An Overview of Acute Stroke Therapy Past, Present, and Future

REVIEW ARTICLE An Overview of Acute Stroke Therapy Past, Present, and Future

Marc Fisher, MD; Wolf Schaebitz, MD

he effort to develop effective therapies for acute ischemic stroke achieved several important successes during the past decade, but also many disappointing failures. The 2 primary successes were related to thrombolysis. The first was the NINDS rt-PA (Na- tional Institute of Neurological Disorders and Stroke Recombinant Tissue-Type Plas- Tminogen Activator) trial reported in 1995. This study demonstrated that initiation of intravenous (IV) rt-PA within 3 hours after the onset of acute ischemic stroke significantly improved outcome at 3 months.1 This study led to the approval of rt-PA initiated within 3 hours of stroke onset as the only currently available acute stroke therapy. The second major success was the demonstration that intra-arterial prourokinase initiated within 6 hours of stroke onset in patients with angio- graphically documented proximal middle cerebral artery (MCA) occlusion also improved outcome at 3 months.2 A third marginally positive acute stroke trial used ancrod, a defibrinogenating agent derived from Malaysian pit vipers.3 Ancrod initiated within 3 hours after stroke onset also improved 3-month outcome but to a lesser degree than either rt-PA initiated within 3 hours or prourokinase initiated within 6 hours. These successful acute stroke therapy trials were out- weighed by a large number of neuroprotective trial failures. Currently, not one of many purported neuroprotective therapies assessed in pivotal clinical trials has demonstrated unequivocal, statis- tically significant improvement in clinical outcome.4 The neuroprotective trials all included patients who presented with a stroke 3 hours after onset, and the therapies used for each patient failed for myriad reasons that will be explored in detail.

In this overview of the current status and terventions, a treatment strategy likely to future direction of acute stroke therapy, we lead to maximal improvement in the great- will discuss in detail the current situation est number of stroke patients. of thrombolytic therapy for acute ischemic stroke, reviewing the results of pub- THROMBOLYTIC THERAPY FOR lished clinical trials, postmarketing expe- ACUTE ISCHEMIC STROKE rience with rt-PA given within the 3-hour window, and future directions of how to po- The NINDS rt-PA trial was the first acute tentially expand this window for IV throm- ischemic stroke trial to unequivocally dem- bolytic therapy. The status of various neu- onstrate that this disorder could be ben- roprotective therapies for acute ischemic efited by any therapeutic intervention.1 In stroke will be reviewed and potential new this trial, 624 carefully selected patients neuroprotective strategies previewed. were randomly and blindly assigned to Last, we attempt to envision likely ap- therapy with rt-PA (0.9 mg/kg) or pla- proaches toward multiple therapeutic in- cebo within 3 hours of stroke onset. Half of the patients were treated within 90 min- From the Departments of Neurology, University of Massachusetts Medical School, utes of onset, an accomplishment by the Worcester (Dr Fisher), and University of Heidelberg, Heidelberg, Germany investigators participating in the trial. The (Dr Schaebitz). Dr Fisher serves as a paid consultant to Bristol-Myers Squibb. patients treated with rt-PA had an abso-

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3196

Downloaded From: https://jamanetwork.com/ on 10/04/2021 Postmarketing Studies of Intravenous rt-PA Studies Compared With the NINDS Trial*

NINDS Trial

Modified Rankin Median Age, y, Median Time Score 0-1 at Symptomatic Study Patients, No. NIHSS Mean ± SD to Treat, min 90 Days, % Mortality, % ICH, % NINDS, placebo1 165 15 66 ± 13 90 26 21 1 NINDS, rt-PA (part 2)1 168 14 69 ± 12 90 39 17 7 Chiu et al7 30 14 ± 8 66 ± 15 157 30 23 7 Egan et al8 33 17 NA 157 36 18 9 Tanne et al9 75 NA 66 ± 15 NA 34 (At discharge) 11 3 Wang et al10 14 15.4 ± 2 (15-88) NA 57 7 7 Grond et al11 100 12 63 ± 11 124 40 12 5 Albers et al12 296 13 68 ± 13 165 35 (Day 30) 12 4

*rt-PA indicates recombinant tissue-type plasminogen activator; NINDS, National Institute of Neurological Disease and Stroke; NIHSS, National Institutes of Health Stroke Scale Score; ICH, intracerebral hemorrhage; and NA, not available.

lute improvement rate of 11% to 13% this time point may not be of proven study, the median baseline NIHSSs at 90 days when compared with the efficacy. were 12 and 13, while in the NINDS placebo patients on various out- Several postmarketing studies of trial the median baseline NIHSS in the come measures that evaluated both IV rt-PA are now available.7-15 Pa- placebo group was 14 in part 1 and neurologic and functional status. tients were included in these stud- 15 in part 2. In other acute stroke tri- The patients treated with rt-PA had ies using the general guidelines for als where the baseline NIHSS was 11, a symptomatic intracerebral hem- treatment used in the NINDS trial. the percentage of patients achieving orrhage rate of 6.4% (almost half the The most important inclusion crite- a Rankin score of 0-1 approximates patients died) within 36 hours of on- rion was initiation of therapy within 37%13 and when the baseline NIHSS set, while the rate was only 0.6% in 3 hours of stroke onset. Most of the was 13, 29% achieved this out- the placebo group. Despite this early studies encompassed relatively small come.14 Comparing the outcomes in hemorrhagic risk, the 90-day mor- numbers of patients, ranging from 14 the study by Grond and colleagues tality rate was 17% in the rt-PA to 75 (Table). However, several and the STARS study with those of a group and 21% in the placebo group. larger studies are available, includ- placebo group that had a similar de- Subsequent analysis of the study data ing the study reported by Grond and gree of baseline severity demon- demonstrated that early computed colleagues11 of 100 patients and the strates an absolute improvement rate tomographic (CT) demonstration of STARS (Standard Treatment With of 3% to 6%, not the approximately extensive edema or hypodensity, his- Activase to Reverse Stroke) study of 12% absolute rate of improvement tory of diabetes mellitus, and el- 296 patients.12 The median time from observed with rt-PA treatment in the evated baseline National Institutes stroke onset to initiation of rt-PA NINDS trial. The postmarketing stud- of Health Stroke Scale Score (NIHSS) therapy ranged from 124 minutes in ies do, however, provide some en- were predictors of poor outcome.5 the study by Grond and colleagues couraging data about the rate of The use of rt-PA was associated with to 165 minutes in the STARS study. symptomatic intracerebral hemor- improved outcome in all stroke sub- The percentage of patients achiev- rhage. The percentage of patients ex- types included in the study, in pa- ing a modified Rankin score of 0-1, periencing this serious complica- tients across the broad range of base- the results defined as a favorable out- tion of thrombolysis ranged from 0% line stroke severity, and in all age come in the NINDS trial, ranged to 19% with only 2 studies observ- groups. The initial analysis of the from 34% to 57%; although, in sev- ing double-digit rates of intracere- study data did not distinguish a dif- eral of the reports, day 90 data were bral hemorrhage. In the 2 largest ference in benefit of rt-PA related to not provided. studies, the intracerebral hemor- time-of-treatment initiation. How- On the surface, the rates of rhage rates were only 4% to 5%. It ever, in a subsequent analysis that favorable functional outcome dem- therefore appears that expanding IV adjusted for baseline severity of the onstrated in these postmarketing rt-PA use into general practice is not neurologic impairment, an earlier studies appear to be quite good, sur- associated with a substantially in- time to initiation of therapy was as- passing in some studies the 39% 0-1 creased risk of intracerebral hemor- sociated with a more favorable out- Rankin rate at 90 days seen in the rhage, if the guidelines for patient se- come, demonstrating an inverse lin- NINDS trial. These results must be lection used in the NINDS trial are ear relationship between time to treat interpreted cautiously because the followed. and the odds ratio of a favorable out- baseline severity of the patients Studies evaluating the efficacy of come.6 The confidence interval for treated in these postmarketing stud- IV rt-PA beyond the 3-hour time win- a favorable outcome crossed 1 in pa- ies were not as severe as in the NINDS dow were conducted. The first clini- tients treated beyond 2 hours 40 trial. For example, in the 2 largest cal trial to evaluate IV rt-PA up to 6 minutes after stroke onset, suggest- postmarketing studies, the study by hours after stroke onset was the Eu- ing that treatment initiated beyond Grond and colleagues and the STARS ropean Cooperative Acute Stroke

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3197

Downloaded From: https://jamanetwork.com/ on 10/04/2021 Study (ECASS-1).14 Patients were ran- One other large IV rt-PA study, the fully selected patients and should ini- domly and blindly assigned to rt-PA Alteplase Thrombolysis for Acute tiate additional attempts to success- (1.1 mg/kg) or placebo within the Non-Interventional Therapy in Is- fully expand the time window for IV 6-hour period after acute stroke on- chemic Stroke (ATLANTIS), evalu- thrombolysis in acute ischemic set in the MCA territory. Predefined ating therapy initiated from 3 to 5 stroke. exclusion criteria included evi- hours after stroke onset is available. Currently, with the restrictive dence of CT hypodensity or sulcal ef- In this study, the patients received 0.9 3-hour time window necessary for the facement involving more than a third mg/kg of IV rt-PA and the primary use of IV rt-PA in stroke therapy only of the MCA territory on the pretreat- outcome measure was the percent- approximately 1% to 2% of acute ment CT scan. The overall results as age of patients achieving a NIHSS of stroke patients are estimated to re- analyzed in the intention-to-treat 0 or 1 at 90 days.15 The study in- ceive this intervention.16 While pa- analysis of the trial were negative, but cluded 547 patients and the pri- tient educational efforts and institu- when protocol violators were ex- mary end point was almost identical tional initiatives might increase the cluded, several outcome measures in the 2 groups. The median base- percentage of patients treated to per- were better in the rt-PA group. The line NIHSS was 11 in the 2 groups haps 5% to 10%, expanding the time CT exclusion criterion (more than a and a day 90 modified Rankin of 0-1 window for successful IV thromboly- third of the MCA territory showing was achieved in 42% of the rt-PA sis should have a much greater im- early infarct signs) occurred in 63 group and 40% of the placebo pa- pact on the number of patients (58%) of 109 protocol violators and tients. One positive result from the treated. How might the therapeutic these patients had a high risk of ECASS-2 and ATLANTIS trials was time window for beneficial IV throm- symptomatic often-fatal intracere- that the rate of symptomatic intrace- bolysis be expanded in acute ische- bral hemorrhages when they re- rebral hemorrhage was 8.8% and mic stroke? There are at least 2 strat- ceived rt-PA. A second ECASS study 7.0%, respectively, not greatly in- egies possible that may be synergistic. was performed using the NINDS dose creased from the 6.4% rate seen with The first strategy is better identifica- of rt-PA, 0.9 mg/kg, and, in this study, rt-PA in the 3-hour window NINDS tion of patients likely to respond to the study investigators received bet- trial. treatment beyond 3 hours after stroke ter CT training to identify hyper- The Intra-arterial PROACT-2 onset. In PROACT-2, the use of an- acute CT changes indicative of early (Recombinant Prourokinase in Acute giography helped to identify pa- infarction.13 Patients were again ran- Cerebral Thromboembolism) study tients and likely led in part to the suc- domized up to 6 hours after stroke was a 0- to 6-hour thrombolysis trial cess of the trial.2 Unfortunately, onset. In ECASS-2, the primary end that demonstrated a significant treat- angiography is time-consuming and point was the percentage of patients ment effect. This study used re- not readily available at many insti- in the 2 treatment groups achieving combinant prourokinase (r-proUK) tutions throughout the day. In addi- a Rankin score of 0-1, and this out- delivered locally into an angiographi- tion, angiography only provides in- come was observed in 40.3% of the cally documented proximal MCA formation about the presence or rt-PA–treated group and 36.6% of the thrombus with low-dose IV hepa- absence of a vascular occlusion and placebo group, a nonsignificant dif- rin.2 In both the active treatment and does not provide information about ference. The median baseline NIHSSs placebo groups, the median time to the status of ischemic injury within were only 11 in the 2 groups. There- treat in PROACT-2 was 5.3 hours and the brain parenchyma. fore, the baseline severity of the stroke the median baseline NIHSS was 17. The new magnetic resonance patients included in ECASS-2 was less The trial included 180 patients ran- imaging (MRI) techniques of diffu- than in the first ECASS study or the domized 2:1 to r-proUK or placebo. sion and perfusion MRI combined NINDS study, likely explaining in At day 90, 40% of the patients treated with magnetic resonance angiogra- part the better outcome observed in with r-proUK achieved the primary phy can provide a wealth of infor- the placebo group of this trial than outcome measure of a Rankin score mation about the extent and loca- in the other 2 IV thrombolysis tri- of 0 to 2, while only 25% of the pla- tion of ischemic injury and the status als. A post hoc analysis of the cebo patients achieved this favor- of perfusion in the microvascula- ECASS-2 data demonstrated a sig- able outcome (P=.04). Secondary ture, and document the presence or nificant difference between the rt- outcome measures also tended to be absence of an occlusion in the ma- PA–treated group and the placebo better in the r-proUK group. Symp- jor intracerebral vessels.17 These MRI group when the Rankin score was di- tomatic intracerebral hemorrhage techniques can be performed in the chotomized into 0 to 2 and greater within 24 hours occurred in 10.2% same approximate time required to than 2. In this analysis, 54.3% of the of the r-proUK group and 1.9% of obtain a CT scan and are becoming rt-PA patients achieved a 90-day out- control patients. Despite the early risk widely available. Preliminary evi- come of 0 to 2, while only 46% of pla- of symptomatic intracerebral hem- dence suggests that acute stroke pa- cebo patients had a 90-day Rankin orrhage, the 90-day mortality was al- tients who have perfusion lesion vol- score of 0 to 2. The difference be- most identical in the 2 groups, 25% umes larger than diffusion lesion tween a score of 1 or 2 on the Rankin in the r-proUK group and 27% in volumes, so-called diffusion- scale is not great and this result controls. The PROACT-2 study dem- perfusion mismatch have ischemic re- speaks to the inherent difficulties in onstrates that thrombolytic therapy gions within the mismatch that are determining the best outcome mea- can be effective when initiated up to more likely to respond to therapeu- sure to use in acute stroke studies. 6 hours after stroke onset in care- tic interventions such as thromboly-

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3198

Downloaded From: https://jamanetwork.com/ on 10/04/2021 sis.18 Several studies demonstrated lysis has been raised in the past with- stroke patients may benefit from the that approximately 70% of acute out direct confirmatory evidence. neuroprotective approach. It is hoped stroke patients evaluated within 6 Recently, both animal and hu- that the first neuroprotectant will hours of stroke onset demonstrate man MRI studies demonstrated that be proven to be effective in the near this pattern of perfusion-diffusion secondary injury after successful future. mismatch.19,20 Follow-up studies in reperfusion does indeed occur, al- Compared with thrombolysis, patients with a diffusion-perfusion though the precise mechanisms re- the neuroprotective approach for mismatch document that the diffu- sponsible remain to be elucidat- stroke treatment is more complex and sion volume expands into the re- ed.25,26 The most likely processes reflects the diversity of the ischemic gion of the perfusion abnormality in responsible for secondary injury af- cascade (Figure 1). Neuroprotec- most patients who are left un- ter successful reperfusion are the re- tive agents have been developed and treated. In 1 preliminary study, pa- cruitment and activation of inflam- tested for nearly all components of the tients with a diffusion-perfusion matory white blood cells (primarily ischemic cascade. As seen with re- mismatch who were successfully re- polymorphonuclear leukocytes), the cently discovered mechanisms such canalized with IV rt-PA had a much release of cytokines, the generation as gene expression or the role of zinc better clinical outcome than pa- of oxygen free radicals and apopto- after stroke, new agents may be de- tients who did not reperfuse.18 Fur- sis. It is quite possible that more than veloped and new therapeutic op- ther studies are needed to docu- one of these potential inducers of tions occur. The discovery that cal- ment if patients with a diffusion- secondary injury are active within cium-induced excitotoxicity occurs perfusion mismatch observed 3 to 6 different regions of the reperfused tis- after ischemia is relatively old (about hours after stroke onset are good can- sue. Prior animal studies demon- 20 years) and was widely accepted as didates for IV rt-PA. strated that drugs inhibiting poly- a key event after cerebral ischemia. Is- Concerns arose that MRI may morphonuclear leukocyte activity or chemia-induced energy failure causes not accurately detect acute intrace- free radical scavengers are almost membrane depolarization and re- rebral hemorrhage. However, sev- exclusively beneficial in reperfu- lease of excitatory amino acids such eral recent reports document that sion models and demonstrate little if as glutamate into the extracellular susceptibility-weighted MRI stud- any effect in permanent occlusion space.28,29 Glutamate receptors be- ies can reliably demonstrate hem- models.4 Combining these types of come activated, resulting in calcium orrhages.21,22 It will therefore be drugs with thrombolysis in animals overload of neuronal cells. This step likely that CT scans will not be re- does indeed extend the therapeutic can effectively be blocked by N- quired when screening patients for time window for successful reperfu- methyl-D-aspartate (NMDA), and inclusion in MRI-based clinical tri- sion.27 Using these types of drugs af- ␣-amino-3-hydroxy-5-methyl-4- als with IV thrombolysis. ter documented reperfusion in- isoxazole propionic acid (AMPA) A second potential approach to duced by IV rt-PA could be another channel antagonists. Water, so- prolonging the therapeutic time win- way to extend the time for benefi- dium, and chloride move intracellu- dow for successful thrombolysis cial thrombolysis in stroke patients, larly via monovalent ion channels into would be to give neuroprotective presumably by inhibiting reperfu- the cell causing so-called cytotoxic therapy before, during, or after the sion injury that would impact on out- edema. The large amount of intracel- infusion of IV rt-PA. Neuroprotec- come measures determined several lular clacium activates proteolytic en- tive therapy initiated before or dur- months after stroke onset. zymes that degradate cytoskeletal and ing the use of IV rt-PA could ex- extracellular matrix proteins. Cal- tend the time that the ischemic PATHOPHYSIOLOGICAL cium also activates phospholipase A2 penumbra, the presumed therapeu- TARGETS FOR and cyclooxygenase producing oxy- tic target for both thrombolysis and NEUROPROTECTION gen free radicals. Nitric oxide (NO) neuroprotection, remains salvage- is synthesized from L-arginine and able.23 Animal studies confirm that Neuroprotective agents and strate- molecular oxygen by calcium- neuroprotective therapy can pro- gies have been studied for years and dependent NO synthase (NOS) and long the time window for success- appear to be effective in a variety of NO reacts with superoxide to form ful reperfusion.24 Currently, there is preclinical stroke models. However, peroxynitrite. The free radical re- no evidence that rt-PA and a neu- none of the drugs have proven con- lease promotes further membrane roprotective drug used in combina- clusively to be effective in humans. damage, and subsequently mitochon- tion act synergistically to extend the It is difficult to translate data regard- drial dysfunction. Free radical activ- time window for IV thrombolysis in ing drug dosage, time window, sex ity can be blocked by antioxidants and stroke patients, but such combina- differences, and, in particular, the free radical scavangers. tion trials are being considered. An- stroke target population from ani- Secondary to these events, ex- other potential therapeutic combi- mals to humans. Misunderstanding pression of proinflammatory genes is nation would be the use of IV rt-PA and misinterpretation of these is- induced by the synthesis of transcrip- followed by an agent designed to in- sues may have caused negative trial tion factors and release of mediators hibit reperfusion injury induced by results and a pessimistic view for neu- of inflammation, such as platelet- successful clot lysis. The possibil- roprotection of stroke in general. Evi- activating factor, tumor necrosis fac- ity of secondarily generated reper- dence from the recent clinical trials tor, and interleukin 1␤ (IL-1␤).30 fusion injury in the brain after clot demonstrates that subpopulations of Consequently, expression of the ad-

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3199

Downloaded From: https://jamanetwork.com/ on 10/04/2021 Embolic or Thrombotic Ischemia-Induced Vascular Occlusion Reduction of CBF and High-Energy Metabolism Failure Reduced CBF Lubeluzole

Failure of High-Energy Ion Dyshomeostasis Excitatory Inflammatory Metabolism Neurotransmitter Response: Release of Activation of Release TNF-α, PAF, and IL-1β Voltage-Gated Excitatory Ion Dyshomeostasis Ca++ Channels Apoptosis Neurotransmitter Na +, Ca ++, Cl –, K +, IP and Reverse Ca-Blocker Release i i i e 3 of Na+-Ca++ Ca++ Exchanger Growth Factors NMDA/AMPA- NMDA Anti-ICAM Anti-ICAM Ca++ Effusion Blocker Apoptosis Molecules Cytokine AMPA Channel Inhibitors Activation NMDA-Blocker Ca++ Infusion Activation of Cellular Enzymes: Proteases, Lipases, Kinases, Activate Cellular Enzymes: Calmodulin, and nNOS Microvascular Kinases, Calmodulin, nNOS, Free Radical Obstruction Proteases, Lipases Scavengers Statins

Free Radical Release Peroxynitrite Citicoline

Free Radical Cell Death Organelle Injury Formation Membrane Damage

Lipid Membrane Inflammatory Peroxidation Response Cell Death

Figure 1. The cascade of events currently thought to be important Figure 2. Diagram depicting where various neuroprotective agents may contributors to focal ischemic brain injury. CBF indicates cerebral blood flow; intervene on the ischemic cascade. TNF-␣ indicates tumor necrosis factor ␣; + ++ − Nai , intracellular sodium; Cai , intracellular calcium; Cli , intracellular PAF, platelet-activating factor; IL-1␤, interleukin 1␤; and ICAM, intercellular + chloride; Ke , extracellular potassium; IP3, inositol 1,4,5-triphosphate; NMDA, adhesion molecule. Other abbreviations are given in the legend to Figure 1. N-methyl-D-aspartate; AMPA, ␣-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid; and nNOS, neuronal nitric oxide synthase.

hesion molecules intercellular ad- Another mechanism that con- clude neuronal sprouting and hesion molecule-1 (ICAM-1), P- tributes to ischemic injury is apop- synaptogenesis, which are part of the selectins and E-selectins occur on the tosis, although the role of apoptotic spontaneous recovery process par- endothelial surface.31-33 After bind- injury in stroke patients remains un- ticularly after smaller focal lesions. ing to adhesion molecules, neutro- certain. Triggered by a number of Enhancement of recovery followed phils adhere to the endothelium, pathophysiological processes, includ- by improvement of behavioral out- cause microvascular obstruction, ing excitotoxicity, free radicals, the come can be achieved in animals cross the vascular wall, and enter the inflammatory reaction, and mito- with drugs such as growth factors brain parenchyma followed by mac- chondrial and DNA damage, apop- and amphetamines.46,47 rophages and monocytes. Blocking of tosis occurs after milder ischemic in- adhesion molecules can prevent these jury, particularly within the ischemic ACUTE TREATMENT events. Activated inflammatory cells penumbra.38 Apoptosis is mediated by STRATEGIES and injured neurons produce a num- a cascade of gene expression, includ- ber of toxic mediators that may ing the caspases, a family of aspartate- Various neuroprotective agents that worsen ischemia. For example, in- specific cysteine proteases, as well as may intervene on the ischemic cas- filtrating neutrophils produce toxic genes that suppress (eg, Bcl-2) or aug- cade are presented in Figure 2. amounts of NO through activation of ment (eg, Bax) cell death.39,40 Caspase inducible NOS (iNOS).34 Pharmaco- activity can be blocked by caspase in- Calcium Antagonists logical blockade of iNOS inhibitors hibitors, reducing infarct size and consequently reduces ischemic brain apoptosis as well as improving out- Calcium-channel antagonists were injury.35 Ischemic neurons also ex- come.41,42 Other agents that may among the first drugs evaluated for press cyclooxygenase 2, an enzyme counteract apoptosis after focal ce- neuroprotection after stroke. They re- that mediates ischemic injury by pro- rebral ischemia via Bcl-2/Bax– duce calcium influx into the cell via ducing superoxide and toxic prosta- dependent mechanisms are growth voltage-sensitive calcium channels. noids.36 Inhibition of the enzyme by factors such as brain-derived neuro- Calcium antagonists were indeed cyclooxygenase 2–blockers signifi- trophic growth factor.43 shown in several experimental stud- cantly reduces ischemic brain dam- Recovery and reorganization ies to be neuroprotective after focal age. Furthermore, injured neurons of the brain after focal ischemic cerebral ischemia.48 The best stud- produce the proinflammatory cyto- injury occurs over weeks and may ied drug is the dihydropyridine com- kine, IL-1␤, that can be blocked by have a major impact on the out- pound nimodipine, which typically IL-1␤ inhibitors to reduce infarct size come after stroke.44,45 The underly- blocks the L-type calcium channel. after ischemia.37 ing mechanisms of recovery in- Nimodipine was tested in at least 10

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3200

Downloaded From: https://jamanetwork.com/ on 10/04/2021 randomized, placebo-controlled clinical studies in a variety of spe- mal stroke models.68 The ␥-amino- stroke trials. Aside from the positive cies demonstrated potent infarct re- butyric acid agonists clomethiazole results with reduction of mortality ducing effects, clinical trials using and muscimol reduce infarct size in and improved neurologic outcome NMDA antagonists (selfotel, elip- animals.69,70 Clomethiazole was stud- at 6 months after stroke in an early rodil, cerestat, dextrorphan) were ied in a European phase 3 trial and trial, all other trials had negative re- negative or stopped due to adverse ef- the overall results were negative.71 sults.49 Adverse effects of IV nimo- fects.58-61 Adverse effects with some However, patients with total ante- dipine include hypotension, which of these drugs occurred before a neu- rior circulation infarcts who repre- was directly correlated to an in- roprotective plasma level could be sented 40% of the study population crease in mortality.50 A meta-anal- achieved. Adverse effects occurred in showed an 11% absolute improve- ysis of 9 trials including 3719 pa- a dose-dependent manner and in- ment rate on the primary functional tients showed a benefit for oral cluded neuropsychiatric symptoms outcome measure, the Barthel In- nimodipine (30 mg over 6 hours) in (agitation, confusion, hallucina- dex. The main adverse events asso- the subgroup treated within 12 hours tion, catatonia, ataxia, dysarthria) and ciated with clomethiazole use were of stroke onset.51 Therefore treat- hypertension. These adverse effects somnolence and rhinitis. Based on ment with oral nimodipine was again are known to be phencyclidine re- these encouraging results, a second studied within 6 hours of symptom lated and should not occur, when tar- efficacy trial restricted to patients with onset in the Very Early Nimodipine geting another subunit of the NMDA large anterior circulation strokes was Use in Stroke (VENUS) trial.52 In the receptor. Indeed, antagonists of the organized and will be completed VENUS study no benefit on out- glycine site of the NMDA receptor soon. come after stroke was observed. Nev- (GV150526, ACEA1021) are gener- ertheless, nimodipine is standard ally better tolerated without neuro- Lubeluzole treatment for prevention of ische- psychiatric adverse effects.62,63 A mic neurologic deficits after sub- phase 3 trial of GV150526 was com- Lubeluzole, a benzothiazole com- arachnoid hemorrhage.53 pleted and the results were nega- pound, is a sodium blocker that pre- tive.64 Another compound that acts vents the presynaptic glutamate re- NMDA-Antagonists on the NMDA receptor and cur- lease and reduces postsynaptic rently is being studied in a multi- excitotoxicity. Lubeluzole also pre- The NMDA antagonists reduce cal- center clinical trial is magnesium. A vents glutamate-mediated increases cium influx into neurons through pilot study showed that the drug was in NO production by inhibiting NOS postsynaptic agonist-operated cal- well tolerated and associated with a activity. Experimental studies dem- cium channels. They were the first trend toward fewer early deaths in a onstrated neuroprotective effects af- drugs that remarkably reduced in- magnesium-treated group.65 ter focal cerebral ischemia.72,73 Two farct size (40%-70%) after experi- large clinical, multicenter, and pla- mental focal cerebral ischemia, pri- AMPA-Antagonists cebo-controlled trials, 1 in North- marily in the ischemic penumbra.54,55 ern America and 1 in Europe-Aus- This effect can be achieved with com- AMPA-receptor antagonists pre- tralia, have been completed.74,75 petitive and noncompetitive NMDA vent sodium influx into the cell by Besides occasional and transient elec- antagonists. Competitive NMDA an- blockage of the AMPA/kainate re- trocardiographic QT prolongations, tagonists such as phosphonates or ceptor and prevent cell depolariza- lubeluzole was well tolerated at the selfotel block the glutamate recognition and the subsequent intracellu- concentrations given in both trials tion site of the receptor. Noncom- lar calcium overload of the cell. The (7.5 mg/h IV followed by 10 mg/d for petitive NMDA antagonists, includ- AMPA receptor antagonists such as 5 days IV). However, there was no ing phencyclidine, ketamine, NBQX and ZK200775 have potent significant effect on the primary end dizocilpine maleate, dextrorphan hy- neuroprotective capacities when point mortality, although the Ameri- drochloride, and cerestat, block the given after experimental focal cere- can trial74,75 showed a significant 7% NMDA-associated ion channel in a bral ischemia66 but adverse effects in- increase in patients who had little or use-dependent manner. Negative clude nephrotoxicity (NBQX) and no disability at 3 months after stroke. modulation of receptor activity can sedation (ZK200775). A meta-analysis of 1375 patients sug- be achieved by zinc and hydrogen. gested a positive effect on mild- Magnesium blocks the channel in a ␥-Aminobutyric Agonists moderate strokes, but no effect on se- voltage-dependent manner and has vere strokes.76 Therefore, a third trial been shown to reduce infarct size af- Enhancing the activity of the ␥-ami- (LUB-INT-13) was performed, de- ter focal cerebral ischemia. Apart nobutyric acid A subtype receptor is signed to detect an approximately 7% from the main recognition site for another mechanism for inducing functional benefit in the active treat- glutamate, the receptor also con- neuroprotection.67 This inhibitory ment group.77 The results were nega- tains a glycine site and inhibition of neurotransmission system hyperpo- tive and development was stopped. the glycine action reduces NMDA re- larizes and stabilizes resting mem- ceptor activity. Antagonists of the gly- brane potential and may therefore Free Radical Scavengers cine site of the receptor also reduce inhibit peri-infarct–depolarizing infarct size after experimental focal events, a phenomenon that is asso- Tirilazad mesylate, a 21-aminosteroid, cerebral ischemia.56,57 Although pre- ciated with infarct evolution in ani- acts as a free radical scavenger and has

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3201

Downloaded From: https://jamanetwork.com/ on 10/04/2021 antioxidant effects. Tirilazad treat- antibodies were shown to reduce in- tery disease.99,100 Since this reduc- ment reduces infarct size after tran- farct size and improve outcome af- tion was independent of the serum sient but not after permanent focal ce- ter transient but not permanent ex- cholesterol lowering, statins may have rebral ischemia. Tirilazad was tested perimental focal cerebral ischemia.89 other effects besides their antiarthero- in several clinical trials with incon- The results of a randomized, pla- sclerotic properties. Prophylactic clusive results. The drug was well tol- cebo-controlled trial (625 patients) statin therapy in animals improves erated at a daily dose of 6 mg/kg.78,79 using the anti-ICAM antibody enli- cerebral blood flow through up- The RANTTASS (Randomized Trial momab were negative. The treat- regulation of endothelial NOS and re- of Tirilazad Mesylate in Patients With ment group received IV 160 mg the duces infarct size (30%).101 Statins Acute Stroke) trial was terminated for first day, 40 mg the next 4 days (time also inhibit the cytokine-mediated lack of efficacy.80 Release of NO and window 6 hours), and had an even (IL-1ß, tumor necrosis factor ␣) up- peroxynitrite can be inhibited by neu- worse outcome and increased mor- regulation of iNOS and production ronal NOS or inducible iNOS block- tality due to higher rates of fever, in- of NO in rat astrocytes and macro- ers. The selective neuronal NOS fection, and pneumonia.90 The ad- phages.102 Furthermore, statins may blocker 7-nitroindazole and 1-(2-tri- verse effects were thought to be due reduce lipoprotein oxidation and at- fluoromethylphenyl) imidazole sig- to a complement-mediated reac- tenuate free radical injury.103,104 More nificantly reduced infarct size after fo- tion triggered by enlimomab. studies are needed to define time win- cal and global cerebral ischemia in dow, dosage, and mechanisms of ac- animals.81,82 Blockage of iNOS can be Inhibition of Cytokines tion before statins should be consid- selectively achieved by aminoguani- ered for a future clinical use in stroke. dines. Aminoguanidines were re- The best-studied cytokines that have ported to be potent neuroprotec- key roles after ischemia and that can RECOVERY TREATMENT tants after focal cerebral ischemia.83 be pharmacologically inhibited are Aminoguanidines seem even to be IL-1␤ and tumor necrosis factor ␣. Growth Factors protective when the treatment is de- IL-1␤ mediates excitotoxicity layed for 24 hours, making them in- through NMDA receptor activation Growth factors are endogenously oc- teresting for future human use.84 and activates surface adhesion mol- curring polypeptides that have not Another promising antioxi- ecules.91 Consequently, overexpres- only neuroprotective but also regen- dant is the seleno-organic com- sion of endogenous IL-1␤ receptor erative and proliferative capacities pound ebselen. Ebselen acts through antagonists or treatment with the and may therefore be unique candi- a glutathione peroxidase–like ef- IL-1␤ antagonist zinc protoporphy- dates for stroke therapy. Several fect. It inhibits the peroxidation of rin reduces infarct and edema size growth factors were shown to be membrane phospholipids and li- with temporary focal ischemia.92,93 neuroprotective after experimental poxygenase in the arachidonate cas- Another cytokine, TFN-␣ may also ischemia in vivo and in vitro. The cade. Ebselen also blocks the pro- exacerbate ischemic and in particu- best-studied growth factors after duction of superoxide anions by lar reperfusion injury.94 Inhibiting tu- focal cerebral ischemia are basic fi- activated leukocytes, inhibits iNOS, mor necrosis factor ␣ improves ce- broblast growth factor,105 brain- and protects against peroxynitrite. rebral blood flow and reduces infarct derived neurotrophic growth fac- Ebselen has been shown to be neu- size after focal cerebral ischemia.95 tor,106 insulin-like growth factor,107 roprotective after transient and Other cytokines involved in the and osteogenic protein-1.108 They are permanent experimental focal reperfusion damage after ischemia are all robust neuroprotectants after cerebral ischemia.85,86 A clinical pi- the IL-1 and IL-6, platelet activating acute stroke and reduce infarct size lot study showed that oral ebselen factor, and transforming growth fac- 35% to 50% in animals. Potential (300 mg/d) was well tolerated and tor-ß. As indicated by preclinical mechanisms of action after stroke in- significantly improves functional studies, inhibition of these cyto- clude attenuation of excitotoxicity, outcome when given within 24 kines may have neuroprotective ef- improvement of cerebral blood flow, hours of stroke.87 Given within 12 fects after cerebral ischemia.96-98 Fur- and reduction of apoptosis.109,110 Re- hours (150 mg/d) after ischemia, ther experimental studies evaluating generative and proliferative capaci- there was a trend toward reduced in- dosage and timing of cytokine an- ties of growth factors after focal is- farct volume and better outcome in tagonists in different models of fo- chemic lesions were convincingly patients treated with ebselen. A sig- cal ischemia and in different species shown for basic fibroblast growth nificant reduction in infarct vol- need to be completed before future factor and osteogenic protein-1. Both ume and outcome occurred only in clinical development. compounds induced significant im- the 6-hour subgroup.88 provement of behavioral outcome Statins without changes in infarct size when Antibodies to Intercellular given 24 hours after ischemia.111,112 Adhesion Molecules Recent trials of 3-hydroxy-3- Growth factor treatment enhanced methylglutaryl coenzyme A reduc- synaptogenesis and dendritic sprout- Monoclonal antibodies against the tase inhibitors (statins) demon- ing.111 Basic fibroblast growth fac- ICAM-1 receptor on the vascular en- strated a significant reduction of tor demonstrated a good safety pro- dothelium prevent leukocyte acti- ischemic stroke incidence in pa- file in a phase 2 study, but phase 3 vation and plugging. The ICAM-1 tients with history of coronary ar- studies have been stopped.113

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3202

Downloaded From: https://jamanetwork.com/ on 10/04/2021 Citicoline and could be future candidates for re- of times and the worst of times.” The covery studies after stroke. D- benefit of IV rt-PA when initiated Citicoline (cytidine 5’-diphosphate amphetamine improved behavioral within 3 hours of stroke onset docu- choline), a naturally occurring en- outcome and memory function up to ments that acute ischemic stroke can dogenous compound that serves as 60 days after focal cerebral ischemia. indeed respond to treatment. The re- an intermediate in the synthesis of This was correlated with enhanced sults of PROACT-2 demonstrate that membrane phosphatidylcholine, is neocortical sprouting and synapto- successful treatment can be ex- thought to have membrane-stabiliz- genesis in the group treated with D- tended to selected patients within 6 ing functions and to reduce free fatty amphetamine.47 hours of stroke onset given an intra- acid formation during stroke. Citi- arterial thrombolytic agent. This news coline reduces the size of infarction Combination Treatment is countered by the current lack of and improves neurologic outcome in documentation that any purported experimental models of focal cere- Exploring effective combination neuroprotective drug significantly im- bral ischemia.114,115 In clinical stud- therapies for stroke seems rational proves outcome when given after ies, treatment with citicoline im- since cerebral ischemia triggers a stroke onset. In addition, the approved cognitive and behavioral multitude of pathophysiological and proved use of IV rt-PA is restricted function in patients with memory biochemical events that affect the to the United States and Canada, deficits.116,117 The drug was well tol- evolution of focal ischemia differ- where only a very small percentage erated in human studies without any ently. Impeding different steps in this of stroke patients receive this inter- known adverse effects.118 Therefore, cascade with different therapeutic vention. Many lessons have been citicoline was tested in a dose-re- agents may not only synergistically learned from the myriad successful sponse study at 0.5, 1, and2gbe- enhance the neuroprotective effect and unsuccessful thrombolytic and ginning within 24 hours after stroke. but may also allow the use of lower neuroprotective trials, suggesting that The outcome based on the Barthel In- doses of each drug and consecu- future, better designed trials will likely dex was significantly better only for tively less adverse effects. This ap- demonstrate significant benefits with the 0.5-g group.119 A randomized, proach demonstrated promise in sev- appropriate safe and effective drug double-blind, placebo-controlled eral experimental studies and may treatments initiated within 6 hours phase 3 study of the 0.5-g dose of citi- serve as a future strategy for stroke of stroke onset. Also, the age of com- coline demonstrated negative re- therapy in humans. For example, bination drug trials is approaching sults, although post hoc analysis in- combining low doses of citicoline rapidly and it is combination treat- dicated that medium and severe with dizocilpine124 or basic fibro- ments directed at both the vascular stroke patients benefited from the blast growth factor125 significantly re- and cellular mechanisms of ische- treatment.120 A study targeted at pa- duces infarct size after focal cere- mic brain injury that are likely to have tients with moderate-to-severe stroke bral ischemia, whereas a low dose of the greatest impact upon stroke dis- was completed recently and did not the compounds alone was not effec- ability. This is not a time to abandon show a significant effect on the pri- tive. Strategies include not only the hope for developing safe and effec- mary outcome measure.121 combination of different neuropro- tive stroke therapies that are benefi- tective agents but also the combina- cial when initiated hours after on- Nootropic Agents tion of thrombolysis and neuro- set. Rather it is a time to reflect on protection. Adding free radical lessons learned from recent scien- Piracetam is a ␥-aminobutyric acid de- scavengers such as tirilazad,126 AMPA tific advances and clinical trials to rivative and nootropic agent with neu- antagonists (NBQX),127 NMDA an- better move forward into the new roprotective capacities mediated tagonists (dizocilpine),128 or citico- millennium. through restoration of cell mem- line129 to rt-PA treatment, extends the brane fluidity and maintenance of time window and enhances the ef- SUMMARY membrane-bound cell functions. Ex- fect of thrombolysis after stroke. perimental studies demonstrated a Combining agents for acute (eg, rt- The acute treatment of ischemic neuroprotective and regenerative PA, NMDA antagonists, free radical stroke to improve neurologic and effect of piracetam after focal le- scavengers) and recovery treatment functional outcome remains a chal- sions.122 However, a randomized, pla- (eg, growth factors or citicoline) lenging task with the potential of tre- cebo-controlled IV trial, including 927 could be an important future ap- mendous rewards for both patients patients, produced negative results for proach to stroke therapy. However, and the health care delivery sys- mortality and outcome 12 weeks af- more preclinical studies need to be tem. Currently, the only approved ter stroke. Only a subgroup of pa- completed before combination thera- therapy is IV tissue-type plasmino- tients treated within 7 hours showed pies should move into clinical devel- gen activator initiated within 3 hours a trend toward better neurologic out- opment. of stroke onset in appropriately se- come.123 A new randomized, placebo- lected patients. Intra-arterial infu- controlled, multicenter trial with a CONCLUSIONS sion of another thrombolytic agent, 7-hour time window has been started. prourokinase, within 6 hours of Amphetamines increase release at the The current status of acute stroke stroke onset also improved out- noradrenergic terminals of norepi- therapy can be appropriately de- come in a single trial, but the drug nephrine, dopamine, and serotonin scribed as a combination of “the best has not yet been approved for gen-

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3203

Downloaded From: https://jamanetwork.com/ on 10/04/2021 eral use. A large number of neuro- stroke in a community-based approach. Stroke. 29. Kristian T, Siesjo BK. Calcium in ischemic cell protective drugs were developed 1998;29:1544-1549. death. Stroke.1998;29:705-718. 12. Albers GW, Bates VE, Clark WM, et al. Intrave- 30. Rothwell NJ, Hopkins SJ. Cytokines and the ner- based on an enhanced comprehen- nous tissue-type plasminogen activator for treat- vous system, II: actions and mechanisms of ac- sion of the mechanisms of focal is- ment of acute stroke: the Standard Treatment with tion. Trends Neurosci. 1995;18:130-136. chemic brain injury. They were stud- Activase to Reverse Stroke (STARS) Study. 31. Lindsberg PJ, Hallenbeck JM, Feuerstein G. Plate- ied in pivotal clinical trials and so far JAMA. 2000;283:1145-1150. let-activating factor in stroke and brain injury. Ann there is no conclusive evidence of ef- 13. Hacke W, Kaste M, Fieschi C, et al. Randomized Neurol. 1991;30:117-129. double-blind placebo-controlled trial of throm- 32. Haring HP, Berg EL, Tsurushita N, Tagaya M, del ficacy with any of these agents. It is bolytic therapy with intravenous alteplase in acute Zoppo GJ. E-selectin appears in nonischemic tis- hoped that effective neuroprotec- ischaemic stroke (ECASS II): Second European- sue during experimental focal cerebral ische- tive therapy for acute ischemic stroke Australasian Acute Stroke Study Investigators. mia. Stroke. 1996;27:1381-1391. will be available soon and this thera- Lancet. 1998;352:1245-1251. 33. Zhang R, Chopp M, Zhang Z, Jiang N, Powers peutic approach can then be com- 14. Hacke W, Kaste M, Fieschi C, et al. Intravenous C. The expression of P- and E-selectins in three thrombolysis with recombinant tissue plasmino- models of middle cerebral artery occlusion. Brain bined with thrombolysis to en- gen activator for acute hemispheric stroke: the Res. 1998;785:207-214. hance the benefits of acute stroke European Cooperative Acute Stroke Study 34. Forster C, Clark HB, Ross ME, Iadecola C. In- therapy. (ECASS). JAMA. 1995;274:1017-1025. ducible nitric oxide synthase expression in hu- 15. Clark WM, Wissman S, Albers GW, et al. Re- man cerebral infarcts. Acta Neuropathol (Berl). combinant tissue-type plasminogen activator (al- 1999;97:215-220. Accepted for publication June 14, 2000. teplase) for ischemic stroke 3 to 5 hours after 35. Iadecola C, Zhang F, Casey R, Nagayama M, Ross We thank Gregory W. Albers, symptom onset: the ATLANTIS trial. JAMA. 1999; ME. Delayed reduction of ischemic brain injury MD, and Elaine J. Skalabrin, MD, for 282:2019-2026. and neurological deficits in mice lacking the in- their helpful comments, and Linda M. 16. Hacke W, Brott T, Caplan LR, et al. Thromboly- ducible nitric oxide synthase gene. J Neurosci. Dickman for her secretarial support. sis in acute ischemic stroke: controlled trials and 1997;17:9157-9164. clinical experience. Neurology. 1999;53(suppl): 36. Nogawa S, Zhang F, Ross ME, Iadecola C. Cyclo- Reprints: Marc Fisher, MD, 119 S3-S14. oxygenase-2 gene expression in neurons con- Belmont St, Worcester, MA 01605. 17. Fisher M, Albers GW. Applications of diffusion- tributes to ischemic brain damage. J Neurosci. perfusion magnetic resonance imaging in acute 1997;17:2746-2755. ischemic stroke. Neurology. 1999;52:1750- 37. Loddick SA, Rothwell NJ. Neuroprotective ef- REFERENCES 1756. fects of human recombinant interleukin-1 re- 18. Jansen O. Early recanalization in acute is- ceptor antagonist in focal cerebral ischaemia in 1. The National Institute of Neurological Disor- chaemic stroke saves tissue at risk defined by the rat. J Cereb Blood Flow Metab. 1996;16:932- ders and Stroke rt-PA Stroke Study Group. Tis- MRI. Lancet. 1999;353:2036-2037. 940. sue plasminogen activator for acute ischemic 19. Tong DC, Yenari MA, Albers GW, O’Brien M, 38. Li Y, Chopp M, Jiang N, Yao F, Zaloga C. Tem- stroke. N Engl J Med. 1995;333:1581-1587. Marks MP, Moseley ME. Correlation of perfu- poral profile of in situ DNA fragmentation after 2. Furlan AJ, Higashida R, Wechsler L, et al. sion- and diffusion-weighted MRI with NIHSS transient middle cerebral artery occlusion in the PROACT II: recombinant prourokinase (r-ProUK) score in acute (Ͻ6.5 hour) ischemic stroke. Neu- rat. J Cereb Blood Flow Metab. 1995;15:389- in acute cerebral thromboembolism initial trial rology. 1998;50:864-870. 397. results [abstract]. Stroke. 1999;30:234. 20. Barber PA, Darby DG, Desmond PM, et al. Pre- 39. Thornberry NA, Lazebnik Y. Caspases: enemies 3. Sherman DG, for the STAT Writers group. De- diction of stroke outcome with echoplanar per- within. Science. 1998;281:1312-1316. fibrinogenation with viprinex (ancrod) for the fusion- and diffusion-weighted MRI. Neurol- 40. Adams JM, Cory S. The Bcl-2 protein family: ar- treatment of acute ischemic stroke [abstract]. ogy. 1998;51:418-426. biters of cell survival. Science. 1998;281:1322- Stroke. 1999;30:234. 21. Patel MR, Edelman RR, Warach S. Detection of 1326. 4. Fisher M. Neuroprotection of acute ischemic hyperacute primary intraparenchymal hemor- 41. Fink K, Zhu J, Namura S, et al. Prolonged thera- stroke: where are we? Neuroscientist. 1999;6: rhage by magnetic resonance imaging. Stroke. peutic window for ischemic brain damage caused 392-401. 1996;27:2321-2324. by delayed caspase activation. J Cereb Blood Flow 5. National Institute of Neurological Disorders and 22. Schellinger PD, Jansen O, Fiebach JB, Hacke W, Metab. 1998;10:1071-1076. Stroke. Generalized efficacy of t-PA for acute Sartor K. A standardized MRI stroke protocol: 42. Endres M, Namura S, Shimizu-Sasamata M, et stroke: subgroup analysis of the NINDS t-PA comparison with CT in hyperacute intracerebral al. Attenuation of delayed neuronal death after Stroke Trial. Stroke. 1997;28:2119-2125. hemorrhage. Stroke. 1999;30:765-768. mild focal ischemia in mice by inhibition of the 6. Marler JR, Tilley BC, Lu M, et al. Earlier treat- 23. Fisher M. Characterizing the target of acute stroke caspase family. J Cereb Blood Flow Metab. 1998; ment associated with better outcome in the therapy. Stroke. 1997;28:866-87224. 18:238-247. NINDS t-PA stroke study [abstract]. Stroke. 1999; 24. Bowers MP, Rothlein R, Fagan SC, Zivin JA. 43. Scha¨bitz WR, Sommer C, Zoder W, Kiessling M, 30:244. Monoclonal antibodies preventing leukocyte ac- Schwab S. Intravenous BDNF reduces infarct size 7. Chiu D, Krieger D, Villar-Cordova C, et al. Intra- tivation reduce experimental neurologic injury and and counterregulates Bax and Bcl-2 expression venous tissue plasminogen activator for acute enhance efficacy of thrombolytic therapy. Neu- after temporary focal cerebral ischemia. Stroke. ischemic stroke: feasibility, safety, and efficacy rology. 1995;45:815-819. 2000;31:2212-2217. in the first year of clinical practice. Stroke. 1998; 25. Li F, Hsu S, Tatlisumak T, et al. Reversal of ap- 44. Nudo RJ, Milliken GW. Reorganization of move- 29:18-22. parent diffusion coefficient abnormalities and de- ment representations in primary motor cortex fol- 8. Egan R, Lutsep HL, Clark WR, et al. Open label layed neuronal death following transient focal ce- lowing focal ischemic infarcts in adult squirrel tissue plasminogen activator for stroke: the Or- rebral ischemia in rats. Ann Neurol. 1999;46: monkeys. J Neurophysiol. 1996;75:2144- egon experience. J Stroke Cerebrovasc Dis. 1999; 333-342. 2149. 8:298-290. 26. Kidwell CS, Saver JC, Mattiello J, et al. Throm- 45. Nudo RJ, Wise BM, SiFuentes F, Milliken GW. 9. Tanne D, Mansbach HH, Verro P, et al. Intrave- bolytic reversal of acute human cerebral ische- Neural substrates for the effects of rehabilita- nous rt-PA therapy for stroke in clinical prac- mic injury demonstrated by diffusion perfusion tive training on motor recovery after ischemic in- tice: a multicenter evaluation of outcome [ab- MRI [abstract]. Neurology. 1999;52:A536. farct. Science. 1996;272:1791-1794. stract]. Stroke. 1998;29:288. 27. Zhang RL, Zhang ZG, Chopp M. Increased thera- 46. Kawamata T, Alexis NE, Dietrich WD, Finkle- 10. Wang DZ, McLean JM, Rose JA, Honings D, Mil- peutic efficacy with rt-PA and anti-CD18 anti- stein SP. Intracisternal basic fibroblast growth brandt JC. The outcome of administering intra- body treatment of stroke in the rat. Neurology. factor (bFGF) enhances behavioral recovery fol- venous rt-PA for acute ischemic strokes at OSF 1999;52:273-279. lowing focal cerebral infarction in the rat. J Cereb Comprehensive Stroke Center [abstract]. Neu- 28. Siesjo BK. Pathophysiology and treatment of fo- Blood Flow Metab. 1996;16:542-547. rology. 1998;50:A436. cal cerebral ischemia, part II: mechanisms of 47. Stroemer RP, Kent TA, Hulsebosch CE. En- 11. Grond M, Stenzel C, Schmulling S, et al. Early damage and treatment. J Neurosurg. 1992;77: hanced neocortical neural sprouting, synapto- intravenous thrombolysis for acute ischemic 337-354. genesis, and behavioral recovery with D-

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3204

Downloaded From: https://jamanetwork.com/ on 10/04/2021 amphetamine therapy after neocortical infarction a phosphonate quinoxalinedione AMPA antago- The neuroprotective efficacy of ebselen (a gluta- in rats. Stroke. 1998;29:2381-2393. nist for neuroprotection in stroke and trauma. Proc thione peroxidase mimic) on brain damage in- 48. Grotta JC. Clinical aspects of the use of calcium Natl Acad SciUSA.1998;95:10960-10965. duced by transient focal cerebral ischaemia in the antagonists in cerebrovascular disease. Clin Neu- 67. Green AR, Cross AJ, Snape MF, deSouza RJ. The rat. Neurosci Lett. 1995;185:65-69. ropharmacol. 1991;14:373-390. immediate consequences of MCA occlusion on 86. Takasago T, Peters EE, Graham DI, Masayasu H, 49. Gelmers HJ, Gorter K, de Weerdt CJ, Wiezer HJ. GABA synthesis in mouse cortex and cerebel- Macrae IM. Neuroprotective efficacy of eb- A controlled trial of nimodipine in acute ische- lum. Neurosci Lett. 1992;138:141-144. selen, an anti-oxidant with anti-inflammatory ac- mic stroke. N Engl J Med. 1988;318:203-207. 68. Hossman KA. Periinfarct depolarizations. Cere- tions, in a rodent model of permanent middle ce- 50. Wahlgren N, MacMahon DG, DeKeyser J, et al. brovasc Brain Metab Rev. 1996;16:202-213. rebral artery occlusion. Br J Pharmacol. 1997; Intravenous Nimodipine West European Stroke 69. Sydserff SG, Cross AJ, Green AR. The neuro- 122:1251-1256. Trial (INWEST) of nimodipine in the treatment protective effect of chlomethiazole on ischemic 87. Yamaguchi T, Sano K, Takakura K, et al. Eb- of acute ischemic stroke. Cerebrovasc Dis. 1994; neuronal damage following permanent middle ce- selen in acute ischemic stroke: a placebo- 4:204-210. rebral artery ischemia in the rat. Neurodegen- controlled, double-blind clinical trial. Stroke. 51. Mohr J, Orgagozzo JM, Harrison MJG, et al. Meta- eration. 1995;4:323-328. 1998;29:12-17. analysis of oral nimodipine trials in acute ische- 70. Lyden PD. GABA and neuroprotection. In: Green 88. Ogawa A, Yoshimoto T, Kikuchi H, et al. Eb- mic stroke. Cerebrovasc Dis. 1994;4:197-203. AR, Cross AJ, eds. Neuroprotective Agents and selen in acute middle cerebral artery occlusion: 52. Horn J, Haas R, Vermuelen M, Limburg M. Cerebral Ischemia. London, England: Aca- a placebo-controlled, double-blind clinical trial. VENUS (Very Early Nimodipine Use in Stroke): demic Press Ltd; 1997:233-258. Cerebrovasc Dis. 1999;9:112-118. preliminary trial results [abstract]. Stroke. 71. Wahlgren NG, Rawasinha KW, Rosolacci T, et 89. Zhang RL, Chopp M, Siang N, et al. Anti- 1999;30:242. al. Clomethiazole Acute Stroke Study (CLASS): intercellular adhesion molecule-1 antibody re- 53. Feigin VL, Rinkel GJE, Algra A, Vermuelen M, Van results of a randomized, controlled trial of clo- duces ischemic cell damage after transient but Gijn J. Calcium antagonists in patients with an- methiazole versus placebo in 1360 acute stroke not permanent middle cerebral artery occlusion eurysmal subarachnoid hemorrhage: a system- patients. Stroke. 1999;30:21-28. in the Wistar rat. Stroke. 1995;26:1438-1442. atic review. Neurology. 1998;50:876-883. 72. Aronowski J, Strong R, Grotta JC. Treatment of 90. Enlimomab Acute Stroke Trial Study Group. The 54. Park CK, Nehls DG, Graham DI, Teasdale GM, experimental focal ischemia in rats with lubelu- enlimomab acute stroke trial final results. Neu- McCulloch J. The glutamate antagonist MK- zole. Neuropharmacology. 1996;35:689-693. rology. 1997;48:A270. 801 reduces focal ischemic brain damage in the 73. Culmsee C, Junker V, Wolz P, Semkova I, Kriegl- 91. Jean WC, Spellman SR, Nussbaum ES, Low WC. rat. Ann Neurol. 1988;24:543-551. stein J. Lubeluzole protects hippocampal neu- Reperfusion injury after focal cerebral ischemia: 55. Choi DW. Methods for antagonizing glutamate rons from excitotoxicity in vitro and reduces brain the role of inflammation and the therapeutic ho- neurotoxicity. Cerebrovasc Brain Metab Rev. damage caused by ischemia. Eur J Pharmacol. rizon. Neurosurgery. 1998;43:1382-1396. 1990;2:105-147. 1998;342:193-201. 92. Betz AL, Yang GY, Davidson BL. Attenuation of 56. Tatlisumak T, Takano K, Meiler MR, Fisher M. A 74. Grotta J, and the the US and Canadian Lubelu- stroke size in rats using an adenoviral vector to glycine site antagonist, ZD9379, reduces num- zole Ischemic Stroke Study Group. Lubeluzole induce overexpression of interleukin-1 recep- ber of spreading depressions and infarct size in treatment of acute ischemic stroke. Stroke. 1997; tor antagonist in brain. J Cereb Blood Flow Metab. rats with permanent middle cerebral artery oc- 28:2338-2346. 1995;15:547-551. clusion. Stroke. 1998;29:190-195. 75. Diener HC, and the European and Australian 93. Kadoya C, Domino EF, Yang GY, Stern JD, Betz 57. Takano K, Tatlisumak T, Formato J, et al. A gly- Lubeluzole Ischaemic Stroke Study Group. Mul- AL. Preischemic but not postischemic zinc pro- cine site antagonist attenuates infarct size in ex- tinational randomized controlled trial of lubelu- toporphyrin treatment reduces infarct size and perimental focal ischemia: postmortem and dif- zole in acute ischaemic stroke. Cerebrovasc Dis. edema accumulation after temporary focal ce- fusion mapping studies. Stroke. 1997;28:1255- 1998;8:172-181. rebral ischemia in rats. Stroke. 1995;26:1035- 1262. 76. Hantson L, Wessel T. Therapeutic effects of lube- 1038. 58. Lees KR. Cerestat and other NMDA antagonists luzole in ischemic stroke [abstract]. Stroke. 1998; 94. Barone FC, Arvin B, White RF, et al. Tumor ne- in ischemic stroke. Neurology. 1997;49:66-69. 29:287. crosis factor-alpha: a mediator of focal ische- 59. Davis SM, Albers GW, Diener HC, Lees KR, Nor- 77. Diener HC. Lubeluzole in acute stroke: lack of ef- mic brain injury. Stroke. 1997;28:1233-1244. ris J. Termination of Acute Stroke Studies In- ficacy in a large phase III study with an 8 hour 95. Dawson DA, Martin D, Hallenbeck JM. Inhibi- volving Selfotel Treatment: ASSIST Steering window [abstract]. Stroke. 1999;30:234. tion of tumor necrosis factor-alpha reduces fo- Committee [letter]. Lancet. 1997;349:32. 78. STIPAS Investigators. Safety study of tirilazad cal cerebral ischemic injury in the spontane- 60. Albers GW, Atkinson RP, Kelley RE, Rosenbaum mesylate in patients with acute ischemic stroke ously hypertensive rat. Neurosci Lett. 1996;25; DM, for the Dextrorphan Study Group. Safety, tol- (STIPAS). Stroke. 1994;25:418-423. 218:41-44. erability, and pharmacokinetics of the N-methyl- 79. Peters GR, Hwang LJ, Musch B, Brosse DM, Or- 96. Panetta T, Marcheselli VL, Braquet P, Bazan NG. D-aspartate antagonist dextrorphan in patients with gogozzo JM. Safety and efficacy of 6mg/kg/d Arachidonic acid metabolism and cerebral blood acute stroke. Stroke. 1995;26:254-258. tirilazad mesylate in patients with acute ische- flow in the normal, ischemic, and reperfused ger- 61. Fisher M, for CNS1102 Investigators. Cerestat mic stroke. Stroke. 1996;27:195. bil brain. Ann N Y Acad Sci. 1989;559:340-351. (CNS 1102), a noncompetitive NMDA antago- 80. RANTTAS Investigators. A randomized trial of 97. Prehn JH, Backhauss C, Krieglstein J. Transform- nist, in ischemic stroke patients: dose escalat- tirilazad mesylate in patients with acute ische- ing growth factor-beta 1 prevents glutamate neu- ing safety study [abstract]. Cerebrovasc Dis. mic stroke. Stroke. 1996;27:1453-1458. rotoxicity in rat neocortical cultures and protects 1994;4:245. 81. Escott KJ, Beech JS, Haga KK, Williams SC, Mel- mouse neocortex from ischemic injury in vivo. 62. Albers GW, Clark WM, Atkinson RP et al. Dose drum BS, Bath PM. Cerebroprotective effect of the J Cereb Blood Flow Metab. 1993;13:521-525. escalation study of the NMDA glycine site an- nitric oxide synthase inhibitors, 1-(2-trifluo- 98. Gross CE, Bednar MM, Howard DB, Sporn MB. tagonist ACEA in acute ischemic stroke [ab- romethylphenyl) imidazole and 7-nitro indazole, Transforming growth factor-beta 1 reduces in- stract]. Stroke. 1997;28:233. after transient focal cerebral ischemia in the rat. farct size after experimental cerebral ischemia 63. Lees KR, Lavelle JF, Hobbinger SF, for the Gain J Cereb Blood Flow Metab. 1998;18:281-287. in a rabbit model. Stroke. 1993;24:558-562. European Study Group. Safety and tolerability of 82. Nanri K, Montecot C, Springhetti V, Seylaz J, Pi- 99. Sacks FM, Pepper MA, Moye LA, et al, for the GV150526 in acute stroke [abstract]. Cerebro- nard E. The selective inhibitor of neuronal nitric Cholesterol and Recurrent Events Trial Investi- vasc Dis. 1998;8:20. oxide synthase, 7-nitroindazole, reduces the de- gators. The effect of pravastatin on coronary 64. Lees KR, for the GAIN International Study Group. layed neuronal damage due to forebrain ische- events after myocardial infarction in patients with Results of the GAIN International Trial. Paper pre- mia in rats. Stroke. 1998;29:1248-1253. average cholesterol levels. N Engl J Med. 1996; sented at: 25th International Stroke Confer- 83. Cockcroft KM, Meistrell M, Zimmerman GA, et 335:1001-1009. ence, American Stroke Association meeting; al. Cerebroprotective effect of aminoguanidine 100. Stalenhoef AF. Scandinavian Simvastatin Study January 2000; New Orleans, La. in a rodent model of stroke. Stroke. 1996;27: (4S). Lancet. 1994;344:1766-1767. 65. Muir KW, Lees KR. A randomized, double- 1393-1398. 101. Endres M, Laufs U, Huang Z, et al. Stroke pro- blind, placebo-controlled pilot trial of intrave- 84. Iadecola C, Zhang F, Xu X. Inhibition of iNOS ame- tection by 3-hydroxy-3-methylglutaryl (HMG)- nous magnesium in acute stroke. Stroke. 1995; liorates cerebral ischemic damage. Am J Physiol. CoA reductase inhibitors mediated by endothe- 26:1183-1186. 1995;268:286-292. lial nitric oxide synthase. Proc Natl Acad Sci 66. Turski L, Huth A, Sheardown M, et al. ZK200775: 85. Dawson DA, Masayasu H, Graham DI, Macrae IM. USA.1998;95:8880-8885.

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3205

Downloaded From: https://jamanetwork.com/ on 10/04/2021 102. Pahan K, Sheikh FG, Namboodiri AM, Singh I. duces infarct size after experimental ischemia in domized efficacy trial of citicoline in patients with Lovastatin and phenylacetate inhibit the induc- a rat model. Stroke. 1995;26:2154-2159. acute ischemic stroke. Stroke. 1999;30:2592- tion of nitric oxide synthase and cytokines in rat 111. Kawamata T, Ren J, Chan TC, Charette M, Finkle- 2597. primary astrocytes, microglia, and macro- stein SP. Intracisternal osteogenic protein-1 en- 121. Gammans RE, Sherman DG, and the ECCO 200 phages. J Clin Invest. 1997;100:2671-2679. hances functional recovery following focal stroke. investigators. ECCO study of citicoline for treat- 103. Hussein O, Schlezinger S, Rosenblat M, Keidar S, Neuroreport. 1998;9:1441-1445. ment of acute ischemic stroke [abstract]. Stroke. Aviram M. Reduced susceptibility of low density 112. Kawamata T, Dietrich DW, Schallert E, et al. In- 2000;3l:278. lipoprotein (LDL) to lipid peroxidation after flu- tracisternal basic fibroblast growth factor en- 122. Coq JO, Xerri C. Acute reorganization of the fore- vastatin therapy is associated with the hypocho- hances functional recovery and up-regulates the paw representation in the rat SI cortex after fo- lesterolemic effect of the drug and its binding to expression of a molecular marker of neuronal cal cortical injury: neuroprotective effects of pi- the LDL. Atherosclerosis. 1997;128:11-18. sprouting following focal cerebral infarction. Proc racetam treatment. Eur J Neurosci. 1999;11: 104. Chen L, Haught WH, Yang B, Saldeen TG, Natl Acad SciUSA.1997;94:8179-8184. 2597-2608. Parathasarathy S, Mehta JL. Preservation of en- 113. The FIBLAST Safety Study Group. Clinical safety 123. De Deyn PP, Reuck JD, Deberdt W, Vlietinck R, dogenous antioxidant activity and inhibition of lipid of intravenous bFGF in acute stroke. Stroke. 1998; Orgogozo JM, for the Piracetam in Acute Stroke peroxidation as common mechanisms of antiath- 29:287. Study (PASS) Group. Treatment of acute ische- erosclerotic effects of vitamin E, lovastatin and am- 114. Schabitz WR, Weber J, Takano K, et al. The ef- mic stroke with piracetam. Stroke. 1997;28: lodipine. J Am Coll Cardiol. 1997;30:569-575. fects of prolonged treatment with citicoline in 2347-2352. 105. Fisher M, Meadows ME, Do T, et al. Delayed treat- temporary experimental focal ischemia. J Neu- 124. O¨ nal MZ, Li F, Tatlisumak T, et al. Synergistic ef- ment with intravenous basic fibroblast growth rol Sci. 1996;138:21-25. fects of citicoline and MK-801 in temporary ex- factor reduces infarct size following permanent 115. Aronowski J, Strong R, Grotta JC. Citicoline for perimental focal ischemia in rats. Stroke. 1997; focal cerebral ischemia in rats. J Cereb Blood Flow treatment of experimental focal ischemia: his- 28:1060-1065. Metab. 1995;15:953-959. tologic and behavioral outcome. Neurol Res. 125. Scha¨bitz WR, Li F, Irie K, et al. Synergistic ef- 106. Scha¨bitz WR, Schwab S, Spranger M, Hacke W. 1996;18:570-574. fects of a combination of low-dose bFGF and citi- Intraventricular brain-derived neurotrophic fac- 116. Spiers PA, Myers D, Hochanadel GS, Lieber- coline after temporary experimental focal ische- tor reduces infarct size after focal cerebral is- man HR, Wurtman RJ. Citicoline improves ver- mia. Stroke. 1999;30:427-432. chemia in rats. J Cereb Blood Flow Metab. 1997; bal memory in aging. Arch Neurol. 1996;53:441- 17:500-506. 448. 126. Meden P, Overgaard K, Pedersen H, Boysen G. 107. Loddick SA, Liu XJ, Lu ZX et al. Displacement 117. Alvarez XA et al. Citicoline improves memory per- Effect of early treatment with tirilazad combined of insulin-like growth factors from their binding formance in elderly subjects. Methods Find Exp with delayed thrombolytic therapy in rat embolic proteins as a potential treatment for stroke. Proc Clin Pharmacol. 1997;19:201-210. stroke. Cerebrovasc Dis. 1996;6:141-148. Natl Acad SciUSA. 1998;95:1894-1898 118. Tazaki Y, Sakai F, Otomo E et al. Treatment of 127. Meden P, Overgaard K, Sereghy T, Boysen G. En- 108. Lin SZ, Hoffer BJ, Kaplan P, Wang Y. OP-1 pro- acute cerebral infarction with a choline precur- hancing the efficacy of thrombolysis by AMPA tects against cerebral infarction induced by MCA sor in a multicenter double-blind placebo- receptor blockage with NBQX in a rat embolic ligation in rats. Stroke. 1999;30:126-133. controlled study. Stroke. 1988;19:211-216. stroke model. J Neurol Sci. 1993;119:209-216. 109. Freese A, Finklestein SP, DiFiglia M. Basic fibro- 119. Clark WM, Warach SJ, Pettigrew LC, Gammans 128. Zivin JA, Mazzarella V. Tissue plasminogen ac- blast growth factor protects striatal neurons in RE, Sabounjian LA, for the Citicoline Stroke Study tivator plus glutamate antagonists improves out- vitro from NMDA-receptor mediated excitotox- Group. A randomized dose-response trial of citi- come after embolic stroke. Arch Neurol. 1991; icity. Brain Res. 1992;575:351-355. coline in acute ischemic stroke patients. Neu- 48:1235-1240. 110. Tanaka R, Miyasaka Y, Yada K, Ohwada T, Ka- rology. 1997;49:671-678. 129. Anderson M, Meden P, Overgaard K, Boysen G. meya T. Basic fibroblast growth factor in- 120. Clark WM, Williams BJ, Selzer KA. For the citi- Effects of citicoline combined therapy in a rat em- creases regional cerebral blood flow and re- coline acute ischemic stroke study group: a ran- bolic stroke model. Stroke. 1999;30:1464-1471.

(REPRINTED) ARCH INTERN MED/ VOL 160, NOV 27, 2000 WWW.ARCHINTERNMED.COM 3206

Downloaded From: https://jamanetwork.com/ on 10/04/2021