CADISS

Incomplete

Markus HS, et al. "Antiplatelet treatment compared with anticoagulation treatment for cervical artery dissection (CADISS): a randomized trial". Lancet Neurology. 2015. 14(4):361-367.
PubMed

Clinical Question

In patients with extracranial carotid or vertebral artery dissection, does treatment with anticoagulants, when compared to antiplatelet agents, reduce the risk of subsequent stroke or death?

Bottom Line

In patients presenting with symptomatic extracranial carotid and vertebral artery dissection, treatment with anticoagulation (heparin or lovenox bridging to warfarin) was not found to significantly lower the risk of subsequent stroke or death when compared to treatment with antiplatelet agents at 3 months (aspirin, dipyridamole, or clopidogrel alone or in combination). Risk of subsequent stroke in this population was found to be low in general, however, and lower than was once thought based on past observational studies.

Major Points

Dissection is an important cause of stroke and TIA in the young, and a common clinical intuition is that treatment with anticoagulation will reduce the risk of secondary stroke in this population. Until CADISS, no data existed to guide the choice of secondary prevention therapies, and it was not known if anticoagulation offered any benefit over potentially less hazardous anti platelet agents such as aspirin or clopidogrel. Prior systematic review of observational studies suggested no benefit to anticoagulation over antiplatelet agent

CADISS enrolled 250 patients (224 of whom presented with stroke or TIA, 26 with symptoms of the dissection itself such as headache, neck pain or Horner's syndrome), and found that the rate of recurrent stroke was only 2% within the 3 month follow-up period; secondary strokes occurred only in patients who had originally presented with stroke or TIA. 1 stroke occurred in the anticoagulation group vs. 3 strokes in the anti platelet group - a non significant difference (odds ratio 0.335, 95% CI 0.006-4.233; P-value=0.63). Notably dissection was radiographically confirmed in only 198 of the 250 patients, thus intention-to-treat vs. per-protocol analyses were particularly relevant to this study and were almost identical.

Guidelines

American Heart Association/American Stroke Association guidelines suggest that either antiplatelet or anticoagulation are reasonable for 3–6 months (class IIa, level B, a weak recommendation).

Design

• Multicenter, open-label, parallel-group, randomized, controlled trial
• N=250
  • Antiplatelet (n=126)
  • Anticoagulant (n=124)
• Setting: 46 hospitals with specialist stroke or neurology services in the UK and Australia
• Enrollment: November 2005 to May 2014
• Mean follow-up: 3 months
• Analysis: Intention-to-treat and Per-Protocol
• Primary outcome: Ipsilateral Stroke (ipsilateral territory stroke for carotid dissection, or vertebrobasilar territory stroke for vertebral dissection); Death

Population

Inclusion Criteria

• Extracranial carotid or vertebral artery dissection
• Onset of symptoms in the past 7 days
• Imaging evidence of definite or probable dissection by MRI, MRA, CTA or conventional angiography

Exclusion Criteria

• Intracranial cerebral artery dissection
• Contraindications to antiplatelet or anticoagulation agents
• Definite indication for or prior use of antiplatelet or anticoagulation agents (i.e., mechanical heart valves)

Baseline Characteristics

All figures listed refer to Intention-to-treat population (n=250)

• Mean age: 49 years
• Sex: 69% Male
• Time to randomization: 3.7 days
• Risk Factors: Smoking (51%), Hypertension (22%), Head/Neck Trauma (21%), Migraine (18%), Hyperlipidemia (17%), Diabetes (4%)
• Presenting Symptoms: Stoke (78%), TIA (19%), Headache (67%), Neck Pain (48%), Horner's Syndrome (24%)

Interventions

• Randomization to antiplatelet or anticoagulation, choice of therapy at the discretion of the local physicians
  • Heparin or Low Molecular Weight Heparin, bridging to Coumadin with goal INR 2-3
  • ASA, Dipyridamole, Clopidogrel, alone or in combination
• 3 month follow-up after randomization
• Repeat imaging with MRA or CTA at 3 months to assess recanalization

Outcomes

Primary Outcomes

Ipsilateral stroke or death (Intention-to-treat)

2% vs. 1% (OR 0.335; 95% CI 0.006-4.233; P=0.63)

Ipsilateral stroke or death (Per-protocol)

3% vs. 1% (OR 0.346; 95% CI 0.006-4.390; P=0.66)

Secondary Outcomes

Ipsilateral stroke, death, or major bleed (Intention-to-treat)

3% vs. 2% (OR 0·673; 95% CI 0·055– 5·983; P=1.00)

Any stroke or TIA (Intention-to-treat)

4% vs. 4% (OR 1·017; 95% CI 0·228–4·540; P=1·00)

Major bleeding (Both Intention-to-treat and Per-protocol)

0% vs. 1%

Subgroup Analysis

None performed

Adverse Events

No significant differences in the following:

Abdominal pain 0 (0%) 1 (1%) Abnormal liver function test 0 (0%) 1 (1%) Allergic reaction 1 (1%) 0 (0%) Chest pain 0 (0%) 1 (1%) Diplopia 1 (1%) 0 (0%) Dizziness 0 (0%) 1 (1%) Haematuria 0 (0%) 1 (1%) Haemoptysis 0 (0%) 1 (1%) Headache 4 (3%) 3 (2%) Hip pain 0 (0%) 1 (1%) Hydrocephalus 0 (0%) 1 (1%) Myocardial infarction 1 (1%) 0 (0%) Nausea or vomiting 3 (2%) 1 (1%) Neck pain 0 (0%) 1 (1%) Numbness 0 (0%) 1 (1%) Ophthalmic nerve neuralgia 1 (1%) 0 (0%) Pneumonia 1 (1%) 1 (1%) Seizure 0 (0%) 2 (2%) Subarachnoid haemorrhage 0 (0%) 1 (1%) Vision loss 0 (0%) 1 (1%) Worsening of ataxia 1 (1%) 0 (0%) Worsening of Horner’s syndrome 0 (0%) 1 (1%)

Criticisms

CADISS likely studied a population of patients with mostly spontaneous dissections, which lead to recurrent stroke at lower rates than do traumatic dissections. The study was also significantly under-powered to detect a statically significant difference in outcome: the authors' own power analysis determines that 10,000 patients would be needed, which given the long recruitment period of this study would seem a nearly impossible sample size to reach. The high-number of patients for which dissection was not confirmed on central review of radiographic imaging (20%) necessitated a per-protocol analysis. Study design also precluded patients with early severe subsequent stroke.

Funding

Stroke Association, a UK based charity Leading authors received research grants from the Stroke Association, otherwise no conflicts of interest