Therapeutic Plasma Exchange for Urgent Rivaroxaban Reversal

Abstract

Direct oral anticoagulants, which include the factor Xa inhibitor rivaroxaban, have some advantages over vitamin K antagonists in regard to stroke prevention in patients with atrial fibrillation. However, no antidotes to reverse the effect of oral anticoagulants are commercially available, which can complicate treating patients in whom reversal is urgent. We faced this challenge in a kidney transplant candidate, a 65-year-old man with end-stage renal disease who had been taking rivaroxaban for paroxysmal atrial fibrillation. When a deceased-donor kidney became available, we needed to rapidly reduce the patient's bleeding risk, while minimizing the cold ischemic time of the donor kidney. Therefore, we decided to take an experimental approach and perform therapeutic plasma exchange. The patient's plasma anti-factor Xa level decreased from 0.4 IU/mL immediately before treatment to 0.21 IU/mL afterward, indicating that rivaroxaban had been actively removed from circulation. Waste fluid showed significant anti-Xa activity, indicating that the risk of rebound anticoagulation had been mitigated. The patient subsequently underwent successful kidney transplantation.

To our knowledge, this is the first report of therapeutic plasma exchange to reverse the effects of rivaroxaban in a patient undergoing urgent surgery. This treatment may also be suitable for patients who have life-threatening, large-volume bleeding, especially in the presence of substantial kidney or liver dysfunction.

Direct oral anticoagulants, including factor Xa inhibitors such as rivaroxaban, are increasingly replacing vitamin K antagonists because they provide a more consistent anticoagulant effect and require less monitoring. However, the lack of effective antidotes to these agents remains a major challenge.¹ Factor Xa inhibitors have an extended half-life, and their use is contraindicated in patients with severe kidney or liver dysfunction because of impaired clearance.^2,3 We report the case of a kidney transplant candidate who had been taking rivaroxaban for atrial fibrillation when a deceased-donor kidney became available. Because the need to reverse anticoagulation was urgent, we performed therapeutic plasma exchange (TPE).

Case Report

A 65-year-old man with end-stage renal disease (ESRD) secondary to diabetes mellitus was admitted to our hospital to undergo deceased-donor kidney transplantation. He also had a history of paroxysmal atrial fibrillation, for which an outside physician had inappropriately prescribed rivaroxaban (20 mg/d). He had taken his last dose 12 hours before admission.

At the time of admission, the patient was in his usual state of health without symptoms of infection or acute illness. His physical examination revealed no pallor and no evidence of ecchymosis or bleeding. His hemoglobin level was 10.6 g/dL (near his baseline level), and his platelet count and coagulation times were within normal range. Because of rivaroxaban's extended half-life and the potential for impaired clearance in our patient with ESRD, we suspected that its remaining anticoagulant effect was appreciable. An anti-factor Xa assay calibrated for low-molecular-weight heparin—performed to evaluate the anticoagulant response to rivaroxaban—revealed a plasma anti-Xa level of 0.49 IU/mL, indicating near-therapeutic anticoagulation (therapeutic range, 0.6–2 IU/mL) (Fig. 1).

Fig. 1.

Graph shows the patient's plasma anti-factor Xa (low-molecular-weight heparin) levels from the time of admission. Time 0 marks 12 hours after the last 20-mg dose of rivaroxaban. Therapeutic plasma exchange was performed from hours 5 to 6.5.

Given the urgency of the operation, we opted to perform TPE to minimize both the cold ischemic time of the donor kidney and the risk of bleeding in our anticoagulated patient. Before TPE (5 hr after the initial assay), the patient's anti-Xa level was 0.4 IU/mL. The patient underwent an abbreviated 90-min session of plasma exchange with 3 L of fresh frozen plasma. After TPE, the serum anti-Xa level decreased to 0.21 IU/mL. Furthermore, the anti-Xa level in the waste fluid was 0.22 IU/mL, confirming that the drug had been removed from circulation.

Immediately after TPE, the patient underwent kidney transplantation, which was completed without complication. During surgery, his estimated blood loss was 250 mL, and he was given 2 units of packed red blood cells. His plasma anti-Xa level 10.5 hours after TPE was undetectable (<0.05 IU/mL), confirming that there was no rebound anticoagulation.

The patient's postoperative course was uneventful. The donor kidney began to produce urine immediately, and the patient's renal clearance progressively improved. During the perioperative period, he had no bleeding complications and remained hemodynamically stable. His hemoglobin level and coagulation times were also stable. At his 2-month follow-up examination, laboratory data showed that his kidney function continued to improve (Table I).

TABLE I.

Laboratory Values at Admission and at 2-Month Follow-Up

Discussion

Because approximately 90% of factor Xa inhibitors remain bound to plasma protein,^3,4 we hypothesized that TPE could be used to expeditiously remove rivaroxaban from the blood in our patient before he underwent kidney transplantation. We evaluated the treatment's effectiveness by measuring the patient's anti-Xa level, which is used as a surrogate for rivaroxaban concentration. Anti-Xa levels vary linearly with the concentration of rivaroxaban in plasma over a wide range of drug levels.^4,5

Therapeutic plasma exchange proved successful in decreasing our patient's anti-Xa levels in a relatively short period of time (Fig. 1). During the 5 hours before TPE, in the presence of impaired kidney function, the anti-Xa clearance rate was 0.02 IU/mL/hr. During the 1.5 hours of TPE, the rate increased to 0.13 IU/mL/hr. On the basis of pharmacokinetic data,² this precipitous non-physiologic increase in the rivaroxaban clearance rate likely indicates the transfer of the drug into the waste fluid. The elevated anti-Xa level in our patient's waste fluid indicates that rivaroxaban was removed from the plasma by plasma exchange rather than by physiologic clearance or degradation.

The reversal of Xa inhibitors has been a long-standing subject of investigation. Recently, andexanet alfa, a recombinant Xa molecule and decoy receptor for rivaroxaban and apixaban, has shown promise as an antidote to these anticoagulants.^6–8 In the Andexanet Alfa, a Novel Antidote to the Anticoagulation Effects of FXA Inhibitor Rivaroxaban (ANNEXA-R) trial, andexanet given as a bolus or as a bolus and infusion effectively suppressed anticoagulant therapy almost immediately in healthy older volunteers who were taking rivaroxaban.⁸ However, the U.S. Food and Drug Administration has not approved andexanet for use because there are some important caveats. For instance, the cohort receiving the bolus dose showed evidence of gradual rebound anticoagulation almost immediately after the infusion ended, with return to the individual's baseline elimination curve over time.⁸

In clinical practice, the indications for andexanet include major and potentially fatal bleeding episodes. However, many patients with large-volume, acute bleeding also have substantial kidney dysfunction, liver dysfunction, or both. In healthy people, approximately 33% of plasma rivaroxaban is cleared renally in its active form, and 66% is metabolized through cytochrome P450-dependent mechanisms into an inactive form.⁹ Thus, severe liver and kidney dysfunction can substantially slow the patient's intrinsic rate of rivaroxaban elimination. Because studies show that current reversal agents inhibit the anticoagulant effect without expediting clearance or degradation, the risk of rebound coagulation may be greater in acutely ill patients than in healthy patients. The full effects of rivaroxaban in patients with ESRD are not known. The Rivaroxaban Once Daily Oral Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET-AF) excluded patients with a creatinine clearance <30 mL/min.¹⁰ In other studies, however, severe declines in kidney function significantly prolonged the areas under the plasma concentration-time curve and the effect-time curve.³

Our experimental method of using TPE to reverse the anticoagulant effect of rivaroxaban resulted in a slower initial decrease in anticoagulation than that reported for andexanet. However, we think that TPE has an advantage over andexanet in that it does not rely on the patient's intrinsic kidney and liver function to clear the drug. A second advantage is that, in selected patients at high risk of thromboembolism, TPE would enable a titratable decrease in the anticoagulant effect of rivaroxaban, which may occasionally be preferable to the rapid reversal of anticoagulation seen with andexanet, given recent evidence of thromboembolic events with use of andexanet in high-risk populations.⁷ Third, for patients with severe bleeding, such as that caused by an intracranial hemorrhage, the permanent removal of rivaroxaban by means of TPE considerably reduces the risk of rebound anticoagulation.

To our knowledge, ours is the first report of the use of TPE to acutely reverse the anticoagulant effect of rivaroxaban. Of note, in 2015, Lam and colleagues¹¹ reported a similar use of TPE in the reversal of the anticoagulant effect of apixaban in a patient with cardiac tamponade.

In summary, TPE may provide a valuable option for reversing the anticoagulant effect of rivaroxaban in patients who have large-volume, life-threatening bleeding or who need urgent procedures. In the presence of substantial kidney or liver dysfunction, TPE may provide some benefit over other therapies under development. Further prospective investigation is necessary to validate the use of TPE for the reversal of hemorrhage during direct oral anticoagulant use.