When Nature Meets Oncology: Unraveling Herb–Drug Interactions in Cancer Therapy

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

• This review summarizes the available literature on human performance indicators, but it offers only limited new insights. Although it comprehensively synthesizes the evidence, the novelty lies more in the clinical framework than in the original synthesis. The authors should clarify how their work advances this field beyond existing reviews (e.g., Chan et al., 2023; Duan et al., 2025).
• Many interactions are described based on preclinical or small-scale studies. Texts sometimes treat these findings as clinically significant without sufficient caution. A clear hierarchy of evidence (case report, clinical trial, meta-analysis) would enhance discussion and help clinicians assess reliability.
• Suggestion: Add a graphic or table that categorizes interactions by strength of evidence (strong clinical, limited clinical, preclinical only).
• While emphasizing risks, the section on the therapeutic potential of natural products is relatively brief and less critical. A more balanced assessment is needed. For example, clinical trials of ginger for treating viral hepatitis (CINV) or ginseng for treating fatigue should be compared with their potential pharmacokinetic risks.
• Tables 1 and 2 are useful, but they could be explained more concisely and visually. Adding a diagram illustrating the interaction pathways between PK and PD would improve reader understanding.
• The risk classification (Table 2) is a valuable contribution, but the methodology for determining risk levels is not described. Were these categories based on evidence or expert opinion?
• The section on natural products and immune checkpoint inhibitors is interesting, but mostly speculative. The authors should clarify the limitations of the current evidence (mostly preclinical and mechanistic) to avoid overestimating the clinical effects.
• The practice recommendations (proactive communication, guideline implementation, and stratification) are valuable. However, the authors could expand on providing more specific guidance on "do not use" versus "use with caution" for high-risk herbs (e.g., St. John's wort, green tea extract, and garlic supplements).
• Abstract: Consider explicitly stating that most interactions lack strong clinical evidence, to set realistic expectations.
• The conclusion adequately highlights evidence gaps. Adding a stronger research agenda (such as the need for pharmacoepidemiological studies, integration of real-world electronic health records, and computer-based predictive models) would enhance the impact.
• Replace informal phrases like “step on the gas pedal” with more scientific terminology.
• Ensure consistent citation style (currently a mix of numerical and narrative forms).
• Abbreviations such as PK, PD, ADME should be defined once and used consistently.

Author Response

Response to the reviewer 1 comments:

 

We appreciate the through review. Please find our responses here as well as changes in Manuscript highlighted in RED text.

 

Comment 1: This review summarizes the available literature on human performance indicators, but it offers only limited new insights. Although it comprehensively synthesizes the evidence, the novelty lies more in the clinical framework than in the original synthesis. The authors should clarify how their work advances this field beyond existing reviews (e.g., Chan et al., 2023; Duan et al., 2025).

 

Response 1: We truly appreciate this thoughtful and constructive feedback. It’s a fair point that we needed to make the distinct contribution of our work more explicit. While earlier reviews, such as Chan et al. (2023) and Duan et al. (2025), have done an excellent job of outlining the underlying mechanisms, our intention was to take that conversation a step further. What we’ve added is a structured clinical risk-stratification framework that weighs the strength of the evidence, connects these interactions directly to modern treatment modalities like TKIs and ICIs, and frames the discussion around practical decisions oncologists face every day.

In response to this comment, we’ve made two targeted changes to the manuscript. First, we inserted a new paragraph at the end of the Introduction (Section 1.4) (Line 89-99) that clearly situates our work in the context of existing literature and clarifies its clinical decision-support focus. Second, we added a new section 10- Knowledge gaps and future perspectives (567-571) to emphasize how this approach helps clinicians move from theoretical concepts to more actionable guidance at the bedside.

Comment 2: Many interactions are described based on preclinical or small-scale studies. Texts sometimes treat these findings as clinically significant without sufficient caution. A clear hierarchy of evidence (case report, clinical trial, meta-analysis) would enhance discussion and help clinicians assess reliability.

 

Response 2: We agree that the distinction between preclinical and clinically validated evidence must be made explicit to avoid overinterpretation. We have added language that clarifies the level of evidence associated with each type of interaction and sets a more cautious interpretive frame. (Lines 22-25, 168-172, 198-210, 444-448)

 

Comment 3: Suggestion: Add a graphic or table that categorizes interactions by strength of evidence (strong clinical, limited clinical, preclinical only).

 

Response 3: We appreciate the suggestion. We have added column focusing on strength of evidence in table 3 (Previously Table 2).

 

Comment 4: While emphasizing risks, the section on the therapeutic potential of natural products is relatively brief and less critical. A more balanced assessment is needed. For example, clinical trials of ginger for treating viral hepatitis (CINV) or ginseng for treating fatigue should be compared with their potential pharmacokinetic risks.

 

Response 4: We agree and have expanded the discussion on therapeutic benefits, highlighting controlled trial data and placing it alongside potential risks to provide a balanced narrative.

Lines 290-298. New table added Table 2

 

Comment 5: Tables 1 and 2 are useful, but they could be explained more concisely and visually. Adding a diagram illustrating the interaction pathways between PK and PD would improve reader understanding.

 

Response 5: We agree that a visual schematic would improve clarity. We have created a simple PK–PD pathway diagram. Figure 1

 

Comment 6: The risk classification (Table 2) is a valuable contribution, but the methodology for determining risk levels is not described. Were these categories based on evidence or expert opinion?

 

Response 6: We appreciate this observation. We have clarified our methodological approach to risk stratification. Updated lines 507-511

 

Comment 7: The section on natural products and immune checkpoint inhibitors is interesting, but mostly speculative. The authors should clarify the limitations of the current evidence (mostly preclinical and mechanistic) to avoid overestimating the clinical effects.

 

Response 7: We fully agree and have tempered the language to clearly state the speculative nature of these findings. Lines 467-470

 

Comment 8: The practice recommendations (proactive communication, guideline implementation, and stratification) are valuable. However, the authors could expand on providing more specific guidance on "do not use" versus "use with caution" for high-risk herbs (e.g., St. John's wort, green tea extract, and garlic supplements).

 

Response 8: We agree and have made Table 2 more prescriptive to enhance clinical applicability. New column added: “Clinical Action”

 

Comment 9: Abstract: Consider explicitly stating that most interactions lack strong clinical evidence, to set realistic expectations.

 

Response 9: We agree with the reviewer and We have modified the abstract accordingly. Line 30-32

 

Comment 10: The conclusion adequately highlights evidence gaps. Adding a stronger research agenda (such as the need for pharmacoepidemiological studies, integration of real-world electronic health records, and computer-based predictive models) would enhance the impact.

Response 10: We agree and have added new section 10 to address Knowledge gaps and future perspectives. Lines 567-586

Comment 11: Replace informal phrases like “step on the gas pedal” with more scientific terminology.

 

Response 11: Thank you for catching that. We have update the language to more scientific terminology.

 

Comment 12: Ensure consistent citation style (currently a mix of numerical and narrative forms).

 

Response 12: We have updated the citation style.

 

Comment 13: Abbreviations such as PK, PD, ADME should be defined once and used consistently.

 

Response 13: We agree with the comment and updated abbreviations use and consistency.

 

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript aimed to evaluate clinical evidence for interactions between natural products (NPs) and cancer therapies. The authors concluded that NPs pose a significant patient safety risk due to their bioactive nature and call for improved clinician-patient communication.

There are some comments:

• It offers superficial, speculative treatment of emerging areas like immunotherapy interactions, lacking concrete clinical data.

• The text presents conflicting stances, particularly on antioxidants and curcumin, confusing readers without providing clear clinical guidance.

• The risk-stratification table 2 ignores critical variables like dosage and formulation, rendering its "use with caution" advice vague and unactionable.

•  NPs are treated as monoliths, failing to distinguish between dietary use and high-dose supplements, which limits practical applicability.

•  The discussion of NPs' therapeutic benefits is isolated from the risk management framework, creating an overall alarmist tone that may deter the use of proven, low-risk supportive therapies.

Author Response

This manuscript aimed to evaluate clinical evidence for interactions between natural products (NPs) and cancer therapies. The authors concluded that NPs pose a significant patient safety risk due to their bioactive nature and call for improved clinician-patient communication.

Response: Thank you so much for the thoughtful review and comments. Please find below for point to point responses and associated manuscript changes are highlighted in blue text.

Comment 1: It offers superficial, speculative treatment of emerging areas like immunotherapy interactions, lacking concrete clinical data.

Response 1: We appreciate this observation. The initial version briefly discussed immunotherapy interactions primarily from a mechanistic standpoint. We have now expanded this section to include a clearer delineation between preclinical findings, early translational studies, and the current evidence gap in clinical data. To add substance, we integrated newly published studies examining the influence of phytochemicals on PD-L1 modulation and gut microbiome–immunotherapy response correlation. We emphasize that, while clinical validation is still limited, these preclinical insights are valuable for hypothesis generation rather than immediate clinical translation. This refinement aligns the tone with evidence strength and avoids overinterpretation. Section 7.2 is updated Lines 457-465

Comment 2: The text presents conflicting stances, particularly on antioxidants and curcumin, confusing readers without providing clear clinical guidance.

Response 2a: We thank the reviewer for pointing this out. We agree that the evidence surrounding antioxidants and curcumin is inherently conflicting, and we intentionally highlighted this ambiguity in Section 4.2 to reflect the current state of the literature. The contradictory findings arise from heterogeneous preclinical data and the absence of adequately powered randomized clinical trials, which limits the ability to offer definitive guidance.

To improve clarity for readers, we have revised this section to make the take-home clinical message more explicit: while mechanistic and early clinical studies suggest both synergistic and antagonistic effects, the lack of robust clinical data precludes routine use of high-dose curcumin or antioxidant supplements during active treatment. We now emphasize that clinicians should approach these agents with caution, avoiding their use in settings where oxidative mechanisms are central to therapeutic efficacy (e.g., platinum chemotherapy, radiation). Lines 360-364

Response 2b: The mention of antioxidants in our manuscript arises primarily from the pharmacodynamic antagonism section (Section 3.2), where we summarize concerns about ROS neutralization by high-dose antioxidant supplements. We have now clarified the clinical implication of these findings — emphasizing that these theoretical risks translate into a recommendation to avoid high-dose antioxidant use during active therapy. Lines 212-216

Comment 3: The risk-stratification table 2 ignores critical variables like dosage and formulation, rendering its "use with caution" advice vague and unactionable.

Response 4: We agree with reviewer comment and we have updated Table footer for Table 3 (Previously Table 2). Line 512-517

Comment 4: NPs are treated as monoliths, failing to distinguish between dietary use and high-dose supplements, which limits practical applicability.

Response 5: We appreciate this important point. We have now added a clarifying subsection that differentiates dietary exposure, nutraceutical doses, and pharmacologic formulations. This contextualization provides clinicians with more pragmatic guidance and prevents overgeneralization that might discourage safe dietary practices. Added a new subsection 1.5 to address above. Lines 103-109

Comment 6:  The discussion of NPs' therapeutic benefits is isolated from the risk management framework, creating an overall alarmist tone that may deter the use of proven, low-risk supportive therapies.

Response 6: We fully concur that the manuscript should maintain balance between caution and therapeutic opportunity. We have therefore revised Section 4 and Section 8 to better integrate the discussion of potential benefits with the risk-management framework. The conclusion now explicitly states that risk communication should empower informed decision-making, not discourage evidence-based supportive use.

Reviewer 3 Report

Comments and Suggestions for Authors

The review addresses interactions between natural products/supplements and oncologic therapies (chemotherapy, tyrosine kinase inhibitors – TKI, immune checkpoint inhibitors – ICI), with a focus on pharmacokinetic/pharmacodynamic (PK/PD) mechanisms, potential symptomatic benefits, and risks of interaction. The topic is clinically relevant, but the current version requires methodological grounding and a clear hierarchy of evidence.

Comments:

1. I recommend completing the "Materials and Methods" section in accordance with PRISMA, PRISMA-ScR, or PRISMA-S guidelines, by including the following essential elements: databases consulted, search period, exact search strings used, inclusion and exclusion criteria, dual independent screening, data extraction process, and assessment of study quality and risk of bias. Additionally, the inclusion of a PRISMA flow diagram and a detailed table of search strategies in the supplementary material is advised.
2. A clear separation between preclinical and clinical studies is suggested, with further subdivision of the clinical studies into pharmacokinetic (PK), randomized controlled trials (RCTs), and observational designs. Including a summary table that indicates the level of evidence and strength of recommendation (A/B/C), categorized by therapeutic class (e.g., cytotoxic agents, TKI, ICI), would be appropriate.
3. I recommend aligning table content with the main text by ensuring accurate referencing and correcting scientific names (e.g., Curcuma longa, Hangeshashinto). Reclassification of specific examples is also advised (e.g., sirolimus should be placed under "targeted/oral agents or other classes"). Controversial areas (e.g., antioxidants; dose/timing/class-related issues) should be explicitly marked.
4. A strict delineation between preclinical and clinical data is necessary, along with a moderation of translational conclusions. Adding a dedicated section on knowledge gaps and ongoing clinical trials, as well as an explicit warning regarding the risks of self-supplementation during ICI therapy, is recommended.
5. Inclusion of a risk stratification algorithm for orally administered supplements—considering factors such as CYP3A4/P-gp metabolism, narrow therapeutic index, and QT prolongation potential—is advised. A checklist and validated resources for clinicians would also be useful.
6. Interindividual variability in drug transporters and CYP polymorphisms should be addressed, with practical examples of clinical management (e.g., ECG monitoring for QT risk) provided.
7. "Completely avoid" type recommendations should be explicitly supported by clinical guidelines; otherwise, they should be rephrased in a more cautious and non-absolute manner.
8. Writing and Formatting Issues

• Citations should be placed before punctuation marks (e.g., “…critical [36].”).
• Redundant punctuation and hyphenation artifacts (e.g., “re-search”, “com-munication”) should be removed.
• Tables should be standardized in terms of headers, alignment, and referencing. Additional columns such as “Dose/formulation (if standardized)” and “Key safety notes” are recommended.
• Binomial names should be italicized with correct capitalization (e.g., Curcuma longa, Panax ginseng), while higher-level taxa (e.g., Rutaceae) should remain unitalicized. Botanical spelling errors should be corrected.
• Unnecessary spaces (double spaces, spaces before punctuation, spaces around citations, trailing spaces) should be removed. Use a single space after periods and a non-breaking space between numbers and units (e.g., 10 mg, 37 °C).

Author Response

We sincerely thank Reviewer for the thorough and constructive evaluation of our manuscript. The reviewer’s insightful recommendations have significantly strengthened both the methodological rigor and the clinical applicability of our review. Kindly find response below and associated manuscript changes are highlighted in green text.

Comment 1: I recommend completing the "Materials and Methods" section in accordance with PRISMA, PRISMA-ScR, or PRISMA-S guidelines, by including the following essential elements: databases consulted, search period, exact search strings used, inclusion and exclusion criteria, dual independent screening, data extraction process, and assessment of study quality and risk of bias. Additionally, the inclusion of a PRISMA flow diagram and a detailed table of search strategies in the supplementary material is advised.

Response 1: We agree and have now added a clearly defined Materials and Methods section describing our literature search strategy, eligibility criteria, data abstraction process, and quality appraisal in alignment with PRISMA-ScR guidelines. Inserted new section labelled Materials and Methods. Lines 111-129

Comment 2: A clear separation between preclinical and clinical studies is suggested, with further subdivision of the clinical studies into pharmacokinetic (PK), randomized controlled trials (RCTs), and observational designs. Including a summary table that indicates the level of evidence and strength of recommendation (A/B/C), categorized by therapeutic class (e.g., cytotoxic agents, TKI, ICI), would be appropriate.

Response 2: Thank you for the recommendation. Table 3 is updated with two new columns labelled Clinical Actions and Evidence level.

Comment 3: I recommend aligning table content with the main text by ensuring accurate referencing and correcting scientific names (e.g., Curcuma longa, Hangeshashinto). Reclassification of specific examples is also advised (e.g., sirolimus should be placed under "targeted/oral agents or other classes"). Controversial areas (e.g., antioxidants; dose/timing/class-related issues) should be explicitly marked.

Response 3: Thank you for carefully reviewing this section and for your helpful suggestions. We have corrected all botanical and fungal scientific names to italic format throughout the manuscript. Sirolimus has been reclassified under the “Targeted/Oral Therapy” category in Table 3 (previously Table 2) to better reflect its pharmacologic classification. We also updated Table 3 by adding a new “Clinical Action” column and revised the footnote to clarify that “Use with caution” recommendations reflect both the dosage typically required to produce clinically relevant interactions and the presence of controversial or dose-dependent evidence.

 

Comment 4: A strict delineation between preclinical and clinical data is necessary, along with a moderation of translational conclusions. Adding a dedicated section on knowledge gaps and ongoing clinical trials, as well as an explicit warning regarding the risks of self-supplementation during ICI therapy, is recommended.

Response 4: Thank you for this thoughtful observation. We have revised the manuscript to temper over-interpretive statements, added a new Section 10: Knowledge Gaps and Future Perspectives, and inserted an explicit caution against unsupervised supplement use during immunotherapy. These changes aim to provide a more balanced interpretation of the evidence and strengthen the clinical applicability of our conclusions. Lines 566-586

Comment 5: Inclusion of a risk stratification algorithm for orally administered supplements—considering factors such as CYP3A4/P-gp metabolism, narrow therapeutic index, and QT prolongation potential—is advised. A checklist and validated resources for clinicians would also be useful.

Response 5: A concise, clinically oriented algorithm (Figure 2) have been added to guide clinicians in assessing oral supplement risks by metabolic pathway and toxicity profile.

Comment 6: Interindividual variability in drug transporters and CYP polymorphisms should be addressed, with practical examples of clinical management (e.g., ECG monitoring for QT risk) provided.

Response 6: We agree. A paragraph has been added in Section 8.3 highlighting CYP3A5/UGT1A1 polymorphisms and practical management examples, including ECG monitoring for QT risk. Lines 519-525

Comment 7: "Completely avoid" type recommendations should be explicitly supported by clinical guidelines; otherwise, they should be rephrased in a more cautious and non-absolute manner.

Response 7: We agree with the reviewer comment- We have updated Table 3- Clinical action tab.

 

Writing and Formatting Issues

Citations should be placed before punctuation marks (e.g., “…critical [36].”).

Redundant punctuation and hyphenation artifacts (e.g., “re-search”, “com-munication”) should be removed.

Tables should be standardized in terms of headers, alignment, and referencing. Additional columns such as “Dose/formulation (if standardized)” and “Key safety notes” are recommended.

Binomial names should be italicized with correct capitalization (e.g., Curcuma longa, Panax ginseng), while higher-level taxa (e.g., Rutaceae) should remain unitalicized. Botanical spelling errors should be corrected.

Unnecessary spaces (double spaces, spaces before punctuation, spaces around citations, trailing spaces) should be removed. Use a single space after periods and a non-breaking space between numbers and units (e.g., 10 mg, 37 °C).

 

Response: We have updated writing and formatting issues as suggested.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

1. While the manuscript now cites recent and relevant literature up to 2025, some older references (such as the BMJ 2002 on St. John's Wort) could be supplemented with more recent pharmacovigilance data or meta-analyses to maintain current trends.

2. Abbreviations such as PK, PD, and ADME have been defined, but their consistent use should be checked for the first time in all sections.

3. Ensure strict adherence to IJMS formatting of citations and references (spaces, punctuation, and capitalization).

4. Given the strong visual component of the revised manuscript, the addition of a concise graphic summary summarizing the risk stratification model would enhance accessibility and interactivity.

Author Response

We sincerely appreciate the reviewers’ time, insightful comments, and valuable suggestions. Their feedback has been instrumental in improving the quality and precision of our work.

1: While the manuscript now cites recent and relevant literature up to 2025, some older references (such as the BMJ 2002 on St. John's Wort) could be supplemented with more recent pharmacovigilance data or meta-analyses to maintain current trends.

Response 1: We agree and have supplemented older citations with recent pharmacovigilance data and meta-analyses reflecting current evidence. Specifically, we have retained the landmark BMJ 2002 citation for historical context while adding updated references (Fasinu et al., Front Oncol 2019; Lam et al., Phytomedicine 2022) that integrate real-world and pharmacoepidemiologic data. This ensures the discussion reflects both foundational and contemporary perspectives on herb–drug interactions in oncology. Also added two additional  recent references in section 3.1.1. Enzyme induction to supplement older reference.

2. Abbreviations such as PK, PD, and ADME have been defined, but their consistent use should be checked for the first time in all sections.

Response 2: Thank you for noting this. We have reviewed the full manuscript to ensure that abbreviations such as pharmacokinetic (PK), pharmacodynamic (PD), and absorption, distribution, metabolism, and excretion (ADME) are consistently defined at their first appearance within each major section and used uniformly thereafter.

3. Ensure strict adherence to IJMS formatting of citations and references (spaces, punctuation, and capitalization).

Response 3: We appreciate this reminder and have carefully standardized all in-text citations and reference entries according to the IJMS author guidelines. Adjustments included proper use of semicolons for multiple citations, consistent capitalization of journal names, and correction of spacing and punctuation inconsistencies.

4. Given the strong visual component of the revised manuscript, the addition of a concise graphic summary summarizing the risk stratification model would enhance accessibility and interactivity.

We appreciate this constructive suggestion and have created a concise graphical summary and uploaded as a graphical abstract. It visually synthesizes the proposed risk-stratification framework. The figure highlights clinical risk tiers (High, Moderate, Low) and aligns them with primary mechanisms (CYP3A4 modulation, PD antagonism, or supportive benefit). This enhances the accessibility of key clinical takeaways for readers.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors answered all comments. The manuscript is improved and accepted for publication

Author Response

We sincerely thank the reviewer for their positive feedback and support. We appreciate the time and effort dedicated to reviewing our manuscript.