Predictive value of Tmax perfusion maps on final core in acute ischemic stroke: an observational single-center study

Fabrizio Giammello*, Sara Rosa Maria de Martino, Luigi Simonetti, Raffaele Agati, Stella Battaglia, Luigi Cirillo, Mauro Gentile, Ludovica Migliaccio, Stefano Forlivesi, Michele Romoli, Ciro Princiotta, Caterina Tonon, Silvia Stagni, Simone Galluzzo, Raffaele Lodi, Giuseppe Trimarchi, Antonio Toscano, Rosa Fortunata Musolino, Andrea Zini

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review


Purpose: To assess utility of computed tomography perfusion (CTP) protocols for selection of patients with acute ischemic stroke (AIS) for reperfusive treatments and compare the diagnostic accuracy (ACC) in predicting follow-up infarction, using time-to-maximum (Tmax) maps. Methods: We retrospectively reviewed consecutive AIS patients evaluated for reperfusive treatments at comprehensive stroke center, employing a multimodal computed tomography. To assess prognostic accuracy of CTP summary maps in predicting final infarct area (FIA) in AIS patients, we assumed the best correlation between non-viable tissue (NVT) and FIA in early and fully recanalized patients and/or in patients with favorable clinical response (FCR). On the other hand, the tissue at risk (TAR) should better correlate with FIA in untreated patients and in treatment failure. Results: We enrolled 158 patients, for which CTP maps with Tmax thresholds of 9.5 s and 16 s, presented sensitivity of 82.5%, specificity of 74.6%, and ACC of 75.9%. In patients selected for perfusion deficit in anterior circulation territory, CTP-Tmax > 16 s has proven relatively reliable to identify NVT in FCR patients, with a tendency to overestimate NVT. Similarly, CTP-Tmax > 9.5 s was reliable for TAR, but it was overestimated comparing to FIA, in patients with unfavorable outcomes. Conclusions: In our experience, Tmax thresholds have proven sufficiently reliable to identify global hypoperfusion, with tendency to overestimate both NVT and TAR, not yielding satisfactory differentiation between true penumbra and benign oligoemia. In particular, the overestimation of NVT could have serious consequences in not selecting potential candidates for a reperfusion treatment.
Original languageEnglish
Pages (from-to)414-425
Number of pages12
JournalRadiologia medica
Issue number4
Early online date2022
Publication statusPublished - Apr 2022
Externally publishedYes

Cite this