Estrogen Receptor-Low Positive (ER-Low) Breast Cancer: A Unique Clinical and Pathological Entity

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Overall, this is a solid and relevant manuscript that requires only minor revisions.

It is a very nice compilation on the topic of ER-low breast carcinoma, highlighting the necessity of using AI for relevant and reproducible assessment.

I have only minor comments.

In row 78, immunohistochemistry is abbreviated in brackets as (IHC). In rows 81–82, the same term is abbreviated as IIC. I do not understand this change.

In row 159, the proliferation index is written as ki67 with a lowercase “k”. It should be written as Ki67 or Ki-67.

In row 412, the authors did not mention the Finnish AI tool Aiforia, which also includes an AI platform for ER assessment.

Author Response

Comments 1: Overall, this is a solid and relevant manuscript that requires only minor revisions. It is a very nice compilation on the topic of ER-low breast carcinoma, highlighting the necessity of using AI for relevant and reproducible assessment. I have only minor comments.

Response 1: Thank you very much for your comments and suggestions. We are delighted that you appreciated our manuscript and we believe that your comments helped us to improve our work. You can find the changes that we have done in our paper written in red.

Comments 2: In row 78, immunohistochemistry is abbreviated in brackets as (IHC). In rows 81–82, the same term is abbreviated as IIC. I do not understand this change.

Response 2: Thank you for your observation. It was a mistake. We have corrected the sentence and used in rows 81-82 and afterwords the same abbreviation (IHC) for immunohistochemistry.

Comments 3: In row 159, the proliferation index is written as ki67 with a lowercase “k”. It should be written as Ki67 or Ki-67.

Response 3: Thank you. We have changed ki67 into Ki67.

Comments 4: In row 412, the authors did not mention the Finnish AI tool Aiforia, which also includes an AI platform for ER assessment.

Response 4: Thank you. We have mentioned Aiforia and added a sentence about this AI tool.

Reviewer 2 Report

Comments and Suggestions for Authors

This review examines the challenging subtype of estrogen receptor-low (ER-low, 1–9%) breast cancer, covering its pathological definition, clinical characteristics, treatment strategies—with a focus on endocrine therapy, chemotherapy, and immunotherapy—and the prospects of artificial intelligence in precision diagnosis. The topic holds significant clinical and academic value, and the review covers a broad scope with up-to-date references. However, the article shows notable shortcomings in several key aspects:

 

1. Some discussions remain superficial and lack in-depth analysis of critical controversies:

(1) When citing the ASCO guidelines, it should more precisely indicate that the current (2020) guidelines define 1–10% as "ER-low positive," pointing out that this is precisely the source of clinical dilemma: the guidelines provide the definition but do not clearly state that its treatment recommendations differ entirely from those for ER-positive cancers with >10% expression (lines 111–113). This is an important contextual detail.

(2) The information on clinical trials mentioned in the future prospects section (NCT02115048, NCT03594396) is too brief. It is recommended to supplement the rationale behind these trial designs (e.g., why test Afatinib or PARPi combined with immunotherapy in ER-low cases) and briefly discuss their potential implications. The mention of CDK4/6 inhibitors is overly simplistic; it could be noted that current data are lacking, but theoretically they might be effective if a functional ER pathway exists, highlighting the need for further research.

(3) The nature of the controversy surrounding endocrine therapy should be more explicitly stated—it is not just about “whether it is effective,” but more importantly “for whom it is effective.” This relates to the heterogeneity of ER expression, ER signaling pathway activity (beyond mere protein expression), and interactions with other pathways (e.g., growth factor receptor pathways). Concepts such as “ER signal dilution” or “functional ER loss” in ER-low tumors could be briefly introduced.

 

2. Transitions within and between certain paragraphs are not smooth, making the text read like a simple listing of research findings without sufficient integration and analysis:

(1) In Part 3, Treatment Strategies, the discussion on endocrine therapy (ET) is central, but the conclusions are somewhat vague. It is suggested to add a summary paragraph at the end of Section 3.1, clearly stating: Current evidence indicates that for the vast majority of ER-low patients (especially those who are PR-negative), ET offers limited and uncertain benefit; however, for patients with ER expression close to 10% (e.g., 6–10%), PR-positive status, or those who do not achieve pCR after chemotherapy, there may still be some benefit, and thus ET should not be automatically excluded. This would better align with the later discussion on the need for AI-based precise stratification.

(2) Part 4 details several AI platforms, while Part 5 mentions "multimodal AI." It is recommended that the beginning of Part 5 clearly explains that multimodal AI represents the next-generation technology beyond the single-task (e.g., ER quantification) AI models described in Part 4. By integrating multidimensional data such as histology, genomics, and imaging, it holds promise for more precise subtyping and treatment prediction for ER-low patients, thereby providing a natural transition.

 

3. Formatting and References:

Some citation formats need to be unified, and it should be confirmed that all references correspond to in-text citations. It is advised to carefully check the formatting of in-text citations and the reference list to ensure compliance with journal requirements.

 

4. Suggested relevant literature to cite:

PMCID: PMC12221439, PMCID: PMC12637072, PMCID: PMC12536554, PMCID: PMC11075339, PMCID: PMC10799417, PMCID: PMC10304265.

Author Response

Comments 1: The English could be improved to more clearly express the research.

Response 1: Thank you. We have checked the manuscript and revised English accordingly.

Comments 2: This review examines the challenging subtype of estrogen receptor-low (ER-low, 1–9%) breast cancer, covering its pathological definition, clinical characteristics, treatment strategies—with a focus on endocrine therapy, chemotherapy, and immunotherapy—and the prospects of artificial intelligence in precision diagnosis. The topic holds significant clinical and academic value, and the review covers a broad scope with up-to-date references. However, the article shows notable shortcomings in several key aspects

Response 2: Thank you very much for your comments. We believe that your suggestions helped us to improve our work. We have modified the manuscript according to your comments and requests. You can find the changes that we have done in our paper written in red.

Comments 3.1: Some discussions remain superficial and lack in-depth analysis of critical controversies:

• Comments 3.1.1.: When citing the ASCO guidelines, it should more precisely indicate that the current (2020) guidelines define 1–10% as "ER-low positive," pointing out that this is precisely the source of clinical dilemma: the guidelines provide the definition but do not clearly state that its treatment recommendations differ entirely from those for ER-positive cancers with >10% expression (lines 111–113). This is an important contextual detail.
• Response 3.1.1: Thank you for your observation. We have specified that the current ASCO guidelines are those of 2020 and added some sentenced underlying the controversy between pathological definition and indications for clinical management of ER-low BC compared to ER positive BC expressing ER in more than 10% of cancer cells.
• Comments 3.1.2: The information on clinical trials mentioned in the future prospects section (NCT02115048, NCT03594396) is too brief. It is recommended to supplement the rationale behind these trial designs (e.g., why test Afatinib or PARPi combined with immunotherapy in ER-low cases) and briefly discuss their potential implications. The mention of CDK4/6 inhibitors is overly simplistic; it could be noted that current data are lacking, but theoretically they might be effective if a functional ER pathway exists, highlighting the need for further research.
• Response 3.1.2: Thank you for your suggestions. We have provided the rationale for the studies mentioned and briefly discussed their implications. We also have gone deeper into the topic of CDK4/6 inhibitors in ER-Low BC patients.
• Comments 3.1.3: The nature of the controversy surrounding endocrine therapy should be more explicitly stated—it is not just about “whether it is effective,” but more importantly “for whom it is effective.” This relates to the heterogeneity of ER expression, ER signaling pathway activity (beyond mere protein expression), and interactions with other pathways (e.g., growth factor receptor pathways). Concepts such as “ER signal dilution” or “functional ER loss” in ER-low tumors could be briefly introduced.
• Response 3.1.3: thank you for your comment. We have briefly described the concept that you mentioned and provided some literature references about ER loss.

Comments 3.2: Transitions within and between certain paragraphs are not smooth, making the text read like a simple listing of research findings without sufficient integration and analysis:

• Comments 3.2.1: In Part 3, Treatment Strategies, the discussion on endocrine therapy (ET) is central, but the conclusions are somewhat vague. It is suggested to add a summary paragraph at the end of Section 3.1, clearly stating: Current evidence indicates that for the vast majority of ER-low patients (especially those who are PR-negative), ET offers limited and uncertain benefit; however, for patients with ER expression close to 10% (e.g., 6–10%), PR-positive status, or those who do not achieve pCR after chemotherapy, there may still be some benefit, and thus ET should not be automatically excluded. This would better align with the later discussion on the need for AI-based precise stratification.
• Response 3.2.1: Thank you for your comment. We have added the paragraph that you suggested at the end of section 3.1.
• Comments 3.2.2: Part 4 details several AI platforms, while Part 5 mentions "multimodal AI." It is recommended that the beginning of Part 5 clearly explains that multimodal AI represents the next-generation technology beyond the single-task (e.g., ER quantification) AI models described in Part 4. By integrating multidimensional data such as histology, genomics, and imaging, it holds promise for more precise subtyping and treatment prediction for ER-low patients, thereby providing a natural transition.
• Response 3.2.2: Thank you for your comment. We have entered the sentences that you suggested.

Comments 4: Formatting and References: Some citation formats need to be unified, and it should be confirmed that all references correspond to in-text citations. It is advised to carefully check the formatting of in-text citations and the reference list to ensure compliance with journal requirements.

Response 4: Thank you. We have checked the references and the text and corrected what was not aligned, as well as the formatting.

 

Comments 5: Suggested relevant literature to cite: PMCID: PMC12221439, PMCID: PMC12637072, PMCID: PMC12536554, PMCID: PMC11075339, PMCID: PMC10799417, PMCID: PMC10304265.

Response 5: Thank you.  We have added the references that you have indicated.

Reviewer 3 Report

Comments and Suggestions for Authors

• Section 3.1 should include a paragraph summarizing when endocrine therapy might be reasonable rather than listing multiple studies sequentially.
• The AI section should better link AI-based ER quantification and multimodal models to practical clinical decision-making, particularly how AI could resolve uncertainty in ER-low classification and treatment selection.
• There are no figures, include the figures.

Author Response

Response: Thank you very much for your comments. We believe that your suggestions helped us to improve our work. We have modified the manuscript according to your comments and requests. You can find the changes that we have done in our paper written in red.

 

Comments 1: Section 3.1 should include a paragraph summarizing when endocrine therapy might be reasonable rather than listing multiple studies sequentially.

Response 1: Thank you for your comment. We have included a paragraph summarizing when endocrine therapy might be considered at the end of this section.

Comments 2: The AI section should better link AI-based ER quantification and multimodal models to practical clinical decision-making, particularly how AI could resolve uncertainty in ER-low classification and treatment selection.

Response 2: Thank you for your suggestion. We have added some sentences to better link AI tools in defining ER quantification with practical clinical decision-making.

Comments 3: There are no figures, include the figures.

Response 3: Thank you. We have added a figure and a table.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The author has corrected the manuscript. However, adding a mechanistic figure will improve the paper quality. 

Author Response

Comments 1: The author has corrected the manuscript. However, adding a mechanistic figure will improve the paper quality. 

Response 1: Thank you very much for your suggestions. We are delighted that you appreciated our manuscript and we believe that your comments helped us to improve our work. We have added a mechanistic figure in the conclusion section.

Round 3

Reviewer 3 Report

Comments and Suggestions for Authors

Add reason(s) for inefficacy or efficacy in Table 1.

Author Response

Comments 1: Add reason(s) for inefficacy or efficacy in Table 1.

Response 1: Thank you for your suggestion.  Since currently the reasons underlying the efficacy or inefficacy of endocrine therapy have not been yet clarified, not only in the studies that we have reported but in general in the ER-low population, as we analyze and discuss in our review, we have added in Table 1 the explanation of why ET is considered effective or not in the studies that we have cited according to their results. Otherwise, only hypothesis and speculations might have been added in the table and they were not available for the references cited. We hope you would agree.

Round 4

Reviewer 3 Report

Comments and Suggestions for Authors

There are some typographical mistakes. Please correct them.