Salvage of the PWI/DWI mismatch up to 48 h from stroke onset leads to favorable clinical outcome


  • Funding: H Ma received NHMRC (Australia) Medical Post-Graduate Scholarship and Pfizer (Australia) CVL grant.


Background and Purpose

In acute ischemic stroke perfusion/diffusion-weighted image, mismatch using magnetic resonance imaging approximates the ischemic penumbra. For early time windows, mismatch salvage improves clinical outcomes, but uncertainty exists at later time epochs. We hypothesized that (a) mismatch may exist up to 48 h; (b) the proportion of mismatch salvage is time independent; and (c) when salvaged, it improves clinical outcomes.


Magnetic resonance imaging was performed within 48 h of ischemic stroke. Perfusion-weighted image was defined by relative Tmax two-second delay. Perfusion/diffusion-weighted image mismatch was the perfusion-weighted image not overlapped by the diffusion-weighted image when coregistered. Infarct volume and disability (modified Rankin Score) were assessed at three-months. Mismatch salvage was the region not overlapped by final infarction. Favorable outcome was defined as modified Rankin Score 0–1.


Sixty-six patients were studied [mean age 69·9 years (standard deviation 13·1), initial median National Institute of Health Stroke Scale 9·0 (interquartile range 6·0, 18·3)]. There was no relationship between time of stroke onset and the proportion of mismatch salvaged (P = 0·73). Age (adjusted odds ratio = 0·92, 95% confidence interval 0·86–0·98, P = 0·01), initial National Institute of Health Stroke Scale (adjusted odds ratio = 0·80, 95% confidence interval 0·70–0·92, P < 0·01), mismatch volume (adjusted odds ratio = 0·98, 95% confidence interval 0·968–0·1, P = 0·05), and percentage of mismatch salvage (adjusted odds ratio = 1·04, 95% confidence interval 0·99–1·07, P = 0·05) were independently associated with favorable outcome.


Using coregistered perfusion/diffusion-weighted image criteria, mismatch persists up to 48 h post stroke. For the whole group, the proportion of mismatch salvage remains independent of time and, although the effect is small, its salvage is independently associated with improved clinical outcomes at three-months. Larger sample sizes are needed to determine the time limit for mismatch salvage.