McTaggart RA, Jovin TG, Lansberg MG, Mlynash M, Jayaraman MV, Choudhri OA, et al. Alberta Stroke Program Early Computed Tomographic Scoring Performance in a Series of Patients Undergoing Computed Tomography and MRI: Reader Agreement, Modality Agreement, and Outcome Prediction. Stroke. 2014
In his writings, Hippocrates first described the phenomenon of sudden paralysis that is often associated with ischemia. The Greek term apoplexy, meaning “struck down with violence”, was used to describe this occurrence. The word “stroke” was derived from this original description and continues to be used to this day.
In addition to providing accurate clinical descriptions of stroke, Hippocrates started to take things a step further in 400 B.C. He wrote, “When persons in good health are suddenly seized with pains in the head and straightaway are laid down speechless and breathe with stertor, they die in seven days.” Essentially, he provided the basis for a predictive model for patients with hemorrhagic stroke. He was able to generate a relatively accurate prognosis for patients who presented with these symptoms based on his clinical observations.
Today, we continue to look for ways to better prognosticate and predict outcome for our patients. This quest is usually driven by patients and their families and is echoed in the phrase most often heard by physicians: “Am I, or is he/she going to be alright?” Or, will this be the end?
For ischemic stroke, previous studies have validated the Alberta Stroke Program Early CT Scoring (ASPECTS) as a straightforward and quick way to use non-contrast CT scans to predict functional outcome and hemorrhage risk in patients eligible for IV tPA. The ASPECTS score was later adapted to the MRI since this modality was known to be more sensitive and specific for diagnosing stroke.
Even though the ASPECTS was originally designed to aid in decision making about which patients were more likely to benefit from IV tPA, it was also used to predict outcome and likelihood of reperfusion in the IMS III study.
So far, we had a score that used the non-contrast CT scan prior to treatment to help us predict outcome for patients about to receive IV tPA. But, the MRI was found to be a better modality for predicting outcome for IV tPA patients. More recently, the non-contrast CT version of this score was also validated for patients who not only received IV tPA, but also underwent endovascular therapy. The most logical next step was to try the MRI version of this score and see which one did a better job of predicting outcome: the non-contrast CT version of the ASPECTS score, or the diffusion-weighted MRI version. That’s what Dr. McTaggart and his colleagues did.
They prospectively evaluated 74 patients with acute ischemic stroke. All patients had a non-contrast CT and an MRI prior to undergoing endovascular treatment within 12-hours of symptom onset. Based on their imaging, these patients were then given an ASPECT score with both the diffusion-weighted MRI and the non-contrast CT. Each patient had two scores, one derived from each modality. Two expert clinicians determined these scores completely independently of each other (i.e. they were blinded to the other’s scoring). A third scorer then compared the scoring of the original two scorers, and chose the more accurate of the two for inclusion in the study. The scores were then compared with the modified Rankin Score (mRS) at 90-days.
To determine if different ASPECT scores correlated with outcome, the scale was evaluated in several different ways: The entire scale (ASPECTS 1-10), two groups (ASPECTS 0-7 and 8-10) and three groups (ASPECTS 0-4, 5-7 or 8-10), were compared with the 90-day mRS.
Although the sample size was not very large, the results were significant. Dr. McTaggart and colleagues found that the DWI-ASPECTS and CT-ASPECTS correlated well with the DWI volume. However, the DWI-ASPECTS correlated better with DWI volume. Furthermore, the area under the ROC curve (AUC) for predicting good outcome for DWI-ASPECTS was 0.705 (95% CI, 0.587-0.822), 0.581 (95% CI, 0.440-0.722) for time interval from stroke onset to procedure start, and 0.548 (95% CI, 0.411-0.685) for the CT-ASPECTS. The AUC difference for DWI-ASPECTS vs. CT-ASPECTS was significant (p=0.026) whereas the difference between DWI-ASPECTS and time interval from stroke onset to cath lab was not (p=0.20).
DWI-ASPECTS grouped into three (8-10, 5-7, and 0-4), correlated with good functional outcome (GFO) at rates of were 53%, 27%, and 10%, (p=0.004 for this trend), whereas CT-ASPECTS did not (corresponding GFO rates were 39%, 37%, and 20%, p=0.534).
While reading this paper, there were two major questions that came up and required answering. Fortunately, these questions were answered in the discussion the authors provided.
The primary question concerned the time difference between when the CT scan of the head was obtained and when the procedure was performed and when the MRI was performed in relation to the treatment and eventual outcome. Since this data was obtained from the DEFFUSE-2 trial, it was pointed out that the CT head was always obtained before the MRI was obtained. So, for evolving strokes, it only makes sense that the MRI would be more accurate since it is capturing the evolution at a later time than the CT scan. However, the authors attempt to dispel this notion by stating that the mean DWI- ASPECTS scores were not significantly lower than the CT scores (7.6 vs 7.0) and the difference between the scores did not increase in relation to the time difference between CT and MRI.
The second question was more of a surprised exclamation ending in a question: “How could time to reperfusion be less important that the DWI-ASPECTS?” The authors answered this question by suggesting that the finding of DWI-ASPECTS being a stronger predictor of outcome than time-to-reperfusion may reflect the variable growth rate of DWI lesions over time, and implies that the volume of irreversible tissue injury is a more potent predictor of clinical outcome than time to reperfusion. Essentially, to quote the movie The Fifth Element, the authors are saying “time not important, only life, important”, the life of viable brain cells, that is.