NIHSS/EIC Mismatch Explains the >1⁄3 MCA Conundrum

>1⁄3 MCA Conundrum To the Editor: The controversy about the significance of early ischemic changes (EIC) on CT after acute stroke by von Kummer, Lyden, and Davis and Donnan1–3 nicely summarizes the issues except for what, in my mind, is the most likely explanation for the discrepancy between the NINDS and ECASS conclusions on this subject. I would also like to propose an alternative way of looking at the “more or less than one third MCA territory” controversy. The aim of treatment with intravenous tPA is to reperfuse tissue that is hypoperfused but still viable (penumbra). While there are undoubtedly other variables that are important, the overwhelming burden of animal and human data have shown that time is critical for the existence of penumbra; penumbral tissue is more plentiful the earlier the patient is treated after the stroke onset. Therefore, when NINDS patients were treated on average 90 minutes after stroke onset, the core of already infarcted tissue, represented by areas of EIC on CT, was in most cases likely still surrounded by substantial salvageable, normally appearing penumbral brain tissue on CT. Such patients could, and did, benefit from reperfusion after intravenous tPA. However, 240 to 300 minutes after the stroke, when the average ECASS patient was treated, the core would be larger (hence more patients with EIC in 1⁄3 MCA), and the regions of EIC would most likely be surrounded by relatively less penumbra. The result of treating such patients, therefore, is risk with less likelihood of much benefit. This logically would explain why EIC by themselves, whatever their extent, did not predict lack of benefit from tPA treatment in the NINDS patients, but why extensive EIC were associated with risk and no benefit in ECASS. This hypothesis is supported by clinical observation of patients seen within 3 hours of stroke onset in the middle cerebral artery territory in whom it is not uncommon to see EIC. In most of these patients with EIC, their clinical deficit is severe, with high NIHSS scores, representing dysfunction in more extensively involved hypoperfused penumbral tissue than would be expected from injury to just the areas of insular cortex, putamen, and capsular hypoattenuation that comprise their EIC. Analogous to the theory (still unproven) regarding the importance of PWI/DWI “mismatch” based on MRI studies, I hypothesize that patients with such “NIHSS/EIC mismatch” might benefit from intravenous tPA more than patients without such NIHSS/EIC mismatch, ie, with lower NIHSS scores that match EIC in 1⁄3 of the MCA territory, or high scores and extensive EIC in most of the MCA territory. This hypothesis could be easily tested by re-review of the NINDS study database. In summary, at least within 3 hours of symptom onset, the data indicate that the extent of EIC is not critical to decision-making regarding the use of tPA. It is more likely that whether there is continued existence of hypoperfused penumbral tissue is what is important. Based on the above reasoning, the differences between the NINDS and ECASS data on this issue make complete sense and should not foster confusion, further “controversy,” or excuse not to treat. For patients who meet published criteria and can be treated within 3 hours of symptom onset, until further data are forthcoming testing the validity of the mismatch theory (either MRI PWI/DWI mismatch or clinical NIHSS/EIC mismatch), the conclusion of the NINDS group should be followed and such patients should be treated even in the presence of EIC.