Diagnostic accuracy of whole-brain CT perfusion in the detection of acute infratentorial infarctions.

Abstract

Although the diagnostic performance of whole-brain computed tomographic perfusion (WB-CTP) in the detection of supratentorial infarctions is well established, its value in the detection of infratentorial strokes remains less well defined. We examined its diagnostic accuracy in the detection of infratentorial infarctions and compared it to nonenhanced computed tomography (NECT), aiming to identify factors influencing its detection rate. Out of a cohort of 1380 patients who underwent WB-CTP due to suspected stroke, we retrospectively included all patients with MRI-confirmed infratentorial strokes and compared it to control patients without infratentorial strokes. Two blinded readers evaluated NECT and four different CTP maps independently for the presence and location of infratentorial ischemic perfusion deficits. The study was designed as a retrospective case-control study and included 280 patients (cases/controls=1/3). WB-CTP revealed a greater diagnostic sensitivity than NECT (41.4 vs. 17.1%, P=0.003). The specificity, however, was comparable (93.3 vs. 95.0%). Mean transit time (MTT) and time to drain (TTD) were the most sensitive (41.4 and 40.0%) and cerebral blood volume (CBV) the most specific (99.5%) perfusion maps. Infarctions detected using WB-CTP were significantly larger than those not detected (15.0 vs. 2.2ml; P=0.0007); infarct location, however, did not influence the detection rate. The detection of infratentorial infarctions can be improved by assessing WB-CTP as part of the multimodal stroke workup. However, it remains a diagnostic challenge, especially small volume infarctions in the brainstem are likely to be missed.