Safety and Tolerability of the Glutamate Antagonist CGS 19755 (Selfotel) in Patients With Acute Ischemic Stroke
Results of a Phase IIa Randomized Trial
Background and Purpose CGS 19755 is a competitive N-methyl-d-aspartate (NMDA) receptor antagonist that limits neuronal damage in animal stroke models. The objectives of this multicenter (7 centers), randomized, double-blind, placebo-controlled, ascending-dose phase IIa study were to evaluate the safety and tolerability of CGS 19755 and obtain pharmacokinetic and preliminary data on its efficacious dose range in patients treated within 12 hours of hemispheric ischemic stroke.
Methods At each dose level, 6 patients were randomized to one or two intravenous bolus doses of CGS 19755, and 2 patients were randomized to placebo. An unblinded safety and monitoring committee evaluated results at each dose before ascending to the next level. All patients at the first level (1 mg/kg) received two doses separated by 12 hours. The first 2 patients at 2 mg/kg received two doses, but adverse experiences occurred in both; subsequent patient groups received single doses of 2.0, 1.75, or 1.5 mg/kg.
Results Adverse experiences (agitation, hallucinations, confusion, paranoia, and delirium) occurred in all 6 patients treated with 2 mg/kg, and in 3 of 5 at 1.75 mg/kg. Similar but milder adverse experiences were noted in 4 of 7 patients at 1.5 mg/kg and 1 of 6 patients at 1.0 mg/kg. Adverse experiences began between 20 minutes and 22 hours (mean, 8 hours) after treatment and lasted 2 to 60 hours (mean, 24 hours). Mortality was 1 of 8 in patients receiving placebo and 3 of 24 in treated patients. In treated survivors, median and mean percent improvement in National Institutes of Health Stroke Scale scores from baseline to terminal visit (mean, 86 days) was comparable at all doses, and 95% of treated patients had Barthel Index scores of ≥70 at the terminal visit.
Conclusions We conclude that a single intravenous dose of 1.5 mg/kg CGS 19755 is safe and tolerable in patients with acute ischemic stroke. An efficacy trial is indicated.
In response to an ischemic or hypoxic insult, levels of many neurotransmitters, including the endogenous excitatory amino acid glutamate, are elevated; this results in excessive neuronal stimulation of postsynaptic N-methyl-d-aspartate (NMDA) receptors.1 Activation of NMDA receptors opens the NMDA-associated calcium ion channel, allowing the entry of ions into the cell. Glutamate-induced neuronal death is believed to result from at least two possibly additive mechanisms of excitotoxic neurotoxicity: (1) rapid swelling caused by depolarization, chloride influx, and water entry and (2) delayed necrosis secondary to intraneuronal calcium accumulation.2 NMDA antagonists bind directly to the NMDA receptor and inhibit the actions of glutamate, thereby preventing the delayed pathological changes that lead to cell death.
CGS 19755 (cis-4-[phosphonomethyl]-2-piperidine carboxylic acid; Selfotel) is a competitive NMDA receptor antagonist that limits neuronal damage in animal stroke models.3 4 5 6 7 8 After global ischemia in rats, the effective dose ranged from 10 to 30 mg/kg IP (multiple injection), and neuroprotection was observed when administration was not delayed more than 30 minutes after the onset of ischemia.4 A single intravenous dose of 10 mg/kg started 5 minutes after permanent middle cerebral artery occlusion reduced infarct size in rats.5 6 7 Efficacy has also been demonstrated in a rabbit model of spinal cord ischemia.8
In normal healthy volunteers, doses up to 1 mg/kg did not produce signs or symptoms. Doses of 2 and/or 3 mg/kg produced transient central nervous system (CNS) effects including sedation, dizziness, motion sickness with disorientation, and nausea, but there were no abnormal findings on neurological examination (data on file at the CIBA-GEIGY Corporation). All subjects with adverse experiences completely returned to normal. There were no significant effects on vital signs.
The objectives of the present study were to evaluate the safety and tolerability of CGS 19755 during and after administration of two intravenous bolus injections ranging from 1 to 5 mg/kg started within 12 hours of symptom onset in patients with acute hemispheric ischemic cerebral infarction. In addition, we wanted to obtain preliminary pharmacokinetic data from patients receiving CGS 19755 in this dose range and to obtain preliminary data on a possible efficacious dose range.
This was a multicenter, double-blind, placebo-controlled, ascending-dose study. There were four different dose groups in the trial, each consisting of 6 patients assigned to CGS 19755 and 2 to placebo. An unblinded independent safety monitoring committee (SMC) composed of two stroke neurologists and a CIBA-GEIGY representative (who were not otherwise involved in the trial) evaluated the adverse experiences 2 days after dose administration in each group of 8 patients and decided whether to increase or decrease the dose for the next group. The first group received two injections of 1.0 mg/kg separated by 12 hours. The second group received 2.0 mg/kg; 3 patients received two doses separated by 12 hours, and 5 received a single dose. The third group was assigned to a single dose of 1.5 mg/kg, and the fourth group received a single dose of 1.75 mg/kg.
All personnel at each study site and at CIBA-GEIGY involved in conducting and monitoring the trial were blinded to the study drug codes.
To be included in the study, patients had to be at least 18 years of age with a diagnosis of acute ischemic cortical or subcortical cerebral infarction presumed to be in the territory of the anterior, middle, or posterior cerebral arteries, with symptoms lasting for a minimum of 60 minutes but beginning within the past 12 hours before dosing. Stroke onset for patients whose onset of symptoms occurred during sleep or were unobserved was assumed to be at the last time patients were seen to be free of stroke. A computed tomography (CT) scan compatible with the clinical diagnosis of acute ischemic stroke was required before randomization, and patients had to be awake or arousable to minor stimulation. All patients or their legal representative signed an informed consent approved by the institutional review board of each study site.
Patients were excluded if their CT scan was inconsistent with acute cerebral infarction, they had a decreased level of consciousness, they had a suspected brain stem or cerebellar localization of their stroke, or they had a seizure between the onset of their stroke symptoms and initial dose of study drug. In addition, patients whose stroke was thought to be related to psychoactive drugs or who required the use of psychoactive drugs during the acute phase of their illness were excluded. Patients with an unstable medical condition or with a preexisting psychiatric diagnosis were also excluded.
The patients were monitored closely for the development of neurological change or behavioral adverse experiences throughout the study. A National Institutes of Health (NIH) Stroke Scale was completed 1 and 12 hours after each dose. Vital signs were obtained every half hour for 2 hours, hourly for 6 hours, and then every 6 hours during the 36 hours after the study drug administration. A follow-up CT scan was obtained between days 4 and 7 after the onset of stroke and at day 90. Physical and neurological examinations (including NIH Stroke Scale) were conducted daily for the first week. The patients were reevaluated with an NIH Stroke Scale and Barthel Index on days 30 and 90 after the stroke.
Pharmacokinetic studies were carried out by drawing blood samples immediately before the first dose and at 2 minutes and 1, 3, 6, and 12 hours after the first dose, as well as at 2 minutes and at 6 hours after the second dose. For patients receiving a single dose, samples were also drawn at 9 and 16 hours after dosing.
The SMC reviewed unblinded adverse experiences as each group of 8 patients completed a dose level. Serious adverse experiences, including deaths, were reviewed by the SMC on an ongoing basis.
As expected, because of the small sample size, there was some variability between treatment groups in patient demographics (Table 1⇓). Overall, the ratio of men to women in the treated group was 2.4 to 1 compared with 1 to 1 in the placebo group. The mean time to treatment in all groups was between 7.9 and 10.3 hours from onset of symptoms.
The blood concentrations of CGS 19755 were proportional to the dose given, with a mean blood half-life of 2 to 3.3 hours (Table 2⇓). Non-CNS adverse experiences reported in more than 1 patient and thought to be trial drug–related by the investigator were vomiting (3 patients) and nausea (2 patients). Brief intermittent apnea occurred in 1 patient treated with 1.75 mg/kg and in 1 placebo patient. Other non-CNS drug-related adverse experiences (occurring in 1 patient each) included hypertension (systolic pressure >160), abdominal pain, hyperventilation, and abnormal vision at 2 mg/kg, whereas bradycardia, atrial fibrillation, and supraventricular tachycardia were reported in placebo patients. There were no significant drug-related effects on vital signs, electrocardiograms, or laboratory values.
CNS adverse experiences that were possibly, probably, or highly probably drug related (as designated by the investigator) were common and appeared to be dose related (Table 3⇓). These adverse experiences included hallucinations, agitation, confusion, dysarthria, ataxia, delirium, paranoid reaction, and somnolence. Onset time from dosing ranged from 20 minutes to 22 hours, and average onset was approximately 8 hours. Symptoms lasted from 2 to 60 hours with an average duration of approximately 24 hours. All symptoms resolved with no permanent sequelae. Some patients required supportive treatment with intravenous lorazepam (0.5 to 4 mg total) or haloperidol (2 to 10 mg total). Adverse experiences developed in a dose-dependent fashion. Two doses of 1.0 mg/kg were well tolerated; however, 2.0 mg/kg given twice, as well as single doses of 2.0 and 1.75 mg/kg, were felt to be too high in this patient population. With high doses, some patients fluctuated between agitation and a state of wakeful unresponsiveness or somnolence. Hallucinations were variable, consisting of a sensation of movement or auditory or visual phenomena. A single dose of 1.5 mg/kg was determined to be the clinically tolerated dose.
Mortality in the overall CGS 19755 group was 3 of 24 (12.5%), and in placebo patients it was 1 of 8 (12.5%). Deaths in the treated group were due to cerebral herniation (1) and pneumonia/sepsis (2) and were not felt to be drug related. The placebo patient died of congestive heart failure.
Among surviving patients, there was improvement in the NIH Stroke Scale score from admission to 3-month follow-up in all patient groups. Overall, treated patients had slightly less severe strokes than placebo patients, making direct comparison difficult (Fig 1⇓, top). Percent change in the NIH Stroke Scale scores from baseline suggested that both the median and the mean improvement was comparable in all the treated groups regardless of dose (Fig 1⇓, bottom). A significantly higher percentage of treated patients achieved an independent Barthel Index score (≥70) at day 90 than placebo patients (Fig 2⇓).
This study demonstrates that when given to acute stroke patients, the competitive NMDA receptor antagonist CGS 19755 caused CNS adverse experiences in a dose-dependent fashion. With doses of 1.5 mg/kg or less, the adverse experiences were relatively mild and were best managed with reassurance, a quiet environment, and sedation with occasional low doses of lorazepam or haloperidol. Because doses of CGS 19755 above 1.5 mg/kg produced severe agitation, hallucinations, and paranoia that required more intense management, 1.5 mg/kg was felt to be the maximum tolerated dose.
Stroke patients had more adverse experiences than normal volunteers at equivalent doses. Although the cause of this is unknown, the stroke population might have a greater tissue bioavailability of drug due to blood-brain barrier disruption or tissue sensitivity due to altered NMDA receptor activity and kinetics.9
Interestingly, the duration of the CNS adverse events lasted for an average of 24 hours, suggesting that neuronal binding of CGS 19755 is longer than the plasma half-life of only 2 to 3 hours. This indicates that a single dose of the drug might have a biological effect at the NMDA receptor for 24 hours or longer. In most microdialysis studies of glutamate release after ischemia, glutamate concentrations have returned to normal well within this time frame.10
The numbers of patients in this study were too small to make any definitive comments about clinical outcome. Furthermore, the treatment groups were not evenly matched for the severity of stroke as measured by the NIH Stroke Scale at the time of randomization. However, it does not appear from the data that long-term outcome was adversely affected by CGS 19755. Furthermore, the maximal tolerated dose of 1.5 mg/kg appeared to be no less effective than higher doses. The suggestion of improved functional outcome in treated patients compared with patients receiving placebo must be confirmed with a larger efficacy study of patients evenly matched for stroke severity.
In conclusion, these data show that although adverse experiences occur with the NMDA receptor antagonist CGS 19755 in a dose-dependent fashion, a single intravenous dose of 1.5 mg/kg is safe and tolerable with adverse experiences that are controllable. An efficacy trial of this dose is indicated in acute ischemic stroke patients.
This study was funded by the CIBA-GEIGY Corporation.
- Received November 28, 1994.
- Revision received January 26, 1995.
- Accepted January 27, 1995.
- Copyright © 1995 by American Heart Association
Simmonds J, Sailer T, Moyer J. The effects of CGS-19755 in rat focal cerebral ischemia produced by tandem ipsilateral common carotid artery and middle cerebral artery occlusion. Soc Neurosci Abstr. 1993;19:1647. Abstract.
Madden K, Clark W, Zivin J. Delayed therapy of experimental ischemia with competitive N-methyl-d-aspartate antagonists in rabbits. Stroke. 1993;24:1068-1071.
Kaku D, Giffard R, Choi D. Neuroprotective effects of glutamate antagonists and extracellular acidity. Science. 1993;260:1516-1518.