Prediction of Malignant MCA Infarction With DWI: Pitfalls in Hyperacute Stroke
To the Editor:
We read with great interest the excellent article by Oppenheim et al1 regarding the use of diffusion-weighted imaging (DWI) for the prediction of malignant middle cerebral artery (MCA) infarction. The authors report impressive sensitivity and specificity rates of 100% and 94%, respectively, for the parameter “acute DWI lesion volume >145 cm3” predicting subsequent malignant MCA infarction. Thus, in their study only patients with an initial DWI lesion volume exceeding 145 cm3 went on to develop malignant MCA infarction. The mean time between symptom onset and the DWI measurements was 6.5±3.5 hours (range 2 to 14 hours).
However, although we strongly agree that acute DWI lesion volume has the potential to become a powerful predictor of subsequent malignant MCA infarction, we would like to point out that the sensitivity of this parameter may be considerably below 100% in the hyperacute phase (<3 hours). We have recently seen 2 patients presenting within 2 hours of symptom onset, 1 with a small (6.3 cm3 ) and 1 without a diffusion deficit (but with persistent MCA occlusion and extensive perfusion deficits), who both went on to develop complete MCA territory infarction. One of these 2 patients (Figure⇓) developed true malignant MCA infarction with signs of beginning uncal herniation and was treated with hemicraniectomy. In the other, somewhat older, patient (73 versus 64 years), only preexisting cerebral atrophy prevented malignant brain swelling leading to herniation, while the increase in DWI lesion size was similarly dramatic.
The 2 cases presented here, although admittedly extreme examples, highlight a more general problem concerning the use of DWI lesion volume as a parameter for the prediction of malignant MCA infarction. As shown by several groups2 3 4 5 and as discussed by Oppenheim et al,1 DWI lesion evolution in acute stroke is a highly dynamic process, particularly in the first hours after symptom onset. DWI lesions can enlarge substantially, and it is easily conceivable that, in a given patient, DWI lesion volume will increase from below 145 cm3 to above that threshold. Thus, in the authors’ opinion it will be necessary to define an optimal time window for the measurement of the DWI lesion volume. Clearly, this “optimal time window” will have to start beyond the time of the early and often unpredictable changes in DWI lesion volume and will have to end before potential treatment forms, such as hemicraniectomy or hypothermia, become less efficacious.
In the hyperacute phase, other parameters, such as a complete MCA territory perfusion deficit or MCA occlusion shown on MR angiography, may be more predictive of malignant MCA infarction than DWI lesion volume.6 7 However, evidence of severe tissue damage more definitive than a perfusion deficit appears desirable before embarking on potentially lifesaving but still experimental measures such as hemicraniectomy. “DWI lesion volume exceeding 145 cm3” may be such a parameter, when measured in the yet-to-be-defined optimal time window.
- Copyright © 2001 by American Heart Association
Oppenheim C, Samson Y, Manai R, Lalam T, Vandamme X, Crozier S, Srour A, Cornu P, Dormont D, Rancurel G, Marsault C. Prediction of malignant middle cerebral artery infarction by diffusion-weighted imaging. Stroke. 2000;31:2175–2181.
Schwamm LH, Koroshetz WJ, Sorensen AG, Wang B, Copen WA, Budzik R, Rordorf G, Buonanno FS, Schaefer PW, Gonzalez RG. Time course of lesion development in patients with acute stroke: serial diffusion- and hemodynamic-weighted magnetic resonance imaging. Stroke. 1998;29:2268–2276.
Barber PA, Darby DG, Desmond PM, Yang Q, Gerraty RP, Jolley D, Donnan GA, Tress BM, Davis SM. Prediction of stroke outcome with echoplanar perfusion- and diffusion-weighted MRI. Neurology. 1998;51:418–426.
Barber PA, Davis SM, Darby DG, Desmond PM, Gerraty RP, Yang Q, Jolley D, Donnan GA, Tress BM. Absent middle cerebral artery flow predicts the presence and evolution of the ischemic penumbra. Neurology. 1999;52:1125–1132.
We thank the Frankfurt group for their very nice comments on our paper regarding the use of DWI for the prediction of malignant MCA infarction.R1 This group raises the issue of DWI predictive value at the hyperacute phase (<3 hours), describing 2 patients with initially small DWI abnormalities and persistent MCA occlusion who ultimately develop very large MCA infarcts. One of the 2 patients develops malignant edema. In our nonmalignant group of patients, the final size of the infarct was often much larger than the initial DWI abnormality, but none of the patients had a life-threatening malignant edema. However, few of them were initially studied before 3 hours, and we totally agree that there is a need for defining an “optimal time window” for DWI-based prediction of malignant MCA infarct beyond the first 2 to 3 hours.
Yet, we believe that “brain-time” will become increasingly more important than “clock-time”; that patients with a high NIHS score, large perfusion defect, small DWI abnormality with moderate ADC decrease,R2 and persistent MCA occlusion should first be considered excellent candidates for intravenous or intra-arterial thrombolysis; and that hemicraniectomy should only be considered several hours later if reperfusion has not been achieved and if the DWI abnormality has been rapidly growing >145 cm3. Finally, our clinical experience is that “slowly growing” infarctions (>24 hours) may achieve very large volume without occurrence of malignant edema. This, again, points to the importance of defining a precise time window for DWI prediction of malignant edema. Future studies are likely to show that the optimal time window is somewhere between 0 to 3 hours and 24 to 48 hours after stroke.
Oppenheim C, Samson Y, Manaï R, Lalam T, Vandamme X, Crozier S, Srour A, Cornu P, Dormont D, Rancurel G, Marsault C. Prediction of malignant middle cerebral artery infarction by diffusion-weighted imaging. Stroke. 2000;31:2175–2181.
Tong DC, Adami A, Moseley ME, Marks MP. Relationship between apparent diffusion coefficient and subsequent hemorrhagic transformation following acute ischemic stroke. Stroke. 2000;31:2378–2384.