D-dimer as a predictor of venous thromboembolism in acutely ill, hospitalized patients: a subanalysis of the randomized controlled MAGELLAN trial

Background D-dimer concentrations have not been evaluated extensively as a predictor of increased venous thromboembolism (VTE) risk in acutely ill, hospitalized medical patients. Objectives To analyze the relationships between D-dimer concentration, VTE and bleeding in the MAGELLAN trial (NCT00571649). Patients/methods This was a multicenter, randomized, controlled trial. Patients aged >= 40years, hospitalized for acute medical illnesses with risk factors for VTE received subcutaneous enoxaparin 40mg once daily for 10 +/- 4days then placebo up to day 35, or oral rivaroxaban 10mg once daily for 35 +/- 4days. Patients (n=7581) were grouped by baseline D-dimer <= 2xor >2xthe upper limit of normal. VTE and major plus non-major clinically relevant bleeding were recorded at day 10, day 35, and between days 11 and 35. Results The frequency of VTE was 3.5-fold greater in patients with high D-dimer concentrations. Multivariate analysis showed that D-dimer was an independent predictor of the risk of VTE (odds ratio 2.29 [95% confidence interval 1.75-2.98]), and had a similar association to established risk factors for VTE, for example cancer and advanced age. In the high D-dimer group, rivaroxaban was non-inferior to enoxaparin at day10 and, unlike the low D-dimer group, superior to placebo at day 35 (P<0.001) and days 11-35 (P<0.001). In both groups, bleeding outcomes favored enoxaparin/placebo. Conclusions Elevated baseline D-dimer concentrations may identify acutely ill, hospitalized medical patients at high risk of VTE for whom extended anticoagulant prophylaxis may provide greater benefit than for those with low D-dimer concentrations.