Bleeding, Reversal & Antiplatelet Decisions

The Big Picture

Major bleeding means hemorrhage that is fatal, involves a critical site (intracranial, intraspinal, pericardial, retroperitoneal), drops hemoglobin by 2 g/dL or more, or requires transfusion of 2 units or more. This threshold, not the visible volume of blood, is what triggers reversal protocols.

The core question every time:

1. Is the patient hemodynamically unstable?

2. What is the cause: drug, dilution, consumption, or structural?

3. Is there a specific antidote, or only supportive and nonspecific reversal?

Classify the Cause

Category	Examples
Anticoagulant	Warfarin, DOACs, heparin/LMWH
Antiplatelet	Aspirin, P2Y12 inhibitors
Consumptive	DIC, massive transfusion, sepsis
Factor deficiency	Liver disease, hemophilia, vitamin K deficiency
Structural	Trauma, ulcer, aneurysm, post-procedural bleed

Key habit: ask which drug, which dose, and when the last dose was taken. As of December 2025, the answer also determines which drug class still has a specific antidote on the US market.

Where Drugs Hit the Cascade

Warfarin blocks synthesis of factors II, VII, IX, X (vitamin K dependent)

Dabigatran is a direct thrombin (IIa) inhibitor

Rivaroxaban, apixaban, edoxaban are direct factor Xa inhibitors

Heparin/LMWH potentiate antithrombin, inhibiting IIa and Xa

Aspirin causes irreversible COX-1 inhibition in platelets

P2Y12 inhibitors block ADP-mediated platelet activation

Reversing Warfarin

Major bleeding or urgent surgery:

4F-PCC, weight- and INR-based, 25 to 50 U/kg IV (fixed-dose protocols commonly use about 1,500 IU)

Plus vitamin K, 5 to 10 mg IV, slow infusion over about 30 minutes. Mandatory, not optional.

Target INR under 1.5

Why both agents? Factor VII in 4F-PCC has only a roughly 6-hour half-life. Vitamin K sustains reversal as the PCC effect wears off. 4F-PCC is preferred over 3F-PCC and over FFP: it corrects INR in minutes rather than hours, at lower volume.

Recheck INR every 6 to 8 hours for 24 to 48 hours. Some patients need a repeat vitamin K dose.

Reversing DOACs, Updated for 2026

Drug	Current US Reversal
Dabigatran	Idarucizumab (Praxbind) 5 g IV, two 2.5 g aliquots. Unaffected by the Andexxa withdrawal.
Apixaban / Rivaroxaban	4F-PCC or aPCC, 25 to 50 U/kg IV, now the default. No FDA-approved specific antidote remains on the US market.
Edoxaban	4F-PCC or aPCC, 25 to 50 U/kg IV, off-label, same as above.

Andexanet alfa remains available outside the US: conditionally approved in the EU, marketed in the UK and Japan under the brand Ondexxya.

Protamine, Platelets & DDAVP

Unfractionated heparin: protamine sulfate, 1 mg per 100 units of heparin given in the last 2 to 3 hours (max about 50 mg, slow IV to avoid hypotension or an anaphylactoid reaction).

LMWH (e.g. enoxaparin): protamine is only about 60 percent effective. Use 1 mg per 1 mg of enoxaparin if within 8 hours.

Antiplatelet bleeding: platelet transfusion, 1 unit, if life-threatening and a recent dose was taken. Desmopressin (DDAVP), 0.3 mcg/kg IV, improves platelet adhesion, useful in uremia or aspirin-related bleeding.

Exception, spontaneous ICH: the general rule above does not hold for intracerebral hemorrhage. PATCH randomized 190 patients with antiplatelet-associated spontaneous ICH to platelet transfusion versus standard care and found transfusion inferior, with higher in-hospital mortality (24% vs 17%). Do not transfuse platelets for this indication outside a surgical bleeding target. Full trial detail in Part II, section 15.

TEG / ROTEM: What Each Value Means

TEG	ROTEM	Reflects	If Abnormal, Give
R	CT	Clotting factors	FFP
K / Angle	CFT / Angle	Fibrinogen	Cryoprecipitate
MA	MCF	Platelet function/strength	Platelets ± DDAVP
LY30	LI30/ML	Fibrinolysis	TXA / aminocaproic acid

Typical adult TEG reference: R 5–10 min, K 1–3 min, angle 53–72°, MA 50–70 mm, LY30 <3%. Varies by lab; use your institution's range.

The Bedside Algorithm: One Look, One Plan

1. Unstable? Activate massive transfusion, call for help.

2. Identify agent and last dose time. Choose a specific antidote if one is still available.

3. Send TEG/ROTEM plus standard coags: PT/INR, aPTT, fibrinogen, platelets.

4. Treat the specific deficit the trace shows. Not a blind shotgun of products.

5. Reassess the trace and labs after each intervention.

New: The Andexxa Withdrawal, What Changed and Why

Andexanet alfa (Andexxa) received FDA accelerated approval in 2018 as the first reversal agent aimed specifically at apixaban and rivaroxaban, based on surrogate anti-factor Xa activity in healthy volunteers rather than confirmed clinical outcomes. The confirmatory trial required by that accelerated pathway, ANNEXA-I, was published in the New England Journal of Medicine in 2024 and showed andexanet improved control of hematoma expansion compared to usual care.

The same trial carried the safety signal. In ANNEXA-I, andexanet alfa was associated with a higher rate of thrombosis (14.6% vs 6.9% with usual care, mostly 4F-PCC) and a higher rate of thrombosis-related death (2.5% vs 0.9%) at 30 days. FDA weighed those results alongside postmarketing reports of thromboembolic events, and on December 18, 2025 issued a safety communication stating the risks, including serious and fatal thromboembolic events, outweighed the benefits.

Timeline

2018: FDA accelerated approval for apixaban/rivaroxaban reversal

2024: ANNEXA-I confirmatory trial published (NEJM): better hematoma control, but thrombosis 14.6% vs 6.9%

Dec 18, 2025: FDA safety communication, risks found to outweigh benefits

Dec 22, 2025: AstraZeneca ends US manufacture and sale; BLA voluntarily withdrawn

AstraZeneca said it remained confident in the drug's benefit-risk profile but could not reach agreement with the FDA on a path from accelerated to traditional approval, and withdrew for commercial reasons under that pressure. The practical result is the same either way: as of December 22, 2025, Andexxa is gone from US hospitals.

Many institutions never formally adopted andexanet alfa, citing cost (tens of thousands of dollars per dose) and the lack of a high-quality trial against 4F-PCC. For those centers, the withdrawal changes nothing operationally. For centers that had it on formulary, this means updating order sets now.

Complications & Pitfalls: Don't Get Burned

Thrombosis risk: 4F-PCC carries a real prothrombotic signal. Restart anticoagulation as soon as safely possible. This same signal, seen postmarketing, is what pulled andexanet alfa from the US market entirely.

Vitamin K overshoot: exceeding 10 mg can make re-warfarinization difficult for days.

Protamine reactions: hypotension, anaphylactoid response. Infuse slowly, keep resuscitation ready.

Retired pitfall: the old andexanet alfa warning about interfering with heparin-based anticoagulation monitoring no longer applies in the US. Do not leave it in order sets.

Reflex platelet transfusion: reaching for platelets in antiplatelet- associated ICH is the intuitive move and the wrong one. See section 15.

Quick Reference Card & Primary Sources

Agent	Reversal, US, July 2026
Warfarin	4F-PCC 25–50 U/kg IV + vitamin K 5–10 mg IV
Dabigatran	Idarucizumab 5 g IV
Apixaban / Rivaroxaban / Edoxaban	4F-PCC or aPCC 25–50 U/kg IV (only option, US)
Unfractionated heparin	Protamine 1 mg per 100 units
LMWH	Protamine 1 mg per 1 mg (~60% effective)
Antiplatelet agents	Platelets ± DDAVP 0.3 mcg/kg IV *Not for spontaneous ICH — see §15

Primary sources checked for this revision:

FDA. Update on the Safety of Andexxa. FDA Safety Communication, Dec 18, 2025.
Connolly SJ, Sharma M, Cohen AT, et al. Andexanet for Factor Xa Inhibitor-Associated Acute Intracerebral Hemorrhage. N Engl J Med. 2024;390(19):1745–1755.
Owens B. Anticoagulant Reversal Drug Andexxa Voluntarily Pulled From US Market. Medscape, Dec 22, 2025.
Neale T. Andexanet Alfa Pulled From US Market. TCTMD, Dec 22, 2025.
Andexanet alfa pulled from US market due to thromboembolic risk. Healio, Dec 23, 2025.
Abuan I, Wong KH, Bolinske B, Hale KS. Andexanet Alfa. StatPearls (archived), NCBI Bookshelf.
UMEM Educational Pearls. University of Maryland School of Medicine, Department of Emergency Medicine, Feb 2026.

STEMI + Primary PCI: Load the Potent Agent

Primary PCI intended: ticagrelor 180 mg or prasugrel 60 mg, loaded before or at first PCI. Fibrinolysis planned instead: clopidogrel remains preferred — neither potent agent has been tested safely alongside lytics.

Trial	Finding
PLATO	Ticagrelor vs clopidogrel, 18,624 ACS patients: composite vascular death/MI/stroke 9.8% vs 11.7% at 12 months; overall major bleeding similar, but more non-CABG major bleeding with ticagrelor.
TRITON-TIMI 38	Prasugrel vs clopidogrel, 3,534 STEMI patients undergoing PCI: HR 0.79 for the ischemic composite, no significant rise in non-CABG major bleeding.

Benefit held regardless of primary vs delayed PCI timing in the TRITON STEMI subgroup — the drug choice matters independent of door-to-balloon time.

NSTE-ACS: Load Prasugrel After the Angiogram, Not Before

The one head-to-head trial of the two potent P2Y12 agents found a real difference, and it hinged on timing as much as drug.

Ticagrelor: 180 mg loaded upstream, as soon as possible — before angiography

Prasugrel: 60 mg loaded downstream, only after anatomy is known and PCI confirmed

Trial	Finding
ISAR-REACT 5	4,018 ACS patients: death/MI/stroke at 1 year 9.3% ticagrelor vs 6.9% prasugrel (p=0.006), no bleeding penalty for the downstream strategy.
NSTE-ACS subgroup	Composite endpoint 8.7% ticagrelor vs 6.3% prasugrel (HR 1.41) — the advantage held in the UA/NSTEMI-only analysis.

For STEMI, load upstream — the anatomy is already decided. For NSTE-ACS, hold prasugrel until the angiogram confirms PCI rather than pretreating everyone at triage.

Minor Stroke / High-Risk TIA: DAPT, Sequenced

1. Rule out reperfusion eligibility — thrombolysis or thrombectomy first.

2. Assess bleeding risk (ABCD2, HAS-BLED) plus clinical judgment.

3. Apply evidence-guided DAPT, duration tailored to ischemic vs bleeding risk.

Trial	Regimen	Finding
CHANCE	Clopidogrel+ASA, 21d	Founding trial for DAPT in minor stroke/high-risk TIA.
POINT	Clopidogrel+ASA, 90d	Recurrence cut 26%, but major hemorrhage rose 0.9% vs 0.4% — the tradeoff CHANCE didn't show.
THALES	Ticagrelor+ASA, 30d	Reduced stroke/death vs aspirin alone; editorial noted a smaller relative ischemic-stroke reduction and larger relative bleeding rise than clopidogrel-based regimens.
CHANCE-2	Ticagrelor vs clopidogrel in CYP2C19 LOF carriers	Post hoc: ticagrelor beat clopidogrel without symptomatic carotid stenosis (5.5% vs 7.8% new stroke at 90 days; HR 0.70); no difference with symptomatic carotid stenosis (12.2% vs 12.2%, 197 patients). The treatment-by-stenosis interaction was not significant (p = 0.405).

Default to clopidogrel-aspirin — deepest evidence base, best relative bleeding profile. Reserve ticagrelor for known/suspected CYP2C19 LOF carriers (section 17).

Antiplatelet-Associated ICH: The Full PATCH Data

Section 6 already flags this: don't transfuse platelets for antiplatelet- associated spontaneous ICH. Here is the trial behind that rule.

PATCH: 190 patients with spontaneous supratentorial ICH on ≥7 days of antiplatelet therapy, randomized to platelet transfusion within 6 hours vs standard care.

Result: transfusion was inferior on the primary functional outcome, with higher in-hospital mortality (24% vs 17%).

TICH-2's antiplatelet subgroup found no specific hemostatic therapy proven beneficial for this population — platelet transfusion remains the one strategy shown to cause harm in a randomized trial. Standard care (BP control, reversal of any concurrent anticoagulant, rapid neurosurgical evaluation) outperforms the reflex to transfuse.

Urgent Surgery on DAPT: Interrupt, Don't Reflexively Bridge

Stop DAPT: lowers surgical bleeding, raises stent thrombosis risk — highest first 4–6 weeks post-PCI.

Continue or bridge: preserves platelet inhibition, but continuation raises surgical bleeding; routine bridging lacks proven benefit for hard outcomes.

Source	Finding
BRIDGE	Cangrelor vs placebo bridging in cardiac surgery patients on oral P2Y12 inhibitors: maintained platelet inhibition without excess major bleeding — proof of concept, not a mandate.
Cangrelor bridging cohorts	Retrospective series (VA, single-center): feasible alternative to holding oral P2Y12 inhibitors for urgent surgery; evidence remains observational.
2024 AHA/ACC guidance	Delay elective surgery ≥30 days post-BMS, ≥6 months post-DES for chronic disease, ≥12 months if DES was ACS-indicated. Routine IV bridging not recommended.

Default: continue aspirin if possible, interrupt the P2Y12 inhibitor, restart as soon as bleeding risk allows. Reserve cangrelor bridging for genuinely high thrombotic risk (recent ACS-indicated DES) with surgery, cardiology, and anesthesiology in explicit agreement.

Clopidogrel Resistance: Test the Genotype

About 30% of people carry a CYP2C19 loss-of-function allele that keeps clopidogrel from becoming its active metabolite. A stent thrombosis or recurrent MI on appropriately dosed clopidogrel is a genotype question before it's an adherence question.

Trial	Finding
TAILOR-PCI	5,300 PCI patients, point-of-care CYP2C19 testing: a 34% relative reduction in adverse events at 1 year that missed statistical significance (HR 0.66, P = .06). The 79% reduction in the first 3 months came from a post hoc analysis.
POPular Genetics	2,488 STEMI patients: genotype-guided care noninferior on net adverse events (5.1% vs 5.9%) with significantly less bleeding.
Pooled meta-analysis	7 RCTs, ~16,000 patients: 30% reduction in ischemic events when LOF carriers were switched off clopidogrel; no benefit switching non-carriers.

Point-of-care genotyping is fast enough to act on during the index admission. Keep normal metabolizers (~70%) on clopidogrel; reserve ticagrelor/prasugrel for confirmed or presumed LOF carriers.

Post-Arrest TTM: Oral P2Y12 Absorption Fails Quietly

Hypothermia slows gastric emptying and splanchnic flow, delaying absorption of any oral drug — worsened by opioids, vasopressors, and gut edema in the post-arrest state. Stent thrombosis after PCI in resuscitated arrest runs roughly 5-fold higher than standard STEMI, largely from this absorption gap.

Study	Finding
Clopidogrel PK/PD under TTM	Decreased bioavailability, significantly increased residual platelet reactivity at 33°C despite standard loading.
Hypothermia vs normothermia, 164 ACS pts	Adequate antiplatelet effect in 0% of hypothermic patients on day 1, only 31% by day 3.
Enteral crushed ticagrelor via NG tube	Reliably inhibited platelet function regardless of hypothermia, though delayed; reflux was a distinct failure mode.
Cangrelor bridge pilot RCT under TTM	Bolus + 4h infusion achieved immediate inhibition, bridging the gap until enteral crushed ticagrelor took effect; no rebound reactivity after discontinuation.

Don't trust an intact oral tablet to be working on schedule in a hypothermic, sedated post-arrest patient. Crush and give enterally, watch for reflux, and consider a cangrelor bridge through the highest-risk early window if thrombotic risk is high.

Quick Reference: Antiplatelet Decisions & Sources

Scenario	Default
STEMI, primary PCI	Ticagrelor or prasugrel
STEMI, fibrinolysis planned	Clopidogrel
NSTE-ACS, invasive strategy	Ticagrelor upstream, or prasugrel after angiogram confirms PCI
Minor stroke/high-risk TIA	Clopidogrel + ASA, duration by risk
Antiplatelet-associated ICH	Standard care — no platelet transfusion
Urgent surgery on DAPT	Continue ASA, interrupt P2Y12, restart ASAP
Recurrent ischemia on clopidogrel	Check CYP2C19 genotype
Post-arrest TTM, oral P2Y12	Crush enterally; consider cangrelor bridge

Primary sources for Part II:

Wallentin L, et al. Ticagrelor versus clopidogrel in ACS (PLATO). N Engl J Med. 2009;361(11):1045-1057.
Montalescot G, et al. Prasugrel vs clopidogrel in STEMI PCI (TRITON-TIMI 38). Lancet. 2009;373(9665):723-731.
Schüpke S, et al. Ticagrelor or prasugrel in ACS (ISAR-REACT 5). N Engl J Med. 2019;381(16):1524-1534.
Valina C, et al. Ticagrelor or prasugrel in NSTE-ACS. J Am Coll Cardiol. 2020;76(21):2436-2446.
Wang Y, et al. Clopidogrel with aspirin in minor stroke/TIA (CHANCE). N Engl J Med. 2013;369(1):11-19.
Johnston SC, et al. Clopidogrel and aspirin in acute stroke/high-risk TIA (POINT). N Engl J Med. 2018;379(3):215-225.
Johnston SC, et al. Ticagrelor and aspirin in acute stroke/TIA (THALES). N Engl J Med. 2020;383(3):207-217.
Xie X, Jing J, Wang A, et al. Dual antiplatelet therapy with ticagrelor vs clopidogrel in patients with TIA or minor stroke with or without symptomatic carotid artery stenosis: a post hoc analysis of the CHANCE-2 trial. Stroke Vasc Neurol. 2025;10(5):551-559. doi:10.1136/svn-2024-003293 (NCT04078737).
Baharoglu MI, et al. Platelet transfusion vs standard care in antiplatelet-associated ICH (PATCH). Lancet. 2016;387(10038):2605-2613.
Outcomes in antiplatelet-associated ICH: TICH-2 exploratory analysis. J Am Heart Assoc. 2021.
Angiolillo DJ, et al. Bridging antiplatelet therapy with cangrelor (BRIDGE). JAMA. 2012;307(3):265-274.
2024 AHA/ACC perioperative guideline, coronary stents and noncardiac surgery. Summarized via OpenAnesthesia, 2026.
Pereira NL, et al. Genotype-guided P2Y12 selection after PCI (TAILOR-PCI). JAMA. 2020;324(8):761-771.
Claassens DMF, et al. Genotype-guided P2Y12 in primary PCI (POPular Genetics). N Engl J Med. 2019;381(17):1621-1631.
Clopidogrel PK/PD in OHCA patients under target temperature management. Resuscitation. 2016.
Antiplatelet effect of clopidogrel reduced by therapeutic hypothermia after arrest. Resuscitation. 2010.
Efficacy of enteral ticagrelor in hypothermic post-arrest patients. Platelets. 2016.
Cangrelor bridge pilot RCT, comatose OHCA survivors under TTM. NCT04005729.