Dabigatran-reversal failure using standard dose of idarucizumab: a systematic review and meta-analysis of cases

Graphical abstract

Idarucizumab is a humanized monoclonal antibody fragment targeting dabigatran with a very high affinity approximately 300 times greater than the affinity of dabigatran for thrombin, thus potently and specifically neutralizing dabigatran and its anticoagulant effect [1,2]. Accordingly, idarucizumab is the specific antidote used for dabigatran reversal in patients with life-threatening/uncontrolled bleeding or those requiring an urgent or invasive procedure. In phase III of the Reversal of Dabigatran Anticoagulant Effect With Idarucizumab (RE-VERSE AD) trial [3], 2 consecutive doses of 2.5 g of idarucizumab within 15 minutes showed a median maximum percent reversal of 100% after 4 hours. This perfect reversal is explained by the fact that a full dose of 5 g theoretically inhibits 700 to 980 ng/mL of unbound dabigatran [4], whereas the median dabigatran plasma levels in the RE-VERSE AD cohort were 110 ng/mL in patients who experienced serious bleeding (group A, n = 301) and 74 ng/mL in patients who required invasive procedures (group B, n = 202) [3]. In clinical practice, contrasting results were reported with several patients experiencing plasma dabigatran rebound defined as a re-elevation of dabigatran level after a successful idarucizumab reversal [5,6]. Previous studies showed that a baseline (before reversal) dabigatran level of >200 ng/mL [4,7] could predict the reappearance or rebound of dabigatran and was associated with less hemostatic effectiveness [7]. After injection, idarucizumab binds the unbound dabigatran in plasma because of its limited distribution volume. In contrast, dabigatran presents a larger distribution volume and a longer elimination half-life than idarucizumab. After elimination or saturation of idarucizumab, dabigatran can freely move from the extravascular compartment to the plasma compartment to re-establish equilibrium [8]. This difference in pharmacokinetic properties probably explains the rebound mechanism. In addition, idarucizumab efficacy seems to be limited with respect to achieving complete reversal in patients with a substantial dabigatran overdose with a detectable dabigatran level despite idarucizumab administration. There is a dose-response relation between dabigatran level at baseline and the ability of idarucizumab to achieve complete reversal [4]. However, no predictive threshold of dabigatran-reversal failure has been published. We aimed to describe all cases of dabigatran-reversal failure in the literature to characterize patients at risk of reversal failure in emergency situations after a complete course of idarucizumab administration.

We performed a systematic literature review in PubMed using the term “dabigatran reversal” covering the period from January 1, 2015, to December 31, 2022, according to the Preferred Reporting in Systematic Reviews and Meta-Analysis Protocols guidelines. We included studies with patients benefiting from idarucizumab for dabigatran reversal and with a reversal failure. To allow comparison, an unbound dabigatran level of >20 ng/mL was considered detectable, and dabigatran determination was performed with ecarin clotting time or diluted thrombin time, as in RE-VERSE AD [3]. We only included articles written in English or French with available individual data. Two investigators (S.M. and N.G.) assessed the articles for eligibility and data extraction. First, we excluded articles based on titles and abstract in order to select eligible articles for full text review. We notably included an article matching the inclusion criteria but presenting missing data, thanks to the author who responded favorably to complete data [9]. Continuous data were expressed as median (5th to 95th percentiles) range.

From 623 articles identified, 528 were excluded on the basis of their title and abstract, and finally, only 4 articles, corresponding to 5 patients, matched our inclusion criteria (Figure A). Two additional patients published in the literature by our center met the inclusion criteria, thanks to the availability of clinical and biological data [7,10].

Figure.

Dabigatran-reversal failure using standard dose of idarucizumab. (A) Flowchart of article selection. (B) Unbound dabigatran level of 7 patients assessed at admission after a first idarucizumab administration of 2.5 g ×2 and at the time of the rebound. All assays were performed with ecarin clotting time or diluted thrombin time (dTT). (C) Evolution of dabigatran levels and endogenous thrombin potential (ETP) in 1 patient followed up for 19 days in our center. Liquid chromatography–tandem mass spectrometry (LC/MS) was performed to assess plasma total and free dabigatran. dTT was used to estimate the unbound dabigatran level on plasma, and ETP was assessed with thrombin generation assay. The yellow area corresponds to a normal ETP range of 1200 to 1500 nM/min obtained from 30 healthy subjects.

In total, 7 patients were included in our analysis [7,[9], [10], [11], [12], [13]]. The median age was 74.0 years (62.0-86.0 years), and all patients were treated with dabigatran for atrial fibrillation (Table). At admission, all patients had renal impairment, with a median creatinine level of 354 μmol/L (151-1030 μmol/L) for 6 patients and an estimated glomerular filtration rate of 19 mL/min for 1 patient [11]. At baseline, dabigatran plasma level was >1000 ng/mL for all patients except 1 (85.7%), and the median level was 1456 ng/mL (555-3337 ng/mL). After reversal, median dabigatran level was 73 ng/mL (46-1440 ng/mL), with a median delay of 1.0 hour (0.75-16.00 hours) between idarucizumab administration and dabigatran measurement. The median percentage of reversal was 89.6% (49.6%-97.5%). The evolution of dabigatran levels after reversal is illustrated in Figure B. After reversal, none of the 3 patients with major bleeding achieved clinical hemostatic efficacy [9,11,12] according to the International Society on Thrombosis and Haemostasis criteria [14]. One patient had perioperative fatal bleeding [12]. Four patients among 5 with available data (80.0%) received >5 g of idarucizumab. Death occurred in 3 (42.9%) patients due to uncontrollable bleeding [12] or circulatory shock [9]. Of note, in 1 case published by our center [10], liquid chromatography–tandem mass spectrometry and thrombin generation assay were performed. Thrombin generation assay was triggered with 1 pM of tissue factor (PPP-Reagent LOW, Diagnostica Stago). Thrombin generation was measured by Calibrated Automated Thrombography and Fluorocan Ascent Fluorometer (Thermoscientific Labsystems). The evolution of different dabigatran levels and endogenous thrombin potential is illustrated in Figure C. Endogenous thrombin potential was clearly inhibited at a high dabigatran level and restored but exceeded the normal range after reversal, as recently described in dabigatran reversal using prothrombin complex concentrate [15]. During rebound, the release of total dabigatran probably corresponded to free dabigatran and dabigatran bound to proteins and/or idarucizumab. Similar results were observed in the phase I trial of idarucizumab [2].

Table.

Clinical and biological characteristics of patients with idarucizumab failure.

Clinical and biological characteristics	Patients (n = 7)	Available data
Age (y), median (range)	74.0 (62.0-86.0)	7
Male, n (%)	4 (57.1)	7
Dose of dabigatran, n (%)		6
150 mg twice a day	5 (83.3)
110 mg twice a day	1 (16.7)
Idarucizumab indication, n (%)		7
Urgent procedure	2 (28.6)
Bleedinga	4 (57.1)
Overdose without bleeding	1 (14.3)
Biology at baseline, median (range)
Creatinine (μmol/L)	354 (151-1030)	6
Hemoglobin (g/L)	81.0 (59.0-92.0)	6
aPTT (s)	150 (113-169)	4
Dabigatran (ng/mL)	1456 (555-3337)	7
Biology after reversal
aPTT (s), median (range)	75 (41-150)	4
Dabigatran (ng/mL), median (range)	73 (46-1440)	7
Delay of dabigatran monitoring after idarucizumab injection (h), median (range)	1.0 (0.75-16.0)	7
Plasma dabigatran rebound, n (%)	6 (100.0)	6
Dabigatran level at rebound (ng/mL), median (range)	728 (142-1586)	6
Outcomes after reversal, n (%)
Need for additional doses of idarucizumab	4 (80.0)	5
Postoperative bleeding	1 (100.0)	1
Hemostatic effectivenessb	0 (0.0)	3
Death	3 (42.9)	7

Data were expressed as median and (5th to 95th percentiles) range.

aPTT, activated partial thromboplastin time.

^aOne major intra-abdominal bleed and 2 major and 1 minor gastrointestinal bleeds.

^bAmong the 3 patients with major bleeding related and according to International Society on Haemostasis and Thrombosis criteria [14].

Several cases of successful idarucizumab reversal in patients with considerable overdose have been reported, but these patients had substantial or multiple rebounds after reversal [5,16]. A case of successful dabigatran reversal by idarucizumab [17] was described in a child with an estimated initial dabigatran level ranging from 1507 to 2040 ng/mL after ingestion of between 30 and 50 capsules of dabigatran 150 mg. No rebound was observed and his renal function was normal. Also, we observed that a 5-g dose of idarucizumab was sufficient to completely reverse dabigatran in patients with normal to severe impairment [18]. However, it appeared that higher initial dabigatran levels were associated with increased risk of reversal failure.

Overall, our study suggested that idarucizumab failure could be suspected in patients with a dabigatran level of >1000 ng/mL, but this risk should be weighed against renal function and the degree of overdose. The dose of 5 g was calculated based on the 99th percentile of a population with a dabigatran level associated with therapeutic dose and without severe renal impairment [19]. In RE-VERSE AD, 15 patients presented with dabigatran levels at baseline of >1000 ng/mL and all presented with renal impairment, including 11 patients with an estimated creatinine clearance of <30 mL/min [18]. In addition, the 3 patients with the highest baseline dabigatran levels all presented with idarucizumab failure and rebound of dabigatran levels. We confirmed that initial determination of dabigatran plasma level is of interest to predict biological and clinical outcomes in patients receiving idarucizumab, as suggested recently by guidelines of the International Council for Standardization in Haematology [20]. We acknowledge several limitations, including the small number of included patients, the lack of complete clinical data about bleeding and outcomes, and the impossibility to estimate frequency of failure due to spontaneous declaration of cases. Moreover, over the same study period, in the Hôpital Européen Georges Pompidou (Paris, France) and the Bichat–Claude-Bernard hospitals (Paris, France), there were 36 and 10 patients, respectively, that required idarucizumab for reversal with dabigatran level measurement prior to and/or after reversal and 0 and 2 patients, respectively, experienced incomplete reversal. Thus, we calculated a frequency of reversal failure of 4.3% in accordance with the Danish study by Haastrup et al. [9] that described 1 case of idarucizumab reversal failure (4.2%) among 24 patients with dabigatran level monitoring.

In summary, despite the small number of cases included, we observed that patients with dabigatran overdose and renal impairment are at high risk of idarucizumab failure with a conventional 2.5-g ×2 dose. Indeed, idarucizumab should be injected without awaiting the laboratory results. Nevertheless, determination of dabigatran level at admission could greatly help in anticipating dabigatran-reversal failure and/or rebound. Then, in case of a baseline dabigatran level of >1000 ng/mL, close monitoring of dabigatran should be performed in order to reduce the risk of failure and/or rebound, and an additional idarucizumab dose may be considered, eg, in case of persistent uncontrolled bleeding or need for an urgent invasive procedure.