[Value of D-dimer measurement in the exclusion diagnosis of pulmonary embolism]

Diagnosis of pulmonary embolism, a common and potentially fatal disease, is first based on clinical probability assessment, and often requires invasive testing such as pulmonary angiography. However, it often represents a diagnosis challenge. The measurement of D-dimer, a specific fibrin degradation product, was recently introduced in the diagnosis strategy. Even if D-dimer levels are highly sensitive in the diagnosis of pulmonary embolism, they are not specific of an on-going venous thromboembolic process. Its high negative predictive value enables to validly exclude diagnosis of pulmonary embolism, particularly in outpatients, in the case of D-dimer levels below a well-defined cut-off value. Prospective management studies confirmed that D-dimer measurement could be validly used as an initial screening test in patients with clinically suspected pulmonary embolism. Using such a diagnosis strategy, imaging tests would be performed only in the case of high D-dimer levels i.e. above the cut-off level. Even if they constitute the gold standard, conventional Elisa are not useful as a routine emergency test. New rapid and automated assays based on various principles (Elisa-derived or micro-latex agglutination) are now available. All demonstrated both high sensitivity (about 100%) and negative predictive value (over 98%), using a well-defined cut-off level (usually defined to be 500 ng/mL). Finally, with the increasing number of new D-dimer assays currently available, a lack of standardization was pointed out. As the result, both the clinical significance and the cut-off level have to be defined in prospective clinical trials, for each individual assay.