Background and purpose: The importance of postmenopausal estrogen replacement therapy for stroke in females remains controversial. We previously showed that female rats sustain less infarction in reversible middle cerebral artery occlusion (MCAO) than their ovariectomized counterparts and that vascular mechanisms are partly responsible for improved tissue outcomes. Furthermore, exogenous estrogen strongly protects the male brain, even when administered in a single injection before MCAO injection. The present study examined the hypothesis that replacement of 17beta-estradiol to physiological levels improves stroke outcome in ovariectomized, estrogen-deficient female rats, acting through blood flow-mediated mechanisms.

Methods: Age-matched, adult female Wistar rats were ovariectomized and treated with 0, 25, or 100 microgram of 17beta-estradiol administered through a subcutaneous implant or with a single Premarin (USP) injection (1 mg/kg) given immediately before ischemia was induced (n=10 per group). Each animal subsequently underwent 2 hours of MCAO by the intraluminal filament technique, followed by 22 hours of reperfusion. Ipsilateral parietal cortex perfusion was monitored by laser-Doppler flowmetry throughout ischemia. Cortical and caudate-putamen infarction volumes were determined by 2,3, 5-triphenyltetrazolium chloride staining and digital image analysis. End-ischemic regional cerebral blood flow was measured in ovariectomized females with 0- or 25-microgram implants (n=4 per group) by (14)C-iodoantipyrine quantitative autoradiography.

Results: Plasma estradiol levels were 3.0+/-0.6, 20+/-8, and 46+/-10 pg/mL in the 0-, 25-, and 100-microgram groups, respectively. Caudate-putamen infarction (% of ipsilateral caudate-putamen) was reduced by long-term, 25-microgram estrogen treatment (13+/-4% versus 31+/-6% in the 0-microgram group, P<0.05, and 22+/-3% in the 100-microgram group). Similarly, cortical infarction (% of ipsilateral cortex) was reduced only in the 25-microgram group (3+/-2% versus 12+/-3% in the 0-microgram group, P<0.05, and 6+/-3% in the 100-microgram group. End-ischemic striatal or cortical blood flow was not altered by estrogen treatment at the neuroprotective dose. Infarction volume was unchanged by acute treatment before MCAO when estrogen-treated animals were compared with saline vehicle-treated animals.

Conclusions: Long-term estradiol replacement within a low physiological range ameliorates ischemic brain injury in previously ovariectomized female rats. The neuroprotective mechanism is flow-independent, not through preservation of residual ischemic regional cerebral blood flow. Furthermore, the therapeutic range is narrow, because the benefit of estrogen in transient vascular occlusion is diminished at larger doses, which yield high, but still physiologically relevant, plasma 17beta-estradiol levels. Lastly, unlike in the male brain, single-injection estrogen exposure does not salvage ischemic tissue in the female brain. Therefore, although exogenous steroid therapy protects both male and female estrogen-deficient brain, the mechanism may not be identical and depends on long-term hormone augmentation in the female.