The role of leukocytes following cerebral ischemia: Pathogenic variable or bystander reaction to emerging infarct?

Data accumulated over the last 10 years have led to the popular hypothesis that neutrophils and other inflammatory cells play a prominent role in the neuropathology of cerebral ischemia. This hypothesis was derived from a large number of studies involving three general observations: (1) leukocytes, particularly neutrophils, are present in ischemic tissue at the approximate time that substantial neuronal death occurs; (2) neutropenia is sometimes associated with reduced ischemic damage; and (3) treatments that prevent leukocyte vascular adhesion and extravasation into the brain parenchyma can be neuroprotective. This review reexamines the literature to ascertain its support for a pathogenic role for neutrophils in ischemia-induced neuronal loss. To accomplish this goal, we employed several logical theorems of "cause- effect" relationships, as they pertain to leukocytes and ischemic brain damage. Since the majority of studies focused on neutrophils as the most likely pathogenic inflammatory cell, this review necessarily does so here. We reasoned that if neutrophils play an important pathogenic (i.e., cause-effect) role in the neuronal damage that follows a stroke, then one should expect to find clear evidence that: (1) neutrophils invade the ischemic area prior to terminal stage infarction, (2) greater numbers of early appearing neutrophils are accompanied by evidence of greater neuronal loss, and (3) dose-related inhibition of neutrophil trafficking or activity produces a corresponding decrease in the degree of brain damage following ischemia. This review of the literature reveals that the existing evidence does not readily support any of these predictions and that, therefore, it consistently falls short of establishing a clear cause- effect relationship between leukocyte recruitment and the pathogenesis of ischemia. While the available evidence does not necessarily rule out a potential pathogenic role of neutrophils and other leukocytes, it nevertheless does expose serious weaknesses in the existing studies intended to support that hypothesis. For this reason we also offer suggestions for additional experiments and the inclusion of control groups that, in the future, might provide more effective or conclusive tests of the hypothesis. (C) 2002 Elsevier Science.