Him ORIGINAL CONTRIBUTION Glycine Antagonist in Neuroprotection for Patients With Acute Stroke GAIN Americas: A Randomized Controlled Trial Ralph L. Sacco, MD.MS_____ Context Elucidation ofthe ischemic cascade has helped stimulate development of Janet T. DeRosa, MPH neuroprotective drugs aimed at limiting brain injury in the hours following an ische- E, Clarke Haley, Je MD mic stroke. To date, none of these drugs has shown clinical efficacy. Objective To examine the efficacy of gavestinl (GV150526), an antagonist ofthe glycine site of the N-methy/-b-aspartate receptor, as a neuroprotective therapy for acute ischemic stroke when administered within 6 hours of symptom onset. Design The Glycine Antagonist in Neuroprotection (GAIN) Americas trial, a random= ized, double-blind placebo-controled trial with enrollment from April 1998 to Octo- ber 1999, Setting One hundred thirty-two hospital centers across the United States and John L. P. Thompson, PhD Canada’ forthe GATN Americas Inver Patients The primary efficacy population consisted of 1367 ischemic stroke pa- evr tents ith a predefined level o im weakness and functional independence prio N RECENT YEARS, CONSIDERABLE EF. stoke, strait at randomization by age (#75 vs 75 years) and intial stroke sever- fortshave been devoted todevelop. iy ational Institutes of Health [NIB] Stroke Sale scores of 2-5, 6-13, or =14) Ingtrcatmentsaimedatrestoringce Intervention Patients were randomly assigned to receive anintravenousloading dose rebralperfusionorlimitingncuronal (800 mg) plu 5 maintenance doses 000 mg every 12 hours) of gavestinel (n= 70%) damage fllowingacutechemic stroke. 0 placebo (7666) for 3 days Recognition of the “ischemic pensin- Main Outcome Measure Functional capability at3 months, measured by the Bar bra,” a region of reduced cerebral blood the! Index (BI), with scores trichotomized as dead/0-55, 60-90, and 95-100, com- flow in which cell death might be pre- pared between the gavestinel and placebo groups vented,’ has focused attention on treat- Results Treatment groups were well matched for baseline characteristics. For each ments that might minimize oF reverse group, median NIH Stroke Seale was 12, median age was 72 years, and median time brain damage when admired soon ftegumentas52 hous No ubstclysgnicant mprovemest on he month Mer woke onset Enlargement of im Bthchotomy was demonstrated or gavesiel =. 79), The proportion who were fun ime lene Ie iesnundl by changes onalyindependent Slscore=95-100) was 30% nthe eves! group and37% 0 fare eum detent by chants The placebo oup. No alstcalysgntcantafereneen3monthsuntval was inmotion case served sig Kaplan Meter curves B11) No ole secondary end port suggested ee cee aia Snadantge gerne ong ih 28 pen Gh) woreda y supply non sve ype peminogen aetvate ee was aso no bene or gavestnel = 33 rane depolatzation, lease of neuro. SRMOSUE-Ypesmnogen attr, aves ( tnsmiters(ncudingglaamatin © Cgpetyrgg, I this sud, gavel administered up to 6 hours ater an ace amounts), accumulation of intracelli- Chemie stroke did not improve functional outcome at 3 months. larcaleiumn, increased production of ni. Chemie stoke did no ven jamacon tric oxide and free radicals, develop. AMA7HRORV7IP A rem ment ofcellilaredems, and fnally.cell a ca ‘he nom Saori wath Eachstepalongibe ischemic cas- AE Afliations: Columbia Unversity andthe Nova Soi (Pip death Each step along the eh New Yor reaytean Horpta New York NY. Member of GAN Arias rite tthe ed of cade offers potential target for thera- (Ore Suet, Levin, Randel, Thompean and Me thea Dek: University of Vigna Heath Sytem, ConerpondingAutorand epi ph Sac, Charoteste (Or He) Case Weleome ine! MD, NS"The Newslog tte, Coumba Ur Comment seep 176, Research Tangle Pare NC (Ors Oraronneas, vey, 710 W 68th, Room 7, New York NY z nt $e P TTC Oe ne Snipes) and Dalhousie University, Halifax, — 10032 (ema fs1Qcolumbiaedu). is cascade in- For editor (©2001 American Medical Association, A rights reserved. (Reprinted) JAMA, Ap 4, 2001 Nol 205, No 131749 Downloaded From: on 0192/2019(GLYCINE ANTAGONIST FOR ACUTE STROKE fel patients, or thet legally ate Table 1. Fy Cera thorized representatives (in accordance hohe cer ‘with local regulations), provided writ- Symptons consistant with acute stoke and presenta ime ofthe study teatment ten informed consent to enter the trial Treamant canbe rtd itn 6 hours of amplom eres i ancwo ener thet LU weckness prover Ith am arlig af sectd, hee must be dit within 10 seconds for All participating centers obtained ap- Thesrmand sescords{arheleg lon | mb alleced heimbmusttouchthebedwitn proval for the consent form and the 10 scone fr hea pacrge ft theleg study protocol from their local ethies Previous) ndependent bledied Farin sere ly p Whiten tomes const gen bythe suet rs legay author representative or institutional research review com Ha woman, must be of narchideearhg ptenal, or of cidbearng poet wth regaive mittees prior to randomizing patients. Pregary testo scteen and corfmation ofsdequae craven se Exclusion eer (bun responding eny with tle motor autonomic fet, ott unresponsive Randomization feck, wot ‘The treatment assignment schedule was Symptoms rapid merovig and kayo ras completely within 24 hours i Serre Sapien aracvoR na pve toate generated by the statistician on the in- ohn serausifevestening nee kay fo ead deathin the nxt 9 mont cpendent safety and efficacy data ‘Symptoms consistent with severe congeative heart falure dependent safety and efficacy d: Presence of maignat hypertension monitoring committee (SEDMC) and ‘no istry of pcan real mpiment or hepatic cosase ‘was given to the centralized random Pariipaton na cca ital wih aninvesbgaional ug or riemaldevce wine past months NS Previous tratment ith govern ization service. Patients were ran- Uri tobe vata for elowup domly assigned (1:1) to receive either gavestinel oF placebo, with stratifica- peutic intervention. To date, several set have improved functional out tion by age (=75 or >75 years) and classes of neuroprotective compounds come (independence) at 3 months,and stroke severity (NIHSS scores catego- hhave been investigated in phase 3 trials to further assess the compound'ssafety rized as 2-5, 6-13, =14) yielding 6 foracute stroke. They have included eal- in the clinical setting of acute stroke. strata, These NIHSS categories were se- ium channel antagonists, N-methyl-D- Here we report the final results ofGAIN — lected based on theit associations with aspartate (NMDA) receptor antago- Americas, which was conducted in the _6-month functional outcome ina larg nists, free radical scavengers, anti-_ United States and Canada. population-based study." Treatmental- intercellular adhesion molecule 1 ——______ locations were obtained by pharma- antibody, GM-1 ganglioside, y-amino- METHODS: cists at the study sites by telephoning butyric acid agonists, and sodium chan- ‘The GAIN Americas trial, conducted atthe randomization service, which used nelantagonists,amongothers.Allofthe 132 medical centersin the United States an interactive voice-response system, trials have yielded disappointing effi- and Canada, wasa randomized, double- cacy results and some showed safety blind, placebo-controlled, parallel- Treatment problems, including increased mortal- group study designed to evaluate theef-_Thestudy drug was provided in prenum- ity or psychotic effects, which resulted ficacy and safety of gavestinelin patients ered packs of amber vials containing in their early termination.** with a clinical diagnosis ofacutestroke. gavestinel sodium or vehicle placebo The compound gavestinel (Glaxo Wellcome Ine, Research Tri- (GV150526) was shown to have high Patients angle Park, NC). The contents of the vi alfinity and high selecuiity forthe gly- The eligibility criteria for the wial are als were diluted in 5% dextrose (DSW) cine site of the NMDA receptor com- outlined in TABLE 1. Sttokeseverity was duc to the poor solubility of gavestinel plex,and inhibited NMDA-induced de- assessed using the National Institates in saline. Since gavestinel slight sensi- polarization alow doses in preclinical of Health Stroke Seale (NIHSS). To- ve, the infusion bag and tubing wes studies” When administered up to 6 al baseline NIHSS scores could range covered with opaque green plastic hhours fom stroke onset, gavestinel re- from 2 t0 41, due to the limb weak- sleeves during preparation and admin- duced infarct sizeby 50% inaratmiddle ness inclusion and level of consciots- istration. The sleeves also helped pre- cerebral artery occlusion model of is- ness exclusion criteria, Eligible pa- vent unblinding of study personnel by chemic stroke.* In phase 2 studies, ents with hemorthagic stroke were keeping them Irom seeing the pale yel- sgavestine! was well olerated when ad inchided in order to provide informa- low color ofthe diluted gavestinel so- ministered at projected neuroprotee- on about the safety and preliminary lution. Accidental unblindings of pes tive doses" The phase 3 Glycine An- efficacy of gavestinel among patients sonnel involved in patient care were tagonist in Neuroprotection (GAIN) with intracerebral hemorthage (ICH). recorded in the ease eport form, tals, GAIN Americas and GAIN In- Patients who received intravenous re- ‘The study drug was given intrave- lemational, were undertaken using combinant tssue-type plasminogen ac-nowsly over 3 days, Patients received nearly identical protocols to deter ivator("-PA) within 3 hoursofstoke ther placebo or a total of 1800 mg of mine whether patients treated with onset, and who met the eligibility er gavestinel administered as a loading gavestinel within 6 hours ofstrokeon- teria after FPA infusion, were en- dose of 800 mg (400 mg in 250 ml. of 4720 JAMA, Apu 4, 2001-Vol 285, No. 13 Repent) (©2001 American Medical Association. All rights reserved Downloaded From: on 0192/2019DSW over 30 minutes, then 400 mg in 250 mL of DSW over 3.5 hours), fol- lowed by 5 maintenance doses of 200 mg each in 150 mL. of DSW over 15 minutes at 12,24, 36, 48, and 60 hours after the start of the loading dose. Assessments Baseline assessments included demo- graphics, medical history, cerebrovas- cular event history, prestroke level of independence by the Modified Rankin Scale (MRS)," stroke severity by the NIHSS, and stroke subtype by the ‘Oxfordshire Community Survey Project classification.” Study personnel wer trained and certified in using the NIHSS through video training procedures. The presence of ICH was determined froma brain computed tomography (CT) sean taken within 18 hours of stroke onset, fora magnetic resonance imaging (MRI) sean taken within 6 hours. Digitized images were read by a blinded commit- tee of 3 independent neuroradiologists. Atday 7 ort hospital discharge, which- ever came sooner, stroke subtype was classified by the criteria used in the Trial of ORG1O172 in Acute Stroke Treat- ment (TOAST)."” Routine hematologic and clinical chemistry parameters ‘were measured at baseline and after treatment completion. central labora- tory performed the analyses. Study treatment was terminated if creatinine values exceeded 2.0 mg/dL. (178.8 jumnol/), or if bilirubin, aspartate asic notransferase,or alanine aminotransfer- ase values exceeded 4 times the upper limitofreferencerangeon these orlocally obtained laboratory tests, Eleetrocardio- grams were obiained at baseline and dur- ing the treatment phase and were read ata central blinded reading center. Primary Outcome The primary outcome was functional capability in activities of daily living at 3 months as measured by the Barthel Index (B1)."” Barthel Index scores were ichotomized as follows: 95-100. independence” (litle or no help r quired), 60-902 “assisted indepen- dence" (some help required), and 0-55="dependence” (help required (©2001 American Medical Association, All rights reserved, Downloaded From: on 0192/2019 GLYCINE ANTAGONIST FOR ACUTE STROKE with most orall activities). Deaths were analyzed along with the 0-55 group, The Bl cut points were chosen based on prior studies that established that scores fof 60 and 95 defined meaningful clin cal subgroups." Guidelines for ad- ministering the Bl were reviewed at pre study training meetings. To minimize bias in the ascessment of the primary outcome, the person performing the Bl at 3 months could not be a person in- volved in caring for the patient during the initial hospitalization, Secondary Outcomes Secondary outcomes included level ofin- dependence determined by Bl at 7 days, or hospital discharge (whichever came sooner) and at | month, and both stroke severity determined by NIHSS and level of handicap determined by MRSat 1 and 3 months, Other outcomes included sur- vival through 3 months anda global test, composed of dichotomies ofthe BI (05- 100, 0-90/dead), NIHSS (0-1, 2-42/ dead), and MRS (0-1, 2-5/dead) mea- sured at3 months, This was modified version ofthe global test used in the Na- ional Institute of Neurological Disor- ders and Stroke (NINDS) trial of t-PA for acute stroke in that it omitted the Glasgow Outcome Seale.” Reduction in {ischemic lesion volume growth from baseline to 3 months by diffusion- \weighted MR health care resource use through 3 months, and functional out- come and quality of life using the Stroke Impact Scale at 1 and 3 months were additional prespecified secondary out- ‘comes that will he reported separately Safety Monitoring Adverse events, both serious and non- serious, were recorded. Certain ad- verse events were designated a common to stroke patients” in the pro- tocol. These included progression of stroke, hemorshagic transformation, brain herniation, aspiration pneumo- nla, deep vein thrombosis, and recur rent stroke, For such events that also met the standard definition ofa serious ad- verse event, the requirement for imme- diate reporting by investigators was waived (with prior regulatory ap- proval) unless the event proved fatal or the investigator otherwise felt the event required urgent notification, Other types of serious adverse events followed usual reporting requirements. TheSEDMC re- viewed all safety data after the first 150 patients were enrolled, and then alter each subsequent 250 patients were rolled. Fatal events were reported to the SEDMC as they occurred. Statistical Methods The primary efficacy analysis in- cludedall ischemic stroke patients who received at least 1 dose of study drug, Randomized patients who did not go ‘on to receive any study treatment wer Following the intent-to- treat principle, patients who re: any medication were analyzed in the treatment group to which they were randomized, regardless of treatment 1 ceived. Treated patients who were lost to follow-up oF withdrew consent prior to their 3-month assessments were as- signed their last posttreatment Bl score. Barthel Index score distributions in ac- live and placebo groups we pared using the extended Mantel Haenszel x’ test (1d) stratified by baseline stroke severity and age group, at the 2-sided, 5% level of significance. The scores used for the extended Mantel- Haenszel test were 2 for BL 95-100, I for BI 60-90, and 0 for BI 0-55 o dead, These scores were optimal under a pro- portional adjacent odds model. The null, hypothesis was that distributions of the lrichotomized outcome would be the same in both treatment groups. Given the scoring, positive z values were in- lerpretable as gavestinel superior to pla ceebo, and negative z values as placebo superior to gavestinel. The trial was pow= cred a 90% at 2-tailed «=.05 for an (un- stratified) outcome distribution for the trichotomized Bl, as shown in TABLE 2 This allowed for a 10 percentage-point difference between treatment groups in the best outcome category in favor of jgavestinel and a6 percentage-point dif- ference in the worst category. The out- ccome distribution in the placebo group ‘was based on the Clomethiazole Acute Stroke Study.” A sample size of 673 exclude ved (Reprinted) JAMA, Ap 4, 2001 Nol 205, No 139724