J. Broderick et al., Apolipoprotein E phenotype and the efficacy of intravenous tissue plasminogen activator in acute ischemic stroke, ANN NEUROL, 49(6), 2001, pp. 736-744
We used stored plasma samples from 409 patients in the National Institute o
f Neurological Diseases and Stroke (NINDS) tissue plasminogen activator (t-
PA) Stroke Trial to examine the relationship between an apolipoprotein (Apo
) E2 or an Apo E4 phenotype and a favorable outcome 3 months after stroke,
the risk of intracerebral hemorrhage, and the response to intravenous t-PA
therapy. For the 27 patients with an Apo E2 phenotype who were treated with
t-PA, the odds ratio (OR) of a favorable outcome at 3 months was 6.4 [95%
confidence interval (CI) 2.7-15.3%] compared to the 161 patients without an
Apo E2 phenotype who were treated with placebo. The 190 patients treated w
ith t-PA who did not have an Apo E2 phenotype also had a greater, though le
ss pronounced, likelihood of a favorable outcome (OR 2.0, 95% CI 1.2-3.2%)
than patients without an Apo E2 phenotype treated with placebo. For the 31
patients with an Apo E2 phenotype treated with placebo, the OR of a favorab
le 3 month outcome was 0.8 (95% CI 0.4-1.7%) compared to the 161 patients w
ithout an Apo E2 phenotype treated with placebo. This interaction between t
reatment and Apo E2 status persisted after adjustment for baseline variable
s previously associated with 3 month outcome, for differences in the baseli
ne variables in the two treatment groups and in the Apo E2-positive and -ne
gative groups, and for a previously reported time-to-treatment x treatment
interaction (p = 0.03). Apo E4 phenotype, present in 111 (27%) of the 409 p
atients, was not related to a favorable 3 month outcome, response to t-PA,
3 month mortality, or risk of intracerebral hemorrhage. We conclude that th
e efficacy of intravenous t-PA in patients with acute ischemic stroke may b
e enhanced in patients who have an Apo E2 phenotype, whereas the Apo E2 phe
notype alone is not associated with a detectable benefit on stroke outcome
at 3 months in patients not given t-PA. In contrast to prior studies of hea
d injury and stroke, we could not detect a relationship between Apo E4 phen
otype and clinical outcome.