Highlight
The study identifies Growth Differentiation Factor-15 (GDF-15) as a significant baseline predictor for increased venous thromboembolism (VTE) risk in ambulatory cancer patients. Additionally, elevated baseline N-terminal pro-B-type natriuretic peptide (NT-proBNP) and C-reactive protein (CRP) were found to be independently associated with a higher risk of clinically relevant bleeding. Perhaps most notably, a dynamic increase in high-sensitivity troponin T (hs-TnT) within the first month of treatment was linked to a nearly two-fold increase in VTE risk. These findings led to the development of clinical nomograms aimed at improving personalized risk assessment for patients undergoing chemotherapy.
Background: The Challenge of Cancer-Associated Thrombosis
Venous thromboembolism (VTE), encompassing deep vein thrombosis and pulmonary embolism, remains a leading cause of morbidity and mortality among patients with cancer. The relationship between malignancy and coagulation is complex, driven by systemic inflammation, direct tumor procoagulant activity, and the side effects of various anticancer therapies. Historically, the Khorana score has been the clinical standard for identifying patients at high risk for VTE. However, its predictive accuracy is often criticized for being suboptimal in certain cancer types and for failing to account for the concurrent risk of bleeding associated with prophylactic anticoagulation.
As precision medicine evolves, there is an urgent need for more refined predictive tools. Biomarkers that reflect systemic inflammation and cardiac stress offer a promising avenue for risk stratification. While biomarkers like C-reactive protein (CRP) and cardiac troponins are well-established in cardiovascular medicine, their role as dual indicators for both thrombotic and hemorrhagic complications in the oncological setting has remained poorly defined until now.
Study Design: A Post Hoc Analysis of the AVERT Trial
This study utilized data from the AVERT trial, a randomized, double-blind, placebo-controlled trial that evaluated the efficacy and safety of apixaban for VTE prevention in high-risk ambulatory cancer patients (Khorana score ≥2). The researchers performed a post hoc analysis on a cohort of 574 patients, with baseline biomarker samples available for 514 individuals. The biomarkers measured included two inflammatory-related markers—CRP and GDF-15—and two cardiac markers—NT-proBNP and hs-TnT.
Measurements were taken at baseline, one month, and three months (CRP only for the three-month mark). The primary objectives were to estimate the associations between these markers and the occurrence of VTE and clinically relevant bleeding. The statistical approach utilized Fine and Gray regression models to account for the competing risk of death, adjusting for variables such as age and the presence of advanced-stage cancer. This methodological rigor ensures that the findings reflect the specific risk of the clinical events in question rather than general mortality.
Key Findings: Biomarkers as Predictors of VTE
The results highlighted distinct roles for different biomarkers in predicting thrombotic events. At baseline, elevated GDF-15 was significantly associated with an increased risk of VTE, with a subdistribution hazard ratio (SHR) of 1.36 (95% CI 1.01-1.84). GDF-15 is a member of the transforming growth factor-beta cytokine superfamily and is known to be upregulated in response to tissue injury, inflammation, and malignancy. Its association with VTE suggests that it may serve as a surrogate for the intensity of the pro-thrombotic inflammatory state within the tumor microenvironment.
The study also revealed the importance of longitudinal monitoring. While baseline hs-TnT levels were not significantly predictive, an increase in hs-TnT from baseline to the one-month mark was strongly associated with a higher risk of subsequent VTE (SHR 1.89, 95% CI 1.14-3.16). This dynamic change may reflect subclinical cardiac strain or vascular injury resulting from chemotherapy, which in turn predisposes the patient to thrombotic complications.
Key Findings: Markers for Clinically Relevant Bleeding
Predicting bleeding is as critical as predicting clots, particularly when deciding on the use of direct oral anticoagulants (DOACs). The study found that baseline levels of NT-proBNP and CRP were specifically linked to bleeding risk. Patients with higher baseline NT-proBNP exhibited an SHR of 1.44 (95% CI 1.08-1.92) for clinically relevant bleeding. Similarly, elevated baseline CRP was associated with an SHR of 1.38 (95% CI 1.07-1.76).
The association between NT-proBNP—a marker of myocardial wall stress—and bleeding is particularly intriguing. It may indicate a level of systemic vascular fragility or be a proxy for occult cardiovascular comorbidities that increase the risk of hemorrhagic events during anticoagulation. CRP’s role likely reflects the broader impact of systemic inflammation on mucosal integrity and vascular permeability, both of which can exacerbate bleeding tendencies.
Clinical Tools: The Development of Nomograms
To translate these statistical associations into clinical practice, the researchers developed nomograms. These visual calculation tools allow clinicians to input biomarker levels and patient characteristics to estimate the individualized probability of VTE and bleeding over time. By providing a more granular risk profile than the Khorana score alone, these nomograms could help oncologists and hematologists identify which patients would benefit most from thromboprophylaxis and which patients might be at excessive risk for bleeding, thereby requiring closer monitoring or alternative strategies.
Expert Commentary: Mechanistic Insights and Clinical Utility
The integration of cardiac and inflammatory biomarkers into oncology represents a significant step toward “onco-cardiology” risk assessment. The finding that GDF-15 and hs-TnT correlate with VTE, while NT-proBNP and CRP correlate with bleeding, suggests that different biological pathways drive these two complications. GDF-15 is increasingly recognized not just as a marker of inflammation, but as a mediator of cancer cachexia and progression, which are inherently pro-thrombotic states.
However, several limitations must be considered. As a post hoc analysis, the study is hypothesis-generating rather than definitive. The patient population was limited to those with a Khorana score of ≥2, meaning the results may not be generalizable to low-risk patients. Furthermore, while the nomograms are promising, they require external validation in independent prospective cohorts before they can be widely adopted in clinical guidelines.
Conclusion: Towards Precision Thromboprophylaxis
This study demonstrates that select inflammatory and cardiac biomarkers can provide critical insights into the risk of VTE and bleeding in cancer patients. By utilizing GDF-15, NT-proBNP, CRP, and dynamic changes in hs-TnT, clinicians can move beyond one-size-fits-all risk scores. The development of predictive nomograms offers a practical framework for personalizing anticoagulation therapy, potentially reducing the burden of VTE while minimizing the danger of treatment-induced bleeding. Future research should focus on validating these markers in broader patient populations and exploring whether biomarker-guided therapy leads to improved clinical outcomes.
Funding and References
This research was supported by institutional grants and utilized data from the AVERT trial. For further reading and detailed statistical data, refer to the original publication:
Roy DC, Wang TF, Mallick R, Burger D, Carrier M, Wells P, Hawken S. Venous thromboembolism and bleeding in cancer patients: role of inflammatory and cardiac biomarkers. Eur Heart J. 2025 Dec 12:ehaf1002. doi: 10.1093/eurheartj/ehaf1002. PMID: 41384379.