2025 年 19 巻 1 号 論文ID: cr.2025-0090
Objective: Heparin-induced thrombocytopenia (HIT) is a prothrombotic, immune-mediated complication of heparin therapy. In urgent neuroendovascular procedures, alternative anticoagulation strategies are crucial for mitigating thrombotic risk. Argatroban is a potential substitute; however, its use in neurointervention remains limited. This report describes a case of successful endovascular retreatment using argatroban in a patient with active HIT and includes a literature review to clarify optimal administration protocols, dosages, and monitoring strategies.
Case Presentation: A 77-year-old man with dural arteriovenous fistula (dAVF) developed HIT following initial endovascular embolization. Due to recurrent cortical venous reflux and a high risk of rebleeding, urgent retreatment was performed using argatroban. An intermittent bolus strategy was employed, with dosing adjusted every 30 minutes based on activated clotting time (ACT) to maintain ACT ≥200 seconds. Complete shunt obliteration was achieved without any ischemic or hemorrhagic complications. HIT antibodies became negative 3 months later.
Conclusion: Argatroban is a viable and safe alternative to heparin in neuroendovascular procedures for patients with HIT. Intermittent bolus administration guided by ACT offers precise, situation-specific control and may be particularly appropriate for hemorrhagic cerebrovascular conditions such as dAVFs.
Heparin-induced thrombocytopenia (HIT) is a serious immune-mediated complication resulting from the formation of antibodies against heparin–platelet factor 4 complexes. This paradoxical prothrombotic condition occurs in approximately 0.9%–5% of patients receiving heparin, with thrombosis developing in up to 75%.1) Current guidelines recommend postponing elective procedures until HIT antibodies become undetectable, typically within 3 months.2) However, in urgent situations, particularly in cerebrovascular cases with a high risk of rebleeding, immediate intervention with alternative anticoagulation is required.
We present a case of successful endovascular retreatment for a dural arteriovenous fistula (dAVF) in a patient with active HIT, using argatroban as the intraoperative anticoagulant. An intermittent bolus strategy was employed, guided by intraoperative activated clotting time (ACT), closely resembling traditional heparin protocols. Additionally, we review the relevant literature to clarify the optimal usage of argatroban, including administration protocol, dosage, and monitoring strategies, in endovascular procedures.
A 77-year-old man presented with sudden right hemiparesis. CT revealed a left frontoparietal subcortical hemorrhage (Fig. 1A). DSA identified a superior sagittal sinus (SSS) dAVF (Cognard type IIa+b)3) fed by bilateral middle meningeal arteries (MMAs), posterior meningeal arteries (PMAs), occipital arteries, and superficial temporal arteries, with severe SSS stenosis and bilateral parietal cortical venous reflux (CVR) (Fig. 1B and 1C). Initial transarterial embolization (TAE) was performed via the right MMA under systemic heparinization, using Onyx18 (Medtronic, Minneapolis, MN, USA) with balloon-assisted protection (Fig. 1D). Near-complete shunt occlusion was achieved. Although slight residual shunt flow remained from the bilateral PMAs, CVR was completely resolved postoperatively (Fig. 1E and 1F). A total of 8000 units of heparin sodium was administered intermittently, with ACT maintained at approximately 250 seconds throughout the procedure.

Five days after treatment, the patient developed pulmonary embolism and thrombocytopenia. HIT antibodies were positive (1.8 U/mL), and argatroban was initiated, leading to clinical improvement.
One month later, follow-up DSA showed increased residual shunt flow from the bilateral PMAs and recurrent CVR (Fig. 2A and 2B). Given the persistent hemorrhagic risk, urgent retreatment was pursued despite ongoing HIT antibody positivity. Argatroban (200 μg/kg intravenous bolus) was administered immediately after sheath placement, followed by additional 100 μg/kg boluses as needed, based on ACT monitoring every 30 minutes, to maintain an ACT of ≥200 seconds. Additionally, argatroban-saline (10 mg/500 mL) was used for continuous catheter system rinse. TAE was performed via the left PMA using N-butyl cyanoacrylate (Fig. 2C), followed by embolization via the right PMA using Onyx18 (Fig. 2D), achieving complete obliteration of the shunt (Fig. 2E and 2F). Figure 3 illustrates the line chart of ACT values and argatroban dosing during the procedure. Ten minutes after the initial 200 μg/kg bolus, ACT increased from 168 to 253 seconds. By 60 minutes post-bolus, ACT had decreased to 166 seconds. Subsequent intermittent 100 μg/kg administrations consistently maintained ACT within the target range of 200–300 seconds for the remainder of the procedure. A total of 30 mg of argatroban was administered during the treatment. No ischemic or hemorrhagic events occurred. HIT antibodies became undetectable 90 days after initial identification.


We conducted a systematic literature review of reported cases involving neuroendovascular treatment with argatroban. Searches of PubMed, Scopus, and Web of Science were performed using combinations of the terms: “argatroban,” “endovascular,” “neurointervention,” “anticoagulation,” “heparin,” “thrombocytopenia,” and disease- or procedure-specific terms such as “dural arteriovenous fistula” or “transarterial embolization.” English-language case reports with sufficient clinical detail were included, while studies lacking clear documentation of clinical context were excluded. Extracted data included patient background, procedure type, and details of argatroban use, including administration method, dosage, and monitoring.
Table 1 summarizes previous reports of argatroban use in neuroendovascular procedures. All cases involved patients with confirmed HIT. Alaraj et al. used a 4 µg/kg/min loading dose over 10 minutes, followed by a 1 µg/kg/min continuous infusion, monitored via activated thrombin time.4) Kim et al. applied a percutaneous coronary intervention (PCI)-based protocol—350 µg/kg bolus and 25 µg/kg/min infusion—for carotid stenting.5) Dean et al. and Fukushima et al. administered systemic infusion for HIT-related cerebral venous sinus thrombosis, achieving good recovery.6,7)
| Author (year) | Case no. | Disease | Treatment | HIT | Administration protocol | Argatroban dose | Monitoring | Complications |
|---|---|---|---|---|---|---|---|---|
| Alaraj et al. (2014)4) | 1 | ICS | CAS | Positive | Initial bolus and continuous infusion | 4 µg/kg/min × 10 min (initial bolus) | ATT | None |
| 2 | Aneurysm | Coil embolization | ||||||
| 1 µg/kg/min infusion (continuous) | ||||||||
| 3 | AIS | Mechanical thrombectomy | ||||||
| Kim et al. (2015)5) | 4 | ICS | CAS | Positive | PCI-based protocol (bolus and continuous) | 350 µg/kg IV (initial bolus) | APTT | None |
| 25 μg/kg/min (continuous) | ||||||||
| Dean et al. (2018)6) | 5 | CVST | Thrombectomy | Positive | Continuous infusion | Dose not specified (continuous) | Not listed | None |
| Fukushima et al. (2019) 7) | 6 | CVST | Balloon sinoplasty | Positive | Continuous infusion | 0.7 μg/kg/min (continuous) | Not listed | None |
| Present Case | 7 | dAVF | TAE | Positive | Initial and intermittent bolus | 200 µg/kg IV (initial bolus) | ACT | None |
| 100 µg/kg IV (intermittent administration) |
ACT, activated clotting time; AIS, acute ischemic stroke; APTT, activated partial thromboplastin time; ATT, activated thrombin time; CAS, carotid artery stenting; CVST, cerebral venous sinus thrombosis; dAVF, dural arteriovenous fistula; HIT, heparin-induced thrombocytopenia; ICS, internal carotid stenosis; IV, intravenous injection; PCI, percutaneous coronary intervention; TAE, transarterial embolization
In the present case, the patient developed HIT during the initial treatment of complex dAVF. Although elective interventions are generally suspended until HIT antibodies become undetectable, this typically requires 50–85 days.2) However, hemorrhagic-onset dAVFs with CVR present a substantial rebleeding risk, up to 35% within 2 weeks and 7.4% annually.8–10) After careful consideration, immediate retreatment was therefore prioritized in light of this high risk.
Argatroban, a direct thrombin inhibitor with hepatic metabolism and a short half-life (39–51 minutes), has often been used as an alternative to heparin in PCIs.11) However, its application in neuroendovascular treatment remains relatively limited. Our literature review revealed that most previously reported neuroendovascular cases involving argatroban employed an initial bolus followed by continuous infusion. However, the specific dosages varied, likely due to the absence of standardized protocols for neuroendovascular applications. Despite this variability, no ischemic or hemorrhagic complications were reported in any of the cases, underscoring the procedural safety and clinical feasibility of argatroban as an alternative anticoagulant in neurointerventional settings.
In contrast to the continuous infusion approaches reported previously, we adopted an initial bolus (200 μg/kg) followed by an intermittent administration (100 μg/kg) strategy, tailored to intraoperative ACT levels. This method, previously employed during percutaneous transluminal angioplasty for arteriovenous fistula stenosis in dialysis patients with HIT, allowed flexible and responsive anticoagulation.12,13) In our case, ACT was consistently maintained within the target range (200–300 seconds), and no perioperative complications occurred. These findings further highlight the clinical utility of argatroban’s short half-life and its compatibility with ACT-guided monitoring in neuroendovascular procedures. By allowing situation-specific control of anticoagulation, this approach may be particularly advantageous in the management of hemorrhagic cerebrovascular conditions such as dAVFs.
Argatroban is a safe and effective alternative to heparin for patients with HIT undergoing urgent neuroendovascular treatment. Both continuous infusion and intermittent bolus strategies are viable when monitored with ACT. Further accumulation of cases is necessary to establish standardized dosing and monitoring protocols.
Prior to this study, the protocol was approved by the Ethics Review Board of Tokyo Women’s Medical University (Approval No. 3996).
All authors declare no conflict of interest.