You are in: eMedicine Specialties > Neurology > Neuro-vascular Diseases Reperfusion Injury in StrokeArticle Last Updated: Jan 17, 2007AUTHOR AND EDITOR INFORMATIONSection 1 of 10
  Author: Wayne M Clark, MD, Director of Oregon Stroke Center, Department of Neurology, Professor, Oregon Health Sciences University Wayne M Clark is a member of the following medical societies: Alpha Omega Alpha, American Academy of Neurology, American Medical Association, American Neurological Association, American Society of Neuroimaging, National Stroke Association, and Oregon Medical Association Editors: Richard M Zweifler, MD, Professor, Director of Stroke Center, Director of Neurosonology Lab, Director of Vascular Neurology Fellowship, Director of Medical Student Education, Department of Neurology, University of South Alabama; Francisco Talavera, PharmD, PhD, Senior Pharmacy Editor, eMedicine; Howard S Kirshner, MD, Professor of Neurology, Psychiatry and Hearing and Speech Sciences, Vice Chairman, Department of Neurology, Vanderbilt University School of Medicine; Director, Vanderbilt Stroke Center; Program Director, Stroke Service, Vanderbilt Stallworth Rehabilitation Hospital; Consulting Staff, Department of Neurology, Nashville Veterans Affairs Medical Center; Matthew J Baker, MD, Consulting Staff, Collier Neurologic Specialists, Naples Community Hospital; Helmi L Lutsep, MD, Associate Professor, Department of Neurology, Oregon Health and Science University; Associate Director, Oregon Stroke Center Author and Editor Disclosure Synonyms and related keywords: reperfusion of ischemic brain tissue, inflammatory response, blood flow, leukocyte adhesion, leukocyte infiltration, free radical release, neuronal membrane breakdown, reperfusion injury in stroke REPERFUSION INJURY IN STROKE: POTENTIAL MECHANISMS AND FUTURE THERAPYSection 2 of 10
  Prompt reperfusion of ischemic brain tissue is critical for restoring normal function. However, this return of blood flow can paradoxically produce a progressive destruction of reversibly damaged cells, thereby leading to tissue dysfunction and infarction. This "reperfusion injury" has a multifactorial etiology but appears to be strongly associated with an inflammatory response; with the return of blood flow, several inflammatory processes may occur to potentiate ischemic injury, including the following:
Much of this inflammatory response appears to be mediated by interleukins (ILs), a multifunctional subclass of cytokines. Leukocytes also appear to play a critical role in reperfusion injury. In addition to injuring endothelium and neurons, leukocytes can obstruct the microcirculation directly. This leukocyte capillary plugging also may be the major mechanism of the "no-reflow phenomenon." First observed by Ames et al in 1968, this phenomenon is defined as incomplete restoration of normal blood flow following a period of ischemia. Areas of parenchyma that are still viable when blood flow returns are not reperfused adequately and ultimately die. This article reviews the mechanisms of reperfusion injury in stroke, focusing on the 2 most important: leukocyte adhesion and the cytokine response. The current state of research in these areas is reviewed, and the status of future potential therapies is discussed. CYTOKINES: ROLE IN REPERFUSION INJURYSection 3 of 10
Much of the inflammatory response involved in central nervous system (CNS) reperfusion injury appears to be mediated by various cytokines. The pro-inflammatory cytokines (including IL-1, tumor necrosis factor alpha [TNFa] and IL-6) can influence the function and synthesis of other cytokines by a complex cytokine network. These pro-inflammatory ILs are produced by various cells (eg, microglial cells, astrocytes, leukocytes); they appear to modulate CNS cell apoptosis, differentiation, and proliferation and to influence subsequent infiltration by leukocytes. Increased levels of IL-1, TNFa, and IL-6 have been observed following experimental CNS ischemia. Indirect evidence for the involvement of ILs in ischemic injury comes from several clinical studies, which found that cerebrospinal fluid (CSF) and plasma levels of IL-6 predict the degree of functional recovery of the patient and correlate with infarct size. Evidence also suggests that cytokines may be key components in the activation and recruitment of leukocytes into the CNS. IL-1b, TNFa, and IL-6 have been shown to activate leukocytes and increase expression of adhesion receptors on the leukocytes (ie, CD-18) and on endothelial and astrocyte cells (ie, intercellular adhesion molecule 1 [ICAM-1]). This evidence suggests that the cytokine response is an important component of CNS ischemic injury. If the cytokine response could be modulated, the associated reperfusion injury may be decreased. Since the inflammatory response appears to be most important in CNS ischemic injury that is associated with reperfusion, discovery of a method of reducing CNS reperfusion injury would have widespread therapeutic potential. This potential for reducing reperfusion injury is especially critical because of the widespread use of thrombolytic agents such as intravenous tissue plasminogen activator (tPA), an approved stroke treatment, and increasing use of direct intraarterial thrombolysis. Cytokines in ischemic brain injury Cytokines are small proteins (molecular weights from 8-30,000 d) that possess multiple biologic activities and are active in low (picomolar) concentrations. Cytokines appear to be produced primarily in response to an external stimulus and may not have a role in normal homeostatic mechanisms. As a direct consequence of the ionic imbalances and free calcium accumulation that are associated with ischemic brain injury, free fatty acids and other pro-inflammatory lipid metabolites are released. These metabolites are believed to promote expression and release of a pro-inflammatory cytokine cascade. This cascade appears to involve the release of IL-1 and TNFa initially; these cytokines then lead to the production of other pro-inflammatory cytokines (eg, IL-6, IL-8), activation and infiltration of leukocytes, and production of anti-inflammatory cytokines (including IL-4 and IL-10, which may produce a negative feedback on the cascade).
IL-1 in CNS ischemia IL-1b, a cytokine with multiple pro-inflammatory actions, is produced in the CNS by endothelial cells, microglia, astrocytes, and neurons. IL-1b acts via specific brain receptors; a naturally occurring IL-1ra has been identified and purified. Several studies have investigated the role of IL-1b in CNS ischemia. These studies found that IL-1b mRNA and protein were induced in rat models of focal stroke; initial up-regulation occurred at 1 hour after ischemia, and a peak response at 6-12 hours after ischemia. This increased expression of IL-1b was seen in ischemic areas and was most pronounced in animals that had an associated reperfusion injury. Injection of IL-1b into the lateral ventricle during transient ischemia increased the infarct size, whereas injection of an anti-IL-1 monoclonal antibody produced a decrease in postischemic infarct size. Injection of IL-1ra directly into the brain also decreased focal experimental CNS ischemic injury, associated cerebral edema, and neutrophil infiltration into the ischemic areas. In a rat model of middle cerebral artery (MCA) occlusion, Relton et al (1996) found that peripherally administered IL-1ra significantly reduced infarct size and associated cerebral edema. This same group recently found that IL-1ra was protective even when administered up to 3 hours after MCA occlusion (Mulcahy et al, 2003). Finally, using the MCA occlusion model in mice, Ohtaki et al (2003) found that IL-1 knock out animals had smaller infarcts and decreased expression of nitric oxide synthase. TNFa in CNS ischemia TNFa has diverse pro-inflammatory functions, including stimulating acute-phase protein secretion and increasing leukocyte infiltration. TNFa is produced in the brain by macrophages, monocytes, and astrocytes. Since it appears to be important in demyelination and blood-brain barrier damage, most prior studies have investigated the role of TNFa in multiple sclerosis and CNS tumors. Studies in CNS ischemia have found that the time course of TNFa expression is similar to that of IL-1b. Using a rat MCA occlusion model, Wang et al (1994) detected TNFa mRNA levels at 1 hour after ischemia, with peak levels between 6-12 hours. Local injection of TNFa into the brain greatly increased leukocyte infiltration to this area, suggesting that TNFa may potentiate CNS reperfusion injury. Other studies found that TNFa promoted neuronal cell death in cultures. Although these studies suggest that TNFa is detrimental during CNS ischemia, Bruce et al (1996) reported a conflicting view. Using TNF receptor knockout mice (TNFR-KO), they found that animals without TNFa had larger infarcts (MCAO) and decreased neuronal survival compared to control animals. With the recent development of high-sensitivity immunoassay, serum TNF can now be measured in stroke patients. Intiso et al (2004) found that serum TNF increased after clinical stroke, peaking at 7 days, but it did not correlate with outcome. Other clinical studies have found that serum TNF levels correlated with the total area of hypoperfusion but not acute infarction size (Montaner et al, 2003). Finally, increased TNF has been found in stroke patients who also had a recent transient ischemic attack (TIA), suggesting that it is involved in ischemic tolerance (Castillo et al, 2003). Further studies are needed to determine the exact role of TNFa and other cytokines in CNS ischemia. IL-6 in experimental CNS ischemia IL-6, a 26-kd molecule, is a pleiotropic cytokine that is involved in regulating many aspects of the inflammatory response. It is secreted by a wide variety of cells, including endothelial cells, microglial cells, astrocytes, and leukocytes. IL-6 stimulates hepatocytes to produce several plasma proteins, including fibrinogen, that contribute to the acute phase response. The synthesis of IL-6 can be up-regulated by various agents, including the cytokines IL-1b and TNFa. IL-6 initiates its biological effects through binding to a high-affinity receptor complex of 2 membrane glycoproteins: an 80-kd membrane-bound receptor (IL-6R) that binds IL-6 with low affinity and a 130-kd component (gp130) that is required for high-affinity binding and for signal transduction. Both components have been cloned, sequenced, and expressed. Like IL-1b and TNFa, IL-6 expression seems to occur in response to several different forms of brain injury. Increased levels of IL-6 have been detected in CSF from patients suffering from viral meningitis, encephalitis, or multiple sclerosis; IL-6 mRNA also has been shown to be up-regulated following focal CNS ischemia. Other studies have suggested that IL-6 mRNA expression is highest in CNS reperfusion models. The role of IL-6 in CNS ischemia has been investigated, using a rat model of 2-vessel forebrain ischemia reperfusion to investigate the time course of immunohistologic changes for IL-6 and concurrent ICAM-1 expression. Low levels of IL-6 immunoreactivity were seen in sham surgery controls. IL-6 reactivity significantly increased in the CA1 pyramidal cell layer in the hippocampus at 24 hours after ischemia and had resolved by 4 days. These results suggested that following CNS ischemia, increased IL-6 production occurs predominantly in areas of neuronal cell loss. To quantify these changes more accurately, the author and his colleagues measured IL-6 mRNA levels in the brain using a mouse model of MCA filament occlusion. In this study, a focal reversible stroke model was used to investigate the time course and relationship to outcome of IL-6 production in plasma, brain, and CSF. Reversible MCA occlusion or sham surgery was produced in 50 adult Swiss Webster mice by advancing a 5-0 filament into the internal carotid artery for 2 hours (sham, 1 minute). At 3, 6, 12, 24, and 72 hours after ischemia, groups of animals (8 each of mice with induced ischemia; 2 each sham) were evaluated on a 28-point clinical scale, blood and CSF were obtained for measurement of IL-6 levels by enzyme-linked immunosorbent assay (ELISA), and the brains were evaluated for infarct volume and IL-6 mRNA levels. Results were as follows:
The overall R2 between plasma and CSF IL-6 was 0.20. The overall R2 between infarct size and serum IL-6 level was 0.13, and that between infarct size and CSF IL-6 level was 0.20. The overall R2 between neurologic function and serum IL-6 level was 0.10, and that between neurologic function and CSF IL-6 level was 0.17. These findings confirm that IL-6 values increase following CNS ischemia; serum and CSF levels peak before brain values. CSF IL-6 levels had a stronger correlation with neurologic function and infarct size than serum levels. Although increased IL-6 production appears to be associated with CNS ischemia, data are conflicting as to whether IL-6 may be exerting neuroprotective or neurotoxic actions. While IL-6 appears to promote neuronal survival and to inhibit N-methyl D-aspartate (NMDA) toxicity in vitro, higher concentrations of IL-6 have been shown to be toxic in neuronal cultures. Transgenic mice overexpressing IL-6 have shown early neurodegenerative changes. IL-6 also has a mitogenic effect on astrocytes and appears to contribute to the reactive gliosis that appears after ischemia (Selmaj, 1990). These results suggest that IL-6 is increased in CNS reperfusion injury, but whether decreasing the IL-6 response would decrease injury is unknown. To test the effects of IL-6 modulation on stroke, the author and his colleagues recently used a model of focal reversible stroke to investigate whether transgenic mice lacking IL-6 were protected against acute ischemic injury. We first developed a strain of transgenic IL-6 C57 black mice (I-129 IL-6 KO backcrossed with C57), including homozygous "knockouts" (kTg), heterozygous littermates (hTg), and normal litter mates (nTg); all genotypes were confirmed by DNA sampling and polymerase chain reaction (PCR) analysis. Reversible MCA occlusion was produced in these animals by advancing a silicone-coated 8-0 filament into the internal carotid artery for 2 hours (experiment 1) or 45 minutes (experiment 2). At 24 hours after this induced ischemia, animals were evaluated on a 28-point clinical scale, blood and CSF were obtained for IL-6 levels, and the brains were evaluated for infarct volume and IL-6 mRNA levels. The results were as follows:
IL-6 in clinical stroke Several clinical studies in patients with stroke have found an association between CSF and plasma IL-6 levels and infarct size and neurologic outcome. Some of these studies have yielded conflicting results. Fassbender et al (1994) found that plasma levels of IL-6 showed a significant increase within the first few hours following stroke onset, peaking at 10 hours. This increased level of IL-6 significantly correlated with the volume of the brain lesion and was associated with a poor functional neurologic outcome. In contrast, serum levels of IL-1b and TNFa did not increase. Different results were obtained by Tarkowski et al (1995). In this study, CSF levels of IL-6 were elevated significantly by day 2 and were correlated with the volume of infarct observed; serum levels of IL-6 did not correlate with the size of brain lesion. This study did not find any measurable levels of plasma IL-1b. Finally, Kim et al (1996) measured IL-6 levels in 29 patients with acute stroke. They found the level of IL-6 to be highest (49±16 pg/mL) at 24 hours after onset, remaining high (14±4 pg/mL) for at least 7 days. The role of cytokines in clinical stroke also has been investigated. Plasma levels in vivo of IL-6 and the naturally occurring IL-1ra were measured using ELISA in 50 patients with acute stroke (4±2 days after onset) and in 20 age-matched healthy controls. Levels of both IL-6 and IL-1ra were significantly higher in the stroke population (4.6±4.2 pg/mL and 354±270 pg/mL, respectively) than in controls (1.0±0.9 pg/mL and 139±113 pg/mL). The levels of both cytokines were significantly higher in patients who had infarcts measuring more than 3 cm on CT scans. We also found that the level of IL-6 is elevated not only in the acute stroke period, but it remains elevated for up to a year (compared to controls) upon longitudinal follow-up (Coull, 1993). The acute IL-6 elevation and its ability to predict neurologic recovery from stroke has been investigated. In this study, plasma levels of IL-6, fibrinogen, and albumin along with white blood cell (WBC) count were measured within 4±2 days of onset in 131 patients with stroke. Standard clinical predictors of outcome also were obtained, including neurologic assessments and head CT infarct sizes. Peak levels of IL-6 were correlated significantly (r=0.18) with 6-month outcome as assessed by the Glasgow Outcome Scale. Taken together, the acute-phase response variables strongly predicted 6-month recovery (R2=0.31) and were nearly as strong a predictor as the standard clinical predictors (R2=0.38). Recent clinical studies have confirmed these findings and found that IL-6 values strongly correlated with C reactive protein (Smith et al, 2004). Acalovschi et al (2003) recently confirmed the robust elevation of serum IL-6 post clinical stroke but found that the soluble receptors (antagonists) for IL-6 were down-regulated. These clinical results suggest that the acute-phase response, in particular IL-6, is strongly associated with acute ischemic stroke; the acute-phase response also appears to be correlated with initial infarct size and degree of long-term recovery. Cytokines - Conclusions Taken together, the aforementioned observations strongly suggest that the pro-inflammatory cytokines are involved in the potentiation of CNS ischemic injury; part of their effect probably involves influences on leukocyte adhesion. Experimental studies in focal CNS ischemic models show that IL-1, TNFa , and IL-6 increase early after onset of ischemia; inhibition of IL-1 has been shown to produce a therapeutic benefit. In clinical studies, IL-6 levels in CSF and plasma appear to correlate with both infarct size and degree of functional recovery. Further studies investigating the efficacy of modulating cytokines during CNS ischemia are needed. The "cytokine cascade" appears to be a fertile area for future neuroprotective investigation. LEUKOCYTES: ROLE IN REPERFUSION INJURYSection 4 of 10
As part of an inflammatory response, leukocytes appear to potentiate CNS reperfusion injury by clogging the microcirculation and infiltrating into the brain, where they release free radicals and other neurotoxic substances. The initial step in these processes is leukocyte-to-endothelial cell adhesion. Through the use of specific monoclonal antibodies directed against leukocyte adhesion receptors, both microcirculation obstruction and brain infiltration can be decreased. Experimental studies in animal models of stroke have found reduced stroke damage through the use of antibodies that bind to either the CD-18 leukocyte adhesion molecule or its endothelial cell ligand, ICAM-1. This section reviews the components of leukocyte-to-endothelial cell adhesion, preclinical studies of anti-inflammatory agents (ie, anti-leukocyte adhesion), and preclinical studies investigating combination therapy (ie, anti-adhesion plus thrombolytics). Leukocyte-to-endothelial cell adhesion The migration of leukocytes from the blood into the brain is a multistep process. First, primary leukocyte-endothelial interaction (ie, rolling) is mediated by selectins involving P-selectin and E-selectin on the surface of the endothelial cells, and L-selectin on leukocytes (Fassbender, 1995). Upon activation, firm adherence of leukocytes to the endothelial lining (ie, sticking) is mediated by a leukocyte membrane glycoprotein receptor complex (termed CD-18 or b2-integrin) and its endothelial ligand, ICAM-1. CD-18 is predominantly responsible for final endothelial adherence. This complex consists of 3 heterodimers; all 3 share an identical beta unit (also frequently called CD-18) and are distinguished from each other by distinct alpha subunits. The 3 alpha subunits are termed leukocyte function antigen (LFA-1 or CD-11a, present on all leukocytes), MAC-1 (CD-11b, present mostly on polymorphonuclear lymphocytes and monocytes), and P150 (CD-11c, present on neutrophils and monocytes). Corresponding counter-receptors for the CD-18 integrin complex are the ICAM family of adhesion molecules. While ICAM-1 is expressed widely on many cells and binds to LFA-1 and MAC-1, ICAM-2 is expressed only on endothelial cells and leukocytes and is recognized by LFA-1 only (Smith, 1989). Unlike ICAM-2, which is expressed constitutively, ICAM-1 is inducible with inflammatory cytokines such as IL-1 and TNFa. CD-18/ICAM-1–mediated neutrophil adhesion has been demonstrated to increase after clinical and experimental stroke. Leukocyte appearance in CNS ischemic tissue previously has been considered to represent a pathophysiological response to existing injury. Recent evidence suggests that leukocytes also may be involved directly in the pathogenesis and extension of CNS reperfusion injury. Two proposed mechanisms of leukocyte involvement in reperfusion injury are direct microvascular occlusion after endothelial and basement membrane adhesion and transendothelial migration with secondary CNS tissue infiltration and neuronal cytotoxic injury. Initial adhesion of leukocytes to microvascular endothelium is essential for initiation of either of these mechanisms (see Picture 1A). One of the strongest areas of support for the involvement of leukocytes in CNS reperfusion injury comes from therapeutic preclinical studies, which reduced CNS injury by preventing leukocyte adhesion and infiltration. Preclinical studies of anti-adhesion therapyLeukocyte depletion The role of leukocyte adhesion and infiltration in potentiating CNS reperfusion injury is supported by older experimental studies. Systemic leukocyte depletion was produced in experimental animals by administering antiserum, cyclophosphamide, or mechlorethamine; the animals then were subjected to CNS ischemia. These studies showed that leukocyte depletion improved blood flow in an embolic stroke model, reduced infarct size in a model of thromboembolism, and improved blood flow in a forebrain ischemia model. Although these studies support leukocyte involvement in stroke, the use of toxic leukocyte depletion treatments cannot be used in clinical stroke therapy. Instead of depleting leukocytes, an alternative approach is to block their adhesion to the endothelium through the use of specific anti-adhesion monoclonal antibodies (see Picture 1B). By blocking leukocyte-to-endothelial adhesion, both capillary plugging and tissue infiltration may be prevented. CD-18 Using specific mouse monoclonal antibodies directed at various components of the CD-18 complex, previous studies have shown a reduction of ischemic injury and decreased leukocyte tissue infiltration in heart, lung, and intestinal ischemia models. Clark et al (1991) were the first to test the therapeutic potential of CD-18 antibodies in two CNS ischemia models in rabbits, using a mouse monoclonal antibody directed against the CD-18 beta complex.
Treatment with monoclonal antibodies against the CD-18 complex also has been found beneficial in a rat model of MCA occlusion. Two hours after the rats' MCAs were occluded with a filament, Chen et al (1994) administered an antibody directed against the MAC-1 component (anti-CD-11b) upon reperfusion; the filament then was removed to permit reperfusion.
Matsuo et al (1994) also found that monoclonal antibodies against the CD-18 complex were effective in the rat model of MCA occlusion. In this study, the investigators actually used 3 separate antibodies: the first was directed against the CD-11a component (LFA-1), the second against the beta unit (CD-18), and the third against ICAM-1. Compared to the responses in animals that underwent sham operation, each of these of antibodies produced significant reductions in infarct size and leukocyte infiltration at 24 hours.
Lindsberg et al (1995) also found that anti-adhesion therapy was useful after reperfusion in a rabbit model of spinal cord injury. A monoclonal antibody directed against the beta unit (CD-18) was given (1 mg/kg) after 30 minutes of reperfusion (ie, 1 hour after onset of ischemia). The animals that received the anti-CD-18 therapy showed significant improvements in spinal cord blood flow, neurologic function, and preservation of the blood-brain barrier (as measured by extravasation of Evans blue albumin); the number of surviving motor neurons at 6 hours after the insult also was improved. Takeshima et al in 1992 published the only negative study about anti-adhesion agents. Using a MCA reperfusion model in cats, 2 mg/kg of an anti-CD-18 (common beta unit) was given to the cats 45 minutes after the onset of MCA ischemia. The animals underwent a total of 90 minutes of ischemia followed by 180 minutes of reperfusion and were then killed. At 180 minutes, neither blood flow to the 2 hemispheres nor infarct volume as measured by triphenyltetrazolium chloride staining were different in experimental and control animals. One of the major differences in this study compared to the prior work is in the timing of the evaluation; in the study of Takeshima et al, the animals were evaluated after 3 hours of reperfusion. Much of the histologically detectable injury that occurs from white cell influx is likely to happen in a more delayed fashion, ie, more than 3 hours after beginning of reperfusion. This may explain the failure to find a benefit at 3 hours. Also, the triphenyltetrazolium chloride staining may be an inexact way of determining infarct size, particularly when this is attempted at 3 hours after ischemia. A study in transgenic mice has confirmed the importance of leukocyte adhesion mediated by CD-18 in reperfusion injury. Mice lacking CD-11b/CD-18 (MAC-1) were found to have a 25% reduction in infarct size and a 50% reduction in leukocyte infiltration (Soriano, 1999). The studies described here suggest that antibodies directed against the common beta unit (anti-CD-18 antibodies) and those directed against the MAC-1 component (CD-11b) are effective at reducing CNS reperfusion injury. To our knowledge, antibodies directed against P150 have not been tested in CNS ischemia. ICAM-1 Previous studies using anti-ICAM-1 monoclonal antibody treatments in animal models have shown reductions in inflammation, asthma, kidney rejection, and lung injury. On the basis of the encouraging results from these studies, in addition to the apparent success of anti-CD-18 therapy in reducing CNS ischemia, Clark et al (1991) tested the ability of the monoclonal antibody directed against ICAM-1 to reduce experimental CNS ischemic injury. In this study, 1 mg/kg of anti-ICAM-1 was given to rabbits 30 minutes before induction of either reversible spinal cord ischemia or irreversible multiple brain emboli. Similar to their anti-CD-18 results, the investigators found that pretreatment with anti-ICAM-1 produced a marked improvement in neurologic function in the model of spinal cord reperfusion CNS injury, whereas no benefit was seen with anti-ICAM-1 treatment in the nonreperfusion cerebral microsphere emboli model. They also found qualitative reductions in spinal cord infiltration in animals treated with anti-ICAM-1. They then tested these results in a model of multiple cerebral embolism with reperfusion. In this study, animals were treated with either saline or 1 mg/kg of anti-ICAM-1 5 minutes after being injected with multiple blood clot cerebral emboli. The difference in this model compared to the nondissolvable microspheres is that autologous blood clots may undergo spontaneous lysis, which allows reperfusion. In the multiple clot model with reperfusion, treatment with anti-ICAM-1 significantly increased the number of emboli that were required to produce neurologic impairment at 18 hours. Taken together, these studies indicate that treatment with anti-ICAM-1 is effective in CNS ischemia, but the beneficial action appears only when the extent of reperfusion involved is significant. The finding that anti-ICAM-1 treatment is beneficial in a reperfusion model but not in a nonreperfusion model is supported by the findings of Zhang et al (1994). In this study, the filament model of MCA occlusion was used.
In addition to reducing neurologic impairment, treatment with anti-ICAM-1 also appears to significantly reduce leukocyte infiltration. Zhang et al used 1 mg/kg of anti-ICAM-1 in the rat model of MCA occlusion.
Histologic studies have shown that ICAM-1 is expressed minimally on endothelial cells in healthy brain. However, it is expressed strongly in areas of focal brain ischemia, beginning as early as 3 hours after vessel occlusion. A potential advantage of this finding is that treatment with monoclonal antibodies directed against ICAM-1 interferes only with leukocyte adhesion at selective areas, where brain ischemia is present. Such antibodies may therefore be less likely to interfere with other leukocyte functions, particularly the ability to fight infections. An increased risk of infection is one of the most concerning potential side effects of anti-adhesion therapy. By interfering with ICAM-1 interactions with CD-18, a condition similar to CD-18 deficiency (LAD) may result; such patients exhibit multiple, frequent infections with little ability to produce an appropriate leukocyte infiltration. To test whether anti-adhesion therapies would increase the risk of infection, a recent study injected animals subcutaneously with doses of Staphylococcus aureus in conjunction with anti-CD-18 therapy. These animals were found to have much larger abscesses than control animals. In contrast, animals treated with anti-ICAM-1 antibody in conjunction with S aureus did not have increased abscess size. A possible explanation of these results is that, besides being an adhesion molecule, CD-18 is also a complement receptor. Clinical update - Anti-adhesion therapyAt the 1997 meeting of the American Academy of Neurology, the author and colleagues presented results of a study of 600 patients with acute stroke that used antibodies against ICAM-1 (Enlimomab Acute Stroke Trial Investigators). A total of 625 patients from the United States and Europe were enrolled, and primary endpoints consisted of functional outcome at 3 months after ischemic event. Unfortunately, patients treated with enlimomab were significantly less likely to have a good functional outcome than those who received placebo. Patients treated with enlimomab also had a higher mortality rate than those treated with placebo (22% vs. 16%) and had a higher incidence of fever (51% vs. 27%) and total infections (55% vs. 42%). Why this clinical trial produced an adverse treatment effect with anti-ICAM therapy in contrast to the multiple animal studies that yielded beneficial effects is unknown. Possible reasons are being investigated. One possibility is that the mouse monoclonal antibody used in this trial could have produced adverse effects. This foreign antibody may actually have increased the inflammatory response, which counteracted the desired leukocyte-blocking effects. Part of the adverse treatment effect also may have been due to the increased incidence of fever in treated patients. In this trial, fever was significantly associated with a poor outcome; since a higher percentage of patients treated with enlimomab had fevers, this could have been a mechanism for some (although not all) of the observed adverse effects. The results of this clinical trial suggest that the leukocyte adhesion system is not as simple as hoped. A phase III trial was then performed with a human antileukocyte antibody, Hu23F2G, developed by ICOS Corporation. Because this agent is a humanized and not a murine antibody, it hopefully avoids the unwanted effects of enlimomab. In the trial, this agent did not appear to produce the immune response seen with enlimomab. However, no clinical benefit was seen with Hu23F26 on any of the planned measures. No further studies with this agent are planned, though smaller antibodies that target specific leukocyte receptors may still have potential. A clinical trial was also recently completed with UK-279, 276, a neutrophil inhibiting factor. Although no side effects occurred, no treatment effects were seen (Krams et al, 2003). Additional anti-adhesion therapies Although these studies suggest that antiadhesion monoclonal antibody therapy may play a role in future stroke treatment, these agents are currently not available outside of ongoing clinical trials. Leukocyte-depleting strategies are not appropriate to the clinical population, and general strategies aimed at suppressing immune function either have not shown any benefit to stroke patients (eg, steroid treatment) or could increase the risk of secondary infection. One currently available treatment that appears to reduce leukocyte adhesion is the antibiotic doxycycline. Doxycycline, a member of the tetracycline family of antibiotics, interferes with leukocyte function by binding divalent cations, which are required for the CD-18 receptor complex to function. In vitro studies have shown that doxycycline inhibits leukocyte adhesion. Clark et al (1994) found doxycycline to be beneficial in a CNS reperfusion injury model. Using a dose of 10 mg/kg of doxycycline (equivalent to 100 mg bid clinical dose), they found that treatment 30 minutes before reversible spinal cord ischemia produced a significant reduction in neurologic impairment at 18 hours. Although doxycycline also may have clinical efficacy in stroke, it has not been studied in a randomized clinical trial. Although the majority of leukocyte-to-endothelial adhesion is mediated by the CD-18 to anti-ICAM-1 receptor interaction, additional leukocyte adhesion is mediated by a family of adhesion receptors, the selectins. Although circulating forms of selectins (eg, sL-selectin, sE-selectin) have been observed after stroke, anti-adhesion monoclonal antibodies directed against selectins have not, to the author's knowledge, been evaluated for potential of reducing CNS ischemic injury. INTERACTION OF CYTOKINES AND LEUKOCYTESSection 5 of 10
One of the primary mechanisms whereby pro-inflammatory cytokines may potentiate CNS ischemic injury is through their direct effects on leukocyte adhesion and infiltration. IL-6, IL-1b, and TNFa all have been shown to increase expression of ICAM-1 on endothelial cells and astrocytes, to facilitate leukocyte infiltration, and to increase leukocyte activation. A recent study found that induction of ICAM-1 by IL-6 was prevented when the animals were pretreated with a neutralizing antibody to IL-6; the induction of ICAM-1 by IL-6 also appears to occur predominantly with reperfusion. In addition to the cytokines increasing leukocyte infiltration, the cytokines and leukocytes also are linked; activated monocytes also appear to be a major source of cytokine production. In vitro studies have found that activated mononuclear cells produce massive quantities of IL-1b, IL-6, and TNFa, while neutrophils produce relatively small quantities of these cytokines. Given that a mononuclear cell infiltrate can be seen within 6-12 hours following CNS focal ischemia, some of the cytokine production in the brain at 12 hours is likely to be secondary to infiltrating monocytes. To date, no reported studies have investigated what effects blocking leukocyte adhesion and infiltration would have on subsequent IL-6 production in the CNS. REPERFUSION THERAPY COMBINED WITH THROMBOLYSISSection 6 of 10
Overall, the experimental animal studies already described support the beneficial effects of using antibodies directed against the CD-18 complex or ICAM-1 in reducing CNS reperfusion injury. In clinical stroke treatment, reperfusion injury may be of particular importance following the use of thrombolytic agents. Consequently, an ideal therapeutic approach would combine a thrombolytic agent and an agent that reduces reperfusion injury. To test this combined approach, Bowes et al studied the use of tPA and anti-CD-18 antibody in a rabbit model of multiple cerebral emboli. In this study, animals were injected with multiple autologous small blood clots. Five minutes after ischemic onset, animals received either saline or 1 mg/kg of anti-CD-18 (which binds to the beta unit); 30 minutes after embolization, the animals received either saline or 3 mg/kg of tPA; neurologic impairment was rated 18 hours later.
Zhang et al (2003) recently studied the effects of lek-279, 276, a neutrophil inhibiting factor, combined with TPA in a MCA clot model in rats. Although not effective on its own, UK-279, 276 extended the therapeutic time window from 2 hours to 4 hours. Leukocytes - Conclusions These studies support the deleterious role of leukocytes in CNS ischemia, demonstrating that monoclonal antibodies directed against either CD-18 or ICAM-1 might reduce CNS ischemic injury. These agents appear to be most effective in CNS reperfusion models and appear to have beneficial effects even when administered as long as 3 hours after the onset of CNS ischemia. These agents may be of particular benefit in clinical situations that are associated with reperfusion (ie, following thrombolysis); some experimental data exist to support this synergistic effect. Future stroke trials combining these agents with TPA are needed before these agents are proven to be clinically useful. SUMMARYSection 7 of 10
Studies indicate that CNS reperfusion injury is involved directly in the potentiation of stroke damage. Components of the inflammatory response, including cytokine release and leukocyte adhesion, appear to play key roles in these deleterious effects. Animal studies using various methods of modulating the cytokine response have found beneficial effects for IL-1 and TNF. Various experimental studies utilizing agents that block leukocyte endothelial adhesion (ie, monoclonal antibodies that block either the adhesion receptor on leukocytes [CD-18] or the corresponding adhesion receptor on the endothelial cell [ICAM-1]) have found beneficial effects in terms of reducing infarct size and improving functional outcome. In general, these experimental studies have found benefit when a period of ischemia is followed by a return of blood flow (reperfusion) but not in studies in which ischemia is permanent. For this reason, these anti-adhesion therapies may prove to be most beneficial clinically when given in association with thrombolytic agents. Although clinical studies using antibodies against ICAM-1 have failed to find a clinical benefit, further investigations of antiadhesion therapies with TPA are ongoing. Given the strong preclinical support for the usefulness of anti–reperfusion injury agents, they are likely to be employed in future "stroke cocktail" therapeutic efforts. ACKNOWLEDGMENTSSection 8 of 10
The authors wish to thank Anne Tillinghast for her assistance in preparing this manuscript. MULTIMEDIASection 9 of 10
REFERENCESSection 10 of 10
| |||||||||||||||||||||||||||||||||||
