[Skip to Content]
[Skip to Content Landing]
Citations 0
April 26, 2000

Intra-arterial Prourokinase for Acute Ischemic Stroke—Reply

Author Affiliations

Phil B.FontanarosaMD, Deputy EditorIndividualAuthorStephen J.LurieMD, PhD, Contributing EditorIndividualAuthor

JAMA. 2000;283(16):2102-2104. doi:10.1001/jama.283.16.2101

In Reply: We agree with Dr Akins that the only therapy approved by the FDA for acute ischemic stroke is IV tPA given within 3 hours of onset. However, PROACT II began in February 1996, 4 months before FDA approval of IV tPA. Following FDA approval, the PROACT investigators were ethically obliged to make eligible patients aware of IV tPA as part of the informed consent process. The PROACT investigators were also specifically advised they could treat eligible patients with IV tPA and, in fact, 180 patients were excluded from PROACT II on this basis.

In response to Drs Schneweis and Grond, it is not clear whether the advantages of intra-arterial thrombolysis (ie, lesion documentation, superior recanalization) outweigh the disadvantages (ie, angiography, time) relative to IV thrombolysis in patients with large vessel occlusions. It is possible that IV and intra-arterial thrombolysis can be combined in selected patients.1 New thrombolytic agents may someday produce better IV recanalization rates than tPA.2 As we reported, less than 2% of patients who underwent angiography experienced major procedural complications. We did not systematically monitor therapy in patients excluded by angiography. The number of angiograms could be reduced by noninvasive testing. In addition to vascular analysis, the choice of best stroke therapy may soon be guided by physiological analysis of tissue salvageability using techniques like xenon computed tomography and functional magnetic resonance imaging.3 As new therapies and technologies for acute stroke emerge, the health care system will have to adapt to improve both patient access and the dismal 2% rate of IV tPA use.4

In response to Dr Tirschwell, there were sound statistical reasons that the PROACT II protocol prespecified a Mantel-Haenszel analysis of the primary end point stratified for initial National Institutes of Health Stroke Scale score. The Mantel-Haenszel approach is generally thought of as a test of the common OR over strata.5 There are many other possible analyses, including so-called precision-based techniques, that combine treatment effects via the log (RR) or risk difference. Although Tirschwell does not cite the specific method he used (the term Mantel-Haenszel adjusted is nonspecific, especially in the context of RR), we presume it is a test of the weighted log RR with weights defined as the inverse approximate variance. Although this test could have been used, PROACT II prespecified the Mantel-Haenszel test, and thus stratum-specific treatment effects were expressed as ORs. Nonetheless, we quoted a 58% relative benefit, which corresponds to the RR of 1.58 that Tirschwell calculated. The OR is a valid measure of treatment effect, and is simply a different way to quantify treatment effect than risk difference and RR. It is simplistic to claim that ORs are only estimates of RR and to decree that they should only be used when a true RR cannot be calculated. It is also inappropriate to imply that by representing the treatment effect as an OR we are overestimating the size of the effect. We believe that Tirschwell's reanalysis does not fundamentally change the strength of our evidence.

Lewandowski  CAFrankel  MTomsick  TA  et al.  Combined intravenous and intra-arterial r-TPA versus intra-arterial therapy in acute ischemic stroke.  Stroke. 1999;30:2598-2605.Google Scholar
Cannon  CP Overcoming thrombolytic resistance: rationale and initial clinical experience combining thrombolytic therapy and glycoprotein I-Iib/IIIa receptor inhibition for acute myocardial infarction.  J Am Coll Cardiol. 1999;34:1395-1402.Google Scholar
Albers  GW Expanding the window for thrombolytic therapy in acute stroke: the potential role of acute MRI in patient selection.  Stroke. 1999;30:2230-2237.Google Scholar
Katzan  ILFurlan  AJLloyd  LE  et al.  Use of tissue-type plasminogen activator for acute ischemic stroke: the Cleveland area experience.  JAMA. 2000;283:1151-1158.Google Scholar
Fleis  JLIn: Statistical Methods for Rates and Proportions. 2nd ed. New York, NY: John Wiley & Sons Inc; 1981:160-176.