Millions of patients take vitamin K antagonists (VKA) for the treatment and prevention of thromboembolic events, but 1-3% of patients who take them have life-threatening or fatal bleeds. Up to 2% suffer intracranial hemorrhage, which results in death in up to 79% of cases.

“There are more than 42,000 [hospital] admissions related to serious bleeds with VKAs in the United States alone,” Ravi Sarode of the department of pathology at University of Texas Southwestern Medical Center, Dallas, said Wednesday morning during a State of the Art lecture at the ISTH 2015 Congress.

The three main options for VKA reversal are vitamin K, plasma, and prothrombin concentrates (PCC). “When vitamin K is given orally, it has a very slow onset of action and, therefore, it’s not a good treatment for patients with a serious bleed,” he said.

Ravi Sarode

Ravi Sarode

The effect of intravenously-administered vitamin K is “very predictable,” Sarode said, adding that it is not an option for very severe bleeding because it takes several hours to act. “A dose of 5-10 mg is typically recommended, however, you may want to use a lower dose of 1-2 mg to avoid VKA resistance in some patients.”

There are also limitations of plasma in urgent warfarin reversal, he said, including the need to match for ABO blood type and thaw frozen plasma. In addition, the exact dose for warfarin reversal is not yet known and ranges from 10-30 ml kg-1, “which can be a large volume in an 80 kg person,” he said. “Most of these patients are elderly and, therefore, they already have compromised cardiac and renal function. Therefore, these [patients] are transfused very slowly. There is also a risk for volume overload and unknown pathogen transmission.”

In a randomized trial, Sarode and his associates evaluated the impact of a 4-factor prothrombin complex concentrate (4F-PCC) in patients on VKAs who presented with a major bleed (Circulation 2013; 128: 1234-43). They found that 4F-PCC was non-inferior to plasma for hemostatic efficacy in the first 24 hours after infusion. However, 4F-PCC was superior to plasma for rapid decrease in INR 30 minutes after infusion.

Sarode said that while 4F-PCC was non-inferior to plasma for hemostatic efficacy, “it seems a more effective treatment.”  Most patients given 4F-PCC achieved vitamin K-dependent coagulation factor hemostatic levels within 30 minutes. Compared with plasma, 4F-PCC required 8 times less volume and was administered 8 times faster, and there were no cases of fluid overload reported. Both groups had similarly low numbers of thromboembolic events (TEEs).

He pointed out, though, that patients with a history of TEEs were excluded from the study. “When I talk to clinicians, they’re still hesitant to use 4F-PCCs, because the main concern they have is thrombolic events,” Sarode said.

In a single-center retrospective analysis, Sarode and his associates evaluated outcomes 60 days following the use of 4F-PCC for warfarin reversal in 113 patients. Overall, 7 TEEs occurred (6.1%) and 17 patients died (15%). “However, we found two patients who had thrombotic complications immediately after PCC infusion and died,” Sarode said. One had a left ventricular assist device and the other had antiphospholipid syndrome (APS), but four independent reviewers noted that their deaths could have been related to PCC.

“Because of these kinds of events, some centers have been using a low dose or fixed dose of PCC,” Sarode said. He cautioned that while it appears PCC is a more effective treatment for VKA-associated bleeding, compared with plasma alone, “concern remains for thromboembolic events,” he said. “Therefore, we need a well-designed randomized, controlled trial to find the ideal hemostatic dose.”

Sarode disclosed that he is a consultant for CSL Behring. He is also is an advisor for Octapharma.

In a separate presentation, Sunny Dzik expressed concern about what he described as the “premature proliferation” of guidelines concerning the reversal of oral factor Xa inhibitors by PCCs.

Sunny Dzik

Sunny Dzik

“There is a chorus of these things,” said Dzik, a hematologist who is co-director of the Blood Transfusion Service at Massachusetts General Hospital, Boston. “It’s unbelievable. I honestly think there are more guideline documents than there are patients who bled because of these new agents, but it’s certainly not hard to find guidelines, and it’s not hard to find advice on what to do.”

The tone of these guidelines has changed over the last two years, starting with phrases such as “PCCs should be administered…” and graduating to stronger statements such as “Emergent bleeding requires utilization of PCCs.” The latter statement “could be interpreted to mean that if you fail to give these agents, you’ve committed some sort of inadequate practice,” he commented.

Interest in this topic can be traced to a 2011 study of 12 healthy volunteers who were given rivaroxaban 20 mg b.i.d. for two days. On day three, they were given one more dose and randomized to receive intravenous saline or
15 IU kg-1 of the 4-Factor PCC Cofact, which is available in Europe but not in North America (Circulation 2011; 124:1573-79). The researchers found that Cofact immediately and completely reversed the anticoagulant effect of rivaroxaban in healthy subjects. “It’s interesting: the guideline statements that are based on this particular paper perhaps assumed that an effect that would be observed with this particular agent would also be observed with all [PCC] agents,” said Dzik, who was not affiliated with the study. “It turns out that’s not the case.”

Dzik highlighted several studies that shed further light on this topic. The first is a study in which PCCs were added to plasma in vitro. Four different doses of rivaroxaban (Kcentra/Beriplex), a product which is marketed in North America) were studied (Thrombosis Res 2014;133:671-81). Regardless of dose, no reversal of prothrombin occurred. “This is interesting, because it differs from the experience we saw with the Cofact agent,” Dzik said.

In a separate study, 12 normal volunteers were given Kcentra/Beriplex 20 mg b.i.d. for four days. On day 5 they were randomized to saline vs. 50 IU kg-1 of Kcentra/Beriplex (J Thromb Hemost 2014;12:1428-36). The investigators measured prothrombin time, activated partial thromboplastin time, thrombin generation, and anti-Xa levels. “This interested me as a reviewer because rivaroxaban is a factor Xa inhibitor,” Dzik noted. “By measuring anti-Xa levels, you’re getting to the heart of the matter much more than you would be [by evaluating] prothrombin time.”

The researchers found that the agent failed to correct prothrombin time, compared with saline, and that it was no more effective than saline for reversal of anti-Xa levels.

Dzik mentioned another trial in which Kcentra/Beriplex had no statistically significant effect on selected punch biopsy bleeding times in volunteers taking edoxaban (Circulation 2015; 131:82-90). He also noted findings from recent abstracts suggesting that andexanet alfa reproducibly reverses anti-Xa activity in volunteers taking oral factor Xa inhibitors.

Dzik cautioned that the “premature proliferation of guidelines” that advise clinicians to consider PCC, “perhaps originally intended for life-threatening situations, may have a tendency to creep toward situations where the strategy is without justification or medical value.”

He had no relevant financial conflicts to disclose.

By Doug Brunk |June 24, 2015