Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This review provides a timely and well-organized overview of immune cell engagers and tumor-restricted “masking” strategies in solid tumors, with particular emphasis on head and neck squamous cell carcinoma. The manuscript covers a broad range of engager platforms, target antigens, conditional activation strategies, biomarkers, and ongoing clinical development. The schematic figure is particularly helpful in illustrating the different immune-effector engager concepts, and the summary tables provide useful information for readers. The manuscript is generally balanced in discussing both the potential and the current limitations of these approaches. I have several relatively minor comments that may further improve the clarity and precision of the review.

1. The manuscript frequently extrapolates findings from other solid tumors to HNSCC. This is reasonable given the relatively limited HNSCC-specific clinical evidence, but the distinction between direct evidence in HNSCC and evidence derived from other tumor types should be made more consistently. In particular, preclinical or early clinical findings from non-HNSCC models should be clearly identified as such when their relevance to HNSCC is discussed.
2. Table 2 would benefit from clarification regarding which efficacy and safety results are specifically derived from HNSCC patients. Some reported values, such as ORR, DCR, PR, SD, or DLT rates, appear to represent the overall solid-tumor study population rather than the HNSCC subgroup. Please specify whether each result is HNSCC-specific or derived from the overall study population and, where available, provide the number of HNSCC patients enrolled.
3. Similarly, the developmental-stage terminology used in Table 1, such as “Clinical” and “Early clinical,” is somewhat broad. It would be useful to clarify whether these stages refer to development in any malignancy, in solid tumors specifically, or in HNSCC. This would help readers better understand the actual maturity of each platform in the context of the present review.
4. The toxicity discussion could be slightly more nuanced across different engager classes. CRS and ICANS are particularly relevant to CD3-directed T-cell engagers, but they are not necessarily equally established or dominant toxicities for NK-cell, γδ T-cell, or myeloid-cell engager platforms. A brief distinction among these classes would improve the accuracy of this section.
5. The manuscript would benefit from additional language editing. Several expressions are grammatically awkward or non-standard (for example, “Below it is provided a list,” and “in head-neck district”). In addition, terminology and notation should be standardized throughout; for example, “INFγ” should be corrected to “IFN-γ”.

Author Response

Comment 1: The manuscript frequently extrapolates findings from other solid tumors to HNSCC. This is reasonable given the relatively limited HNSCC-specific clinical evidence, but the distinction between direct evidence in HNSCC and evidence derived from other tumor types should be made more consistently. In particular, preclinical or early clinical findings from non-HNSCC models should be clearly identified as such when their relevance to HNSCC is discussed.

Response:
We thank the Reviewer for this important observation. We agree that the distinction between direct HNSCC evidence and evidence extrapolated from other tumor types should be clearly stated, particularly where clinical evidence remains limited. We therefore re-examined the relevant sections of the manuscript. Most preclinical sections already specify the experimental tumor context and use appropriately cautious language when translating these findings to HNSCC. However, we identified the discussion of NKCE as a passage in which the distinction could be made clearer. We therefore revised the paragraph describing AFM24 to explicitly state that the available early clinical experience derives from patients with advanced EGFR-positive solid tumors rather than from an HNSCC-specific clinical population, and that the clinical efficacy of NKCEs, specifically in HNSCC, remains to be established. This modification more clearly separates demonstrated clinical evidence from the biological rationale for applying NK-cell redirection to HNSCC.

Location: Section 5, Pages 15-16, Lines 524–540.

Comment 2: Table 2 would benefit from clarification regarding which efficacy and safety results are specifically derived from HNSCC patients. Some reported values, such as ORR, DCR, PR, SD, or DLT rates, appear to represent the overall solid-tumor study population rather than the HNSCC subgroup. Please specify whether each result is HNSCC-specific or derived from the overall study population and, where available, provide the number of HNSCC patients enrolled.

Response:
We fully agree with the Reviewer and performed a systematic re-evaluation of the studies included in Table 2, with particular attention to the population from which the reported efficacy and safety outcomes were derived. The table has therefore been revised to distinguish HNSCC-specific results from outcomes derived from the overall solid-tumor study population. Where HNSCC-specific data and patient numbers were publicly available, these have been explicitly reported. Conversely, when results were derived from the overall study population or when HNSCC-specific outcomes had not been separately reported, this is now clearly indicated in the corresponding entry. This re-evaluation also prompted us to reconsider the eligibility of individual agents for inclusion in Table 2. In particular, petosemtamab (MCLA-158; NCT03526835) was removed. Although this EGFR×LGR5 bispecific antibody has demonstrated clinically relevant activity in HNSCC, it neither directly redirects an immune-effector cell nor represents a masked/conditionally activated therapeutic platform and therefore does not meet the scope criteria used for this table. For the remaining programs, the table now more clearly identifies whether HNSCC represents a dedicated expansion cohort, an eligible tumor type without separately reported results, or a population for which HNSCC-specific efficacy data are available. 

Location: Table 2, Pages 23–24.

Comment 3: Similarly, the developmental-stage terminology used in Table 1, such as “Clinical” and “Early clinical,” is somewhat broad. It would be useful to clarify whether these stages refer to development in any malignancy, in solid tumors specifically, or in HNSCC. This would help readers better understand the actual maturity of each platform in the context of the present review.

Response:
We thank the Reviewer for pointing out this potential ambiguity. Rather than modifying the individual developmental-stage categories, we added an asterisk to the “Developmental stage” heading in Table 1 and included a corresponding explanatory note below the table. The note now clarifies that the developmental stage refers to the most advanced stage reached by each platform across oncology and does not necessarily indicate clinical development specifically in HNSCC.

Location: Table 1, Pages 13-16, Lines 488–490.

Comment 4: The toxicity discussion could be slightly more nuanced across different engager classes. CRS and ICANS are particularly relevant to CD3-directed T-cell engagers, but they are not necessarily equally established or dominant toxicities for NK-cell, γδ T-cell, or myeloid-cell engager platforms. A brief distinction among these classes would improve the accuracy of this section.

Response:
We agree with the Reviewer. The toxicity section has been revised to more clearly distinguish the safety profiles associated with different ICE classes. CRS is now specifically described as a major class-associated toxicity of CD3 TCE, while ICANS is discussed in the context of T-cell-redirecting platforms rather than being generalized to all ICEs. We further clarify that the clinical safety profiles of NKCE, γδTCE, and myeloid-cell engager platforms are either different or substantially less well characterized and should not be assumed to reproduce the toxicity spectrum of conventional CD3-directed TCEs. Early clinical experience with NKCE is discussed separately, whereas the limited clinical maturity of γδTCE and myeloid-cell engager platforms is explicitly acknowledged. Finally, we refined the HNSCC-specific discussion of local inflammatory toxicity, since edema, tumor necrosis, or bleeding may have particularly important clinical consequences when tumors are located close to the upper airway and swallowing structures.

Location: Section 3.4, Page 8, Lines 278–307.

Comment 5: The manuscript would benefit from additional language editing. Several expressions are grammatically awkward or non-standard (for example, “Below it is provided a list,” and “in head-neck district”). In addition, terminology and notation should be standardized throughout; for example, “INFγ” should be corrected to “IFN-γ”

Response:
We thank the Reviewer for identifying these language and terminology issues. The manuscript underwent an additional language-editing pass to improve grammar, readability, and consistency. The expressions specifically highlighted by the Reviewer were revised. For example, “Below it is provided a list” was replaced with the more idiomatic formulation “Several ICE modalities have been developed, as summarized below and illustrated in Figure 1,” while the non-standard expression “in head-neck district” was removed during revision of the toxicity section. Terminology and notation were also systematically checked throughout the manuscript. In particular, all incorrect occurrences of “INFγ” and “IFNγ” were standardized to “IFN-γ”. Additional grammatical and stylistic inconsistencies identified during this review were also corrected.

Location: throughout the revised manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,

The aim of the review proposed for evaluation was to examine the biological rationale, the engineering principles, the current preclinical evidence, and the early clinical development of ICEs with particular emphasis on their potential application in HNSCC.

The choosed topic was original.

Specifying databases, the number of studies analyzed and the times may improve the clarity and credibility of the study.

The study offered important informations regarding The HNSCCs’ TME: barriers and opportunities for ICEs, ICEs: concepts, classes, and design principles, TCEs for HNSCCs: modalities and innovations,  NKCEs in HNSCC, Gamma-delta T-cell engagers, target antigens in HNSCC for ICEs, masking immunotherapy in HNSCC, synergy with ICB and TME modulation, biomarkers for patient selection and pharmacodynamic monitoring, the clinical landscape: checkpoint blockade as the foundation for ICE integration, delivery innovations and TME conditioning.

The conclusions and future perspectives were clearly formulated. They are fully supported, argued by the observations presented earlier.

I congratulate the authors for the excellent work !

Author Response

Comment: Specifying databases, the number of studies analyzed and the times may improve the clarity and credibility of the study.

Response:
We thank the Reviewer for this helpful suggestion. We agree that providing additional information regarding the literature search improves the transparency of the review. We have therefore added a brief statement at the end of the Introduction specifying that the literature search was primarily conducted using PubMed/MEDLINE and was supplemented by ClinicalTrials.gov for ongoing and completed clinical trials. We also specified that the literature and clinical-trial information were last updated in September 2026. As this manuscript was designed as a narrative rather than a systematic review, study selection was not conducted according to a predefined systematic-review protocol. Consequently, a formal number of screened and included studies cannot be reliably reported retrospectively, and we have explicitly clarified this point in the manuscript rather than providing a potentially misleading study count. We thank the Reviewer again for the encouraging assessment of our work and for this suggestion.

Location: page 3, lines 107-113.

Reviewer 3 Report

Comments and Suggestions for Authors

Review Report: Manuscript ID: cancers-4600080

 

 

Title: Immune Cell Engagers and “masking immunotherapy” for solid tumors: focus on head and neck cancers

 

The topic is timely and the authors bring together immune cell engagers, conditional activation, and the safety challenges of head and neck cancer. The review would be stronger with a more precise account of what has been demonstrated in HNSCC, clearer boundaries between platform types, and correction of several claims that currently overstate the evidence. At this stage I give a major revision and resubmit the work.

 

Major comments

1. Please distinguish HNSCC-specific evidence from evidence obtained in other tumor types throughout the review. This distinction is particularly important in the sections on emerging engager formats and masked platforms. Where HNSCC data are unavailable, please state that the proposed relevance is an extrapolation.
2. Please clarify the scope of “masking immunotherapy.” Masked cytokines and antibody–drug conjugates illustrate conditional activation, but they are not immune cell engagers. Consider presenting these platforms separately from ICEs in Table 2 and in the discussion of clinical evidence.
3. The manuscript sometimes describes protease-dependent masking as confining activation to the tumor and reducing off-target toxicity. Please distinguish the intended mechanism from effects demonstrated experimentally or clinically. The possibility of activation in inflamed normal tissue and variation in protease activity between lesions also merits discussion.
4. Please qualify the statements that ICEs act independently of MHC presentation. This applies to many engagers directed against surface antigens, whereas the peptide–HLA-directed approaches discussed later in the manuscript depend on antigen processing and the relevant HLA type.
5. Please revise Table 2 so that readers can identify the HNSCC cohort size, where available, and determine whether each reported response or adverse-event result comes from that cohort or the overall study population. Please also distinguish trial termination from discontinuation of an agent’s broader development program.
6. The discussion of target antigens would benefit from a more balanced assessment of expression in normal tissues. In particular, the description of EGFR expression as “restricted” in normal tissue appears inconsistent with the later discussion of on-target, off-tumor toxicity. Similar care is warranted for CD44v6 and B7-H3.
7. Several sections repeat the same points about immune exclusion, tumor microenvironment conditioning, and the promise of masking. Condensing these passages would create room for a more critical comparison of the available preclinical and clinical evidence.
8. The abstract refers to “targeted searches.” Please briefly describe how the literature was identified or revise this wording if no defined search approach was used.
9. Please qualify proposed safety advantages of CD8-selective, logic-gated, and half-life-extended constructs where comparative evidence is limited.
10. Please standardize abbreviations and correct typographical errors, including instances of “INFγ” in place of “IFN-γ.”
11. Sections 4.2 and 8.2 should distinguish logic gating demonstrated in soluble engagers from concepts established primarily in engineered cell therapies.
12. The statement that longer circulating TCE formats may reduce cytokine-release toxicity (Section 4.1) needs qualification; improved pharmacokinetics alone does not establish improved safety.
13. The manuscript would benefit from a thorough language edit. Several passages are repetitive or awkwardly phrased, and typographical errors occur throughout.
14. Please revise for clarity and concision, standardize terminology and abbreviations (including IFN-γ and HNSCC), and check the text, tables, and figure legends for consistency.

 

 

Author Response

Comment 1: Please distinguish HNSCC-specific evidence from evidence obtained in other tumor types throughout the review. This distinction is particularly important in the sections on emerging engager formats and masked platforms. Where HNSCC data are unavailable, please state that the proposed relevance is an extrapolation.

Response:
We thank the Reviewer for this important recommendation. Throughout the revision process, we have more clearly distinguished direct HNSCC evidence from findings obtained in other solid tumors, particularly where clinical evidence in HNSCC remains limited. In the NKCE section, for example, we now explicitly state that the available clinical experience with AFM24 was not HNSCC-specific and that its clinical efficacy in HNSCC remains to be established. More generally, claims regarding emerging platforms have been qualified where their relevance to HNSCC is based predominantly on biological rationale or extrapolation from other tumor types.

Location: Section 5, Pages 15-16, Lines 524–540.

Comment 2: Please clarify the scope of “masking immunotherapy.” Masked cytokines and antibody–drug conjugates illustrate conditional activation, but they are not immune cell engagers. Consider presenting these platforms separately from ICEs in Table 2 and in the discussion of clinical evidence.

Response:
We agree and have clarified that masking represents an engineering strategy rather than a distinct class of ICEs. Non-ICE masked platforms are retained because they provide proof-of-concept for conditional tumor activation, but they are now explicitly distinguished from ICEs. Table 2 has been reorganized into separate blocks for “ICE Drugs” and “Masked Immunotherapeutic Drugs”, allowing readers to distinguish ICEs from non-ICE conditionally activated platforms.

Location: Table 2, Pages 23–24.

Comment 3: The manuscript sometimes describes protease-dependent masking as confining activation to the tumor and reducing off-target toxicity. Please distinguish the intended mechanism from effects demonstrated experimentally or clinically. The possibility of activation in inflamed normal tissue and variation in protease activity between lesions also merits discussion.

Response:
We thank the Reviewer for this important point. We have revised the discussion to avoid equating protease-dependent activation with absolute tumor specificity. The manuscript now explicitly acknowledges that proteases capable of cleaving masking domains may also be active in inflamed or remodeling normal tissues and that inter-patient and inter-lesional variability in protease activity may result in heterogeneous drug activation. We therefore emphasize that any improvement in therapeutic index depends on sufficient differential protease activity between tumor and normal tissues and requires construct-specific clinical validation. Related statements regarding masked IL-12 and conditionally activated TCEs were also qualified to describe preferential rather than absolute tumor-restricted activation.

Location: Section 3.4, Page 8, Lines 278-307; Section 4.6, Page 12, Lines 456–462; Section 8.1, Page 18, Lines 660–662; Section 8.2, Page 19, Lines 690-693.

Comment 4: Please qualify the statements that ICEs act independently of MHC presentation. This applies to many engagers directed against surface antigens, whereas the peptide–HLA-directed approaches discussed later in the manuscript depend on antigen processing and the relevant HLA type.

Response:
We agree that MHC independence should not be generalized to all ICE platforms, particularly because peptide–HLA-directed engagers require antigen processing, peptide presentation, and the appropriate HLA allele. We have therefore qualified the general statement in the Conclusions by stating that ICEs act “often independently of MHC presentation”, rather than presenting MHC independence as a universal property. 

Location: Conclusions, Page 25, Line 877

Comment 5: Please revise Table 2 so that readers can identify the HNSCC cohort size, where available, and determine whether each reported response or adverse-event result comes from that cohort or the overall study population. Please also distinguish trial termination from discontinuation of an agent’s broader development program.

Response:
We agree and have comprehensively revised Table 2. HNSCC cohort size and HNSCC-specific efficacy results are now reported where publicly available, while results derived from the overall study population are clearly distinguished from HNSCC-specific data. When HNSCC was an eligible tumor type but subgroup-specific results were not reported, this is explicitly stated. According to the related published articles, we intended "discontinuation" as the interruption due to excessive toxicity of an experimental drug, while "termination" as the interruption due to limited anti-cancer activity or other causes not related to toxicity.

Location: Table 2, Pages 23–24.

Comment 6: The discussion of target antigens would benefit from a more balanced assessment of expression in normal tissues. In particular, the description of EGFR expression as “restricted” in normal tissue appears inconsistent with the later discussion of on-target, off-tumor toxicity. Similar care is warranted for CD44v6 and B7-H3.

Response:
We agree that target expression in normal tissues requires careful qualification. For this reason we thought it more appropriate to modify the CD44v6 section to explicitly state that its expression is not tumor-exclusive and can occur in normal epithelial tissues, with the resulting potential for on-target, off-tumor toxicity. The manuscript also explicitly discusses physiological EGFR expression as a source of on-target, off-tumor toxicity.

Location: Section 7, Page 17, Lines 599–604; Section 8.2, Page 19, Lines 690-693.

Comment 7: Several sections repeat the same points about immune exclusion, tumor microenvironment conditioning, and the promise of masking. Condensing these passages would create room for a more critical comparison of the available preclinical and clinical evidence.

Response:
We agree. Several repetitive passages concerning immune exclusion, TME conditioning, antigen heterogeneity, anatomical vulnerability, and tumor-selective activation have been condensed throughout the manuscript. This allowed us to retain the main biological concepts while making the discussion more concise and creating greater emphasis on the critical comparison of preclinical and clinical evidence.

Location: throughout the manuscript.

Comment 8: The abstract refers to “targeted searches.” Please briefly describe how the literature was identified or revise this wording if no defined search approach was used.

Response:
We agree and have clarified the literature-identification approach. The Introduction now specifies that the review was based on the authors’ comprehensive literature knowledge supplemented by targeted searches of PubMed/MEDLINE and ClinicalTrials.gov, with literature and clinical-trial information last updated in September 2026. We also clarify that, because this is a narrative review, study selection was not based on a predefined systematic-review protocol.

Location: Introduction, Page 3, Lines 107-113.

Comment 9: Please qualify proposed safety advantages of CD8-selective, logic-gated, and half-life-extended constructs where comparative evidence is limited.

Response:
We agree that the proposed safety advantages of emerging constructs should be distinguished from clinically demonstrated benefits. The CD8-selective TITAN section has therefore been revised to state that the platform is designed to limit CD4⁺ T-cell engagement and potentially reduce cytokine production, while explicitly acknowledging that whether this translates into a clinically meaningful reduction in CRS or an improved therapeutic index remains to be established. Similar caution has been introduced for logic-gated and half-life-extended formats.

Location: Section 4.3, Page 10, Lines 384–391; Section 8.2, Page 20, Lines 720-726.

Comment 10: Please standardize abbreviations and correct typographical errors, including instances of “INFγ” in place of “IFN-γ.”

Response:
We thank the Reviewer. Abbreviations and cytokine nomenclature have been standardized throughout the manuscript, including correction of all residual instances of “INFγ/IFNγ” to the standard notation “IFN-γ”.

Location: throughout the manuscript.

Comment 11: Sections 4.2 and 8.2 should distinguish logic gating demonstrated in soluble engagers from concepts established primarily in engineered cell therapies.

Response:
We agree. The revised manuscript now explicitly distinguishes Boolean logic gating established primarily in engineered cellular therapies from the less mature evidence available for soluble T-cell engagers. We state in section 8.2 that logic gating is most extensively established in engineered cellular therapies, particularly CAR-T cells, whereas related principles have been demonstrated in soluble TCEs but remain predominantly supported by preclinical evidence.

Location: Section 8.2, Page 20, Lines 720-726.

Comment 12: The statement that longer circulating TCE formats may reduce cytokine-release toxicity (Section 4.1) needs qualification; improved pharmacokinetics alone does not establish improved safety.

Response:
We agree that improved pharmacokinetics should not be interpreted as evidence of improved safety. We have therefore revised Section 4.1 to distinguish the pharmacokinetic advantages of HLE-TCEs from their safety profile. The revised text states that HLE formats may prolong systemic exposure and reduce the need for continuous infusion, while explicitly noting that improved pharmacokinetics alone does not establish reduced CRS, which remains dependent on construct design, target biology, dose, and treatment schedule.

Location: Section 4.1, Page 10, Lines 357–361.

Comment 13: The manuscript would benefit from a thorough language edit. Several passages are repetitive or awkwardly phrased, and typographical errors occur throughout.

Response:
We thank the Reviewer. The manuscript has undergone an additional language-editing and concision pass. Repetitive and awkwardly phrased passages have been revised throughout, while preserving the scientific content.

Location: throughout the manuscript.

Comment 14: Please revise for clarity and concision, standardize terminology and abbreviations (including IFN-γ and HNSCC), and check the text, tables, and figure legends for consistency.

Response:
We agree. Terminology, abbreviations, and notation have been standardized across the main text, tables, and figure legends, and consistency between these components has been rechecked.

Location: throughout the revised manuscript, including Tables 1–2.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The revised manuscript provides a useful overview of immune-cell engagers and conditionally activated therapeutic platforms in HNSCC. The added discussion of protease variability, the pharmacokinetic caveats, and the separation of ICEs from non-ICE masked therapies improve the review. However, several concerns remain only partially addressed.

1. MHC dependence: The abstract and Sections 3.1 and 3.5 still describe MHC independence as a general property of ICEs. Please consistently distinguish surface-antigen-directed engagers from peptide–HLA-directed constructs, which require antigen processing, presentation, and the appropriate HLA allele.

2. Normal-tissue expression: Section 7 retains the description of EGFR expression in normal tissues as “restricted,” despite acknowledging physiological EGFR expression and off-tumor toxicity later. The statement regarding “minimal” normal-tissue B7-H3 expression also requires qualification and supporting evidence.

3. Conditional activation: The new limitations discussed in Section 4.6 are appropriate, but other passages still imply exclusive tumor activation or established safety improvements. Please revise the Simple Summary, abstract, main text, and Table 1 consistently to distinguish intended tumor preference from experimentally or clinically demonstrated selectivity.

4. Logic gating: Section 4.2 continues to present AND/NOT gating and reduced off-tumor activation broadly as established properties of soluble engagers. Please identify the relevant constructs, specify their evidence level, and distinguish them from concepts established primarily in engineered cellular therapies.

5. Clinical evidence and Table 2: Clearly separate overall-study safety findings from HNSCC-specific enrollment and efficacy results. Provide denominators for adverse-event findings, including the CX-2029 DLT data. HNSCC eligibility should not be described as an enrolled HNSCC cohort unless enrollment is documented.

6. Trial and development status: The response letter incorrectly distinguishes “discontinuation” from “termination” according to toxicity versus efficacy. Please report trial status, the documented reason for stopping, and the status of the broader development program separately.

7. Development stages: Table 1 lists TITAN as “early clinical,” whereas the cited reference is a platform-development conference abstract. Please provide supporting clinical evidence or revise the development-stage designation.

8. Presentation and consistency: Improve table column widths and page breaks to avoid fragmented drug names and mechanism descriptions. Please also ensure that the response letter accurately reflects the changes implemented throughout the manuscript.

These revisions would strengthen the scientific precision and critical value of the review.

Comments on the Quality of English Language

The manuscript is understandable, but it needs further language editing before publication. Several passages remain repetitive or awkwardly phrased, and terminology is not consistently standardized despite the stated language revision.

Author Response

Comment 1 (MHC dependence): The abstract and Sections 3.1 and 3.5 still describe MHC independence as a general property of ICEs. Please consistently distinguish surface-antigen-directed engagers from peptide–HLA-directed constructs, which require antigen processing, presentation, and the appropriate HLA allele.

Response: We thank the Reviewer for this important clarification. We have revised the manuscript to distinguish conventional surface-antigen-directed immune-cell engagers, which can bypass classical MHC-restricted antigen presentation, from peptide–HLA-directed TCR-based or TCR-mimic engagers, which remain dependent on intracellular antigen processing, presentation of the cognate peptide, and expression of the appropriate HLA allele. This distinction has been incorporated into the Abstract and the main text, including the explanation of fundamental ICE mechanisms and their relevance to HNSCC. 

Locations: pages 1-2, lines 38-41; page 5, lines 207-215; page 9, lines 340–347.

 

Comment 2 (Normal-tissue expression): Section 7 retains the description of EGFR expression in normal tissues as “restricted,” despite acknowledging physiological EGFR expression and off-tumor toxicity later. The statement regarding “minimal” normal-tissue B7-H3 expression also requires qualification and supporting evidence.

Response: We thank the Reviewer for emphasizing that tumor-associated antigens should not be described as strictly tumor-specific. We have revised the discussion of EGFR and B7-H3 to acknowledge their expression outside malignant tissues and the consequent possibility of on-target, off-tumor toxicity. In particular, we now explicitly state that EGFR is physiologically expressed in several epithelial and endothelial tissues. We have also clarified that B7-H3 expression is not restricted to malignant cells and may occur in selected normal tissues and non-malignant stromal compartments. These considerations are now linked to the rationale for developing conditionally activated and other tumor-preferential targeting strategies in HNSCC.

Location: page 16, lines 595–601; page 17, lines 628–631.

 

Comment 3 (Conditional activation): The new limitations discussed in Section 4.6 are appropriate, but other passages still imply exclusive tumor activation or established safety improvements. Please revise the Simple Summary, abstract, main text, and Table 1 consistently to distinguish intended tumor preference from experimentally or clinically demonstrated selectivity

Response: We thank the Reviewer for highlighting the need to distinguish the intended biological advantages of masked or conditionally activated platforms from clinically demonstrated improvements in safety. We have revised the Simple Summary and Abstract to clarify that preclinical findings supporting tumor-preferential activation do not establish clinically meaningful tumor selectivity or an improved therapeutic index in patients. We have also modified the safety discussion to emphasize that conditionally activated TCEs are designed to preferentially restore activity under TME-associated conditions, while the magnitude of any resulting clinical safety benefit remains to be determined. Furthermore, the manuscript discusses important limitations of protease-dependent activation, including protease activity in inflamed or remodeling normal tissues, interpatient and interlesional heterogeneity, and the need for platform-specific validation. Related qualifications have been incorporated into the discussion of checkpoint-inhibitory TCEs and into Table 1.

Location: page 1, lines 19–25; page 2, lines 43–50; page 8, lines 321–325; page 11, lines 431–437; pages 13-14, Table 1 (COBRA/Probody/Logic-Gated ICEs)

 

Comment 4 (Logic gating): Section 4.2 continues to present AND/NOT gating and reduced off-tumor activation broadly as established properties of soluble engagers. Please identify the relevant constructs, specify their evidence level, and distinguish them from concepts established primarily in engineered cellular therapies

Response: We thank the Reviewer for this important observation. We have clarified the distinction between co-stimulatory signaling and genuine combinatorial antigen recognition. In Section 4.2, we now explain that SMITE-based strategies combine T-cell redirection with additional co-stimulatory signals, but that the presence of two signaling modules does not necessarily establish strict AND-gate behavior or prevent off-tumor activation. We have also expanded the discussion of experimental soluble T-cell engager architectures designed to favor activation in the presence of multiple tumor-associated antigens. We explicitly acknowledge that evidence supporting AND-like recognition in soluble TCEs remains predominantly preclinical and that improvements in clinical selectivity and safety have not been established. Finally, we distinguish these approaches from inhibitory NOT-gating and more complex programmable antigen-sensing circuits, which have been developed predominantly in engineered cellular therapies, including CAR-T cells.

Location: page 10, lines 385–406; page 19, lines 732–736; pages 13-14, Table 1 (SMITEs)

 

Comment 5 (Clinical evidence and Table 2): Clearly separate overall-study safety findings from HNSCC-specific enrollment and efficacy results. Provide denominators for adverse-event findings, including the CX-2029 DLT data. HNSCC eligibility should not be described as an enrolled HNSCC cohort unless enrollment is documented.

Response: We thank the Reviewer for emphasizing the importance of distinguishing eligibility for HNSCC enrollment from actual enrollment and separately reported HNSCC-specific outcomes. We have extensively revised Section 11 and Table 2 to improve the accuracy and transparency of clinical evidence reports. In particular, we now distinguish overall-study safety and efficacy findings from results specifically reported for patients with HNSCC. We also clarify that eligibility for an HNSCC cohort does not necessarily establish that patients with HNSCC were enrolled or treated. For CX-2029, we have separated the overall Phase I dose-escalation safety results from the sponsor-reported HNSCC expansion data. The table now specifies that 45 patients were included in the overall dose-escalation population, whereas the HNSCC expansion included 29 treated patients and 25 efficacy-evaluable patients at the October 29, 2021 cutoff. One confirmed partial response was reported in the HNSCC efficacy-evaluable population, corresponding to an ORR of 4.0%, with a disease control rate of 56.0%. For CX-2009, we have clarified that the overall study included 99 treated patients, with 80 response-evaluable patients. We have also specified that the HNSCC subgroup comprised nine patients and that the reported partial response was unconfirmed. Accordingly, this observation is not presented as an established confirmed HNSCC-specific ORR. For the remaining agents, we have clarified whether HNSCC was an eligible population and whether HNSCC-specific enrollment or efficacy outcomes have been publicly reported. The revised table notes further explain the interpretation of study populations, denominators, and available clinical evidence.

Location: pages 22–24, Table 2 (all results sections) and lines 840-896

 

Comment 6 (Trial and development status): The response letter incorrectly distinguishes “discontinuation” from “termination” according to toxicity versus efficacy. Please report trial status, the documented reason for stopping, and the status of the broader development program separately.

Response: We thank the Reviewer for highlighting the need to distinguish clinical-trial termination from broader decisions regarding the continuation or discontinuation of drug-development programs. We have revised Table 2 to report trial registry status and the available reasons for termination or discontinuation more precisely, avoiding the assumption that all discontinued programs failed because of insufficient antitumor efficacy. For TAK-186, we now report the terminated trial status, the available overall-study results, and the sponsor's decision to discontinue further development following limited monotherapy antitumor activity. For JANX008, we distinguish the registered trial status from the sponsor's decision not to pursue further clinical development following Phase Ia evaluation. For GEN1044, we have clarified that the trial was stopped during dose escalation following a safety-committee assessment of the overall safety profile and emerging safety signals, and that the planned expansion phase was not initiated. We distinguish this trial-level decision from the subsequent development-program outcome. For CX-2009, we have clarified that trial termination was associated with the COVID-19 pandemic and business-related restructuring rather than a documented conclusion of insufficient antitumor efficacy. For studies without a clearly documented discontinuation decision, we have avoided inferring an unsupported reason.

Location: pages 23-24, Table 2 (all results sections)

 

Comment 7 (Development stages): Table 1 lists TITAN as “early clinical,” whereas the cited reference is a platform-development conference abstract. Please provide supporting clinical evidence or revise the development-stage designation

Response: We thank the Reviewer for requesting clarification of the developmental status of the TITAN platform. For this reason we added in Section 4.3 the clinical-development example of AZD5492, a CD20-directed TITAN construct being investigated in a Phase 1/2 study in B-cell malignancies (NCT06542250). We clarify that this clinical program does not constitute evidence of clinical efficacy in HNSCC. The corresponding development-stage information has also been clarified in Table 1.

Location: page 10, lines 418–420; page 13, Table 1 (TITAN)

 

Comment 8 (Presentation and consistency): Improve table column widths and page breaks to avoid fragmented drug names and mechanism descriptions. Please also ensure that the response letter accurately reflects the changes implemented throughout the manuscript.

Response: We thank the Reviewer for this helpful suggestion. We have revised the formatting of Tables 1 and 2 to improve readability, particularly for Table 1 we have added internal cell borders. We have prepared point-by-point responses with explicit references to the relevant manuscript sections, pages, and line numbers, so that the changes described in the response letter can be readily located and verified. In addition, we have also added six new references (n° 59, 61, 72, 76, 120 and 121) to better support our revision following your comments.

Round 3

Reviewer 3 Report

Comments and Suggestions for Authors

Comments for the authors

The revised manuscript provides a more balanced discussion of immune-cell engagers and conditionally activated immunotherapies in solid tumors, with particular emphasis on HNSCC. The clarification of HLA dependence, limitations of protease-mediated tumor selectivity, and differences between trial eligibility and documented HNSCC outcomes strengthens the review. I recommend minor revision to address the following remaining points:

1. In Section 8.2, lines 701–703, please correct the distinction between Probody and COBRA platforms. Probody technology is not restricted to conventional monoclonal antibodies and also encompasses antibody–drug conjugates and T-cell-engaging bispecifics. Please ensure consistency with Table 1.

2. In Section 4.6, lines 463–475, some statements still imply that masked constructs become active exclusively within tumors and provide established safety benefits. Please consistently describe activation as preferential rather than exclusively tumor-restricted, and distinguish intended therapeutic-index improvements from demonstrated clinical benefits.

3. In Section 8.2, lines 736–740, please refine the explanation of logic gating. Dual-antigen recognition does not necessarily establish strict AND-gate behavior or eliminate activation against normal tissues. The discussion should distinguish experimentally validated soluble-engager designs from gating mechanisms developed primarily in engineered cellular therapies.

4. Please verify the clinical-trial terminology in Table 2. In particular, “active, not yet recruiting” for JANX008 is ambiguous. Report the exact registry status and access date, separately from sponsor decisions concerning further program development.

5. In Section 12, lines 906–910, the statement that radiotherapy “synergizes naturally with ICEs” is too definitive. Please provide direct supporting evidence or describe this combination as a mechanistic rationale requiring further validation, particularly in HNSCC.

6. Please expand the Figure 1 legend to explain the illustrated immune-cell populations, engager-mediated interactions, and relevant receptor–ligand pairs, and define the abbreviations.

Comments on the Quality of English Language

The manuscript is generally understandable, but further English language editing is required before publication. Remaining issues include subject–verb agreement, awkward sentence construction, inconsistent terminology, and overly absolute wording for findings supported mainly by preclinical evidence. For example, “Other platforms has been investigated” should be corrected to “Other platforms have been investigated.” Please review the clean manuscript after accepting all tracked changes to ensure consistent terminology and abbreviation usage, correct grammar and spacing, and reduce repetitive statements. Tracked deletions should not themselves be treated as language errors in the intended final text.

 

Author Response

Comment 1: In Section 8.2, lines 701–703, please correct the distinction between Probody and COBRA platforms. Probody technology is not restricted to conventional monoclonal antibodies and also encompasses antibody–drug conjugates and T-cell-engaging bispecifics. Please ensure consistency with Table 1.

Response: We agree with the Reviewer. Section 8.2 has been revised to clarify that Probody technology encompasses several protease-activatable antibody formats. COBRA is now described as a distinct conditionally activated T-cell-engager platform. The terminology has been aligned with Table 1.

 

Comment 2: In Section 4.6, lines 463–475, some statements still imply that masked constructs become active exclusively within tumors and provide established safety benefits. Please consistently describe activation as preferential rather than exclusively tumor-restricted, and distinguish intended therapeutic-index improvements from demonstrated clinical benefits

Response: We agree and we have revised Section 4.6 to consistently describe protease-dependent activation as preferential rather than exclusively tumor-restricted. We have also distinguished the intended improvements in therapeutic index from benefits supported by preclinical evidence and those not yet clinically established. The revised text explicitly acknowledges potential activation in inflamed or remodeling normal tissues.

 

Comment 3: In Section 8.2, lines 736–740, please refine the explanation of logic gating. Dual-antigen recognition does not necessarily establish strict AND-gate behavior or eliminate activation against normal tissues. The discussion should distinguish experimentally validated soluble-engager designs from gating mechanisms developed primarily in engineered cellular therapies.

Response: We thank the Reviewer for this clarification. The discussion has been revised to explain that dual-antigen recognition does not necessarily establish strict AND-gate behavior or eliminate off-tumor activation. We now distinguish experimentally investigated soluble-engager architectures from programmable logic-gating circuits developed predominantly in engineered cellular therapies, particularly CAR-T cells.

 

Comment 4: Please verify the clinical-trial terminology in Table 2. In particular, “active, not yet recruiting” for JANX008 is ambiguous. Report the exact registry status and access date, separately from sponsor decisions concerning further program development.

Response: We thank the reviewer for this observation. We have accessed today clinicaltrial.gov and we have seen that the recruitment status is “Active, not recruiting”, so we wrote the exact terminology in Table 2. The sponsor's decision to discontinue further clinical development is still reported separately from the registered status of the clinical trial. 

 

Comment 5: In Section 12, lines 906–910, the statement that radiotherapy “synergizes naturally with ICEs” is too definitive. Please provide direct supporting evidence or describe this combination as a mechanistic rationale requiring further validation, particularly in HNSCC.

Response: We agree that the previous wording was overly definitive. Section 12 has been revised to present radiotherapy–ICE combinations as a biologically plausible therapeutic strategy rather than an established synergistic interaction. The revised discussion acknowledges that direct evidence, particularly in HNSCC, remains limited and that efficacy, treatment sequencing and safety require further validation

 

Comment 6: Please expand the Figure 1 legend to explain the illustrated immune-cell populations, engager-mediated interactions, and relevant receptor–ligand pairs, and define the abbreviations

Response: We have written a paragraph describing the illustrated immune-effector cells, the principal engager-mediated interactions, and relevant receptor–ligand pairs. Abbreviations have also been reported.