Stroke: Ticlodipine decreases stroke, MI or vascular death.

Clinical bottom line (level 1b)

  1. Patients with thromboembolic stroke who are given ticlopidine are less likely to have a stroke, MI or vascular death, than those given placebo (NNT = 19 at 3 years) .
  2. Patients given ticlopidine are more likley to have adverse effects that will cause discontinuation of therapy, than those given placebo (NNH = 11 at 3 years) .
  3. There is no clear difference in stroke, MI or death, stroke and stroke death, vascular death or death alone.
Gent et al: Lancet 1989; : 1215-1220
Expires March 2003

The study

Double-blinded concealed randomised trial with intention-to-treat
Setting: 25 clinical centres, Canada and USA

1072 patients (aged mean 66 years, 60% male) Thromboembolic stroke no less than 1 week or more than 4 months before study entry. Diagnosis was based on a neurological evaluation and assessment of clinical course, and required a sudden onset of a new neurological deficit with demonstrable residual effects at the time of randomisation. A CT scan was also required.

Excluded if
  • likely to remain bedridden
  • severe comorbidity
  • contraindications to ticlopidine
  • requirement of long-term treatment with anticoagulants or antiplatelets
  • cardioembolic stroke


    • Note:
  • Code broken for 4 patients during the course of the study at the insistence of local physicians (2 patients in each group)
  • Randomisation schedules were generated separately for each centre.


    • Control Group: (n = 528, 528 analysed): placebo given two times daily as study drug
    Experimental Group: (n = 525, 525 analysed): ticlopidine two 125 g tablets twice daily

    98% followed for 3 years Follow-up was for a mean of 24 months, with a maximum of 3 years.

    The evidence

    Outcome Time to outcome CEREERRRR
    (95% CI)    		ARR
    (95% CI)    		NNT
    (95% CI)
    stroke, MI or vascular death 3 years 134
    (25.4%)    		 106
    (20.2%)    		 20.0%
    (0.00% to 0.36%)    		 5.19%
    (0.13% to 10.2%)    		 19
    (10 to 760)
    adverse effects severe enough to cause discontinuation of drug 3 years 15
    (2.84%)    		 62
    (11.8%)    		 -316%
    (-621% to -140%)    		 -8.97%
    (-12.1% to -5.87%)    		 -11
    (-17 to -8)
    adverse experience at any time during study 3 years 181
    (34.3%)    		 283
    (53.9%)    		 -57.0%
    (-81.0% to -36.0%)    		 -19.6%
    (-25.5% to -13.7%)    		 -5
    (-7 to -4)
    stroke, MI or death 3 years 156
    (29.6%)    		 134
    (25.5%)    		 14.0%
    (-5.00% to 29.0%)    		 4.02%
    (-1.37% to 9.41%)    		 25
    (NNT = 73 to infinity;
    NNH = 11 to infinity)
    stroke and stroke death 3 years 98
    (18.6%)    		 80
    (15.2%)    		 18.0%
    (-7.00% to 37.0%)    		 3.32%
    (-1.20% to 7.84%)    		 30
    (NNT = 83 to infinity;
    NNH = 13 to infinity)
    vascular death 3 years 41
    (7.77%)    		 32
    (6.10%)    		 22.0%
    (-23.0% to 50.0%)    		 1.67%
    (-1.40% to 4.74%)    		 60
    (NNT = 72 to infinity;
    NNH = 21 to infinity)
    death 3 years 64
    (12.1%)    		 64
    (12.2%)    		 -1.00%
    (-39.0% to 27.0%)    		 -0.07%
    (-4.02% to 3.88%)    		 -1444
    (NNT = 25 to infinity;
    NNH = 26 to infinity)

    Comments

    1. Given the side-effect profile of ticlopidine, is is generally reserved for those who cannot take aspirin.
    2. 19 patients were found to be ineligible after randomisation.

    Citation

    1. Gent M, Blakely JA, Easton JD, et al: The Canadian American ticlopidine study (CATS) in thromboembolic stroke. Lancet 1989; : 1215-1220
    Contributor: Clare Wotton and Musab Hayatli, December 1999
    Reviewer: Daniel Sontheimer

    Clinical Question.
    Patient thromboembolic stroke
    Intervention or Exposure ticlopidine
    Comparison placebo
    Outcome subsequent stroke, MI or vascular death