Poor outcomes after dabigatran-associated intracranial hemorrhage despite idarucizumab reversal

Abstract Intracranial hemorrhage (ICH) is the most deadly bleeding complication associated with anticoagulation. The efficacy of idarucizumab in treating dabigatran-associated ICH in the real world is uncertain. We sought to assess patient outcomes in this sick population. This was a 2-year prospective observational study of functional neurologic status in patients who received idarucizumab following dabigatran-associated ICH across three tertiary Canadian hospitals. The primary outcome was disability on the modified Rankin scale thirty days after antidote administration. Five patients received idarucizumab for dabigatran-associated ICH. The median time to idarucizumab administration was 43 minutes (range: 2–163 minutes). Four patients were dead at 30 days. The fifth patient was in a minimally conscious state with hemiparesis requiring full nursing care. Three patients were transitioned to palliative care based on their advanced directives and dismal prognosis as determined by the treating team. High quality care should not include idarucizumab when it is unlikely to achieve patients’ previously stated goals of care. However, rapid administration of this expensive antidote is often necessary when information is incomplete.


Introduction
Intracranial hemorrhage (ICH) is the most deadly bleeding complication associated with anticoagulation from dabigatran, a direct thrombin inhibitor [1]. Idarucizumab is a humanized monoclonal antibody fragment against dabigatran that produces rapid reversal of coagulation parameters. While Health Canada approved idarucizumab in March 2016 for severe dabigatran-associated bleeding, its impact on clinical outcomes remains imprecise due to the lack of high quality randomized studies and uncertainty regarding the therapeutic window for its administration [2]. The decision to administer idarucizumab in clinical practice is further complicated by the drug's high cost and perceived medicolegal risk to provider if withheld [3]. The effectiveness of idarucizumab in treating ICH likely differs from the results reported by a large industry-funded study such as Reversal of Dabigatran Anticoagulant Effect With Idarucizumab (RE-VERSE AD) [4]. We sought to assess patient outcomes following administration of the drug for ICH in clinical practice.

Study design
This was a prospective, observational study of functional neurologic outcome following idarucizumab administration for ICH across 3 urban, academic Canadian hospitals between June 1, 2016 and June 1, 2018. Combined, these centers have 700 inpatient beds, 95,000 annual emergency patient visits, and referrals for specialized care in trauma, neurosurgery, and interventional neuroradiology. Ethics approval was obtained from the McGill University Health Centre institutional review board.

Study enrolment and data collection
Physicians ordering idarucizumab were required to submit an evaluation form with the following patient data to obtain pharmacy release of the drug: patient demographics, clinical indication for idarucizumab, and available coagulation parameters. All patients with ICH with a request for idarucizumab were included. There were no exclusion criteria. The  following data were collected for each patient on a standardized data abstraction tool by two independent reviewers who resolved discrepancy by consensus discussion: type of bleed, initial Glasgow coma scale (GCS), partial thromboplastin time (PTT) before and after idarucizumab administration, medications, renal impairment, and time to idarucizumab administration. The primary outcome was functional neurologic status at thirty days as assessed using the modified Rankin Scale (mRS), a clinician-reported score of global disability [5]. Patients who were discharged or transferred before 30 days were followed by phone.

Analysis
The study population was small and non-parametric descriptive statistics were used when appropriate.

Results
The pharmacy received 12 evaluation forms for idarucizumab administration during the study period. Seven requests were for non-ICH indications. Five patients with ICH received idarucizumab and were thus included in the study. Patients had both traumatic and spontaneous bleeds (Table 1). There was 100% follow-up and median mRS for patients at 30 days was 6 (range 5-6). All four patients enrolled from the emergency department died despite presenting with initial GCS ! 12. The surviving patient was enrolled from a medicine ward and remained hemiparetic in a minimally conscious state at 30 days.
The median time from emergency department presentation or symptom onset to idarucizumab administration was 43 minutes (range: 2-163 minutes).
The PTT was elevated in all patients upon presentation and normalized when repeated after antidote administration. Patients 1 and 2 did not receive repeat partial thromboplastin tests; one due to palliative care choice by the family and the other due to transfer to another institution. One patient (#1) was on concurrent aspirin therapy. Substitute decision makers for patients 1, 2, and 4 requested palliative care within 2 hours of idarucizumab administration. Treating teams believed the prognosis was dismal for each of these patients based on the speed of clinical deterioration and the extent of ICH on the initial scan.

Discussion
Our results suggest that patients who receive idarucizumab for dabigatran-reversal in ICH fare poorly in spite of normalized coagulation parameters. This contrasts the findings of RE-VERSE AD in which only 16 of 98 patients with ICH died at 30 days (16.4%). Given the relationship between industry funding and the reporting of positive findings, this discrepancy raises questions about the effectiveness of idarucizumab in clinical practice [6,7]. We could not compare our functional neurologic outcomes with those of RE-VERSE AD because of their incomplete reporting. Many patients in our study had advanced directives which resulted in early de-escalation of care. Each of these decisions, however, occurred after discussion with the treating physician and reflects the critical intersection of individual patient values and medical evidence. The treating teams did not believe the patients were capable of achieving a meaningful neurologic recovery based on rapid deterioration and the initial imaging studies despite receiving idarucizumab. This raises questions on the therapeutic window for idarucizumab and whether with certain, extensive ICH its administration is futile.

Limitations
Although PTT is an imperfect measure of thrombin inhibition, it is widely available and correlates well with both plasma dabigatran concentration and dilute thrombin values in the therapeutic range [8]. All participating hospitals were referral centers which may have led to the disproportionate enrolment of sicker patients at risk for worse outcome. Conversely, the better outcomes reported in RE-VERSE AD may have been due to overrepresentation of patients with minor bleed or even artifactual ICH mimics on computed tomography. Our study design precludes assessment of patients who may have benefited from idarucizumab but did not receive it. Lastly, the small number of study subjects limits the generalizability of our findings.

Conclusions
Clinicians striving to provide high value care need specific criteria to determine when idarucizumab administration is likely to improve patient-centered outcomes after ICH. We believe it is judicious to rapidly administer idarucizumab to reverse dabigatran in ICH patients with a potential for neurological recovery. However, our results suggest that patients with dabigatran-associated extensive ICH do poorly despite reversal and that high quality care should not include the expensive administration of idarucizumab when it is unlikely to achieve previously stated goals of care.