Effect of Therapeutic Plasma Exchange on Apixaban Plasma Levels

Abstract

Objective: Therapeutic plasma exchange (TPE) may enhance the elimination of drugs from human plasma. Removal of apixaban during TPE has not been extensively described previously. Case: This is a retrospective report of a 76-year-old man admitted to the hospital on apixaban with acute worsening of respiratory and bulbar symptoms due to myasthenia gravis. On hospital day (HOD) 1, TPE was initiated for management of myasthenic crisis. Apixaban levels were obtained before and after a TPE session on HOD 3. On HOD 3, the patient’s apixaban plasma level decreased from 88 ng/mL before TPE to 81 ng/mL after. Apixaban displayed an elimination half-life of 65.5 hours with intervening TPE. Discussion/conclusions: We report no clinically significant apixaban removal in a patient undergoing TPE for myasthenic crisis. This differs from previous reports. Supplemental dosing or rescheduling of apixaban doses around TPE sessions may be unnecessary, though more data are needed.

Objective

Therapeutic plasma exchange (TPE) is a therapeutic strategy that involves the removal of a patient’s own plasma, replacing it with either albumin or another source of plasma. It is used to treat various auto-immune and neurologic conditions. Therapeutic plasma exchange is known to remove endogenous substances, including plasma proteins, immunoglobulins, and coagulation factors. Few data exist regarding drug removal by TPE, particularly for those that are plasma protein bound, such as apixaban. We report a patient case that includes serial apixaban plasma level measurements around TPE and review the relevant literature describing apixaban removal by this process.

Case

A 76-year-old man presented to the hospital with acute worsening of respiratory and bulbar symptoms associated with seronegative myasthenia gravis exacerbation. Over the preceding year, the patient had experienced progressive weight loss, dyspnea, dysphagia, and general functional decline. His past medical history also included prostate cancer, hyperthyroidism, hypertension, and gout. Additionally, the patient was anticoagulated with apixaban 5 mg by mouth twice daily for a remote history of deep vein thrombosis (DVT) and pulmonary embolism (PE). Patient clinical data during hospitalization are reported in Table 1.

Table 1.

Patient Clinical Data.

	Hospital day 0	Hospital day 1	Hospital day 2	Hospital day 3	Hospital day 4
Height, cm	193	—	—	—	—
Weight, kg	84.8	—	—	—	—
SCr, mg/dL	0.63	0.52	0.61	0.56	0.51
BUN, mg/dL	25	19	16	14	16
AST/ALT, U/L	41/94	—	—	—	—
Alk phos, U/L	72	—	—	—	—
Total bilirubin, mg/dL	0.7	—	—	—	—
Albumin, g/dL	3.9	—	—	—	—
Hgb	16.8	15.0	15.3	11.7	11.7
PLT	147	130	127	93	96

Abbreviations: Alk phos, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BUN, blood urea nitrogen; Hgb, hemoglobin; PLT, platelets; SCr, serum creatinine.

Due to significant hypercapnia and impending neuromuscular respiratory failure on HOD 0, the patient was transferred to the neurocritical care unit (NCCU) for intubation and mechanical ventilation. Following intubation, a chest radiograph revealed a new right-sided pneumothorax. The patient’s apixaban was discontinued on HOD 2 in anticipation of the need for a chest tube. The patient did not receive bridging anticoagulation leading up to this procedure due to the remote nature of his DVT and PE and presumed low risk for recurrence off apixaban. Multiple apixaban plasma levels were obtained HOD 2-4 to help direct timing of chest tube insertion while minimizing bleed risk. The patient’s plasma apixaban levels were measured using a chromogenic anti-Xa assay (Siemens Healthineers USA) and are reported in Table 2. This chromogenic anti-Xa assay is calibrated with an apixaban standard to provide an accurate quantification of apixaban drug concentration in plasma. ¹ The patient’s initial trough apixaban plasma level on HOD 2 was therapeutic at 142 ng/mL (expected apixaban trough levels 22-177 ng/mL).

Table 2.

Apixaban Plasma Levels.

Hospital day (time)	Apixaban level (ng/mL)	T1/2 (hours)
2 (1957)	142	—
3 (0537)	88	14.0
3 (1327)	81	65.5
4 (0544)	61	39.8
4 (2332)	51	68.9

Abbreviations: T_1/2, half-life.

Last apixaban dose administered HOD 2 at 0828. Expected peak levels 59-302 ng/ml and trough levels 22-177 ng/ml for chromogenic anti-Xa assay. TPE was performed on HOD 3 from 0845 to 1017.

For management of myasthenic crisis, a femoral dialysis catheter was placed on HOD 1 for initiation of TPE, and the patient’s first TPE session was completed without complication. On HOD 3, a second TPE session with exchange of one plasma volume was performed over 92 minutes. Collected plasma volume was 3149 mL and replacement fluid included 2900 mL of 5% human albumin. Anticoagulation during the procedure consisted of 101 mL of Anticoagulant Citrate Dextrose Solution, Solution A. The patient remained off apixaban or any systemic bridging anticoagulation at this time. The patient tolerated this second TPE session without complication. The patient’s apixaban plasma level negligibly decreased from 88 ng/mL before TPE to 81 ng/mL after. Timing of TPE in relation to apixaban levels is depicted in Figure 1.

Figure 1.

Apixaban plasma levels during TPE.

The patient’s chest tube was ultimately placed on HOD 4 following a measured apixaban plasma level of 61 ng/mL. After chest tube insertion, anticoagulation was reinstituted with a heparin infusion on HOD 5, which was maintained through most of the patient’s remaining TPE treatment course and NCCU stay. The patient received a total of 5 TPE sessions while hospitalized, with a final session completed on HOD 11. Additional treatments for myasthenia included immunomodulation with corticosteroids, multiple courses of intravenous immune globulin, and rituximab. Pyridostigmine was also administered for symptomatic management. Once medically stabilized, the patient’s heparin infusion was transitioned back to apixaban on HOD 39. After prolonged hospitalization of approximately 2 months duration, the patient was transferred to an inpatient rehabilitation facility.

Discussion/Conclusion

Therapeutic plasma exchange is a therapeutic strategy that involves the removal of a patient’s own plasma, replacing it with either albumin or another source of plasma. Therapeutic plasma exchange can be used for a variety of indications such as myasthenia gravis, desensitization before certain organ transplantation, and thrombocytopenic purpura. ² Therapeutic plasma exchange is known to remove endogenous substances such as coagulation factors (antithrombin III, prothrombin, factor VIII, fibrinogen, and platelets), ³ yet little data exists regarding drug removal. This can often present a clinical conundrum when trying to evaluate the need for supplemental dosing or alteration of drug administration timing around TPE sessions. Those drugs with narrow therapeutic indexes pose a larger concern, as removal during TPE may result in levels falling below therapeutic drug concentrations, leading to negative outcomes for the patient. High-risk medications such as therapeutic anticoagulants present such a situation with an increased risk of thrombosis with subtherapeutic drug concentrations and increased risk of bleeding with supratherapeutic drug concentrations.

In general, one expects those agents with high protein binding (>80%) and low volumes of distribution (<0.3 L/kg) to be most at risk of significant drug removal during TPE sessions. ⁴ Ideally serum drug concentrations can be utilized to evaluate if significant drug removal is taking place. Unfortunately, drug concentration assays are not available for all medications. In cases that drug concentration monitoring is attainable, one must consider timing of dose in relation to TPE sessions, as well as timing of levels after TPE sessions to account for a redistribution phase of the drug. In this specific case, apixaban has high protein binding at 87% and a relatively low volume of distribution of 21 liters. ⁵ This would potentially make apixaban an agent that is readily removed by TPE.

Most data we have regarding drug removal in TPE are in the form of case reports. To date we are aware of 4 case reports specifically addressing the use of apixaban with TPE treatment.^{6 -9} A recent case report by Holt and colleagues describes a patient who had been receiving plasma exchange for thrombotic thrombocytopenic purpura. ⁶ The patient was admitted to the hospital within 48 hours of their last plasma exchange session because of an acute ischemic stroke. The authors hypothesized that the patient’s apixaban levels, which he had been taking for stroke risk secondary to atrial fibrillation, may have been altered by his plasma exchange session prior to admission. Unfortunately, this case report was not able to provide real time apixaban anti-Xa levels to support this hypothesis. Alternatively, the authors also disclosed that the patient would sometimes miss apixaban doses.

Lam and colleagues describe an 82-year-old man who underwent TPE to decrease the anticoagulation effect of apixaban prior to a pericardiocentesis due to potential bleeding. ⁷ Of note, the patient’s serum creatinine was elevated from his admission baseline, indicating an acute kidney injury. Since apixaban is partially renally eliminated (27% of total clearance), the patient’s reduced renal function created potential for decreased apixaban clearance. ⁵ The patient declined use of prothrombin complex concentrate for apixaban reversal, but instead agreed to receiving a single 2-hour TPE session. Unfortunately, the authors were only able to report anti-Xa levels calibrated for unfractionated heparin (UFH) and low-molecular-weight heparin (LMWH) around TPE. The author’s only comment was that the patient’s initial anti-Xa levels for UFH and LMWH were elevated roughly a day after the last dose of apixaban was administered. Subsequently, the authors reported that TPE reduced the plasma anti-Xa levels for UFH and LMWH. This case also reported a LMWH anti-Xa level of 0.82 IU/mL in the plasma waste removed during TPE, which was noted to be similar to the patient’s pre-TPE plasma LMWH anti-Xa level. The patient underwent a successful pericardiocentesis after the TPE session.

A case by Francisco and colleagues describes a 63-year-old woman on apixaban therapy who developed an intestinal obstruction requiring surgical intervention. ⁸ The patient’s clinical course was also complicated by hemorrhage and need for continuous renal replacement therapy (CRRT). As discussed previously, apixaban is partially renally eliminated. Additionally, apixaban undergoes some fecal elimination. ⁵ The patient’s need for CRRT and her intestinal obstruction likely contributed to decreased apixaban clearance and elevated apixaban anti-Xa levels several days after the last apixaban administration. The authors report a nearly 37% reduction in apixaban anti-Xa levels after the first session of TPE. Although the exact anti-Xa level was not included, the patient’s apixaban anti-Xa level remained elevated the next day, and another session of TPE was used to expedite further clearance. This case report demonstrated a decrease in apixaban anti-Xa levels from 172 to 87 ng/mL with the use of 2 separate sessions of TPE over the course of 2 days.

Finally, Pilková and colleagues ⁹ present a case of a 76-year-old man admitted for dysphagia secondary to myasthenia gravis. The patient’s clinical course included 5 TPE sessions occurring every other day. The patient had been taking apixaban 5 mg twice daily for a history of PE. This case report differed from previous reports as it aimed to describe the pharmacokinetics of apixaban around TPE sessions. The authors were able to obtain serum apixaban levels as well as levels from the TPE discharge (plasma and fluid to flush the circuit). A calculated apixaban concentration of 365 ng/mL in the removed plasma matched the patient’s plasma concentration post-TPE. The authors estimated that TPE removed only approximately 1 mg of apixaban, or 10% of a total daily dose. Pilkova and colleagues concluded that apixaban pharmacokinetics were similar around TPE sessions as they were in the absence of TPE sessions. They proposed that, although further data may be needed, apixaban could prove a better option during TPE compared to other anticoagulants, such as low-molecular-weight heparins and unfractionated heparin. The authors did also note that by day 14 the patient did not have either bleeding or clotting complications.

We did not observe that apixaban was significantly removed by TPE in the case we present (Table 2 and Figure 1). The patient received their last apixaban dose on HOD 2 at 0828. On HOD 3, TPE was performed for 92 minutes from 0845 to 1017. A plasma apixaban level of 88 ng/mL was measured at 0537 prior to TPE. Several hours after TPE, a plasma apixaban level of 81 ng/mL was measured at 1327, which remained therapeutic (expected apixaban peak levels 59-302 ng/ml and trough levels 22-177 ng/mL). Apixaban half-life calculations were performed using standard equations for drugs following first-order kinetic models. ¹⁰ The patient’s apixaban levels demonstrated an elimination half-life of 65.5 hours with intervening TPE. The patient’s apixaban half-life on HOD 3 exceeded the normally expected 12 hours and slow elimination persisted on HOD 4 without TPE. ⁵ Our case lacked any obvious explanation for delayed apixaban clearance HOD 3-4, such as renal or hepatic dysfunction. The patient did not receive any concurrent medications that would have inhibited apixaban elimination.

Our case differs from the 4 cases previously described in several important ways. Our patient did not have supratherapeutic apixaban anti-Xa levels or bleeding necessitating initiation of TPE. Our patient did not appear to have any reason for delayed clearance of apixaban such as the renal dysfunction or bowl obstruction reported in other cases. Additionally, the case report by Lam and colleagues used anti-Xa levels that were not calibrated specifically to apixaban. This could have confounded the observed impact of TPE on drug removal, whereas our case involved chromogenic anti-Xa assays calibrated specifically for apixaban. While Francisco and colleagues did not report the specific timing of apixaban anti-Xa levels in relation to TPE, Lam and colleagues evaluated an anti-Xa immediately after TPE. Anti-Xa levels may be lower prior to redistribution of drug after TPE, which could lead to overestimation of clearance. Our post-TPE plasma apixaban level was collected over 3 hours after completion of TPE, which should have allowed for adequate drug redistribution. Finally, 2 cases reported use of plasma as their replacement fluid during TPE whereas our case used albumin. Factor X is present in plasma which could theoretically have impacted the post-TPE lab values in other cases. We would like to note that the lack of apixaban clearance in our case appears similar to the case presented by Pilkova and colleagues.

Conclusion

Although prior cases have reported the use of TPE for urgent removal of apixaban, we observed non-clinically significant drug removal of apixaban during TPE. Further studies involving pre- and post-TPE drug levels are needed to fully understand TPE’s role in apixaban removal. More robust pharmacokinetic studies utilizing drug levels would greatly add to clinical decision making when dosing medications during TPE.