Background: Deep vein thrombosis (DVT) is a common complication in cancer patients associated with significant morbidity and mortality. D-dimer is a widely used screening tool for suspected DVT; however, its accuracy may be influenced by body mass index (BMI). We evaluated whether BMI modified the diagnostic utility of D-dimer levels in patients with cancer and DVT.
Methods: We retrospectively analyzed 439 patients with newly diagnosed solid cancers suspected of having DVT between January 2013 and December 2020. DVT was confirmed or excluded by using computed tomography or echography. D-dimer was measured, and patients were classified by BMI as low- (< 18.5 kg/m2), normal- (18.5-24.9 kg/m2), or high-BMI (≥ 25.0 kg/m2). D-dimer levels, positive and negative predictive values (PPV and NPV), and overall survival (OS) were compared.
Results: Of 439 patients, 175 (39.9%) had DVT. BMI did not differ significantly between the DVT-positive and DVT-negative patients. In the normal and high BMI groups, D-dimer levels were significantly higher in patients with DVT than those without (p < 0.01), whereas patients with a low BMI did not show this difference (p = 0.12). Using a 1 µg/mL cut-off, PPV was 32% in low-, 52% in normal-, and 49% in high-BMI patients; NPV was 83%, 91%, and 97%, respectively. OS did not differ among the BMI groups, and no deaths were directly related to DVT.
Conclusion: Diagnostic performance of D-dimer level may diminish in underweight patients with cancer. Clinicians should carefully interpret the D-dimer levels in low-BMI populations and consider additional diagnostic strategies.
Keywords: Body mass index; Cancer patient; D-dimer; Deep vein thrombosis.
© 2025. The Author(s) under exclusive licence to Japan Society of Clinical Oncology.