• Modified the primary efficacy objective and primary efficacy end point to include changes in anti-fXa activity. The 2 primary efficacy end points were to be tested sequentially, with the proportion achieving hemostatic efficacy tested only if the change in anti-fXa activity was first demonstrated. • Modified the secondary efficacy objective and secondary efficacy end point to assess the relationship between the 2 primary efficacy end points: to evaluate the relationship between change from baseline to the evaluation period in anti-fXa activity and effective hemostasis. • Eliminated the original Study Day 1, 24-hour study data point and updated time points to Day 1, 8-hour and Day 1, 12-hour data points. • Clarified the requirements, inclusion, and exclusion criteria regarding acute major bleeding. • Changed the definition of history prior to Screening from 1 month to 2 weeks. • Modified follow-up to include rescue therapy for participants with a poor or no response. • Provided specific detail on the measurement of closed bleeds to document hemostatic control. • Increased the number of sites allowed in North America and Europe from 60 to 120.

• Modified the duration of safety follow-up from 45 to 30 days to align follow-up period with standard clinical practice for intracranial hemorrhage recovery timelines. • Clarified that (visible) bleeding must be overt to qualify for inclusion in this trial. • Clarified that intracranial hemorrhage bleeds could be diagnosed and assessed with either computed tomography or magnetic resonance imaging. • Included edoxaban as 1 of the FXa inhibitors that could qualify a participant for this trial. Also, clarified that the list of FXa inhibitors being studied in this study was restricted to apixaban, edoxaban, rivaroxaban, and enoxaparin. • Clarified that participants who were scheduled for surgery to occur within the first 12 hours after receiving andexanet should not be enrolled in this study. • Clarified the blood and blood-related products were allowed as well as the time frames allowed. • Increased the size of allowable hematoma volume from 30 to 60 cubic centimetres to enable the participant population to be more representative of what might be expected in the clinical setting. • Shortened and/or consolidated the restrictions around duration of prior exposure to medications to within 7 days of andexanet treatment, based on expected duration of effect. • Provided specificity around rating of hemostatic efficacy for different subtypes of intracranial hemorrhage.

• Increased sample size from 250 participants to 350 participants. • Enriched participant population for intracranial hemorrhage; ensured a minimum of 110 efficacy evaluable intracranial hemorrhage participants, including 50 participants at high risk for hematoma expansion. • Added a requirement for a reasonable expectation that a participant would be treated with andexanet within 2 hours after a baseline scan (intracranial hemorrhage participants only). • Excluded participants with visible, intra-articular, and musculoskeletal bleeding. • Excluded participants for whom the Investigator believed that the hemoglobin would drop below 8 grams/decilitre after volume resuscitation. • Changed threshold of efficacy evaluability for enoxaparin participants from 0.5 to 0.25 international units/millilitre (IU/mL). • Added clinical criteria for re-bleeding and guidance for re-dosing of andexanet in the event of re-bleeding. • Added re-bleeding, tissue factor pathway inhibitor (TFPI), antithrombin III, anti-IIa activity, Glasgow Coma Scale, Modified Rankin Score, and National Institutes of Health Stroke Scale as exploratory efficacy end points. • Converted thrombin generation from a safety end point to an efficacy end point, and mortality from an efficacy end point to a safety end point. • Added additional time points through 72 hours post-andexanet dosing for TFPI levels and thrombin generation. • Updated andexanet dosing recommendations.

• To establish lower anti-fXa activity threshold for participants taking edoxaban to reflect contemporary understanding of risks and benefits of edoxaban. • Efficacy evaluable participants was re-defined as follows: All patients must have a central laboratory-determined anti-fXa activity ≥ 75 ng/mL for patients receiving apixaban and rivaroxaban, ≥ 40 ng/mL for patients receiving edoxaban, and ≥ 0.25 IU/mL for patients receiving enoxaparin. All other criteria stayed the same. • Exclusion criterion #5 was updated as follows: The patient has a recent history (within 2 weeks) of a diagnosed TE as follows: Venous Thromboembolism (VTE; e.g., deep venous thrombosis, PE, cerebral venous thrombosis), MI (including an isolated troponin elevation), DIC, cerebral vascular accident, TIA, unstable angina pectoris hospitalization, or severe peripheral vascular disease within 2 weeks prior to Screening (see Protocol Amendment 5, Appendix E for DIC scoring algorithm). • Clarified definition of re-bleeding to be consistent throughout protocol. • Updated and clarified investigational product return and destruction.

• Corrected an error in the Synopsis and updated for accuracy and to align with the protocol body. • Clarified that adverse events and survival was to be followed through the study. • Updated to simplify the description of the study periods related to safety monitoring. • As a surrogate for elevated anti-fXa activity, the eligibility criteria restrict enrollment to participants who received their last dose of FXa inhibitor within 18 hours, if the timing is known. (If the timing of the last dose of FXa inhibitor is unknown, the andexanet bolus must begin as soon as possible–following signing of the informed consent form [ICF] and completion of pretreatment procedures–but no later than 3 hours following signing of the ICF). • Clarified exclusion criterion 4.2 to include language about the timing of additional surgery (12 hours after end of andexanet infusion). • Updated the andexanet dosing to reflect accurate ranges of < 30 mg (low dose) and ≥ 30 mg (high dose) for andexanet dosing in participants receiving edoxaban. • Added the Sponsor in Japan as Bristol-Myers Squibb K.K. in collaboration with Portola Pharmaceuticals, Inc. • Eliminated or revised all sentences related to enoxaparin participants because the dose of enoxaparin allowed in the study was not approved in Japan. • Modified andexanet dosing regimen for each FXa inhibitor. • Revised sentences related to drug supply and accountability from managing by Portola Pharmaceuticals, Inc. to managing by Bristol-Myers Squibb K.K. • Increased sample size to 500 participants to ensure adequate enrollment of various subgroups within the enrolled population.