Clinical and sonographic patterns of tandem internal carotid artery/middle cerebral artery occlusion in tissue plasminogen activator-treated patients.

BACKGROUND AND PURPOSE The National Institutes of Health Stroke Scale (NIHSS) is predictive of thrombus presence but has limited ability to identify occlusion location in the anterior circulation. We describe clinical and sonographic patterns that are associated with tandem internal carotid artery (ICA) and middle cerebral artery (MCA) occlusions. METHODS Consecutive acute ischemic stroke patients receiving intravenous tissue plasminogen activator (TPA) were studied. Pretreatment NIHSS scores and bedside transcranial Doppler (TCD) were obtained for all patients. RESULTS A total of 95 patients treated with intravenous TPA at 132+/-60 minutes from stroke onset were studied. On TCD, 48 had isolated MCA occlusion (mean NIHSS 16.8+/-5.8, median 17, range 5 to 28); and 16 had tandem ICA/MCA occlusion (mean NIHSS 18.8+/-5.8, median 22, range 8 to 29; P=NS). In the MCA occlusion and tandem ICA/MCA occlusion groups, 19% and 11%, respectively, had NIHSS scores <12 points. Compared with the NIHSS scores in patients with hemiplegia, forced gaze deviation, and complete neglect, the lower NIHSS scores were attributable to partial arm and/or leg paresis, gaze preference, and partial neglect. In those patients, TCD showed > or =2 major collateral channels and low-resistance flow at the M1 origin, suggesting perfusion of perforating arteries. Although TCD cannot differentiate between high-grade ICA stenosis or occlusion, collateral flow patterns and stenotic signals at the terminal ICA differentiated tandem lesions from isolated MCA occlusion (P<0.01). CONCLUSIONS Tandem ICA/MCA occlusion was found on TCD in 17% of TPA-treated patients. NIHSS scores were similar in patients with isolated MCA and tandem occlusions. Lower NIHSS scores were seen in patients with a higher number of major collateral flow channels and higher Thrombolysis in Brain Ischemia (TIBI) flow grades at the MCA origin.