PONE-D-24-28161Latent representation of H&E images retains clinical information in a breast cancer cohortPLOS ONE

Dear Dr. Benmussa,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please accept our sincere apologies for the delay on this review. Please address the reviewers' comments below.

Please submit your revised manuscript by Nov 25 2024 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Amy McCart Reed

Academic Editor

PLOS ONE

Comments from the Journal Office

Please note that a second reviewer was invited to assess your manuscript, but was unable to submit their comments formally through Editorial Manager. However, they have provided valuable feedback on your manuscript, and their comments are available at the end of this email. Please provide your response to the comments of both reviewers in your rebuttal. 

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please note that PLOS ONE has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, we expect all author-generated code to be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse.

3. Thank you for stating the following financial disclosure:

“The RESCUER project has received funding from the European Union’s Horizon 2020 Research and Innovation Programme under Grant agreement No. 847912.

A.P. received funding from Fundació La Marató TV3 201935-30, Fundación CRIS contra el cáncer PR\_EX\_2021-14, Agència de Gestó d'Ajuts Universitaris i de Recerca 2021 SGR 01156, Fundación Fero BECA ONCOXXI21, Instituto de Salud Carlos III PI22/01017, Asociación Cáncer de Mama Metastásico IV Premios M. Chiara Giorgetti, Breast Cancer Research Foundation BCRF-22-198 and BCRF-23-198, and RESCUER, funded by European Union's Horizon 2020 Research and Innovation Programme under Grant Agreement No. 847912. F.B-M. received funding from Fundación científica AECC Ayudas Investigador AECC 2021 (INVES21943BRAS).”

Please state what role the funders took in the study.  If the funders had no role, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

If this statement is not correct you must amend it as needed.

Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf.

4. Thank you for stating the following in the Competing Interests section:

“I have read the journal's policy and the authors of this manuscript have the following competing interests: A.P. reports advisory and consulting fees from Roche, Pfizer, Novartis, Amgen, BMS, Puma, Oncolytics Biotech, MSD, Guardan Health, Peptomyc and Lilly, lecture fees from Roche, Pfizer, Novartis, Amgen, BMS, Nanostring Technologies and Daiichi Sankyo, institutional financial interests from Boehringer, Novartis, Roche, Nanostring, Sysmex Europa GmbH, Medica Scientia inno. Research, SL, Celgene, Astellas and Pfizer; stockholder and consultant of Reveal Genomics, SL; patents filed PCT/EP2016/080056, PCT/EP2022/086493, PCT/EP2023/060810, EP23382703 and EP23383369. Z.Y. is consulting at Verily Inc. F.B-M. has patents filed: PCT/EP2022/086493, PCT/EP2023/060810, EP23382703 and EP23383369. The rest of the authors has no competing interests.”

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

5. In the online submission form, you indicated that [Code available upon request. For data access, please contact fbraso@clinic.cat. And for code access, contact chloebenmussa@gmail.com].

All PLOS journals now require all data underlying the findings described in their manuscript to be freely available to other researchers, either 1. In a public repository, 2. Within the manuscript itself, or 3. Uploaded as supplementary information.

This policy applies to all data except where public deposition would breach compliance with the protocol approved by your research ethics board. If your data cannot be made publicly available for ethical or legal reasons (e.g., public availability would compromise patient privacy), please explain your reasons on resubmission and your exemption request will be escalated for approval.

6. Thank you for stating the following in the Acknowledgments Section of your manuscript:

“We thank the RESCUER consortium members for useful discussions and for the collaboration. We thank the Yakhini Research Group, especially Alon Oring, for important discussion and comments. The RESCUER project has received funding from the European Union’s Horizon 2020 Research and Innovation Programme under Grant agreement No. 847912. A.P. received funding from Fundaci´o La Marat´o TV3 201935-30, Fundaci´on CRIS contra el c´ancer PR EX 2021-14, Ag`encia de Gest´o d’Ajuts Universitaris i de Recerca 2021 SGR 01156, Fundaci´on Fero BECA ONCOXXI21, Instituto de Salud Carlos III PI22/01017, Asociaci´on C´ancer de Mama Metast´asico IV Premios M. Chiara Giorgetti, Breast Cancer Research Foundation BCRF-22-198 and BCRF-23-198, and RESCUER, funded by European Union’s Horizon 2020 Research and Innovation Programme under Grant Agreement No. 847912. F.B-M. received funding from Fundaci´on cient´ıfica AECC Ayudas Investigador AECC 2021 (INVES21943BRAS).”

We note that you have provided funding information that is currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

“The RESCUER project has received funding from the European Union’s Horizon 2020 Research and Innovation Programme under Grant agreement No. 847912.

A.P. received funding from Fundació La Marató TV3 201935-30, Fundación CRIS contra el cáncer PR\_EX\_2021-14, Agència de Gestó d'Ajuts Universitaris i de Recerca 2021 SGR 01156, Fundación Fero BECA ONCOXXI21, Instituto de Salud Carlos III PI22/01017, Asociación Cáncer de Mama Metastásico IV Premios M. Chiara Giorgetti, Breast Cancer Research Foundation BCRF-22-198 and BCRF-23-198, and RESCUER, funded by European Union's Horizon 2020 Research and Innovation Programme under Grant Agreement No. 847912. F.B-M. received funding from Fundación científica AECC Ayudas Investigador AECC 2021 (INVES21943BRAS).”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

7. We notice that your supplementary figure 1 is included in the manuscript file. Please remove them and upload them with the file type 'Supporting Information'. Please ensure that each Supporting Information file has a legend listed in the manuscript after the references list.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors describe a computational approach to classify regions of Haematoxylin and eosin stained Whole Slide Images of breast cancer with corresponding biological/molecular labels and to use autoencoders to generate a latent (lower dimension) representation of the data, highlighting how this impacts retention of clinically relevant information. The study is relevant and the approach appears to be sound but I suggest that the term "clinical" in the title be replaced as the majority of the data appears to be directly tumour-related (pathological/molecular or biological).

In areas the language is somewhat unclear e.g. lines 273/274 making it difficult to understand. The paper would benefit from some language revision of the text in areas to improve readability and understanding.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

**********

Addition Feedback from a Second Reviewer

Manuscript Summary:

This paper investigates the use of convolutional autoencoders to reduce the compute requirements for performing deep learning classification on tumor histology images. First, this paper evaluates deep learning prediction capabilities for a series of biomarkers in the specific breast cancer dataset used. These results are reported, and the authors speculate on tumor heterogeneity using tile-level predictions of PAM50 subtype. Second, a series of convolutional autoencoders were trained to reduce the dimensionality of the tile images and evaluate whether the biomarker label Ki67 could be learned from this lower-dimensional data. Across a range of sizes of latent embeddings, Ki67 could still be predicted, although performance was diminished by reducing the latent embedding dimensions. Additionally, to investigate whether biomarker labels can be inferred from the latent embedding, the authors compare the inertia for a subset of tiles with biomarker labels to random labeling and find lower inertia for biomarker labeled tiles.

Significance:

Understanding the heterogeneity of cancer from a deep learning computation on readily-available H&E stained slides could be clinically significant, particularly for predicting which patients may have treatment-resistant cancers and adjusting their treatment strategy. Previous studies have explored the use of deep learning applied to tumor histology images to infer heterogeneity, including Levy-Jurgenson et al (2020, https://doi.org/10.1038/s41598-020-75708-z) and Zormpas-Petridis et al (2021, https://doi.org/10.3389/fonc.2020.586292) which used whole-slide labels to infer heterogeneity of tumors. 

Additionally, Shahamatdar et al (2024, https://doi.org/10.1111/his.15180) learned local tile predictions from whole-slide labels and compared these predictions to spatial labels from laser capture microdissection (LCM) to find that deep learning predictions were not consistent with LCM ground truth. This paper applies published methods to a published dataset.

Additionally, gaining a greater understanding of the variables that deep learning models use to predict clinical features from histology images is of critical importance for the translation of such black box models to the clinical setting. Prior work has used autoencoders to extract features from breast cancer histology data, including work from Xu et al (2015, 10.1109/TMI.2015.2458702) and Xie et al (2019, 10.3389/fgene.2019.00080). The use of low dimensional latent representations allow the authors to gain some insight into the variables performant models are utilizing for prediction, like tile hue for 2 latent dimensions. The major advance that this paper offers is a study of how compressing histology images using a convolutional autoencoder effects deep learning classification results.

Major points and limitations:

#While an investigation into the effects of dimensionality reduction could be valuable to better understand prediction-associated features I believe this presented work only superficially explores this topic.

#The data towards assessing heterogeneity is PAM50 subtype tile-level predictions and values for patient-level model accuracy. It is difficult to make any conclusions from predictions of one label shown for 6 slides. While characterizing the heterogeneity of a tumor is valuable and deep learning applied to readily available H&E scanned slides would be an exciting tool for gaining such valuable information, this paper does not generate results towards this end. Ground truth spatial labels would be needed to evaluate whether this work does “support the assessment of heterogeneity” [346].

#The authors note that their motivation for using an autoencoder to reduce the dimensionality of the data is to lessen the compute resources required to train classification models [227]. While it does take substantial compute resources to train large classification models, the most performant models in the field are massive multi-institutional foundation models. Therefore, hospitals implementing classification models for patient care would only be evaluating models for the patient at hand, a relatively trivial computational task. Additionally, the degradation of performance with dimensionality reduction is a significant concern for the clinical setting where the accuracy of tests is paramount. I feel that this is not the proper motivation for the subsequent work.

#Ki67 is the only biomarker with results reported that allow for critical evaluation of how dimensionality reduction impacts performance. The performance for all labels should be reported.

#In Figure 9 it is expected that some information relevant to label classification is lost at the latent dimension is reduced, however, the percentage of instances for which the random inertia is above the label inertia remains relatively stable. From table 3 we can note validation accuracy >0.5, AUROC > 0.5, and F1 score >0.5, meaning that some information is lost but at 2 latent dimensions some relevant information for Ki67 classification remains. A deeper exploration of the information encoded in the latent space that allows for better than random prediction is needed. Examples of supporting data would be dimensionality reduced plots (PCA, UMAP, or tSNE) of the latent space colored by class label side by side with the same scatterplot but with tile images plotted for each point. This should be done for each dimensionality of latent space.

#The data is developed from a multi-centre trial, and interpretation of latent space reduced dimensions could rely on site-specific batch effects (Howard Nature Communications 2021). Please provide baseline model performance for the primary endpoints of interest using relevant meta-data variables as solitary inputs. A model understanding predictive performance of trial site as the only inclusion variable is essential.  

# Many figures are of low quality, and have text which are either unreadable or very low resolution (example: Figure 5). 

Minor points:

# All text should be past tense e.g. [60, 61, 98, 100, 130, 183]

# Clarification of how tiles are generated, whether they are over the whole tissue section, a pathologist-annotated tumor region, and if white-space is excluded is needed [153-155].

#The trait descriptions can be shorted and would benefit from including the assay used to generate the label.

_______________________

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PONE-D-24-28161R1Latent representation of H&E images retains biological information in a breast cancer cohortPLOS ONE

Dear Dr. Benmussa,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. Upon review, we noticed that several of the figures appear blurry and low in resolution, which makes it difficult to interpret the data clearly. Could you please replace all figures with high-quality, high-resolution versions (ideally ≥300 dpi) so that labels, axes, and graphical details are crisp and easily readable?We appreciate your attention to this matter and look forward to receiving the revised manuscript with improved figures.

Please submit your revised manuscript by Jul 11 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Amgad Muneer

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Latent representation of H&E images retains biological information in a breast cancer cohort

PONE-D-24-28161R2

Dear Dr. Benmussa,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. If you have any questions relating to publication charges, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Amgad Muneer

Academic Editor

PLOS ONE