Next Article in Journal
Clinical Validation of the Belay Ascent™ Test to Report on Chromosomal Arm-Level Aneuploidy and Gene-Level Copy Number Variants in Cerebrospinal Fluid Using Low-Pass Whole-Genome Sequencing
Previous Article in Journal
How Laboratory Innovations Are Shaping the Future of Multiple Myeloma Care
 
 
Review
Peer-Review Record

Myosteatosis and Sarcopenic Obesity in Men Receiving Androgen Deprivation Therapy for Prostate Cancer: Rationale for Mechanism-Driven Multimodal Intervention

Cancers 2026, 18(8), 1276; https://doi.org/10.3390/cancers18081276
by Nagi B. Kumar 1,2,*, Nathan Parker 3, Jingsong Zhang 2, Julio Pow-Sang 2, Jong Y. Park 1,2 and Michael J. Schell 4
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3: Anonymous
Cancers 2026, 18(8), 1276; https://doi.org/10.3390/cancers18081276
Submission received: 17 March 2026 / Revised: 14 April 2026 / Accepted: 14 April 2026 / Published: 17 April 2026
(This article belongs to the Section Cancer Survivorship and Quality of Life)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Authors submitted a narrative review that addresses a clinically important survivorship toxicity of ADT. Overall, I believe the topic is highly relevant and timely. However, I have some suggestions/comments as follows:

1) Abstract: I suggest adding 1-2 phrases that explicitly state the review is narrative and that myosteatosis endpoints are rarely included in ADT trials.

2) Introduction: I recommend tightening the claims about oncologic outcomes (biochemical recurrence, CRPC, mortality) and they should specify whether evidence is ADT-specific vs broader PCa survivorship literature. Also, I suggest adding a concise definition box early (or a small table) distinguishing sarcopenia (mass/function), myosteatosis (quality/radiodensity/intramuscular fat), sarcopenic obesity (coexistence), and which imaging metrics you recommend (e.g., SMI, muscle radiodensity, intermuscular fat).   

3) Materials and Methods: I recommend polishing and clarifying to ensure all search terms are formatted consistently (quotation marks and AND/OR logic), fix minor grammar issues (“studies were excluded…” begins lowercase) and clarify what you mean by “related databases” and whether any trial registries were screened. 

4) Mechanisms and conceptual framework: I think the conceptual framework is weakened by the unfinished figures. Please finalize Figure 1 and ensure it is visually coherent with a concise legend and clear directional arrows. I also recommend reducing repetition between 3.1 and 3.2 and adding one short “testable predictions” paragraph (e.g., “ADT → early decline in muscle radiodensity precedes overt lean mass loss,” etc.).

5) Current strategies: I recommend you add a summary table listing the key ADT interventions (duration, sample size, components, outcomes measured, and whether any muscle quality measure was included). 

6) Figure 2: It appears as raw text rather than a finalized figure with a proper legend, formatting, and consistent biomarker list.  Please revise this into a polished schematic and ensure the biomarker panel is realistic and prioritized (e.g., NF-κB activity is not routinely measured clinically; specify whether it is a research endpoint). 

7) Conclusions and Future directions: I recommend adding a concise “minimum core outcome set” (e.g., muscle radiodensity/myosteatosis, strength, functional test, metabolic biomarkers and patient-reported outcomes).

Author Response

We thank the reviewers for a thorough review of our manuscript and their valuable critique and recommendations to improve the work.   We are pleased to resubmit a revised manuscript entitled “Myosteatosis and Sarcopenic Obesity in Men Receiving Androgen Deprivation Therapy for Prostate Cancer: Rationale for Mechanism-Driven Multimodal Intervention” for consideration for publication in Cancer. This manuscript presents a structured, translational narrative review with an integrated conceptual framework.

Please see our specific responses (R ) to each of the reviewer’s concerns/critique (C ).  All changes appear in red in the tracked version of the manuscript.

Reviewer 1:

Authors submitted a narrative review that addresses a clinically important survivorship toxicity of ADT. Overall, I believe the topic is highly relevant and timely. However, I have some suggestions/comments as follows:

C1. Abstract: I suggest adding 1-2 phrases that explicitly state the review is narrative and that myosteatosis endpoints are rarely included in ADT trials.

R1.  We thank the reviewer for this helpful suggestion. We have revised the Abstract to explicitly state that this is a structured narrative review and to highlight that myosteatosis-related endpoints are infrequently incorporated into clinical trials of men receiving ADT. This clarification has been added to improve transparency and contextual framing.

C2. Introduction: I recommend tightening the claims about oncologic outcomes (biochemical recurrence, CRPC, mortality) and they should specify whether evidence is ADT-specific vs broader PCa survivorship literature. Also, I suggest adding a concise definition box early (or a small table) distinguishing sarcopenia (mass/function), myosteatosis (quality/radiodensity/intramuscular fat), sarcopenic obesity (coexistence), and which imaging metrics you recommend (e.g., SMI, muscle radiodensity, intermuscular fat).   

R2. We have revised the Introduction to more clearly distinguish evidence derived from ADT-specific populations versus broader prostate cancer survivorship literature when discussing oncologic outcomes. In addition, we have added a concise definition table (Table 1) early in the manuscript distinguishing sarcopenia, myosteatosis, and sarcopenic obesity, along with commonly used imaging metrics (e.g., skeletal muscle index, muscle radiodensity, and intermuscular adipose tissue).

C3. Materials and Methods: I recommend polishing and clarifying to ensure all search terms are formatted consistently (quotation marks and AND/OR logic), fix minor grammar issues (“studies were excluded…” begins lowercase) and clarify what you mean by “related databases” and whether any trial registries were screened. 

R3. We have revised the Methods section to ensure consistent formatting of search terms, corrected minor grammatical issues, and clarified that “related databases” refers to citation tracking and reference list screening. We have also clarified that trial registries were not systematically searched, consistent with the narrative review design, and this limitation is now acknowledged.

C4. Mechanisms and conceptual framework: I think the conceptual framework is weakened by the unfinished figures. Please finalize Figure 1 and ensure it is visually coherent with a concise legend and clear directional arrows. I also recommend reducing repetition between 3.1 and 3.2 and adding one short “testable predictions” paragraph (e.g., “ADT → early decline in muscle radiodensity precedes overt lean mass loss,” etc.).

R4. We agree and have substantially revised this section. Figure 2 (previously Figure 1) has been finalized with clear directional relationships and incorporation of key mechanistic pathways and a concise legend. Redundant text between Sections 3.1 and 3.2 has been reduced. We have also added a brief “testable predictions” paragraph to strengthen the conceptual framework and translational relevance.

C5. Current strategies: I recommend you add a summary table listing the key ADT interventions (duration, sample size, components, outcomes measured, and whether any muscle quality measure was included). 

R5.  Please see addition of Table 2.

C6. Figure 2: It appears as raw text rather than a finalized figure with a proper legend, formatting, and consistent biomarker list.  Please revise this into a polished schematic and ensure the biomarker panel is realistic and prioritized (e.g., NF-κB activity is not routinely measured clinically; specify whether it is a research endpoint). 

R6. Figure 3 (previously Figure 2) has now been revised with a legend and a consistent biomarker list with a realistic biomarker panel that are clinically feasible or can be obtained as part of standard of care imaging in this patient population (CT abdomen) or as a research biomarker.

C7. Conclusions and Future directions: I recommend adding a concise “minimum core outcome set” (e.g., muscle radiodensity/myosteatosis, strength, functional test, metabolic biomarkers and patient-reported outcomes).

R7. We have added concise outcomes set and revised the conclusion paragraph as recommended.

 

Thank you

Reviewer 2 Report

Comments and Suggestions for Authors

This narrative review examines the development of sarcopenic obesity in men receiving androgen deprivation therapy (ADT) for prostate cancer, emphasizing the underrecognized role of myosteatosis as a key determinant of muscle quality and metabolic dysfunction. The authors synthesize current evidence on exercise and nutrition-based interventions.

Issues to address:

-The manuscript is presented as a “structured narrative review,” but lacks key elements expected even for high-quality narrative reviews, such as, clear search strategy reproducibility, study selection flow, or systematic assessment framework...

-The proposed central role of myosteatosis as a driver, and not just a marker, of sarcopenic obesity is conceptually interesting but insufficiently supported by causal human data. Indeed, much of the mechanistic discussion relies on preclinical or associative evidence, so the authors should avoid implying causality.

-Studies are summarized descriptively without discussing risk of bias, heterogeneity, or strength of evidence.

-The proposed “multimodal bundled intervention” (exercise, diet, omega-3s, phytochemicals…) is broad and not evidence-ranked, mixing well-supported (exercise) with highly preliminary strategies (phytochemicals, fasting)

Author Response

We thank the reviewers for a thorough review of our manuscript and their valuable critique and recommendations to improve the work.   We are pleased to resubmit a revised manuscript entitled “Myosteatosis and Sarcopenic Obesity in Men Receiving Androgen Deprivation Therapy for Prostate Cancer: Rationale for Mechanism-Driven Multimodal Intervention” for consideration for publication in Cancer. This manuscript presents a structured, translational narrative review with an integrated conceptual framework.

Please see our specific responses (R ) to each of the reviewer’s concerns/critique (C ).  All changes appear in red in the tracked version of the manuscript.

Reviewer 2:

Issues to address:

C1. The manuscript is presented as a “structured narrative review,” but lacks key elements expected even for high-quality narrative reviews, such as, clear search strategy reproducibility, study selection flow, or systematic assessment framework...

R1. We agree that additional transparency is needed and have now expanded the Methods section to include a detailed and reproducible search strategy, clarified study selection criteria, and added a flow diagram summarizing the identification and inclusion of studies. We have also incorporated a structured approach to evidence synthesis. Please see our revisions in the methods section, in addition to Table 1.

C2. The proposed central role of myosteatosis as a driver, and not just a marker, of sarcopenic obesity is conceptually interesting but insufficiently supported by causal human data. Indeed, much of the mechanistic discussion relies on preclinical or associative evidence, so the authors should avoid implying causality.

R2. We thank the reviewer for this important point. We have revised the manuscript to avoid implying causality and now consistently describe myosteatosis as a clinically relevant marker and a potential mechanistic mediator, supported primarily by associative human data and complementary preclinical evidence. We have clarified where evidence is derived from preclinical versus human studies and tempered language accordingly.

C3. Studies are summarized descriptively without discussing risk of bias, heterogeneity, or strength of evidence.

R3. We agree and have added discussion of study heterogeneity, including variability in imaging methods, thresholds, populations, and study design. We have also incorporated Table 2 demonstrating the limitations of current studies and the limitations and risk of bias within the narrative synthesis.

C4. The proposed “multimodal bundled intervention” (exercise, diet, omega-3s, phytochemicals…) is broad and not evidence-ranked, mixing well-supported (exercise) with highly preliminary strategies (phytochemicals, fasting)

R4.  We have revised the intervention section to more clearly differentiate between evidence-based components (e.g., resistance and aerobic exercise) and emerging or exploratory strategies (e.g., nutritional approaches, time-restricted eating, SARMs etc). The level of supporting evidence is now stated for each component (Table 3) and are framed as hypothesis-generating components requiring further prospective validation.

 

 

Thank you

Reviewer 3 Report

Comments and Suggestions for Authors

< !-- obsidian -->

This narrative review aims to summarize current evidence on myosteatosis and sarcopenic obesity in men receiving androgen deprivation therapy (ADT) for prostate cancer, and to propose a mechanism-based multimodal intervention framework. The topic is clinically relevant and of interest. However, there are important concerns regarding data accuracy, mechanistic coverage, literature completeness, and the quality and presentation of figures and tables. These issues limit the overall strength and contribution of the manuscript. Major revision is therefore recommended.

Major issues:

  1. The epidemiological claim on lines 202–204 that ADT reduces bone mineral density by 8.5% within 6 months (Ref. 35, Daniell et al. 2000) is not supported by the cited source. Daniell reported 2.4% loss in year 1 and 7.6% over 2 years, with a range of 2.5% to 17.0%. The 8.5% at 6 months is not present, and the 17% value reflects an upper bound rather than a typical estimate. These statements should be corrected with accurate attribution.

  2. The mechanistic discussion in Section 3.1 (lines 153–175) is largely limited to NF-κB signaling and inflammatory cytokines. Key catabolic pathways, particularly the TGF-β superfamily myokine axis including myostatin, GDF11, and activins, are not addressed. This omission is significant for a mechanism-focused review and should be corrected.

  3. The manuscript does not consider androgen receptor-mediated mesenchymal stem cell lineage switching. Testosterone withdrawal may shift differentiation from myogenic toward adipogenic pathways, providing a mechanistic basis for concurrent muscle loss and fat infiltration. This concept should be incorporated into Section 3.2 (lines 176–191) and Figure 1.

  4. Hepatic contributions to protein catabolism are not discussed. Beyond direct muscle effects, systemic metabolic changes such as increased hepatic urea production and whole-body protein oxidation may play an important role. This aspect should be included, as it has implications for interpreting intervention strategies.

  5. The reference list contains multiple duplications. Ref. 35 and Ref. 38 are identical; similar duplication occurs for Ref. 22 and 51, Ref. 23 and 52, Ref. 24 and 53, Ref. 33 and 56, Ref. 8 and 57, Ref. 17 and 55, and Ref. 16 and 46. In addition, Ref. 26 is missing the first author’s surname. These issues indicate insufficient reference checking and should be corrected.

  6. Several relevant areas of evidence are not covered, including associations between ADT-related muscle loss and mortality risk, imaging-based changes in muscle quality, and evidence supporting multimodal interventions. Expanding this section would improve completeness and clinical relevance.

  7. The manuscript treats ADT as a homogeneous exposure and does not address differences among treatment regimens. Emerging evidence suggests that different agents may have distinct effects on body composition. This limitation should be acknowledged and discussed.

  8. The discussion of pharmacological strategies in Section 3.4 is limited. Additional therapeutic directions under investigation, including metabolic-targeting agents, ActRII pathway modulation, SARMs, and nutritional supplementation strategies, should be considered to strengthen Section 5.

Minor issues:

  1. The manuscript uses both "PCa" and "PCA" to refer to prostate cancer. This abbreviation should be standardized throughout.

  2. The citation given for Huggins & Hodges (1941) in Ref. 2 appears to correspond to the 2002 reprint in J Urol rather than the original Cancer Research article. The original reference should be carefully verified.

  3. The statement "Patents: Pending" on line 438 is too vague. If this is relevant to the manuscript, it should be disclosed more clearly in the Conflicts of Interest section.

  4. The two tables (Table 1 and Table 2) on the final page are not cited in the main text and do not contain structured tabular content. As they are primarily narrative in form, their presentation as tables is not appropriate and should be reconsidered.

Author Response

We thank the reviewers for a thorough review of our manuscript and their valuable critique and recommendations to improve the work.   We are pleased to resubmit a revised manuscript entitled “Myosteatosis and Sarcopenic Obesity in Men Receiving Androgen Deprivation Therapy for Prostate Cancer: Rationale for Mechanism-Driven Multimodal Intervention” for consideration for publication in Cancer. This manuscript presents a structured, translational narrative review with an integrated conceptual framework.

Please see our specific responses (R ) to each of the reviewer’s concerns/critique (C ).  All changes appear in red in the tracked version of the manuscript.

Reviewer 3:

C1. The epidemiological claim on lines 202–204 that ADT reduces bone mineral density by 8.5% within 6 months (Ref. 35, Daniell et al. 2000) is not supported by the cited source. Daniell reported 2.4% loss in year 1 and 7.6% over 2 years, with a range of 2.5% to 17.0%. The 8.5% at 6 months is not present, and the 17% value reflects an upper bound rather than a typical estimate. These statements should be corrected with accurate attribution.

R1. We thank the reviewer for this careful observation. We have revised the statement to accurately reflect the findings reported by Daniell et al. Specifically, we now report that bone mineral density declined by approximately 2.4% in the first year and 7.6% over two years, with a reported range of 2.5%–17.0%, rather than the previously stated value. The text has been corrected to ensure accurate attribution and interpretation of the cited study.

C2. The mechanistic discussion in Section 3.1 (lines 153–175) is largely limited to NF-κB signaling and inflammatory cytokines. Key catabolic pathways, particularly the TGF-β superfamily myokine axis including myostatin, GDF11, and activins, are not addressed. This omission is significant for a mechanism-focused review and should be corrected.

R2. We appreciate this important suggestion. We have expanded Section 3.1 to include key catabolic pathways beyond NF-κB signaling, specifically the TGF-β superfamily myokine axis, including myostatin, GDF11, and activins. The revised text now highlights their roles in regulating muscle protein degradation, inhibiting myogenesis, and contributing to muscle atrophy and adipogenic infiltration, thereby strengthening the mechanistic and biological framework of this review.

C3. The manuscript does not consider androgen receptor-mediated mesenchymal stem cell lineage switching. Testosterone withdrawal may shift differentiation from myogenic toward adipogenic pathways, providing a mechanistic basis for concurrent muscle loss and fat infiltration. This concept should be incorporated into Section 3.2 (lines 176–191) and Figure 1.

R3. We thank the reviewer for this insightful point. We have incorporated discussion of androgen receptor–mediated mesenchymal stem cell lineage allocation into Section 3.2. The revised section now addresses how testosterone deprivation may shift progenitor cell differentiation from myogenic toward adipogenic pathways, providing a mechanistic basis for concurrent muscle loss and intramuscular fat accumulation. This concept has also been integrated into Figure 2 (Previously Figure 1).

C4. Hepatic contributions to protein catabolism are not discussed. Beyond direct muscle effects, systemic metabolic changes such as increased hepatic urea production and whole-body protein oxidation may play an important role. This aspect should be included, as it has implications for interpreting intervention strategies.

R4. We agree that systemic metabolic adaptations, including hepatic contributions, are highly relevant. We have expanded the manuscript to include discussion of increased hepatic urea production and whole-body protein oxidation as contributors to net protein loss in ADT-treated patients, including implications for nitrogen balance and interpretation of protein metabolism endpoints. This addition provides a more comprehensive systems-level perspective and has implications for interpreting intervention strategies.

C5. The reference list contains multiple duplications. Ref. 35 and Ref. 38 are identical; similar duplication occurs for Ref. 22 and 51, Ref. 23 and 52, Ref. 24 and 53, Ref. 33 and 56, Ref. 8 and 57, Ref. 17 and 55, and Ref. 16 and 46. In addition, Ref. 26 is missing the first author’s surname. These issues indicate insufficient reference checking and should be corrected.

R5. We thank the reviewer for identifying these issues. The reference list has been carefully reviewed and corrected to remove duplicate entries (Refs. 35/38, 22/51, 23/52, 24/53, 33/56, 8/57, 17/55, and 16/46). Missing author information (Ref. 26) has also been corrected. The revised reference list has been thoroughly cross-checked for accuracy and consistency.

C6. Several relevant areas of evidence are not covered, including associations between ADT-related muscle loss and mortality risk, imaging-based changes in muscle quality, and evidence supporting multimodal interventions. Expanding this section would improve completeness and clinical relevance.

R6. We have expanded the relevant sections to include:
(i) evidence linking ADT-associated muscle loss and adverse body composition to mortality and clinical outcomes;
(ii) emerging data on imaging-based measures of muscle quality, including skeletal muscle radiodensity and intramuscular adipose tissue; and
(iii) evidence supporting multimodal intervention strategies combining nutrition, exercise, and metabolic targeting approaches.

These additions enhance the clinical relevance, mechanistic depth, and completeness of the review.

C7. The manuscript treats ADT as a homogeneous exposure and does not address differences among treatment regimens. Emerging evidence suggests that different agents may have distinct effects on body composition. This limitation should be acknowledged and discussed.

R7. We agree with the reviewer that ADT is not a uniform exposure. We have added a discussion acknowledging heterogeneity across ADT regimens, including GnRH agonists, antagonists, and combination therapies, and their potentially differential effects on body composition and metabolic outcomes. This limitation is now explicitly discussed.

C8. The discussion of pharmacological strategies in Section 3.4 is limited. Additional therapeutic directions under investigation, including metabolic-targeting agents, ActRII pathway modulation, SARMs, and nutritional supplementation strategies, should be considered to strengthen Section 5.

R8. We have included additional therapeutic strategies under investigation, including metabolic-targeting agents, modulation of the activin receptor type II (ActRII) pathway, selective androgen receptor modulators (SARMs), and targeted nutritional supplementation approaches. This strengthens the translational and forward-looking aspects of the manuscript.

Minor issues:

 

C9. The manuscript uses both "PCa" and "PCA" to refer to prostate cancer. This abbreviation should be standardized throughout.

R9. The abbreviation has been standardized to “PCa” throughout the manuscript.

C10. The citation given for Huggins & Hodges (1941) in Ref. 2 appears to correspond to the 2002 reprint in J Urol rather than the original Cancer Research article. The original reference should be carefully verified.

R10. The reference has been corrected to cite the original 1941 publication by Huggins and Hodges in Cancer Research, rather than the later reprint.

C11. The statement "Patents: Pending" on line 438 is too vague. If this is relevant to the manuscript, it should be disclosed more clearly in the Conflicts of Interest section.

R11. The statement “Patents: Pending” has been clarified and appropriately expanded within the Conflicts of Interest section to ensure full transparency.

C12. The two tables (Table 1 and Table 2) on the final page are not cited in the main text and do not contain structured tabular content. As they are primarily narrative in form, their presentation as tables is not appropriate and should be reconsidered

R12. We have removed the original narrative Tables 1 and 2 and replaced them with structured tables aligned with journal expectations and explicitly referenced in the main text.

Thank you.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

All comments have been addressed. 

Author Response

We greatly appreciate your valuable feedback and the time you have dedicated to evaluating our work.

Reviewer 2 Report

Comments and Suggestions for Authors

ISSUES HAVE BEEN ADDRESSED

Author Response

We greatly appreciate your valuable feedback and the time you have dedicated to evaluating our work.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have made substantial and commendable revisions addressing all major concerns raised during review. The epidemiological claim regarding ADT-induced bone mineral density loss has been corrected to accurately reflect the original findings reported by Daniell et al. The mechanistic discussion has been significantly expanded to include the TGF-β superfamily myokine axis, specifically myostatin, activins, and ActRII/SMAD2/3 signaling. Androgen receptor–mediated mesenchymal stem cell lineage switching has been incorporated into a newly created section on systemic endocrine and inflammatory drivers, with an updated conceptual framework figure now depicting hormonal/anabolic suppression pathways. The evidence base for clinical outcomes has been strengthened through new structured tables distinguishing myosteatosis, sarcopenia, and sarcopenic obesity, a comprehensive intervention trial summary, and expanded discussion of mortality and imaging-based muscle quality data. The pharmacologic strategies section is substantially strengthened by the addition of emerging approaches including ActRII pathway inhibition, SARMs, and metabolic-targeting agents, with a new evidence strength classification. 

All minor issues have also been resolved. The reference list has been thoroughly revised, eliminating all identified duplicate pairs, and the Huggins & Hodges citation has been corrected to the original 1941 Cancer Research article. The abbreviation "PCa" has been standardized throughout, the patent disclosure has been expanded with full USPTO details, and the original narrative tables have been replaced with structured, text-referenced tables. New additions including a study selection flow diagram and testable hypotheses derived from the conceptual framework further strengthen the manuscript.

However, two minor points would benefit from brief clarification. First, the hepatic contributions to protein catabolism mentioned in the authors' response could not be readily located in the revised text (Comment 4); the authors may wish to confirm its placement or add a brief sentence. Second, Ref. 26 still appears to be missing the first author's surname "Batsis" (Comment 5). These are minor clarifications that can be addressed in a final revision, after which the manuscript would be acceptable for publication.

Author Response

Reviewer 3:

Minor clarifications:

 

C1. The hepatic contribution to protein catabolism mentioned in the author’s response could not be readily located in the revised text (Comment 4): The authors may wish to confirm its placement or add a brief sentence.

R4. We agree that systemic metabolic adaptations, including hepatic contributions, are highly relevant. We have expanded the manuscript to include discussion of increased hepatic urea production and whole-body protein oxidation as contributors to net protein loss in ADT-treated patients, including implications for nitrogen balance and interpretation of protein metabolism endpoints. These additions appear in Section 3.1.

C2. Ref 26 still appears to be missing the first author’s surname “Bastis” (Comment 5).

 

R2. We have now corrected this error- Reference 26.  Batsis JA, Villareal DT. Sarcopenic obesity in older adults: aetiology, epidemiology and treatment strategies. Nat Rev Endocrinol. 2018 Sep;14(9):513-537. doi: 10.1038/s41574-018-0062-9. PMID: 30065268; PMCID: PMC6241236.

 

Back to TopTop