Dabigatran Reversal With Idarucizumab in Patients Requiring Urgent Surgery

A Subanalysis of the RE-VERSE AD Study

Jerrold H. Levy, MD; Joanne van Ryn, PhD; Frank W. Sellke, MD; Paul A. Reilly, PhD; Amelie Elsaesser, PhD; Stephan Glund, PhD; Jörg Kreuzer, MD; Jeffrey I. Weitz, MD; Charles V. Pollack Jr., MD, MA


Annals of Surgery. 2021;274(3):e204-e211. 

In This Article


Patient Demographics

From June 2014 through July 2016, a total of 202 patients were enrolled in group B at 177 sites in 37 countries. Of this cohort, 69% required surgery (n = 140), 28% urgent invasive procedures (n = 57), and 2.5% (n = 5) had their procedure cancelled after receiving idarucizumab, mainly because hemodynamic instability precluded surgical intervention. One of these 5 patients required dialysis after an intentional overdose and idarucizumab was given to allow placement of the dialysis catheter. Dabigatran was completely reversed after idarucizumab and the procedure was no longer needed.[11] The 140 surgical procedures performed included 35% abdominal (n = 49), 32% orthopedic (n = 45), 24% vascular (n = 34), 6% neurosurgical (n = 8), and 3% renal-gynecological (n = 4). The 57 interventional procedures included drainage (35%, n = 20), catheter insertion (51%, n = 29), or diagnostic (14%, n = 8) interventions. A list of the most common surgeries and procedures in each group is shown in Table 1.

Patients requiring abdominal surgery included those with ischemic or perforated bowel and peritonitis or sepsis, cholecystitis, or incarcerated/strangulated hernia requiring urgent repair. Indications requiring orthopedic interventions included 31 hip and/or femur and compound wrist/arm or ankle fractures, more than 80% of which were due to falls. Eleven of the vascular procedures required cardiopulmonary bypass, including 6 ruptured aortic aneurysms or dissections, and 5 with heart transplantation or valve replacement. Permanent pacemaker placement due to complete A-V block or severe bradycardia was performed in 8 patients. Drainage procedures included chest tube insertion for pneumothorax or hemothorax; abscess drainage or joint debridement accounted for the remainder. Central line catheter placement was predominantly for pericardiocentesis, or dialysis catheter insertion for acute renal failure. Diagnostic procedures included angiography, colonoscopy, and bronchoscopy for patients with suspected, but unconfirmed bleeding or infection, or decompression of tumors.

The patient characteristics were similar across all groups; median age ranged between 70.5 and 80 years in the surgery group, with vascular surgery patients being the youngest and those requiring orthopedic interventions the oldest (Table 2). Patients requiring drainage procedures, mostly with chest tubes, were the oldest (Table 3). The majority of patients (≥94%) in all groups were receiving dabigatran for stroke prevention on the background of atrial fibrillation. Creatinine clearance was normal or mildly impaired in all groups, except those in the renal-gynecological group, who had renal injury, in the abdominal group or those in the catheter group, who mostly required central catheter placement for dialysis as a result of acute kidney injury, and was compounded by sepsis in many of these patients.

The median patient-reported time from the last dabigatran dose to idarucizumab administration ranged between 11 and 25 hours (Table 2 and Table 3). Median dabigatran levels at enrollment ranged between 41 and 77 ng/mL in the surgery patients. Patients in the abdominal group, including acute abdominal ischemia, peritonitis, or other abdominal inflammatory conditions had the highest levels. In patients undergoing procedures, median dabigatran levels ranged from 67 to 118 ng/mL; the group requiring central catheters had the highest dabigatran levels of the procedural group, correlating with their degree of renal impairment.

Almost all patients had at least 1 cardiovascular comorbidity, including hypertension, congestive heart failure, or coronary artery disease or had a history of stroke or transient ischemic attack. In addition, in all procedural subgroups, more than 10% of patients had active cancer, as did those in the abdominal and urological-gynecological surgical groups.

Reversal of Dabigatran Anticoagulation

At study entry, 185 of the 202 patients (90.1%) had a prolonged aPTT, dTT, or ECT. Median maximum reversal within 4 hours measured using these tests was 100% (95% confidence intervals, 100%–100%). This was achieved in all patients in the surgery and procedural arm, except for 2 patients in the abdominal surgery group where reversal was more than 50% in the first 4 hours using the dTT. Overall complete reversal was achieved in more than 95% of patients when measured as the dTT. When measured using the ECT, at least 80% reversal was achieved in all patients, except for one in the abdominal surgery group; 72% or more achieved 100% reversal. The dTT and thrombin times normalized by 30 minutes in 100% of patients studied, and ECT in 94% of the patients studied (Figure 1).

Figure 1.

Reversal of dabigatran plasma levels after administration of idarucizumab (arrows) in patients requiring surgical (graphs A–E) or procedural (graphs F–H) interventions. Bar represents median and quartiles, whiskers represent the 10 to 90 percentiles. The number of patients (n) with at least 1 measurable sample are shown.

When using the aPTT to measure reversal, a similar pattern was seen (Figure 2). Median maximum reversal of 100% (95% confidence intervals, 100%–100%) was achieved in approximately 90% of patients in all groups, except in the neurosurgery group, where only 3 of 8 patients had an elevated aPTT at enrollment and 100% reversal was achieved in 2 of these patients. In both the vascular surgery group and diagnostic procedures, heparin was also given during the interventions in some patients, that is, for cardiopulmonary bypass of ruptured aneurysm repair, which also prolonged the aPTT (Figure 2) and partially explains the increased variation in this assay. All patients had detectable levels of dabigatran upon enrollment and the extent of reversal based on unbound dabigatran concentrations in plasma is shown for each subgroup in Figure 1. When comparing graphs C or H in Figures 1 and 2, it becomes evident that the aPTT prolongation in some patients was not due to elevated dabigatran levels.

Figure 2.

Reversal of the activated partial thromboplastin time (aPTT) after administration of idarucizumab (arrows) in patients requiring surgical (graphs A–E) or procedural (graphs F–H) interventions. Bar represents median and quartiles, whiskers represent the 10 to 90 percentiles. The number of patients (n) with at least 1 measurable sample are shown. upper limit of normal (ULN) is the upper limit of normal of the assay (39.8 seconds).

In a small subset of patients a re-elevation in dabigatran levels was observed after 24 hours. In addition, a second dose of idarucizumab was given to 4 (2.0%) of the 202 patients. Two patients were enrolled for abdominal procedures and 2 with acute renal failure requiring catheter placement for dialysis. Re-elevation of dabigatran was also accompanied by an increase in anticoagulation, as can be seen in Figure 2 when measured as the aPTT.

Clinical Outcomes

Time to Surgery. The time from administration of the first vial of idarucizumab to the initiation of the surgery or procedure is shown in Table 4. In all groups, the median time to surgery or procedure was between 1.2 and 1.9 hours, except in those requiring neurosurgery, where it was 3.3 hours. This value includes 1 patient, a 90-year old man, who presented with a subdural hematoma and had a decompressive craniotomy performed 5 days after administration of idarucizumab. The time to surgery would be consistent with the other groups, 1.9 (1.2, 15.5, median IQR) hours in the other 7 patients.

Periprocedural Hemostasis. Periprocedural hemostasis was achieved in more than 91% of patients undergoing surgery and in more than 93% of patients undergoing procedures (Table 4). Several patients experienced mildly abnormal hemostasis (oozing) during abdominal, orthopedic, or vascular surgery and in 1 patient in both the catheter and drainage groups. Moderately abnormal hemostasis (controllable bleeding) occurred in 3 patients in total; 2 in the surgical cohort and 1 in the procedural cohort. There was no severely abnormal hemostasis during any surgery or procedure.

Blood loss in the first 24 hours occurred in 6 of 140 patients after surgery and 1 in 57 patients postprocedure; none was fatal. Fresh frozen plasma and packed red blood cells were the most frequently administered blood products; volume expanders were used in 11% to 25% of patients (Table 5). In the surgical patients, fresh frozen plasma was transfused in 9% to 25% of patients, whereas red blood cells were transfused in 13% to 42% of patients, except for those undergoing gynecology-urology procedures who received no blood products. Platelets were administered in 0 to 15% of patients depending upon the type of surgery as shown in Table 5. Surgical patients were hospitalized for a median of 9 days after abdominal and vascular procedures, and for 15 and 19 days after orthopedic and neurosurgical procedures, respectively, including a median time of 2 to 4 days in the ICU (Supplemental file Table 6, http://links.lww.com/SLA/B789). Neurological patients were in the ICU for a median of 8 days. Procedural patients were in the hospital for a median of 6.5 to 9 days, with a median of 5 to 7 of these in the ICU, with the exception of diagnostic procedures, who were in the ICU for a median of 2 days.

Postreversal Thrombotic Events, Anticoagulation Postsurgery, and Mortality. There were no thrombotic events in patients undergoing abdominal, urogenital, or neurosurgery or for those too unstable for surgery (Supplemental file Table 6, http://links.lww.com/SLA/B789). In the first 5 days post-treatment there were 2 wound infections, 2 wound hematomas that included 1 abdominal wall hematoma and 1 incision site hematoma. There were 3 events each in patients with orthopedic and vascular surgery within 30 days. The overall 30 day rate in the surgery group was 4.3%. Of the 57 patients in the procedural group, 4 (7%) had a thrombotic event within 30 days; 2 in the drainage group and 2 in the catheter group.

Anticoagulation was restarted in the majority of patients within days of their surgery or procedure, with parenteral anticoagulants started in 62% and 75% of patients, respectively, within 1 day postprocedure (Supplemental file Table 7, http://links.lww.com/SLA/B789). Oral anticoagulation was started in 52% to 100% of patients, of these, dabigatran was restarted in 62% to 100% post-surgery and procedures. The exception was patients undergoing neurosurgery. Parenteral anticoagulation was started much later postprocedure, and 5 of 8 such patients restarted antithrombotic therapy, one with dabigatran.

Mortality at 30 days in the catheter group was 34.5%; related to septic shock, multiple organ failure, worsening peritonitis, and/or circulatory shock in approximately one-third of these patients (Supplemental file Table 6, http://links.lww.com/SLA/B789). In addition, 2 of the 5 patients too unstable for surgery upon enrollment died within days of peritonitis and circulatory shock. Overall 30-day mortality rates in the surgical and procedural groups were 7.9% and 21%, respectively.