Intern Emerg Med. 2021 Apr 10. doi: 10.1007/s11739-021-02730-y. Online ahead of print.
Recent guidelines for diagnosing acute pulmonary embolism (PE) are based on clinical decision rules and D-dimer. D-dimer measurement is recommended only for patients who are 'PE-unlikely'. We aimed to assess the current guidelines for cancer patients and to determine an optimal D-dimer cut-off level. This retrospective observational study was conducted in the emergency department of Asan Medical Center (Seoul, Korea) between 02/2017 and 09/2017 for the development cohort and between 06/2018 and 02/2019 for the validation cohort. Among adult active cancer patients with suspected PE, we included those who were 'PE-unlikely' according to Wells' criteria and who underwent D-dimer testing and computed tomographic pulmonary angiography (CTPA). A total of 498 patients (227 in the development cohort and 271 in the validation cohort) were included, and PE was diagnosed in 8.8% and 18.5% of patients, respectively. The optimal D-dimer cut-off level was 2.0 μg/mL. This elevated cut-off level showed a much higher specificity of 21.3% (95% confidence interval [CI] 16.2-27.3%) and 21.7% (95% CI 16.8-7.6%) in the development and validation sets, respectively, compared with the specificity of 4.4% (95% CI 2.3-8.1%) and 4.1% (95% CI 2.2-7.6%) using the age-adjusted cut-off. The new D-dimer cut-off value identified unnecessary CTPA for 21.3% of patients (absolute difference, 16.9%, 35 of 207) in the development cohort and 21.7% (absolute difference, 17.6%, 39 of 221) of patients in the validation cohort compared to using the standard age-adjusted cut-off. The elevated D-dimer cut-off value combined with Wells' criteria might reduce unnecessary CTPA in active cancer patients with a 'PE-unlikely' classification. Further clinical trials are warranted to improve the PE diagnostic strategy in cancer patients.