Cancer-Associated Thrombosis in Patients Treated with Immune Checkpoint Inhibitors

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This review entitled “Cancer-Associated Thrombosis in Patients Treated With Immune Checkpoint Inhibitors” addresses a timely and clinically relevant topic, namely the potential link between immune checkpoint inhibitors (ICIs) and cancer-associated thrombosis (CAT), within the context of thromboinflammation. While the manuscript provides a broad overview of existing concepts, it remains largely descriptive and lacks critical integration of clinical evidence, mechanistic specificity, and translational relevance. Substantial revisions are required to strengthen the scientific rigor and clinical impact of the review.

Major comments:

1. Lack of a clear association analysis between ICIs and CAT risk. While the review aims to explore the possible causative relationship between immune checkpoint inhibitors (ICIs) and cancer-associated thrombosis (CAT), the manuscript does not provide a sufficiently structured analysis of clinical evidence linking ICIs to venous thromboembolism (VTE). The authors are encouraged to systematically summarize and discuss reported VTE events in patients receiving ICIs, including whether comparative data exist among different classes of ICIs (e.g., anti–PD-1, anti–PD-L1, anti–CTLA-4). Without such analysis, it remains unclear whether ICIs per se confer an increased thrombotic risk beyond that of cancer or prior chemotherapy.
2. The current design of Figure 1 lacks a clear logical flow. Thrombosis appears to be presented as a parallel component rather than a downstream consequence of ICI-induced immune and inflammatory alterations. Conceptually, thrombosis should be depicted as the final outcome resulting from immune activation, endothelial dysfunction, platelet activation, and coagulation cascade engagement triggered by ICIs. The authors are strongly encouraged to reorganize this figure to better reflect the proposed pathophysiological sequence.
3. Limited discussion of the clinical utility of predictive biomarkers. While several potential predictive biomarkers are mentioned, the manuscript does not adequately address their relevance to clinical practice. The authors should clarify how these biomarkers could be used for risk stratification, early prediction of CAT, or guidance of thromboprophylaxis in patients receiving ICIs. A more critical discussion distinguishing exploratory or associative biomarkers from those with potential clinical applicability is needed.
4. Inadequate discussion of clinical implications and management strategies. Given the increasing use of ICIs across multiple cancer types, the manuscript would benefit from a more in-depth discussion of the clinical implications of a potential association between ICIs and CAT. The authors should elaborate on how current evidence may inform patient monitoring, risk assessment models, and thromboprophylaxis strategies in the immunotherapy setting, while clearly acknowledging the limitations of existing data.

Overall, the manuscript does not clearly articulate the central message or key perspective that the authors intend to convey. As a result, the review lacks a well-defined focus, and the main concepts are not sufficiently highlighted or prioritized throughout the text.

Author Response

We thank the Reviewer for the constructive and insightful evaluation of our manuscript. In response, we have substantially revised the text to better integrate clinical evidence, clarify mechanistic pathways, and strengthen the clinical relevance and focus of the review.

Comment 1:Lack of a clear association analysis between ICIs and CAT risk. While the review aims to explore the possible causative relationship between immune checkpoint inhibitors (ICIs) and cancer-associated thrombosis (CAT), the manuscript does not provide a sufficiently structured analysis of clinical evidence linking ICIs to venous thromboembolism (VTE). The authors are encouraged to systematically summarize and discuss reported VTE events in patients receiving ICIs, including whether comparative data exist among different classes of ICIs (e.g., anti–PD-1, anti–PD-L1, anti–CTLA-4). Without such analysis, it remains unclear whether ICIs per se confer an increased thrombotic risk beyond that of cancer or prior chemotherapy.

Response 1: We thank the Reviewer for this important comment. In response, we have expanded and better structured the analysis of clinical evidence linking immune checkpoint inhibitors to venous thromboembolism. Specifically, we added a more detailed discussion of reported VTE events across different ICI classes (anti–PD-1, anti–PD-L1, and anti–CTLA-4) and across different patient populations. This expanded analysis has been incorporated into the “Understanding CAT” section (lines 120–179). 

Comment 2: The current design of Figure 1 lacks a clear logical flow. Thrombosis appears to be presented as a parallel component rather than a downstream consequence of ICI-induced immune and inflammatory alterations. Conceptually, thrombosis should be depicted as the final outcome resulting from immune activation, endothelial dysfunction, platelet activation, and coagulation cascade engagement triggered by ICIs. The authors are strongly encouraged to reorganize this figure to better reflect the proposed pathophysiological sequence.

Response 2: We thank the Reviewer for this constructive comment. In response, we revised Figure 1 to improve its logical structure and to depict thrombosis as the downstream outcome of immune and inflammatory alterations induced by ICIs. The figure has been reorganised to illustrate a sequential pathway starting from T-cell activation, followed by intracellular inflammatory signalling and cytokine-mediated activation of the innate immune system, leading to endothelial dysfunction, platelet activation, NET formation, and tissue factor release. These events converge on activation of the coagulation cascade, inhibition of fibrinolysis, and ultimately thrombus formation. Accordingly, thrombosis is now presented as the consequence, rather than a parallel event, of upstream pathogenic mechanisms. The figure legend has also been updated to clarify the proposed mechanism (lines 432–450).

Comment 3: Limited discussion of the clinical utility of predictive biomarkers. While several potential predictive biomarkers are mentioned, the manuscript does not adequately address their relevance to clinical practice. The authors should clarify how these biomarkers could be used for risk stratification, early prediction of CAT, or guidance of thromboprophylaxis in patients receiving ICIs. A more critical discussion distinguishing exploratory or associative biomarkers from those with potential clinical applicability is needed.

Response 3: We thank the reviewer for this valuable comment. Thanks to the very useful observation made on the clinical role of predictive biomarkers for thrombotic risk, we have decided to clarify their role in routine clinical practice. We have added a dedicated subsection (Section 5.6, lines 607-622) that critically discusses the clinical applicability of the biomarkers described, clearly distinguishing those with potential short-term clinical relevance from biomarkers that have an exclusively exploratory role. In the Discussion (lines 725-732), we underline that early biomarker changes during immunotherapy may reflect reversible immune activation rather than stable thrombotic risk, limiting the use of static thresholds and supporting a longitudinal, monitoring-oriented interpretation until prospective validation is available.

Comment 4: Inadequate discussion of clinical implications and management strategies. Given the increasing use of ICIs across multiple cancer types, the manuscript would benefit from a more in-depth discussion of the clinical implications of a potential association between ICIs and CAT. The authors should elaborate on how current evidence may inform patient monitoring, risk assessment models, and thromboprophylaxis strategies in the immunotherapy setting, while clearly acknowledging the limitations of existing data.

Response 4: We thank the Reviewer for this constructive and clinically relevant comment. In response, we expanded the Discussion section to better address the clinical implications of a potential association between ICIs and CAT. In particular, we elaborated on patient monitoring strategies, the role and limitations of available risk assessment models (RAMs), and considerations regarding thromboprophylaxis in selected patients treated with ICIs. Throughout this section, we explicitly acknowledge the limitations of the current evidence and the lack of ICI-specific prospective data to guide routine clinical practice (lines 696-698; 710-712; 732-735).

The line numbers referring to the revisions correspond to the Word document uploaded as “Revised Manuscript".

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have reviewed the evidence for cancer-associated thromboembolism (CAT) incidence and risk as developed for various cancers and anti-cancer therapies and then applied this information to assessing the CAT risk for immune checkpoint inhibitor (ICI) therapy.  They conclude that, although there is compelling mechanistic evidence that ICI therapy carries an even greater risk of CAT than conventional cancer therapies, there is as yet an insufficient means of assessing clinical risk with these agents.  A number of suggestions are made for improved assessment using biomarkers and metabolites, many of them not immediately applicable, but the obvious question is whether this review is premature.  Perhaps "Barriers to Assessing" should be put before the present title.  The problem is that ICIs potentiate immune mechanisms that increase coagulability.  This reviewer remains skeptical that thromboembolism cannot be reliably assessed in these patients.  How does the assessment of TE in ICI patients compare to that in DIC and COVID-19?  I think that consideration of these questions would improve the usefulness of the review.  There are many acronyms in this manuscript.  It could benefit from a table of abbreviations.

Author Response

Comment: The authors have reviewed the evidence for cancer-associated thromboembolism (CAT) incidence and risk as developed for various cancers and anti-cancer therapies and then applied this information to assessing the CAT risk for immune checkpoint inhibitor (ICI) therapy.  They conclude that, although there is compelling mechanistic evidence that ICI therapy carries an even greater risk of CAT than conventional cancer therapies, there is as yet an insufficient means of assessing clinical risk with these agents.  A number of suggestions are made for improved assessment using biomarkers and metabolites, many of them not immediately applicable, but the obvious question is whether this review is premature.  Perhaps "Barriers to Assessing" should be put before the present title.  The problem is that ICIs potentiate immune mechanisms that increase coagulability.  This reviewer remains skeptical that thromboembolism cannot be reliably assessed in these patients.  How does the assessment of TE in ICI patients compare to that in DIC and COVID-19?  I think that consideration of these questions would improve the usefulness of the review.  There are many acronyms in this manuscript.  It could benefit from a table of abbreviations.

Response: We thank the Reviewer for this thoughtful comment. The points raised regarding the timeliness of the review and the possible comparison with other thromboinflammatory conditions, such as DIC and COVID-19, were carefully discussed among all authors and are certainly worth consideration. However, we considered that a detailed comparison would be too complex to be adequately addressed within the scope of a narrative review focused on ICI-treated cancer patients and could risk shifting the focus away from the main topic. Finally, in response to the comment on acronyms, we included a dedicated Abbreviations section listing only acronyms, reported as continuous text at the end of the manuscript, immediately before the Conflicts of Interest section, in order to improve readability without interrupting the flow of the main text (lines 789-839).

The line numbers referring to the revisions correspond to the Word document uploaded as “Revised Manuscript".

Reviewer 3 Report

Comments and Suggestions for Authors

Report for “Cancer-Associated Thrombosis in Patients Treated With Immune Checkpoint Inhibitors".

In this manuscript, Ilari et al. provide a narrative review that brings together the current evidence on cancer-associated thrombosis (CAT) in patients who receive immune checkpoint inhibitors (ICIs). It covers the epidemiology of CAT, reviews the strengths and limitations of existing risk assessment models (RAMs) developed for chemotherapy cohorts, suggests possible mechanisms linking ICIs to thrombosis, and surveys new biomarkers that could improve risk stratification and assist with prophylaxis. The manuscript argues for risk models specific to ICIs and approaches driven by biomarkers for prevention and management.

The manuscript is generally well structured, guiding the reader from epidemiology to RAMs, mechanistic biology, biomarkers, and clinical management. The conceptual figure of ICI-related thrombo-inflammatory pathways is useful, and the biomarker tables both host-derived and tumor-derived, provide clear overviews.

However, I have several comments and concerns that need to be addressed before the manuscript is suitable for publication, and to enhance its overall quality:

1. Please include a methods section that outlines your literature search strategy. This should cover databases, date ranges, keywords, inclusion/exclusion criteria, and how you assessed study quality to strengthen the review's rigor.
2. The manuscript lists differing incidence rates for VTE/ATE associated with ICIs (e.g., 2.7% vs. 9–24% VTE). Can you clarify these figures by specifying time frames, ascertainment methods, tumor types, and how you handled competing risks for each cited source?
3. The manuscript reports the extended apixaban dose as 2.5 mg once daily. Can you verify this dose or correct it to the standard 2.5 mg twice daily? Also, please cite the most recent cancer-specific trial on extended therapy.
4. Please clarify the section on DOAC metabolism and drug interactions. Specify which DOACs are substrates of CYP3A4 and provide examples of common oncology drugs that may pose interaction risks.
5. For biomarker claims (e.g., D-dimer, sP-selectin, CRP flare, NET markers, EV-TF), could you provide effect sizes (hazard ratios with confidence intervals, areas under the curve) and their timing in relation to ICI initiation? Include details on assay standardization status and any clinically actionable thresholds.
6. The L2HSDK score is highlighted for ICI-treated Asian cohorts. Has it been validated in populations outside Asia or across different tumor types? How does its performance compare to Khorana and COMPASS-CAT in cohorts treated only with ICIs?
7. The conclusions of the network meta-analysis suggest fewer arterial events when a single ICI is added to chemotherapy, possibly due to cytotoxic immunosuppression. Please provide specific references, effect estimates, and discuss any alternative explanations or confounders.
8. The Vienna CATS score is mentioned as validated in contemporary cohorts that include immunotherapy. Please cite the study details for validation, including the population, time points, and performance metrics. Also, comment on its practical implementation, such as the availability of the P-selectin assay.
9. Can you add a practical clinical algorithm, either as text or a figure, that proposes how to integrate clinical factors and serial biomarkers for deciding on prophylaxis in ICI recipients? This should include factors that influence bleeding risk and suggestions for monitoring intervals.
10. Please provide more detail on the PRIN project (NCT07288632): its design, primary endpoints, population, biomarkers collected, and expected timelines.

Author Response

We thank the Reviewer for the positive and constructive evaluation of our manuscript and for the helpful suggestions to further improve its quality.

Comment 1: Please include a methods section that outlines your literature search strategy. This should cover databases, date ranges, keywords, inclusion/exclusion criteria, and how you assessed study quality to strengthen the review's rigor.

Response 1: We thank the Reviewer for this comment. We have added a description of the literature search strategy at the end of the Introduction (lines 57–68).

Comment 2: The manuscript lists differing incidence rates for VTE/ATE associated with ICIs (e.g., 2.7% vs. 9–24% VTE). Can you clarify these figures by specifying time frames, ascertainment methods, tumor types, and how you handled competing risks for each cited source?

Response 2: We agree with this comment. We added references with different incidence rates and specified possible reasons in the “Understanding CAT” section (lines 149-157).

Comment 3: The manuscript reports the extended apixaban dose as 2.5 mg once daily. Can you verify this dose or correct it to the standard 2.5 mg twice daily? Also, please cite the most recent cancer-specific trial on extended therapy.

Response 3: Thank you for your comment. We corrected the dose of apixaban, cited the recent trial in the “Discussion” section and added the reference (lines 687; 690-692).

Comment 4: Please clarify the section on DOAC metabolism and drug interactions. Specify which DOACs are substrates of CYP3A4 and provide examples of common oncology drugs that may pose interaction risks.

Response 4: Thank you for pointing this out. We included the recommended information in the section “Discussion” (lines 670-671; 673-675).

Comment 5: For biomarker claims (e.g., D-dimer, sP-selectin, CRP flare, NET markers, EV-TF), could you provide effect sizes (hazard ratios with confidence intervals, areas under the curve) and their timing in relation to ICI initiation? Include details on assay standardization status and any clinically actionable thresholds.

Response 5: We thank the Reviewer and agree that quantitative effect sizes and clinically actionable thresholds are important. However, for most biomarkers associated with ICI-related thrombosis, the available evidence is limited and heterogeneous. Therefore, we reported quantitative estimates only where supported by ICI-specific data, namely for early CRP changes (“CRP flare”), and added Supplementary Table S1 to highlight gaps in evidence regarding timing, assay standardization, and validated thresholds.

Comment 6: The L2HSDK score is highlighted for ICI-treated Asian cohorts. Has it been validated in populations outside Asia or across different tumor types? How does its performance compare to Khorana and COMPASS-CAT in cohorts treated only with ICIs?

Response 6: Thank you for your observation. We added a more detailed comment on this score (lines 279-284).

Comment 7: The conclusions of the network meta-analysis suggest fewer arterial events when a single ICI is added to chemotherapy, possibly due to cytotoxic immunosuppression. Please provide specific references, effect estimates, and discuss any alternative explanations or confounders.

Response 7: Thank you very much for your suggestion. We provided the suggested information, possible explanations, and added references (lines 120-179).

Comment 8: The Vienna CATS score is mentioned as validated in contemporary cohorts that include immunotherapy. Please cite the study details for validation, including the population, time points, and performance metrics. Also, comment on its practical implementation, such as the availability of the P-selectin assay.

Response 8: Thank you for pointing this out. We added details and comments as suggested (lines 199-221).

Comment 9: Can you add a practical clinical algorithm, either as text or a figure, that proposes how to integrate clinical factors and serial biomarkers for deciding on prophylaxis in ICI recipients? This should include factors that influence bleeding risk and suggestions for monitoring intervals.

Response 9: We thank the reviewer for this valuable suggestion and agree that a practical clinical algorithm would be clinically useful. However, current evidence in patients treated with ICIs is largely exploratory and lacks prospective validation, standardised biomarker thresholds, defined monitoring intervals, and an integrated assessment of bleeding risk. Proposing an algorithm under these conditions could be ambiguous. Therefore, in light of ongoing prospective studies, such as the PRIN project, we decided not to include a clinical algorithm in this work.

Comment 10: Please provide more detail on the PRIN project (NCT07288632): its design, primary endpoints, population, biomarkers collected, and expected timelines.

Response 10: We thank the Reviewer for this comment. We added the details to the Conclusion section (lines 757-761).

The line numbers referring to the revisions correspond to the Word document uploaded as “Revised Manuscript".

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have carefully addressed the major and minor concerns raised in the previous round of review. The revised manuscript has been substantially improved in terms of clarity, organization, and depth of discussion. In particular, the authors have strengthened the mechanistic links between immune checkpoint inhibitor therapy and cancer-associated thrombosis.

Overall, the revisions have significantly enhanced the scientific rigor and readability of this review. I have no further major concerns, and I believe that the manuscript is now suitable for publication in its current form.

Author Response

Comment 1: The authors have carefully addressed the major and minor concerns raised in the previous round of review. The revised manuscript has been substantially improved in terms of clarity, organization, and depth of discussion. In particular, the authors have strengthened the mechanistic links between immune checkpoint inhibitor therapy and cancer-associated thrombosis. Overall, the revisions have significantly enhanced the scientific rigor and readability of this review. I have no further major concerns, and I believe that the manuscript is now suitable for publication in its current form.

Response 1: We thank the Reviewer for the positive evaluation of the revised manuscript and for considering it suitable for publication.