Response to Letter by Guedes and Ferro
We appreciate the attention of Guedes and Ferro to our meta-analysis of randomized controlled trials on the efficacy and safety of anticoagulant treatment in acute cardioembolic stroke.1 Indeed, our paper was submitted to Stroke on April 26, 2006, and thus before the ENS meeting held in September 2 to 5, 2006. After updating the meta-analysis by adding the ARGIS-1 study,2 Guedes and Ferro confirmed that the routine use of anticoagulants in the first 48 hours after an ischemic stroke of presumed cardioembolic origin cannot be recommended because of a bleeding excess. However, they found a favorable effect of anticoagulants on the prevention of recurrent ischemic stroke or stroke of unknown cause (odds ratio 0.64, CI between 0.47 to 0.88).
We would like to make a comment on the statistical analysis performed in this update meta-analysis. The authors used a fixed-effect model. Our understanding is that random effects models are recommended if heterogeneity through the studies is anticipated. This approach is endorsed by the Cochrane Collaboration. The trials on anticoagulants in acute stroke have remarkable differences in their design such as time to treatment, dose, comparator (control or aspirin), and study population. Therefore, heterogeneity is likely to be present. The test for heterogeneity used by Guedes and Ferro is probably underpowered and the fact that the probability value is nonsignificant does not mean that heterogeneity is absent. By using a random effect model in this update meta-analysis, the odds ratio for recurrence of stroke is 0.68 with an CI between 0.44 to 1.06 (P=0.09). Thus, the advantage in efficacy is not evident, confirming the results reported in Table 2 of our paper in absence of the ARGIS-1 results.
Guedes and Ferro call for a new pragmatic randomized clinical trial on anticoagulation in acute stroke. Our view is that there are consistent and solid data that allow to conclude that anticoagulants cannot be recommend in any subtype of acute ischemic stroke. One of the issues that are still open is the advantage of the very early administration of heparin (within 3 hours from stroke onset).3 For this reason, a trial within this time window but comparing anticoagulants with thrombolysis could be prospected.
G.A. received honoraria as a member of the Speaker Bureau of AstraZeneca and Bayer.
Paciaroni M, Agnelli G, Micheli S, Caso V. Efficacy and safety of anticoagulant treatment in acute cardioembolic stroke: a meta-analysis of randomized controlled trias. Stroke. 2007; 38: 423–430.
LaMonte MP, Nash ML, Wang DZ, Woolfenden AR, Schultz J, Hursting MJ, Brow PH. Argatroban anticoagulation in patients with acute ischemic stroke (ARGIS-1): a randomized, placebo-controlled safety study. Stroke. 2004; 35: 1677–1682.
Camerlingo M, Salvi P, Belloni G, Gamba T, Cesana BM, Mamoli A. Intravenous heparin starter within the first 3 hours after onset of symptoms as a treatment for acute nonlacunar hemispheric cerebral infarctions. Stroke. 2005; 36: 2415–2420.