New drugs slash catheter-related clot risk

4 minute read


Two experimental factor XI inhibitors more than halved the rate of venous thromboembolism following PICC insertion, without an increase in major bleeding, a phase 2 trial has found.


Two investigational anticoagulants have cut the risk of catheter-related thrombosis by more than half, raising hopes of a safer approach to preventing blood clots in patients requiring long-term intravenous treatment. 

The phase 2 ROXI-CATH trial found that a single infusion of either amrecibart or cenvacibart reduced venous thromboembolism (VTE) rates to 20%, compared with 43% among patients receiving placebo. 

Neither treatment was associated with major bleeding during the primary 14-day assessment period, suggesting that targeting factor XI (FXI) could offer an alternative to conventional anticoagulation. 

The findings, published in The Lancet Haematology, could have particular implications for patients with cancer, who face an increased risk of thrombosis associated with peripherally inserted central catheters (PICCs) used to deliver chemotherapy and other treatments. 

Despite the frequency of catheter-related thrombosis, routine pharmacological prophylaxis is not currently recommended because of concerns about bleeding and uncertainty over the net clinical benefit of existing anticoagulants. 

The international, double-blind, placebo-controlled trial enrolled 195 adults undergoing PICC insertion at 12 hospitals across Europe and North America between September 2024 and December 2025. 

Of these, 62% had active cancer, making the findings particularly relevant to haematologists managing patients requiring prolonged intravenous therapy. 

Participants were randomly assigned to receive a single intravenous infusion of amrecibart 300mg, cenvacibart 250mg, or placebo within 24 hours of catheter placement. 

Both drugs are monoclonal antibodies targeting FXI, although they inhibit different domains of the protein involved in coagulation. 

Amrecibart targets the FXI Apple 2 domain, blocking activation mediated by factor XIIa, while cenvacibart targets the catalytic domain, inhibiting FXI activity induced by FXIa, FXIIa or thrombin. 

At 14 days, adjudicated VTE had occurred in 13 of 65 patients (20%) receiving amrecibart and 13 of 66 (20%) receiving cenvacibart, compared with 27 of 63 patients (43%) in the placebo group. 

The odds ratios were 0.37 for amrecibart and 0.36 for cenvacibart, with a 99.8% posterior probability of benefit for each treatment compared with placebo. The results corresponded to an absolute reduction of approximately 23 percentage points in thromboembolic events. 

A secondary analysis specifically examining PICC-associated thrombosis showed similarly encouraging results, with events occurring in 18% of amrecibart recipients and 20% of cenvacibart recipients, compared with 43% of placebo recipients. 

Deep vein thrombosis occurred in 14%, 8%, and 32% of patients, respectively. 

However, the primary endpoint included asymptomatic thromboses detected through routine ultrasound screening, and the trial was not powered to establish whether treatment reduced clinically symptomatic events. 

Symptomatic VTE occurred in four patients receiving amrecibart, none receiving cenvacibart, and three receiving placebo. 

The safety findings were also encouraging, with no major bleeding events reported in any group during the first 14 days. 

Clinically relevant non-major bleeding occurred in 3% of participants in each treatment arm and the placebo group. 

Treatment-emergent adverse events through day 100 were reported in 69% of amrecibart recipients, 64% of cenvacibart recipients, and 76% of placebo recipients. 

One participant receiving cenvacibart experienced a traumatic subdural haemorrhage following a fall on day 28 and subsequently died. Although investigators considered the death unrelated to treatment, the researchers acknowledged that residual FXI inhibition could not be excluded as a contributing factor. 

The researchers said the findings supported FXI inhibition as a promising strategy for preventing thrombosis triggered by contact between blood and artificial surfaces, including central venous catheters. 

Unlike conventional anticoagulants, FXI inhibitors aim to interrupt thrombosis-associated coagulation pathways while having less effect on normal haemostasis. 

The researchers also noted that the similar efficacy of the two antibodies, despite their different mechanisms, suggested that FXIIa-mediated FXI activation could play a central role in catheter-related thrombosis. 

“Our results align with a previous phase 2 study of FXI inhibition in patients with cancer undergoing CVC placement, in which catheter-associated thrombosis was detected in 13% of gruticibart-treated participants compared with 40% of participants in the control cohort, with no significant adverse or bleeding-related events reported,” they wrote. 

“Gruticibart, like amrecibart, binds the FXI A2 domain and selectively inhibits FXIIa-mediated FXI activity, whereas cenvacibart targets the FXI catalytic domain and inhibits FXI activity induced by FXIa, FXIIa, or thrombin.” 

However, the researchers cautioned that the relatively small ROXI-CATH trial had important limitations, including the detection of asymptomatic thromboses of uncertain clinical significance. 

The study population was also predominantly white, with most participants recruited in Europe, potentially limiting the generalisability of the findings, they said. 

Patients requiring therapeutic anticoagulation or most antiplatelet treatments were excluded, as were those with several conditions associated with increased bleeding or thrombotic risk. 

The researchers said larger trials would be needed to establish whether the reductions in catheter-associated thrombosis translated into meaningful clinical benefits without increasing bleeding complications. 

“This will be addressed in the ROXI-PEAK phase 3 clinical trial (NCT07697599), which is currently underway to further evaluate the efficacy and safety of amrecibart and cenvacibart in preventing PICC-associated thrombosis,” they wrote. 

The Lancet Haematology, November 2026 

End of content

No more pages to load

Log In Register ×