Spontaneous Retrobulbar Hemorrhage

During his inpatient stay, the patient remained hemodynamically stable and did not require transfusion of blood or platelet products. All home anticoagulant medications were discontinued. The patient continued to have some oozing the following morning, and was given an additional dose of IV PCC (50 U/kg), which resolved the bleeding. He remained hemostatic and, based on his history of AF, he was discharged home on warfarin without bridge therapy. Both rivaroxaban and daily aspirin therapy were discontinued. The lateral canthotomy and cantholysis healed without need for surgical intervention. An ophthalmology follow-up clinic visit 1 week after discharge from the hospital revealed an already self-healed incision without ectropion or retraction and with only mild laxity. Given the patient’s history of AF with TIAs while off anticoagulants, the ophthalmologist did not recommend any other surgical intervention that would have required discontinuing the warfarin.

Discussion

With any retrobulbar hematoma, one must be concerned for orbital compartment syndrome. Orbital hemorrhage is the most common cause of orbital compartment syndrome, usually occurring secondary to trauma, surgery, or retrobulbar injection. In this case, spontaneous hemorrhage due to anticoagulation was believed to be the cause—albeit a rare one—of orbital compartment syndrome. Because the orbital space is enclosed and cannot expand, it is vulnerable to compartment syndrome, and subsequent ischemia can lead to permanent vision impairment or complete loss of vision.⁵ Early recognition and treatment is imperative to preserve vision as an elevated intraorbital pressure for 60 to 100 minutes can lead to permanent visual sequelae.

Management

Treatment of retrobulbar hemorrhage includes lateral canthotomy and cantholysis, which have been shown to reduce IOP an average of 14.2 mm Hg.⁶ In our patient, IOP in the affected eye was reduced by 18 mm Hg.  In addition to the patient’s high IOP at presentation, another concern was the continued hemorrhaging from both the incision site and its potential to exacerbate the underlying retrobulbar hematoma. Management of this condition posed a challenge because this patient was taking a newer anticoagulant, for which there is currently no specific reversal agent. After consultation with hematology services, the patient was given PCC because small studies have suggested that PCC may reverse rivaroxaban-induced anticoagulation.⁷ While more expensive than fresh frozen plasma, PCC has a high safety profile and should be considered in cases of life-threatening bleeding—especially in patients who have renal failure, as rivaroxaban is renally excreted. The half-life of rivaroxaban is 5 to 9 hours and its effects may last up to 12 hours. An IV dose of 50 U/kg PCC can be effective in reversing rivaroxaban; this dose can be repeated every 12 hours until hemorrhaging abates or until rivaroxaban is cleared. 

Potential Factor Xa Reversal Agent

Phase IV trials are underway in the ANNEX-A (Andexanet Alfa a Novel Antidote to the Anticoagulant Effects of FXa inhibitors – Apixaban) and ANNEX-R (Andexanet Alfa a Novel Antidote to the Anticoagulant Effects of FXa inhibitors – Rivaroxaban) studies assessing andexanet alpha, an FXa inhibitor reversal agent and potential FXa inhibitor antidote. Andexanet alpha is a decoy protein that binds to FXa inhibitors in the active site, restoring endogenous FXa and reducing anticoagulant activity.⁸ This serves as another promising reversal agent for apixaban, edoxaban, and rivaroxaban. With the development of these new FXa reversal agents, EPs will have more options for reversal of anticoagulation in patients with unique hemorrhagic presentations.

Conclusion

Rivaroxaban has the potential to replace warfarin as a “novel” oral anticoagulant of choice for multiple indications, especially as more insurance companies cover the use of the FXa inhibitors. As a result of their increased use, the EP is likely to see an increasing number of patients who present with hemorrhagic consequences of the FXa inhibitors, and in turn must be familiar with the properties of this class of anticoagulants—including potential reversal strategies.

Our case of spontaneous retrobulbar hemorrhage may be one of these new patterns of bleeding to be expected from a novel FXa inhibitor. Therefore, it is imperative that EPs consider retrobulbar hemorrhage and other possible bleeding locations in patients on an FXa inhibitor.