Editorial Comment—Neuroimaging After TIA: A Crystal Ball?
Previous studies of the Kaiser-Permanente Northern California database have indicated that 10.5% of patients will develop stroke in the 90 days after a TIA and that approximately half of these events will occur within 48 hours.1 In this issue of Stroke, Douglas et al query the same database to provide novel information regarding the short-term risk associated with a new infarction on head CT after symptoms of TIA.2 In their study, 4% of 322 patients had this finding, which was associated with an approximately 4-fold increase in the risk of stroke in the next 90 days, after adjusting for all other known risk factors. In absolute terms, 38% (5 of 13) of patients in the group with new infarction on initial head CT had a new stroke in the following 90 days compared with nearly 10% (30 of 309) of patients who had no such findings on initial imaging. Other imaging variables studied (including old infarction, periventricular white matter disease, atrophy, and vascular calcification) had no statistical relationship with follow-up events. An additional finding in this study, comparable to previous studies, was that 1% of patients had a nonischemic etiology as a presumed cause of symptoms.
What implications will these findings have on clinical practice? Considering that the current AHA guidelines already recommend imaging for patients after TIA,3,4 there will likely be no significant change among neurologists who see these patients primarily. However, many patients with TIA are not seen by neurologists and frequently do not undergo neuroimaging as part of the diagnostic evaluation. The findings in this study may prompt non-neurologists to pursue neuroimaging more routinely.
What implications will these findings have on clinical research? It seems likely that this study will be replicated using more sensitive markers of cerebral injury, such as diffusion-weighted imaging. Diffusion-weighted imaging changes are seen in approximately 50% of patients with TIA.5 In this study, new infarction on head CT correctly predicted recurrent stroke in 5 of 35 patients, representing a sensitivity of 14%. A CT negative for new infarction had a specificity of 97% (279 of 287). Therefore, the positive and negative predictive values for head CT in TIA, based on this study, are 38% (5 of 13) and 90% (279 of 309), respectively. Perhaps diffusion-weighted imaging will improve on these numbers and help focus research and diagnostic efforts on those truly at risk. The use of imaging variables plus clinical variables may ultimately help to create a powerful risk assessment tool.
Although Douglas et al have answered the question about the predictive power of head CT after TIA, an outstanding issue is the therapeutic utility of such testing. While an argument might be made that the correct diagnosis of TIA mimics leads to improved outcomes for the 1% of patients with other findings, there is no proof at this time that neuroimaging leads to a difference in outcomes for the other 99% of patients. Indeed, in this study, an identical percentage of patients (10.9%) who had CT testing (n=322) compared with those who did not (n=156) went on to develop a new stroke. Nevertheless, the work by Douglas et al represents an important step toward the goal of optimizing diagnostic testing and developing an effective prediction model for patients with TIA.
Douglas VC, Johnston CM, Elkins J, Sidney S, Gress DR, Johnston SC. Head computed tomography findings predict short-term risk after transient ischemic attack. Stroke. 2003; 34: 2894–2899.
Feinberg WM, Albers GW, Barnett HJM, Biller J, Caplan LR, Carter LP, Hart RG, Hobson RW II, Kronmal RA, Moore WS, et al. Guidelines for the management of transient ischemic attacks: from the Ad Hoc Committee on Guidelines for the Management of Transient Ischemic Attacks of the Stroke Council of the American Heart Association. AHA medical/scientific statement: special report. Circulation. 1994; 89: 2950–2965.
Culebras A, Kase CS, Masdeu JC, Fox AJ, Bryan RN, Grossman CB, Lee DH, Adams HP, Thies W. Practice guidelines for the use of imaging in transient ischemic attacks and acute stroke: a report of the Stroke Council, American Heart Association. Stroke. 1997; 28: 1480–1497.
Kidwell CS, Alger JR, Di Salle F, et al. Diffusion MRI in patients with transient ischemic attacks. Stroke. 1999; 30: 1174–1180.