The Vienna CATScore for predicting cancer-associated venous thromboembolism: an external validation across multiple...

Area of Science:

• Oncology
• Hematology
• Medical Statistics

Background:

• Cancer patients on systemic therapy face a high risk of venous thromboembolism (VTE).
• Existing risk assessment models often underperform in contemporary cancer populations.
• There is a need for models that assess VTE risk at multiple time points during therapy.

Purpose of the Study:

• To validate the Vienna CATScore, a nomogram-based VTE risk assessment model.
• To evaluate the model's performance in patients initiating contemporary systemic anticancer therapies.
• To test the model's validity at baseline, 3 weeks, and 3 months after therapy initiation.

Main Methods:

• Prospective cohort study of 598 cancer patients.
• Validation of the Vienna CATScore, incorporating tumor type and D-dimer levels.
• Assessment of model discriminatory ability (c-statistics) at multiple time points.

Main Results:

• The Vienna CATScore demonstrated good discriminatory ability (c-statistics: 0.69 at baseline, 0.68 across time points).
• The model effectively distinguished between low- and high-risk groups for VTE.
• A 6-month predicted VTE risk threshold of 8% identified patients who could benefit from thromboprophylaxis.

Conclusions:

• External validation confirms the Vienna CATScore's effectiveness in predicting VTE risk.
• The model accurately assesses risk in patients receiving state-of-the-art anticancer therapies.
• The Vienna CATScore is useful for risk stratification across multiple time points during treatment.