Diagnosis of deep vein thrombosis by plasma-soluble fibrin or D-dimer

The present study was designed to determine the cutoff values of D-dimer and soluble fibrin (SF) for the diagnosis of deep venous thrombosis (DVT) and pulmonary embolism (PE) in Japanese patients. Plasma levels of these molecules were measured in 243 patients suspected of having DVT and 100 healthy volunteers (controls). Out of 243 patients, 20 patients were diagnosed with DVT. In the control group, plasma levels of D-dimer and SF did not show normal distribution, and the 95% confidence intervals (CI) of D-dimer and SF were 2.45 mu g/mL and 6.92 mu g/mL, respectively. Plasma levels of D-dimer and SF of patients with DVT were significantly higher than of those without DVT. In patients with DVT, the minimum values of D-dimer and SF were 1.71 and 1.44 mu g/mL, respectively. When the cutoff value was set at the average + 1 SD of those of the control (D-dimer, about 1.8 mu g/mL; SF, about 1.4 mu g/mL), 1 and 0 patient with DVT was overlooked, respectively. The sensitivity and specificity of D-dimer and SF for DVT were 95% and 100%, and 61.9% and 53.8%, respectively. When the cutoff value was set at 95% CI of the control (D-dimer, 2.5 mu g/mL; SF, 6.9 mu g/mL), 2 and 9 patients with DVT were overlooked, respectively. The sensitivity and specificity of D-dimer and SF were 90% and 50%, and 77.6% and 88.3%, respectively. When the cutoff values set at 2.5 mu g/mL of D-dimer or 6.9 mu g/mL of SF, 1 DVT patient was overlooked, with sensitivity and specificity of 95% and 69.5%. Our data suggest that both D-dimer and SF are useful markers for the diagnosis of DVT and that measurement of both D-dimer and SF increases the sensitivity and specificity for the diagnosis of DVT/PE.