Impact of inferior vena cava thrombus extension on thrombolysis for acute iliofemoral thrombosis

Abstract

Background: Inferior vena cava (IVC) thrombosis may occur in patients with iliofemoral deep venous thrombosis (DVT), and its impact on thrombolysis outcomes is poorly defined. This study compared outcomes of patients undergoing thrombolysis for acute iliofemoral DVT with and without IVC involvement.

Methods: Patients who underwent thrombolysis for iliofemoral DVT between May 2007 and March 2014 were identified from a prospectively maintained database and divided into two groups: those with IVC involvement and those without. End points were technical and clinical success (≥50% lysis or freedom from 30-day DVT recurrence), long-term DVT recurrence, and post-thrombotic syndrome (PTS; Villalta score ≥5). Multivariate regression models were used to determine predictors of anatomic and clinical failures.

Results: There were 102 patients (127 limbs) treated with various combinations of catheter-directed or pharmacomechanical thrombolysis. In 46 patients, thrombus extended into the IVC (54.3% extended up to the renal veins; 87% had ≥50% luminal reduction; 50% occurred in association with an indwelling thrombosed IVC filter). The caval group had fewer women and more previous DVTs but otherwise was similar to the noncaval group. Pharmacomechanical thrombolysis was used more frequently in the caval thrombus group (97.8% vs 82.1%; P = .011), and iliac vein stenting was used more often in the noncaval group (41.3% vs 62.5%; P = .033). Clinical success was similar between the two groups (88.7% for caval vs 89.3% for noncaval; P = .921). All failures in the caval group occurred in patients with an indwelling thrombosed IVC filter. Primary patency at 2 years for the caval and noncaval groups was 76.7% and 78.0%, respectively (P = .787). Valve reflux and PTS at 2 years were higher in the noncaval group (50.8% and 34.3% vs 23.3% and 11.5% in the caval group; P = .013 and P = .035). On multivariate analysis, incomplete lysis was predictive of recurrence (hazard ratio [HR], 22.7; P < .001) and PTS (HR, 5.59; P = .010), whereas caval involvement (HR, 0.22; P = .005) was protective from PTS.

Conclusions: IVC thrombosis does not have an impact on the technical success of thrombolysis in patients with iliofemoral DVT; the presence of a thrombosed IVC filter, though, may make failure more likely. Caval thrombosis may not affect primary patency but is associated with a lower incidence of PTS after successful lysis.