiple analytical frameworks strengthens the internal validity of the findings. The identification of a nonlinear relationship and the cautious interpretation of the exploratory cutoff are particularly well handled, avoiding common pitfalls associated with dichotomization .

The discussion is balanced and appropriately contextualized within the existing literature. The authors correctly avoid overinterpretation and acknowledge the limitations inherent to the retrospective design, including potential residual confounding and the influence of clinical decision-making on the outcome definition. The consideration of the composite nature of tooth extraction as an endpoint is especially pertinent and demonstrates critical methodological awareness.

The limitations section is thorough and transparent, addressing key issues such as the lack of tooth-level dosimetry, potential variability in clinical decision-making, and the need for external validation of the identified threshold . This strengthens the credibility of the conclusions.

From a reporting perspective, adherence to STROBE guidelines is explicitly documented and appears complete, although the inclusion of a participant flow diagram could further enhance transparency, as acknowledged by the authors in the checklist.

Overall, the manuscript meets high standards of scientific rigor, clarity, and clinical relevance. The integration of systemic biomarkers into risk stratification frameworks represents a meaningful contribution to the field and may have important implications for personalized dental management in oncology settings.

Author Response

Response to Reviewer 1

Dear Reviewer,

We sincerely thank you for your comprehensive, thoughtful, and highly constructive evaluation of our manuscript entitled “The Comprehensive Repair–Inflammation Index (CRII) Predicts Tooth Extraction After Chemoradiotherapy.” We greatly appreciate your positive assessment of the scientific rigor, methodological transparency, and clinical relevance of our study.

Below, we provide a point-by-point response to your comments.

Comment 1: The manuscript entitled ‘The Comprehensive Repair–Inflammation Index (CRII) Predicts Tooth Extraction After Chemoradiotherapy’ presents a well-conducted retrospective analysis addressing a clinically relevant and insufficiently explored dimension of post-chemoradiotherapy dental outcomes. The study is methodologically robust, clearly structured, and statistically sophisticated, particularly in its use of continuous modeling and formal assessment of nonlinearity.”

Response 1: We sincerely thank the reviewer for this positive and encouraging assessment of our work. We appreciate the recognition of the study’s methodological rigor, clarity of presentation, and the use of advanced analytical approaches.

Comment 2: “The introduction provides a coherent and up-to-date rationale, appropriately framing the limitations of dose-centric approaches and justifying the exploration of host-related factors. The research objective is clearly articulated and aligns well with the analytical strategy implemented.”

Response 2: We sincerely thank the reviewer for this positive and thoughtful assessment of the Introduction. We appreciate the recognition of the study rationale and the alignment between the research objective and the analytical approach.

Comment 3: “The methodological section is comprehensive and transparent. The inclusion and exclusion criteria are well defined, and the clinical and radiographic assessment protocols are described with sufficient detail to ensure reproducibility. The definition and operationalization of the CRII are clearly presented, and the decision to model the predictor as a continuous variable—supplemented by spline-based nonlinear analysis—represents a significant methodological strength. The statistical approach is appropriate for the study design, and the efforts to address overfitting, collinearity, and model calibration are commendable.”

Response 3: We sincerely thank the reviewer for this positive and detailed assessment of the methodological aspects of our study. We appreciate the recognition of the transparency, reproducibility, and rigor of the study design and statistical approach.

Comment 4: “The results are clearly reported and logically organized. The consistency of the association between CRII and tooth extraction across multiple analytical frameworks strengthens the internal validity of the findings. The identification of a nonlinear relationship and the cautious interpretation of the exploratory cutoff are particularly well handled, avoiding common pitfalls associated with dichotomization.”

Response 4: We sincerely thank the reviewer for this positive and encouraging assessment of the Results section. We appreciate the recognition of the clarity of presentation, the consistency of the findings across analytical approaches, and the careful handling of the nonlinear analysis and exploratory cutoff.

Comment 5: The discussion is balanced and appropriately contextualized within the existing literature. The authors correctly avoid overinterpretation and acknowledge the limitations inherent to the retrospective design, including potential residual confounding and the influence of clinical decision-making on the outcome definition. The consideration of the composite nature of tooth extraction as an endpoint is especially pertinent and demonstrates critical methodological awareness.”

Response 5: We sincerely thank the reviewer for this positive and thoughtful assessment of the Discussion. We appreciate the recognition of the balanced interpretation, methodological considerations, and the emphasis on the composite nature of tooth extraction as an endpoint.

Comment 6:“The limitations section is thorough and transparent, addressing key issues such as the lack of tooth-level dosimetry, potential variability in clinical decision-making, and the need for external validation of the identified threshold. This strengthens the credibility of the conclusions.”

Response 6: We sincerely thank the reviewer for this positive and thoughtful assessment of the limitations section. We appreciate the recognition of its clarity and transparency and its role in strengthening the overall interpretation of our findings.

Comment 7: “From a reporting perspective, adherence to STROBE guidelines is explicitly documented and appears complete, although the inclusion of a participant flow diagram could further enhance transparency, as acknowledged by the authors in the checklist.”

Response 7: We thank the reviewer for this constructive suggestion. In response, we have now included a flow diagram of patient selection (Figure 1) to enhance transparency in accordance with STROBE recommendations. The diagram summarizes the initial cohort, applied exclusion criteria, and final study population, providing a clear overview of participant selection and data availability.

Comment 8: “Overall, the manuscript meets high standards of scientific rigor, clarity, and clinical relevance. The integration of systemic biomarkers into risk stratification frameworks represents a meaningful contribution to the field and may have important implications for personalized dental management in oncology settings.”

Response 8: We sincerely thank the reviewer for this positive and encouraging assessment of our work. We appreciate the recognition of the study’s scientific rigor, clarity, and potential clinical relevance.

Once again, we thank the reviewer for the careful evaluation of our manuscript and for the insightful and encouraging comments, which have further strengthened the quality and clarity of our work.

Sincerely,

Efsun Somay, DDS, PhD

On behalf of all authors

Reviewer 2 Report

Comments and Suggestions for Authors

This is a well‑designed retrospective study addressing a clinically relevant problem. The CRII is a novel composite biomarker, and your use of restricted cubic splines plus segmented regression to explore nonlinearity is a clear strength. The overall findings are plausible and potentially useful for risk stratification. However, several issues related to presentation, language clarity, and some methodological details need attention before the manuscript is ready for publication. Below I list them in order of importance.

Major comments (methodology and interpretation)

1. Duplicated section (editorial error)

In the PDF, section 2.2 (Baseline Clinical Oral Examination) appears twice verbatim (page 4). This is likely a copy‑paste mistake. Please remove the duplicate.

2. Clarify the CRII calculation and scaling

You state that CRII was scaled by 10⁻³ (divided by 1000) for presentation.

Question: Was the same scaled value used in all regression analyses (logistic, splines, segmented)?
Suggestion: Add a clear sentence in Methods: "All regression analyses used the scaled CRII (divided by 1000) to improve coefficient interpretability."

3. Missing key covariates from the multivariable model

In section 2.2 you describe detailed baseline oral health variables (number of teeth, decayed teeth, residual roots, periodontal history). None of these were included in the regression models.

Concern: These are strong potential confounders. A patient with poor baseline dentition may need extractions regardless of CRII.
Request: Either (a) include the most relevant ones (e.g., number of decayed teeth, number of remaining teeth) in an additional sensitivity model, or (b) explicitly justify their exclusion (e.g., limited events per variable, or high collinearity) and state this as a limitation more prominently.

4. Dosimetric parameters justification

You used V50 and V50 as binary indicators (≥1 cc vs <1 cc).

Question: Why 1 cc? Is this a standard threshold in the literature?
Suggestion: Provide a reference or a brief rationale. If it was an arbitrary choice, state that clearly and add a sensitivity analysis using different cutoffs (e.g., ≥0.5 cc or ≥2 cc) to show robustness.

5. Wording of the conclusion about mandibular dose

In the Abstract and main conclusion you write: "Mandibular dose parameters were not significant."

Problem: This may mislead readers into thinking that radiation dose does not matter at all. Your dosimetric assessment was at the mandibular organ level, not tooth‑ or socket‑specific.
Suggested revision: Change to: "Mandibular dose parameters (mean dose, V50, V60) were not significantly associated with TE in our organ‑level analysis, but this does not rule out tooth‑level dose effects."
Do this in the Abstract, Results (end of section 3.4), and Discussion.

Minor comments (clarity and presentation)

6. English language

The English is generally understandable but contains multiple stylistic issues, overly long sentences, odd prepositions, and some unidiomatic expressions.

Examples:
- "nutritional compromise" → better "nutritional deficits"
- "not simply a routine procedure" → "not merely a routine procedure"
- Overly long sentence on page 3 (lines 66–78) describing previous studies → break into two or three shorter sentences.
Recommendation: Please have the manuscript reviewed by a native English speaker or a professional scientific editing service (many MDPI journals offer this). This is not grounds for rejection, but it does affect readability.

7. Table 2 formatting

The row for "CRII (continuous, per 10‑unit increase)" does not clearly indicate whether the reported OR is unadjusted or adjusted.

Suggestion: Split this into two rows: one for unadjusted OR and one for adjusted OR, or add a footnote clarifying.

8. Figures and Graphical Abstract

Figure 2 is good, but consider adding a rug plot (histogram) along the x‑axis to show the distribution of CRII values, so readers see where data are sparse at extremes.
Graphical Abstract: Your PDF shows a blank space. Make sure to upload the actual figure, and specify whether it was created with AI assistance (see ethical comment below).

Ethical and integrity comments

9. AI use declaration (need more detail)

You state that ChatGPT was used "to assist in the preparation of schematic figures".

Clarity needed: Which figures? If only the Graphical Abstract or conceptual diagrams, that is fine. However, if Figures 1 or 2 (which display actual study data) were generated or substantially modified by AI, that would be problematic because data figures must be fully reproducible by the authors.
Request: Specify in the declaration: *"ChatGPT was used only to draft the layout of the graphical abstract. All figures showing study data (Figures 1 and 2) were created independently by the authors using R/SPSS."*

10. Self‑citations (not inappropriate, but a suggestion)

You cite several of your own previous indices (GLUCAR, CARWL, SII). These are relevant, so no misconduct. However, to improve balance, consider citing at least one study from another independent group that has used composite inflammatory indices in nasopharyngeal carcinoma (e.g., NLR or PLR studies from Asian or European centers not involving the same co‑authors). This would show that the field is not exclusively yours.

Comments on the Quality of English Language

The manuscript is written by non-native English speakers (affiliations in Turkey). The scientific content is understandable, but grammatical awkwardness, odd prepositions, excessively long sentences, and occasional unidiomatic expressions reduce clarity and give a less professional impression than the science deserves. A light to moderate professional editing service would bring the language to an acceptable standard for publication in JCM.

Below I provide specific, actionable examples grouped by error type, followed by a short list of corrected sentences.

1. Overly long and convoluted sentences

Original (Introduction, page 3):

"Previous studies evaluating such composite markers have demonstrated associations with radiation‑induced dental complications in patients with LA‑NPC [14‑16]. However, these studies have predominantly relied on categorized or cutoff‑based analyses and have not evaluated associations within a continuous modeling framework or formally assessed potential nonlinear effects, approaches that, according to established statistical guidance, may obscure underlying risk gradients and limit the ability to capture the complexity of biological responses [17‑19]."

Problem: The second sentence runs for nearly 50 words with nested clauses.
Suggested split:

"However, these studies have predominantly relied on categorized or cutoff‑based analyses. They have not evaluated associations within a continuous modeling framework or formally assessed potential nonlinear effects. According to established statistical guidance, such approaches may obscure underlying risk gradients and limit the ability to capture the complexity of biological responses [17‑19]."

3. Unidiomatic or imprecise word choices

Original	Suggested correction	Reason
"nutritional compromise"	"nutritional deficits"	"Compromise" is vague in this context
"not simply a routine procedure"	"is not merely a routine procedure"	More formal academic tone
"tooth extraction is a clinically consequential event"	"tooth extraction represents a clinically consequential event"	More natural verb choice
"we retrospectively analyzed"	Fine, but consider "we conducted a retrospective analysis of"	Slightly more natural
"the relationship was not strictly linear"	Fine	–
"robust biological and clinical evidence"	"substantial biological and clinical evidence"	"Robust" is overused; "substantial" works better

4. Missing or unnecessary articles (a/an/the)

Original (page 4):

"Eligible patients were aged ≥18 years, had histopathologically confirmed squamous cell carcinoma"

Suggestion:

"Eligible patients were aged ≥18 years and had histopathologically confirmed squamous cell carcinoma" (no article needed – correct as is)

But here (page 5):

"Extraction decisions were made by the treating dental team following comprehensive clinical and radiographic evaluation, with consideration of the impaired healing capacity"

Suggestion:

"with consideration of impaired healing capacity" (remove "the" – it is a general concept)

5. Verb tense consistency

Original (Discussion, page 9):

"In this study, CRII demonstrated a robust association with post-CCRT TE, with consistent findings across multiple analytical approaches. The relationship between CRII and TE was not strictly linear, with evidence of a nonlinear pattern and an exploratory threshold beyond which TE risk increased substantially, and this association persisted after adjustment."

Problem: Mixing past tense (demonstrated, persisted) with present tense (is) is fine, but the sentence remains too long.
Suggested revision:

"In this study, CRII demonstrated a robust association with post‑CCRT TE, with consistent findings across multiple analytical approaches. The relationship between CRII and TE was not strictly linear; we found evidence of a nonlinear pattern and an exploratory threshold beyond which TE risk increased substantially. This association persisted after adjustment for clinical and dosimetric variables."

6. Punctuation errors (commas, semicolons)

Original (page 8):

*"Assessment of the functional form using restricted cubic splines indicated deviation from linearity compared with a linear model (χ² = 8.18, df = 3; p = 0.042). Segmented logistic regression identified an exploratory breakpoint at a CRII value of 145.7, with modest improvement in model fit (χ² = 4.66, df = 1; p = 0.031)."*

Comment: This is actually fine. But earlier on page 7 (Table 2 footnote):

"No evidence of multicollinearity was observed among included covariates (all variance inflation factors <1.5). Model calibration was adequate based on the Hosmer‑Lemeshow goodness‑of‑fit test (p > 0.05)."
Suggestion: Change the period before "Model calibration" to a semicolon or keep as separate sentence – it is already correct. No change needed.

7. Redundant or repetitive phrasing

Original (Discussion, page 10):

"From a clinical perspective, these findings suggest that CRII may serve as a practical tool for risk stratification in patients undergoing CCRT, enabling identification of individuals at increased risk for post‑treatment TE. The observed gradient in risk across the CRII spectrum, supported by both continuous modeling and exploratory stratification, indicates that CRII captures clinically meaningful variability not reflected by the evaluated conventional mandibular dosimetric parameters."

Problem: "clinical perspective" / "clinically meaningful" – repetitive within two sentences.
Suggestion: Remove "From a clinical perspective" and start directly: "These findings suggest that CRII may serve as a practical tool for risk stratification..."

Author Response

Response to Reviewer 2

Dear Reviewer,

We sincerely thank you for your careful evaluation of our manuscript entitled “The Comprehensive Repair–Inflammation Index (CRII) Predicts Tooth Extraction After Chemoradiotherapy.” We greatly appreciate your constructive comments and valuable suggestions, which have helped improve the scientific balance, methodological clarity, and overall transparency of the manuscript.

Below, we provide a detailed point-by-point response.

Major Comments

Comment 1. Duplicated section (editorial error) “In the PDF, section 2.2 (Baseline Clinical Oral Examination) appears twice verbatim (page 4). This is likely a copy-paste mistake. Please remove the duplicate.”

Response 1: We thank the reviewer for this careful observation. We have re-examined the manuscript and confirm that section 2.2 (“Baseline Clinical Oral Examination”) is not duplicated in the current version of the manuscript. It is possible that this duplication was related to a previous draft or a formatting issue in the PDF generated during submission. Nevertheless, we have carefully reviewed the entire document to ensure that no duplicated sections or formatting inconsistencies remain.

Comment 2. Clarify the CRII calculation and scaling: “You state that CRII was scaled by 10⁻³ (divided by 1000) for presentation. Question: Was the same scaled value used in all regression analyses (logistic, splines, segmented)?”*

Response 2: We thank the reviewer for this important clarification. CRII values were scaled by a factor of 10⁻³ (i.e., divided by 1000) prior to analysis, and this scaled variable was used consistently across all statistical models, including logistic regression, restricted cubic spline modeling, and segmented regression. To avoid ambiguity, we have added an explicit statement in the Methods section clarifying this point.

Comment 3. Missing key covariates from the multivariable model: “Baseline oral health variables may represent important confounders.”

Response 3: We thank the reviewer for this important observation and agree that baseline oral health status may influence the risk of post-treatment tooth extraction.

In our institutional protocol, all patients underwent pre-treatment dental evaluation, and teeth deemed unrestorable were extracted prior to the initiation of chemoradiotherapy. In addition, patients received standardized dental support and follow-up during and after treatment. This approach was intended to reduce variability related to baseline dental status and to minimize the influence of pre-existing non-restorable disease on post-CCRT tooth extraction. However, we acknowledge that more granular oral health variables—including tooth-level restorability, periodontal status, and oral hygiene behaviors—were not systematically captured and therefore could not be included in the multivariable model. Residual confounding from these factors cannot be excluded. This limitation has been clarified in the Discussion section.

Comment 4. Dosimetric parameters justification: “You used V50 and V60 as binary indicators (≥1 cc vs <1 cc). Why 1 cc?”

Response 4: We thank the reviewer for this important question and agree that clarification of this choice is warranted.

Mandibular V50 and V60 were selected based on prior evidence supporting the relevance of dose–volume exposure in the 50–60 Gy range for radiation-related mandibular complications. In the present analysis, these variables were dichotomized at ≥1 cc versus <1 cc as a pragmatic approach to represent the presence versus absence of appreciable high-dose mandibular exposure.

In clinical dosimetric analyses, small absolute volumes (e.g., on the order of 1 cc) are often used pragmatically to distinguish between minimal focal high-dose exposure (“hot spots”) and more appreciable volumes of irradiated tissue; however, this does not represent a validated biological or clinical threshold. In our cohort, the distribution of high-dose mandibular volumes was limited, and this binary specification allowed stable estimation while avoiding overparameterization.

To address this point more transparently, we have added a clarifying statement in the Methods section and incorporated this consideration into the Discussion (limitations), emphasizing that the ≥1 cc cutoff is pragmatic and should not be interpreted as a definitive dosimetric threshold.

Comment 5. Wording of the conclusion about mandibular dose: “The statement ‘Mandibular dose parameters were not significant’ may be misleading.”

Response 5: We thank the reviewer for this important observation. We agree that the original wording may have been overly absolute and could be misinterpreted. Accordingly, we have revised the relevant statements throughout the manuscript to clarify that mandibular dose parameters were not significantly associated with tooth extraction in the present analysis, rather than implying a lack of relevance. In addition, we have refined the wording in the Abstract and Conclusion to avoid overinterpretation and to emphasize that systemic host-related factors may contribute to tooth extraction risk alongside conventional dosimetric parameters, rather than suggesting a hierarchical relationship.

Minor Comments

Comment 6. English language: “The English is generally understandable but contains stylistic issues and overly long sentences.”

Response 6: We thank the reviewer for this valuable and detailed assessment of the language and style of the manuscript. We agree that clarity and readability are essential for effective scientific communication. In response, we have carefully revised the manuscript throughout to improve sentence structure, reduce unnecessary complexity, and correct grammatical and stylistic issues. Specifically, we have:

Shortened overly long and complex sentences,
Improved word choice and idiomatic expression,
Ensured consistency in verb tense and article usage, and
Removed redundant or repetitive phrasing.

We believe that these revisions have substantially improved the clarity, readability, and overall presentation of the manuscript.

Comment 7. Table 2 formatting: “The row for ‘CRII (continuous, per 10-unit increase)’ does not clearly indicate whether the OR is adjusted or unadjusted.”

Response 7: We thank the reviewer for this helpful comment. To improve clarity, we have added an explicit statement in the Table 2 footnote indicating that odds ratios for CRII (per 1-unit and per 10-unit increase) are presented separately for univariate and multivariable models in the corresponding columns. This clarification resolves the potential ambiguity.

Comment 8. Figures and Graphical Abstract: Consider adding a rug plot and ensure the Graphical Abstract is uploaded properly.”

Response 8: We thank the reviewer for this helpful suggestion. The graphical abstract and spline figure have been revised to improve clarity and alignment with the analytical framework. Specifically:

A rug plot has been added to the spline panel to illustrate the distribution of CRII values.
The spline visualization has been refined by removing the point marker and retaining a dashed vertical line to indicate the exploratory breakpoint, avoiding any implication of a discrete threshold.
The graphical abstract has been revised accordingly.

We hope these revisions enhance both the interpretability and the accuracy of the figure.

Ethical and Integrity Comments

Comment 9. AI use declaration: Please clarify which figures involved AI assistance.”*

Response 9: We thank the reviewer for this important comment. We have clarified the use of artificial intelligence tools in the manuscript.

AI assistance was limited to schematic elements of the graphical abstract only. All data-driven figures, including the spline-based analyses and statistical visualizations, were generated directly from the original dataset using standard statistical software and did not involve any AI assistance. No AI tools were used for data analysis, interpretation, or generation of scientific content. This clarification has been incorporated into the Acknowledgments section.

Comment 10. Self-citations: Consider citing studies from independent groups using inflammatory indices.”

Response 10: We thank the reviewer for this valuable suggestion. To provide a broader and more balanced context, we have incorporated additional references from independent groups evaluating systemic inflammation-based indices in oncology.

We would like to note that studies specifically examining inflammation-related predictors of tooth extraction following chemoradiotherapy are currently scarce outside our research group. For this reason, we have retained the relevant prior studies from our group, as they represent the limited available literature directly addressing this clinical endpoint. Importantly, these citations are used to contextualize the present findings rather than to support them exclusively.

To address the reviewer’s concern, we have now supplemented these with independent, widely cited studies demonstrating the prognostic relevance of systemic inflammation indices across cancer populations. This provides a broader framework while preserving the specificity of the existing literature on dental outcomes. These additions have been incorporated into the Discussion section.

Once again, we sincerely thank the reviewer for the highly constructive comments and detailed methodological suggestions. We believe these revisions have substantially improved the clarity, transparency, and scientific quality of the manuscript.

English Editing

The manuscript is written by non-native English speakers (affiliations in Turkey). The scientific content is understandable, but grammatical awkwardness, odd prepositions, excessively long sentences, and occasional unidiomatic expressions reduce clarity and give a less professional impression than the science deserves. A light to moderate professional editing service would bring the language to an acceptable standard for publication in JCM.

Below I provide specific, actionable examples grouped by error type, followed by a short list of corrected sentences.

Comment 1. Overly long and convoluted sentences

Original (Introduction, page 3):

Problem: The second sentence runs for nearly 50 words with nested clauses.
Suggested split:

Response 1. We thank the reviewer for this helpful observation. We agree that the original sentence was overly long and contained nested clauses that reduced readability. The sentence has been revised by splitting it into shorter, clearer statements while preserving the original meaning. Similar revisions have been applied throughout the manuscript to improve clarity and sentence structure.

Comment 2. Unidiomatic or imprecise word choices

Original	Suggested correction	Reason
"nutritional compromise"	"nutritional deficits"	"Compromise" is vague in this context
"not simply a routine procedure"	"is not merely a routine procedure"	More formal academic tone
"tooth extraction is a clinically consequential event"	"tooth extraction represents a clinically consequential event"	More natural verb choice
"we retrospectively analyzed"	Fine, but consider "we conducted a retrospective analysis of"	Slightly more natural
"the relationship was not strictly linear"	Fine	–
"robust biological and clinical evidence"	"substantial biological and clinical evidence"	"Robust" is overused; "substantial" works better

Response 2. We thank the reviewer for these precise and constructive suggestions regarding word choice and phrasing. We agree that refinement of terminology improves clarity and readability. Accordingly, we have revised the manuscript to incorporate the suggested corrections where appropriate, including replacing imprecise or less idiomatic expressions (e.g., “nutritional compromise” with “nutritional deficits” and “not simply a routine procedure” with “is not merely a routine procedure”). We have also refined verb usage to improve fluency (e.g., “represents” in place of “is” where contextually appropriate).

In addition, we have reviewed the manuscript more broadly to ensure consistent and natural academic language, while preserving precision and avoiding unnecessary alterations where the original phrasing was already appropriate.

Comment 3 . Missing or unnecessary articles (a/an/the)

Original (page 4):

"Eligible patients were aged ≥18 years, had histopathologically confirmed squamous cell carcinoma"

Suggestion:

"Eligible patients were aged ≥18 years and had histopathologically confirmed squamous cell carcinoma" (no article needed – correct as is)

But here (page 5):

"Extraction decisions were made by the treating dental team following comprehensive clinical and radiographic evaluation, with consideration of the impaired healing capacity"

Suggestion:

"with consideration of impaired healing capacity" (remove "the" – it is a general concept)

Comment 3. We thank the reviewer for these helpful comments regarding article usage. The suggested corrections have been implemented, including revision of sentence structure for clarity (e.g., addition of “and” in the eligibility criteria) and removal of unnecessary articles (e.g., “the” before “impaired healing capacity”). The manuscript has also been carefully reviewed to ensure consistent and appropriate use of articles throughout.

Comment 4 . Verb tense consistency

Original (Discussion, page 9):

Problem: Mixing past tense (demonstrated, persisted) with present tense (is) is fine, but the sentence remains too long.
Suggested revision:

Response 4. We thank the reviewer for this helpful suggestion. The sentence has been revised to improve clarity and readability by splitting it into shorter statements and reducing complexity, while maintaining a consistent and formal academic tone. Similar refinements have been applied throughout the Discussion section.

Comment 5 . Punctuation errors (commas, semicolons)

Original (page 8):

Comment: This is actually fine. But earlier on page 7 (Table 2 footnote):

Response 5. We thank the reviewer for this comment. The examples provided were already appropriately punctuated; however, we have carefully reviewed the manuscript to ensure consistent and correct use of punctuation throughout.

Comment 6 . Redundant or repetitive phrasing

Original (Discussion, page 10):

Response 6. We thank the reviewer for this helpful observation. The suggested revision has been implemented to eliminate redundant phrasing and improve conciseness. The manuscript has also been reviewed to identify and correct similar instances of repetition.

Once again, we sincerely thank the reviewer for the thoughtful and constructive comments. We believe these revisions have significantly strengthened the scientific rigor, clarity, and balance of the manuscript.

Sincerely,

Efsun Somay, DDS, PhD

On behalf of all authors

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript addresses an important topic and is potentially suitable for publication. However, substantial revisions are required:

Comment 1: the topic is relevant, but the novelty of CRII should be presented more cautiously because it overlaps with prior biomarker studies in the same setting.

Comment 2: the retrospective design limits causal inference. Please specify which oral variables were not available and how this may have influenced extraction decisions.

Comment 3: tooth extraction is a clinically relevant endpoint, but it partly depends on provider judgment. Please clarify whether extraction criteria were applied uniformly.

Coment 4: the spline analysis is a strength. However, the cutoff of 145.7 should be described only as exploratory, not as a clinical threshold.

Comment 5: the lack of association with mandibular dose may reflect limited dosimetric detail. Please temper the conclusion that host factors are more important than dose.

Comment 6: the multivariable model is simple and may still be affected by overfitting. Please justify covariate selection more explicitly.

Comment 7: follow-up duration is not clearly reported. Please add the median follow-up time and the timing of tooth extraction events.

Comment 8: the conclusions are somewhat stronger than the data allow. Please soften claims about prediction until external validation is available.

Comment 9: the study needs clearer limitations, please.

Author Response

Response to Reviewer 3

Dear Reviewer,

We sincerely thank you for your careful evaluation of our manuscript entitled “The Comprehensive Repair–Inflammation Index (CRII) Predicts Tooth Extraction After Chemoradiotherapy.” We greatly appreciate your constructive comments and valuable suggestions, which helped us improve the scientific balance, methodological clarity, and overall transparency of the manuscript.

Below, we provide a detailed point-by-point response.

Comment 1:“The topic is relevant, but the novelty of CRII should be presented more cautiously because it overlaps with prior biomarker studies in the same setting.”

Response 1: We thank the reviewer for this thoughtful and constructive comment and agree that the novelty of CRII should be presented with appropriate caution.

In the revised manuscript, we have refined the framing of CRII to emphasize that it represents an integrative composite index building upon established inflammation- and nutrition-based biomarkers, rather than a wholly distinct construct. To reflect this more balanced positioning, we have moderated the language throughout the manuscript and avoided statements that could be interpreted as overstating novelty.

Specifically, we have incorporated a clarifying statement in the Discussion section, noting that CRII should be interpreted as an extension of existing biomarker frameworks. These revisions ensure that the study’s contribution is presented as the integration of multiple host-related domains and their evaluation using continuous and nonlinear modeling approaches, rather than the introduction of an entirely new biomarker paradigm.

Comment 2: “The retrospective design limits causal inference. Please specify which oral variables were not available and how this may have influenced extraction decisions.”

Response 2: We thank the reviewer for this important and insightful comment and agree that further clarification is warranted.

As detailed in the revised Discussion section, although all patients underwent standardized pre-treatment dental evaluation with extraction of unrestorable teeth and received structured dental support, several granular oral health–related variables were not systematically captured and therefore could not be included in the multivariable model. These include tooth-level restorability, periodontal status, oral hygiene behaviors, and provider-level clinical decision-making factors. These variables may influence both the underlying susceptibility to dental deterioration and the clinical threshold for recommending tooth extraction. Accordingly, because post-CCRT tooth extraction represents a composite endpoint reflecting both biological tissue vulnerability and real-world decision-making, residual confounding from these unmeasured factors cannot be excluded. We have revised the limitations section to explicitly enumerate these variables and clarify their potential influence on extraction decisions.

Comment 3: “Tooth extraction is a clinically relevant endpoint, but it partly depends on provider judgment. Please clarify whether extraction criteria were applied uniformly.”

Response 3: We thank the reviewer for this important comment and agree that the role of provider judgment in tooth extraction decisions warrants clarification.

In our study, extraction decisions were made according to predefined clinical and radiographic criteria, including non-restorable caries, residual roots, persistent or symptomatic apical or periodontal infection, advanced periodontal breakdown, severe mobility, and failure of conservative management. These criteria are described in the Methods section (Section 2.6).

In addition, all patients underwent standardized baseline and follow-up dental evaluations using consistent clinical and radiographic protocols (Sections 2.2 and 2.5), and extraction decisions were made by an experienced dental team within the same institutional framework. This approach was intended to ensure consistency in assessment and management across patients. We have clarified this point in the Methods section to emphasize the uniform application of extraction criteria.

Comment 4: “The spline analysis is a strength. However, the cutoff of 145.7 should be described only as exploratory, not as a clinical threshold.”

Response 4: We thank the reviewer for this valuable comment and fully agree with this recommendation.

In the revised manuscript, the CRII cutoff of 145.7 is consistently described as an exploratory, data-driven threshold, and we have taken care to avoid any interpretation of this value as a definitive or clinically validated cutoff. Specifically:
In the Methods and Results sections, the cutoff is explicitly referred to as an exploratory breakpoint identified using segmented regression.
In the Discussion, we emphasize that this cutoff should be interpreted as descriptive rather than definitive, in line with established statistical guidance regarding the limitations of dichotomization and cohort-specific thresholds.

We also clarify that the cutoff is intended to provide illustrative risk stratification, while the primary inference is based on continuous and nonlinear modeling of CRII, to ensure that the cutoff is presented appropriately and not as a clinically actionable threshold.

These revisions ensure that the cutoff is presented appropriately and not as a clinically actionable threshold.

Comment 5:“The lack of association with mandibular dose may reflect limited dosimetric detail. Please temper the conclusion that host factors are more important than dose.”

Response 5: We thank the reviewer for this important and constructive comment and agree that the interpretation of dosimetric findings should be appropriately cautious.

In the revised manuscript, we have tempered the interpretation to avoid implying a hierarchical relationship between host-related factors and radiation dose. Specifically, we now emphasize that the absence of a significant association with mandibular dose parameters in the present analysis should be interpreted within the constraints of organ-level dosimetric assessment, which may not capture spatial heterogeneity at the level of individual teeth or extraction sites.

We have revised the Discussion and Conclusion sections to clarify that systemic host-related factors, as reflected by CRII, should be interpreted as providing complementary explanatory insight alongside dosimetric parameters, rather than indicating that they are more important than radiation dose. We also explicitly note this limitation in the Discussion.

Comment 6: “The multivariable model is simple and may still be affected by overfitting. Please justify covariate selection more explicitly.”

Response 6: We thank the reviewer for this important comment and agree that explicit justification of covariate selection is essential.

In the present study, covariates were selected a priori based on clinical relevance and prior literature, rather than through data-driven variable selection procedures. To reduce the risk of overfitting, the number of covariates was deliberately constrained relative to the number of observed events, consistent with established recommendations for logistic regression modeling. Specifically, the model included a limited set of clinically relevant variables (age, sex, smoking status, alcohol use, and mandibular dose parameters), while avoiding the inclusion of a larger number of potentially correlated or sparsely distributed predictors. No stepwise or automated selection methods were applied, and model complexity was intentionally limited to ensure parsimony and stability. We have clarified this rationale in the Methods section.

Comment 7: “Follow-up duration is not clearly reported. Please add the median follow-up time and the timing of tooth extraction events.”

Response 7: We thank the reviewer for this important comment. We have revised the Results section to explicitly report the median follow-up duration and to clarify the timing of tooth extraction events after completion of CCRT. Specifically, we now provide the median follow-up time and describe the temporal distribution of extraction events during follow-up.

Comment 8: “The conclusions are somewhat stronger than the data allow. Please soften claims about prediction until external validation is available.”

Response 8: We thank the reviewer for this important comment and agree that the conclusions should be interpreted with appropriate caution. In the revised manuscript, we have softened the language in the Conclusion section to avoid overstating predictive implications and to more clearly reflect the observational nature of the study. Specifically, we have replaced terms such as “predictor” with “associated with” and emphasized that external validation is required before clinical application.

Comment 9:“The study needs clearer limitations, please.”

Response 9: We thank the reviewer for this helpful comment. The manuscript already included a detailed limitations section; however, we agree that greater clarity in presentation would improve readability. In the revised version, we have refined the structure and wording of the limitations section to more explicitly delineate key limitations related to study design, unmeasured confounding (including oral health variables and provider-level decision-making), dosimetric resolution, and generalizability.

Sincerely,

Efsun Somay, DDS, PhD
On behalf of all authors