Plasma D-dimer in the diagnosis of acute aortic dissection

We read with interest the recent article by Sodeck et al., who concluded that D-dimer testing is a reliable and sensitive diagnostic marker for exclusion of acute aortic dissection (AAD). In particular, an exclusion criterion is suggested, based on D-dimer levels <0.1 µg/mL, where the sensitivity of the test is 100%.1 AAD represents a medical emergency that, despite its rare occurrence, is characterized by severe morbidity and mortality. A timely and accurate diagnosis is the key to achieve a positive outcome. According to the current clinical data, diagnosis is mostly based on a high index of suspicion and diagnostic imaging. Similar to deep vein thrombosis and pulmonary embolism,2 plasma D-dimer has recently been suggested as a valuable ‘rule-out’ test in the diagnostic work-up of AAD.1,3 In fact, because of the high negative predictive value of a negative result, the role of D-dimer testing in patients admitted to the emergency department with suspected AAD might greatly contribute to the diagnostic reasoning, in that it would allow a rapid and reliable exclusion of AAD and other thrombotic pathologies without performing further invasive and expensive analyses.1–4 However, we are rather surprised to read that the threshold recommended by Sodeck et al. for ruling out AAD is 0.1 μg/mL. Reference studies indicate that plasma D-dimer levels vary markedly among healthy individuals and the suggested inter-quartile distribution ranges from 0.175 to 0.345 µg/ml.5 Using the same analytical technique (STA-Liatest D-dimer assay), Ghanima et al. observed that the exclusion rate for venous thrombo-embolism at a higher cut-off point (0.3 μg/mL) than that suggested by Sodeck et al. is <25%.6 Extrapolating the results from the receiver operating characteristic curve of the above mentioned article, the specificity at the 0.1 μg/mL cut-off is lower than 10%. Although studies on AAD and venous thrombo-embolism patients can not be directly compared, the diagnostic specificity is always influenced by the number of false-positive tests in the patients under evaluation. Regardless of the analytical technique to measure D-dimer, we suspect that adoption of a very low cut-off value, such as 0.1 µg/mL, would not be helpful to exclude from the diagnostic workout of AAD most of, if not all, outpatients referred to the emergency department. Probably, complete standardization of D-dimer results will never be possible, because D-dimer is not a single entity in plasma but a mixture of heterogeneous fibrin degradation products, and different antibody specificities from manufacturers.7 Therefore, a highly specific cut-off must be identified for each D-dimer assay used to rule out AAD, as currently suggested for venous thrombo-embolism.8