safety of CT perfusions per FDA

For more information, please see FDA's Safety Investigation of CT Brain Perfusion Scans:  Initial Notification at:

  http://www.fda.gov/MedicalDevices/Safety/AlertsandNotices/ucm185898.htm <http://www.fda.gov/MedicalDevices/Safety/AlertsandNotices/ucm185898.htm>

d-dimer predicts cardioembolic stroke

D-dimer assay may differentiate subtypes of ischemic stroke; study of 126 patients hospitalized with acute ischemic stroke;
stroke subtypes were cardioembolic in 34 (27%),
atherothrombotic in 34 (27%),
lacunar in 31 (25%),
unknown in 27 (21%);

mean D-dimer levels on day 1 were 2.96 mcg/mL with cardioembolic stroke, 1.34 mcg/mL with atherothrombotic stroke and 0.67 mcg/mL with lacunar stroke; similar results at days 6 and 12; D-dimer > 2 mcg/mL had 59% sensitivity and 93% specificity for predicting cardioembolic stroke, D-dimer < 0.54 mcg/mL had 61% sensitivity and 96% specificity for predicting lacunar stroke

(Arch Intern Med 2002 Dec 9/23;162(22):2589)

Biomarkers for stroke

Foerch C, Montaner J, Furie KL,Ning MM, Lo EH. Searching for oracles: blood biomarkers for acute stroke . Invited Article. Neurology 2009; 73: 393-399.


Article discusses batteries of biomarkers used together to diagnose acute stroke or intracranial hemorrhage. They note that CT excludes hemorrhage, but does not include stroke, so a panel of markers may play a useful role.

To diagnose acute ischemic stroke, authors suggest a need for higher sensitivity than specificity to avoide missing treatable strokes.

NMDA antibodies are typically high early in stroke (first 3 hours) but may be false positive in patients with atherosclerosis or old strokes. New research focuses on fragments of NMDA rather than antibodies.

Reynolds et al. (Clin Chemisty2003) tested 50 markers and found a panel of four or five significantly predicted ischemic stroke. They were S100b, B type neurotropic growth factor, VWF, MMP-9, and monocyte chemotactic protein 1.They had a combined sens/spec of 91/97 for IS within 12 hours of stroke onset. The same group in a 2d study found S100b, MMP-9. vascular adhesion molecule, and VWF to have a combined sens/spec of 90/90. (Lynch JR Stroke 2004). Laskowitz found a panel of BNP, CRP, d-dimer, MMP-9, and S100b has a sens/spec of 81/70. Laskowitz et al. published a prospective multicenter trial (Stroke 2009) of 1100 patients using d-dimer, BNP, MMP-9, and S100b within 24 hours of stroke onset and found a s/s of 86/37.

For intracranial hemorrhage, markers include GFAP, which is also high in gliomas. In another analysis, RAGE and S100b best differentiated group from controls. ApoC1 and especially ApoC3 differentiate ICH from controls.

Montaner et al (Circulation 2003) found that MMP-9 was a powerful predictor of hemorrhagic risk in patients given alteplase. MMP-9 is also a predictor of HT in non lysed patients. Fibronectin has a similar pattern.

S100B which tends to correlate with infarct size, was also a predictor of malignant edema in one study (Foerch et al., Stroke 2004).