Description: Purpose: To assess the predictive value of reperfusion indices, recanalization, and important baseline clinical and radiologic scores for good clinical outcome prediction. Materials and Methods: The study was approved by the local research ethics board. Written consent was obtained from all participants or their caregivers. Baseline computed tomography (CT) perfusion less than 4.5 hours after stroke symptoms, follow-up CT perfusion at 24 hours or less, and 5–7-day magnetic resonance images were obtained for 114 patients. Baseline imaging was assessed blinded to outcome. Recanalization status was determined at follow-up CT angiography. Reperfusion index was calculated on baseline and on follow-up at-risk tissue volume. Kruskal-Wallis, Mann-Whitney rank sum, and Spearman correlation were used for group comparisons and correlation studies. Univariate and multivariate logistic regression tested the association of clinical and imaging parameters with good outcome. Models with and without recanalization and reperfusion were compared by using Akaike information criterion. Results: Reperfusion indices were significantly higher in patients with recanalization than in those without ( P 〈 .001). Despite significance of recanalization at univariate analysis, only reperfusion, age, and National Institutes of Health Stroke Scale score were significant after multivariate analysis ( P 〈 .01). Time to maximum reperfusion index had the highest accuracy (area under the receiver operating characteristic curve, 0.70) for good outcome, and reperfusion was defined as time to maximum volume of 59% or greater. Patients with reperfusion but no recanalization had significantly lower total infarct volume ( P = .001) and infarct growth ( P = .004) and had higher salvaged penumbra ( P = .009) volumes than patients without reperfusion and recanalization. A final model with reperfusion but not recanalization was the most prognostic model of good clinical outcome. Conclusion: Reperfusion showed stronger association with good clinical outcome than did recanalization. © RSNA, 2013

Description: Purpose: To evaluate whether treatment with an aromatase inhibitor (AI) influences background parenchymal enhancement (BPE) or amount of fibroglandular tissue (FGT) at breast magnetic resonance (MR) imaging in postmenopausal women with prior history of breast cancer. Materials and Methods: A waiver of authorization and patient consent was granted by the institutional review board for this HIPAA-compliant retrospective study. Postmenopausal women with breast cancer and MR imaging findings of the contralateral unaffected breast, before and during 6–12 months of AI treatment (anastrozole, letrozole, or exemestane), between August 1999 and June 2010 were retrospectively identified ( n = 149). Two readers performed blinded side-by-side comparison of BPE and MR imaging–depicted FGT before and during treatment. BPE and FGT were classified as the same or greater on one of the two MR studies and by using categorical scales: minimal, mild, moderate, or marked for BPE and fatty, scattered, heterogeneously dense, or dense for FGT. Consensus was reached in cases of disagreement. The sign test was used to conduct a side-by-side comparison of BPE and FGT before and during AI treatment. Results: A decrease in BPE occurred in 33.9% (37 of 109) of women during anastrozole treatment, while an increase occurred in only one ( P 〈 .0001); 28 of 37 decreases resulted in a category change of BPE. A decrease in MR imaging–depicted FGT occurred in 5.5% (six of 109) of women, while no increases occurred ( P = .031). During letrozole treatment, a decrease in BPE occurred in 46% (15 of 33), while an increase occurred in one woman ( P = .0003); a decrease in FGT occurred in only one woman, and no increases occurred. Similar results were seen when women also undergoing chemotherapy were excluded. Only seven women were treated with exemestane. Conclusion: Treatment with 6–12 months of anastrozole or letrozole was associated with decreases in BPE, which occurred in a greater proportion of women than decreases in FGT. © RSNA, 2012