Four-factor prothrombin complex concentrate improves thrombin generation and prothrombin time in patients with bleeding complications related to rivaroxaban: a single-center pilot trial

Background Direct oral anticoagulants (DOACs) pose a great challenge for physicians in life-threatening bleeding events. The aim of this study was to test the efficacy of reversing the DOAC rivaroxaban using four-factor PCC (prothrombin complex concentrate), a non-specific reversing agent. Methods Patients with life-threatening bleeding events during rivaroxaban treatment were included and administered 25 U kg−1 of PCC. Blood samples were collected immediately prior to as well as after PCC treatment at predefined time intervals. The primary endpoint was defined as the difference in thrombin generation (TG) parameters ETP (endogenous thrombin potential) and Cmax (peak thrombin generation) prior to and ten minutes subsequent to PCC treatment. Results Thirteen patients, of whom the majority suffered from intra-cranial haemorrhage (ICH) or subdural haemorrhage (SDH), were included and administered PCC. The results show that the ETP (TG) significantly (p = 0.001) improved by 68% and Cmax (TG) by 54% (p = 0.001) during PCC treatment. In addition, the Quick value (prothrombin time: QuickPT) significantly improved by 28% and the activated partial thromboplastin time (aPTT) was decreased by 7% ten minutes after PCC administration. Cmax was reduced at baseline, but not ETP, aPTT or QuickPT. Lag time until initiation (TG, tlag), thromboelastometry clotting time (CTEXTEM) and time to peak (TG, tmax) correlated best with measured rivaroxaban levels and were out of normal ranges at baseline, but did not improve after PCC administration. In 77% of the patients bleeding (ICH/SDH-progression) ceased following PCC administration. During the study three participants passed away due to other complications not related to PCC treatment. The possibility of thrombosis formation was also evaluated seven days after administering PCC and no thromboses were found. Conclusions This study shows that use of PCC improved ETP, Cmax, QuickPT and aPTT. However, of these parameters, only Cmax was reduced at the defined baseline. It can be concluded that CTEXTEM, tlag and tmax correlated best with the measured rivaroxaban levels. The study drug was found to be safe. Nonetheless, additional studies specifically targeting assessment of clinical endpoints should be performed to further confirm these findings. Clinical trial registration EudraCT trial No. 2013–004484-31. Electronic supplementary material The online version of this article (10.1186/s12959-017-0158-9) contains supplementary material, which is available to authorized users.


Background
Rivaroxaban (Xarelto®, Bayer, Germany) is used as thrombosis prophylaxis and therapy instead of anticoagulants such as unfractionated heparin, low molecular weight heparin or vitamin K antagonists, because it is considered to have a wider therapeutic range and a more predictable dose-response [1][2][3]. Although bleeding and bleeding complications in trauma and surgery are the most common complications/side-effects [4], no evidence-based reversal strategy is available for a bleeding patient taking rivaroxaban [5,6]. A specific antidote called Andexanet alfa is in development, but not yet clinically available [7]. Therefore, the off-label use of coagulation factor concentrates, particularly four-factor prothrombin complex concentrate, in the management of rivaroxaban-related, life-threatening bleeding is the most recommended/used therapy [8,9]: In healthy male volunteers, PCC was seen to reverse the anticoagulant effect of rivaroxaban [10,11]. Especially for the treatment of intracranial haemorrhage (ICH), PCC seems to be a promising potential reversing agent [12][13][14]. However, managing life-threatening bleeding events under DOAC remains a challenge for physicians since in vivo data regarding efficacy of anticoagulation reversal are still lacking.
Therefore, the aim of this study was to evaluate the effect of PCC on coagulation status after life-threatening bleeding events in patients treated with rivaroxaban. Written informed consent was obtained from all study participants. Patients, who for medical reasons (unconscious, shock, sedated, etc.) were not able to give their consent to participate in this clinical trial, were enrolled based on the principle of deferred consent according to § 43a (1), Austrian Medicinal Products Law ("Österreichisches Arzneimittelgesetz").

Ethics committee approval
The study was performed in compliance with the Declaration of Helsinki guidelines regarding ethical principles for medical research involving human subjects and followed Good Clinical Practice as defined by the International Conference on Harmonization (ICH-GCP).

Study design and setting
The present study is a single-centre, analytic, observational, prospective, open-labelled, single-armed, non-commercial clinical pilot trial.

Selection of participants
Inclusion criteria were age over 18 years, patients receiving rivaroxaban and with life-threatening bleeding and the need for reversal and/or patients needing acute reversal of the anticoagulation effects of rivaroxaban. Patients who had already been administered pro-coagulant therapies including recombinant activated factor seven, activated prothrombin complex concentrates or other coagulation factor concentrates (except fibrinogen concentrate, DDAVP and tranexamic acid) were not included. Further exclusion criteria were: a greater risk for thromboembolic events than for bleeding, pregnancy, suspected or confirmed sepsis, recent history of thromboembolic events, and active participation in another clinical trial or refusal to participate.

Interventions
The investigational medicinal product (four-factor PCC) was generously provided by CSL Behring, Marburg, Germany (Beriplex® P/N 500 IU). PCC was administered i.v. at a dosage of 25 IU/kg BW as bolus and/or continuous infusion. If bleeding continued after the first dosage, a second dosage of 25 IU/kg BW had to be administered. However, for all of the included patients only the first dosage of PCC was necessary. The maximum total allowed dose was 5000 IU.
Study assessments were performed as follows: Visit 1 (V1, baseline) was performed immediately before administration of four-factor PCC (25 IU/kg BW). Ten minutes after the end of PCC administration Visit 2 was conducted (V2). Thereafter, Visits were calculated as V3: +1 h, V4: +3 h, V5: +6 h, V6: +12 h, V7: +24 h, V8: +2 days, V9: +4 days, V10: +7 days, V11 + 30 days. At Visit 11 (30 days after V2) the patient was interviewed to evaluate his health status. Between days 5 and 7 an ultrasound examination of the leg arteries and veins was performed to screen for thrombosis. At every visit bleeding and thrombosis status were assessed, laboratory measurements performed and vital signs documented.
Blood coagulation assays were performed and blood coagulation factor concentrations determined using vacutainer tubes containing 1.106 mol l −1 trisodium citrate solution; blood samples for blood cell count were taken using vacutainer tubes containing 1.6 mg ml −1 3 K EDTA. All blood sampling tubes were purchased from Sarstedt, Nümbrecht, Germany.
All coagulation analyses were performed immediately after blood draw, except thrombin generation. Plasma for thrombin generation assays was immediately frozen at −80°C (or at −20°C for a maximum of seven days) and thawed immediately before being analysed.

Methods and measurements
ROTEM® parameters were determined using a ROTEM® gamma analyser (TEM Innovations GmbH, Munich, Germany). All tests were performed according to manufacturer's instructions in whole blood immediately after blood sampling and using the specific liquid reagents provided by the manufacturer for EXTEM (extrinsically activated assay with tissue factor) and INTEM (intrinsically activated test using kaolin) measurements.
Thrombin generation measurements were performed using the Innovance ETP assay (Siemens, Marburg, Germany) on an automated coagulation analyser (BCS XP, Siemens, Marburg, Germany). Coagulation was activated by adding phospholipids, human recombinant tissue factor (in a concentration of 600-850 pmol l −1 ) and calcium ions to platelet-poor plasma. The generated thrombin cleaves a chromogenic substrate (H-b-Ala-Gly-Arg-pNA) and the turnover of the substrate is recorded over time. The final substrate concentration is 733 nmol l −1 and 19 mmol l −1 CaCl. The original curve is corrected for the estimated α macroglobulin-bound thrombin activity. From this curve the following parameters can be obtained: total amount of generated thrombin present in the reaction from the point of initiation until return to baseline, which is measured by the calculated increase in the extinction rate (measured in milliextinctions or mE), and also known as "endogenous thrombin potential" (ETP), sometimes referred to as "area under the curve" (AUC), peak thrombin generation (C max ) (the maximum of the first derivation of the ETP), lag phase until initiation (t lag ), and time to peak thrombin activity (t max ).

Outcomes
Primary objective was assessment of the coagulation status after administration of PCC. Coagulation measurements included PT, aPTT, endogenous thrombin potential (ETP), thromboelastometry (ROTEM®), single factor concentrations, blood cell count and the rivaroxaban level.
Secondary endpoints were the development of the following parameters from V1 to V10 and their correlation with measured rivaroxaban levels: thrombin generation, single factor profiles, standard coagulation tests PT, aPTT, fibrinogen, AT, and thromboelastometry (ROTEM®).

Analysis
A descriptive analysis of all measured blood characteristics was performed. The Wilcoxon signed-rank test was used to evaluate the primary endpoint and all other differences in coagulation measurements between baseline and samples from the same patient at V2 to V10. To evaluate correlations between rivaroxaban level and blood coagulation assays and coagulation factor activities Pearson's correlation was used. P values ≤0.05 were considered statistically significant. Statistical analyses were performed using STATISTICA 10 software (StatSoft Europe GmbH, Hamburg, Germany).

Characteristics of study subjects
Fourteen patients (5 female, 9 male) aged between 47 and 96 years (median 80 years) were enrolled at the Medical University of Innsbruck. One patient (male) had to be excluded from analysis because he withdrew consent. Median weight was 76 kg (range 52 kg to 99 kg) and median height 170 cm (range 156 cm to 183 cm). Average body mass index was 25 (range 18 to 34).
The first visit to the first patient (FPFV) was in August 2014, last patient last visit (LPLV) was in October 2016. Underlying disease was mostly intracerebral haemorrhage (ICH, six patients), four patients suffered from subdural hematoma (SDH), one patient suffered from haemorrhage after removing the urinary catheter, one during laparotomy, and one suffered from gastrointestinal bleeding.
Median amount of administered PCC was 2000 IU (range 1500 to 2400 IU).

Safety
Three subjects died during study participation, one from septic shock (multi-organ failure), one due to progressive cancer and one from ICH. All deaths were considered to be unrelated to the study medication. Of the thirteen study patients three showed signs of re-bleeding (progressive ICH) after administration of PCC. Other serious adverse events (SAEs) were an epileptic event, ischemia caused by thromboembolic closure of the femoral artery bifurcation (left), sepsis and embolic arteria cerebri anterior infarct (one patient). Both thromboembolic SAEs occurred only after day 7 (thrombotic screening via duplex ultrasound).

Main results
Baseline For baseline values of laboratory parameters other than coagulation measurements, please refer to Table 1.
Mean rivaroxaban level at V1 was 120 ng ml −1 (95% confidential interval 48 to 192 ng ml −1 ). The course of measured rivaroxaban levels can be seen in Fig. 1. Standard laboratory parameters aPTT and PT (Quick value) were in normal range. Thromboelastometry EXTEM clotting time (CT EXTEM ) was prolonged, C max (TG) was reduced and ETP (TG) was in normal range. Moreover, C-reactive protein (CRP) levels were increased, as were d-dimers, FVIII levels and vWF levels, whereas erythrocyte and leukocyte counts were below reference ranges for healthy subjects. Blood loss was 800 and 750 cm 3 for patients who were bleeding externally, whereas in the case of ICH intracerebral blood volume was 0.02 cm 3 , 8 cm 3 , 12 cm 3 , 34 cm 3 , 43 cm 3 and 99 cm 3 , respectively. Regarding SDH, blood loss volume was 3 cm 3 , 11 cm 3 , 33 cm 3 and 116 cm 3 , respectively. In the case of gastrointestinal bleeding blood loss was not measured. All baseline coagulation laboratory measurements can be seen from Table 2.

Primary endpoint
All included patients reached the primary endpoint (difference in thrombin generation between V1 and V2). The ETP significantly (p = 0.001) increased by 68% as a result of administration of PCC (median amount of administered PCC: 2000 IU [range 1500-2400 IU]) from 316 mE at V1 to 530 mE at V2. C max significantly (p = 0.001) improved by 54% from 80 mE min −1 (V1) to 123 mE min −1 (V2). t lag and t max did not show a significant difference between V1 and V2 ( Fig. 2a and b).

Secondary endpoints
Regarding thrombin generation in addition to reaching the primary endpoint (compare above), the ETP remained significantly elevated until 48 h after PCC administration (V8) and C max until 12 h after PCC administration (V6). Compare Table 1 and Fig. 2a/b.
None of the measured ROTEM® parameters significantly improved after administration of PCC. MCF INTEM significantly decreased by 3% (p = 0.019) after PCC administration and remained significantly decreased until one hour after PCC administration (V3, data not shown). CT EXTEM was significantly decreased only after V5 (compare Fig. 2c). Assessment of coagulation status following the administration of PCC showed a significant increase in Quick PT (+28%) ten minutes after PCC administration (V2). The Quick value remained significantly increased, as compared to baseline (V1), until V9 or four days after administration of PCC (compare Fig.  2d). Moreover, PCC caused a significant decrease of 7% in aPTT. aPTT also remained significantly decreased until V2 (compare Table 2). The single factor profiles showed a significant increase in activation of all coagulation factors contained in PCC (factors II, VII, IX and X) until V3 (FVII), V4 (FII) and until V7 (FIX and FX) as compared to baseline V1. All other coagulation factors showed no increase in activity following PCC administration (compare Additional file 1: Table S1).
Regarding correlation of rivaroxaban level with coagulation parameter, best correlation was found between t lag and rivaroxaban level (single measurements for every patient from V1 to V10 correlated with the respective rivaroxaban level) followed by CT EXTEM , CT INTEM , Quick value, and C max , MCF EXTEM , and t lag . All other coagulation parameters showed no significant correlation with rivaroxaban levels (compare Table 3).
When comparing mean values of rivaroxaban and the mean values of all measured coagulation parameters at every time point (V1 -V10), t lag , CT EXTEM , t max , MaxVt FIBTEM , MaxV-t EXTEM , MCF FIBTEM , CT INTEM and FI clauss showed a significant correlation with rivaroxaban levels. All other coagulation parameters showed no significant correlations with rivaroxaban levels (compare    Fig. 3). For values of all measured laboratory parameters, please refer to Additional file 1: Table S1.

Discussion
The present study enrolled patients with the need for acute reversal of the effects of rivaroxaban to investigate the effects of PCC on the coagulation status. The majority of included patients (77%) suffered from lifethreatening bleeding events in the central nervous system (ICH or SDH). Administration of PCC resulted in improved thrombin generation (C max , ETP), aPTT and Quick PT , whereas ETP, aPTT and Quick PT were in normal range already before treatment. Therefore, C max was the only parameter reduced at baseline (out of normal range) that improved by PCC administration. T lag , CT EX-TEM and t max correlated best with measured rivaroxaban levels and were out of normal ranges at baseline, but did not respond to PCC administration. Study design and the small number of included patients do not allow a profound safety evaluation to be conducted. However, ultrasound examination seven days after administration of four-factor PCC revealed no thrombotic events. The occurrence of thrombosis found in the present study was 14% during the observation period (30 days), which is comparable to the frequency found in the ANNEXA-4 study with an incidence of 18% [21].
To date, no studies regarding the efficacy of PCC in patients with life-threatening bleeding events, associated with DOACs, are available. Yoshimura and colleagues recently published data from an observational study evaluating PCC in ten patients with bleeding complications associated with DOAC treatment: PT also significantly improved after PCC administration, but in contrast to our study they found no significant changes in aPTT [22]. This might be attributed to the lower dose of PCC administered (median 1000 IU vs. 2000 IU in the present study). A recent retrospective study by Grandhi and colleagues investigated the effect of PCC on ICH in 16 patients under rivaroxaban treatment. They collected no data about coagulation status but found PCC to be effective in reducing haemorrhagic complications and hematoma expansion [12]. Erenberg and colleagues conducted a randomized, double-blind, placebo-controlled study with 12 healthy volunteers who Fig. 2 Effect of four-factor PCC on coagulation parameters. ETP (a), C max (b), CT EXTEM (c) and Quick PT (d) for all patients from V1 to V10. Normal (healthy persons without anticoagulation) values are indicated (red area). Significant differences from baseline were determined using the Wilcoxon signed-rank test (*p < 0.05, **p < 0.01). Abbreviations: ETP, endogenous thrombin potential; C max , thrombin generation, peak thrombin generation; CT, ROTEM®, clotting time; EXTEM, ROTEM®, extrinsic coagulation pathway; Quick PT , Quick value (prothrombin time) received rivaroxaban for two and a half days (2 × 20 mg) and subsequently a PCC bolus (50 U kg −1 ): effects of rivaroxaban and PCC were evaluated using the PT and ETP. Both parameters were affected by rivaroxaban and improved by PCC [10]. A similar study set up by Levi and colleagues with 35 healthy volunteers, who were randomized into a 4-factor PCC, a 3-factor PCC (50 U kg −1 for both) and a placebo (saline) group after four days of rivaroxaban (2 × 20 mg) treatment, also revealed that 4-factor PCC improved PT, aPTT and ETP [18]. In a recent study, we already demonstrated the effects of PCC on CT EXTEM and TG ex vivo [16]: PCC was able to improve both parameters ex vivo, and rivaroxaban levels correlated with Quick PT , CTs, C max , aPTT and t lag .
The main difference between the present study and previous investigations is the different study set-up (in vivo). Consequently, rivaroxaban levels were more heterogeneous than in controlled ex vivo studies with patient blood or in vitro studies with healthy volunteers. For example, in the present study ETP was not reduced (as compared to normal values) at baseline V1, but the exact time of rivaroxaban intake was mostly not known. Consequently, rivaroxaban levels were lower than in comparable studies and ETP was not impaired.
The majority of patients were admitted with acute intracerebral haemorrhage. Early hematoma enlargement occurs more frequently in patients on anticoagulant treatment, and is associated with worsened long-term outcome [23]. Therefore, immediate reversal of any anticoagulant is recommended. Although we did not include a clinical or radiographic outcome parameter in the current study, we were able to show that PCC can be used to improve coagulation in these patients.
Interestingly, factor VIII and VWF levels were shown to be strongly increased (>200%) at all investigated time points. We already observed this phenomenon in previous investigations with patients who were not bleeding under rivaroxaban [16]. A potential explanation is that blood coagulation factor VIII and VWF levels significantly increase during an acute phase response [24].
Limitations of the present study are the lack of randomization / a control group and the small number of patients, resulting in insufficient data on mortality or effective haemostasis. However, this is the first prospective clinical study investigating the effect of PCC on rivaroxaban-related, life-threatening bleeding in vivo. Previous investigations have focused on healthy volunteers receiving an oral Xa inhibitor followed by PCC. To obtain these data at precise time points in emergency patients is challenging and not possible in a large number of patients except in an industry-sponsored study with research staff available around the clock. Especially the measured laboratory parameters add valuable knowledge to this topic.
Nevertheless, great care has to be taken since patients concomitantly took other medication, which probably influenced the results. Moreover, the influence of rivaroxaban on PT, aPTT, coagulation factor levels and thromboelastometry assays differs widely depending on the reagents used and must be evaluated for each particular reagent. Consequently, effects of rivaroxaban on the various assays can also be expected to differ from reagent to reagent.
Consequently, additional studies focusing on these outcomes must be conducted. Even though a specific antidote for the reversal of rivaroxaban is in development, there is still little knowledge about the real efficacy of Andexanet Alpha: preliminary analysis of the ANNEXA-4 study showed 79% efficacy. However, four hours after administration of the antidote rivaroxaban levels returned almost to pre-treatment values [21]. Moreover, it is still unclear whether smaller hospitals will be able to store these, presumably very expensive, antidotes. Additionally, specific antidotes that directly target and eliminate the anticoagulant activity of the DOACs do not possess pro-coagulant properties, which might be necessary in patients suffering from severe haemorrhage. This indicates that coagulation factor concentrates should still be part of the management strategy in bleeding patients.

Conclusions
No controlled clinical studies in humans involving potential reversing agents in bleeding situations are available yet, which illustrates a difficulty in evaluating the potential reversing agents and proposing them for the management of life-threatening bleeding in rivaroxabananticoagulated patients. Our data indicate that fourfactor PCC might be a valuable alternative to specific antidotes for the reversal of rivaroxaban-induced and life-threatening bleeding events. Further studies with clinical endpoints will have to confirm these findings.
Availability of data and materials All data generated or analysed during this study are included in this published article (and its supplementary information files). Moreover, the datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.