ScienceDirect - Pharmacology & Therapeutics : Nitrones as neuroprotective agents in cerebral ischemia, with particular reference to NXY-059

Pharmacology & Therapeutics
Volume 100, Issue 3 , December 2003, Pages 195-214

doi:10.1016/j.pharmthera.2003.07.003
Copyright © 2003 Elsevier Inc. All rights reserved. Full Text (PDF)

Associate Editor: W.C. Bowman

Nitrones as neuroprotective agents in cerebral ischemia, with particular reference to NXY-059

A. Richard Green^a^,, Tim Ashwood^b, Tomas Odergren^b and David M. Jackson^c

^a AstraZeneca R&D Charnwood, Bakewell Road, Loughborough, Leics LE11 5RH, UK
^b AstraZeneca R&D Södertälje, Sodertalje S 151-85, Sweden
^c Department of Pharmacology and Toxicology, School of Medical Sciences, University of Otago, P.O. Box 9131, Dunedin, New Zealand

Available online 27 November 2003.

Abstract

Stroke is a major clinical problem, and acute pharmacological intervention with neuroprotective agents has so far been unsuccessful. Recently, there has been considerable interest in the potential therapeutic benefit of nitrone-derived free radical trapping agents as neuroprotective agents. Nitrone compounds have been shown to be beneficial in animal models of various diseases, and the prototypic compound alpha-phenyl-N-tert-butylnitrone (PBN) has been extensively demonstrated to be neuroprotective in rat models of transient and permanent focal ischemia. The nitrone radical trapping agent disodium 2,4-disulfophenyl-N-tert-butylnitrone (NXY-059) has also been shown to be neuroprotective in these models. Furthermore, it has recently been shown to improve neurological function and reduce infarct volume in a primate model of permanent focal ischemia even when given 4 hr postocclusion. While radical trapping activity is demonstrable with NXY-059 and other nitrone compounds such as PBN, this activity is weak. Arguments for and against ascribing radical trapping as the therapeutic mechanism of action are discussed. This compound is well tolerated in human stroke patients and can be administered to produce plasma concentrations exceeding those effective in animal models; crucially, at the same time, it has also been shown to be effective in animal models. NXY-059 may thus be the first compound to be examined in stroke patients using drug exposure and time to treatment that have been shown to be effective in animal models of stroke.

Author Keywords: Nitrones; PBN; NXY-059; Free radical trapping agents; Neuroprotection; Stroke

Abbreviations: 4-POBN, alpha-(4-pyridyl-1-oxide)-N-tert-butylnitrone; 5-HT, 5-hydroxytryptamine or serotonin; AR-R15896AR, S-(+)- alpha-phenyl-2-pyridine ethanamine dihydrochloride; Cl_CR, creatinine clearance; DMPO, 5,5-dimethyl--pyrroline N-oxide; ICH, intracerebral hemorrhage; MCA, middle cerebral artery; MPP⁺, 1-methyl-4-phenylpyridinium; MPTP, 1,methyl-4-phenyl-1,2,3,6-tetrahydropyridine; MRI, magnetic resonance imaging; NIHSS, National Institutes of Health Stroke Scale; NMDA, N-methyl--aspartate; NXY-059, disodium 2,4-disulfophenyl-N-tert-butylnitrone; PBN, alph-phenyl-N-tert-butylnitrone; SAE, serious adverse events; S-PBN, 2-sulfophenyl-N-tert-butylnitrone; STAIR, Stroke Therapy Academic Industry Roundtable; TBI, traumatic brain injury; TEMPO, 2,2,6,6-tetramethylpiperidine-N-oxyl; t-PA, tissue plasminogen activator

Article Outline

1. Introduction
2. alpha-Phenyl-N-tert-butylnitrone, 2-sulfophenyl-N-tert-butylnitrone, and neuroprotection
3. MDL 101,002, 2,2,6,6-tetramethylpiperidine-N-oxyl, and neuroprotection
4. Disodium 2,4-disulfophenyl-N-tert-butylnitrone and animal models of stroke
5. Neuroprotective mechanism of action of nitrone compounds
6. Clinical studies with disodium 2,4-disulfophenyl-N-tert-butylnitrone
7. General conclusions
References

1. Introduction

1.1. Stroke and neuroprotective agents

Stroke is the third leading cause of death in major industrialized countries [Bonita, 1992], with an incidence of ~350 per 100,000 population aged 45–89 [Murray & Lopez, 1997]. While the rate of stroke has fallen in the last 15 years, the absolute number has increased and will continue to rise due to the increasingly elderly population. Approximately 20% of stroke patients do not survive longer than 1 month, and a third of those who are alive after 6 months are dependent on others [Warlow, 1998]. Stroke is therefore a major cause of long-lasting disability, and this has major repercussions not only for the survivor but also for the family and society as a whole.

We now have a reasonable understanding of the biochemical consequences of an acute ischemic stroke based on many years of detailed investigation in the brain of experimental animals subjected to an acute ischemic stroke and appropriate in vitro studies. The fact that the pathological consequences of stroke, in terms of both histology and functional outcome, appear to be similar in experimental animals and humans has led to the view that the biochemical mechanisms following the ischemic insult are likely to be similar in humans and animals. Consequently, substantial numbers of experimental drugs have been investigated over the last 20 years, which have been targeted at various parts of the "ischemic cascade" (i.e., the biochemical chain of events that is initiated by the cerebral ischemia) (Fig. 1). The hope has been that by attenuating or preventing some of the biochemical consequences of stroke, the neurological damage will be lessened; thus, the functional disabilities that occur will also be prevented or reduced. Compounds designed to interfere with the mechanisms of the neurodegenerative process have been named neuroprotective agents (or drugs). This identifies them as being in a separate category from the thrombolytic or "clot buster" drugs whose mechanism is that of restoring blood flow to the compromised region.

(21K)

Fig. 1. A simplified diagram of the ischemic cascade showing compounds developed to interfere with mechanisms and provide neuroprotection.

While it is probable that the cerebral tissue in the area immediately surrounding the infarct (the core) is probably damaged beyond recovery, the surrounding area (the penumbra), although with compromised perfusion, is probably capable of recovery, given the right conditions. Without appropriate treatment, however, this penumbral tissue will also become severely damaged (see, e.g., [Snape et al., 1993]) and thus presumably worsen the clinical outcome further.

To date, no neuroprotective approaches have been approved in the United States or Europe despite the success of a variety of compounds in animal models. Reasons for the failures to establish clinical efficacy have been the subject of a substantial number of reviews, including some questioning the value of animal models [De Keyser et al., 1999 and Gladstone et al., 2002]. There appears, however, to be opportunities to employ greater stringency in the use of animal models to aid the selection of new candidate neuroprotectants and guide clinical design. In many trials, particularly those investigating N-methyl--aspartate (NMDA) antagonists, adverse events prevented the administration of doses sufficient to produce plasma levels in patients that were neuroprotective in animal studies. A recent study has further demonstrated that some compounds that have been examined clinically even produced marked adverse events in rodents when given at doses required to produce neuroprotection [Dawson et al., 2001].

Recent experimental studies, which we have conducted in rats, have demonstrated that the plasma levels that were adequate to protect in animal models of transient (or reperfusion) focal ischemia were still not sufficient to protect in a permanent focal ischemia model [Sydserff et al., 2002]. Given the fact that many infarcts reperfuse slowly, if at all [Ringelstein et al., 1992], it is reasonable to assume that a drug must be given clinically at doses that are effective in rat permanent ischemia models if it is to be broadly effective in stroke.

Because of the many failures in clinical trials of putative neuroprotective agents, an academic industrial roundtable group [Stroke Therapy Academic Industry Roundtable, 1999] met and devised guidelines for drug development (the Stroke Therapy Academic Industry Roundtable [STAIR] criteria), which included selection criteria that should be met before a compound is progressed to clinical development [Stroke Therapy Academic Industry Roundtable, 1999]. These criteria are outlined in Table 1. It can be seen that adequate dose-response data and use of permanent ischemic models to demonstrate efficacy are included. However, an additional criterion that is also listed is the use of time window studies and we would suggest that effective use of such data will be a major determinant for future clinical success. This therapeutic window of opportunity, specifically the time between the occurrence of the stroke and the time that treatment is initiated, has, until recently, often been assumed to differ in animals and humans. That is, there had been a view that damage develops more slowly in the cerebral tissue of humans and that a short time window in a rat model did not preclude giving the drug after a longer interval between stroke and drug administration in humans. A good example of this is the clinical investigation of NMDA antagonists. Despite substantial evidence for these compounds only providing protection when given shortly (60–90 min) after the ischemic insult (Fig. 2) [Massieu et al., 1993], they have nevertheless been administered to stroke patients up to 6 hr after the onset of the stroke [Davis et al., 2000]. One may well ask why. The predominant reason is probably that of practicality because it is difficult to get patients to hospital and diagnosed within 90 min of stroke onset whereas 6 hr is a reasonable time frame for presentation and treatment. Indeed, the problems of carrying out a clinical trial with a short time window are substantial. However, the success of the tissue plasminogen activator (t-PA) trial with a 3-hr time window [NINDS t-PA Stroke Study Group, 1995] shows that such studies are possible. It is noteworthy that t-PA is also efficacious in animal stroke models in the same time frame [Brinker et al., 1999], which suggests that animals and humans may be similar in their time window of opportunity.

Table 1. The STAIR criteria
(<1K)

Adapted from the criteria proposed in the paper of [Stroke Therapy Academic Industry Roundtable, 1999].

(24K)

Fig. 2. The time course of neuroprotection following administration of the NMDA antagonist AR-R15896AR in a permanent MCAO model (Sydserff & Cross, unpublished observations).

If we look at the simplified model of the neurodegenerative cascade (Fig. 1), it can be seen that most compounds investigated to date act on the early events (glutamate antagonists, ion channel compounds, clomethiazole, etc.). Consequently, one can speculate that these drugs must be given rapidly after the ischemic insult if they are to be of any value. If, as we believe, the time window of neurodegenerative events is similar in experimental animals and humans, then we have to use 1 of 2 approaches. (1) Administer the drug very soon after the stroke; an approach that is practically very difficult. (2) Develop a compound acting on a later part of the ischemic cascade that can be given in experimental animals at a time after the ischemic insult, which can be also achieved practically in clinical practice. Data from animal stroke models suggest that nitrone radical trapping compounds allow us to employ the second approach, as these compounds have a large therapeutic window of opportunity in experimental animals.

1.2. The nitrone radical trapping agents

The nitrone-derived free radical trapping agents were originally developed as tools for studying free radical chemistry and named spin traps. They allowed the indirect detection of short-lived free radical species, which, because of their reactivity, never accumulate in sufficient concentration to allow direct measurement [Janzen & Blackburn, 1968]. The nitrone compound reacts with the free radical to form a compound called a spin adduct (Fig. 3). Once the adduct is formed, it is relatively stable and the radical thus becomes inactivated and unable to damage cellular tissues or biochemical processes.

(4K)

Fig. 3. NXY-059 and the formation of an adduct.

Free radicals have been implicated in the pathology of a substantial number of disease processes and even normal aging. Any compound that interferes with free radical formation may therefore have wide utility. A variety of compounds have been developed (Fig. 4) and examined for their biological activity. Because this review focuses on cerebral ischemia, discussion of the studies of nitrone radical traps and other applications is out of place. However, the list shown in Table 2, which is adapted from that published by [Hensley et al., 1997], gives a feel for the many pharmacological actions of nitrone-based radical trapping compounds and their possible therapeutic applications.

(6K)

Fig. 4. Structures of some nitrone compounds investigated for biological activity.

Table 2. Some reported pharmacological actions of nitrone-based spin trapping compounds in addition to those relating to neuroprotection following cerebral ischemia
(<1K)

2. -Phenyl-N-tert-butylnitrone, 2-sulfophenyl-N-tert-butylnitrone, and neuroprotection

2.1. Effect of -phenyl-N-tert-butylnitrone in global and focal ischemia models in rats

The first reports on the neuroprotective action of -phenyl-N-tert-butylnitrone (PBN) in experimental models of stroke came in 1990 with publications from 2 unrelated laboratories. [Oliver et al., 1990] reported on the protective efficacy of PBN in a gerbil reperfusion model, although damage was not measured by conventional histological techniques but rather by measurement of protein carbonyl derivatives and glutamine synthase activity. PBN (300 mg/kg) injected prior to ischemia attenuated the ischemia-induced rise in carbonyl derivatives and loss of glutamine synthase activity. In addition, [Floyd, 1990] reported that PBN administration decreased the rate of mortality following prolonged global ischemia in gerbils. [Phillis & Clough-Helfman, 1990] observed that PBN (100 mg/kg) both prevented the ischemia-induced rise in locomotor activity and significantly reduced damage to hippocampal pyramidal cell layers produced by global ischemia in Mongolian gerbils. Histological evidence for protection in this model was subsequently also reported by [Yue et al., 1992].

In focal reperfusion models, PBN has also been shown to be effective, reducing the size of the infarct produced by occlusion of the middle cerebral artery (MCA) followed by reperfusion [Zhao et al., 1994, Schulz et al., 1997, Mori et al., 1998, Kuroda et al., 1999, Pazos et al., 1999 and Li et al., 2001]. [Schulz et al., 1997] observed that PBN protected the striatum in addition to the cortex. [Pazos et al., 1999] also demonstrated that the neuroprotective efficacy of PBN could be separated from the hypothermic effect that this drug can have in rodents.

There appears to have been only one study on the neuroprotective actions of PBN in permanent focal ischemia model using occlusion of the MCA. [Cao & Phillis, 1994] examined its activity in rats when administered at various times following the start of a permanent MCA occlusion (MCAO) electrocoagulation. The investigation demonstrated the substantial neuroprotective effect of the compound even when given 12 hr after the start of the ischemic episode. A marked reduction in edema was also observed (Table 3).

Table 3. Cerebral infarct volume and brain edema following PBN administration
(11K)

Data taken from [Cao & Phillis, 1994].

2.2. Effect of -phenyl-N-tert-butylnitrone and 2-sulfophenyl-N-tert-butylnitrone in embolic stroke

Most studies on focal ischemia use models that occlude the MCA by use of either an intraluminal thread or electrocoagulation. However, recently, [Yang et al., 2000] reported on the use of a model that involves injection of an autologous thrombus into the MCA, a model that is said to closely mimic the clinical situation in stroke.

This model was used to examine the effect of both PBN and 2-sulfophenyl-N-tert-butylnitrone (S-PBN) in attenuating ischemic damage. Both compounds were given as a 100-mg/kg i.p. injection once daily for 3 days starting 2 hr after the introduction of the clot, and both produced a significant reduction in infarct size (Table 4).

Table 4. Effect of focal ischemia in rats on the volume of infarct and the effect of PBN and S-PBN
(69K)

Data taken from [Yang et al., 2000].

A rather different approach to the utility of radical trapping compounds was employed by [Ashai et al., 2000]. This group examined the effect of PBN administration on the incidence of intracerebral hemorrhage (ICH) following t-PA. At present, thrombolytic stroke therapy with t-PA is complicated by the risk of secondary ICH following thrombolysis [Wardlaw et al., 1997 and Jean et al., 1998]. The reperfusion that occurs may also be associated with both further damage and free radical production [Pahlmark & Siesjö, 1996, Facchinetti et al., 1998 and Nakashima et al., 1999]. Therefore, studies on the effect of t-PA combined with a radical trapping agent might appear to be a logical approach.

[Ashai et al., 2000] introduced clot emboli to occlude the MCA in spontaneously hypertensive rats and observed high rates of cerebral hemorrhage 24 hr later when t-PA had been administered 6 hr postischemia. Infarction and neurological deficits were also worsened by t-PA. When PBN was combined with t-PA, the t-PA-induced hemorrhage volumes were reduced by 40%, with a reduction in infarction and neurological deficits also occurring. Parallel studies with Wistar-Kyoto rats failed to demonstrate similar negative effects of t-PA administration, which suggests that blood pressure is an important correlate of t-PA-induced hemorrhage. Crucially, the study suggests that PBN and other radical trapping agents might reduce the severity of t-PA-induced hemorrhage and brain injury following cerebral ischemia.

A somewhat related study was performed by [Lapchak et al., 2001] using a rabbit model of focal embolic stroke. This group also observed a marked increase in hemorrhage rate if t-PA was administered 60 min after injection of blood clots into the MCA. They also observed that pretreatment with either PBN or another nitrone, 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO), 5 min after the blood clot injection decreased the rate of hemorrhage. However, this group also reported that administration of PBN alone increased hemorrhage rate, an effect not seen with TEMPO. These data cannot unfortunately be fully compared with the rat study because [Ashai et al., 2000] did not study the effect of PBN alone on the hemorrhage rate.

2.3. Effect of -phenyl-N-tert-butylnitrone in a model of hemorrhagic stroke

While the majority of strokes are thromboembolic, 10–15% of strokes in Western populations [Anderson et al., 1994 and Gillurn, 1995] and up to 30% in Oriental populations [Burchfiel et al., 1994 and Lo et al., 1994] are due to ICH. Hemorrhagic transformation can also occur in a significant number of patients presenting with ischemic stroke [Lyden & Zivin, 1993].

It is known that iron compounds can markedly accelerate free radical processes by catalyzing the formation of hydroxyl radical compounds [Halliwell, 1992], and iron compounds such as hemoglobin and its degradation products are present in high concentration during ICH [Sadrzadeh et al., 1987]. Iron compounds could therefore worsen the situation during ICH because free radicals are already being formed as a consequence of the ischemia.

[Peeling et al., 1998] examined 2 free radical inhibitors, dimethylthiourea and the nitrone PBN, in a collagenase-induced model of ICH. Effects of drug treatment were examined by use of behavioral models, magnetic resonance imaging (MRI) and histopathology. Administration 2 hr after the ICH of either of the compounds resulted in an improved neurological deficit score. However, treatment did not reduce edema, resolution of the hematoma, or neuronal damage in the tissue adjacent to the hemorrhage.

2.4. Effect of -phenyl-N-tert-butylnitrone and 2-sulfophenyl-N-tert-butylnitrone on traumatic brain injury

[Li et al., 1997] examined the effect of PBN on compression injury to rat spinal cord (the compound being given both before and subsequent to the injury) but found no protective effect in either functional tests or histological outcome measures. However, relatively low doses of the drug were given in comparison with those required in cerebral ischemia models and this could have influenced the outcome.

In contrast, in a model of traumatic brain injury (TBI), both PBN and S-PBN have been demonstrated to reduce the loss of ipsilateral hemispheric tissue when the nitrones were administered 30 min after a fluid percussion injury [Marklund et al., 2001a]. Interestingly, both compounds attenuated increased free radical production although S-PBN, in contrast to PBN, could not be detected in cerebral tissue. The authors concluded that a major site of free radical production was in TBI at the blood-endothelial interface [Marklund et al., 2001b].

2.5. Effect of -phenyl-N-tert-butylnitrone and 2-sulfophenyl-N-tert-butylnitrone on neurotoxin-induced damage to cerebral tissue

A variety of neurotoxins are considered to induce damage through free radical-mediated events, and it is not surprising therefore that nitrones have been extensively investigated for their effects in neurotoxin-induced damage. A full review of this area is outside the brief of this article. However, given the association between cerebral ischemia and excitotoxicity, the effects of PBN and S-PBN in some studies on neurotoxin-induced cerebral damage will be mentioned.

Two studies by [Schulz et al., 1995a and Schulz et al., 1995b] have demonstrated the efficacy of S-PBN against neurotoxin-induced damage in mice. In one study, a comparison was made of the protective properties of dizocilpine, lamotrigine, and S-PBN against damage produced by the mitochondrial toxin malonate [Schulz et al., 1995a]. While dizocilpine and lamotrigine were only effective when given up to 1 hr following the malonate injection, S-PBN was effective up to 6 hr later, supporting the view that nitrones have a large window of opportunity. In the second study, the same group showed that PBN and S-PBN administration attenuated malonate, 1-methyl-4-phenylpyridinium (MPP⁺), and NMDA-induced damage and also lessened the malonate-induced rise in free radical production in the rat brain [Schulz et al., 1995b]. Available evidence from other models points to S-PBN having poor brain penetration [Marklund et al., 2001b and Dehouck et al., 2002], which raises questions as to why it appears to be acting as a radical trapping agent within the brain. One explanation may be that the neurotoxins were injected via a probe, thereby damaging the blood-brain barrier. The mechanism of action of nitrone compounds is discussed further in Section 5.

The same laboratory has also studied the effect of S-PBN on the dopamine neurotoxin MPP⁺ in the rat and again found a protective effect of the nitrone compound [Fallon et al., 1997]. Because MPP⁺ is injected via intracerebral injection, this also might be expected to damage the blood-brain barrier.

[Guidetti & Schwarcz, 1999] examined whether PBN would protect against striatal damage produced by -3-hydroxykynurenine, which is a metabolite of tryptophan and is known to be a generator of free radicals. S-PBN was found to be an effective neuroprotectant when this neurotoxin was injected directly into the cerebral tissue.

One neurotoxic compound that has been shown to produce free radicals in cerebral tissue when injected systemically is the recreationally used amphetamine derivative 3,4-methylenedioxymethamphetamine (MDMA or ecstasy). This compound has been found to produce free radicals in the striatum and also produce a subsequent degeneration of hydroxytryptamine or serotonin (5-HT) nerve endings [Colado et al., 1997]. PBN administration attenuated the free radical production [Colado et al., 1997] and markedly decreased the degree of damage to 5-HT nerve endings as measured by both binding of [³H]-paroxetine to 5-HT nerve endings and loss of tissue 5-HT content [Colado & Green, 1995, Colado et al., 1997 and Yeh, 1999].

3. MDL 101,002, 2,2,6,6-tetramethylpiperidine-N-oxyl, and neuroprotection

While there have been a variety of nitrone-derived spin traps synthesized (Fig. 4), published data on the biological activity of these compounds are scarce with the exception of PBN. Of the other compounds examined, probably the most investigated is the cyclic nitrone MDL 101,002, which is one of a series of cyclic nitrone spin traps that have been synthesized [Thomas et al., 1996]. All the cyclic compounds examined in this study were more potent as in vitro inhibitors of lipid peroxidation than PBN, with the unsubstituted variant MDL 101,002 being ~8-fold more potent than PBN. In general, the potency correlated with lipophilicity [Thomas et al., 1997 and Craig et al., 1997].

MDL 101,002 has been found to be neuroprotective in both transient and focal ischemia models in rats. In a reperfusion model, administration of the compound at the start of the reperfusion period following a 3-hr occlusion provided a dose-dependent neuroprotection against ischemic damage of up to 70% [Johnson et al., 1998], and protection of up to 90% was observed in a mixed permanent/transient distal MCAO model. A reduction in damage of 40% was also observed in a permanent MCAO model when the drug was given 30 min after the start of the occlusion [Johnson et al., 1998].

MDL 101,002 is also efficacious in a rat embolic stroke model when administered 5 min after embolization, both decreasing the volume of ICH and improving the behavioral deficit score related to vehicle-treated animals [Hu et al., 1999].

Finally, administration of MDL 101,002 has been found to produce dose-dependent protection against malonate-induced striatal lesions and also significant protection against the depletion of dopamine and its metabolite induced by the neurotoxin 1,methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) [Mathews et al., 1999].

Some other compounds have been studied in models of cardiac ischemia and therefore will not be discussed here. However, TEMPO was studied in a rabbit model of embolic stroke and is discussed in Section 2.2.

4. Disodium 2,4-disulfophenyl-N-tert-butylnitrone and animal models of stroke

4.1. Effect of disodium 2,4-disulfophenyl-N-tert-butylnitrone in transient focal ischemia models in rats

The first report of the neuroprotective effect of disodium 2,4-disulfophenyl-N-tert-butylnitrone (NXY-059) in a rat model of transient focal ischemia was that of [Kuroda et al., 1999] who compared its effect with PBN. The model used the insertion of an intraluminal filament to occlude the MCA for 2 hr. The first experiments established the dose-response relationship of NXY-059 in this stroke model. NXY-059 was administered 1 hr following a 2-hr MCAO. Rats were injected with a loading dose of NXY-059 (0.3, 3.0, or 30 mg/kg, respectively) followed immediately by a 24-hr i.v. infusion (0.3, 3.0, or 30 mg/kg/hr). NXY-059 produced dose-dependent neuroprotection (Fig. 5) when damage was assessed histologically following 48-hr reperfusion. NXY-059 (3.0 mg/kg/hr) was markedly more effective than PBN given at an equimolar infusion dose (bolus: 1.4 mg/kg and 1.4 mg/kg/hr) because this dose of PBN was not neuroprotective (Fig. 5). NXY-059-treated rats also displayed fewer neurological deficits than the control or PBN-treated rats both 24 and 48 hr after reperfusion.

(5K)

Fig. 5. Effect of NXY-059 and PBN in a transient MCAO model. Results show data recalculated and redrawn from results published in [Kuroda et al., 1999]. The PBN dose was 1.4 mg/kg, which is equimolar to a NXY-059 dose of 3 mg/kg. Individual animal infarct size shown together with horizontal bar illustrating the median value of the group.

The main findings of the [Kuroda et al., 1999] study were supported by the observations of [Sydserff et al., 2002]. This group administered doses of NXY-059 of 3, 10, or 30 mg/kg/hr by i.v. infusion starting 5 min after the end of the 2-hr occlusion. No loading dose was given. The dose-dependent degree of neuroprotection reported by Sydserff et al.was similar to that observed by Kuroda et al. (Fig. 6). The increased efficacy seen by Kuroda et al. at low doses of NXY-059 may have been due to the initial loading dose used in that study and which was not used by Sydserff et al. The data on plasma concentrations presented in Fig. 6 have all been extrapolated from dose-plasma concentration values obtained in the Sydserff et al. study and may thus be slightly in error in the Kuroda et al. projections.

(6K)

Fig. 6. Dose-dependent neuroprotective effect of NXY-059 in models of transient focal ischemia as observed in the studies of [Kuroda et al., 1999] and [Sydserff et al., 2002].

The [Kuroda et al., 1999] study also included investigation into the therapeutic time window in the transient MCAO model. NXY-059 provided effective neuroprotection when administered 5 hr after MCAO (3 hr after the start of reperfusion) and the protective effect was almost significant 8 hr after occlusion. Finally, this group also made a brief study of the duration of protection afforded by NXY-059 following the ischemic insult. Similar protection was observed when outcome was measured either 2 or 7 days after the ischemic episode, indicating that damage was probably being prevented rather than the drug slowing the development of damage.

4.2. Effect of disodium 2,4-disulfophenyl-N-tert-butylnitrone in permanent focal ischemia models in rats

A major study on the effect of NXY-059 in the permanent MCAO model has recently been reported [Sydserff et al., 2002]. Doses of 30, 50, and 70 mg/kg/hr for 24 hr were administered by s.c. osmotic minipumps circumventing the problem of keeping an indwelling i.v. line patent for 24 hr, although it increased the problem of delivering the drug reliably from the moment of implantation. As before, an appropriate loading dose (30, 50, or 70 mg/kg, respectively) was given at the start of the treatment period. NXY-059 produced a dose-dependent neuroprotection in permanent MCAO. Protection was demonstrated not only when measured as a total volume of damage but also when measured in terms of cortical and subcortical damage (Fig. 7).

(24K)

Fig. 7. The relationship between the plasma free concentration of NXY-059 and the degree of neuroprotection in a rat model of permanent focal ischemia. Reproduced from [Sydserff et al., 2002], with permission of Stockton Press.

The plasma concentration at 24 hr appeared to be linearly related to dose, and results suggested a linear relationship between dose and neuroprotection (Fig. 7). Extrapolation of these values (given the linear relationship between dose and plasma concentration) suggested that around 80% neuroprotection could be achieved when the plasma "free" concentration was around 150 mol/L at 24 hr of infusion (Fig. 8) [Sydserff et al., 2002].