Restarting anticoagulants after an intracerebral hemorrhage (ICH) is one of the challenges that no physician feels happy to face: the available evidence is very weak to support any strong suggestion. Unfortunately, this is a rather frequent scenario given the large number of patients who are treated with anticoagulants, and the fact that intracerebral hemorrhage is not a rare complication of anticoagulants. Given the lack of any relevant randomized controlled trial, one needs to weigh the thromboembolic risk against the risk for re-bleeding to reach a clinical decision. Obviously, the long-term thromboembolic risk depends on the underlying condition with the most frequent causes being atrial fibrillation (AF), mechanical valves and venous thromboembolism (VTE). Moreover, one should keep in mind that a post-ICH patient is frequently bed-ridden, which further increases the risk for VTE regardless of and in addition to the underlying condition that initially required treatment with anticoagulants. On the other hand, the risk of ICH recurrence seems to be associated with ICH location [1], age [2], apolipoprotein E ε2 or ε4 alleles [3] and presence of microbleeds on T2*-weighted gradient-echo MRI [4].

Unfortunately, there are no randomized data available to guide our decision about reinitiating anticoagulation when encountering a patient with anticoagulant-associated ICH. Only a few observational studies have investigated whether reinitiation or avoidance of anticoagulants is the preferable strategy. Recently, the CHIRONE study shows that the rate of ICH recurrence after reinitiation of vitamin-K antagonists is 2.56 per 100 patient-years during a median follow-up of 16.5 months, (of which, only 0.4 cases per 100 patient-years were fatal) [5]. Of note, there was no hemorrhage outside the central nervous system during the follow-up of 778 patient-years [5]. A recent report by the Registry of the Canadian Stroke Network shows that reinitiation of warfarin after ICH does not increase mortality or bleeding in patients with high thrombotic risk [6]. In similar, another single-center study in 52 patients who survived a warfarin-associated ICH shows that there is no statistically significant difference in outcome between patients who restart warfarin and patients who do not [7]. Of note, the rate of recurrent ICH after reinstitution of warfarin therapy is numerically lower than the rate of recurrent thromboembolic events in patients who do not restart warfarin therapy [7].

Of course, one should be very cautious when trying to draw conclusions from these studies given the inherent limitations of any observational non-randomized study. This is also reflected in the Guidelines of the European Stroke Organization about the management of spontaneous ICH in which no specific evidence-based suggestion is made due to the very low quality of available data [8]. But still, these are the best data available and at the bottom line, any inferences need to be drawn from these data. And these data imply that the risk for ICH recurrence in patients who reinitiate warfarin is not higher than the thromboembolic risk in those who do not reinitiate warfarin.

Nevertheless, our choice about our warfarin-associated ICH patients may finally be even more straightforward in the era of the new oral anticoagulants: during recent years, four new oral anticoagulants were launched based on a series of large randomized controlled trials. Apixaban, dabigatran, edoxaban and rivaroxaban were all shown to be at least as effective as warfarin for primary and secondary stroke prevention in patients with AF [9–14], as well as for extended treatment of VTE [15–19]. Importantly, the huge advantage of these drugs against warfarin is their excellent safety profile: they are all associated with a large reduction of bleeding risk, and in particularly ICH risk [13, 14], which provides further support to the strategy of reinitiating anticoagulation in an anticoagulant-ICH patient, in particular using one of these new agents. Unfortunately, one should keep in mind that this is the case for patients with AF or VTE, but not for patients with mechanical heart valves, for whom dabigatran was recently shown to be associated with worse outcome compared to warfarin [20].

Hopefully, more evidence-based suggestions may become possible when the ongoing RESTART trial (www.RESTARTtrial.org) is completed. Until then, current data show that there is weak evidence that reinitiation of anticoagulants in patients with warfarin-associated ICH does not increase mortality and bleeding risk, and in this context, they argue in favor of restarting anticoagulants in these patients. Moreover, the substitution of warfarin by one of the new oral anticoagulants seems to further enhance the safety of this strategy for patients with AF or VTE, but not for patients with mechanical valves.

Needless to say, of course, that the patient should definitely be a part of the decision process after being thoroughly informed about all potential risks and strategies.