We read with interest the article by Hamilton et al 1 in the March issue of Stroke. Although we were very encouraged by the finding that tissue plasminogen activator (t-PA) treatment reduced infarct size in this study, we have concerns regarding their finding of equal efficacy of intra-arterial and intravenous t-PA administration. Previously, using a similar rabbit cerebroemboli model, we found intra-arterial treatment to be significantly more efficacious than intravenous treatment as assessed

more »

... y vessel recanalization. 2 Because determining the optimal method of drug administration has important clinical implications, we feel that a comparison between the two studies is warranted. The main difference between the two studies is the method used to assess efficacy of thromborytic treatment. Hamilton et al 1 used a measurement of area of whole brain ischemic injury as the primary end point. We used Doppler ultrasound to determine the exact time course of vessel recanalization. Although Hamilton et al were able to detect a significant therapeutic benefit using the ischemic area technique, the inherent variation in this method makes it difficult to detect smaller differences in treatment efficacy. They found the area of ischemic injury in the intra-arterial administration group to be 4.6±4.1%; in the intravenous administration group the area was 3.4±2.6%. Because the standard deviation is approximately that of the entire area, the likelihood of a type 2 error would appear to be very high. In addition, the use of triphenyltetrazolium chloride staining to assess infarct area has been shown to overestimate actual infarct size. 3 In our study the intravenous t-PA infusion group had a 50% vessel recanalization rate at 2 hours, compared with a rate of 100% in the group treated with an equivalent amount of intra-arterial t-PA. A second potentially important difference is the method of thrombolytic agent administration. In the study by Hamilton et al 1 1 mg of t-PA was given initially, followed by an infusion of 1 mg/kg per hour over 2 hours. In our study t-PA for both intra-arterial and systemic infusion was given in a square-wave infusion, with 20% given as a loading dose and the remaining drug given over 30 minutes for a total dose of 3 mg/kg. This square-wave infusion over 30 minutes was significantly more effective than a longer continuous infusion. 3 Although in both studies approximately 3 mg/kg was the total dose, our findings suggest that the method of administering the thrombolytic agent can have important implications for the success of thromborysis. Several other differences between the two studies could potentially be of importance. They include differences in thrombus size, difference in age of the thrombus, and difference in the technique of injecting the emboli. Finally, in both studies an injection of t-PA through the external carotid artery was the technique used to give "intra-arterial" t-PA. Although this method delivers the drug reasonably close to the emboli, it is different than the clinical method currently employed, whereby a selective catheter is actually directed into the clot itself. 4 -3 In addition to delivering the drug at the site of the thrombus, the catheter could also produce direct mechanical disruption of a clot. This may provide additional therapeutic benefit in terms of rapid clot lysis. Because neither of these experimental studies evaluate this technique, caution must be exercised in extrapolating the results of these animal studies to clinical use. References 1. Hamilton MG, Lee JS, Cummings PJ, Zabramski JM. A comparison of intra-arterial and intravenous tissue-type plasminogen activator on autologous arterial emboli in the cerebral circulation of rabbits. Stroke. 1994^25:651-655. 2. Russell D, Madden KP, Clark WM, Zivin JA. Tissue plasminogen activator cerebrovascular thromborysis in rabbits is dependent on the rate and route of administration. Stroke. 1992;23:388-393. 3. Lin TN, He YY, Wu G, Khan M, Hsu CY. Effect of brain edema on infarct volume in a focal cerebral ischemia model in rats. Stroke. 1993;24:117-121. 4. Zeumer H, Freitag H-J, Knospe V. Intravascular thromborysis in central nervous system cerebrovascular disease. Neuroimaging Clinics of N Am. 1992;2:359-369. 5. Barnwell SL, Clark WM, Nguyen TT, O'Neill OR, Wynn ML, Coull BM. Safety and efficacy of delayed intra-arterial urokinase therapy with mechanical clot disruption for thromboembolic stroke. AJNR. In press. Response We thank Drs Clark, Russell, Madden, and Zivin for their comments and observations. We have reviewed their article 1 and would like to respond to their concerns. In the study by Russell et al, 1 venous clot aged for 2 hours was used to embolize the internal carotid artery to simulate embolic stroke in a rabbit model. Treatment with t-PA was initiated 5 minutes after embolization by intra-arterial or intravenous routes with either 10 mg/kg or 3 mg/kg t-PA and results were compared with those of a control group receiving saline. A Doppler flow probe was used to document vessel recanalization and define the efficacy of thrombolytic therapy in their report. We disagree with the use of vessel recanalization as a definitive end point for defining "success" in the treatment of stroke. Although we agree that vessel recanalization is the principle aim and mechanism of action for treatment with t-PA, we do not feel that it can be used alone to define success. Russell et al reported that 100% recanalization was demonstrated by Doppler ultrasound in the rabbits receiving a 3 mg/kg intra-arterial t-PA infusion, compared with 67% recanalization in the animals undergoing a 3 mg/kg intravenous infusion, yet there was no difference in the incidence of infarction: infarction was documented in two of the six animals in by guest on July 22, 2018 http://stroke.ahajournals.org/ Downloaded from