In Vitro Skin Models for Skin Sensitisation: Challenges and Future Directions

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This review systematically recapitulates how reconstructed human skin models have been used as platforms for skin sensitization testing, including the latest approaches using organ-on-a-chip and microfluidic technologies. This review is aimed at providing ideas and references for the development of next-generation organotypic skin models with increased complexity and monitoring capabilities. However, the article still needs some improvement to be a higher quality review. And there are some concerns for authors to correct.

1. Keep the background information in the abstract and introduction brief and provide more details about the main content and key findings of the
2. The use of abbreviations was not standardized. They should be stated along with the full name when they firstly appear and used consistently thereafter. Abbreviations should not be used if they occur fewer than three times.
3. The article should provide more figures, which may include the testing methods used in the examples, relevant mechanisms, or other supporting information.
4. Since the Skin-on-a-chip was reported to provide more advanced tools for non-animal risk assessment, it is recommended to include more examples, especially those from the past two years.
5. In the conclusion of the article, it is recommended to provide the author's insights or suggestions on the construction of skin models for skin sensitisation.
6. The citation format for references should be consistent, and some citation details are missing. This should be carefully checked and corrected.

Author Response

REVIEWER 1 COMMENTS

We sincerely thank the reviewer for their valuable and constructive comments, which have greatly contributed to enhancing the quality of our manuscript. All new or revised text in the manuscript has been highlighted in yellow for your convenience.

COMMENTS TO THE AUTHORS:

1. Keep the background information in the abstract and introduction brief and provide more details about the main content and key findings of the
ANSWER: We appreciate this feedback. However, as the reviewer's comment appears to be incomplete or missing information, we were unable to fully address this suggestion.

2. The use of abbreviations was not standardized. They should be stated along with the full name when they firstly appear and used consistently thereafter. Abbreviations should not be used if they occur fewer than three times.
ANSWER: We have thoroughly re-checked the manuscript and, following your advice regarding abbreviation usage to improve clarity and readability, have removed all abbreviations that appeared only once in the text. For abbreviations used two or more times, we have ensured they are correctly defined upon their first appearance and used consistently thereafter.

3. The article should provide more figures, which may include the testing methods used in the examples, relevant mechanisms, or other supporting information.
ANSWER: We appreciate the suggestion to include more figures. However, the journal's guidelines for review articles limit the inclusion to a maximum of three figures or tables. After careful consideration, we concluded that the tables would provide readers with more comprehensive and valuable detailed information compared to an additional figure, given these constraints.

4. Since the Skin-on-a-chip was reported to provide more advanced tools for non-animal risk assessment, it is recommended to include more examples, especially those from the past two years.
ANSWER: Following the reviewer’s advice, Section 5.1 has been substantially expanded to include several recently published SoC models (from the past two years), emphasising advances in vascularisation, immune cell integration, dynamic perfusion, and multi-organ crosstalk. We also now explicitly address challenges in SoC development, including scalability, cost, reproducibility, and regulatory validation, to provide a more balanced and realistic appraisal of their current status and translational potential.

5. In the conclusion of the article, it is recommended to provide the author's insights or suggestions on the construction of skin models for skin sensitisation.
ANSWER: The reviewer's observation is well-taken. The conclusion section has been revised to incorporate more specific suggestions and future perspectives regarding in vitro skin models for sensitization assessment.

6. The citation format for references should be consistent, and some citation details are missing. This should be carefully checked and corrected.
ANSWER: We appreciate this advice and have diligently corrected references where citation information was missing or formatted inconsistently. We acknowledge that some older references lack DOI numbers, and certain sources, such as OECD guidelines or market studies, do not adhere to conventional reference styles. We have ensured these are presented as consistently as possible within these inherent limitations.

Reviewer 2 Report

Comments and Suggestions for Authors

The review discusses in vitro models for skin sensitization and integrates regulatory, mechanistic, and technological perspectives. It’s a nicely written review, but perhaps can be made better. Here are my suggestions:

1. In the abstract, please include and mention of the limitations of current in vitro methods and a clear statement on what gap this review addresses.
2. Same thing in the introduction, rather than just stating the regulatory shift, emphasize the scientific challenge of mimicking skin immunocompetence in vitro.
3. In section 2, can the authors add discussion on inter-laboratory variability and the gap in addressing pro-/pre-haptens.
4. Discuss briefly why the LLNA, despite ethical issues, remains the benchmark for potency prediction.
5. Similarly for section 4, Add a paragraph analyzing why methods like SensCeeTox and RHE/IL-18 failed to progress in validation. Was it technical limitations or variability?
6. Expand on challenges of scaling up SoC systems and their current regulatory acceptance, Maybe Include discussion of cost barriers and technical reproducibility.
7. The part on omics and In silico: These are presented well but could be tied back to how they complement RHE-based models (e.g., multi-omics to identify novel KE markers). Also, perhaps Add a sub-section projecting where the field is heading in the next 5–10 years.

Author Response

REVIEWER 2 COMMENTS

We really appreciate the valuable and insightful comments from the reviewer, which have significantly helped us to improve the quality of this review article.

All new or revised text in the manuscript has been highlighted in yellow for your convenience.

COMMENTS TO THE AUTHORS:

1. In the abstract, please include and mention of the limitations of current in vitro methods and a clear statement on what gap this review addresses.

2. Same thing in the introduction, rather than just stating the regulatory shift, emphasize the scientific challenge of mimicking skin immunocompetence in vitro.

ANSWER: (Addressing comments 1 and 2 together) We appreciate this pertinent feedback and have revised both the abstract and introduction to emphasize the scientific challenges inherent in mimicking skin immunocompetence in vitro, rather than solely focusing on regulatory shifts. The abstract has been modified to include a statement of the main limitations of individual methods and link these limitations to the potential advantages offered by in vitro skin models. Also, a mention to the use of skin models as platforms for testing keratinocyte as well as dendritic cell responses has been added.

In the introduction, we have introduced new sentences to highlight the current scientific challenges faced in developing in vitro skin models for sensitisation assessment. These additions, alongside corresponding revisions in the abstract, aim to underscore the current challenges and clearly articulate the gap that this review addresses. We believe these modifications enhance the key message regarding the limitations and ongoing efforts in in vitro skin models for skin sensitization studies.

3. In section 2, can the authors add discussion on inter-laboratory variability and the gap in addressing pro-/pre-haptens.

ANSWER: Following the reviewer’s recommendation, Section 2 has been expanded to include a comprehensive discussion on the issue of inter-laboratory variability in the application of NAMs, emphasizing how differences in protocols, equipment, and operator expertise can impact the reproducibility and reliability of results. Additionally, we discuss the challenges associated with the assessment of pro- and pre-haptens, particularly the need for controlled activation steps and standardized procedures to capture their sensitization potential accurately."

 

4. Discuss briefly why the LLNA, despite ethical issues, remains the benchmark for potency prediction.

ANSWER: We appreciate the reviewer's observation. In response, we have briefly clarified in section 2 the most important reasons why the LLNA remains the benchmark for sensitization potency prediction.

5. Similarly for section 4, Add a paragraph analyzing why methods like SensCeeTox and RHE/IL-18 failed to progress in validation. Was it technical limitations or variability?

ANSWER: This is an interesting point, but difficult to address based solely on the published information. Rather than adding a separate paragraph, because the validation status was mentioned in the test describing each individual method, we have added some of the causes that can explain the failure of SensCeeTox and RHE/IL18 to progress in the validation process in their respective sections. We have limited to mention that can be found, or we think can be inferred from published literature, emphasizing that lack of a validation study report limits the conclusions that can be drawn.

6. Expand on challenges of scaling up SoC systems and their current regulatory acceptance, Maybe Include discussion of cost barriers and technical reproducibility.

ANSWER: Following reviewer comments, we have included a new subsection, Section 5.4: Future Directions and Integration Pathways, which outlines anticipated developments in skin sensitisation testing over the next decade. This includes the integration of multi-modal NAMs, such as SoC–omics–in silico platforms, and references recent initiatives such as the FDA ISTAND programme, EU-PARC, and ONTOX, which are likely to shape regulatory acceptance and implementation of next-generation models.

7. The part on omics and In silico: These are presented well but could be tied back to how they complement RHE-based models (e.g., multi-omics to identify novel KE markers). Also, perhaps Add a sub-section projecting where the field is heading in the next 5–10 years.

ANSWER: In Sections 5.2 and 5.3, we clarified the complementary roles of omics and in silico approaches in enhancing mechanistic insight and predictive performance of RhE-based models. We also strengthened the discussion of how omics technologies are being integrated into more complex systems (e.g., SoC platforms) and how computational tools can support weight-of-evidence approaches within IATA.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I find the comments well addressed and the manuscript should be accepted for publication