Basic and Clinical Evidence for Perioperative Immunotherapy in Resectable HNSCC

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I enjoyed reading this manuscript and have the following comments:

1. While the writing is clear, I think you have the opportunity to make the paper more succinct - some paragraphs are repetitive as is the wording.
2. I like the rationale discussions but wonder if an initial overview around antigen load and presentation relative to ICI use (maybe a first paragraph in section3) would be of benefit.  A figure showing differences in antigen load prior to, post op, and post CRT might be useful.
3. Please confirm (or discuss) that you're not advocating for a cessation for node dissection if postop ICI is anticipated.
4. Any thoughts on stratifying ICI applications by HPV status?
5. Some of the clinical trials cited were early on, small numbers - you might make note.
6. I liked the suggestion of precision use based on antigen load (lines 318-325).

Author Response

Reviewer 1. I enjoyed reading this manuscript and have the following comments

Answer: We thank Reviewer 1 for the positive and encouraging evaluation of our manuscript and for the constructive comments. We have carefully revised the manuscript accordingly and believe that these changes have improved the clarity, balance, and overall quality of the review. Our detailed responses are provided below.

Q1. While the writing is clear, I think you have the opportunity to make the paper more succinct - some paragraphs are repetitive as is the wording.

Answer: We agree that several concepts, particularly those related to tumor-draining lymph nodes, antigen presentation, systemic immune surveillance, and the distinctions between neoadjuvant and adjuvant immunotherapy, were described repeatedly throughout the manuscript. We revised multiple sections to reduce redundancy and simplify repetitive wording where appropriate.

Q2. I like the rationale discussions but wonder if an initial overview around antigen load and presentation relative to ICI use (maybe a first paragraph in section3) would be of benefit. A figure showing differences in antigen load prior to, post op, and post CRT might be useful.

Answer: We agree that a clearer conceptual overview of antigen availability would strengthen the manuscript. Accordingly, we added an introductory paragraph and Figure 1 at the beginning of Section 3 to clarify how antigen load may differ between the neoadjuvant, postoperative, and post-chemoradiotherapy settings. (Revised manuscript: Lines 131–138 and Figure 1.)

Q3. Please confirm (or discuss) that you're not advocating for a cessation for node dissection if postop ICI is anticipated.

Answer: We thank the reviewer for this important clarification. We would like to emphasize that we are not advocating omission or cessation of lymph node dissection when postoperative immunotherapy is anticipated. Cervical lymph node dissection remains a critical component of curative treatment for many patients with resectable HNSCC, particularly in the presence of clinically evident nodal disease or established pathological risk factors.

Our intention was instead to discuss the immunological significance of tumor-draining lymph nodes and the possibility that treatment sequencing may influence antitumor immune priming. We have revised the manuscript to clarify that neck dissection of metastatic lymph nodes remains necessary, even in the era of adjuvant ICI therapy. (Revised manuscript: Lines 448–466.)

Q4. Any thoughts on stratifying ICI applications by HPV status?

Answer: We thank the reviewer for this important question. At present, the role of HPV status in stratifying perioperative ICI therapy remains unclear. In both KEYNOTE-689 and NIVOPOSTOP, HPV-positive (p16-positive) HNSCC accounted for only a small proportion of enrolled patients, limiting definitive subgroup interpretation.

In the recurrent or metastatic setting, HPV status has not consistently been associated with major differences in responsiveness to ICIs. Although HPV-positive tumors express viral antigens that are theoretically highly immunogenic, HPV-negative HNSCCs also frequently exhibit substantial neoantigen burden, partly related to TP53 alterations, and are generally considered immunologically active hot tumors. This may partly explain why the anticipated immunological advantage of HPV-positive disease has not translated into clearly superior clinical responses to ICIs. We have added a discussion regarding this issue in the revised manuscript. (Revised manuscript: Lines 414–424.)

Q5. Some of the clinical trials cited were early on, small numbers - you might make note.

Answer: We agree that several of the early perioperative immunotherapy studies cited in this review were exploratory phase I/II trials with relatively small sample sizes and limited follow-up. We have revised the manuscript to more clearly acknowledge these limitations and to avoid overinterpretation of preliminary pathological and survival findings from early-stage studies.

Q6. I liked the suggestion of precision use based on antigen load (lines 318-325).

Answer: We are pleased that the reviewer found the discussion regarding the precision use of immunotherapy based on tumor antigen load to be valuable.

Reviewer 2 Report

Comments and Suggestions for Authors

Major revision recommendation: The manuscript provides a timely overview of perioperative immunotherapy in resectable HNSCC; however, its scientific rigor remains insufficient for publication in its current form. Mechanistic interpretations are frequently descriptive rather than critically synthesized, biomarker discussions lack translational depth, and the comparative analysis between KEYNOTE-689 and NIVOPOSTOP remains underdeveloped. Greater emphasis on unresolved controversies, treatment sequencing, resistance mechanisms, and evidence limitations is required to support the manuscript’s claimed clinical significance.

1. Several recent reviews have already discussed KEYNOTE-689 and neoadjuvant ICIs. Why is another review warranted, and what unmet gap does this manuscript specifically address?
2. The title suggests “Basic and Clinical Evidence,” yet mechanistic immunology receives greater emphasis than translational integration. Can the authors justify this imbalance?
3. The manuscript repeatedly describes perioperative immunotherapy as “practice-changing.” Is this statement premature given immature overall survival data and unresolved patient-selection issues?
4. The discussion on tumor-draining lymph nodes is compelling but largely extrapolated from preclinical studies. Where is the direct human evidence demonstrating causality in HNSCC?
5. The manuscript heavily relies on KEYNOTE-689 and NIVOPOSTOP. Does this overemphasize two studies while underrepresenting negative perioperative trials?
6. Were differences in HPV status, PD-L1 CPS, smoking history, and tumor site adequately considered when interpreting response heterogeneity?
7. The manuscript references improved pathological response with chemoimmunotherapy combinations. How do toxicity burdens alter the risk-benefit ratio?
8. The manuscript acknowledges absence of validated biomarkers but offers limited discussion on actionable biomarker development strategies. How should future trials address this gap?
9. PD-L1 expression is inconsistently predictive across trials. Should reliance on CPS continue in perioperative settings?
10. The manuscript proposes de-escalated surgery following neoadjuvant immunotherapy. Is this suggestion premature and potentially unsafe without randomized evidence?

 

Author Response

Reviewer 2. The manuscript provides a timely overview of perioperative immunotherapy in resectable HNSCC; however, its scientific rigor remains insufficient for publication in its current form. Mechanistic interpretations are frequently rather than critically synthesized, biomarker discussions lack translational depth, and the comparative analysis between KEYNOTE-689 and NIVOPOSTOP remains underdeveloped. Greater emphasis on unresolved controversies, treatment sequencing, resistance mechanisms, and evidence limitations is required to support the manuscript’s claimed clinical significance.

Answer: We thank the reviewer for the thoughtful and constructive critique. We agree that the original manuscript placed substantial emphasis on mechanistic rationale and that several clinically important unresolved issues required more balanced and critical discussion. In response, we substantially revised the manuscript to improve its scientific rigor and translational perspective.

Specifically, we expanded the discussion regarding the limitations of current biomarker approaches, including HPV status, the impact of spatial and temporal tumor heterogeneity, and limitations of PD-L1 assessment. We also strengthened the comparative discussion between KEYNOTE-689 and NIVOPOSTOP by emphasizing that these trials represent biologically and clinically distinct perioperative strategies rather than directly interchangeable treatment models. We also expanded discussion regarding the negative phase III experiences combining definitive chemoradiotherapy with ICIs and discussed possible biological explanations. Finally, throughout the revised manuscript, we moderated several statements that may have appeared overly definitive and clarified when evidence was primarily preclinical or exploratory.

Q1. Several recent reviews have already discussed KEYNOTE-689 and neoadjuvant ICIs. Why is another review warranted, and what unmet gap does this manuscript specifically address?

Answer: We agree that several recent reviews have discussed KEYNOTE-689, including those cited in our manuscript (PMID: 40907366). We comprehensively reviewed nearly 20 previously published review articles on perioperative immunotherapy in HNSCC. However, most primarily focused on summarizing clinical trial outcomes, whereas few provided a clear discussion of dynamic changes in tumor antigen load during treatment or the immunological rationale underlying neoadjuvant approaches. Furthermore, limited attention has been paid to therapeutic strategies for recurrent disease following perioperative ICI exposure. Given these unresolved and clinically relevant gaps in the current literature, we considered it important to provide a comprehensive review integrating both clinical and immunological perspectives on perioperative immunotherapy in HNSCC. We believe that this integrated basic, translational, and clinical perspective provides added value beyond existing reviews focused primarily on neoadjuvant/adjuvant ICI trial outcomes.

Q2. The title suggests “Basic and Clinical Evidence,” yet mechanistic immunology receives greater emphasis than translational integration. Can the authors justify this imbalance?

Answer: We thank the reviewer for this important comment. We agree that the original manuscript placed relatively strong emphasis on mechanistic immunology. Our intention was to provide a biological framework for understanding why the timing and sequencing of perioperative immunotherapy may matter in resectable HNSCC.

In response, we revised the manuscript to better balance basic and clinical evidence. Specifically, we strengthened the translational discussion by linking the mechanistic rationale to clinically relevant issues, including interpretation of KEYNOTE-689 and NIVOPOSTOP, biomarker development, tumor heterogeneity, surgical decision-making after neoadjuvant therapy, and treatment sequencing after recurrence during or after perioperative ICI exposure. We believe that these revisions better justify the title “Basic and Clinical Evidence.”

Q3. The manuscript repeatedly describes perioperative immunotherapy as “practice-changing.” Is this statement premature given immature overall survival data and unresolved patient-selection issues?

Answer: We agree that repeated use of the term “practice-changing” may have conveyed an overly definitive interpretation, particularly because overall survival data remain immature and optimal patient selection has not yet been fully established. We replaced “practice-changing” with “practice-relevant.”

Q4. The discussion on tumor-draining lymph nodes is compelling but largely extrapolated from preclinical studies. Where is the direct human evidence demonstrating causality in HNSCC?

Answer: We agree that direct causal evidence demonstrating the role of tumor-draining lymph nodes in human HNSCC remains limited. In clinical practice, unlike in mouse models, lymph nodes suspected of harboring metastasis cannot ethically or oncologically be preserved; therefore proving causality in humans is inherently more difficult than in experimental models.

Nevertheless, several lines of human evidence are beginning to support the clinical relevance of tumor-draining lymph nodes in HNSCC. For example, activation of tumor-reactive T cells within draining lymph nodes has been associated with longer overall survival, suggesting that immune activation in the nodal compartment may contribute to antitumor control (PMID: 33462898). In addition, prior neck dissection has been reported to attenuate the efficacy of immune checkpoint inhibitors at recurrence or metastasis, indirectly supporting the importance of preserved cervical lymphatic and nodal immune function (PMID: 41617453). As perioperative immunotherapy becomes more widely implemented, analyses of cervical lymph nodes obtained after neoadjuvant ICI therapy may provide an important opportunity to clarify the immunological and clinical significance of tumor-draining lymph nodes in human HNSCC. We have incorporated this perspective into the revised manuscript. (Revised manuscript: Lines 215–225.)

Q5. The manuscript heavily relies on KEYNOTE-689 and NIVOPOSTOP. Does this overemphasize two studies while underrepresenting negative perioperative trials?

Answer: We agree that excessive emphasis on KEYNOTE-689 and NIVOPOSTOP could give the impression that the current evidence for perioperative immunotherapy is uniformly positive. In the revised manuscript, we incorporated a more detailed discussion of the JAVELIN Head and Neck 100 trial and KEYNOTE-412, both of which failed to demonstrate a clear benefit from adding ICIs to definitive chemoradiotherapy, as well as the IMvoke010 study, in which maintenance atezolizumab after definitive multimodal treatment did not improve clinical outcomes. We also discussed how these negative results highlight the importance of treatment timing, patient selection, and the potential differences between definitive CRT-based strategies and perioperative approaches. (Revised manuscript: Lines 83–86, 343–353.)

Q6. Were differences in HPV status, PD-L1 CPS, smoking history, and tumor site adequately considered when interpreting response heterogeneity?

Answer: We agree that response heterogeneity to perioperative immune checkpoint inhibitors may be influenced by multiple clinicopathological factors, including HPV status, PD-L1 CPS, smoking history, and primary tumor site, and that these factors were not sufficiently discussed in the original manuscript. In the revised manuscript, we added a more detailed discussion of these variables.  (Revised manuscript: Lines 387–424.)

Q7. The manuscript references improved pathological response with chemoimmunotherapy combinations. How do toxicity burdens alter the risk-benefit ratio?

Answer: Although the addition of ICIs to chemotherapy may enhance tumor antigen release and contribute to improved pathological responses, it may also increase the risk of adverse events such as pneumonitis and renal dysfunction, which can lead to treatment discontinuation. Thyroid dysfunction was among the most common immune-related adverse events associated with ICIs and was generally manageable. In contrast, renal dysfunction may be induced by both ICIs and cisplatin, and this toxicity is clinically important because it may compromise the feasibility of subsequent systemic chemotherapy. Although the overall safety profiles of the NIVOPOSTOP and KEYNOTE-689 trials were considered acceptable, we agree that these safety considerations warrant further emphasis and have incorporated this discussion into the revised manuscript. (Revised manuscript: Lines 332–341.)

Q8. The manuscript acknowledges absence of validated biomarkers but offers limited discussion on actionable biomarker development strategies. How should future trials address this gap?

Answer: In the revised manuscript, we emphasized that future perioperative immunotherapy trials should prospectively integrate pre-treatment biomarker analyses. Potential strategies include standardized PD-L1 CPS assessment, as well as evaluation of tumor mutational burden, immune gene signatures, and interferon-related signatures. Although single-cell RNA sequencing and spatial transcriptomic analyses may provide highly detailed characterization of the tumor immune microenvironment, their widespread implementation in routine clinical practice remains challenging because of financial and technical limitations. To establish clinically applicable biomarkers that can meaningfully guide treatment selection, biomarker strategies should ideally be based on pretreatment specimens and utilize approaches that are feasible in terms of both cost and technical accessibility. From this perspective, future studies should first focus on treatment stratification using biomarkers that are already clinically applicable, including PD-L1 immunohistochemistry, histopathological evaluation of tumor-infiltrating lymphocytes and tertiary lymphoid structures, and peripheral blood–based immune markers such as the neutrophil-to-lymphocyte ratio. We believe these additions provide a more practical framework for future biomarker-driven perioperative immunotherapy trials in HNSCC. (Revised manuscript: Lines 483–502.)

Q9. PD-L1 expression is inconsistently predictive across trials. Should reliance on CPS continue in perioperative settings?

Answer: We agree that reliance on PD-L1 CPS alone in the perioperative setting remains problematic. PD-L1 expression in HNSCC can show marked intratumoral heterogeneity, and CPS assessment based on small pretreatment biopsy specimens may not fully reflect the immune status of the entire tumor.

In KEYNOTE-689, patients with higher CPS appeared to derive greater pathological benefit. However, the number of patients with absent or very low CPS expression was limited, and the trial was not designed to establish CPS-based treatment selection. Therefore, although CPS remains the most clinically accessible and widely used biomarker for anti-PD-1 therapy in HNSCC, its utility as a standalone biomarker for perioperative ICI selection remains insufficient. In the revised manuscript, we clarified that CPS should be interpreted as one component of a broader biomarker framework rather than as an exclusive determinant of treatment indication. (Revised manuscript: Lines 385–403 and 486–489.)

Q10. The manuscript proposes de-escalated surgery following neoadjuvant immunotherapy. Is this suggestion premature and potentially unsafe without randomized evidence?

Answer: We agree that de-escalated surgery after neoadjuvant immunotherapy remains premature and potentially unsafe in the absence of prospective randomized evidence. We therefore revised the manuscript to clarify that surgical resection should generally be performed according to the pretreatment tumor extent as defined by imaging studies. De-escalated surgical approaches should currently be considered only within carefully designed prospective clinical trials incorporating standardized imaging reassessment, pathological validation, functional outcome measures, locoregional control, and long-term survival endpoints. We believe these revisions provide a more cautious and clinically appropriate interpretation of response-adapted surgical strategies following neoadjuvant immunotherapy. (Revised manuscript: Lines 444–466.)

Reviewer 3 Report

Comments and Suggestions for Authors

This narrative review provides a valuable overview of perioperative immunotherapy in resectable HNSCC and appropriately highlights the need for predictive biomarkers to guide patient selection. However, the review does not adequately address how tumor heterogeneity—a well-recognized barrier to reliable biomarker development in HNSCC [4†L21-L26]—complicates this goal. HNSCC exhibits marked inter‑ and intra‑tumoral heterogeneity in cellular composition, immune contexture, and treatment response [10†L8-L11][11†L10-L12], but the review does not discuss how these complexities impact the identification of clinically useful predictive biomarkers.

The authors should expand the Discussion to explicitly address this limitation and, referencing PMID: 40092490 and PMID: 38896289, propose integration of single‑cell.

Any papers recommended in the report are for reference only. They are not mandatory. You may cite and reference other papers related to this topic. 

Author Response

Reviewer 3. This narrative review provides a valuable overview of perioperative immunotherapy in resectable HNSCC and appropriately highlights the need for predictive biomarkers to guide patient selection. However, the review does not adequately address how tumor heterogeneity—a well-recognized barrier to reliable biomarker development in HNSCC [4†L21-L26]—complicates this goal. HNSCC exhibits marked inter‑ and intra‑tumoral heterogeneity in cellular composition, immune contexture, and treatment response [10†L8-L11][11†L10-L12], but the review does not discuss how these complexities impact the identification of clinically useful predictive biomarkers. The authors should expand the Discussion to explicitly address this limitation and, referencing PMID: 40092490 and PMID: 38896289, propose integration of single‑cell.

Answer: We thank the reviewer for this insightful and important comment. We agree that tumor heterogeneity represents a major challenge for biomarker development in HNSCC. HNSCC exhibits substantial inter- and intra-tumoral heterogeneity in cellular composition, immune microenvironment, genomic alterations, and treatment responsiveness, all of which may complicate identification of robust predictive biomarkers for perioperative immunotherapy. We have also previously reported that the expression of EGFR and PD-L1 on tumor cells can vary even within the same patient, depending on disease recurrence and the antibody clone used for evaluation (PMID: 41905239 and 40156197).

In response to the reviewer’s suggestion, we expanded the Discussion section to address how the spatial and temporal heterogeneity of both tumor cells and the immune microenvironment may limit the reliability of conventional bulk biomarker approaches, including PD-L1 expression and immune gene signatures. We additionally discussed that biomarker assessment based on a single biopsy specimen may not fully capture the dynamic and heterogeneous immune landscape of HNSCC. Furthermore, we incorporated discussion of recent advances in single-cell and spatial transcriptomic analyses, including the references suggested by the reviewer (PMID: 40092490 and PMID: 38896289). (Revised manuscript: Lines 490–502.)

Reviewer 4 Report

Comments and Suggestions for Authors

Previously, immune checkpoint inhibitor (ICI) therapy for head and neck cancer was primarily indicated for recurrent or metastatic cases that were unresectable. However, in February 2026, the anti-PD-1 antibody pembrolizumab officially received manufacturing and marketing approval in Japan for perioperative (neoadjuvant and adjuvant) therapy in patients with locally advanced head and neck squamous cell carcinoma. Therefore, this review is highly relevant and timely. The authors explain the biological rationale and clinical evidence for perioperative ICI treatment in head and neck squamous cell carcinoma (HNSCC), providing readers with insightful information to better understand the development of novel therapeutics for HNSCC. Figure 1 is particularly impressive, as it clearly illustrates the distinctive mechanisms of action between neoadjuvant and adjuvant immunotherapy. 

Comment; Based on the authors' rationale for HNSCC immunotherapy, it appears necessary to minimize lymph node dissection during surgery. Could the authors consider discussing potential modifications or improvements to lymph node dissection to better align with the upcoming era of immunotherapy? 

Author Response

Reviewer 4. Previously, immune checkpoint inhibitor (ICI) therapy for head and neck cancer was primarily indicated for recurrent or metastatic cases that were unresectable. However, in February 2026, the anti-PD-1 antibody pembrolizumab officially received manufacturing and marketing approval in Japan for perioperative (neoadjuvant and adjuvant) therapy in patients with locally advanced head and neck squamous cell carcinoma. Therefore, this review is highly relevant and timely. The authors explain the biological rationale and clinical evidence for perioperative ICI treatment in head and neck squamous cell carcinoma (HNSCC), providing readers with insightful information to better understand the development of novel therapeutics for HNSCC. Figure 1 is particularly impressive, as it clearly illustrates the distinctive mechanisms of action between neoadjuvant and adjuvant immunotherapy.

Answer: We thank the reviewer for the positive and encouraging comments regarding our manuscript. We also appreciate the reviewer’s favorable evaluation of Figure 1 and are pleased that the figure effectively conveyed the conceptual differences between neoadjuvant and adjuvant immunotherapy.

Q1. Based on the authors' rationale for HNSCC immunotherapy, it appears necessary to minimize lymph node dissection during surgery. Could the authors consider discussing potential modifications or improvements to lymph node dissection to better align with the upcoming era of immunotherapy?

Answer: We thank the reviewer for this important clarification. Although we consider cervical lymph nodes to play an indispensable role in perioperative antitumor immunity, de-escalation of neck dissection cannot currently be recommended because of insufficient clinical evidence. We speculate that tumor antigen–mediated T-cell priming within cervical lymph nodes is largely completed within approximately 1–2 weeks after ICI administration, after which memory T cells may become systemically distributed. From this perspective, standard neck dissection following neoadjuvant ICI treatment may be acceptable when sufficient time has elapsed after ICI administration.

From an immunological standpoint, super-selective neck dissection targeting only radiologically or pathologically confirmed metastatic lymph nodes may represent a preferable strategy for preserving regional lymphatic immune function. However, further advances in imaging modalities to detect metastatic lymph nodes are required to safely implement such approaches in clinical practice. At present, treatment sequencing strategies designed to preserve and exploit lymph node immune function, such as neoadjuvant approaches, are likely to represent the most clinically applicable strategy. We have revised the manuscript accordingly to clarify these points. (Revised manuscript: Lines 444–466.)

Reviewer 5 Report

Comments and Suggestions for Authors

This review discusses the rapidly evolving role of perioperative immunotherapy in resectable HNSCC and appropriately highlights the impact of KEYNOTE-689 and NIVOPOSTOP. The topic is timely and clinically relevant. However, several important aspects require deeper critical analysis and more balanced interpretation.

COMMENTS:

1. The manuscript strongly promotes the biological rationale for neoadjuvant immunotherapy, particularly the importance of preserving tumor-draining lymph nodes and intact antigen exposure. However, the review occasionally presents these concepts as clinically established rather than translational hypotheses. While the mechanistic rationale is compelling, current clinical evidence does not yet definitively demonstrate superiority of neoadjuvant over postoperative-only strategies. Several statements in Sections 3.3 and 3.4 should therefore be moderated.

2. The discussion of KEYNOTE-689 is insufficiently critical. The manuscript should better acknowledge that the study design does not allow distinction between the contribution of neoadjuvant pembrolizumab and prolonged adjuvant pembrolizumab exposure. Moreover, all patients received postoperative radiotherapy, raising concerns regarding overtreatment in lower-risk patients. These limitations deserve more explicit discussion.

3. The review inadequately discusses the negative phase III experiences combining ICIs with definitive chemoradiotherapy, particularly KEYNOTE-412 and JAVELIN Head and Neck 100. Since the manuscript emphasizes treatment sequencing as biologically critical, a deeper analysis of why concurrent CRT-immunotherapy failed whereas perioperative approaches appear promising would considerably strengthen the review.

4. The pathological response discussion should be interpreted more cautiously. Major pathological response and pathological complete response are increasingly used translational endpoints, but they are not yet fully validated surrogate markers for long-term survival in HNSCC. The manuscript currently gives the impression that these endpoints are already clinically established.

5. The section regarding de-escalated surgery after neoadjuvant chemoimmunotherapy is overly speculative. The cited evidence remains retrospective and highly selected, and there is currently insufficient evidence to support modification of surgical extent based on immunotherapy response outside clinical trials. The authors should clearly emphasize this limitation.

6. The manuscript would benefit from a dedicated summary table comparing the major perioperative immunotherapy trials, including design, treatment schedule, pathological response, EFS/DFS outcomes, and toxicity. Currently, the large amount of information is difficult to synthesize for the reader.

7. Several concepts are repeated excessively throughout the manuscript, such as preservation of tumor-draining lymph nodes, systemic immune surveillance, and distinctions between neoadjuvant and adjuvant strategies.

8. Although the review extensively discusses the biological rationale and oncologic outcomes of perioperative immunotherapy, it lacks a sufficiently practical surgical perspective. In particular, the manuscript does not adequately address how perioperative ICIs may concretely influence surgical decision-making in daily clinical practice.

Author Response

Reviewer 5. This review discusses the rapidly evolving role of perioperative immunotherapy in resectable HNSCC and appropriately highlights the impact of KEYNOTE-689 and NIVOPOSTOP. The topic is timely and clinically relevant. However, several important aspects require deeper critical analysis and more balanced interpretation.

Answer: We thank the reviewer for the positive evaluation of the timeliness and clinical relevance of our review, as well as for the constructive criticism. We agree that several aspects of the original manuscript required deeper critical analysis and a more balanced interpretation. In response, we substantially revised the manuscript to better acknowledge current evidence limitations and unresolved clinical questions.

Q1. The manuscript strongly promotes the biological rationale for neoadjuvant immunotherapy, particularly the importance of preserving tumor-draining lymph nodes and intact antigen exposure. However, the review occasionally presents these concepts as clinically established rather than translational hypotheses. While the mechanistic rationale is compelling, current clinical evidence does not yet definitively demonstrate superiority of neoadjuvant over postoperative-only strategies. Several statements in Sections 3.3 and 3.4 should therefore be moderated.

Answer: We agree that the biological rationale for neoadjuvant immunotherapy, including preserved tumor antigen exposure and tumor-draining lymph-node function, should be presented as a translational hypothesis rather than as clinically established evidence. We also agree that current clinical data do not definitively demonstrate the superiority of neoadjuvant immunotherapy over postoperative-only strategies. We revised Sections 3.3 and 3.4 to moderate several statements that may have appeared overly definitive.

Q2. The discussion of KEYNOTE-689 is insufficiently critical. The manuscript should better acknowledge that the study design does not allow distinction between the contribution of neoadjuvant pembrolizumab and prolonged adjuvant pembrolizumab exposure. Moreover, all patients received postoperative radiotherapy, raising concerns regarding overtreatment in lower-risk patients. These limitations deserve more explicit discussion.

Answer: We agree that the use of postoperative radiotherapy in all patients raises an important concern regarding potential overtreatment, particularly among patients who may have had favorable pathological features after surgery and could potentially have been managed with surgery alone in real-world practice. This issue further emphasizes the need for better risk stratification by imaging and predictive biomarkers to identify patients who truly require perioperative immunotherapy and postoperative treatment intensification. We expanded the discussion of potential overtreatment and the need for future trials evaluating response-adapted treatment strategies, biomarker-guided patient selection, and possible de-escalation of adjuvant therapy in appropriately selected low-risk patients.

We also agree that the study design of KEYNOTE-689 does not allow the relative contributions of neoadjuvant pembrolizumab, adjuvant pembrolizumab, and their combined perioperative administration to be distinguished. Therefore, the clinical benefit observed in this trial should not be interpreted as definitive evidence that the neoadjuvant component alone is responsible for the improved outcomes. We revised the manuscript to more clearly acknowledge these limitations. (Revised manuscript: Lines 328–331, 378–384 and 325–441.)

Q3. The review inadequately discusses the negative phase III experiences combining ICIs with definitive chemoradiotherapy, particularly KEYNOTE-412 and JAVELIN Head and Neck 100. Since the manuscript emphasizes treatment sequencing as biologically critical, a deeper analysis of why concurrent CRT-immunotherapy failed whereas perioperative approaches appear promising would considerably strengthen the review.

Answer: We expanded the discussion to address potential biological and clinical differences between definitive CRT-ICI strategies and perioperative approaches. Although KEYNOTE-412, JAVELIN Head and Neck 100, and NIVOPOSTOP all incorporated ICIs in combination with chemoradiotherapy, an important distinction is the presence or absence of macroscopic tumor burden at the time of treatment. In definitive CRT trials, ICIs were administered in the presence of substantial unresected tumor burden, whereas in NIVOPOSTOP, nivolumab was added after surgical removal of the primary tumor and involved lymph nodes, when the treatment target was primarily residual microscopic disease. Consistent with this concept, clinical studies in melanoma have suggested that a high tumor burden may reduce the efficacy of immunotherapy (PMID: 29685882). These considerations suggest that simply adding ICIs concurrently to definitive CRT may not be equivalent to integrating ICIs into a perioperative strategy after tumor debulking. (Revised manuscript: Lines 345–352 and 360–376.)

Q4. The pathological response discussion should be interpreted more cautiously. Major pathological response and pathological complete response are increasingly used translational endpoints, but they are not yet fully validated surrogate markers for long-term survival in HNSCC. The manuscript currently gives the impression that these endpoints are already clinically established.

Answer: We agree that pathological response endpoints, including major pathological response and pathological complete response, have not yet been fully validated as surrogate markers for long-term survival outcomes in HNSCC. We revised the manuscript to clarify that major pathological response and pathological complete response should currently be regarded as translational and exploratory endpoints that may correlate with favorable outcomes, but require further validation in larger prospective studies with long-term follow-up. (Revised manuscript: Lines 251–259.)

Q5. The section regarding de-escalated surgery after neoadjuvant chemoimmunotherapy is overly speculative. The cited evidence remains retrospective and highly selected, and there is currently insufficient evidence to support modification of surgical extent based on immunotherapy response outside clinical trials. The authors should clearly emphasize this limitation.

Answer: We agree that de-escalated surgery after neoadjuvant chemoimmunotherapy remains highly speculative and should not be presented as a clinically established strategy.

In response, we revised the manuscript to more clearly emphasize this limitation. We clarified that, at present, surgical resection should generally be planned according to the pretreatment tumor extent and standard oncologic principles, regardless of radiological or clinical response after neoadjuvant therapy. (Revised manuscript: Lines 444–466.)

Q6. The manuscript would benefit from a dedicated summary table comparing the major perioperative immunotherapy trials, including design, treatment schedule, pathological response, EFS/DFS outcomes, and toxicity. Currently, the large amount of information is difficult to synthesize for the reader.

Answer: We have added a dedicated summary table comparing major perioperative immunotherapy trials, including study design, treatment schedule, pathological response, survival outcomes, and toxicity profiles (Table 1). We believe that this new table improves readability and allows readers to more easily compare the available clinical evidence across different perioperative strategies.

Q7. Several concepts are repeated excessively throughout the manuscript, such as preservation of tumor-draining lymph nodes, systemic immune surveillance, and distinctions between neoadjuvant and adjuvant strategies.

Answer: We agree that several concepts, particularly those related to preservation of tumor-draining lymph nodes, systemic immune surveillance, and the biological distinctions between neoadjuvant and adjuvant immunotherapy, were described repeatedly throughout the manuscript. To improve conciseness and readability, we revised Sections 3.1–3.4 and Conclusion to streamline overlapping explanations, reduce redundancy, and simplify repetitive wording while preserving the overall conceptual framework of the review.

Q8. Although the review extensively discusses the biological rationale and oncologic outcomes of perioperative immunotherapy, it lacks a sufficiently practical surgical perspective. In particular, the manuscript does not adequately address how perioperative ICIs may concretely influence surgical decision-making in daily clinical practice.

Answer: We thank the reviewer for this important comment, and revised the manuscript to add a more practical surgical perspective. Specifically, we discussed that, at present, the extent of primary tumor resection and neck dissection should continue to be determined according to pretreatment tumor extent, rather than radiological or clinical response to neoadjuvant ICIs. We also added that, in patients with rapidly progressing tumors, clinicians may be inclined to choose cytotoxic chemotherapy rather than ICI monotherapy as neoadjuvant treatment because prompt tumor reduction may be required. However, there is currently insufficient evidence to determine the optimal neoadjuvant strategy for such patients. Therefore, future studies should clarify whether clinical features such as tumor growth kinetics, symptom progression, airway or swallowing compromise, and risk of treatment delay can guide selection between neoadjuvant ICI monotherapy and conventional cytotoxic chemotherapy.

We further emphasized that patient selection remains a major unresolved issue in daily practice. Because the eligibility criteria of KEYNOTE-689 were broad, it remains unclear whether all patients with operable stage III/IV HNSCC should routinely receive neoadjuvant immunotherapy. Similarly, although patients with hypopharyngeal cancer appeared to have a poorer prognosis in KEYNOTE-689, such a tendency was not clearly observed in NIVOPOSTOP, and current evidence does not allow reliable stratification of patients according to tumor site, HPV status, or PD-L1 CPS alone. In practical terms, most patients with operable locally advanced HNSCC may potentially meet the indication for neoadjuvant immunotherapy, but robust biomarkers to identify those who truly require treatment intensification have not yet been established.　Likewise, it remains unclear whether nivolumab should be added to postoperative chemoradiotherapy in all patients with high-risk pathological features, because biomarkers for selecting patients most likely to benefit from postoperative ICI intensification are still lacking. In the revised manuscript, we therefore highlighted the urgent need for clinically applicable biomarkers that can guide appropriate patient selection, avoid overtreatment, and help determine whether neoadjuvant immunotherapy, postoperative ICI intensification, or standard surgery-based treatment is most appropriate for individual patients. We believe these revisions provide a more practical surgical perspective and address the reviewer’s concern. (Revised manuscript: Conclusion.)

Round 2

Reviewer 5 Report

Comments and Suggestions for Authors

No other comments. The manuscript is now suitable for publication.