Conference Coverage

Aspirin use linked to increased ICH in trauma patients



WAIKOLOA, HAWAII – Among a group of anticoagulated trauma patients, those on aspirin had the highest rate and risk of intracranial hemorrhage (ICH), while those on novel oral anticoagulants were not at higher risk for ICH, ICH progression, or death, a multicenter study found.

“The number of patients on warfarin and antiplatelet agents has significantly increased over time,” Leslie Kobayashi, MD, said at the annual meeting of the American Association for the Surgery of Trauma. “These oral antithrombotic agents have been associated with poor outcomes following traumatic injury, including increased rates of intracranial hemorrhage, increased progression of intracranial hemorrhage, and increased mortality.”

Dr. Leslie M. Kobayashi

Dr. Kobayashi, a surgeon with the University of California, San Diego’s Division of Trauma, Surgical Critical Care, and Burns, noted that in the past 5 years, several novel oral anticoagulants (NOAs) have been approved by the Food and Drug Administration for the treatment of venous thromboembolism as well as for stroke prevention among patients with atrial fibrillation. In general, the safety profile and efficacy of NOAs favorably compares with warfarin in most large randomized controlled trials. “Because of this, NOAs are now recommended as the first-line therapy for non-malignancy–related venous thromboembolism by the American College of Chest Physicians in their 2016 guideline updates,” she said. “However, we don’t have any idea what the effects of the NOAs are following traumatic injury. Several case reports exist in the literature describing catastrophic hemorrhage in patients on these NOAs following traumatic injury. Because of these case reports, several lawsuits have been brought against the makers of these medications, including one that settled against the makers of Pradaxa for $650 million in 2014.”

In a prospective, multicenter observational study conducted by the AAST’s Multi-institutional Trials Committee, Dr. Kobayashi and her associates set out identify injury patterns and outcomes in trauma patients taking the NOAs, and to test their hypothesis that patients taking NOAs would have higher rates of ICH, ICH progression, and death, compared with patients taking traditional oral anticoagulant therapies (OATs). Patients were included if they were admitted to the trauma service on warfarin, aspirin, clopidogrel, dabigatran, apixaban, or rivaroxaban. Pregnant patients, prisoners, and minors were excluded from the study. Data collected included demographics, mechanism of injury, vitals on admission, injuries/injury severity scores, labs, interventions, and reversal agents used such as vitamin K, prothrombin complexes, dialysis, and transfusion of fresh frozen plasma (FFP). Outcomes studied included ICH, ICH progression, and death.

In all, 16 Level 1 trauma centers enrolled 1,847 patients over a 2-year period. Their average age was 75 years, 46% were female, 77% were white, their median Injury Severity Score (ISS) was 9, and 99% sustained a blunt mechanism of trauma. The top two causes of injury were falls (71%) and motor vehicle crashes (15%). One-third of patients (33%) were on warfarin, while the remainder were on aspirin (26%), clopidogrel (24%), NOAs (10%), and 7% took multiple or other agents.

The mechanism of injury pattern was similar between patients taking NOAs and those taking OATs, with the exception of patients on aspirin being significantly less likely to have sustained a fall. Patients on aspirin also had a significantly higher median ISS. “Patients on NOAs presented more frequently in shock as defined by a systolic blood pressure of less than 90 mmHg, but this was not associated with increased need for packed red blood cell transfusion, bleeding requiring an intervention, need for surgical procedure, hospital LOS, complications, or death,” Dr. Kobayashi said.

About 30% of all patients studied underwent an attempt at reversal. The types of agents used to reverse the patients differed depending on drug agent, with antiplatelet patients more frequently getting platelets, and patients on warfarin more frequently receiving FFP, vitamin K, and prothrombin complex. “Interestingly, patients on the anti-Xa inhibitors more frequently received prothrombin complex as well,” she said. “This likely reflects some of the recent literature which suggests that there may be a therapeutic benefit to using prothrombin complex in patients taking the oral anti-Xa inhibitors but not in patients on dabigatran.”

Overall, bleeding, need for surgical procedure, need for neurosurgical procedure, complications, length of stay, and death were similar between those on NOAs and those on OATs. However, the rate of ICH was significantly higher in patients on aspirin. “What is even more surprising is that 89% of the patients in the aspirin-only group were on an 81-mg baby aspirin rather than the larger 325-mg dose,” Dr. Kobayashi said. This difference was significant on univariate analysis and was retained after multivariate logistic regression adjusted for differences between populations, with an OR for aspirin of 1.7 and a P value of .024. “This is not to suggest that patients on aspirin are doing markedly worse, compared to their counterparts, but I think most of us would have assumed that aspirin patients would have done better,” she commented. “I think we’ve definitively shown that is not the case.” Other independent predictors of ICH were advanced age (OR, 1.02), Asian race (OR, 3.1), ISS of 10 or greater (OR, 2.2), and a Glasgow coma score (GCS) of 8 or less (OR, 5.6).

Despite their increased risk for ICH, patients on aspirin were significantly less likely to undergo an attempt at reversal with any type of agent, at 16% with a P value of less than .001, on univariate analysis. “This was significantly lower than all other medications and was retained after multivariate logistic regression, with an OR of 0.3 and a P value of less than .001,” she said.

Progression of ICH did not differ by medication group. Other independent predictors included intraparenchymal location of hemorrhage (OR, 2.2), need for a neurosurgical procedure (OR, 5.1), an attempt at reversal (OR, 2.3) and a GCS of 8 or lower at admission (OR, 4.3). Similarly, multivariate analysis of death showed no significant differences between the different medication groups. Independent predictors included advanced age (OR, 1.06), GCS of 8 or less (OR, 13), progression of head injury (OR, 10), bleeding (OR, 2.3), and complications (OR, 2.1).

Dr. Kobayashi acknowledged that the study’s observational design is a limitation, as well as the fact that it lacked a control group of age-matched patients who were not taking anticoagulants. “Additionally, we had a relatively low number of patients on NOAs, at only 10% of the study population,” she said. “Lastly, there is potential for enrollment bias as all sites involved in this study were level one trauma centers.” She reported having no financial disclosures.

Next Article: