Anticoagulation therapy is the cornerstone of VTE treatment. However, active bleeding is considered a contraindication to anticoagulation therapy. Decision-making on anticoagulation therapy after spontaneous ICH evokes debate among clinicians and the inadequacies in the evidence, for the reason that such patients have been excluded from randomized trials of anticoagulation therapy for VTE, makes the decision-making challenging. The three questions that need to be answered are as follows: first, whether anticoagulation therapy should be applied on such patients; second, when is the appropriate timing to initiate anticoagulation; and third, which anticoagulant is the optimal choice.
Anticoagulation or not
Limited data are available on the pros and cons of anticoagulation therapy for VTE following spontaneous ICH. Whether anticoagulation should be used depends on the individual risk-benefit ratio of anticoagulation therapy, that is, the risk of recurrent ICH and hematoma expansion versus the risk of VTE progression.
No reliable methods have been established to predict the risk of recurrent ICH and hematoma expansion. The pathophysiological mechanism of hematoma expansion remains unclear. A number of studies have demonstrated that previous use of anticoagulants or antiplatelet agents, advanced age, systolic hypertension, hyperglycemia, high NIHSS score or Glasgow Coma Scale score are related clinical risk factors for hematoma expansion [
52‐
61]; CTA spot sign, blend sign, black hole sign, island sign, and iodine sign are the radiological risk factors [
62‐
68]. The annualized rate of recurrent spontaneous ICH is approximately 2% [
1,
69]. Spontaneous ICH principally results from small vessel diseases that are mainly composed of hypertensive arteriopathy and cerebral amyloid angiopathy (CAA). Hypertension has been identified as one of the most crucial and modifiable risk factors for recurrent ICH. Lowering blood pressure is associated with a reduced recurrence of ICH [
70‐
72]. Amyloid angiopathy has a predilection for the cortical arteries, hence CAA is a major contributor of spontaneous lobar ICH, which has a significantly higher risk for recurrence compared to deep ICH [
73‐
75]. In addition, a number of studies have demonstrated that older age, cerebral microbleeds (CMB) and cortical superficial siderosis (cSS) on MRI, and Apolipoprotein E (APOE) genotype (ε2 or ε4) are associated with a higher recurrence risk [
76‐
78]. In conclusion, anticoagulation therapy should be more careful in patients with the risk factors mentioned above. Besides, the modifiable risk factors, such as hypertension and hyperglycemia, should be minimized if anticoagulants are to be used in this patient population.
Patients with spontaneous ICH have been excluded from prospective studies and randomized trials on anticoagulation therapy for VTE. The available literature is almost exclusively case reports. Ajmeri et al. [
79] presented a case of a 68-year-old male who suffered from recent ICH and recurrent PE. The anticoagulation therapy was initiated with UFH after a CT scan was done to eliminate active bleeding, and then altered to enoxaparin. The patient was discharged from the hospital in a stable condition without neurological deficits. Similar cases were also reported by Becattini et al. [
80] and Lee et al. [
81]
A Danish large-database retrospective study (
n = 2978) which was focusing on patients with spontaneous ICH, demonstrated that oral anticoagulant (OAC) resumption was associated with decreased risk of thrombotic events and not increasing the risk of recurrent ICH [
82]. Current studies on anticoagulation therapy are mostly focused on the OAC resumption after anticoagulation-related ICH. In addition to VTE, the indications for anticoagulation resumption include atrial fibrillation, mechanical valves, myocardial infarction, etc. The majority of studies have tended to reach a consensus that anticoagulation resumption results in a clinical benefit in terms of thromboembolic event reduction without increasing the risk of recurrent ICH or hematoma expansion (Table
4). These results were confirmed by a meta-analysis (8 studies, 5306 patients with ICH) that demonstrated restarting OAC decreased the risk of thrombotic events (6.7% vs. 17.6%, RR 0.34, 95%CI 0.25–0.45) without significantly increased risk of recurrent ICH (8.7% vs. 7.8%, RR 1.01, 95%CI 0.58–1.77) [
87]. What is noteworthy is that only the patients with smaller hemorrhage volumes and mild functional changes have been eligible for anticoagulation resumption in many studies, consequently decreasing the risk of recurrent ICH.
Table 4
Studies on anticoagulation resumption after ICH
(2010) | Retrospective | Anticoagulation-related | 234 | AF, MV, VTE | Median 5.6 weeks | Warfarin | Without AC | Not mentioned | 8 vs. 10 (HR 5.6, 95%CI 1.8–17.2) |
(2012) | Retrospective | Anticoagulation-related | 284 | AF, MV, VTE | Within a month | Warfarin | Without AC | Not mentioned | 15.4% vs. 15.0% (P = 0.94) |
(2015) | Retrospective | Anticoagulation-related | 719 | AF, MV, VTE | Median 31 days | OAC | Without AC | 5.2% vs. 15.0% (P < 0.001) | 8.1% vs. 6.6% (P = 0.48) |
(2015) | Retrospective | Anticoagulation-related | 160 | AF, MI, MV, IS, VTE | Median 14 days | Warfarin | Without AC | 3.7% vs. 12.3% (P = 0.092) | 7.6% vs.3.7% (P = 0.497) |
(2016) | Retrospective | Spontaneous | 2978 | AF, IS, MI, MV, PAD, VTE | Not mentioned | OAC | Without AC | Lower (HR 0.58, 95%CI 0.35–0.97) | Not increased (HR 0.90, 95%CI 0.44–1.82) |
Nevertheless, some opposite results exist. A retrospective study with 79 patients who had brain tumors and following anticoagulation-related ICH showed that anticoagulation resumption with LMWH or DOACs was associated with a significantly lower risk of recurrent VTE (8.1% vs. 35.3%,
P = 0.003) but a higher risk of recurrent ICH (6.1% vs. 4.2%), especially in patients with primary brain tumors and major ICH [
88]. Different study designs and research objects may explain the opposite results. Several retrospective studies and meta-analysis studies have shown that anticoagulation is associated with an increased risk of ICH in patients with primary brain tumors, however, the association seems not to be found in patients with metastatic brain tumors [
89‐
91]. Hence, anticoagulation resumption in patients with ICH and primary brain tumors must be cautious.
Anticoagulant choice
No comparative studies on different anticoagulants in patients with VTE after spontaneous ICH have been published. Existing studies mostly focus on the safety of anticoagulants and ICH is one of the crucial indicators. The relevant studies almost exclusively referred to vitamin-K antagonists used for anticoagulation resumption after ICH. However, the role of novel oral anticoagulants (NOACs), including dabigatran (inhibitor of factor IIa), rivaroxaban, apixaban, and edoxaban (inhibitors of factor Xa), may be more promising than vitamin-K antagonists for the reasons as follows: (1) NOACs are more convenient to use for lack of monitoring requirements and less interaction with food and other drugs; (2) a number of large-scale retrospective studies and randomized trials have proved that NOACs are associated with a reduction in the risk of ICH compared with vitamin K antagonists
(Table
5); (3) NOACs-related ICH is less severe than warfarin-related ICH, with smaller hematoma volume, lower rate of hematoma expansion, favorable functional and vital outcomes, and lower mortality [
98‐
100]; (4) the availability of antidotes (idarucizumab and andexanet) that allow an immediate and complete reversal of the anticoagulant effect of NOACs [
101‐
103]. We expect high-quality and large-scale studies on patients with spontaneous ICH to provide support.
Table 5
Studies about NOACs vs. warfarin in patients with VTE
(2009) | Randomized trial | 2564 | Dabigatran | Warfarin | 2.4% vs. 2.1%, HR 1.10 (0.65–1.84) | 0 vs. 3 |
(2012) | Randomized trial | 4832 | Rivaroxaban | Enoxaparin, followed by warfarin | 2.1% vs. 1.8%, HR 1.12 (0.75–1.68) | 3 vs. 12 |
(2013) | Randomized trial | 5395 | Apixaban | Enoxaparin, followed by warfarin | 2.3% vs. 2.7%, HR 0.84 (0.60–1.18) | 3 vs. 6 |
(2013) | Randomized trial | 4921 | Edoxaban | Warfarin | 3.2% vs. 3.5%, HR 0.89 (0.70–1.13) | 0 vs. 6 |
(2018) | Retrospective | 218,620 | NOACs | Warfarin | Not mentioned | 1/1000 vs. 3.3/1000 (P < 0.0001) |