The Hard Bargain of Anticoagulation after Intracranial Hemorrhage, in the Setting of Venous Thromboembolism: Between a Rock and a Hard Place

Few medical scenarios are more apt for Homeric metaphors about steering between dragons and whirlpools[1] than starting therapeutic anticoagulation in someone who had serious bleeding or so recently nearly bled to death. The optimal assessment and mitigation of bleeding risk has recently been the focus of a position paper in this journal.[2] Resumption of antithrombotic therapy after bleeding is a key component of this debate.[3] [4]

The metaphor is even more apt in a survivor of intracranial hemorrhage (ICH) who goes on to develop a venous thromboembolism (VTE), truly leaving the patient and clinician between a rock and a hard place. The lack of large, randomized trials supporting the use of anticoagulation after ICH has led to considerable variation in recommendations on whether and when to take this weighty step. Clinical trialists are hard at work on remedying this evidence gap, but, in the meantime, we must do with what we have.

In this issue, Schulman and colleagues[5] have offered us a retrospective series of ICH cases who then developed VTE. This model mitigates some of the confounding by indication in retrospective datasets. Mild cases are more likely to be anticoagulated and sooner and are also more likely to have better outcomes. It is impossible to tease apart the effect of anticoagulation from the selection bias, thus the need for prospective data. However, the VTE tips the clinician's hand, making her perhaps more aggressive in restarting than she might otherwise have been with larger hematomas more likely to expand. Of note, the authors excluded from analysis 27 cases who were anticoagulated at the time of the incident ICH, eliminating some of the higher thrombotic risk in the remaining cohort and making this focus on a “start” rather than “restart” analysis. Also, the small size of the dataset precluded multivariate and time-dependent modeling used in other cohorts.[6] Regardless, this model allows us to see what happens from a rebleeding perspective when these high-risk patients are anticoagulated.

Here we see that anticoagulants behave as expected. They are associated with a lot less VTE (20 vs. 58%) and a little more bleeding (18 vs. 13%), none of these differences being statistically significant in a cohort of 104 cases. Of course, the net clinical benefit is not so easily calculated as 38% VTE reduction minus 5% increase in ICH expansion. The latter is on average quite a bit more consequential, leading to thorny questions such as how many deep vein thromboses equal an ICH? Utility Theory suggests about five.[7]

The composite outcome the authors construct to reflect net clinical benefit is interesting but suffers from the same problem. Summing recurrent/progressive VTE + recurrent/expanding ICH + extracranial major bleed + death fails to reflect the varying severity of the components of the composite. This is not to suggest such adjustment is easy, even in a prospective model, but it should give one serious pause in using the findings to inform practice.

The authors suggest that the data support “stepwise escalation” of anticoagulation as preferable to restarting full-dose anticoagulation. This seems premature based on such a small retrospective sample with so much potential bias. Restarting anticoagulation after ICH remains an individualized decision based on perceived risk and benefit, which is just another way of saying that the evidence is not compelling either way. The evidence on when or how to restart is even less so. The most pressing question is whether to ever restart anticoagulation.[8] [9] Large clinical trials such as ENRICH AF (NCT03950076: edoxaban vs. usual care in ICH survivors with atrial fibrillation) and ASPIRE (NCT03907046: apixaban vs. aspirin), will report out in the coming years and shed some light.

The next step will be a timing trial randomizing restart intervals.[10] Beyond the “whether to start …” and “when to start …,” there may be a role for examining the “how to start ….,” such as escalation versus full dose, in a prospective randomized model. It would have to be a large trial to detect a clinically meaningful difference.

The authors should be congratulated for recapitulating Byrne's model[11] in traumatic ICH with subsequent VTE in a spontaneous ICH population, as these two distinct diseases have different causes, courses, treatments, recurrence, and thrombotic risks.[12] The findings generate some very interesting hypotheses for testing in future prospective models.

Publication History

Received: 17 May 2023

Accepted: 19 May 2023

Accepted Manuscript online:
22 May 2023

Article published online:
16 June 2023

• References

• 1 Homer. The Odyssey. London: New York: W. Heinemann; G.P. Putnam's Sons; 1919
• 2 Gorog DA, Gue YX, Chao TF. et al. Assessment and mitigation of bleeding risk in atrial fibrillation and venous thromboembolism: executive summary of a European and Asia-Pacific Expert Consensus Paper. Thromb Haemost 2022; 122 (10) 1625-1652
• 3 Geisler T, Poli S, Huber K. et al. Resumption of antiplatelet therapy after major bleeding. Thromb Haemost 2023; 123 (02) 135-149
• 4 Chi G, Lee JJ, Sheng S, Marszalek J, Chuang ML. Systematic review and meta-analysis of thromboprophylaxis with heparins following intracerebral hemorrhage. Thromb Haemost 2022; 122 (07) 1159-1168
• 5 Gorman J, Candeloro M, Schulman S. Anticoagulant management and outcomes in nontraumatic intracranial hemorrhage complicated by venous thromboembolism: A retrospective chart review. Thromb Haemost 2023; 123 (10) 966-975
• 6 Milling Jr TJ, King B, Yue P. et al; ANNEXA-4 Investigators. Restart of anticoagulant therapy and risk of thrombosis, rebleeding, and death after factor Xa inhibitor reversal in major bleeding patients. Thromb Haemost 2021; 121 (08) 1097-1106
• 7 Hogg K, Kimpton M, Carrier M, Coyle D, Forgie M, Wells P. Estimating quality of life in acute venous thrombosis. JAMA Intern Med 2013; 173 (12) 1067-1072
• 8 Paciaroni M, Agnelli G. Should oral anticoagulants be restarted after warfarin-associated cerebral haemorrhage in patients with atrial fibrillation?. Thromb Haemost 2014; 111 (01) 14-18
• 9 Ivany E, Lotto RR, Lip GYH, Lane DA. Managing uncertainty: physicians' decision making for stroke prevention for patients with atrial fibrillation and intracerebral hemorrhage. Thromb Haemost 2022; 122 (09) 1603-1611
• 10 Milling Jr TJ, Warach S, Johnston SC. et al. Restart TICrH: an adaptive randomized trial of time intervals to restart direct oral anticoagulants after traumatic intracranial hemorrhage. J Neurotrauma 2021; 38 (13) 1791-1798
• 11 Byrnes MC, Irwin E, Roach R, James M, Horst PK, Reicks P. Therapeutic anticoagulation can be safely accomplished in selected patients with traumatic intracranial hemorrhage. World J Emerg Surg 2012; 7 (01) 25
• 12 Nielsen PB, Larsen TB, Skjøth F, Lip GY. Outcomes associated with resuming warfarin treatment after hemorrhagic stroke or traumatic intracranial hemorrhage in patients with atrial fibrillation. JAMA Intern Med 2017; 177 (04) 563-570