Muhammad Rizwan Husain, MD
@RIZWANHUSAINMD
Cervical artery dissection (CAD) has an annual incidence of about 2.9% per 100,000, though that seems to be under-reported, as many patients usually do not present for evaluation or undergo routine vessel imaging for local symptoms like pain or headache. At the same time, even though CAD accounts for about 1-2% of total ischemic strokes, it can be the cause of up to 25% of strokes in the young population.
The Cervical Artery Dissection in Stroke Study (CADISS) is the only prospective randomized clinical trial to date that evaluated if there was a reduction in subsequent stroke in patients treated with either antiplatelet or anticoagulation. It also looked at the presence of arterial recanalization between the two groups.
Inclusion in the trial required all patients to have an extra-cranial carotid or vertebral artery dissection with onset of symptoms within 7 days. Patients had to have a probable or definite diagnosis of dissection on MRA, CTA or conventional angiogram to be enrolled. Ultrasonography could be done to randomize patients in the trial, but this had to be confirmed later with CTA or MRI/MRA. Patients that were excluded from the trial included those that had an intracranial cerebral artery dissection, those that had contraindications to either antiplatelet or anticoagulation use and those that had an underlying medical condition requiring use of any particular group of anti-thrombotic. A total of 250 patients between November 2005 and May 2014 were enrolled. 126 were assigned to the antiplatelet and 124 to the anticoagulation group. Patients were followed up for 3 months. In the antiplatelet group, 55% were on a single agent while 45% were on dual antiplatelet. Choice of antiplatelet or its combination (Aspirin, clopidogrel, Dipyridamole, aspirin + Dipyridamole and Aspirin + Clopidogrel) was left at the discretion of the treating physician. The anticoagulation group was treated initially with either unfractionated heparin or a therapeutic dose of low-molecular-weight heparin followed by warfarin with a target INR of 2-3. The direct oral anticoagulants were not used as they were not approved for use when the trial was conducted. Both groups were followed for 3 months with repeat imaging with CTA or MRA.
Out of the initial 126 and 124 patients assigned to the antiplatelet group and anticoagulation group, respectively, several patients were excluded in both groups as they did not meet imaging criteria for dissection. The protocol analysis eventually included 101 patients in the antiplatelet group and 96 in the anticoagulation group. The initial trial results, reported first in 2015 (Lancet Neurology), demonstrated that in the intention to treat population, there was no difference in the risk of ipsilateral stroke or death between either group; 2% (Antiplatelet) vs. 1% (Anticoagulation) (OR 0.335; 95% CI 0.006-4.233; P=0.63), nor was this seen in the per protocol analysis: 3% (Antiplatelet) vs. 1% (Anticoagulation) (OR 0.346; 95% CI 0.006-4.390; P=0.66). Also, the 3-month risk of recurrent stroke was only 1.6%.
All these patients were subsequently followed up for up to 1 year, and these results were published as the final trial results in 2019. Between the three- to 12-month follow up period, there was only one recurrent stroke in each group (antiplatelet vs anticoagulation). Again, no significant difference was seen in the risk of recurrent stroke between patients on either antiplatelet or anticoagulation (OR 0.56 95% CI 0.10-3.21, P=0.51). In the intention to treat population, the rate of recurrent stroke at 1 year was also low at 2.4%. What was also important to note was that on follow up imaging after 3 months, no difference was noted in residual narrowing or occlusion between the antiplatelet vs the anticoagulation group (p=0.97).
The CADISS trial helped answer many questions regarding clinical management of extra-cranial cervical artery dissection by showing no difference between choice of antithrombotic use and risk of recurrent stroke, nor did it show any difference in angiographic recanalization between the two groups.
A few limitations of this trial are also important to note. The study sample size was small to detect a clear-cut difference in outcomes between the two groups (antiplatelet vs anticoagulation) and, hence, was under-powered. Different diagnostic imaging techniques (MRA, CTA) were used to evaluate for vessel recanalization, both imaging techniques having their own pitfalls. It would have been good to use one standard imaging modality. Furthermore, on the final imaging analysis, dissection could not be confirmed in about 20% of patients. Lastly, none of the novel anticoagulants (that are in common use today) were used.
