CADISS: a feasibility trial that answered its question

Anticoagulation after ischaemic stroke was a topic of major controversy for decades until a series of randomised clinical trials consistently showed no net benefi t of heparin or warfarin compared with aspirin. Any reduction in the risk of recurrent ischaemic stroke was off set by an increase in risk of intracerebral haemorrhages. One cause of stroke for which the eff ect of anticoagulation remains unclear is cervical artery dissection. The classic dogma was that a tear in the intima leads to formation of a fi brin-rich thrombus and poses a very high risk of artery-to-artery embolism or occlusion, making early anticoagulation the treatment of choice. 3 Furthermore, this risk was thought to persist for months until the artery heals. Countless experts recommended this approach, but acknowledged the lack of dependable data. Observational studies have yielded confl icting results. 4–6 A systematic review and meta-analysis 7 of 36 observational studies showed no signifi cant diff erences between anticoagulation and antiplatelet treatment for clinical outcomes or risk of recurrent stroke. For doctors inclined to use anticoagulation, the point estimates suggested a potential benefi t. For those who were sceptical, the neutral results provided no compelling reason to use anything other than antiplatelet treatment. American Heart Association/ American Stroke Association guidelines 8,9 left open both options, suggesting that either approach is reasonable for 3–6 months (class IIa, level B, a weak recommendation), although these guidelines otherwise strongly recommend antiplatelet treatment over anticoagulation for acute management and secondary prevention after non-cardioembolic stroke (class I, level A, their highest endorsement). The Cervical Artery Dissection in Stroke Study (CADISS) was the fi rst randomised clinical trial to compare anticoagulation with antiplatelet treatment for cervical artery dissection. It was designed as a pilot study, aiming to estimate the risk of recurrent stroke reliably and establish whether a defi nitive trial would be feasible. The main results, published in The Lancet Neurology, 10 are much the same as those from the other anticoagulation trials—any reduction in ischaemic stroke was counterbalanced by an increase in haemorrhages. However, events were rare, with only four ischaemic strokes (three in the antiplatelet group and one in the anticoagulant group) and one subarachnoid haemorrhage (in the anticoagulant group) among the 250 participants. The rarity of events meant that the study did not answer the question of which strategy was superior overall, but that was not expected from this pilot trial. Nevertheless, CADISS informs …