Normalization rates of compression ultrasonography in patients with a first episode of deep vein thrombosis of the lower limbs: association with recurrence and new thrombosis

Abstract

Background and objectives: Delayed thrombus regression after a first episode of deep vein thrombosis (DVT) of the inferior limbs has been implicated in the development of the post-thrombotic syndrome. Whether normalization of vein segments involved in the index DVT has prognostic significance with respect to the probability of DVT recurrence or new thrombosis is currently unknown. In this study, we prospectively monitored thrombus regression in consecutive patients with symptomatic and asymptomatic DVT. Factors affecting normalization rates and the relationship between previous normalization and DVT recurrence or new thrombosis were explored.

Design and methods: One hundred and seventy-nine patients with a first episode of symptomatic DVT of the lower limbs (38 with cancer) and 104 patients with DVT occurring after hip replacement surgery were serially monitored by real time B-mode compression ultrasonography (C-US) over a period of 12 months (months 1, 3, 6 and 12). C-US normalization of popliteal and femoral venous segments was arbitrarily assigned to be residual thrombus occupying, at maximum compressibility, less than 40% of the vein area in the absence of compression.

Results: In patients with no DVT recurrence or new thrombosis, C-US normalization was observed at 12 months in 100% of 99 patients with post-operative DVT, in 59% of 134 cancer-free symptomatic DVT outpatients and in 23.3% of 30 symptomatic DVT outpatients with cancer (p = 0.0001). Independent negative effects on the probability of C-US normalization were observed for younger age (p <0.05), for the outpatient presentation of the index DVT (p 0.017), for DVT involving the entire femoro-popliteal axis (p 0.05), and for the presence of cancer (p 0.05). DVT recurrence or new thrombosis was observed in 5 patients with post-operative DVT (4.8%), in 7 cancer-free patients with symptomatic DVT (5.0%) and in 8 patients with cancer (21.1%). Only 4 of these patients had shown normalization of their index DVT prior to the event. The presence of cancer was the only significant predictor of DVT recurrence and/or new thrombosis occurring within 3 months from the index DVT (OR = 4.90, p = 0.002). The absence of previous C-US normalization was the only predictor of recurrence or new thrombosis occurring after 3 and 6 months from the index DVT (OR 5.26, p 0.027).

Interpretation and conclusions: Absence of C-US normalization after a first episode of DVT appears to be a factor favoring recurrence or new thrombosis and may be relevant to the optimal duration of oral anticoagulant treatment.