Life-threatening bleeding due to persistent dabigatran effect in a patient with sepsis despite idarucizumab therapy and haemodialysis

Abstract

A 58-year-old man presented with necrotising fasciitis and septic shock requiring urgent surgical debridement. Idarucizumab was used preoperatively to reverse the effects of dabigatran, which he was taking for chronic atrial fibrillation. He developed multiorgan failure including an oliguric acute kidney injury and was given continuous venovenous haemodiafiltration. Adjunctive intravenous immunoglobulin therapy was used in addition to his antibiotic therapy for necrotising fasciitis. Significant clinical and laboratory coagulopathy continued for over 12 days with evidence of a persistent dabigatran effect. Here, we discuss the potential impact of the immunoglobulin therapy, the patient’s weight on the degree of redistribution of dabigatran seen and the oliguria in the context of an acute kidney injury on the apparent lack of the effectiveness of idarucizumab.

Background

Dabigatran is a direct thrombin inhibitor which is now in common use for the prevention of ischaemic stroke in patients with atrial fibrillation (AF) and for the treatment and prevention of venous thromboembolic disease. It does not require routine monitoring of drug levels.¹ When concerned about dabigatran toxicity or the risk of bleeding prior to surgical procedures, the thrombin clotting time (TCT, also known as the thrombin time) gives the clearest indication of the presence of dabigatran on a routine coagulation screen. Although it is recognised that the direct thrombin time and ecarin clotting time will give a more accurate reflection of dabigatran activity, these are not universally available.² In the past, there has been no specific reversal agent for patients taking dabigatran. Methods used to reverse the activity of dabigatran have included the use of fresh frozen plasma (FFP), prothrombin complex concentrates (eg, prothombinex), activated prothrombin complex concentrates (eg, factor eight inhibitor bypassing fraction (FEIBA))³ and haemodialysis.⁴ Idarucizumab is now available for use as a dabigatran reversal agent. This is a humanised monoclonal antibody fragment, which has structural similarities to thrombin, and binds with a high affinity to dabigatran.⁵ It has a rapid onset of action, and is renally excreted, mostly within the first 4 hours.⁵

Clinical experience of the effective use of idarucizumab is still in its infancy. Here, we present a case where dabigatran reversal was incomplete and examine the potential reasons for this.

Case presentation

A 58-year-old man, weighing 110 kg, was referred from a secondary centre for ongoing treatment of a severe soft-tissue infection of his left lower limb. He was presented 48 hours prior with a 1-day history of left leg erythema, groin swelling, fever and rigours, in the context of a chronic left leg ulcer. He had failed to improve with intravenous flucloxacillin, clindamycin and metronidazole, and was referred to our tertiary centre for specialist infectious diseases and surgical input. A blood culture taken at the time of admission was positive for group A streptococcus.

He had a history of AF and was taking dabigatran, 150 mg two times a day for ischaemic stroke prophylaxis. He had known peripheral vascular disease and a previous superficial femoral artery angioplasty, an inguinal hernia, hypothyroidism and osteoarthritis. There was no prior history of problematic bleeding perioperatively. His other medications were levothyroxine 100 μg daily, amitriptyline 50 mg daily, atorvastatin 40 mg daily, controlled release metoprolol 95 mg daily, omeprazole 20 mg daily, enalapril 10 mg two times a day and topical urea cream.

On arrival at our centre, his initial observations showed a heart rate 118 beats/min, blood pressure 123/66 and oxygen saturations of 80% on room air, with an improvement to 92% with the application of 3 L of oxygen via nasal cannula. His temperature was 36.1°C. Examination of the lower limb showed cellulitis affecting the lower leg and extending to the medial proximal thigh, with deep seated ulceration over the anterior lower leg. There were multiple blisters around the foot and ankle, the largest of which was over the dorsum of the foot, measuring 8×5 cm and had a necrotising edge. Following specialist review, he was diagnosed with necrotising fasciitis and the decision was made to proceed urgently to the operating theatre for the debridement of the necrotic tissue. He had received his last dose of dabigatran 48 hours previously, and had received therapeutic dose enoxaparin in the interim period, with the last dose given 12 hours earlier. His coagulation studies showed ongoing anticoagulant effect, with an elevated TCT of >150 s, supporting an ongoing dabigatran effect. Our centre was involved in the RE-VERSE AD trial⁶ at the time, and the patient was enrolled. He received idarucizumab (2 vials of 2.5 g administered within 15 min) as part of group B of the study (requirement for an urgent procedure). He was also prescribed 1 mg/kg of intravenous immunoglobulin (IVIg) as an adjunctive therapy for septic shock, with administration starting 90 min after his dose of idarucizumab.

Postoperatively, he was admitted to the intensive care unit (ICU) for the treatment of multiorgan failure secondary to sepsis. By this time, he was requiring vasopressor support for shock. He had an oliguric acute kidney injury (AKI), with a creatinine of 289 μmol/L (baseline creatinine 60 μmol/L). His urine output was 6 mL/hour. He was given continuous veno venous haemodiafiltration (CVVHDF) 6 hours after the time of idarucizumab administration.

Despite the surgical team achieving good haemostasis intraoperatively, reporting only mild ooze at the end of the procedure, the wounds continued to bleed throughout the subsequent hours. He required two return trips to theatre within the first 24 hours and received multiple blood product transfusions. The details of the treatments provided are discussed in more detail below.

Although of varying degree of clinical significance, there was evidence of an ongoing dabigatran effect on this man’s coagulation studies and clinical evidence of bleeding for over 2 weeks. This contributed to multiple return trips to the operating theatre for haemostasis reasons. He also experienced malaena and was found to have a perforated sigmoid colon on day 12 of his admission. He underwent an emergency laparotomy and Hartmann’s procedure for this, which was complicated by significant bleeding from the vascular bed, leading to severe hypotension, and ventricular fibrillation (successfully defibrillated). His coagulation studies during this period remained consistent with an ongoing dabigatran effect.

Investigations

At the time of enrolment into the REVERSE-AD study, the patient had an International Normalised Ratio (INR) 1.6, activated prothrombin time (APTT) 118 s, TCT>150 s, fibrinogen 7.8 g/L and platelets 156×10⁹/L.

His APTT and TCT improved to 82 and 29 s, respectively, on testing at the end of idarucizumab administration. However, his TCT was again elevated at >150 s at 20 hours after idarucizumab administration. His APTT by this time had risen to >250 s (figure 1). His platelet count had fallen to 65x10⁹/L, and his fibrinogen declined but remained within normal limits at 2.0 g/L (local laboratory reference range 1.5–4.0 g/L). He received blood products including platelets and cryoprecipitate during this 24-hour period, so it is possible that the blood results do not fully represent the severity of disseminated intravascular coagulopathy (DIC) as a contributory factor.

Figure 1.

Thrombin clotting time (TCT), activated prothrombin time (APTT) and dabigatran assay results over time.

The TCT remained >150 s until day 8 of his ICU admission, and over 100 s until day 17. His platelet count had fully corrected at day 7.

The dabigatran level was checked 45 hours after the administration of idarucizumab (day 3) and was found to be 51 ng/mL (local reference ranges for the mean therapeutic window, taken from the Randomised Evaluation of Long-Term Anticoagulation Therapy trial,⁷ are 91 ng/mL for a trough measurement, and 180 ng/mL for a peak measurement). It was checked again on day 5 (202 ng/mL), day 6 (101 ng/mL) and day 12 (62 ng/mL). Day 12 was the day of his laparotomy, complicated by significant bleeding. His TCT on this day remained elevated at 104 s despite the dabigatran level taken after surgery being subtherapeutic according to the proposed normal ranges.

Differential diagnosis

Haematology specialists were contacted for advice on multiple occasions throughout the treatment of this patient’s coagulopathy. In addition to an ongoing dabigatran effect, other potential contributory factors were considered and were empirically treated where possible, including DIC and dilutional coagulopathy.

Treatment

During his first 24 hours in our ICU, this patient received 30 units of red blood cells, 29 units of FFP, 4 pools of platelets, 14 units of cryoprecipitate, 9 mg Novoseven (recombinant factor VIIa) and 2 g tranexamic acid in an attempt to treat his coagulopathy. He was also given a total of 51 g Intagram P (IVIg) as part of his treatment of necrotising fasciitis. Following publication of the RE-VERSE AD⁶ trial, it was evident that some patients were re-entered for repeat dosing of idarucizumab. Neither the treating team nor the primary investigator for the institution was aware that this was a potential option, and repeat dosing was therefore not used.

On day 3, further blood products were indicated due to ongoing bleeding from his surgical sites. This included a further 12 units of red blood cells, 2 pools of platelets, 2 units of FFP, 5×500 units of Prothombinex (a 4 factor prothrombin complex concentrate) and 5000 units FEIBA (an activated prothrombin complex concentrate).

Throughout the rest of his ICU stay, he received intermittent blood transfusions guided by his haemoglobin. He also received further FFP at the time of the onset of malaena.

He was given CVVHDF within 6 hours of receiving the idarucizumab, although the primary indication for its use was his oliguric AKI. Given the ongoing problems with haemostasis, anticoagulation was not used within the CVVHDF circuit. The blood flow rate was 150 mL/min. Substitution fluids were delivered in predilution (1.2 L) and postdilution (1.2 L) with no net ultrafiltration. He was transitioned onto sustained low efficiency dialysis on day 7, which he received intermittently for a further 11 days.

Outcome and follow-up

In total, this patient required 19 surgical procedures. He was treated in the ICU for 32 days, before spending a further 28 days as a surgical inpatient and 65 days in inpatient rehabilitation. He has since been discharged home. He was not restarted on any anticoagulant therapy. His renal function has returned to his premorbid baseline.

Discussion

This is a case of significant coagulopathy with re-emergence of dabigatran effect despite attempts to neutralise the drug using the novel specific reversal agent idarucizumab, and the concurrent use of haemodiafiltration, which has been accepted as a method for dabigatran removal in cases of toxicity.^{4 8 9} There were confounding factors related to the coagulopathy in this case, particularly in the first 24–48 hours when it is extremely likely that DIC and dilutional coagulopathy were coexistent; however, laboratory values demonstrate an elevated level of dabigatran present in the blood at a much later date.

One unknown factor in this patient’s case is the potential effect of administering IVIg therapy as part of the treatment for necrotising fasciitis. As idarucizumab is a fragmented monoclonal antibody, there is biological plausibility that IVIg could bind to idarucizumab, interfering with its therapeutic interaction with dabigatran. During the REVERSE-AD trial, anti-idarucizumab antibodies were detected, including antibodies present prior to administration. These were not found to have a detectable effect on idarucizumab activity, but titres described were low.⁶ IVIg is a product of pooled human plasma, and could contain any number and variety of immunoglobulins that have cross reactivity with idarucizumab. There does not appear to be any current published literature on this potential drug interaction.

Dabigatran is active within plasma. It undergoes 35% protein binding and has a moderate volume of distribution of 50–70 L.¹ It has been documented that there is a rebound increase of plasma dabigatran levels following both idarucizumab treatment and dialysis.^{4 5} This is thought to be due to redistribution of extravascular dabigatran down its concentration gradient along with uncoupling from plasma proteins.^4–6 Obesity increases the expected volume of distribution of drugs, and we postulate that in this case, the patient’s high body weight will have contributed to a higher level of peripheral dispersal and therefore slower rate of redistribution and subsequent clearance of the drug.

CVVHDF was used in this case approximately 6 hours after idarucizumab administration due to oliguria. Idarucizumab is almost completely renally excreted in healthy subjects within 4 hours.⁵ This patient’s anuria will have prevented excretion. It has been suggested that delayed excretion would only enhance its activity by increasing its ‘area under the curve’,¹⁰ although the evidence is taken from studies of patients without acute oliguric kidney injury.

Drugs measuring over 15 000 Da are not expected to be removed by any modality of continuous renal replacement therapy.⁹ Idarucizumab has a molecular weight of 48 kDa⁵ and neither idarucizumab or the idarucizumab–dabigatran compound would therefore be expected to be removed by CVVHDF.¹¹ In a patient with oliguria or anuria, late dabigatran–idarucizumab uncoupling may therefore become clinically relevant, and may account for a later rebound anticoagulant effect of dabigatran.

Our knowledge of how additional factors, including renal impairment and other drug therapies, affect the efficacy of idarucizumab treatment will continue to evolve as clinical experience grows. Case reports such as this one may provide an important insight into this group of patients, where controlled studies are often difficult to design and implement. This case represents ‘real-world’ clinical usage of idarucizumab in a critically ill patient with septic shock, organ failures and life-threatening bleeding. We would urge caution with the concurrent use of adjunctive IVIg in such patients.

Learning points.

• Idarucizumab is an effective strategy for the reversal of dabigatran, but factors such as weight, other drug therapies and renal function may impact on the ongoing effectiveness of treatment.

• There appears to be biological plausibility in a drug interaction between idarucizumab and intravenous immunoglobulin therapy that warrants further investigation.

• Obesity increases the volume of distribution of dabigatran and will lead to a more prolonged redistribution effect of the drug after idarucizumab treatment and/or extracorporeal enhanced elimination.

• Consideration of repeat dosing of idarucizumab in patients with anuria may help to offset the effect of uncoupling from idarucizumab and dabigatran.