Recurrent PE risk after long-term warfarin therapy remains high (PADIS-PE)
Not long ago, doctors were taught that 6 months of anticoagulation was plenty for patients with unprovoked pulmonary embolism.
That standard was never based on long-term outcomes studies. And as longer-range data started to come in -- gulp -- it was clear that large numbers of people treated with 6-month warfarin courses after unprovoked PE experienced recurrent pulmonary embolism, even in the absence of any detectable thrombophilia. For thousands, their recurrent PEs were fatal or disabling.
Experts advised longer courses of warfarin after unprovoked PE, with the ACCP's flagship antithrombotic updated guidelines recommending an "indefinite" period of anticoagulation be considered for patients with unprovoked pulmonary embolism.
Now, a landmark study published in JAMA confirms the prevention benefits of continuing warfarin for longer periods after unprovoked pulmonary embolism. However, after long-term warfarin was finally stopped, recurrent PEs and DVTs remained disturbingly common. And clot prevention came at a price of higher rates of major bleeding among those taking warfarin for extended periods.
Authors randomized 371 patients with unprovoked PE at 14 French hospitals to receive either warfarin for another 18 months (2 years total warfarin), or placebo (6 months total warfarin). They were also followed for ~2 years after the study ended. The trial was christened PADIS-PE, and took 7 years to complete.
The trial further confirmed warfarin's benefits in long-term prevention of recurrent VTE: during the 18-month treatment period, only 3 patients (1.7%) taking warfarin developed symptomatic recurrent DVT or PE, compared to 25 (14%) taking placebo. These events were mostly nonfatal PEs with a few DVTs (there were no fatal PEs in either arm during the treatment period).
However, major bleeding (fall in Hb >2 g/dL) occurred in 4 patients taking warfarin, vs. 1 taking placebo.
Investigators continued to follow patients for about 2 years after the treatment period ended (~3.5 years overall). In this period, all patients were intended to be off warfarin.
The risk of recurrent DVT and PE remained high long-term: 25 patients (~16%) who had been originally treated with long-term warfarin developed recurrent DVT/PE in the ~2 years after warfarin discontinuation; most were PEs, four were fatal. In the placebo group, 10 patients (~9%) had recurrent DVT/PEs after study discontinuation (most were PEs, none were fatal).
Risks of major bleeding are probably higher in real world practice, where warfarin monitoring and medical care in general will not match the intensity of a clinical trial.
Warfarin's long-term prevention benefits in PADIS-PE echo those of a Cochrane review of 11 trials conducted 1987-2009 (n=3,716), comparing warfarin courses after acute DVT or PE.
Patients taking long-term warfarin had a composite rate of recurrent VTE of 2%, compared to 9% taking shorter courses of warfarin. Major bleeding occurred in 2% with long-term warfarin and 0.4% with short-term warfarin, with overall mortality of ~4.5% in both groups. Risk of pulmonary embolism did decline over years, but warfarin's bleeding risks did not, bringing the risks into an uncertain equipoise.
Despite warfarin's apparent superiority over placebo in long-term prevention of PE, the PADIS-PE authors were careful to point out its limits:
Regardless of whether patients had been treated for 6 or 24 months, about 80% of recurrences comprised another episode of symptomatic pulmonary embolism, of which 8% (4 of 52 episodes) were fatal. Moreover, the vast majority of recurrences were unprovoked in both study groups (>80%), as in other studies.7- 9 Thus, extending the length of secondary anticoagulant prophylaxis to 18 months did not modify the clinical presentation of recurrent venous thromboembolism. Most recurrences were unpreventable and represented the most severe form of venous thromboembolism.
It's hard to interpret their precise intended meaning, since warfarin clearly reduced the rate of recurrent DVT/PE during the treatment period (maybe it made more sense in French.) They seem to mean warfarin did not modify the clinical presentation of recurrent VTE after warfarin is stopped, which is not surprising.
PADIS-PE seems to provide support for ACCP's recommendation for considering "indefinite" (i.e., potentially lifelong) anticoagulation in patients with unprovoked pulmonary embolism, especially after severe initial PEs.
In patients with unprovoked deep venous thrombosis, who are more likely to experience recurrence of DVT than to develop PE, risks of discontinuation of warfarin may be lower than risks of eventual major bleeding, but even this is hard to say with confidence.
Clinical takeaway: The long-term management of people with unprovoked deep venous thrombosis and pulmonary embolism is one of the most challenging problems in medicine. These studies all seem to confirm one thing -- 6 months treatment with warfarin after unprovoked PE is too little for most patients -- but the ideal treatment period after unprovoked DVT/PE remains unknown.