D-DImer Dosing for Risk Prediction of Central Venous Catheter Thrombosis (The DIReCT Trial).

Cancer patients receiving chemotherapy often need central venous catheters (CVC). There is no laboratory marker to identify patients at high risk of thrombosis who may benefit from thromboprophylaxis. D-Dimer dosing is a well studied test that may correlate with thrombosis risk. The aim of this study is to determine if d-dimer levels are a predictor of CVC thrombosis. Adult cancer patients needing a new CVC for more than 1 month and without an ongoing indication for anticoagulation were included. D-dimer dosing was done before (0–3 days), 24–48h after and 7 days after insertion of CVC. Duplex ultrasonography (US) was performed approximately 30 days after installation. We planned to look at quantitative d-dimer variation in patients to establish its prognostic value in relationship with eventual thrombosis occurrence. Thirty five patients are included in this analysis. There were 63% males and 37% females and the median age was 57. Eighty nine percent had solid tumors (64% gastrointestinal, 23% head and neck, 13% breast) and 11% had an hematologic malignancy. Fifty one percent received concomitant radiotherapy. Seventy nine percent had continuous infusion chemotherapy. Twenty two patients underwent US and 27 patients had at least 2 out of three planned d-dimer values. The incidence of venous thrombosis in the whole cohort was 31% (5 symptomatic, 6 asymptomatic) whereas 24 patients were considered free of thrombosis (11 confirmed by duplex, 13 clinically asymptomatic). In patients with available US results, baseline d-dimer level was similar for those with or without thrombosis (924 and 902). For these patients, the absolute increase of d-dimer was not significantly different in the two groups (194 vs 115 at day 1–2; 421 vs 63 at day 7) as well as relative increase (51% vs 31% at day 1–2; 145% vs 40% at day 7). Baseline d-dimer level was similar for the five patients with symptomatic thrombosis compared to the other 30 patients (1038 and 859), but higher at day 7 (1616 vs 905 p=0,048). Different strategies were analysed to try to define a high and a low risk group and two were selected. First, the population with a cutoff value of d-dimer levels at day 7 of > 1100 had a 40% risk of symptomatic thrombosis vs 0% for those below 1100. Second, an absolute increase in d-dimer values at day 1–2 of more than 500 was associated with a 66,7% risk of symptomatic thrombosis vs 0% for the others. Although numbers are small, patients with symptomatic thrombosis were more frequently receiving G-CSF, red cell transfusions, platelets transfusions and were less likely to be receiving continuous infusion chemotherapy. Our pilot study suggests that d-dimer testing after CVC insertion might permit risk stratification for thrombotic complications. If confirmed by larger trials, this strategy might be useful to plan more effective prophylaxis studies in these populations.