Headaches Occurring During Sleep

Secondary causes of nocturnal headaches include drug withdrawal, temporal arteritis, sleep apnea, oxygen desaturation, pheochromocytomas, primary and secondary neoplasms, communicating hydrocephalus, subdural hematomas, subacute angle-closure glaucoma, and vascular lesions. Migraine, cluster, hypnic, and chronic paroxysmal hemicrania are other primary headaches that can cause awakening from sleep. Hypnic headaches only occur during sleep. Migraine typically has associated symptoms and very uncommonly only occurs during sleep. Cluster headaches have autonomic symptoms and may occur during the day as well as during sleep. Chronic paroxysmal hemicrania occurs both during the day and at night, lasts for less than 30 minutes, and occurs 10 to 30 times a day.

Obstructive Sleep Apnea

Snoring and excessive daytime sleepiness are the most common symptoms of obstructive sleep apnea (OSA). Morning headaches are three times more common in those with OSA than in the general population and may occur in 36% of those with OSA.

Sleep Bruxism

Bruxism during sleep occurs in up to 20% of the population as determined by visible tooth wear. Sleep bruxism is most common in children between the ages of 3 and 12 years and in adults between the ages of 19 and 45 years. Individuals are unaware of this behavior, which produces audible sounds in about 20% of episodes. Some patients report morning jaw discomfort and tension-type headaches that improve as the day goes on. Causes include dental factors, psychologic and emotional factors, and systemic disorders. Occlusal bite splints protect against damage. Medications that can be helpful at bedtime include propranolol, L-dopa, bromocriptine, and, for acute exacerbations, diazepam.

Sleep Deprivation and Sleeping In

Lack of sleep can trigger migraine and tension-type headaches. In one study, 38.8% of medical and dental students reported headaches due to sleep deprivation. Some migraineurs find sleeping later than their usual time of awakening is a trigger.

Sleep to Relieve Migraine

Many migraineurs obtain relief from acute attacks by sleeping. In one study, 28% could terminate a migraine with sleep.

Parasomnias and Migraine

Somnambulism occurs in 28% of children with migraine and 5% of controls. Children with migraine also have a greater incidence of night terrors.

Exploding Head Syndrome

Episodes of exploding head syndrome awaken people from sleep with a sensation of a loud bang in the head, like an explosion. Ten percent of cases are associated with the perception of a flash of light. The episodes take place in healthy individuals during awakenings without evidence of epileptogenic discharges.

Gene Therapies for Parkinson's Disease

BACK

INTRODUCTION

Neurodegeneration normally occurs in the brain during development and aging. However, diseases such as Parkinson's, Alzheimer's, Lou Gehrig's and Huntington's. in which neuronal death occurs at an accelerated rate, have caught the attention of both scientists and the public, because of the devasting outcome of these diseases on the individual and their high cost to society. Current therapies for neurodegenerative diseases do not offer cures, and are ameliorative rather than restorative. However, neuroscience research has led to the identification of genes involved in neuronal differentiation, growth and survival, as well as to genes involved in genetic forms of these diseases. Consequently, gene transfer into the CNS is an emerging field likely to lead to novel therapeutic approaches for neurodegenerative diseases in which disease progression is retarded or even reversed.

The pathology of Prakinoson's disease is characterized by the loss of dopamine (DA) neurons in the substantia nigra pars compacta, with consequent depletion of DA in the striatum. With this pathology being focal and toxins being available that selectively kill DA neurons, well-characterized animal models of Parkinson's disease have long been available, and have been used in the evaluation of gene therapy strategies. Such studies represent several approaches, including delivery of genes to increase the synthesis of DA or DA receptors and genes that code for neuroprotective and growth-promoting molecules.

PARKINSON'S DISEASES

Parkinson's disease (PD) affects approximately 1% of people over the age of 50, with nearly 500,000 patients in the United States. The cardinal symptoms of PD include bradykinesia or akinesia, rigidity, resting tremor, and postural instability. Rigidity is clinically detected as increased resistance to passive movement and is caused by tonic contraction of muscles. Resting tremor is caused by a rapid alternating contraction of opposing muscle groups, often observed in the forearm and hand as a "pill rolling" tremor. Life expectancy in PD patients is similar to that of age-matched controls; however, even with medical therapy, 80% of patients are disabled within 10 yr of diagnosis.

The etiology and pathogenesis of PD are unknown. Although the majority of cases are sporadic, are large kindred with autosomal dominant inheritance of PD has shown linkage to chromosome 4q21-q23. The causative gene, once identified, should provide insight into the pathogenesis of both hereditary and sporadic cases. DA neurons may be in PD because of a combination of oxidative stress, excitotoxicity, mitochondrial abnormalities, and calcium toxicity. Reactive oxygen species, generated from auto-oxidation and metabolism of DA, can cause DA neuron death by damaging cell membranes, proteins, and DNA. Excitatory amino acids such as glutamate can increase intracellular calcium (Ca) concentrations. Mitochondrial abnormalities, which have been observed in patients with PD, can impair energy production and cause loss of ion gradients, including Ca. Elevated intracellular Ca may lead to cell death through activation of Ca-dependent proteases.

Current pharmacologic therapies for PD are aimed at ameliorating symptoms, primarily through augmenting DA neurotransmission. L-DOPA (1-dihydroxyphenylalanine), in conjunction with the peripheral decarboxylase inhibitor carbidopa, has been the mainstay of PD therapy for decades. L-DOPA crosses the blood-brain barrier and is converted to DA by DOPA decarboxylase in remaining DA neurons, non-DA neurons, and glia. DA levels can also be augmented by inhibition of DA metabolism with deprenyl, a monoamine oxidase B inhibitor, or by inhibition of catechol-O-methyl-transferase. DA agonists, such as bromocriptine, directly activate D1 and D2 receptors on striatal neurons. Other drugs that improve symptoms such as anticholinergics and amantadine, indirectly compensate for reduced DA levels by altering neurotransmission in the basal ganglia.

L-DOPA and other therapies do not prevent the continued degeneration of DA neurons, and side effects, such as dyskinesias and the on-off phenomenon occur in the majority of patients. As a result, additional therapies are being studied in animal models and PD patients. Experimental therapies include pallidotomy, a surgical procedure that creates a compensatory lesion in basal ganglia circuitry. Patients with unilateral pallidotomy show reductions in L-DOPA-induced dyskinesias and on-off fluctuations, as well as some improvement in tremor, rigidity, and bradykinesia, but not postural instability. Transplantation of DA neuron containing fetal ventral mesencephalon to the striatum has improved symptoms in some patients, and may work by local, continuous release of DA in the striatum and reconstruction of some circuitry. Transplantation of cells genetically modified to produce L-DOPA or DA, or direct modification of host tissue with genes encoding L-DOPA or DA synthesizing enzymes, are alternative strategies to transplantation of fetal tissue that are being investigated in animal models of PD. Molecules with neuroprotective potential, including antioxidants, antagonists of NMDA or other excitatory amino acid receptors, inhibitors of nitric oxide synthase (NOS), and neurotrophic factors, are also under investigation.

ANIMAL MODELS OF PARKINSON'S DISEASE

Several well-characterized animal models of parkinson's disease are available for evaluating novel therapies. Transection of the medial forebrain bundle, which contains nigrostriatal and mesolimbic DA axons, causes a reduction of DA phenotypic markers and cell death in the substantia nigra and ventral tegmental area. Injection of the neurotoxin 6-hydroxydopamine into the striatum, medial forebrain bundle, or substantia nigra results in a selective loss of DA neurons. 6-OHDA is an analog of DA that is concentrated in DA neurons through uptake by the high-affinity DA transporter. 6-OHDA undergoes auto-oxidation, generating hydroxyl radical, hydrogen peroxide, and superoxide anion, molecules that damage lipids, proteins, and DNA, and lead to cell death. Injection of 6-OHDA into the rat striatum produces a progressive loss of DA neurons over several weeks; injection into the medial forebrain bundle or substantia nigra produces a rapid loss of DA neurons. MPTP causes parkinsonian symptoms in humans, nonhuman primates, and mice, following its oxidation to MPP+ by MAO-B. MPP+ enters DA neurons by high-affinity uptake through the DA transporter and interferes with ATP production by inhibiting complex I of the mitochondrial electron transport chain.

Unilateral chemical or physical lesions of the nigrostriatal pathway result in an imbalance in the levels of DA and DA receptors between the two striatae, with behavioral sequelae. For example, DA stores are depleted and striatal postsynaptic DA receptors are upregulated on the lesioned side. Animals exhibit rotational behavior in response to amphetamine and DA agonists, such as apomorphine, that is readily quantifiable. In addition, unilaterally lesioned animals exhibit deficits in contralateral limb use in several spontaneous behaviors. MPTP-lesioned nonhuman primates exhibit symptoms similar to humans with parkinson's disease, including bradykinesia or akinesia, resting tremor, and rigidity.

In addition to chemical or physical lesions of DA neurons, another model of Parkinson's disease is the weaver mutant mouse. Following normal DA innervation of the striatum, approx 75% of DA neurons in the substantia nigra degenerate, with a parallel loss of striatal DA. The weaver phenotype results from a point mutation in the pore region of the G protein activated inwardly rectifying potassium channel subunit, GIRK2. Potassium channels containing the weaver GIRK2-subunit are not selective for potassium, but also conduct sodium, which may cause depolarizatioon leading to excitotoxicity in DA neurons.

NEUROTRANSMITTER REPLACEMENT GENE THERAPIES

As DA neurons die in Parkinson's disease, the synthesis of DA is reduced. The rate limiting enzyme for DA biosynthesis is tyrosine hydroxylase, which catalyzes the conversion of the amino acid L-tyrosine to L-DOPA. This reaction requires the cofactor tetrahydrobiopterin, the synthesis of which is rate-limited by GTP-cyclohydrolase I (GC). L-DOPA is rapidly converted to DA by AADC. DA is subsequently concentrated into synaptic vesicles by the vascular amine transporter. DA acts as an end product inhibitor of TH by competing with BH4 for the cofactor binding site. In Parkinson's disease, orally administered L-DOPA is absorbed into the blood stream, crosses the blood-brain barrier, and is converted to DA by AADC in remaining DA nerve terminals and in other neurons and glia. The fluctuations in response to oral L-DOPA to the striatum, and thus the irregular production of DA. Replacement of the TH gene, perhaps along with other DA biosynthetic enzymes, by gene therapy approaches may lead to continuous local production of DA. This strategy may not produce the motor fluctuations and peripheral side effects with orally administered L-DOPA.

Ex Vivo Enzyme Replacement Gene Therapy

A variety of cell lines and primary cells, including fibroblasts, myoblasts, and astrocytes, are readily genetically modified in vitro by transfection methods or by viral vectors. The TH gene has been introduced into a variety of cells, including the fibroplast cell lines 208F, NRK-49F, NIH3T3, and CVI; a rat endocrine cell line; rat primary fibroblasts; rat primary astrocytes; an immotralized human fetal astrocyte cell line, SVG; and neural cells derived from the rat ventral mesencephalon immorrtalized with a temperature-sensitive oncogene. In most studies, retroviral vectors with the TH cDNA driven by the cytomegaloviral (CMV) promoter or a retroviral long-terminal repeat (LTR) promoter were utilized to genetically modify the cells, although calcium phosphate transfection and infection with a defective herpes simplex virus-1 (SV-1) vector have also been used. Production and release of L-DOPA by TH-expressing fibroblasts in vitro occurred only in BH4 is added to the medium, consistent with the lack of expression of GC in fibroplasts. Retroviral transduction of both the TH and GC genes into fibroplasts resulted in synthesis of BH and L-DOPA. Unlike fibroblasts, primary astrocytes expressing TH release L-DOPA in the absence of exogenous BH4. The amount of L-DOPA released by cells in vitro in the medium is much higher than that contained in cells, suggesting that L-DOPA is constitutively released by fibroblasts. Cocultured TH-expressing fibroblasts and AADC-expressing fibroblasts produced and released DA in vitro in the presence of BH4, indicating that L-DOPA and DA cab readily diffuse in and out of fibroblasts.

Transplantation of cells genetically modified to express TH into the 6-OHDA denervated striatum of rats reduced opamorphine-induced rotation behavior, but transplantation of unmodified cells or cells expressing a reporter gene failed to do so. These results suggest that grafted cells produce and release L-DOPA in the presence of host-derived BH4 L-DOPA is subsequently converted to DA by host AADC, and the increased DA reduces DA-receptor supersensitivity on the lesioned side, thus reducing apomorphine-induced rotation behavior. L-DOPA and DA levels were increased by TH-expressing NIH3T3fibroblasts, as measured by in vivo microdialysis. However, transplanted NRK-49F fibroblasts expressing TH did not synthesize detectable L-DOPA unless BH4 was perfused into the striatum. No DA and only low levels of L-DOPA are observed following transplantation of TH-expressing primary fibroblasts. Thus, endogeneous BH4 levels in the DA-depleted striatum might, in some situations, be inadequate to activate TH. Transplantation of fibroblasts modified to express both TH and GC resulted in measurable L-DOPA and DA production and reduce apomorphine rotation; however, in this study, fibroblasts expressing only TH or unmodified firboblasts reduced apomorphine rotation to the same extent. Retroviral-mediated TH-and GC-transgene expression is rapidly downregulated in vivo. Only a small percentage of grafted astrocytes and fibroblasts were TH-immunoreactive 2 wk after grafting, and L-DOPA production decreased over 95% 2 wk after grafting of TH-and GC-expressing fibroblasts. Additonally, the behavioral improvement observable at 2 wk is reduced at 6 - 8 wk.

Table 1 : Gene Therapy in Animal Models of Parkinson's Disease

Paradigm Biological effect Refs

I. Ex vivo Gene Therapy Approaches

Neurotransmitter Systems

Neurotrophic Factors
MPP + Rat I - BDNF Protection of soma 63
MPP + Fibroblast-BDNF ? SN DA levels 64
6-ODHA, BHK-GDNF Sprouting of fibers 68
partial Astrocyte-BDNF Improvement in rotation 67
6-ODHA, Fibroblast-BDNF Protection of soma, fibers 66
striatal
6-ODHA, Fibroblast-GDNF Protection of soma 69
MFB Fibroblast-BDNF No effect 65
Unlesioned Fibroblast-BDNF ? DA turnover, sprouting 65
Chromaffin/NGF-astrocyte cografts Protective/stimulatory 71.72
Chromaffin/NGF-fibroblast cografts ? Survival/differentiation 73.74
PC12cells+NGF Protective ( survival) 76.77
II.In vivo Gene Therapy Approaches
Neurotransmitter Systems
6-OHDA HSV-TH ?Apomorphine rotation 27
6-OHDA AAV-TH ?Apomorphine rotation 28
6-OHDA Lipofectin-TH ?Apomorphine rotation 30
6-OHDA Ad-TH ?Sensorimotor asymmetry 29
Normal Ad-D2R ? Receptor density 31

Neurotrophic Factors
6-OHDA Ad-GDNF Protection of soma 81,82,
86
6-OHDA AAV-GDNF ? DA, improved behavior, 85
? DA transporters
6-OHDA AAV-GDNF Protection of soma 87

Ex vivo gene therapy approaches also have been applied to a small number of non-human primates. A temperature-sensitive immortalized neural cell line derived from rat embyonic d 14 ventral mesencephalon was modified to express TH with a retrovirus vector. These cells were grafted into the striatum of two Macaco mulata monkeys immunosuppresed with cyclosporin, and previously rendered hemi-parkinsonian with unilateral intracarotid infusion of MPTP. Apomorphine rotation behavior improved in both monkeys, and grafted cells survived and expressed TH, as demonstrated by immunocytochemistry. Autologous primary fibroblasts retrovirally transduced with a TH gene under the Moloney murine leukemia virus (MMLV) LTR and grafted into the striatum of MPTP-lesioned monkeys expressed TH 4 mo after grafting, as demonstrated by immunocytochemistry and in situ hybridization.

In vivo Enzyme Replacement Gene Therapy

An alternative approach to grafting cells modified ex vivo is to genetically modify host tissue by injecting vectors encoding the transgene of interest to the CNS. Long-term behavioral effects have been observed in the 6-OHDA-lesioned rat following injectiono f an HSV-1 amplicon vector and an adeno-associated virus (AAV) vector encoding TH. HSV-TH improved apomorphine rotation behavior 65% from 2 wk to 1 yr after vector injection; HSV-LacZ, encoding the reporter enzyme - galactosidase, was without effect. DA release induced by high extracellular potassium was 300% greater in HSV-TH, compared to HSV-LacZ-or vechicle-injected rats, as measured by in vivo microdialysis at 4-6 mo. Despite considerable behavioral improvement, transgene expression was limited. Six to 16 mo after injection of HSV-TH, 5-300 TH-immuno reactive cells were observed in the striatum. TH mRNA was amplified by reverse transcription polymerase chain reaction from 3 of 10 rats 1 mo after HSV-TH. Injection of AAV-lacZ and AAV-TH resulted in transgene expression for at least 3 and 4 mo, respectively, as detected by X-gal histochemistry and TH immunocytochemistry, respectively. In 6-OHDA-lesioned rats, AAV-TH reduced apomorphine rotation behavior approx 35% at 5 and 10 wk after injection, compared to AAV-LacZ and vehicle-injected rats. An adenovirus-encoding TH decreased sensorimotor asymmetry in lesioned rats. Striatal injection of a lipofectin-plasmid DNA complex encoding TH under the simian viral (SV40) early promoter reduced apomorphine turning by 46% 3 - 15 d after injection in the 6-ODHA lesioned rat; injection of plasmid alone or lipofectin alone had no effect. Only in lipofectin TH plasmid-injected rats were TH-immunoreactive cells and TH mRNA amplified by RT-PCR observed in the striatum. In monkey, injection of AAV-TH or AAV-LacZ resulted in expression in neurons for at least 3 mo. At early time-points, 14,000 - 31,000 - ßgalactosidase-expressing cells were observed per injection site.

Another in vivo gene therapy approach applicable to Parkinson's disease is to increase dopamine D2 receptor expression in the striatum, which is known to be reduced in late-stage Parkinson's disease. An adenoviral vector encoding the D2R, under control of the CMV promoter injected into the normal rat striatum, increased D2R density in a focal area round the injection site, as demonstrated by receptor autoradiography with [3H]- spiperone ligand.

GENE THERAPIES WITH NEUROTROPHIC FACTORS

DA Neurotrophic Factors

Early studies of DA neurons in culture showed that neuronal growth and morphology can be differently influenced by signals from glial cells in target vs nontarget brain regions. Subsequently, the plasticity and potential for regeneration and sprouting inherent to DA neurons in adult brain were realized from transplantation and injury studies. These studies prompted the search for specific DA neurotrophic factors. To date, more than 20 neurotrophic factors have been reported for DA neurons. These include members of several growth factor families operating through different intracellular signaling mechanisms, including the TGFß- superfamily, the neurotrophins, cytokines, and mitogenic growth factors. Some factors, such as glial cell line derived neurotrophic factor and brain-derived neurotrophic factor, act directly on DA neurons in vitro; others, such as fibroblast growth factor, mediate effects on DA neurons through astrocytes.

The identification of factors with potent DA trophic activities in vitro led to the concept of using these as therapeutic agents for Parkinson's disease, as reviewed previously. Results of studies utilizing rat, mouse, and nonhuman primate models of Parkinson's disease, in which large amounts of recombinant DA neurotrophic factor proteins have been injected or infused in to the brain, support the therapeutic efficacy of these substances. For example, GDNF, the most potent DA neurotrophic factor yet identified, protects DA neurons from death and loss of phenotypic markers, and ameliorates parkinsonian behaviors in several animal models of Parkinson's disease. In the MPTP-treated mouse, GDNF partially prevented MPTP-induced decline in striatal DA levels, DA cell number, and TH-IR fiber density. GDNF-treated mice also exhibited increased motor behavior. In rat, GDNF maintained the DA phenotype against 6-ODHA-induced damage. GDNF infused into the substantia nigra 5 wk after 6-ODHA completely reversed apomorphine-induced rotation, increased the number of TH-IR neurons, and normalized DA levels in the substantia nigra. Repeated injections of GDNF just above the substantia nigra in rats with a progressive DA lesion resulting from striatal injection of 6-ODHS also protected DA neurons from cell death. In the hemi-parkinsonian Rhesus monkey, GDNF injected into substantia nigra, caudate nucleus, or lateral ventricle improved bradykinesia, rigidity, and tremor. Moreover, unilaterally injected GDNF protected against the effects of MPRP bilaterally, increasing DA somal size and the density of TH-IR fibers. DA levels in the substantia nigra also were increased in hemi-Parkinsonian monkeys after an intraventricular injection of GDNF. In addition, GDNF injection into normal Rhesus monkey caudate upregulated the DA system. In rats with a physical lesion of the medial forebrain bundle, repeated injections of GDNF near the substantia nigra increased DA cell survival to 85%, compared to 53% in control injected rats. Although the reported effects of GDNF are remarkable, other DA neurotrophic factors, such as BDNF, FGF, and NT-3, also have effects on DA neurons in vivo, BDNF increased DA turnover and DA neuronal activity, and both BDNF and NT-3 ameliorated the development of parkinsonian behaviors in the rat. Several mitogenic factors, including EGF, aFGF, and bFGF, also promoted recovery in the 6-OHDA-lesioned rat and the MPTP-treated mouse.

Although the therapeutic efficacy of DA neurotrophic factors is well founded in animal models of Parkinson's disease, there are practical considerations to their application in humans. Because Parkinson's disease is progressive, it is likely that long-term trophic support for the diseased DA neurons will be required. Neurotrophic factors are labile substances that are unable to cross the blood - brain barrier in significant amounts. Therefore, their therapeutic use for Parkinson's disease will require development of methods for delivering these factors continuously to the DA neurons in the substantia nigra in a manner that is safe, minimally invasive, and that does not elicit effects on other types of neurons. Alternatively, pharmacological approaches targeted to neurotrophic factor receptors may be developed; however, these would be expected to affect all cell types that express the receptor. Repeated injections of recombinant neurotrophic factors into the human brain are unlikely to be practical, and are likely to elicit deleterious side effects over the long term. In this respect, clinical trials in which neurotrophic factors were administered in large amounts in the periphery were stopped because of unanticipated, intolerable side effects. In addition, one patient with Alzheimer's disease experienced the side effects of weight loss, pain, and sleep disturbances following intraventricular infusion of nerve growth factor. Gene therapies for delivering neurotrophic factors to the CNS have the potential to circumvent or even eliminate the drawbacks of recombinant protein therapies. By transferring neurotrophic fact genes to the CNS, there is the potential of producing these vital substances in a continuous manner, or even in a regulatable manner through the use of regulatable promoters. Moreover, expression of a factor may ultimately be confined to a specific cell type through the use of a cell-specific promoter.

Exo Vivo Neurotrophic Factor Gene Therapy

The first studies applying gene therapy with neurotrophic factors to Parkinson's disease used ex vivo gene therapy, implanting various types of cells into the striatum following retroviral transduction with neurotrophic factor genes in vitro. Cell lines, as well as primary fibroblasts, myoblasts, and astrocytes, including human astrocytes, are amenable to transfection and retroviral transduction with neurotrophic factors. Rat-I fibroblasts are primary rat astrocytes secrete bioactive BDNF following retroviral infection, as shown by an in vitro bioassay of embryonic DA neurons. Implantation of BDNF-secreting cells in animal models of Parkinson's disease ameliorate the effects of neurotoxins acting through either oxidative stress or mitochrondrial toxicity. For example, BDNF-secreting fibroblasts grafted near the rat substantia nigra protected approx 80% of the DA neurons from MPP+ injected into the striatum, compared to only 35% protection in rats grafted with control fibroblasts. In addition, DA levels in the substantia nigra were increased, although DA turnover remained unaffected. BDNF-secreting fibroblasts grafted into the striatum of unlesioned rat increased DA turnover. Implantation of BDNF-secreting fibroblasts 1 wk prior to 6-ODHA lesion completely protected DA cell bodies in the substantia nigra for up to 3 wk, and partially protected DA terminals in the rat striatum, as determined by binding of 3H-mazindole to DA uptake sites. In contrast, implantation of BDNF-secreting fibroblasts, either into the striatum or substantia nigra did not protect DA neurons from 6-ODHA injected into the medial forebrain bundle. Astrocytes have also been used to deliver BDNF tot he 6-ODHA lesion of the substantia nigra, reduced motor asymmetry as evaluated by apomorphine-induced rotational behavior in the absence of effects on DA cell survival or fiber density in the striatum.

Ex vivo gene therapy with GDNF has been tried with rat fibroblasts and encapsulated baby hamster kidney cells. Following calcium phosphate transfection of a GDNF expression plasmid, BHK cells were encapsulated in a polymer. Bioactive GDNF was secreted from these capsules, as judged by increased survival and neurite outgrowth of embryonic DA neurons grown in medium conditioned by the capsules. In rats partially lesioned with 6-OHDA, implantation of capsules into the striatum induced in growth of DA fibers into the capsules; however, behavioral improvement was not apparent and effects on DA cell survival were not studied. In another study, rat firboblasts transduced with the human GDNF gene and grafted near the substantia nigra protected DA neurons from a medial forebrain lesion with 6-OHDA, as judged by counting neurons positive for c-ret, a component of the GDNF receptor.

Another ex vivo gene therapy approach is that of combining genetically engineered cells with grafts of other cell types. For example, adrenal chromaffin cells are catecholaminergic cells that secrete high levels of DA when grown in the absence of glucocorticoids. Although chromaffin cells were for a period of time considered to be an ideal replacement for DA neurons, and were experimentally used in human with Parkinson's disease, these cells survive poorly when grafted into the brain. However, if provided with a source of NGF, chromaffin cell survival in the brain is improved. Taking this observation to the gene therapy level, rat chromaffin cells cografted with astrocytes or primary fibroblasts retrovirally transduced with the NGF gene improved survival and neuronal differentiation of the chromaffin cells. In addition, chromaffin cells cografted with NGF-astrocytes improved amphetamine-induced rotational behavior in the 6-OHDA-lesioned rat. Polymer-encapsuled BHK fibroblasts, transfected with an NGF construct under control of the metallothionein promoter and transplanted into the rat striatum, increased grafted chromaffin cell survival 20-fold from young and old donors, and improved apomorphine rotation behavior. Futhermore, chromaffin cell survival and apomorphine rotation behavior were improved with striatal, but not intraventricular, implants of the polymer-encapsulated NGF-BHK cells. Neuronal differentiation and survival of PC12 pheochromocytoma cells grafted into the 6-OHDA-lesioned rat striatum were also improved by transducing the cells with an NGF-retrovirus or NGF under control of a zinc-inducible metallothionein promoter.

Grafting human fetal mesencephalon is an experimental approach presently being tried in Parkinson's disease patients. Although reports of improved symptoms in these patients are encouraging, the survival of DA neurons in these grafts is only 5-6%. In rat, providing exogeneous BDNF to fetal mesencephalic grafts enhanced DA neuronal survival and process outgrowth. Similarly, injection of 4.5 µg GDNF every third day for 3 wk adjacent to fetal mesencephalic grafts increased the survival of DA neurons twofold, increased the density of TH+ fibers 50 - 100%, and enhanced graft function based on amphetamine-induced rotation behavior. No one has yet applied gene therapy to improve survival of grafted DA neurons, although this is an interesting possibility.

In Vivo Neurotrophic Factor Gene Therapy

In vivo gene therapy with neurotrophic factors for Parkinson's disease is an area ripe for investigation. Several classes of viral vectors are being investigated as means for delivering neurotrophic factor support for DA neurons, including Aav, Ad, and HSV-1; however, few publications on this topic have appeared to date. In our studies, we have observed a significant effect of an adenovirus harboring human GDNF on protecting DA neurons from degeneration. Ad-GDNF was injected either just above the rat substantia nigra or into the striatum, 1 wk prior to a progressive 6-ODHA striatal lesion. At 6 wk after the lesion, the number of surviving DA neurons was increased approximately threefold in rats injected with Ad-GDNF, compared to rats injected with Ad-LacZ or an Ad-mutant GDNF lacking bioactivity or rats that remained untreated. In addition, these studies showed that nanogram quantities of GDNF were produced in the injection site when 3x107 plaque-forming units of Ad-GDNF were injected. This level of GDNF is well above that required to activate GDNF receptors. Injection of Ad-GDNF into the striatum in this model also resulted in improvement in amphetamine-induced rotation at 12 d after 6-OHDA. Using a variation of the progressive lesion model, in which 6-OHDA was injected bilaterally, During and colleagues injected AAV harboring the rat GDNF gene unilaterally into the striatum. The AAV-GDNF treated rats, but not the AAV-LacZ injected rats, developed asymmetry in response to amphetamine and apomorphine, and showed improved motor abilities on several behavioral tasks. In addition, in vivo microdialysis at 12 wk after vector injection showed potassium and nomifensine - induced increases in DA in the AAV-GDNF group, suggesting an enhancement of DA neurotransmission in remaining DA terminals. The same vector was also injected into the caudate of two African Green monkeys partially lesioned with MPTP. ß- CIT spectroscopy to image DA transporters showed small increases of 9 and 19% over values obtained before the therapy in the two monkeys, again suggesting a protective effect on DA terminals. Similar effects have been reported for Ad-GDNF and AAV-GDNF. HSV vectors have not yet been specifically applied to Parkinson's disease. However, Federoff and colleagues have injected as HSV vector harboring LacZ under control of the TH promoter into the striatum. Retrograde transport of the vector to the SN was observed, with transgene expression specifically limited to DA neurons. This approach may be well suited to providing neurotrophic support to DA neurons in an autocrine or paracrine manner.

CONCLUSIONS

Gene therapy strategies for Parkinson's disease, aimed at replacing DA synthesizing enzymes, or rejuvenating DA neurons and slowing the progression of the disease through neurotrophic factors, have been reviewed. The latter approach is relevant to neurodegenerative diseases in general, in which increased levels of neurotrophic factors may not only slow the disease process, but may stimulate regeneration or sprouting of remaining neurons. Further technological advances are required to realize the potential of gene therapy for Parkinson's disease. There is no vector presently available that provides long-term, stable gene expression in the brain in the absence of cytotoxic effects. Consequently, new generation vectors need to be developed that not only lead to stable gene expression, but that also minimize hot cellular and humoral responses. For the current state of the art of vectors, the reader is referred to other chapters in this volume and to several recent reviews. It is imperative that vectors be shown to be safe and efficacious in nonhuman primate brain, and, ultimately, in the human brain.

The cellular complexity of the CNS provides another challenge. Vectors harboring genes driven by cellular promoters have the potential of expressing a transgene specifically is one phenotypically defined cell population. In the case of neurotrophic factor therapies for Parkinson's disease, this cell could be the diseased DA neuron itself, although it is not known whether neurotrophic factors acting in an autocrine manner are effective in CNS neurons. The neurotrophic factor could be expressed in DA target neurons, such as enkephalin-or somatostatin-synthesizing striatal neurons, where it would be released, taken up by DA terminals, and retrogradely transported. In this respect, several DA neurotrophic factors are retrogradely transported. In this respect, several DA neurotrophic factors are retrogradely transported. In this respect, several DA neurotrophic factors are retrogradely transported by DA neurons, including GDNF, BDNF, and bFGF. Trophic factor expression could also be targeted to astrocytes, which would in turn secrete trophic support in the vicinity of the DA neuron. It is not presently known which cellular site offers the best trophic support for CNS neurons. Another potential of viral vectors is the possibility of including regulatable promoters for the regulation of transgene expression through peripheral drug administration or through endogenous molecules whose levels are altered by disease or injury. For some diseases or injuries to the nervous system, transient increases in neurotrophic support may be optimal; for others, chronic support may be needed. However, affective or cognitive disorders could potentially result from chronically increased levels of neurotrophic factors in the CNS, and these are difficult to predict from animal studies. If such side effects of gene therapy were to occur, the inclusion of a regulatable promoter or a killer gene in the vector construct could be used to turn off expression or to kill infected cells. For these reasons, systematic studies on optimal ways to deliver and regulate genes in the CNS, expecially in primate CNS, are crucial.

In the parkinsonian brain, DA neurons degenerate over a prolonged period, and the etiology underlying this degeneration is unknown. The elucidation of genes involved in hereditary forms of PD, as well as genes that increase risk for this disease, will result in additional targets of gene therapy and novel animal models. Although the therapies reviewed here show promise, it is not known how closely current rodent and nonhuman primate models of Parkinson's disease, in which DA neurons are chemically or physically damaged, mimic idiopathic Parkinson's disease. The answer to this can come only form the application of these therapies to the human parkinsonian brain.

Hemodynamic Manipulation in the Treatment of Brain Ischemia

BACK

Brain ischemia is a common pathophysiologic entity. In some circumstances, its role in producing a neurologic deficit is unequivocal, as in embolization from an ulcerated carotid atheroma to an intracranial artery or severe systemic hypotension from cardiac asystole. However, its impact spreads far beyond such obvious clinical disorders. Unrecognized brain ischemia probably contributes to the development of neurologic deficits in many other settings, such as an expanding intracerebral hematoma or other mass lesion and severely elevated intracranial pressure (ICP).

The treatment of acute brain ischemia continues to receive considerable attention from clinical and laboratory investigators. Although a comprehensive regimen for preventing brain infarction has not been defined, much has been learned the can improve the clinical outcome. The primary objectives of treatment are to improve blood flow to the ischemic area and to increase the resistance of brain tissue to metabolic injury. Improving circulation reduces the size of the ischemic area and enhances the delivery of agents intended to increase metabolic resistance to ischemia.
This chapter focuses on methods for increasing blood flow to ischemic areas. The primary emphasis is on the treatment of incomplete focal ischemia, the kind most often encountered clinically by neurologists and neurosurgeons.

BASIC PRINCIPLES OF BLOOD FLOW

Poiseuille's law states that volume flow rate is directly proportional to the product of the pressure differential between the ends of the tube and the fourth power of the radius of the tube, and is inversely proportional to the product of the length of the tube and the viscosity of the fluid. In simpler terms, this mans that the volume flow rate can be altered by changing the driving pressure, the radius or length of the tube, or the viscosity of the fluid. Although Poiseuille's law generally applies to blood flow, it must be remembered that blood is non-Newtonian fluid and that perfusion is normally pulsatile. Thus, blood viscosity varies with the velocity of flow (29). Furthermore, brain blood flow increases with conversion from nonpulsatile to pulsatile perfusion in the presence of a constant range of perfusion pressures (48).
Circulation of blood appears to be more complex in the microvasculature than in larger vessels. Fahraeus and Lindquist (12) found that Poiseuille's law did not accurately predict the flow of blood through tubes of progreassively smaller diameter: As vessel diameter decreased to the 40 mm range (that of arterioles), blood viscosity decreased. They ascribed this phenomenon to a reduction in hematocrit. Dintenfass (11) subsequently found that when the radius of the tube was 5 to 7 mm (approximately that of capillaries), blood viscosity stopped decreasing; with further reductions, viscosity increased markedly.
Our understanding of the mechanisms that regulate intracranial circulation is incomplete. Under normal circumstances, the brain regulates blood flow in accordance with its metabolic needs (38). Regional blood flow is regulated by constriction or dilatation of precapillary arterioles; it increases or decreases in response to neuronal metabolic activity. These regional fluctuations are thought to be partly the result of changes in local pH, carbon dioxide tension, and tissue metabolite levels. Brain blood flow is also greatly influenced by changes in systemic arterial CO2 (i.e.CO2 reactivity). Under normal circumstances, brain blood flow remains within a constant range throughout the physiologic range of systemic arterial blood pressure. During severe ischemia, the mechanisms controlling blood flow break down as the arterial system reaches a state of maximal dilatation. As blood flow decreases, CO2 reactivity and autoregulation are progressively impaired and cease when the ischemic threshold is crossed (2,18). Impairment of CO2 reactivity and autoregulation may persist in varying degrees for months or years after the ischemic event whether or nor maximum vasodilatation persists (33, 34).
Methods of enhancing blood flow to ischemic areas are based on out knowledge of the principles of tubular flow and hemorrheology. It is currently possible to increase perfusion pressure, dilate conducting arteries, and alter blood viscosity. The results of some experimental studies support the validity of these approaches in the treatment of brain ischemia, but many questions ramain regarding their effectiveness in reducing infarct size and their application in specific clinical settings.

HYPERTENSIVE THERAPY

Increasing systemic arterial blood pressure to raise cerebral perfusion pressure would appear to be a logical component of the treatment of brain ischemia. Arteries in an area of focal ischemia due to proximal vascular occlusion are connected, through collateral channels that vary greatly in size and number, to arteries in the nonischemia. Despite the presence of these interconnecting vessels, perfusion pressure is lower in the ischemia area that in the nonischemic area (45). An increase in systemic arterial blood pressure would be expected to increase pressure in the collateral arterial conducting system proximal to the occluded artery and in surrounding areas, thereby increasing the pressure differential and augmenting flow into the ischemia area. Whether or not this phenomenon reduces the extent of ischemic tissue damage depends on the adequacy of the collateral circulation and the severity of the ischemia.
The effects of changes in systemic arterial blood pressure on the brain circulation in experimental ischemia have been investigated extensively, but the treatment of ischemia with hypertension has received surprisingly little attention. Waltz (49) studied the effects of varying systemic arterial blood pressure on cortical blood flow in cats undergoing acute middle cerebral artery (MCA) occlusion. In nonischemic cortex, blood flow remained relatively constant despite changes in systemic arterial blood pressure. In ischemic cortex, blood flow paralleled blood pressure at normotensive and hypontensive levels; at hypertensive levels (means arterial blood pressure to 120 mm Hg,) it increased but never reached the baseline levels of nonischemic cortex. With further increases in systemic arterial pressure, cortical blood flow decreased.
Symon et al (43) evaluated the effects of pharmacologically induced hypertension on cerebral blood flow (CBF) in baboons undergoing acute MCA occlusion. These studies clearly showed a link between the degree of ischemia and the extent of autoregulatory impairment. Autoregulation was partially preserved in ischemic area with greater than 40% of basal flow (ollateral zones) and was absent in areas with less than 20% of basal flow (core area of ischemia). Hypotension induced by exsanguination further reduced CBF in the ischemic zone, whereas pharamacologically induced hypertension increased flow. Like Waltz (49), Symon et al (43) found a limit to the favorable effects of hypertension on CBF. When the mean arterial blood pressure rose 50 mm Hg or more above normal preexisting pressure, CBF in the ischemic zone decreased.
Symon et al (43) made additional important observation during reperfusion. CBF was imparied during reperfusion after severe ischemia. CBF was also measured after restoration of normal systemic arterial blood pressure in baboons subjected to MCA occlusion followed by 2 hours of hypotension. CBF levels after restoration of blood pressure did not recover to levels observed immediately after MCA occlusion and before induction of hypotension; but CBF levels after restoration of blood pressure correlated with the severity of ischemia observed during the hypontensive period. Clearly, hypotension had an additive deleterious effect on blood flow during reperfusion. These findings confirmed the observations of other (1) that changes in the microcirculation in areas of severe ischemia may prevent adequate reperfusion. This secondary impairment in blood flow has been called the "no-reflow phenomenon" (43).
Hope et al (16) measured somatosensory evoked potentials and local CBF in baboons subjected to MCA occlusion. Increasing the mean systemic arterial blood pressure pharamacologically by an average of 40 mm Hg significantly improved the evoked potentials and local CBF. Hypertensive therapy elicited a similar response in a small group of patients with ischemic neurologic deficits after aneurysm surgery.
Denny-Brown (10) was probably the first to report that hypertensive therapy improved acute brain ischemia. Others have described similar observations, but most of the studies were uncontrolled, and the reports anecdotal. The effect of hypertensive therapy on acute focal brain ischemia has not yet been systematically evaluated in a clinical setting.
During the past few years, pharmacologically induced hypertension has be come a major component in the treatment of ischemic neurologic deficit after subarachnoid hemorrhage (SAH). The current tendency to operate on a ruptured intracranial aneurysm within hours or days of SAH has largely eliminated the potential risk of rebleding due to hypertensive therapy. The results of studies in which hypertension has been used to treat this condition have been encouraging (4,30,34). In most of these reports, hypertension was combined with hypervolemia. Beneficial effects were not achieved in some patients until very high blood pressure levels (e.g. 240 mm Hg systolic blood pressure) were reached.
Careful monitoring and manipulation of systemic arterial blood pressure are essential in patients with acute brain ischemia. Allowing systemic arterial blood pressure to decrease below initially observed levels further reduces blood flow in ischemic areas and must be avoided. Although the role of hypertensive therapy is less well defined than in patients with SAH, the weight of the evidence strongly favors its application for patients with SAH. The guidelines for its use in patients with brain ischemia due to vascular occlusion have yet to be defined.
Hypertensive therapy is not without risk. For example, pharmacologically induced hypertension may not be tolerated in patients with heart disease and could precipitate rebleeding from an untreated intracranial aneurysm. Hypertensive therapy initiated after prolonged delay could worsen brain edema or produce hemorrhage, particularlly if infarction has already occurred or if reperfusion is induced in a previously severely ischemic area, such as after dissolution of an embolism.

HYPERVOLEMIA

Total body blood volume varies with many factors including age, sex, bodyweight, fat content, activity, position, medications, and disease. Blood volume may by reduced in many persons already at risk for stroke. In patients with SAH, both red blood cell mass and total blood volume are often substantially reduced (27). These changes are thought to be partly the result of bed rest, supine duresis, negative nitrogen balance, decreased erythropoiesis, and iatrogenic blood loss. Systemic secretion of casoconstricting catecholamines that accompanies SAH is also considered an important factor in reducing intravascular volume. Most patients with acute ischemic stroke are hypovolemic at admission (13).
Maintaining intravascular volume appears to be an important factor in improving the outcome of patients with acute ischemia after a recent SAH. Hypovolemia increases the likelihood of relative or absolute systemic hypotension, which could further reduce blood flow in the ischemic area. However, the role of hypervolemia in treating brain ischemia has not yet been established. On the basis of Poiseuille's law, it is difficult to explain how hypervolemia per se would improve brain blood flow without a concomitant increase in perfusion pressure, dilatation of conducting arteries, or reduction of blood viscosity.
Some experimental studies (20) have related improved brain blood flow and outcome in ischemia to an increase in cardiac output induced by hypervolemia. In this setting, pulse pressure is augmented by the increased cardiac stroke volume. Raising pulse pressure might have favorable effects on blood flow in ischemic brain (48). However, another series of experimental studies showed that expanding intravascular volume without hemodilution did not improve cortical blood flow in areas of focal cerebral ischemia despite a significant rise in the cardiac output (51).
Decreased blood volume after SHA appears to increase the risk of symptomatic vasospasm. Solomon et al (37) found subnormal blood volume in 86% of patients with symptomatic vasospasm but in only 13% of patients with asymptomatic angiographic vasospasm. A number of clinical reports, although largely anecdotal, appear to show a considerable reduction of morbidity from vasospasm when aggressive steps are taken in increase intravascular volume (4,30,37). But there is no compelling evidence that hypervolemia is better than normovolemia in treating ischemic symptoms after SAH.
Systemic hypervolemia is unlikely to have any direct effect on blood volume in the ischemic area. expansion of blood volume in the brain microcirculation is an early and consistent response to ischemia. When perfusion pressure and blood flow drop below the normal range, the brain arterial system dilates, spontaneously producing a state of reactive hypervolemia in the ischemic area. as blood approaches the ischemic threshold, vasodilatation and volume expansion of the microcirculation are already maximal and autoegulation and CO2 reactivity are lost. Consequently, it is unclear what benefit systemic volume expansion would have on the microcirculation in and around an ischemic area of brain, except indirectly by its effect on blood pressure.

VASODILATATION

Cerebral arteries and arterioles in an ischemic area lose their autoregulatory capacity and Co2 reactivity and becomes maximally dilated (38,43). Conducting arteries in surrounding areas retain their reactivity either fully or partially depending on their proximity to the ischemic focus. These observations gave rise to the hypothesis that blood flow in the ischemic area could be anhanced by dilating the conducting collateral channels. The findings of subsequent studies, however, have not supported this hypothesis.
The early studies focused on the effects of CO2 on the brain circulation in experimental and clinical ischemia. In 1968, Lassen and Palvolyi (22) reported a further decrease in blood flow "in some brain regions in patients with acute cerebral vascular diseases." They attributed this response to a reduction in perfusion pressure in collateral vessels produced by arterial dilatation in nonischemic brain. This response was called the "intracerebral steal syndrome." They also observed an inverse response, the "Robin Hood" or "inverse steal" syndrome, where hypocapnic provoked an increase in blood flow shunted to poor ischemic areas in association with vasoconstriction and decrese in flow in rich nonischemic areas. Most subsequent reports have confirmed these observation.
Symon (41,42) and Brawley and his associates (7,8), measured the luminal pressure in arteries beyond the point of experimental occlusion of a mojor brain artery. During inhalation of CO2, arterial pressure decreased. Using a thermocouple technique, Brawaley et al (7,80 also showed a concomitant further reduction of cortical blood flow in the ischemic area.
A few studies have show improved blood flow in ischemic areas during CO2 inhalation. For example, Yamamoto et al (52) performed fluorescein angiography and measured microregional CBF in dogs undergoing occlusion of a cortical artery. When the dogs breathed 5% CO2 and 95% oxygen, collateral flow improved and the ischemic area was consistently reduced in size. Hyperventilation that lowered arterial CO2 made the ischemia area larger by reducing collateral flow. Using autoadiography to study rats subjected to MCA occlusion, Jones* found a significant increase in cortical blood flow in the territory of the occluded vessel with high arterial CO2 levels (i.e. 60 torr). The results of such studies indicate that the effects of elevated arterial CO2 levels on the brain circulation in ischemia are variable and may be unpredictable. Clinical studies have never demonstarted significant benefit from induced hypercarbia.
Interest in the use of vasodilating agents to improve blood flow in ischemia brain was rekindled with the introduction of prostacyclin, a prostaglandin derivative, and the dihydropyridine group of calcium-entry blocking agents. Prostacyclin is a very potent vascular smooth muscle relaxant that is produced primarily in endothelial cells; it also reduces platelet adhesiveness. Intravenous injection or topical application to normal cortex causes intense arterial dilatation and a marked increase in blood flow. Unfortunately, prostacyclin has not consistently improved blood flow in experimental studies of brain ischemia (5), nor has it reduced the neurologic morbidity of ischemia in controlled clinical trials (17,23,28).
Nimodipine and nicardipin are dihydropyridine derivatives that have been extensively evaluated in experimental studies of ischemia. Both agents selectively dilate brain arteries and increase brain blood flow under normal circumstances. Their effects on blood flow during ischemia, however, have varied; the most consistent benefit is seen during recirculation after temporary focal ischemia (6,15). In studies of ischemia after SHA from aneurysm rupture (3,26), nimodipine improved the clinical outcome despite the lack of any angiographic evidence or reduction in the severity of vasospasm. Nimodipine, it has ben surgested, may exert a beneficial effect in brain ischemia by preventing toxic calcium accumulation in the cytoplasm of ischemic neurons and glia. The relative lack of binkding affinity of dihydropyridine derivatives to neuronal and glial plasma membranes, however, reduces the likelihood of that possibility (36).

HEMODILUTION

Viscosity is a physical property of fluids that determines the internal resistance to shear forces. Blood viscosity is not constant (29). Blood is a non-Newtonian fluid: Its viscosity varies with the rate of flow. Slow-moving blood has a higher viscosity that the same blood moving rapidly. Hematocrit, erythrocyte rigidity, and plasma fibrinogen concentration also affect blood viscosity. Blood viscosity increases logarithmically with increasing hematocrit. Hematocrit, and consequently viscosity of blood, in the cerebral microcirculation is normally 70% to 80% of that in the large vessels (21).
Brain blood flow decreases with hematocrit levels above 50% and increases with hematocrit levels below 30% (14,46). This compensatory increases allows adequate oxygen delivery in normal subjects even when hamatocrit is as low as 20%. In a patient with an occluded brain artery and limited collateral circulation, a similar compensatory increase would not be possible; more severe impairment of oxygen delivery and greater tissue damage could result.
The relationship of blood viscosity to the pathogenesis and treatment of brain ischemia is of considerable interest. Studies have shown that a hematocrit level greater than 50% increases the risk of stroke (19,47). Other clinical and experimental studies have suggested that optimizing blood viscosity can limit tissue damage during an ischemic event (13,50,51). Hematocrit might also be a factor in the impairment of reperfusion after ischemia. Yield stress - the minimal force required to start blood flowing once it has been stationary - increases in relation to the third power of the hematocrit (29). Consequently, an elevation of hematocrit might prevent the restoration of blood flow after transient ischemia.
The optimal hematocrit for patients with brain ischemia has not been determined. Recent studies suggest that it is probably about 35% (13,14,29,49,50,51). To achieve this level, hematocrit can be quickly reduced by the intravenous administration of low-molecular-weight dextran, albumin, or saline. Blood can be removed concomitantly by venesection if normovolemic hemodilutioh is desired. Transfusions of packed erythrocytes can be given to raise the hematocrit when it is below 35%.
Experimental studies have consistently demonstrated improved blood flow and reduced infarct size after treatment with low-molecular-weight dextran (50). These beneficial effects have been attributed o the lowering of blood viscosity through hemodilution and to the reduction of platelet adhesiveness. The clinical use of low-molecular-weight dextran in acute brain ischemia has many advocates, but most of the reports have been anecdotal.
The Scandinavian Stroke Study Group conducted a stratified randomized multicenter trial in 15 hospitals (35). Patients who had an acute ischemic stroke within 48 hours of admission and a hematocrit of 38% to 50% were randomized to hemodilution and control groups. The results showed no benefit from normovolemic hemodilution (mean hematocrit reduction, 6.9%) maintained by venesectiona and in fusion of low-molecular-weight dextran. The failure to improve neurologic outcome does not mean that this form of therapy is without potential benefit in other situations. Undoubtedly, many of the patients treated in the Scandinavian study already had irreversible brain injury when treatment was initiated. Further studies are needed toe evaluate this therapeutic approach in a more acute setting and to define potentially respnsive subgroups of patients.

HYPEROSMOLARITY

Mannitol is the hyperosmolar agent used most frequently in the treatment of acute brain ischemia. Experimentally, this agent has been shown to have a beneficial effect on the microcirculation and infarct size when it is given early in the course of the ischemic event (24,25). These favorable effects are not seen when mannitol is given after ischemic breakdown of the blood - brain barrier or irreversible brain injury has occurred.
Several mechanisms have been proposed to explain the action of mannitol. It increases blood osmolality, which appears to retard early brain swelling and maintain flow through the microcirculation (24,25). Mannitol reduced erythrocyte transit through the capillary bed (9). The rapid administration of mannitol transiently reduces blood viscosity by lowering the hematocrit. Lower viscosity would also improve flow in ischemic areas. Muizelaar and associates (31,32) suggested that mannitol decreases blood viscosity and increases blood flow in nonischemic brain and that the arteries in nonischemic areas undergo secondary constriction to keep blood flow constant. They postulated that arterial constriction in nonischemic areas, rather than water conduction deiven by an osmotic gradient, is primarily responsible for the reduction of ICP after mannitol infusion. These changes could also act to improve blood flow in ischemic areas by decreasing ICP and thereby increasing perfusion pressure and by producing an inverse steal syndrome.
Report of the clinical efficacy of mannitol in brain ischemia are preliminary and anecdotal. It is unlikely, based on experimental evidence, that the delayed administration of mannitol would have a protective effect. The clinical applications for treatment of acute stroke therefore appear to be limited. Mannitol has frequently been given to patients by neurosurgons when temporary artery occlusion is needed for clipping of an intracranial aneurysm. Suzuki et al (39,40) have used mannitol extensively in combination with dexamethasone and vitamin E (the "Sendai cocktail") with good results in both expeimental and uncontrolled surgical studies.

PERSPECTIVE

Brain ischemia is a frequently encountered problem. In the past, patients with impending or evolving infarction were generally believed to be beyond help and were managed supportively. Although it is difficult to prove a beneficial effect in any given patient, the findings of experimental and clinical studies suggest that the morbidilty and mortality from acute ischemia might be reduced by optimizing systemic circulatory factors. Brain ischemia must be treated early if tissue necrosis is to be prevented or limited. In the usual clinical setting, treatment is often started too late to be effective. In patients undergoing cerebral vascular surgery with a potential risk of ischemia, the circulatory and hemodynamic state of the patient can be optimized in anticipation of the ischemic challenge.

References

1. Amea A, Wright RL, Kowada MD, et al., Cerebral ischemia. II. The no-reflow phenomenon. Am J Pathol 52:437-453, 1968.
2. Astrup J, Siesjo BK, Symon L: Thresholds in cerebral ischemia - the ischemic penumbra. Stroke 12:723-725, 1981.
3. Auer LM: Acute operation and preventative nimodipine improve outcome in patients with ruptured cerebral aneurysms. Neurosugery 15:57-66, 1985.
4. Awad IA, Carter P, Spetzler RF, et al: Clinical vasospasm after subarachnoid hemorrhage: Response to hypervolemic hemodilution and arterial hypertension. Stroke 18:365-372, 1987.
5. Awad IA, Little JR, Lucas F, et al: Modification of focal cerebral ischemia by prostacyclin and indomethacin. J Neurosurg 58:714-719, 1980.
6. Barnett GH, Bose B, Little JR, et al: Effects of nimodipine on acute focal cerebral ischemia. Stroke 17:884-890, 1986.
7. Brawley BW: The pathophysiology in intracerebral steal following carbon dioxide inhalation (abstract). Scand J Clin Lab Invest 22 (Suppl 102):13B, 1968.
8. Brawley BW, Strandness DE, Kelly WA: The physiologic response to therapy in experimental cerebral ischemia. Arch Neurol 17:108-187, 1967.
9. Burke Am, Quest DO, Chien S, et al: The effects of mannitol on blood viscosity. J Neuosurg 55:550-553, 1981.
10. Denny-Brown D: The treatment of recurrent cerebrovascular symptoms and the question of "vasospasm." Med Clin North Am 35:1457-1474, 1951.
11. Dintenfass L: Inversion of the Fahraeus-Lindquist phenomenon in blood flow through capillaries of diminishing radius. Nature 215:1099-1100, 1967.
12. Fahraeus R, Lindquist T: The viscosity of blood in narrow capillary tubes. Am J Physiol 96:562-568, 1931.
13. Grotta JC, Pettigrew LC, Allen S, et al: Baseline hemodynamic state and response to hemodilution in patients with acute cerebral ischemia. Stroke 16:790-795, 1985.
14. Haggendal E, Nilsson JN, Norback B: Efect of blood corpuscle concentration on cerebral blood flow. Acta Chir Scand Suppl 364:3-12, 1966.
15. Harris RJ, Branston NM, Symon L, et al: The effect of the calcium antagonist, nimodipine, upon physiological responses of the cerebral vasculature and its possible influence upon focal cerebral ischemia. Stroke 13:759-766, 1982.
16. Hope DT, Branston NM, Symon L: Restoration of neurological function with induced hypertension in acute experimental cerebral ischaemia. Acta Neurol Scand 56 (Suppl 64):506-507, 1977.
17. Huczynski J, Kostka-Trabka E, Sotowska W, et al: Double-Blind controlled trial of the therapeutic ischemic stroke. Stroke 16:810-814, 1985.
18. Jones TH, Morawet RB, Crowell RM, et al: Thresholds of focal ischemia in awake monkeys. J Neurosurg 54:73-782, 1981.
19. Kannel WB, Gordon T, Wolf PA, et al: Hemoglobin and the risk of cerebral infarction: The Framingham Study. Stroke 3:409-420, 1972.
20. Keller TS, McGillicuddy JE, LaBond VA, et al: Modification of focal cerebral ischemia by cardiac output augmentation. J Surg Res 39:420-432, 1985.
21. Lammertsma AA, Brooks DJ, Beaney RP, et al: In vivo measurement of regional cerebral hematocrit using positron emission tomography. J Cereb Blood Flow Metab 4:317-322, 1984.
22. Lassen NA, Palvolgyi R: Cerebral steal during hypercapnia and the inverse reaction during hypocapnia observed by the xenon-133 technique in man. Scand J Lab Clin Invest Suppl 102:13D, 1968.
23. Linet O, Hsu CY, Faught RF, et al: Epoprostenol in acute stroke (abstract). Stroke 16:149, 1985.
24. Little JR: Modification of acute focal ischemia by treatment with mannitol and high-dose dexamethasone. J Neurosurg 49:517-524, 1978.
25. Little JR: Morphological changes in acute focal ischemia. Response to osmotherapy. Adv Neurol 28:443-457, 1980.
26. Ljungren B, Brandt L: Timing of aneurysm surgery. Clin Neurosurg 33:159-175, 1986.
27. Maroon JC, Nelson PB: Hypovolemia in patients with subarachnoid hemorrhage: Therapeutic implications. Neurosurgery 4:223-226, 1979.
28. Martin JF, Hamdy N, Nichole J, et al: Double-blind controlled trial of prostacyclin in cerebral infarction. Stroke 16:386-390, 1985.
29. Merrill EW: Rheology of blood. Physiol Rev 49:863-888, 1969.
30. Muizelaar JP, Becker DP: Induced hypertension for the treatment of cerebral ischemia after subarachnoid hemorrhage. Surg Neurol 25:317-325, 1986.
31. Muizelaar JP, Lutz HA, Becker DP: Effect of mannitol on ICP and CBF and correlation with pressure autoregulation in severely head-injured patients. J Neurosurg 61:700-706, 1984.
32. Muizelaar JP, Wei Ep, Kontos HA, et al: Mannitol causes compensatory cerebral vasoconstriction and vasodilatation in response to blood viscosity changes. J Neurosurg 59:822-828, 1983.
33. Paulson OB: Regional cerebral blood flow in apoplexy due to occlusion of the middle cerebral artery. Neurology 20:63-77, 1970.
34. Ritchie WL, Weir B, O erton TR: Experimental subarachnoid hemorhage in the cynomolgus monkey: Evaluation of treatment with hypertension, vloume expansion and ventilation. Neurosurgery 6:57-62, 1980.
35. Scandanavian Stroke Study Group: Multicenter trial of hemodilution in acute ischemic stroke. I. Results in the total patient population. Stroke 18:691-699, 1987.
36. Snyder SH, Reynolds IJ: Calcium-antagonist drugs: Receptor interactions that clariy therapeutic effects. N Engl J Med 313:995-1002, 1985.
37. Soloman RA, Post KD, McMurtry JG: Depression of circulating blood volume in patients after subarachnoid hemorrhage: Implications for the management of symptomatic vasospasm. Neurosurgery 15:354-361, 1987.
38. Strandgaard S, Paulson OB: Cerebral autoregulation. Stroke 15:413-416, 1984.
39. Suzuki J, Fujimoto S, Mozoi K, et al: The protective effect of combined administration of anioxidants and perfluorocarbons on cerebral ischemia. Stroke 15:672-679, 1984.
40. Suzuki J, Yoshimoto T, Kayama T: Surgical treatment of middle cerebral artery aneurysms. J Neurosurg 61:17-23, 1984.
41. Symon L: Observaions on the leptomeningeal collateral circulation in dogs. J Physiol 154:1-14, 1960.
42. Symon L: Experimental evidence for "intracerebral steal" following CO2 inhalation (abstract). Scand J Clin Lab Invest 22 (Suppl 102): 13A, 1968.
43. Symon L, Braston NM, Strong AJ: Autoregulation in acute focal ischemia. Stroke 7:547-554, 1976.
44. Symon L, Crockard HA, Dorsch NWC, et al: Local cerebral blood flow and vascular reactivity in a chronic stable stroke in baboons. Stroke 6:482-492, 1975.
45. Symon L, Ishikawa S, Meyer JS: Cerebral arterial pressure changes and development of leptomeningeal collateral circulation. Neurology 13:237-250, 1963.
46. Thomas DJ: Whole blood viscosity and cerebral blood flow. Stroke 12:285-287, 1982.
47. Tohgi H, Yamanouchi H, Murakami M, et al: Important of the hematocrit as a risk factor in cerebral infarction. Stroke 9:369-374, 1978.
48. Tramner FI, Gross CE, Kindt GW, et al: Pulsatile versus nonpulsatile blood flow in the treatment of acute cerebral ischemia. Neurosurgery 19:724-731, 1986.
49. Waltz AG: Effect of blood pressure on blood flow in ischemic and non-ischemic cerebral cortex. Neurology 18:613-621, 1968.
50. Wood JH, Simeone FA, Fink EA, et al: Hypervolemic hemodilution in experimental focal cerebral ischemia. Elevations of cardiac output, regional cortical blood flow and ICP after intravascular volume expansion with low-molecular-weight dextran. J Neurosurg 59:500-509, 1983.
51. Wood JH, Snyder LL, Simeone FA: Failure of intravascular volume expansion without hemodilution to elevate cortical blood flow in regions of experimental focal ischemia. J Neurosurg 56:80-91, 1982.
52. Yamamoto YL, Phillips KM, Hodge P, et al: Microregional blood flow changes in experimental cerebral ischemia. Effects of arterial carbon dioxide studied by fluorescein angiography and xenon clearance. J Neurosurg 35:155-166, 1971.

BACK

Drug Therapy In Epilepsy A Resume

Anti-epileptic medication
Indications for treatment

• Recurrent unprovoked epileptic seizures that are not separated by years.
• Single seizure in patients with:
  - A relevant underlying brain lesion or neurologic abnormality, or
  - A specific epileptic syndrome and an unprovoked seizure that will eventually
  need treatment, such a juvenile myoclonic epilepsy.
• The risks and benefits of both treatment and no treatment have been discussed extensively with the patient.
• The patient understands that the drug must be taken regularly and continued for at less 3 years from the time of the last seizure, and the about 40% of patients will need to take anti-epileptic medication lifelong, depending on the seizure syndrome and other factors.

Principles of anti-epileptic drug therapy

Choice of anti-epileptic drug:

• Select the single most appropriate drug for the seizure type and patient
  (Tables 9,10).
  § Carbamazepine and v
• proate are the first line treatments for focal epliepsies.
• Valproate is the first line treatment for generalized epliepsies.
• Lamotrigine is also very effective for generalized epliepsies.
• Phenytoin and clonazepam still have a place, but barbiturates are now used less.
• Vigabatrin, lamotrigine, topiramate, tiagabine and gabapentin are of proven benefit in 25-50% of patients with refractory partial seizures. It is not clear which is the most effective drug but vigabatrin is not recommended because it may cause constriction of the visual fields.
• Ethosuximide is only used for juvenile abscence epilepsy.
• The cost of newer anti-epileptic drugs (e.g. lamotrigine, topirmate, gabapentin, tiagabine, vigabatrin) is about 10 times greater than older drugs (phenobarb, phenytoin).

Sodium valproate (Epilim)

• Uncertain mechanism of action. Increases synthesis and slows the breakdown of GABA.
• Effective for the primary generalized epilepsies, particularly tonic-clonic seizures, abscences and myoclonus.
• As effective as carbamazepine and phenytoin for partial seizuers, especially is secondary generalization.
• Available as a 40 mg/ml syrup for pediatric or nasogastric tube use, crushable 100 mg tablets, and 200 mg and 500 mg eneric-coated tablets.
• Plasma half-life: 6-9 hours, so may be administered as two or three divided doses daily.
• Starting dose in adults: 200 mg twice daily (bd), given after meals to minimize gastric irritation, and increasing to 400 mg bd after 2 days, if tolerated. The full pharmacologic action may not occur for some weeks.
• Dose can be increased slowly to I g three times daily (tds) if necessary and tolerated.
• Monitoring of blood levels is often not necessary as three is a poor relation between serum concentrations and anti-epileptic efficacy or toxicity.
• Drug interactions: increases concentrations of other anti-epileptic drugs such as carbamazepine (and its metabolities) and phenobarbitone (phenobarbital), by inhibiting their metabolism, and decreases plasma levels of phenytoin; it is a minor inhibitor of oxidative metabolism but is not a liver enzyme induce.
• Adverse effect: tremor, weight gain due to appetite simulation, and thinning or loss of hair. Congnitive impairment is uncommon. Stupor and encephalopathy, although potentially dangerous, are rare idiosyncraic effects and may be due to acumulation of ammonia. The prothrombin time can be prolonged but not sufficient to cause bleeding. Liver toxicity, in the form of a microvesicular steatosis, may occur idiosyncratically, particularly, in young children and when combined with other (anti-epileptic) drugs. Other adverse efects include anorexia, dyspepsia, nausea, vomiting and rash (predictable), and acute pancreatitis, thrombocytopenia, hyperammonemia, teratogenicity (idiosyncratic).

Carbamazepine (Tegretol)

• First used in 1964.
• Effective for the prophylaxis of generalized tonic-clonic and partial seizures but not for absence o myoclonic epilepsy (which it may aggravate).
• Available as 100 mg and 200 mg tablets, or 200 and 400 mg controlled-release tablets that are helpful for avoiding peak-dose adverse effects.
• Plasma half-life: 24-45 hours initially but after continued long term use it falls to about 9 (8-24) hours.
• The drug must be introduced in a low dose to offset mild neurotoxicity (sedation, vertigo, ataxia, diplopia, nausea, headache). The usual starting dose in adults in 100 mg three times daily, or preferably, half a 200 mg or 400 mg controlled-release tablet twice daily for 2 days, followed by half a tablet in the morning and a whole tablet at night, and further increases if necessary, aiming to maintain the plasma level within the therapeutic range and control seizures (rather than waiting for another seizure to occur). Even with this cautious approach and the development of tolerance some patients are unable to remain on carbamazepine because of neurotoxicity. In addition a morbilliform skin rash limits its usefulness in 5-8% of patients.
• Adverse effects:
  - Dose-related: neuroptoxicity (dizziness, double vision, unsteadiness), nausea,
  vomiting, cardiac arrhythmia and orofacial dyskinesia.
  - Idiosyncratic: skin rash (5-8% of patients), agranulocytosis, aplastic anemia,
  syndrome of antidiuretic hormone secretion (leading to fluid retention and
  hyponatremia) hepatotoxicity, photosensitivity, Stevens-Johnson syndrome,
  lupus-like syndrome, thrombocytopenia and pseudolymphoma.
• A major inducer of hepatic cytochorme P450 activity. Variable autoinduction of metabolism accounts for the wide range of doses and for the substantial interindividual variation in concentration found with the same dose.
• Drug interactions: carbamazepine accelerates the clearance of itself (i.e. it induces its own metabolism), ethosuximide, clonazepam, clobazam, corticosteroids, theophylline, haloperidol, warfarin and hormones. So, most women taking the oral contraceptive pill require daily estrogen in a dose of 50 ug. Mutual enzyme induction or inhibition with phenobarbitone (phenobarbital), phenytoin, or primidone can result in a small rise or fall in steady-state concentrations or either of both drugs. The metabolism of carbamazepine is inhibited, causing neurotoxicity, by sodiukm valproate, cimetidine, danazol, dextropropoxphene (propoxyphene), diltiazem, erythromycin, issoniazid, and verapamil.

Phenytoin (Dilantin)

• First used as an antiepileptic drug in 1938.
• Effective for generalized tonic-clonic and partial seizures.
• No longer a drug of first choice, particularly in young women, because it may cause cosmetic changes (gum hyperplasia, acne, hirsutism, and facial coarsening), as well as sedation and unfavorable effects on cognitive function (e.g. attention, memory).
• Available as a 6 mg/ml suspension, as chewable 50 mg tablets, and as capsules of 30mg and 100 m.
• Plasma half-life: about 24 (9-40) hours.
• One of only a few drugs with zero order kinetics at theraoeutic dosage - as the concentration P - 450 enzyme system to metabolize the drug becomes saturated (usually at around 300mg/day), and so a small increment in dose cn poduce a large rise in serum level. Conversely, the circulation concentration may fall precipitously when the dose is modesly reduced.
• Starting dose for children: 5mg/kg daily, for adults starting and maintenance dose: 300mg daily, given as a single dose or in divided doses. (Note, this is not the loading dose that is required for status epilepticus.)
• If seizures continue, an increment of 30mg is appropriate particularly if the serum concentration is above 12mg/1 (60 umol/1).
• Adverse effects: mental slowing, unsteadiness of gait, slurred speech and tremor, and physical examination reveals gaze-evoked nystagmus and tandem gait ataxia. Other predictable adverse effects include nausea, anorexia, vomiting, dyspepsia, cognitive impairment, depression, aggression, drowsiness, headache, paradoxical seizures, megaloblastic anemia, hyperglycemia, hypocalcemia, osteomalacia and neonatal hemorrhage. Prolonged use is associated with coarsening of facial features, gum hyperplasia (98), acne and hirsutism.
• Idiosyncratic effects include blood dycrasia, lupus-like syndrome, reduced serum IgA, pseudolymphoma (Iymphadenopahy), rash, Stevens-Johnson syndrome, Dupuytren's contracture, hepatomegaly and hepatotoxicity and teratogenicity. Long term use may cause peripheral neuropathy, cerebellar degeneration due to purkinje cell loss and osteomalacia.
• Drug interaction: an enzyme inducer and may reduce the efficacy of many lipid-soluble drugs such as other anti-epileptic drugs, anticoagulants, coticosteroids, cyclosporine, oal contraceptives, and therophylline. Its metabolism may be inhibited, causing neurotoxicity by enzyme inhibitors such as allopurinol, amiodarone, chloramphenicol, cimetidine, imipramine, isoniazid, metronidazole, phenothiazines and sulfonamides.

Barbiturates
Phenobarbitone (phenobarbital)

Has been used an anti-epileptic drug since 1912.

• Inexpensive, widely available and as good as carbamazepine and phenytoin in controlling generalized tonic-clonic and partial seizures.
• Main drawback: its effect on cognition and behavior: fatigue, listlessness and tiredness in adults and insomnia, hyperkinesia, and aggression in children (and sometimes in elderly patients). Subtle impairments of mood, memory and learning can occur in all age groups.
• Usual dose varies from 90 to 300mg/day.
• Long plasma half life of 3-4 days; can therefore be taken as a single daily dose.
• In adults, it should be restricted to patients who cannot tolerate first-line anti-epileptic drugs or as an adjunct to first line therapy in refractory epilepsy.
• Withdrawal can lead to temporary increase in seizure frequency.

Methylphenobarbitone (prominal) (mephobarbital)
Methylphenobarbitone (mephobarbital) is metabolized to phenobarbitone (phenobarbital), and is given it twice the dose of phenobarbitone but co gers no special advantage.

Primidone (Mysoline)
Primidone is metabolized to phenobarbitone (phenobarbital) and phenylethlmalonamide, both of which are pharmacologially active. Efficacy is similar to that of phenobarbitone.

Clonazepam (Rivotril)

• A 1,4 benzodiazepine (like diazepam).
• Effective for generalized tonic-clonic and myoclonic seizures, and generalized absence (petit mal) epilepsy and partial seizures.
• Has more sustained and effective anti-epileptic activity than diazepam (Valium) but tolerance develops.
• Available as 0.5mg and 2.0mg tablets.
• Dose varies from 0.5-4.0mg three times daily.
• Long plasma half life of 20-40 hours.
• Adverse effects include sedation, irritability, and aggression.
• Few patients benefit from long term treatment and nearly half have an exacerbation of seizures when the drug is withdrawn, particularly if pre-existing brain damage.

Clobazam (Frisium)

• A 1,5 benzodiazepine.
• Less sedative than clonazepam and diazepam but still commonly cause tiredness as well as depression and iritability.
• Has a limited role as a long term anti-epileptic drug (like clonazepam) but can do effective as a short term treatment to 'cover' for special events such as holidays, weddings and surgery and for catamenial exacerbations.
• A single dose if 10-30mg can also be helpful it taken immediately after the first seizure in patients who have regular clusters of generalized tonic-clonic and partial seizures.
• Visual field defects are the most common serious adverse effect, which limits the use of this drug.

Lamotrigine (Lamictal)

• Inhibits voltage-gated sodium channels and reduces the release of glutamate, an excitatory amino acid neurotransmitter implicate in the pathophysiology of epilepsy.
• Effective in both partial and generalized tonic-clonic and absence seizures in adults and children.
• A weak inhibitor of dihydrofolate reductase, so long term therapy may disturb folate metabolism.
• Does not significantly induce or inhibit hepatic oxidative drug-metabolizing enzymes, nor affect the plasma concentrations of concomitant anti-epileptic drug.
• Metabolism induced by anti-epileptic drugs that induce liver enzymes such as carbamazepine, phenytoin and phenobarbiton (phenobarbital) (half-life about 15 hours).
• Metabolism inhibited by sodium valproate (half-life about 60 hours).
• Plasma half-life: about 29 hours (mean).
• Eliminated largely as an N-glucuronide conjugate.
• Available as 5mg dispersible, 25mg, 50mg, 100mg, and 200mg standard tablets.
• Starting dose in patients not taking sodium valproate is 25mg once a day for 2 weeks, followed by 100 mg per day given in two divided doses for 2 weeks. Thereafter, the dose should be increased to achieve the optimal response. The usual maintenance dose is 200-400 mg per day in two divided doses.
• Starting dose in patients taking sodium valproate is 25mg every alternate day for 2 weeks, followed by 25mg once a day for 2 weeks. The usual maintenance dose is 100-200mg a day, given once a day or in two divided doses.
• Adverse effects: a generalized maculopapular skin rash appears within 4 weeks of generally starting treatment in about 2-5% of patients. The incidence of rash is proportional to the rapidity with which the drug is commenced. It usually resolves after immediate withdrawal of lamotrigine, after which the drug can be re-introduced slowly. Rarely, serious skin rashes such as Stevens-Johnson syndrome and angiodema have been reported. Some patients complain of insomnia which can be minimized by giving the second dose in the afternoon. Dose-related adverse effects include dizziness, headache, diplopia, ataxia, somnolence, nausea, asthenia, blurred vision and vomiting.

Gabapentin (Neurontin)

• A GABA-related amino acid.
• Mechanism of action: uncertain; it may after L-type voltage-depandent calcium channels.
• Effective when used in doses of 1800mg/day and as an 'add-on' treatment in reducing the frequency of seizures by more than half in about 40% of patients with complex partial seizures and 60% with secondarily generalized tonic-clonic seizures.
• Not effective for absence seizures.
• Available as 300mg and 400mg capsules.
• Usual maintenance dose: 1200-2400 mg/day, but higher doses may be more effective.
• Does not seem to interact with other anti-epileptic drugs.
• Adverse effects on cognitive function may arise with higher doses.

Ethosuximide

• Used only for absence seizures; not effective against generalized tonic-clonic seizures.
• Infants require a dose of 20-40 mg/kg per day but lower weight-related doses are used in adults. In children over 6 years of age, I is started with a dose of 250mg capsules twice daily, increasing if necessary to three of four capsules daily.
• Adverse effects include drowsiness and bone marrow depression.

Topiramate (Topamax)

• A sulfamate derivative structurally unique among anti-epileptic drugs.
• Reversibly decreases the number of action potentials in spontaneous epileptiform bursts and reduces burst duration in cultured hippocampal neurons, suppressing intrinsic bursting of proximal subiculum neurons, and reducing voltage-gated sodium currents in cultured cerebellar granule cells.
• Reduces elevated levels of excitatory amino acids, glutamate and aspartate.
• Reversibly enhances post-synaptic GABA receptor currents.
• Effective as adjunctive therapy for partial seizures, in Lennox-Gastaut syndrome and possibly some primary generalized epilepsies such as drop attacks. It is not helpful for, and may aggravate, absence seizures.
• Needs to be commenced gradually, with 25mg alternate days at night or daily at night, other wise adverse cognitive effects may be prohibitive. Cognitive disturbances may manifest as slowed speech (and even mimic dysphasia and dysnomia) and a vulnerability to behavioral disturbances (i.e. cranky, not the same person). Other adverse effects include weight loss, renal stones in 1% of patients and a theoretical risk of teratogenicity.

Tiagabine

• A derivative of nipecotic acid the potently and selectively inhibits neuronal and glial GABA uptake. It specifically inhibits the GABA transporter GAT-1 and interacts only weakly with GABA receptors.
• Effective against partial and generalized convulsive seizures.
• May be contraindicated in generalized absence epilepsy.
• Failed monotherapy
  A single agent (monotherapy) generally achieves satisfactory seizure control without significant adverse effects in about half of patients. Although another
  15-20% attain better control with the addition of a second anti-epileptic drug, it is often better to strive for seizure control with monotherapy by gradually introducing another agent and then gradually withdrawing the initial agent. Overall, about 70% of patients can be controlled with monotherapy.

Withdrawal of antiepileptic medication

• Anti-epileptic drug withdrawal should be considered in patients who have been free of seizures for 2 years or longer; taking anti-epileptic drugs long term is incovenient and costly, and associated with adverse cognitive and behavioral effects, so it is essential to ascertain whether they are still necessary or not.
• Drug withdrawal is not likely to be successful (i.e. seizures recur) if:
  - Late age to onset (>16 years of age).
  - Mental retardation.
  - Symptomatic epilepsy (i.e. underlying untreated cause).
  - Certain epileptic syndromes, such as juvenile myoclonic epilepsy, when
  unprovoked seizures have occurred.
  - Family history of epilepsy.
  - Slow wave activity or focal epileptiform activity on EEG prior to medication
  withdrawal.
  - History of atypical febrile seizures.
• There are no definite predictors of recurrence but EEG findings of epileptiform activity prior to, or during, drug withdrawal are highly predictive of a further seizure if anti-epileptic medication is ceased. Relapses may also occur in patients with normal EEGs however.
• Withdrawal of any anti-epileptic drug must always be gradual in a step-wise fashion, over at least 6 weeks (and probably months for barbiturates) because abrupt withdrawal may provoke rebound seizures.

BACK

ANTIPHOSPHOLIPID SYNDROME
AND OTHER PROTHROMBOTIC STATES
(THROMBOPHILIAS)

Disorders of the blood that predispose to recurrent venous and possibly arterial thrombosis.

ANTIPHOSPHOLIPID SYNDORME

Definition

A heterogeneous disorder, both in terms of its clinical manifestations and range of autoantiboides, which is characterized by thrombosis, recurrent miscarriage, or both, in association with persistent positive laboratory tests for antiphospholipid antibody : lupus anticoagulant (LA), anticardiolipin antibody (ACA), or both, on repeated studies. Thrombocytopenia is an occasional feature.

Antiphospholipid antibodies (APAs)

• APA are a family of antibodies which are specific for several plasma proteins, such as human prothrombin and ß2 glycoproteins, which may bind to phospholipid surfaces.
• LA and ACA are different APAs but occur together in about 60% of patients with the phospholipid antibody syndrome; in the remaining 40%, only one is present.
• Las are immunoglobulins which interfere with one or more of the in vitro phospholipid-dependent tests of coagulation activated partial thromboplastic time, dilute Russell viper venom time, dilute prothrombin time, and kaolin clotting time. They therefore slow the rate of thrombin generation and therefore clot formation in vitro.
• ACAs are detected by immunoassay, most commonly enzyme-linked immunosorbent assay.

Epidemiology

• Prevalence 1 - 40% or so of ischemic stroke/TIA patients have raised circulating IgG or IgM anticardiolipin antibodies and/or the lupus anticoagulant, depending on the selection of patients, the timing of the blood sample after the onset of cerebral ischemia, the laboratory methods, and what level is deemed 'normal'. However, only a few of these patients have some or all of the constellation of features known as the APA syndrome.
• Age : any age.
• Gender : F>M.

Pathology

Thrombi are typically 'bland' and may be found in vessels of any size and on heart valves, such as the mitral valve leaflets. There is no evidence of inflammation of the vascular wall.

Pathophysiology of Thrombosis

The paradoxical association between the presence of autoantibodies with in vitro anticoagulant effects and the occurrence of a prothrombotic state is not fully understood. Patients with antiphospholipid syndrome have evidence of persistent coagulation activation and, as stated above, vascular occlusion is due to thromboembolism or embolization from sterile vegetation on heart valves, and not vasculitis.

It is unlikely that a single prothrombotic mechanism operates. Possible mechanisms include :

• Perturbation of the protein C system : the most likely cause of venous thrombosis : APAs interfere with activated protein C down-regulation of factors Va and VIIIa, resulting in an acquired activated protein C resistance.
• Alteration of the antithrombin III heparin sulfate down regulation of serine proteases : some APAs have specificity for heparin sulfate found on the surface of endothelial cells.
• Up-regulation of tissue factor expression by endothelial cells.

In some cases the APAs may represent an epiphenomenon.

Etiology
Primary
No evidence of other underlying disease.

Secondary

• Associated with rheumatic and connective tissue disorders : systemic lupus erythematosus (SLE) : about 30 - 40% of patients with SLE have LA ; rheumatoid arthritis ; systemic sclerosis ; temporal arteritis ; Sjogren's syndrome ; psoriatic arthropathy ; Behect's syndrome; others.
• Associated with other conditions :
  - Infections : viral ; bacterial ; parasitic.
  - Drug exposure : chlorpromazine ; hydralazine ; quinidine; quinine ; antibiotics ; phenytoin ; valproate ; procainamide.
  - Lymphoproliferative diseases ; malignant lymphoma; paraproteinemia.
  - Miscellaneous conditions : autoimmune thrombocytopenia; autoimmune hemolytic anemia ; sickle-cell disease; intravenous drug abuse; livedo reticularis; Guillian - Barre syndrome.

Clinical Features

Any vascular site can be affected so there is a wide range of clinical manifestation :

Dermatologic

• Sneddon's syndrome : livedo reticularis, stroke-like episodes and hypertension.
• Non-healing ulceration of the ankles and skin necrosis.

Neurologic

• TIA, stroke, or multifocal encephalopahty due to arterial or venous thrombosis in any sized vessel. Venous thromboembolic events account for about 70% of cases and arterial events the remaining 30%. The most common site for arterial thrombi is the cerebral circulation.
• Migraine - like headaches.

Obstetric

• Recurrent spontaneous miscarriage/fetal loss due to intrauterine death in the latter part of the first trimester or early second trimester.
• Intrauterine foetal growth retardation.
• Early - onset pre - eclampsia.
• Prematurity

Pediatric
Post - infectious thromboembolic events.

Investigations for Antiphospholipid Syndrome or any other
Suspected Procoagulant state (Thrombophilia)

Indications

• Young (<50 years) and no other cause found for TIA or ischemic stroke.
• Past history or family history of premature arterial or venous thrombosis, especially if unusual sites (cerebral, mesenteric, hepatic veins).
• Past history of recurrent miscarriages.
• Abnormal full blood picture or blood film (e.g., thrombocytopenia).
• VDRL/RPR positive : may be a false positive in the presence of ACAs because cardiolipin is the antigen used in the VDRL assay. However, the VDRL is positive in only about 25% of patients with APLAb.

Thrombophilia diagnostic tests

• Full blood count and film.
• ESR.
• Prothrombin time.

Coagulation assays for lupus anticoagulant (LA)

• Activated partial thromboplastin time (APTT) : lacks sensitivity but remains the most appropriate screening test for the LA. A prolonged APTT that fails to correct when affected plasma is mixed with normal plasma implies inhibition of the clotting system rather than deficiency of a component, and is the laboratory hallmark of LA.
• Dilute Russell's viper venom time (dRVVT) : prolongation of the dRVVT will not correct with the addition of normal plasma in the presence of LA. For confirmation, a platelet neutralization procedure, to show the dependence of phospholipid inhibitors, should be performed.
• Kaolin clotting time test : also detects LA, but it is not as sensitive as the dRVVT, and sensitivity is highly reagent dependent.
• Tissue thromboplastin inhibition test.

These are indirect coagulation assays sensitive to the phospholipid-dependent steps of blood coagulation. At least tow assays, with sensitive reagents and techniques must be used, most commonly the APTT with another test.

Immunologic assays

• Anticardiolipin antibody (ACA) : detected by means of solid-phase immunoassay, such as enzyme linked immunosorbent assay (ELISA) or raido immunoassay (RIA), employing cardiolipin or other negatively charged phospholipids as the antigen to measure antibody concentration and binding avidity.
• Other phospholipids, e.g., phosphatidyl serine.
• Antibodies to ß2 GPI.

Other thrombophilia diagnostic blood tests

• Antithrombin activity
• Protein C activity.
• Protein S antigen level.
• Factor V Leiden genetic analysis or activated protein C resistance functional analysis.
• Prothrombin G202 10A mutation.
• Antinuclear antibodies.
• Serum protein levels, serum protein electrophoresis and plasma viscosity : indicated in patients with elevated ESR or suspected hyperviscosity syndrome.
• Hemoglobin electrophoresis/sickle test : indicated in appropriate racial groups to detect sickle cell trait or disease.
• Fibrinogen.
• Platelet aggregation.
• Fasting plasma homocysteine.

Diagnosis
Thrombophilia

The diagnosis is usually established from routine first line investigation and standard thrombophilia screening and diagnostic tests.

Antiphospholipid antibody syndrome

Cannot be diagnosed on the basis of a single raised titer of ACA in the serum. The titer must be substantially raised on several occasions and associated with not just cerebral ishcemia but also with some combination of deep venous thrombosis, recurrent miscarriage, livedo reticularis, cardiac valvular, thrombocytopenia and migraine.

Lupus anticoagulant (LA)

Four sequential steps are necessary to establish the diagnosis of LA :
1. Demonstration of an abnormal phosopholipid (PL) dependent coagulation test (e.g., APTT, dRVVT).
2. Proof that the abnormality in step 1 is due to an inhibitor
3. Establishing the PL-dependence of the inhibitor.
4. Ruling out other coagulopathies.

Ischemic stroke due to a procoagulant state

It can be very difficult to attribute confidently the cause of the ischemic stroke to the hematologic disorder if the patient has other disease that could have caused the stroke.

More often than not, the hematologic disorder is one of several factors predisposing to thrombus formation, such as coexistent activated protein C resistance, or coexistent athero-thrombosis, trauma or dehydration. Irrespective, the hematologic disorder frequently needs treating in its own right.

Treatment

Uncertain ; no controlled trials of aspirin, anticoagulants, or immunosuppressive therapy have been undertaken.

Thromboembolic events

Long term anticoagulant therapy, INR : 2.5, if risks considered to be outweighed by benefits. Otherwise antiplatelet therapy.

Pregnancy

Heparin 24 000 units per day s.c. and aspirin 75 mg per day. Low molecular weight heparin may cause les sosteoporosis, heparin-induced thrombocytopenia, and bleeding events than standard heparin and does not require blood monitoring.

Prognosis

Uncertain, but limited studies suggest a high rate of recurrent stroke and other vascular events in patients with ischemic stroke and the antiphospholipid syndrome.

Other Procoagulant states
Quantitive abnormalities of formed blood
Elements

Erythrocytes

• Polucythemia rubra vera (primary polycythemia) :
  - Hematocrit >0.50 in males and >0.47 in females, provided the patient is rested, normally hydrated, and the blood has been taken without venous occlusion.
  - Raised red cell mass.
  - Increased whole blood viscosity, raised platelet count, and enhanced platelet activity, may cause TIAs, cerebral infarction and intracranial hemorrhage.
• Anemia (iron deficiency and presumably other types) :
  - If severe, usually causes non-specific neurologic symptoms such as generalized weakness, poor concentration and faintness but in the presence of severe arterial disease, it may provoke focal cerebral ischemia.

Leukocytes

• Leukemia :
  - May predispose to cerebral arterial or venous occlusion because of increased whole blood viscosity.
  - More commonly is associated with intracranial hemorrhage because of the hemostatic defect or CNS leukemic infiltration. L-asparaginase treatment for leukemia can cause both cerebral ischemia and hemorrhage.
• Malignant angioendotheliosis :
  - can present like a stroke but the neurologic involvement soon becomes diffuse and progressive.

Platelets

• Essential thrombocythemia
  - Platelet count >500 x 1009/1, causes arterial and venous thrombosis.
  - Headache with transient focal and non focal neurologic disturbances are the most common neurologic symptoms.
  - Occasionally, platelet function is defective predisposing to bleeding.
  - Other causes of thrombocytes should be excluded: malignancy, splenectomy or hyposplenism, surgery and other trauma, hemorrhage, iron deficiency, infections, polycythemia rubra vera, myelofibrosis and leukemia.

Qualitative abnormalities of formed blood elements

• Erythrocytes
• Sickle cell (SC) disease :
  - Tschemic stroke is common in homozygotic children and young adults ; intracranial hemorrhage may also occur sometimes.
  - Stroke is rare in heterozygotic adults, and usually in the context of a hypoxia - provoked sickle-cell crisis.
  - Small and large arteries, and veins, are occluded by thrombi as a result of the rigid red blood cells, raised whole blood viscosity, thrombocytosis, impaired fibrinolytic actiivty, and arterial stenosis due to fibrous profileration of the intima; the last can be detected in the MCA with transcranial doppler and is predictive stroke.
• Hemoglobin SC disease
  - may also be complicated by stroke.
• Paroxysmal nocturnal hemoglobinuria :
  - very rare :
  - intracranial venous thrombosis is more common than arterial thrombosis.
  - Anemia, abdominal pain, recurrent deep venous thrombosis, dark urine, hemolysis and a low platelet and granulocyte count are recognized features.

Hyperviscosity

Paraproteinemias

• Waldenstrom's macroglobulinemia, multiple myeloma and perhaps cryoglobulinemia :
  - Most commonly cause the 'hyperviscosity syndrome', a global encephalopahty characterized by headache, ataxia, diplopia, dysarthria, lethargy, poor concentration, drowsiness and coma, visual blurring, and deafness. The retina shows dilatation and tortuosity of the veins, venous occlusions, papilledema and hemorrhages ; similar symptoms may also be due to uremia, hypercalcemia or lymphoma complicating the paraproteinemia.
  - Arterial or venous cerebral infarction may occur if vessels become occluded by acidophilic material which are probably precipitants of the abnormal plasma proteins.
  - Intracranial hemorrhage also occurs due to reduced number and
  Impaired reactivity of platelets, perhaps as a result of uremia.

Coagulation disorders

Antithrombin III

• Inactivates thrombin and activated factors, X, IX, XI, and XII (but not VII).
• An important mediator of the anticoagulant effect of heparin ; heparin increases the activity of antithrombin III by 100-fold. Indeed, heparin resistance is a clue to low antithrombin III activity.
• Deficiency may be inherited, or acquired as a result of severe liver disease, intravascular thrombosis, the nephrotic syndrome, or use of medications such as L-asparaginase of the oral contraceptive pill.

Protein C

• A vitamin K-dependent plasma protein that is synthesized in the liver in an inactive form.
• When activated by thrombin in the presence of calcium, it inhibits the procoagulant activity of factor Va and factor VIIIa and inactivates plaminogen activator inhibitor 1.
• Deficiency may be inherited as an autosomal dominant trait, or acquired as a result of warfarin therapy or severe liver disease, disseminated intravascular coagulation and acute thrombosis. Accounts for about 10% of cases of venous thrombosis.
• Resistance to activated protein C is commonly used by an autosomal dominant inherited mutation in the factor V gene, which was disocvered in Leiden, the Nehterlands, in 1994; hence the inherited defect is called the Leiden factor V mutation. Accounts for 15-50% of cases of venous thrombosis.

Protein S

• A cofactor of protein C ; free protein S increases the affinity of protein C for phospholipid and enhances the inactivation of factors Va and VIIIa by activated protein C.
• Deficiency may be inherited as an autosomal dominant trait, or acquired as a result of warfarin therapy, severe liver disease, and the nephrotic syndrome. Accounts for about 5% of cases of venous thrombosis.

Herediatry deficiency of coagulation inhibitors, activated protein C resistance, and hereditary abnormalities of fibrinolysis : plasminogen deficiency and / or abnormality

• Conditions in which spontaneous and recurrent venous thrombosis and rarely, arterial thrombosis are presenting or complicating features.
• The relevance of low levels of these natural anticoaulants in the etiology and prognosis for many patients with ischemic stroke of 'no other cause' is uncertain so the potential benefits of any treatment are uncertain.
• Difficulties arise attributing the cause of the stoke to one of these conditions :
  - Familial deficiency of antithrombin III, protein C and protein S are not uncommon in the general population so the hypercoagulation state may be coincident and not a sole or contributing cause of the stroke. Therefore, the patient must be properly assessed and other potential causes of ischemic stroke must be considered and excluded if possible.
  - Low levels of these coagulation factors may be caused by the acute stroke itself or its treatment the acute stroke itself or its treatment so the results must be confirmed by repeated testing on several later occasions.
  - Acute stroke is associated with activation of procoagulant and fibrinolytic pathways, as manifest by a significant decrease in functional antithrombin III and plasminogen and an increase in thrombin-antithrombin complex, total protein S, tissue plasminogen activator, plasminogen activator inhibitor 1, and D-dimer.

Factor II (prothrombin) mutation (G20210A)
A mutation in the prothrombin gene results in elevated prothrombin levels and carries a nearly threefold increased risk of venous thrombosis, but does not appear to be associated with an increased risk of arterial thrombosis and ischemic stroke.

Immunologic disorders
Antiphospholipid syndrome

Prothrombotic States of Uncertain cause

Thrombotic thrombocytopenic purpura (TTP)

• Rare.
• Microangiopathic hemolytic anemia, thrombocytopenia, fever, and renal failure are characteristic.
• Platelet microthrombi cause infarcts in many organs, including the brain
• Neurologic symptoms in 90% of cases, and are the presenting symptoms in 60%.
• Fluctuating encephalopathy, rather than a stroke syndrome, is the usual presentation.
• The patient is unwell with malaise, fever, skin purpura, renal failure, proteinuria, hematuria, thrombocytopenia, hemolytic anemia and fragmented red blood cells.
• Brain CT may be normal, or show infarcts, or occasionally intracerebral hemorrhage, more commonly due to heparin treatment than the disease itself.
• Non-convulsive status epilepticus is another treatable cause of altered mental status in these patients.

Cancer

Laboratory abnormalities of coagulation of firbinolysis are commonly found, particularly in patients with metastases.

About 2% of patients with cancer have a TIA or stroke at some stage.

Possible causes of ischemic stroke or TIA include :

• A coagulopathy mediated by intravascular mucinosis, low-grade disseminated intravascular coagulation, or the patient's chemotherapy.
• Embolism of non-infected heart valve vegetations.
• Tumor or septic emboli, sometimes with aneurysm formation.
• Sepsis.
• Hemostatic failure
• Hyperviscosity syndrome
• Neoplastic compression or invasion of neck arteries.
• Irradiation of neck arteries.

Disseminated intravascular coagulation

• Manifests as an acute or subacute global encephalopathy, rather than stroke-like episodes, in a very sick patient.
• CT brain scan reveals widespread hemorrhagic infarcts and hemorrhages.
• Low platelet count, low plasma fibrinogen, raised fibrin degradation products and raised D-dimer point to the diagnosis.

Pregnancy and the puerperium

The relative risk of stroke in the pregnant woman is 13 times the risk in the non-pregnant woman of the same age, but he absolute risk of stroke in the last trimester of pregnancy and the puerperium is no more than 30 per 100 000 deliveries. About three-quarters of ischemic strokes are due to arterial occlusion and one quarter venous occlusion. Causes include :

• Paradoxical embolism from the venous system of the pelvis of legs.
• Valvular heart disease.
• Cardiomyopathy of pregnancy.
• Arterial dissection during labor.
• Hematologic disorders.

Less commonly :

• Amniotic fluid embolism
• Air or fat embolism
• Metastatic choriocarcinoma.

Estrogens/oral contraceptives

• High estrogen dose oral contraceptive use is associated with a threefold increased risk of TIAs and stroke. But the absolute risk is very small.
• The risk is greater in women who are older, who smoke and who have other vascular risk factors such as hypertension.
• Exogenous high doses of estrogen given to elderly men for treatment of prostatic cancer and male survivors of MI increase their risk of vascular death.

Heparin-induced thrombocytopenia

Heparin may paradoxically lead to thrombus formation and thrombocytopenia by two mechanisms :

• Type I consists of a transient decrease in platelet count 1 - 5 days after heparin is started and is thought to be due to reversible clumping of platelets.
• Type II is a persistent depression of the platelet count beginning 3-22 days after the introduction of heparin which is thought to be mediated by IgG antibodies against a heparin -platelet membrane complex.
  Nephrotic syndrome

Can be complicated by ischemic stroke, perhaps due to 'hypercoagulability' : loss of antithrombotic proteins in the urine.

Desmopressin and intravenous immunoglobulin

May cause hypercoagulability and ischemic stroke, perhaps by altering blood viscosity, hemorrheology, platelet aggregability or clotting factor levels.

Snake bite

May cause ischemic stroke but is more likely to cause defibrination and bleeding.

BACK

CARDIOEMBOLIC STROKE

Definition

Embolism of material from the heart to the brain causing ischemia or infarction of a part of the brain or eye, with or without hemorrhagic transformation of the infarct.

Epidemiology

Embolism from the heart probably accounts for about 20% of ischemic stroke and TIAs.

Etiology and Pathophysiology

Cardiac Sources of embolism in anatomic sequence
Right to left shunt (paradoxical emboli from the venous system) via

• Patent foramen ovale.
• Atrial septal defect.
• Ventricular septal defect.
• Pulmonary arteriovenous malformation.

Left atrium

• Thrombus : AF* ; sinoatrial disease (sick sinus syndrome) ; atrial septal aneurysm.
• Myxoma and other tumors*.

*Substantial risk of embolism.

Mitral valve

• Rheumatic endocarditis (stenosis* or regurgitation).
• Infective endocarditis*.
• Mitral valve prolapse
• Non-bacterial thrombotic (marantic) endocarditis.
• Antiphospholipid antibody syndrome.
• Prosthetic heart valve*.
• Papillary finroclastoma.

Left Ventricle

• Mural thrombus :
  - Acute myocardial infraction
  - Left ventricular aneurysm or akinetic segment.
  - Dilated cardiomyopathy*.
  - Mechanical 'artificial' heart*.
  - Blunt chest injury (myocardial contusion).
• Myxoma and other tumors*.
• Hydatid cyst.
• Primary oxalosis.

Aortic Valve

• Rheumatic endocarditis (stenosis or regurgitation).
• Infective endocarditis*.
• Syphilis.
• Non-infective thrombotic (marantic) endocarditis.
• Libman - Sacks endocarditis.
• Antiphospholipid antibody syndrome.
• Prosthetic heart valve*.
• Calcific stenosis / sclerosis/ classification.

Congenital heart disease (particularly with right to left shunt)

Cardiac manipulation / surgery / catheterization / valvuloplasty / angioplasty

*Substantial risk of embolism.

Prevalence of potential cardiac sources of embolism in patients with first-ever ischemic stroke*

• Any AF : 13%.
  - Without rheumatic heart disease : 12%
  - With rheumatic heart disease : 1%.
• Mitral regurgitation : 6%.
• Recent (<6 weeks) myocardial infarction : 5%.
• Prosthetic valve : 1%.
• Mitral stenosis : 1%.
• Paradoxical embolism : 1%.
• Any of the above : 20%.
• Other sources of uncertain significance : aortic stenosis/ sclerosis ; mitral annulus calcification, mitral valve prolapse and son on : 11%.

*Sandercock PAG, Warlow CP, Jones LN, Starkey IR (1989) Predisposing factors for cerebral infarction : The Oxford Community Stroke Project. BMJ, 298 : 75 - 80.

Note :

• Not all cardiac sources of embolism pose equal threats.
• Not all emboli are of the same size of the same material.

Atrial fibrillation
The most common cause of cardioembolic stroke, accounting for up to 12% of all ischemic strokes, and an even greater proportion of ischemic strokes in the very elderly where its frequency in the population is highest. Atrial fibrillation is the cause of stroke in many of these patients but it is not always the cause because :

• Other possible causes of stroke, which may also be the cause of the AF, such as ischemic heart disease and hypertension, are present in about 20% of fibrillating stroke patients.
• Some AF patient have lacunar syndromes.
• 'Only' about 13% of non-rheumatic fibrillating patients have detectable thrombus in the left atrium and it is unknown if these patients have a higher stroke risk than those without detectable thrombi.
• In a few case the AF is caused by the stroke.

The average absolute risk of stroke is uncoagulated non-rheumatic AF patients is about 5% per annum and about 12% per annum in unanticoagulated firbillating TIA / stroke patients. The risk of stroke among patients in AF is variable ; some are at particularly high risk and others at particularly low risk of embolization.

Risk factors for embolization in AF patients

Low risk : no other detectable heart disease.

High risk :

• Rheumatic mitral valve disease.
• Previous cerebral or systemic embolic event.
• Increasing age.
• Hypertension
• Diabetes.
• Left ventricular systolic dysfunction.
• Enlarged left atrium defined by echocardiography.
• Spontaneous echo contrast in the left atrium, probably a consequence of blood stasis.
• Left atrial thrombi, left atrial appendage size and dysfunction, and various hemostatic variables are perhaps adverse risk factors also.

Uncertain risk :

• Recent onset AF.
• Paroxysmal AF : probably depends on frequency and duration of episodes of AF
• Thyrotoxic AF.

Coronary heart disease

Coronary heart disease is common in patients with TIA and ischemic stroke : about 20 - 40% have a past history of MI or current angina. Stroke may occur in up to 5% of patients with recent acute myocardial infarction, due to :

• Embolism of left ventricular mural thrombus.
• Systemic hypotension.
• Intracerebral hemorrhage secondary or thrombolysis, anticoagulants of aspirin.
• Embolism of catheter thrombus during coronary angioplasty / stenting.
• Concurrent non-cardiac cause of stroke.

After the acute period, the risk of stroke is much lower, about 1% in the first year, perhaps higher if there is persisting left ventricular thrombus. Chronic left ventricular aneurysm after MI often contains thrombus but embolization is uncommon.

Prosthetic heart valves

• The risk of embolism is above 2% per annum for all prosthetic valves, provided patients with mechanical valves are on anticoagulants.
• Mechanical valves have a higher risk of embolism that tissue valves, but there is no difference in stroke risk between the difference in stroke risk between the different types of mechanical valve.
• Some Bjork - Shiley tilting disc valves have disintegrated and embolized pieces to the brain.
• Prosthetic mitral valves are more prone to thrombosis than arotic valves.
• Infective enodcarditis is a potential risk for any type of prosthetic valve.

Rheumatic Valvular disease

• Rheumatic mitral stenosis / regurgitation is a well recognized cause of left atrial dilation causing thrombus formation and embolism to the brain.
• The valves also degenerate so even patients in sinus rhythm who have no thrombus in the left atrium are at risk of embolism of degenerate and sometimes calcific fragments of valve into the circulation.
• Stroke may also occur as a result of infective endocarditis and intracerebral hemorrhage due to anticoagulation in these patients.

Non-rheumatic sclerosis/ calcification of the arotic and mitral valves
These may also be a source of embolism in some patients but unless calcific emboli are seen in the retina or on CT it is difficult to attribute confidently the TIA or ischemic stroke to this condition, which is very common in normal elderly people.

Mitral valve (or leaflet) prolapse

Uncomplicated mitral valve prolapse is not a cause of embolism from the heart to the brain. It is only likely to be relevant to the etiology of an ischemic stroke or TIA if it is complicating infective endocarditis, AF, gross mitral regurgitation, or thrombus in the left atrium. Prolapse may be familial and associated with various inherited disorders of connective tissue.

Non-bacterial thrombotic endocarditis

Primary cardiomyopathies are well recognized causes of intracardiac thrombus, particularly the dilated type rather than hypertrophic subaortic stenosis. Many are familial.

Sinoatrial disease

• May be associated with intracardiac thrombus and embolism, particularly if bradycardia alternates with tachycardia, or the patient is in AF.
• May be familial.

Atrial septal aneurysm

• Uncommon
• May be complicated by thrombus and embolism to the brain.
• Often associated with a patent foramen ovale and so has the potential for paradoxical embolism from the venous system. The presence of both atrial septal aneurysm and patent foramen ovale increases the risk of recurrent stroke.

Paradoxical embolism from the venous system, or right atrium

Patent foramen ovale

• Common ; present in about 10% of normal people.
• An the uncommon cause of ischemic stroke ; although bubbles can be shown to move from the right to the left side of the heart frequently, it must be rare for venous thrombus to do so without also going to the lungs, or at least to be able to make a certain diagnosis of such an event during life.
• A recent study of 581 patients, aged 18 - 55 years, with ischemic stroke of unknown origin in the preceding 3 months has shown that after 4 years the risk of recurrent stroke was 2.3% among patients with patent foramen ovale alone, 15.2% among patients with both a patent foramen and atrial septal aneurysm, and 4.2% among patients with neither of these cardiac abnormalities. There were no recurrences among patients with atrial septal aneurysm alone.

Atrial septal defect

Ventriculoseptal defect(rarely)

Pulmonary arterial venous malformation

• Occurs in isolation or as part of hereditary hemorrhagic telangiectasis
• Clues are clubbing, cyanosis, hemoptysis, bruit over the chest and a 'coin lesion' on the chest x-ray.

Intracardiac tumors

• Rare.
• Myxomas are the most common heart tumor and most occur in the left atrium.
• Some are familial.
• Myxomas may cause :
  - Constitutional upset : malaise, weight loss, anemia, raised ESR and hypergammaglobulinemia.
  - TIA or ischemic stroke : embolism of tumor or complicating thrombus.
  - Primary intracerebral or subarachnoid hemorrhage : myxomatous emboli to sites of earlier symptomatic or even asymptomatic embolic occlusions can cause fusiform and irregular aneurysmal dilatations at these sites which can rupture.
• Myocardial hydatid cysts and intracranial calcification due to primary oxalosis are even rarer causes of embolism to the brain, the former with the subsequent development of intracranial cysts.

Myocardial contusion

• Blunt chest injury :
• May be associated with left ventricular thrombus and embolization.

Clinical Features

• Major stroke occurs when large emboli impact permanently in the internal carotid artery or trunk of the MCA.
• Partial anterior circulation infarction occurs when smaller emboli impact in a distal branch of the anterior of MCA.
• Posterior circulation infarction occurs when moderate or small sized emboli impact in the tip of the basilar artery or one of its branches.
• TIA of the brain or eye.
• Asymptomatic : Some emboli do not occlude arteries completely or, if they do, there is sufficient collateral circulation to maintain tissue integrity.

The following strongly suggest embolism from the heart

• Non-lacunar infarcts
• AF.
• Recent acute MI.
• Valvular heart disease.
• Embolic, particularly calcific emboli, visible in the retina.
• Embolic infarction in other organs.

Investigations
CT brain Scan

• Single or multiple wedge-shaped cortical /subcortical infarcts, with or without hemorrhage transformation.
• High density in the middle or basilar artery on the non-contrast CT suggestive of blood clot or calcium.
• Infarction in the territory of the tip of the basilar artery or superior cerebellar artery.

EKG (electrocardiograph) : AF, IHD

Chest x-ray

Echocaardiography

Indications

• Multiple cortical / subcortical cerebral infarcts in different arterial territories.
• Wedge-shaped cortical / subcortical infarct or hemorrhage infarct, or striatocapsular infarct and no carotid lesion on carotid ultrasound.
• Clinical, EKG or CXR evidence of heart disease.
• Age <45 years and no cause found for TIA or stroke.

Preferred echocardiographic technique for detecting various cardiac disorders
Transthoracic echocardiography

• Left ventricular thrombus
• Left ventricular dyskinesis.
• Mitral stenosis.
• Mitral annulus calcification.
• Aortic stenosis.

Tansesophageal echocardiography

• Atrial thrombus
• Atrial appendage thrombus
• Spontaneous echo contrast
• Intracardiac tumors
• Atrial septal defect*
• Atrial septal aneurysm
• Patent foramen ovale*
• Mitral and aortic valve vegetations.
• Prosthetic heart valve malfunction.
• Aortic arch atherothrombosis / dissection.
• Mitral leaflet prolapse.

* Transcranial doppler. Detection of intravenously injected air bubbles is less invasive, more specific, but not quite so sensitive ; galactose particle suspension increases sensitivity.

Diagnosis
Patients may have two or more competing causes of cerebral ischemia, such as carotid stenosis and atrial fibrillation, so one cannot always be sure which is the cause in an individual patient.

Embolism from the heart to the brain or eye
More likely to be the cause of ischemic stroke or TIA if

• An identified cardiac source of embolism, particularly one with a substantial embolic risk.
• Ischemic events in more than one arterial territory, particularly if more than one organ is involved.
• No evidence clinically or by angiography of arterial disease in the neck.
• Calcific emboli in the retina.
• Calcific emboli on brain CT
• No vascular risk factors
• Age <50 years.
• No other explanation for the stroke.

Less likely if

• Lacunar syndrome
• Low flow infarction / ischemia

Uncertain if

• Hemorrhagic transformation of the infarct.
• Past TIA.

High risk of embolism

• Non-rheumatic or rheumatic AF.
• Infective endocarditis.
• Prosthetic heart valve.
• Recent MI.
• Dilated cardiomyopathy
• Intracardiac tumors
• Rheumatic mitral stenosis

Low risk of embolism

• Mitral valve prolapse
• Mitral annulus calcification
• Patent foramen ovale with no evidence of deep venous thrombosis.
• Atrial septal aneurysm.
• Aortic sclerosis.

Treatment
Of acute ischemic stroke of suspected cardioembolic origin.

Immediate anticoagulants very likely to be worthwhile
Start as soon as possible
TIA or ischemic stroke with complete recovery within 1 -2 days + high risk of embolism.

Best time to start unclear
Non-disabling ischemic stroke, no hemorrhagic transformation of the infarction, and AF.

Immediate anticoagulants probably worthwhile
Best time to start anticoagulants unclear

• Acute MI within past few weeks and confirmed ischemic stroke of TIA.
• Disabling ischemic stroke and AF.

Aspirin or no antithrombotic therpay ; anticoagulants not worthwhile or contraindicated
Law risk of recurrent cardioembolic stroke without anticoagulants

• Heart disease with a low risk of embolism
• Other, non-cardiac lesion more likely cause of cerebral infarct : severe ipsilateral carotid stenosis ; likely disease of intracranial small vessels.

Little to gain from long term anticoagulation

• Suspected cholesterol embolization syndrome
• Already severely disabled before the stroke.
• Moribund or predicted to be severely disabled in the long term.

High risk of cererbal hemorrhage on anticoagulants

• Infective endocarditis.
• Large cerebral infarct with midline shift and / or evidence of major hemorrhagic transformation on brain CT.
• Likely to comply poorly with anticoagulation therapy and its monitoring after discharge.
• Severe, uncontrolled hypertension.

Contraindication to anticoagulants

These depend on individual circumstances and are seldom absolute.

• History of bleeding disorder.
  - Hemophilia.
• Uncorrected major bleeding disorder :
  - Thrombocytopenia.
  - Hemophilia.
  - Liver failure.
  - Renal failure.
• Uncontrolled severe hypertension :
  - Systolic pressure over 26.7 kPa (200 mgHg).
  - Diastolic pressure over 16 kPa (120 mgHg).
• Potential bleeding lesions :
  - Active peptic ulcer.
  - Esophageal varices.
  - Intracranial aneurysms
  - Proliferative retinoapthy.
  - Recent organ biopsy.
  - Recent trauma or surgery to head, orbit, spine.
  - Recent stroke, but patient has not had a brain CT scan or MRI.
  - Confirmed intracranial or intraspinal bleeding.
  - History of heparin - induced thrombocytopenia or thrombosis.
• If warfarin planned :
  - Homozygous protein C deficiency.
  - History of warfarin - related skin necrosis.
  - Uncooperative / or unreliable patients.
  - Risk of falling.
  - Unable to monitor INR.

Adverse effects of heparin
Local minor complications of subcutaneous heparin at injection site

• Discomfort.
• Bruising.

Local complications of intravenous heparin at cannula site (or elsewhere)

• Pain at cannula site.
• Infection at cannula
• Reduced patient mobility because of infusion lines and pump.

Major systemic complications

• Intracranial bleeding :
  - Hemorrhagic transformation of cerebral infarct
  - Intracerebral hematoma.
  - Subarachnoid hemorrhage.
  - Subdural hematoma
• Extracranial hemorrhages :
  - Subcutaneous
  - Visceral
• Thrombocytopenia :
  - Type I : dose and duration related, reversible, mild, usually asymptomatic, not serious and often resolves spontaneously.
  - Type II : idiosyncratic, allergic, severe. Affects 3 - 11% of patients treated with intravenous heparin and less than 1% of patients treated with subcutaneous heparin.
• Osteoporosis
• Skin necrosis
• Alopecia.

BACK

ESSENTIAL TREMOR (ET)

Definition

A low frequency postural tremor which is absent at rest and not associated with the clinical signs of parkinsonism or other neurologic deficits.

Epidemiology

• Incidence 8 (95% CI : 4 - 14% 100 000/year.
• Prevalence (lifetime) : 0.3 - 1.7%.
• Age of onset : bimodal : young adults (median about 15 years) and elderly.
• Gender : M = F.

Pathology
No characteristic pathological or biochemical findings.

Etiology

• Hereditary : autosomal dominant inheritance (50% of patients) via a penetrant autosomal dominant gene. The responsible gene(s) remain unknown.
• Sporadic : probably the same entity as hereditary ET.

Pathophysiology

• Unknown
• A central source of oscillation that is influenced by somatosensory reflex pathways.
• Enhanced olivocerebellar oscillation and activation of cerebellothalamocortical pathways probably have an important role in the generation and transmission of the tremor because :
  - functional imaging studies reveal increased olivary glucose metaboism and increased blood flow bilaterally in the cerebellum and red nuclei and contralaterally in the globus pallidus, thalamus and primary sensorimotor cortex of patients with essential tremor.
  - Lesions of the ipsilateral cerebellum diminish essential tremor.
  - Ventrointermediate thalamotomy or microstimulation reduce tremor.
• It is thought that the cerebellum introduces an error in the timing of muscle bursts during voluntary movement, and repeated corrective movements lead to tremor.

Clinical Features

Tremor

• Postural or action tremor of the finer, hands, and forearms when they are help outstretched against gravity, or used for specific, usually visually-guided, manual tasks.
• Variable kinetic component
• 4 - 12 cycles per second.
• Bilateral, but may be asymmetric.
• Other body parts involved in about half of cases : the head, tongue, lips, voice, face, and trunk; and postural tremor of lower limbs.
• Visible.
• Persistent, but the amplitude may fluctuate.
• Exacerbated by emotional upset such as excitement and anger.
• Improved temporarily by alcohol, in small quantities, in about half of patients.
• Relatively longstanding.
• Froment's sign (a rhythmic resistance to passive movements of a limb about a joint when there is a voluntary action of another body part).

Differential Diagnosis

Intention of 'terminal' tremor due to cerebellar disease

• Slower (3-5 cycles per second).
• Principally in a horizontal plane.
• Not present at rest.
• Increases with action and progressively increases during a voluntary movement.
• Unlike kinetic tremor, which is tremor during any form of movement; 'intention' or 'treminal' tremor is the pronounced exacerbation of kinetic tremor toward the end of goal-directed movement.
• The head may be affected, but usually as titubation
• May be incapacitating.
• Responds to propranolol and diazepam.

Parkinosonian tremor

• Typically 6 (3-7) cycles per second.
• Present at rest.
• Not increased with posture or action.
• Involves the limbs and head.
• 'Pill rolling' character.
• Exacerbated by emotion.
• Responds to levodopa and anticholinergic medications.
  Alcohol withdrawal
• 6 -10 cycles per second
• Head and limbs involved.
• Periodic : appears 8 - 12 hours after last drink.
• Responds to diazepam and chlordiazepoxide.

Metabolic derangements, such as thyrotoxicosis

• 10 - 20 cycles per second.
• Head and upper limbs involved
• Symptoms and signs of metabolic disturbance.
• Responds to correcting metabolic disturbance.

Enhanced Physiologic Tremor (EPT)

A rapid 8 - 12 Hz small amplitude, barerly visible, postural tremor, typically of the upper limbs, which occurs as a normal phenomenon during muscle contraction. EPT manifests as apostural and kinetic hand tremor. iT can be difficult to distinguish between EPT and early stages of hereditary ET in a young person. There is no current reliable way of making this distinction but the family history can be a clue.

Causes

• Drugs :
  Adrenergic drugs. Amiodarone.
  Amphetamines. Antipsychotic drugs.
  Caffeine. Cimetidine.
  Corticosteroids. Cyclosporine A.
  Lithium. Oral hypoglycemics.
  Serotonin reuptake inhibitors.
  Sodium valproate. Theophylline.
  Thyroxine. Tricyclin antidepressants.
• Drugs withdrawal :
  Alcohol. Barbiturates.
  Benzodiazepines. Opiates.
• Metabolic :
  Hypoglucemia. Metabolic encephalopathies.
  Pheochromocytoma. Thyrotoxicosis.
• Exaggerated physiologic response :
  Anxiety. Fatigue.
  Fight. Strenuous excretion.

Management is directed to correction of the underlying medical illness
or cessation of any contributing drug. If idiopathic EPT becomes inconvenient or socially embarrassing, propranol may be helpful.

Primary orthostatic tremor

• Isolated, high frequency bilaterally synchronous tremor of legs and trunk when standing still.
• May not be detected easily without palpation of the leg muscles.
• Patients may only complain of unsteadiness when standing.
• Responds to clonazepam and, in some cases, pramipexole ; ET does not respond to clonazepam.

Isolated position - specific or task-specific tremors (including occupational tremors and primary writing tremor)

• Have the appearance of localized ET but the task specificity of dystonia.
• May respond to anticholinergics.

Rubral tremor

• A several tremor with features of a parkinsonian rest tremor and
  marked exacerbation during movement.
• Not always associated with lesions of the red nucleus but may result from lesions of the ipsilateral cerebellar dentate nucleus or superior cerebellar peduncle in the midbrain.
• Most commonly seen in multiple sclerosis.

Dystonia

• Highly asymmetric postural tremor is probably a form of dystonia.
• Isolated voice tremor may be due to laryngeal dystonia and other dytonias of the vocal apparatus.

Psychogenic

• Unphysiologic variations in tremor frequency.
• Unusual and inconsistent behavioral characteristics.
• Spontaneous remissions.
• Psychiatric or social factors

Diagnosis
A clinical diagnosis based on the long duration of symptoms ; positive family history ; lack of rigidity, bradykinesia or other neurologic signs ; symmetry; and alcohol responsiveness.

Diagnostic criteria (predominantly for research purposes)
Definite ET

• Characteristic bilateral tremor on maintaining posture, typically of the outstretched upper limbs of more than 5 years duration. It is absent at rest, not made strikingly worse with movement and is not associated with extrapyramidal or cerebellar signs.
• No definite evidence of sudden onset.
• No direct or indirect trauma to the brain, spinal cord or relevant part of the peripheral nervous system in the preceding 3 months.
• No recent exposure to tremorgenic drugs.
• Not in a state of drug withdrawal.
• Normal neurologic examination other than tremor.
• No history or clinical evidence suggestive of psychogenic origins of tremor.
• No evidence of stepwise deterioration.

Probable ET

• Tremor : same as above
• Duration more than 3 years
• No evidence of isolated or localized tremors such as primary orthostatic tremor, isolated voice tremor, isolated position-specific or task-specific tremor, or isolated tongue or chin tremor.

Possible ET

Type 1
Patients satisfy the criteria for definite or probable tremor but exhibit other neurologic disorders or other neurologic signs of uncertain significance.

Type II
Monosymptomatic and isolated tremors of uncertain relation to ET.

Task-specific tremors and isolated tremors of the voice, tongue, head and legs have all been considered part of the spectrum of ET but it may be that the label 'ET'' is not appropriate for these diverse disorders whose etiologies are not yet known.

Treatment
Not all patients require treatment ; only those severely affected.

Medical
Propranolol and primidone have been shown in controlled trials to afford a partial reduction in tremor amplitude in about two-thirds of cases, and are the mainstay of treatment.

• Propranolol 10 - 40 mg mane or bd, or metroprolol 25 mg or bd, or metoprolol 25 mg twice daily with 25 mg increments to effect or until a maximum of 50 mg three times daily, leads to variable improvement. About half of patients experience a reduction in tremor amplitude of up to 50%. The effect is greatest on postural limb tremor. However, not all patients are helped and the tremor is seldom abolished. There is little point in prescribing a nocte dose. Adverse effects and relative contraindications for propranolol include heart failure, bradycardia, hypotension and asthma. The mechanisms of action of propranolol is thought to be blockade of peripheral skeletal muscle ß2 receptor. ß1 blockade is less effective than propranolol.
• Primidone 50 - 250 mg daily if necessary, if propranolol is ineffective or contraindicated. May be as effective as beta blockers but adverse effects are common. Care is required when introducing primidone because of nausea, sedation and unsteadiness, which may be sufficiently severe to warrant stopping the drug. A low starting dose minimizes adverse effects, particularly if taken in the evening before retiring. There appears to be no added benefit to increasing the daily dose beyond 250 mg.
• Clonidine in doses of 0.1 - 0.9 mg daily.
• Patients with alcohol responsive tremors may find judicious use of alcohol to be helpful before social engagements.
• Others drugs that may be helpful but also have adverse effects include clonazepam, alprazolam, flunarizine, clozapine, nicardipine, and carbonic anhydrase inhibitors such as methisnozole and acetazolamide. Gabepentin is currently undergoing evaluation for the treatment of tremor.

Botulinum toxin
Botulinum toxin can be injected locally into the splenius capitus to reduce 'no-no' tremor without adverse effects, but it does not work well when injected into forearm muscles for wrist tremor.

Surgical treatment

• Stereotactic thalamotomy is usually employed only for the most severe tremors because of the significant risks of adverse effects, particularly with bilateral operations. About 70-90% of patients experience some relief from tremor.
• Thalamic stimulation with chronically implanted electrodes, positioned with the use of intraoperative recordings in the nucleus ventralis intermedius of the thalamus, confers lower risk. High frequency stimulation, controlled by a subcutaneously implanted box on the chest wall, is thought to produce its beneficial effect by inducing depolarization block. Good results have been achieved in the vast majority of patients. It is hoped that the benefits will be maintained with longer term follow-up.

A randomized comparison of thalamotomy and thalamic stimulation
found that both were effective in relieving drug resistant tremor but thalamic stimulation produced greater functional improvement. Bilateral thalamic stimulation may have a lower complication rate than bilateral thalamotomy. Dysarthria and aphonia complicate up to 20 - 25 % of bilateral thalamic lesions, though with stimulation speech improves when the stimulator is switched off.

Prognosis

• ET is slowly progressive but seldom becomes severe.
• Other neurologic impairment do not occur.

BACK

LOCKED - IN - SYNDROME

By K. Gireesh

Definition

A de-efferented state whereby patients are aware of themselves and their environment but are unable to respond due to loss of motor and speech function.

Pathogenesis

• A supranuclear lesion of the descending corticospinal tracts, usually in the ventral portion of the brain stem, below the level of the IIIrd cranial nerve nuclei, causes paralysis of the muscles innervated by the lower cranial nerves and peripheral nerves.
• A widespread nuclear or infranuclear (lower motor neuon) disease of motor nerves.

Etiology
Ventral brain stem lesion

• Infarction or hemorrhage
• Tumor
• Demyelination
• Central pontine myelinolysis, following profound hyponatremia
• Head injury

Polyneuropathy

• Critical illness polyneuropathy.
• Acute onset post-infectious polyradiculoneuropathy

Clinical Features

• Unable to speak
• Unable to move the limbs
• Awareness and consciousness are preserved because the brainstem tegmentum, including the reticular formation and oculomotor nerves and pathways are spared.
• Able to open the eyes and move them, and blink, in order to try and communicate.

Investigations :

• CT or MRI brain scan : ventral pontine or midbrain lesion.
• Other investigations, as appropriate, to ascertain the cause.

Diagnosis

Diagnosis is clinical, based on the presence of total paralysis of the limbs and muscle innervated by the lower cranial nerves, but with the ability of the patient to open and close the eyes voluntarily and in response to commands, and to respond to verbal and sensory stimuli by blinking.

Treatment

General

Patients can see, hear and feel everything so are sensitive to what staff are saying. They are also very frustrated that they cannot move.

• Prevention of complications of immobility : pneumonia, deep vein thrombosis, contractures, urinary tract infection.
• Rehabilitation : physiotherapy, swallowing and speech therapy, occupational therapy, psychologic support and therapy.

Prognosis

Prognosis is poor. Some patient recover, usually with residual limb spasticity.

BACK

MIGRAINE

Definition
A symptoms complex, or syndrome, than manifests as discrete episodes of headache associated with other features of sensory sensitivity.

Epidemology
Prevalence
Lifetime
· Women : 33% (95% CI:31 - 37%).
· Men : 13% (95% CI : 12 - 16%).

1 year
· Women : 25% (95% CI : 23 - 29%).
· Men : 7.5% (CI : 7 - 9%).
· Higher in Caucasians than Africian Americans, than Asians.

Age
· Onset is nearly always before age 50 years; 25% begin in childhood.
· Peak incidence at age 10 - 12 for males and 14 - 16 years for females.
· Peak prevalence at age 50 years for men and 35 years for females.
· Attacks commonly increase in frequency at the menopause, but may decrease.

Gender
· Children : M- F
· Adolescents and adults : F>M = 2 - 3:1.

Etiology
Unknown

Migraine without aura
A combination of genetic factors and environmental factors ; first degree relatives of probands with migraine without aura have a twofold increased risk of migraine without aura compared with the general population ; the probandwise concordance rate is higher in monozygotic than dizygotic twins.

Migraine without aura
Largely genetic ; first degree relatives of probands with migraine with aura have a fourfold increased risk or migraine with aura compared with the general population ; the probandwise concordance rate is higher in monozygotic than dizygotic twins. However, environmental factors are also important as the pairwise concordance rate is less than 100% in MZ twin pairs.

Familial hemiplegic migraine (FHM)
A rare autosomal dominant subtype of migraine with aura. Genes for FHM map of chromosomes 19p 13 and 1q but some families with FHM do not link to either locus, indicating genetic heterogeneity of FHM. The CACNa1A gene at 19p13 encodes the a subunit of a brain specific P/Q type voltage-dependent calcium channel, suggesting that migraine may be a 'cerebral calcium channelopathy'.

Pathophysiology
Triggered by the action of a multitude of environmental and biochemical factors on the cerebral cortex or hypothalamus; the premonitory symptoms of elation, yawning or a craving for sweet foods, experienced by about 25% of patients, suggest hypothalamic activation.

Triggers

• Emotional stress and tension.
• Relaxation after stress.
• Fatigue.
• Hormonal changes : fall in estradiol levels at menstruation and midcycle.
• High dose estrogen - containing contraceptives.
• Strong sensory stimulation : bright or flickering light ; loud noise ; strong smells ; occipital nerve compression.
• Head trauma, such as heading the ball in soccer
• Food idiosyncrasies / allergies : rich foods, red wines, specific dietary amines.
• Missing meals.
• Sleeping late in the morning.
• Meterologic changes.
• Vasodilators, such as alcohol, monosodium glutamate, and anti-anginal agents.
• Substance misuse.
• Physical activity.

Trigeminovascular reflex

The activated cerebral cortex and hypothalamus stimulate brain stem nuclei, dorsal raphe nuclei and locus coeruleus and tirgger the trigeminovascular reflex which constitutes serotonergic and noradrenergic pathways that project from the brain stem to the cortical microcirculation and the spinal trigeminal nucleus and spinal cord.

Axons of the first division of the trigeminal nerve, which innervate the pain-sensitive intracranial structures, depolarize as a result of direct neuronal activation or vasodilation of dural and cerebral arteries, or both, leading to central transmission of nociceptive pain signals to bipolar neurons in the trigeminal ganglion and on to the trigeminal nucleus in its most caudal extent in the caudal medulla and the dorsal horn of the spinal cord at C1 and C2. Impulses are then transmitted to the ventroposteriomedial nucleus of the thalamus via the quintothalamic tract, from where they are relayed to the cortex.

Stimulation of the trigeminal ganglion leads to the release of powerful vasodilator neuropeptides such as calcitonin gene-related peptide from trigeminal neurons that innervate the cranial circulation. This peptide is not only a vasodilator but it also mediate a sterile neurogenic inflammation within the dura mater.

Migraine aura and headache

At the onset of aura, regional cerebral blood flow to the clinically involved part of the brain is reduced by about 20% and reduced neuronal activity spreads in a wave across the cerebral cortex, usually beginning in the occipital region and slowly moving forward.

Migraine headaches begins while regional cerebral blood flow is reduced. Platelets in the blood release serotonin, and this leads to platelet aggregation. During the headache, the level of a vasodilator peptide. GGRP increases in the external jugular venous blood, and some intracranial arteries become dilated and inflamed. Vascular dilatation and neurogenic inflammation is believed to be responsible for the pulsatile nature of the headache.

Migraine attacks can be ameliorated by activating 5-hydroxytryptamine 1 D presynaptic receptors within the vessel wall, thus blocking release of vasoactive neurpeptides, causing vasoconstriction of certain cerebral and dural arteries, and inhibiting depolarization of trigeminal axons, functionally blocking activation of trigeminal perivascular nerve terminals.

Clinical Features
Precipitating factors

Phase one : prodrome

• Occurs in 25 - 50% of migraineurs.
• Gradual onset and evolution over up to 24 hours.
• Lightheadedness, dulled perception, irritability, withdrawal, cravings for particular foods , frequent yawning, elation and speech difficulties.

Phase two : aura

• 15 - 25% of migraine attacks are associated with aura.
• Visual symptoms most commonly : blurred vision, flashing lights or shimmering zigzag lines of light, sometimes around an area of impaired vision or blindness in a part of the visual field of one or both eyes.
• Somatosenosry : tingling or pins and needles, or less commonly numbness, in the face, arm, hand or leg.
• Dysphagia : difficulty understanding and expressing specch.
• Gradual onset, symptoms 'build up' or progress over 5 - 10 minutes, then subside within 5 - 60 minutes.
• Followed within 60 minutes by headache ; the aura may continue to the headache phase.

Phase three : headache
Present in most, but not all migraine attacks.

Site

• Unilateral in two-thirds of patients, and bilateral in one-third.
• Frontotemporal region commonly, spreading to occipital region.

Quality

• Throbbing / pulsatile.
• Moderate to severe.

Aggravating factors

• Physical activity / movement.
• Bright light
• Loud noise

Associated features

• Scalp tenderness on the affected side.
• Nausea (90% of patients).
• Vomiting (60%)
• Diarrhea (20%)
• Heightened awareness of sensation such as smell and noise.
• Fluid retention at onset, and polyuria as headache subsides.

Duration

• 4 - 72 hours.
• Commonly 2 - 6 hours in children, and 6 - 24 hours in adults.

Phase four : postdrome
For up to 24 hours after the headache has subsided, most migraineurs feel tired 'drained' or 'washed out', with aching muscles. Others however, become euphoric for a period of time.

Periodicity with recurrence
Migraine is paroxysmal ; clearly defined episodes recur as often as 4 - 6 times each month.

Family history
Family history of migraine is present in more than half of patients.

Special Forms
Migraine variants
Less than 5 % of migraineurs,

Retinal migraine
Monocular, rather than binocular hemianopic visual disturbance.

Ophthalmoplegic migraine :

• Paralysis of > 1 of the ocular cranial nerves, usually the IIIrd nerve, at the height of a migraine headache.
• The paralysis usually resolves but may persist after recurrent episodes.
• This entity does not embrace a transient dilatation or constriction of one pupil as this is quite commonly seen during severe migraine attacks.
• The cause of the cranial neuropathy in ophthalmoplegic migraine is probably transient ischemia of the cranial nerve.

Vertebrobasilar migraine
Gradual onset and evolution over several minutes of brainstem, cerebellar and visual disturbances, often acompanied or followed by headache and syncope.

Hemiplegic migraine

• Hemiparesis preceding or occurring with a migraine headache.
• A family history of hemiplegic migraine is often present and the gene is located on chromosome 19 or 1.

Migrainous infarction

• Permanent focal neurologic symptoms persisting beyond 24 hours after the cessation of migraine headache. Cranial CT or MRI scan shows features consistent with cerebral infarction.
• The cause is probably arterial thrombosis, provoked by arterial spasm and a procoagulant state.

Menstrual migraine
Just before menstruation, plasma estradiol levels fall rapidly below about 20 ng/ml which sets in motion a series of changes that culminate in the onset of migraine in about 60% of women migraineurs and exclusively at that time in about 14%. Migraine is relieved by pregnancy in about 60% of women, many, but not all, of whom have a history of menstrual migraine.

Migraine in childhood

• Headache and vomiting are common but the child may be unable to describe the symptoms and may simply appear pale, ill, limp, and inert, complaining of poorly localized abdominal pain.
• Fever up to 38.5 C may be present so that the suspicion of appendicitis or mesenteric adenitis often arises.
• Rather than accept a label of 'bilious attack' or 'periodic syndrome', recurrent headaches or vomiting attacks in children which may develop at times of excitement or stress should be considered as possibly migrainous and not psychosomatic.

Differential Diagnosis
Abrupt onset of headache ('thunderclap headache')

Primary

• Cluster headache
• Benign exertional or sex headache.
• Idiopathic stabbing headache.

Secondary

• Subarachnoid hemorrhage.
• 'Sentinel headache'
• Intracranial hemorrhage.
• A precipitous rise in blood pressure : drug induced headache.
• Head injury.
• Acute obstruction of the CSF pathways.

Uilateral headache

• Cluster headache.
• Temporal arteritis.
• Glaucoma.
• Temperomandibular joint disease.
• Internal carotid or vertebral artery dissection.
• Structural intracranial lesion.

Continuous or daily headache

Primary

• Tension headache : just headache, and sometimes mild photophobia and phonophobia but no other features of sensory sensitivity.
• Mixed migraine / tension headache.

Secondary

• Drug - rebound headache : a periodic daily bilateral headache that has gradually increased in frequency, and changed in character from the typical migraine headaches, in concurrence with increasing consumption and misuse of analgesic drugs, particularly those which also contain caffeine.
• Systemic infection
• Giant cell arteritis.
• Raised intracranial pressure : idiopathic intracranial hypertension, brain tumor.
• Vertebrobasilar migraine
  Neurocardiogenic syncope

Migraine aura without headache (acephalgic migraine)

• TIA.
• Epileptic seizure.
• Arteriovenous malformation
• Mitochondrial DNA disorders.
• Cerebral autosomal dominant arteriopathy with subcortical infarction and leukoencephalopathy.

Investigations

• Should only be necessary if headache is suspected to be secondary to another disorder.
• 'Alarm symptoms' include :
  - Onset above age 50 years.
  - Aura without headache.
  - Aura symptoms of acute onset without spread.
  - Aura symptoms that are very brief or unusually long.
  - Aura symptoms that are stereotyped.
  - Sudden increases in migraine frequency or change in migraine characteristics.
  - High fever.
  - Abnormal neurologic examination.
• The role of imaging in patients with suspected migraine is to exclude structural causes for the headache such as AVMs or tumors. A contrast enhanced CT scn is satisfactory for this, and is usually normal. If MRI is performed, the T2W image occasionally shows areas of altered signal in the white matter which may be residual ischemic changes following a recent or prolonged attack, or may rarely be due to CADASIL. The appearance is non-specific however.

Diagnosis

• At least 5 attacks.
• Attacks last 4-72 hours if untreated or unsuccessfully treated.
• At least two of :
  - Unilateral headache.
  - Pulsating headache.
  - Moderate or severe headache.
  - Headache aggravated by routine physical activity.
• At least one of :
  - Nausea, with or without vomiting.
  - Photophobia.
  - Phonophobia.
• Treatment
  Avoid precipitating / triggering factors
  Identify these by keeping a dairy if necessary.

Treatment of the acute migraine attack
Ancillary measures

• Rest in a quiet dark room.
• Intravenous fluids if severely dehydrated.

Non-specific analgesics and antiemetic / prokinetic compounds

Treat as early as possible, and wait 40 minutes. If headache persists, try a specific treatment such as ergotamine 1 mg capsule or sumatriptan 50 mg tablet, and wait > 1 hour. If headache persists, repeat.

Anti-emetic and prokinetic compounds if nausea and vomiting are a problem. Metoclopramide is preferable because it improves the oral absorption of other drugs and may have a favorable central effect. If vomiting is severe, suppositories of domperidone, prochlorperazine, or chlorpromazine may be helpful.

Simple analgesic drugs :
- Aspirin 2 or 3 x 300 mg chewable tablets orally.
- Paracetamol 2x500 mg tablets orally.
- Compound codeine - containing analgesic but may cause or exacerbate nausea.

Non-steroidal anti-inflammatory drugs :
- Ibuprofen.
- Naproxen : oral, rectal.
- Diclofenac : oral, intramuscular.
- Ketorolac : intramuscular.

Other non-specific drugs :
- Chlorpromazine : intramuscular, but long term considerations.
- Narcotic analgesic use is highly controversial, not evidence-based, and is associated with prominent adverse effects and a high risk of dependency. Most patients who require narcotics are misusing analgesics or ergots.
- Lignocaine infusion : may be indicated for prolonged severe migraine unresponsive to other therapy or for rebound headache. Procedure: a 12-lead EKG is obtained and examined before and 30 - 60 minutes after starting the infusion. Lignocaine is delivered by a pump device at a rate of 2 mg/ min. The patient is attached to a bedside cardiac monitor, and a rhythm strip is obtained every 5 minutes for the first 30 minutes, then every 15 minutes for 3 hours, and thereafter every 2 hours. Pulse rate and blood pressure are measured every 5 minutes for 3 hours, and thereafter every 2 hours while the patient is awake. The infusion is maintained until the patient has been headache free for at least 12 hours. The duration of infusion should not exceed 14 days. Contraindications include significant heart disease, epileptic seizures, or allergic reaction to lignocaine.

Specific antimigraine agents

Ergot alkaloids :

• Alpha adrenergic agoinsts with potent 5-HT1 receptor afinity. Also stimulants of dopamine D2 receptors in brainstem and gut, and vascular adrenergic and 5-HT2 receptors
• Ergotamine exists in several forms :
  - 1 mg capsule of ergotamine tartrate, or combined with caffeine.
  - 2 mg tablet of ergotamine tartrate combined with caffeine and an antiemetic.
• Dihydroergotamine mesylate can also be given orally, or more effectively, by suppository, inhalation, sublingual, intramuscular and intravenous routes.
• Effective in about half of cases.
• If two doses at intervals of 2 - 6 hours are ineffective, no more should be given for that attack.
• Limitations include poor oral and rectal bioavailability ; frequent, long-lasting adverse effects; and risk of headache and ergotism with chronic recurrent use.
• The drug of choice in a limited number of migraine sufferers who have infrequent or long duration headaches and are likely to comply with dosing restrictions. For most migraine sufferers requiring specific migraine treatment, a triptan is generally a better option from both an efficacy and side-effect perspective.

Triptans :

Selective and potent agonists of 5-HTIB, ID, IF, and to some extent 5-HTIA receptors. Antimigraine effects are mediated by :

• Inhibition of firing of cells in trigeminal nuclei.
• Inhibition of dural neurogenic inflammation and plasma extravasation.
• Vasoconstriction of meningeal, dural, cerebral or pial vessels.

First - generation triptans : sumatriptan :

• A specific and selective agonist of 5-HTID presynaptic receptors on cranial blood vessels, inhibiting trigeminal neuronal firing at the trigeminal nerve ending.
• Available as subcutaneous injection, oral tablets, nasal spray, and rectal preparations.
• Subcutaneous sumatriptan injection :
  - Bioavailability : 96%.
  - Therapeutic plasma levels : within 10 minutes.
  - 79% of patients improved at 2 hours after injection ; 71% improved within 1 hour.
  - 60% of patients pain-free at 2 hours after injection; 43% pain-free after 1 hour.
  · Oral sumatriptan tablets :
  - Bioavailability : 14%.
  - Therapeutic plasma levels : within 30 - 90 minutes.
  - 59% of patients improved at 2 hours after tablets.
  - 29% of patients pain-free at 2 hours.
• Oral sumatriptan is more effective than conventional treatment with aspirin and metoclopramide or oral ergotamine plus caffeine, particularly in the second and third attacks, suggesting greater consistency for sumatriptan.
• Recurrence of headache occurs iwthin 24 - 48 hours in about one-third of responders to sumatriptan. Repeated drug administration is usually effective, but he headache may recur again.
• Adverse effects of sumatriptan are common but are usually mild and short-lived. The most frequent are tingling, pareshtesias, and warm sensations in the head, neck, chest, and limbs; less frequent are dizziness, flushing, and neck pain or stiffness. The risk and intensity is greater with the fixed subcutaneous formulation. 'Chest-related symptoms' include short-lived heaviness or pressure in the arms and chest, shortness of breath, chest discomfort, anxiety, palpitations, and, very rarely, chest pain. The mechanism is unknown. The risk of sumatriptan-induced myocardial ischemia in the absence of coronary artery disease appears to be acceptable.

Second -generation triptans :

• Zolmitriptan 2.5 mg, 5 mg : similar to oral sumatriptan.
  
• Naratriptan 2.5 mg : slower action and perhaps fewer and less severe adverse effects, but lower efficacy.
• Rizatriptan 10 mg, 40 mg : better efficacy and consistency, and similar tolerability.
• Almotriptan 12.5 mg : similar efficacy, better consistency and tolerability.
• Eletriptan 20 mg, 40 mg and 80 mg : 80 mg orally is more effective than sumatriptan 100 mg orally with similar consistency but lower tolerability.
• Frovatriptan : possibly lower efficacy than oral sumatriptan.

Advantages over oral sumatriptan :

• Higher bioavailability : 45-75%.
• More rapid therapeutic plasma levels : within 30 - 60 minutes.
• Greater potency at 5-HTIB / D receptor sites.
• Increased lipophilicity and brain penetratiion (hence, direct attenuation of excitability of cells within the trigeminal nuclei of the brainstem, as well as vasoconstriction and peripheral inhibition of trigeminal perivascular terminals).
• Cheaper alternatives for patients who do not respond to oral sumatriptan.

None of these agents is consistently effective in all patients and all attacks, and some cause disturbing adverse effects. Rizatriptan 10 mg, eletriptan 80 mg, and almotriptan 12.5 mg provide the highest likelihood of success. Ergotamine and sumatriptan should not be prescribed for patients with suspected coronary artery disease, Prinzmetal variant angina, or uncontrolled hypertension.

PREVENTION

Non-Pharmacologic

• Avoid precipitating factors
• Stress reduction through relaxation exercises and tapes, meditation, yoga, swimming and similar strategies will reduce migraine frequency in many patients.
• Regular exercise such as swimming.
• Acupuncture in short courses by an experienced therapist can be a useful adjunct to other strategies in some patients.

Pharmacologic

• The indication for prophylactic therapy is when the patient needs it. This is usually when the migraine attacks are frequently interfering with their life and recurring every 2 weeks or so and not responding quickly and adequately to acute treatment.
• Efficacy is limited : at most about half of patients will have a reduction in attack frequency of half or more.
• Adverse effects occur commonly.
• The choice of prophylactic agent is primarily determined by the patient and which potential adverse effects are most acceptable.
• Discuss the adverse effect profile of each drug with the patient and determine their preference. Asthma and weight gain and are by far the major concerns.
• Establish realistic expectations with the patient before starting : the medication may reduce the frequency of attacks but uncommonly abolishes attacks, and so occasional breakthrough attacks requiring acute treatment will occur.
• Start slowly, with dosage increments every 7 - 10 days to minimize adverse effects.
• Encourage patients to persist for at least 3 months to adequately trial the drug and because most adverse effects become less prominent with time.
• Follow on with a drug free interval to reassess the frequency and severity of migraine attacks.

Menstrual migraine

• Try standard prophylactic therapy, as above, before hormone manipulation.
• Continuous bromocriptine therapy, 2.5 mg tds, added to the existing prophylactic regime may be beneficial. Adverse effects, such as light-headedness and nausea can be minimized by gradual introduction of medication, beginning with 1.25 mg daily and followed by daily incremental 1.25 mg increases over 1 week to full dosage.
• Non-steroids anti-inflammatory drug, such as diclofenac 50 mg bd commencing 24 hours before anticipated menstruation .
• Application of a gel containing 1.5 mg estradiol to the skin 48 hours before the expected onset of menstruation.
• Subcutaneous implantation of estradiol pellets, starting with 100 mg, inhibits ovulation and maintains estradiol levels, while regular monthly periods can by induced by cyclical oral progestogens. Depoprovera, a different oral contraceptive pill, or 3 monthly cycles of the oral contraceptive pill are alternative strategies.
• Tamoxifen citrate, 10-20 mg daily preceeding and during menstruation may help; tamoxifen competes at estrogen an anti-estrogen binding sites and is a calcium channel blocker. However, the anti-estrogenic effect in young women, such as osteoporosis, is an obvious problem with this strategy.

Clinical Course

• Migraine is paroxysmal. Clearly defined episodes of migraine recur as often as three or six times each month but sufferers remain symptom-free between attacks.
• The frequency of migraine attacks may increase until it develops into chronic daily headache, often as the result of stress or the over-use of ergotamine or analgesics.
• Migraine symptoms frequently change over time.
• The severity of the attacks often diminish with time and in some patients the attacks cease in latter years, particularly after the menopause in women.
• In some women however, attacks increase in frequency at the menopause.
• Remission occurs in 70% of pregnancies.
• For young women, below the age of 25 years, the relative risk of ischemic stroke among migraineurs is increased, particularly among those taking the oral contraceptive pill, but the absolute risk is extremely by small.

BACK

PERMANENT VEGETATIVE STATE (PVS)

Definition :

• A condition of 'wakefulness without awareness'.
• The absence of any adaptive response to the external environment and any evidence of a functioning mind which is either receiving or projecting information, in a patient who has long periods of wakefulness.

The vegetative state

• A clinical condition of unawareness of self and environment in which the patient brethes spontaneously, has a stable circulation, and shows cycle of eye closure and eye opening which may stimulate sleep and waking.
• May be transient stage in the recovery from coma or it may persist until death.

The continuing or persistent vegetative state

• A diagnosis based on a high degree of clinical certainty that the continuing vegetative state is irreversible; usually a continuing vegetative state for more than 12 months after head injury and more than 6 months following other causes of brain damage.
• Avoiding this state is one of the most important aspects of attempting to assess prognosis early in coma.

Epidemology

Prevelance : 10 000 to 25 000 adults and 400 - 1000 children in the USA.

Pathophysiology

• Several damage to part or all of the cerebral hemispheres, with an intact brain stem.
• Can occur with damage to the more rostral part of the brainstem.

Pathology

Three main patterns :

• Diffuse axonal injury, typically as a sequel to severe closed head trauma, giving rise to degeneration of the white matter throughout the cerebral hemispheres.
• Extensive laminar necrosis of the cerebral cortex, following global cerebral hypoxia or ischemia.
• Thalamic necrosis, occasionally.

Etiology

• Head injury
• Hypoxic-ischemic encephalopathy : cardiac arrest, carbon monoxide poisoning.
• Stroke : ischemic or hemorrhagic.
• Hypoglycemia.
• Intracranial infection.
• Brain tumor.
• End stage of degenerative brain disorders : Alzheimer's disease.

Clinical Features

• Inattentive and aware of the surroundings but breathes spontaneously, without mechanical support, and has stable circulation.
• At times the eyes are closed and the patient appears asleep, at other times the eyes are open and they seem to be awake.
• May be aroused by painful or prominent stimuli, opening the eyes if they are closed, increasing the respiratory rate, or occasionally grimacing or moving the limbs.
• When the eyes are open the eyelids may blink in response to any threat to the eye.
• A range of spontaneous movement may occur such as roving eye movements, chewing, teeth-grinding, groaning, grunting and even swallowing is possible. More distressingly, patients may smile, shed tears, moan or scream, without any discernible reason. Sometimes, the head and eyes turn fleetingly to follow a moving object or sound.
• Body posture may be decorticate or decerebrate.
• Brainstem reflexes, oculocephalic are usually preserved.
• Primitive reflexes such as pouting and sucking reflexes, grasp reflex, and withdrawal reflexes to pain may be present.
• Painful stimuli may provoke an extensor or a flexor response.
• Plantar responses are commonly extensor.

Diagnosis

Two medical practitioners experienced in assessing disturbances of consciousness and awareness should separately assess the patient (this includes discussion with other medical and nursing staff, relatives and carers about the reactions and responses of the patient, and to ensure that the patient is not sentient) and document their findings and conclusions in the medical record. If there is nay uncertainty about the diagnosis of the permanent vegetative state, then a re-assessment should be undertaken at a later date.

Diagnostic Criteria

Preconditions

• A cause for the syndrome has been established
• Persisting effects of sedative, anesthetic and neuromuscular blocking drugs have been excluded by the passage or time or by appropriate analysis of body fluids
• Reversible metabolic causes have been corrected or excluded as the cause.

Clinical criteria

All three of

• No evidence of awareness of self or environment at any time. No volitional response to visual, auditory, tactile or noxious stimuli, and no evidence of language comprehension or expression.
• Cycles of eye closure and eye opening which may stimulate sleep and waking shall be present.
• Hypothalamic and brain stem function is sufficiently preserved to ensure the maintenance of respiration and circulation.

Other clinical feature include :

• Spontaneous blinking.
• Inconsistent, non-purposeful reflex movements in response to external stimuli :
  - Apparent smiling.
  - Facila 'grimacing' to painful stimuli.
  - Watering of the eyes,
  - Startle myoclonus
  - Occasional movements of the head and eyes towards a peripheral sound or movement.
  - Purposeless movement of the limbs and trunk.
• Retained pupillary and corneal responses
• Absence of visual fixation and ability to track moving objects with the eyes or show a 'menace' response.
• Roving eye movements may be present.
• Conjugate or dysconjugate tonic eye movement, without corrective saccades in response to ice water caloric testing.
• Variable deep tendon reflexes and plantar responses.
• Clonus and other signs of spasticity may be present.
• Incontinence of bladder and bowel.

Differential Diagnosis

• Locked in syndrome : a brainstem a lesion disrupts the voluntary control of movement but arousal and the content of movement but arousal and the content of awareness are not abolished. Patients are able to communicate by movement of the eyes or eyelids.
• Coma : most patients who are in 'coma' for weeks, months or years are in a continuing vegetative state.
• Brain death : implies the irreversible loss of all brainstem functions. It is in a sense, the converse of PVS, in which brainstem functions survives while the function of the cerebral hemispheres is lost or gravely impaired. Brain death is followed, within hours or days, despite intensive care, by cardiac arrest.
• Akinetic mutism : a state of profound apathy with evidence of preserved awareness and attentive visual pursuit, giving an unfulfilled 'promise of speech'. The responsible lesions often involve the medical frontal lobes.
• Psychogenic unresponsiveness.

Investigations

• CT or MRI head scan : may show non-specific focal or diffuse abnormalities, brain atrophy, and hydrocephalus.
• Single photon emission computed tomography (SPECT) and positron emission tomography )PET) : show a reduction in cerebral metabolism.
• Neurophysiology studies : EEG, somatosensory evoked potentials (SSEPs), and electrodermal techniques ; very variable results. Often these studies provide evidence of some cortical activity, reminding us that an accurate clinical diagnosis of vegetative state does not imply cortical silence. They are of little value in predicting outcome, other than early EEG evidence of burst suppression and SSEP evidence of an absent N20 potential.

No findings are diagnostic of the permanent vegetative state.

Prognosis

• Determined by the age of the patient, underlying cause and the duration of the vegetative state.
• The outlook is better in children, and after traumatic brain injury. If the cause is not head injury, there is very little hope of recovery of sentience after 3 months and none after 6 months.
• If the cause is head injury, a longer time should elapse before being confident that the chances of recovery are extremely low.

Management

Establish the diagnosis of a permanent vegetative state by (1) identifying the clinical state of the patient, (2) the cause for the syndrome, and (3) the lapse of time. Because many patients entering a vegetative state emerge from it within a few weeks or months, supportive early management is usually appropriate. The diagnosis of a permanent vegetative state implies that recovery cannot be achieved and further therapy is futile. It merely prolongs an insentient life for the patient and a hopeless vigil for relatives and carers.

Medical care

• Appropriate nursing or home care
• Maintain oxygenation, circulation and nutrition.
• Correct complicating factors such as infection and hypoglycemia.
• Sensitive discussion with, education of, relatives and carers about the cause, clinical state, 'hopeless' prognosis, artificial means of administering food and fluid.
• Decisions to withdraw nutrition, hydration and other life-sustaining medication such as insulin for diabetes, should currently be referred to the court before any action is taken.
• Decisions not to intervene with cardio-pulmonary resuscitation or prescribed antibiotics are clinical decisions, but they should take account of, and respect, the views of the relatives, carers and patient, if known, whether formally recorded in a written document or not. \
• The role of sensory stimulation remains uncertain, despite the writings for Hippocrates : 'the patient in state of coma should be spoken to in a loud voice, splashed with cold water and exposed to bright light'.

BACK

Ageing and the Brain

Ageing of the brain is an inevitable and natural process and is not necessarily accompanied by an intellectual impairment. Often, however, the brain undergoes considerable macroscopic and microscopic changes that are invariably accompanied by alterations in neurochemistry and function. A clear answer as to a possible link between ageing of the brain and pathological processes that underlie dementia, particularly Alzhemier's disease, has not yet been established. However, in many ways Alzheimer's disease can be considered as an acceleration of the ageing process.

Macroscopic changes, brain weight and volume

At postmortem the dura, which is often very densely adherent to the calvaria, is slack because of reduction in the size of the brain. Over the age of 60 years, there is often thickening of the lepotmeninges, particularly in the parasagittal region where they often have a distinctly gelatinous appearance. An inconstant feature is prominence of the arachnoid granulations which may in the posterior frontal and parietal regions produces shallow identations of the inner table of calvaria.

A degree of cerebral atrophy is commonly present over the age of 60 years but is not an invariable accompaniment in the ageing process. When present, there is narrowing of gyri and widening of sulcei, particularly at the vertex in relation there is also an excessive amount of cerebrospinal fluid.

The weight and volume of the normal brain are maximal between the ages of 15 and 60 years; thereafter, there is a gradual onset of this atrophic process beginning earlier in women than in men. After the age of 50 years this loss of brain weight amounts of approximately 2 - 3% per decade over the next four decades. An assessment of the amount of atrophy can be guaged from the ration of brain to skull volume which remains constant at about 95% upto the age of 60 years. Thereafter, even in intellectually normal people, there is variation and this ratio may fall to 80% by the 10 th decade.

The adult brain weighs between 1200 and 1600 g with an average of 1400 g in men and 1250men and 1250 g in women. The weight remains fairly constant throughout middle age, but after the age of 65 years it tends to decline, the mean loss being about 100g. As the Intracranial volume is not routinely measured post mortem, a considerable reliance is placed on the weight of the brain at autopsy as an index of cerebral atrophy. However, it should be noted that both weight and volume change during fixation in 10% formol saline, both increasing by some 10% over a 2 - 3 week period.

An additional difficulty in interpretation of volume and weight loss in the ageing brain in the recognition of the so called secular effect, as a result of which mean body heights and brain weights have increased progressively over the past 50 - 100 years. In some elderly subjects, therefore, a reduction in brain weight may atleast in part be due to the individual having had a small brain. The best way of assessing the true significance of a reduced weight is to determine the ration between brain volume and intracranial cavity volume, as both are closely related to each other between the ages of 15 and 60 years. A useful index as to whether or not a brain is atrophied is to measure the ration between the weight of the whole fixed brain and that of the whole fixed hindbrain the former usually being between eight and 10 times greater than the latter.

In vivo imaging of the ageing brain has shown widening of sulci and some enlargement of the ventricular system. However, hydrocephalus is not a consistent finding. Sometimes in the small brain the ventricles maintain their normal relative size, while in other cases of cerebral atrophy, the brain is not much reduced in size yet the ventricles are enlarged. This is one of the reasons why it is not possible to make a confident diagnosis of normal pressure hydrocephalus post mortem. In general, however, the volumes of the lateral and third ventricles increase progressively with age, rising from a mean of about 15ml in teenagers to 55ml over the age of 60 years. Likewise, autopsy studies have shown that the size of the ventricles increases with age, particularly above 60 years, and that in general the ventricles of the elderly are larger than those of young subjects.

In the absence of obstruction, increases in the volume of the ventricles and subarachnoid space are indicative of a reduction in the volume of the brain, but the precise contributions of this atrophy by grey and white matter have been difficult to establish, although increasing precision is being obtained by quantitative techniques. For example, it has been calculated that there is a progressive reduction in hemispheric volume from the age of 20 years with a greater than in women. Initially, there is a greater reduction in the volume of grey matter than in white matter, but thereafter more in white than in grey. This greater loss of white matter is probably accounted for by loss of neurons, with their myelinated axons having a greater volume than the space occupied by the cell body and dendrites. Such changes are clearly of importance when considering the neuropathology of dementia, and although in Alzheimer's disease there is generalized atrophy of the cortex with commensurate enlargement of the ventricular system, there are undoubted exceptions, particularly in the very elderly when the changes may be no greater than those found in intellectually normal subjects.

Further consideration will be given to the vascular changes found in the ageing brain, but at this stage it should be noted that small recent or old infarcts are not infrequently present in the brains of intellectually normal subjects. This should indicate the need for great caution in attributing dementia, even in the oldest subjects, solely to one or a small number of cerebral infarcts.

Microscopic changes

On histological examination there may be a variety of changes affecting neurones, glia and blood vessels. The findings, however, are by no means constant; for example, there is variable loss of neurones and special silver impregnations have demonstrated that there is often some loss of dendritic processes in the intact neurones. An increased amount of lipofusion in neurones is the rule. An almost invariable finding is an increase of astrocytic nuclei in the subcortical digitate white matter, but staining of myelin is usually preserved. Other features are a subpial and subependymal astrocytosis, and the presence of numerous corpora maylaceae throughout the central nervous system, but particularly, immediately deep of the ependyma and throughout the central nervous system, but particularly, immediately deep of the ependyma and throughout the spinal cord. Commonly, there are deposits of calcium in the walls of small blood vessels in the basal ganglia. Other more specific abnormalities include the presence of neuritic plaques in grey matter, the occurrence of occasional examples of granulovacuolar degeneration, neurofibrillary degeneration and Hirano bodies. Some of these changes are described in more detail below.

Changes in neurones

It seems reasonable to presume that marked atrophy associated with ventricular enlargement in the ageing brain is due to changes in the numbers of nerve cells and their size. However, the assessment is difficult because of variability in the same region in different individuals. However, certain trends can be identified in which particular nuclei are prone to atrophy with increasing age, that is, particular areas of the cerebral cortex and Purkinje cells in the cerebellum, whereas others nuclear groups retain their original numbers and size, that is brainstem nuclei.

NEURONAL LOSS

Many of the findings are still uncertain because of wide variations in the observations from different studies. However, it seems certain that some neuronal populations show cell loss and shrinkage. Others may show shrinkage without significant loss and yet others show neither shrinkage nor loss. The development of computerized image analysers, particularly when combined with editing capability, has allowed more precise studies to be undertaken from which it has been concluded that there is a loss of neurones in some areas of the neocortex with increasing age. In spite of wide variations in total neuronal counts, it has been concluded from a detailed study in a group of elderly patients that there was a 10% reduction in nerve cell counts and that in addition, more consideration had to be given to neuronal shrinkage. More recently, further consideration has been given to the fact that neuronal shrinkage and not neuronal loss is one of the more significant changes which occur in the neocortex of the ageing brain.

Quantative studies have also been carried out on brain areas other than the neocortex. For example, it is generally accepted that there is a loss of pyramidal cells in the hippocampus with increasing age. There is, however, some difference of opinion as to the amount of this loss and whether all sectors of the hippocampus are equally involved. For example, it has been calculated that there is 31% loss of neurones in the hilum of the dentate fascia and 52% loss of neurones in the subiculum between the ages of 13 and 85 years, and the neuronal population within the CA1 sector of the hippocampus reduces by between 3.6 and 6.2% per decade.

In view of the likely importance of the subcortical nuclei that produce neurotransmitters required in cognition, various attempts have been made to compare changes in similar sites between age- and sex - matched controls, and in dements. One area of particular interest is that of the basal nucleus of Meynert, which is the main source of cholinergic fibres to the cortex. Differences of opinion exist, however, ranging from minimal neuronal loss in adult life to a steady decline with age.

The hindbrain is not exempt from changes either. For example, considerable loss of Purkinje cells has been found in the cerebellum of individuals over the age of 60 years. Within the brainstem some nuclear groups, such as the inferior olive and the nuclei of the abducent and trochlear nerves, are stable with increasing age, whereas others such as the locus ceruleus - the main supply of cortical noradrenergic fibres - show variable loss.

ALTERED DENDRITIC PATTERN

Loss of function may be due not only to neuronal loss, but also atrophy of the soma and the dendritic tree, as well as loss of synapses. The use of human postmortem material has created difficulties in interpretation largely because so-called dendritic changes in the elderly have also been seen in young control subjects. Nevertheless, a possible progressive series of changes has been described in the dendritic trees of the pyramidal cells in the cortex of the temporal and frontal lobes. These include loss of dendritic spines, swellings, varicosities and distortions of the horizontal branches, followed by progressive swelling of the cell body, loss of basal dendrites and of branches of the apical shaft, and terminal branches. Finally, the apical shaft is lost, the cell body disappears and there is an astorcytosis. Use of the Golgi-Cox method, which is relatively free from postmorterm artefact, has shown that the pyramidal neurones of the parahippocampal gyrus can undergo considerable dendritic growth in normal old age: a similar conclusion has been reached for pyramidal cells of the cortex. The application of immunocytochemistry has identified changes in synaptic density with age. For example, a 20% reduction in presynaptic terminals of subjects over the age of 60 years has been established using antibodies to synaptophysin. Such studies serve to confirm the general observation of a 20% reduction in presynaptic terminals in the frontal cortex with ageing by electron microscopy. Such studies have suggested that despite such losses, the synaptic density may be maintained by the remaining neurones undergoing sprouting. An additional mechanism by which function is maintained in some regions of cortex is an increase in synaptic contact length for those synapses that remain.

CHANGES IN PIGMENTATION

Lipofuscin appears to increase in amount in neurones at certain sites. It is first seen in the inferior olivary neurones in infancy and may be found in the spinal cord of children. The amount tends to rise with increasing age, particularly in the cranial and spinal motor nuclei, the red nucleus, parts of the thalamus and globus pallidus, and in the dentate nucleus of the cerebellum. In contrast, relatively small amounts of lipofuscin are seen in the cells of the occipital cortex and in purkinje cells of the cerebellum.

Lipofuscin is a cytoplasmic organelle which measures about 1µm in diameter and is a type of lysosome in which non-metabolizable substances accumulate: it stains red with Sudan dyes and is periodic acid-Schiff (PAS) positive. It is membrane - bound and usually has an electron - dense and electron - light component, which presumably is responsible for its light yellow colour. It is a normal organelle and it has been suggested that as the amount of lipofuscin increases there is a loss of Nissl substance followed by marked reductions in cytoplasmic RNA, sufficient perhaps to lead to cell atrophy and death. However, there does not appear to be any relationship between the amount of lipofuscin and neuronal loss, as loss of neurones in old age is also common in sites where lipofusin is present in only small amounts. That there may indeed be a relationship between the accumulation of lipofuscin and loss of function is suggested by the finding that there is an excessive amount of lipofuscin in the brains of dements compared with matched controls.

In addition to lipofuscin there is an accumulation of neuromelanin with ageing, particularly in the substantia nigra and locus ceruleus. As the amount of neuromelanin increases with advancing years, there may be up to a 50% loss of pigmented neurones as part of the normal ageing process. The neuronal loss appears to be greatest in those containing the most pigment, but whether or not the accumulation of the pigment is directly associated with changes that presage cell death is nor clear.

A more recently described change is an increase in the prevalence of granular bodies that react with antibodies to ubiquitin. These bodies are located principally in the medial portion of each temporal lobe and they are considered to be located in dystrophic neurites. Immunoreactive structures that are again ubiquitin positive are found increasingly in the glia of white matter. Other changes that have been described include an altered expression of phsophorylated neurofilament antigen and an increase of a - and ß - crystallin cytoplasmic molecules thought to act as a chaperone and as a stabilizer of the cytoskeleton.

ALZHEIMER'S NEUROFIBRILLARY DEGENERATION

This intraneuronal degenerative change is difficult to see in haematoxylin and eosin-stained sections. Other techniques have therefore been employed, one of the more successful being that of Congo red, which stains tangles a deep pink colour and at the same time renders them birefringent under polarized light. Various silver impregnation techniques also readily identify them and, more recently, antibodies to the various constituents of the tangle have been employed. The great majority of tangles are particularly well demonstrated by antitau antiserum.

By light microscopy the configuration of the tangles is largely determined by the site and type of neurone affected. For example, in the small pyramidal neurones of the cortex, tangles are seen to extend from the base of the cell towards the apical dendrite. In larger pyramidal cells, many resemble a skein of wool, whereas tangles in the hippocampus may have a more complex configuration. In subcortical structures, including the upper brainstem, globoid forms are commonly seen. As the tangle enlarges, the nucleues and any pigments become displaced, until eventually the boundaries of the cell become ill defined and only the tangle remains. In these circumstances the tangle often acquires additional immunoreactive properties staining with ß-A4 amyloid.

Electron microscopic studies have shown that neurofibrillary tangles are made up of filaments that measure 20 mm across with a regular constriction of 10 mm occurring every 80 mm. Although initially thought to be twisted tubules, later studies showed that the appearances were due to paired filaments wound in a double helix. Although tangle formation in normal old age consists predominantly of paired helical filaments they may also be associated with straight tubules within the same neurone or straight filaments. Paired helical filaments derive from components in the normal neuronal cytoskeleton, containing not only sequences from neurofilaments and microtubule-associated proteins, but also antigenic determinants which are unique to them. Current evidence suggests that the abnormal phosphorylation of the tau protein could play an important role in tangle formation.

Neurofibrillary tangles not only occur in normal ageing but also in Alzheimer's disease and in a variety of other neurodegenerative disorders, such as progressive supranuclear palsy. Within the normal ageing process, they are uncommon in non-demented subjects, being found in greatest numbers in the corticomedial portion of the amygdaloid nucleus and in the cortex of the anteromedial part of the temporal lobe: the number increases with age. Particular emphasis has been placed recently on the occurrence of neurofibrillary tangle formation in the entorhinal cortex in normal ageing and the fact that it is unusual to find them to any great extent in the neocortex of normal old age in the absence of dementia.

Neurofibrillary tangles are a prominent feature of Alzheimer's disease, but are also found in adults with Down's syndrome where they are present in most subjects who are over 30 years of age. They are also present in large numbers in the parkinsonian demmentia complex of Guam and in the amyotrophic lateral sclerosis of Guam, as well as in other instances of motor neurone disease. They are also found in subacute sclerosing panencephalitis, head injury and dementia pugilistica. In all these disorders the neurofibrillary tangle shows the same configuration of paired helical filament. The only exceptions are certain instances of motor neurone disease and in progressive supranuclear palsy, in which conditions the tangles consist predominantly of straight fibrils rather than paired helical filaments.

The nature and origin of the tangles remain unclear, although it has been suggested that tangles contain the same protein as the amyloid of blood vessels and plaques, despite the fact that ultrastructurally the neurofibrillary tangle does not resemble amyloid fibrils. Conversely, immunocyto-chemical studies indicate that tangles share antigenic determinants for neurofilaments with microtubules and tau protein. These studies tend to suggest that neurofibrillary tangles form as a result of defective assembly of micro-tubules and/or neurofilaments, which result from abnormal phosphorylation of tau or neurofilaments.

GRANULOVACUOLAR DEGENERATION

This change, which is largely restricted to the pyramidal cells of CA1 of the hippocampus, consists of one or more intracytoplasmic vascuoles measuring some 2 - 5 m in diameter. Multiple vacuoles are common when they may displace the nucleus and normal cytoplasmic organelles: they occasionally occur in association with neurfibrillary tangles. They are easily seen in haematoxylin and eosinstained sections but are strikingly obvious when present in silver-stained preparations. Electron microscopy shows a dense granular core embedded in a translucent matrix, which in turn appear to be separated from the rest of the cytoplasm.

Immunocytochemical studies have shown that some of the granules react with antibodies to phosphorylated neuro-filaments, tubulin, tau and ubiquitin. Such an antigenic profile suggests that the vacuoles are autophagic structures in which cytoskeletal components are being degraded.

Quantitative studies have shown that granulovacuolar degeneration is uncommon before the age of 65 years, but its frequency increases even in non-demented subjects, to the extent that it is present in some 75% by the ninth decade. The number of pyramidal cells in the hippocampus showing this change increases to some 20% in demented subjects. They have been described in other diseases, including young adults with Down's syndrome and in the amyotrophic lateral sclerosis and parkinsonian dementia complex of Guam. Granulovacuolar degeneration may also be present in tuberous sclerosis. Whereas in these conditions the changes are limited to the hippocampus, granulovacuolar degeneration is present in the nuclei of the brainstem in progressive supranuclear palsy.

HIRANO BODIES

In a haematoxylin and eosin-stained sections avoid structures measuring between 10 and 30 m in length and 9 m across may be seen easily, although often they are mistaken for columns of red cells. They are present most commonly in the pyramidal cells of the hippocampus. Up to middle age only the occasional body is seen, but late in life they are numerous. Since their original description they have been found in a variety of disease including Pick's disease, and they are particular abundant in Alzheimer's disease.

Electron microscopically they are made up of parallel filaments 60 - 100 m in length which alternate with lengths of sheet-like material. Immunohistochemistry has shown that Hirano bodies share epitopes for actin and the actin-associated proteins tropomycin, - actionin and vinculin and that a small proportion of the bodies also react with antibodies to tau protein, These observations suggest that Hirano bodies result from an abnormal configuration are microfilaments.

Changes in the neuropil

THREADS

Histologically, these are thread - like structures found in the neurophil of grey matter. In normal ageing they are usually restricted to structures in the medial parts of the temporal lobes. They are found in the dendrities of neurons that contain neurfibrillary tangles and electron microscopically they contain straight straight tubules. Immunocytochemically, their profile is similar to that of neurofibrillary tangles.

NEURITIC PLAQUES

Known also as dendritic and amyloid plaques, these name emphasize the two most striking components of may plaques found in old age, since most consists principally of a central core of amyloid-like material surrounded by swollen abnormal neurites. Large numbers of these structures are found in the brains of patients with Alzheimer's disease, and small numbers of so-called neuritic plaques and large numbers of related non-neutritc plaques are found in the brains of non- demented older people

The appearance of plaques differs depending upon the staining method used. For example, they are difficult to see in haematoxylin and eosin-stained preparations, but are easily seen in either frozen or paraffin-embedded sections using silver impregnation techniques. The amyloid component of these plaques can be seen readily by Congo red or thiflavin S techniques. More recently, immunohiso-chemistry has revealed immunostaining for - A4 amyloid protein in non - neuritic plaques in many aged non-demented subjects. Similarly, - A4 amyloid plaques have also been demonstrated in dementia pugilistica and progressive supranuclear palsy, two conditions that had previously been thought to be characterized by neurofibrillary tangle formation in the absence of plaques.

A classic neuritic plaque measures between 5 and 20 mm in a diameter and in silver-stained preparations consists of a dark central core surrounded by an irregular clear halo, beyond which there is an envelope of granular filamentous or rod-like structures, which like the core, are argophilic. Their size and configuration may vary depending on the state of development of the plaque and the plane of section. Plaques are commonly discrete, whereas in other areas they appear to fuse together into large irregularly shaped structures and less commonly, the appearances may be of ill-defined areas of faintly granular background. Ultrastructural studies have shown that the central core is composed of amyloid fibrils which stain positively with Congo red, while the outer rim consists of a mixture of abnormal distended neuritic processes intermingled with astrocytes and microglia. Since the earliest studies it seems likely that the central part of the plaque always contains fibrils of amyloid, but many of the larger neuritic processes contain paired helical filaments, and that neuritic processes may well originate from dendritic sprouting of neurones affected by tangle formation.

Three stages in the light microscopic formation of a neuritic plaque have been described. In the first stage the primitive plaque consists of abnormal neurites intermingled with the fibre-forming astrocytes and microglial cells. The second stage is the mature plaque which has all the typical constituents of the central core of the amyloid, neurites, microglia and astrocytes. The last stage, which is the 'burn-out' plaque, is composed mainly of amyloid. Recently another form of plaque has been described in which there is a diffuse deposition of amyloid unassociated with central compacted amyloid or abnormal unassociated with central compacted amyloid or abnormal neurites. These lesions have been called diffuse plaques, preamyloid deposits, senile plaque like structures and diffuse senile plaques. This type of plaque is commonly found throughout the cortical mantle occurring amongst other discrete classic plaques. They are also found, however, in areas of the brain where classical plaques are few, such as the brainstem and cerebellum. Such plaques have also been found in dementia pugilistica and in progressive supranuclear palsy - both conditions which previously had been thought to be characterized by neurofibrillarly tangle formation in the absence of plaques. The relationship between this diffuse type of plaque and classic plaque formation is not known.

Non - neuritic senile plaques are rarely found in the brains of normal young and middle-aged subjects. However, they are commonly present in small numbers in those over 60 years of age. Immunocytochemical techniques using antibodies to the -A4 peptide show evidence of plaque formation rising with age from about 20% in the sixth decade to 90 - 100% in centenarians. In the most mildly affected cases plaques are found in isolated areas in the cortex of the frontal and anterior temporal lobe and in the structures comprising the medial parts of the temporal lobes. Neuritic plaques, when present, tend to occur in the deeper layers of the cortex and in the non-neuritic cases in the superficial layers where they are located between vertically orientated clusters of neuronal apical dendrites. In a recent study Sparks et al. showed that brains from non-demented subjects were significantly more likely to contain senile plaques if the subject had died of coronary heart disease than those who had died from other causes. Some studies have found numerous neocortical plaques in elderly non-demented subjects sufficient to have met Khachaturian criteria forms of Alzheimer's disease, especially when examining material obtained after autopsy which has been incompletely assessed clinically.

Of possible importance in the light of what has been noted already is that when many plaques are seen in non-demented subjects, neurofibrillary tangles within cortical neurones were either rare or absent, whereas in the majority of cases of Alzheimer's disease, tangle formation is extensive throughout the cortex.

Changes in glia

Both astrocytes and microglia become more prominent in the human brain with increasing age. For example, age-associated increase in the numbers of astrocytes as demonstrated by glial fibrillary acidic protein immunohistochemistry becomes evident in the eighth decade. Such changes are found not only in cortex but also in subcortical structures and in relation to blood vessels and the ventricular system. Activated astrocytes elaborate in neurotrophic cytokine, S 100 protein, that has been implicated in the development of neuritic plaques in Alzheimer's disease. With normal ageing there are increased numbers of S 100 protein immunoreactive astrocytes in human cerebral cortex.

Microglia also show age-associated changes, activated forms expressing the cytokine interleukin - 1 being significantly increased in numbers in the brains of non-demented individuals with an age of 60 years. The numbers of enlarged cells with processes increase with age while no significant increase is seen in the number of non-enlarged, non-activated forms. Concomitant with these changes in microglial number and morphology, there are significant age-associated increases in tissue levels of interleukin - 1 messenger RNA. As the microglial over-expression of interleukin - 1 has been implicated in the pathogenesis of Alzheimer's disease, these age-associated increases may contribute to the increasing incidence of Alzheimer's disease with advancing age.

Corpus amylaceae are round, basophilic PAS-positive structures 5 - 20 m in diameter that lie within astrocytic processes in the subependymal and subpial areas especially in the basal ganglia, medial parts of the temporal lobe and posterior columns of the spinal cord. They are unusual in the first decade of life, but are universal by the age of 40 years. Although they are essentially identical to the polyglucosan bodies and the Lafora bodies, their significance in normal ageing remains obscure.

Changes in neurotransmitter activity

That cholinergic neurones of the central nervous system play an important role in learning and/or memory has been stimulated by two separate lines of evidence. Firstly patients with Alzheimer's disease sustain severe loss of cortical choline acetyltransferase; and secondly, there is atrophy of the basal nucleus of Meynert, and it is form the subcortical nucleus that most of the cholinergic innervation on the cerebral cortex originates. Such reports have led to the formulation of the cholinergic hypothesis that 'these disturbances play an important role in the memory loss and related cognitive problems associated with old age and dementia'.

Neurochemical and neuropathological studies have shown that there are neurotransmitter abnormalities other than in the cholinergic system. For example, there are deficits in both the noradrenergic and serotoninergic systems that corresond with the loss of noradrenergic cells from the locus ceruleus and of serotoninergic cells from the raphe nuclei. There have also been reports of reduction in certain neuropeptides.

These and other studies clearly show that there are disturbances in several neurotransmitter systems and that the cholinergic hypothesis of memory loss, even in the ageing process and Alzheimer's disease, may prove to be an over-simplification. Nevertheless, the neuroanatomical inter-relationship between the frontal cortex, nucleus basalis, hippocampus and amygdala has been implicated as contributing a functional role in the loss of memory in the elderly and Alzheimer's disease. It is to be expected, therefore, that in vivo imaging techniques such as positron emission tomography will play an increasingly prominent role in the demonstration of regional cerebral metabolic abnormalities and their correlation overtime with neuropsychological performance.

Vascular Changes

Vascular changes are found commonly in the ageing brain. However, very few quantitative studies have been undertaken, although comparison have been made on the amount and distribution of infarction between dements and non-demented old people. Cerebrovascular changes are common causes of admission to hospital in the elderly. The incidence of stroke rises rapidly with increasing age, some 80% of cases occurring in patients over the age of 65 years.

Atheroma

Quantitative studies have shown that in normotensive subjects atheroma does not usually affect cerebral blood vessels less that 2 mm in diameter, such as those supplying the basal ganglia and thalamus.

Hyaline arteriolosclerosis

Changes similar to those found in hypertension occur in the elderly, comprising fibrous replacement of muscle and fragmentation of the elastic tissue. As a consequence the walls of arteries become thickened, longer, more tortuous and rigid. In blood vessels of 1 mm diameter or less, in addition to hypertrophy of the media, the intima becomes thickened by a concentric increase in connective tissue. In the smallest arteries the intimal change predominates and may result in narrowing of the lumen. In contrast to the dilatation seen in the large arteries there may be hyaline thickening of arterioles which gradually extends over the whole circumference and when severe replaces all structures except the endothelium.

Lacunae

These are small cavities measuring between 3 and 20 mm in diameter occurring principally within the diencephalon and brainstem. Lacunae of small diameter are commonly seen in the basal ganglia of the ageing brain. Although some 90% of lacunae are associated with hypertension, they may be found in some 9% of normotensive subjects. They consist of expanded perivascular spaces consequent upon rarefaction and disintegration of the neuropil around the blood vessel. The cavity is normally empty, containing very few cells and is limited by a narrow band of astrocytosis. The term etat lacunaire is used for the cavities if numerous in the grey matter, and etat crible for similar cavities if numerous in the centrum semi ovale and other richly myelinated regions. The pathogenesis of lacunae is not clear, although a number of mechanisms are probably operating. For example, it has been suggested that they are the result of spiral elongations of small intracerebral arteries under the effects of raised blood pressure, whereas Fisher, in a series of publications, found that they could be attributed to occlusive vascular disease, the small lesions appearing to be due to lipohyalinosis and the larger to atheroma or emboli.

Microaneurysms

Micro (miliary) aneurysms have been demonstrated using radiological techniques in both hypertensive and normotensive patients post mortem. Such techniques have demonstrated microaneurysms in small arteries and are seen as out-pouching to the vessel wall. Such aneurysms are uncommon under the age of 50 years and occur most commonly in the brains of hypertensive patients. However, some microaneurysms have been observed in the brains of normotensive patients but not to the same extent as in elderly hypertensive. This suggests that microaneurysms form as part of the normal ageing process, but that this is accentuated by hypertension.

Infarction and leukoaraiosis

In many instances these are found at autopsy and appear not to have been associated with clinical signs or symptoms and in particular without any loss of, or deterioration in, higher mental function. Often these lesions are small and are most commonly found in the basal ganglia and brain stem. The larger the lesion the more of them there are, and their strategic location predisposes to the entity of multifarct dementia which is responsible for some 15% of cases of dementia over the age of 65 years. They also contribute to a further 10% of cases of dementia in association with Alzheimer's disease.

The term leukoaraiosis is used to describe changes in periventricular white matter seen on computed tomography in both demented and in elderly normal subjects. These changes are usually symmetrical and appear histologically as hyaline arteriosclerosis of blood vessels, astrocytosis and partial loss of myelinated axons and oligodendrocytes. These appearances are thought to represent incomplete infarction confined to white matter, possibly consequent to hypoperfusion due to vascular disease. some of these patients are demented, but many show no intellectual deficits. However, patients with leukoaraiosis are more likely to have had strokes in the past and are more likely to have strokes in the future, suggesting an association with vascular disease.

Congophilic angiopathy

Amyloid - like material occurs not only in the centre of neuritic plaques but also in the walls of small blood vessels, meninges and cortex, staining brightly eosinophilic in haematoxylin and eosin-stained preparations. If the complete circumference of the blood vessel is involved, the artery appears as a thickened homogeneous tube, the walls of which become congophilic when stained with Congo red. Such staining also makes it briefringent in polarized light.

Amyloid angiopathy is uncommon in normal subjects under the age of 60 years, the prevalence thereafter rising to about 30%. The cortex of the parietal and occipital lobes is more commonly affected than that of the frontal region, although rarely there is involvement of the brainstem. Changes in normal old age are usually mild and it is much more frequent in cases of Alzheimer's disease than in long surviving Down's syndrome.

Age - matched control brain

There is increasing awareness of the importance of selecting normal control material for tissue-based studies of neurological disorders of the elderly. A described above, some age-associated changes are clearly pathological, for example cerebral infarction, while others are not, for example the presence of diffuse plaques and neurofibrillary tangles and corpora amylacea. Therefore, the selection of material for any given study requires the application of inclusion/exclusion criteria that need to be defined at the outset.

Additional consideration include race and gender differences in the ageing brain. For example, the age-associated changes of plaques and tangles of normal elderly subjects is said to be similar in American Caucasian and East African black populations but is less frequent in the Chinese population of Hong Kong. There is also evidence that the vascular changes comprising leukoaraiosis and small vessel disease are more common in the brains of elderly Japanese, Chinese and American black than in white people. Differences in the brain between men and women have also been noted. It is difficult to identify sex-specific changes that are age associated, although there is some evidence that the neuroendocrine and neurotransmitter functions of certain nuclei in the hypothalamus are sex specific.

In the context of brain banks the importance of control material has been stressed. A recent review entitled 'What makes the brain bank go round?' details the most important aspects that need to be taken into accounting when collecting and providing postmortem brain samples for research. Not only should there be accurate matching for anti- and postmortem factors, but there should also be standardization of the brain area sampled, matched in addition of age and sex, agonal state, lateralization, seasonal variations, time of death, medication, postmortem delay, fixation and storage time. Although the analysis of postmortem human data is difficult, there is nevertheless an increasing appreciation that rapid freezing techniques of materials derived from patients who have died suddenly are eminently suitable for neurochemical and molecular biological techniques and as long as the confounding influences are recognized and built into the design of the study, comparison of disease tissue with appropriate matched controls yields worthwhile information.

Morphometric methods

In recent years there has been an increasing requirement to quantify changes based on reporducible objective and precise methods in order to allow comparisons between the normal and psychometric and neurochemical data derived from the diseased brain. Such methods have been reviewed with a particular emphasis on methods for measuring the volume of the cranial cavity, the brain and its component parts, and methods for estimating the volume fraction of numerical density of particular structures in microscopical sections. The more traditional methods have now largely been replaced.

BACK

Blood - Brain Barrier

The definition 'blood-brain barrier' is somehow restrictive, as it only emphasizes the role that the microvascular endothelial cells perform in the brain in preventing access to the nervous tissue. In recent years, however, it has become increasingly evident that the endothelial cells of the brain microvascular do not just isolate the brain from the blood, but play a pivotal role in maintaining the constancy of the internal milieu of the brain . This is achieved by preventing the entry into the brain of substances which could interfere with neurotransmission, but also by active regulation of the transport of those compounds which are essential for the maintenance of normal brain function. The aim of this chapter is to extend the concept of the barrier beyond that of an impermeable wall and to try to analyse the reason why the microvascular endothelium is essential for the maintenance of normal brain function.

Origin and development of the concept of a barrier

The observation, at the end of the nineteenth century, that 'something' was preventing the access of substances from blood to brain led tot he development of the concept of the blood-brain barrier. In the 1880s Ehrlich noticed that trypan blue injected intravenously stained most of the organs in the body excluding the brain. He concluded that the brain had low affinity for the dye he had injected. Whilst his experiment demonstrated the fundamental property of the barrier, his conclusions were simplistic. Goodman, 20 years or so later, proved that when the same dye was injected into the cerebrospinal fluid the brain became blue, but the other organs did not. This time the conclusion was reached that a permeability barrier separated the blood and the brain. In between those two crucial experiments much work continued showing the different effect of neurotoxins injected into the blood stream or directly into the brain. In particular. Lewandowsky studied the effect of sodium ferrocyanide, a known lipid-insoluble compound. This was non-toxic when injected intravenously, but highly toxic when injected directly into the brain. Perhaps he was the first to put forward the concept of a barrier as it is now under stood. However, almost 60 years passed before the anatomical basis of the blood-brain barrier could be fully elucidated.

Anatomy of the barrier

In the late 1960s a classical series of studies by Rese and Karnovsky and Brightman and Reese using electron microscopy demonstrated the fundamental anatomical properties of the brain endothelial cells which create the passive barrier isolating the blood from the brain interstitial fluid. This passive barrier results from the presence of tight junctions that seal together the endothelial cells impeding intercellular passage of substances. Furthermore, the brain capillaries do not have fenestrations, and they contain very few pinocytotic intracellular vesicles, suggesting that transcellular vesicular transport is almost absent.

While the site of the barrier resides in the particular properties of the endothelial cells, a number of specialized structures participate in the maintenance of the barrier complex. Around the capillaries a basement membrane, rich in collagen, supports the endothelial cells. Within the context of the basement membrane specialized cells, the pericytes, are in strict contact with the endothelial cells. The pericytes, which extend their processes around the capillaries, are contractile cells, but it is believed that under some circumstances they may have a phagocytic function. The perivascular cells, also in close contact with the basement membrane, are, most probably, pure phagocytic cells, as they are part of the resident brain microglia. The astrocytes, through their foot processes, are, however, the cellular elements which share most of the contact with the outer aspect of the basement membrane and endothelial cells.

This close relationship between the astrocytes and the anatomical side of the barrier has been the object the numerous studies which have suggested that astrocytes are essential in the indication and maintenance of the barrier properties. Both in Vitro and in vivo the presence of astrocytes induces endothelial cells to acquire properties of a permeability barrier. Conversely, in brain tumours changes in the astrocyte-endothelial relaitonship correlate with the well known alteration in the permeability of cerebral capillaries. The blood-brain barrier is not present in the entire brain as there are structures outside the barrier, such as the circumventricular organs and the pituitary gland. In these regions, specialized neurones, involved in the regulation of the hormonal systems, come in contact with the blood and can detect and respond to the levels of circulating hormones in the body.

Development of the blood-brain barrier

The belief that the blood-brain barrier develops during the early years of life has been amply revisited in recent years. Most of the difficulties in determining the timing of blood-brain barrier development have been generated by the different gestational periods of the various animal models, different time of barrier development in relation to the anatomical site, and the tracer and technique used ot assess permeability. The vascular plexus on the surface of the embryonic brain, which will give origin to the intracerebral capillaries, does not seem to possess blood-brain barrier properties. When the vessels migrate into the brain and the astroglial foot processes envelope the capillaries, the gap between the endothelial cells closes and the tight junctions develop. Although the tightness of the junction and the exclusion from the brain of plasma proteins appear very early, other barrier properties develop later at different times. The immature brain seems to be more permeable to small molecules and amino acids than the adult brain. Whether this differential permeability is just expression of the progressive maturation of the blood-brain barrier, or whether it responds to clear development needs, is not clear. There are suggestions, however, that the early isolation of the brain from plasma protein is essential for the proper development of neuronal connectivity, while continued permeability to small molecules would allow diffusion of essential substrates required by the fetal and neonatal brain. The endothelial proliferation and the induction of barrier properties appear to be different processes regulated by different factors. While the former is probably under the control of soluble angiogenic substances, the latter occurs only when the endothelial cells penetrate the brain parenchyma and are enveloped by the astrocyte foot processes.

Substances freely permeable to the blood-brain barrier

Of the factors that influence the ability of a solute to cross the blood-brain barrier, the lipid solubility is probably the most important. Lipid-soluble compounds dissolve into the lipid membrane of the endothelial cells enter the brain by simple diffusion. It must be emphasized that diffusion does not require energy and that the limiting factor to diffusion is the difference in concentration of either side of the brain endothelial cells, as molecules will move from an area of higher to one of lower concentration. Most neuroactive compounds, such as ethanol, diazepam and nicotine, are highly lipid soluble and enter the brain so rapidly that the only limiting factor to their uptake is the cerebral blood flow. In fact, the passage of any substance from blood to brain will depend not only upon the permeability of that particular substance, but also upon the surface area of the capillaries available for the exchange with the brain. Therefore the PS product defines the rate at which a product enters the brain.

For highly permeable substances which are almost completely extracted from the blood in a single passage through the brain, the factor that mostly controls the diffusion into the brain is how much, and how fast, that substance can be delivered to the brain itself. Therefore, in this paradigm the limiting factor is, infact, the cerebral blood flow. This is the conceptual basis for the use of highly permeable substances such as xenon or iodoantipyrine for measurement of cerebral blood flow. It should not be forgotten, however, that although highly lipid soluble, a substance can be bound to plasma proteins and therefore prevented from crossing the blood-brain barrier. In this situation the permeability, predicted on the basis of the lipid solubility, will not correspond to the actual penetration on the brain. Anticonvulsants such as phenobarbital and phenytoin belong to this group of compounds. Brain function is dependent upon a constant supply of oxygen which is highly permeable and rapidly diffuse into the brain. Similarly, the free, rapid diffusion of carbon dioxide compared to the more controlled entry of hydrogen ions means that the pH of the brain interstitial fluid is dependent on the concentration of carbon dioxide.

Water exchanges extremely rapidly across the blood-brain barrier by diffusion. Whilst is the peripheral circulation, the hydorstatic pressure drivers water out of the capillary bed, in the cerebral circulation, due to the tightness of the capillary endothelium, osmotic forces assume much more relevance than the hydrostatic ones. The use of mannitol in clinical practice is based on this principle. Mannitol crosses the blood-brain barrier extremely slowly, and thus by increasing the osmolarity of the plasma on the luminal side of the barrier, water will more unidirectionally from brain to blood. It is obvious that for this mechanism to be operational, an intact blood-brain barrier is required since an alteration of the permeability of the endothelial cells will lead to an increased passage of mannitol from blood to brain, resulting in an increased brain osmolarity and a consequent abolition of the antioedema effect.

Regulated permeability of non-diffusible molecules

Many substances which are essential to the functioning and the structural integrity of the brain are polar compounds and therefore non-lipid soluble. Specialized carrier-mediated transport on the luminal side of the endothelial cells ensures that these substances are infact transported in the brain. The carrier-mediated transport for glucose is the best studied. It does not require energy and therefore does not work against concentration gradient and is stereo-specific, since D-glucose but not L-glucose can be transported. The Glut- 1 transporter has a differential expression with expression of the transporter since in chronic hypoglycaemia it is upregulated, while on the contrary, in models of non-obese diabetes, the transporter is downregulated. Therefore, the level of Glut-1 activity in endothelial cells forming the barrier is modulated by blood glucose levels and by the metabolic activity of the area served by the capillaries ensuring that the metabolic requirements of individuals brain areas are realized.

A smiliar arrangements allows large neural L-amino acids such as tryptophan and phenylalanine to enter into the brain quickly. These are essential for the synthesis of neuro-transmitters within the central nervous system and cannot be synthesized directly within brain tissue. In contrast, the entry of small neutral amino acids, such as glycine and ?- aminobutyric acid, is restricted. These compounds are synthesized by neurones for use as neurotransmitters, and hence the entry of circulating glycine and GABA is prevented. Similarly, the acidic amino acids, such as glutamate and asparate, which have an essential role in neurotransmission and are implicated in excitotoxic damage, cross the blood-brain barrier extremely slowly, their levels within the brain being maintained by metabolic processes intrinsic to the brain. The basic amino acids lysine and arginine are transported across the barrier by a carrier similar to that utilized by the large neutral amino acids. There is therefore a free passage for amino acids required as precursors and metabolic intermediates, whilst those amino acids with direct neuromodulatory action are restricted by the absence of a specific carrier.

Metabolic barrier

The first suggestion that the brain endothelial cells do not play a merely passive role in limiting the passage of substances, but might have an active metabolic function in regulation the brain environment, derives from an anatomical observation. Oldendoff et al. demonstrated that the cerebral endothelial cells have five times more mitochondria than capillaries elsewhere in the body suggesting a high metabolic rate. Active, energy-requiring transport at the blood-brain barrier contributes to the regulation of the ionic milieu of the brain. The particular position of the Na+/K+ - adenosine triphosphatase on the abluminal side of the brain endothelial cells confers to the endothelium almost an epithelial - like property. By exchanging sodium and potassium, the capillary endothelial cells can secrete fluid into the brain and contribute to the composition of the intersitial fluid. This system permits the maintenance of a potassium level within the extracellular space of the parenchyma lower than that of the plasma even during periods when plasma potassium levels are changing. As changes in extracellular potassium will directly affect neuronal excitability, it is essential that this is highly regulated.

The presence within the brain endothelial cells of specific enzymatic systems confers the property which has been defined as the enzymatic blood-brain barrier. This is particularly relevant for the detoxification of potentially damaging compounds or for the metabolic activation of prodrugs. The most important area in which the enzymatic properties of the barrier play a role is in the treatment of Parkinson's disease using 3,4-dilhydroxy-L-phenylalanine. Dopamine is not transported across the blood-brain barrier, but L-dopa, the precursor to dopamine, is rapidly taken up via the neutral amino acid transporter and converted to dopamine by dopa-decarboxylase within the endothelial cells. The formation of dopamine in the periphery is minimized by simultaneous administration of the dopa-decarboxylase inhibitor carbidopa thus reducing the requirement for high systemic doses of L-dopa. This does not pass through the barrier, but inhibits dopa-decarboxylase in other tissues. The endothelial cells contains a number of enzymes utilized by the liver to metabolizes drugs and toxins and thus protect the brain form the harmful effects of such compound.

Furthermore, it is conceivable that the endothelial cells would play a role in the metabolism of neurotransmitters derived from neuronal activity. The endothelial cells contain monoamine oxidase, cholinesterase and GABA transaminase, all of which are responsible for the breakdown of neurotransmitters, and thus the enzymatic capability of the endothelial cells is utilized by the brain itself to regulate neuronal activity. Last, but not least, the endothelial cells can produce substances which may influence the activity and properties of neighboring cells.

Because of their size and polarity, proteins are not expected to cross the blood-brain barrier and, infact, most proteins are excluded from the brain. The formation of vasogenic oedema following the breakdown of the blood-brain barrier is produced by the increased osmotic pressure produced by unregulated protein accumulation in the brain parenchyma. However, specific transporters have been described for substances such as insulin, transferrin and interleukin - 1. When these substances bind to the transporter, the luminal membrane of the endothelial cells will invaginate allowing a vesicular transport to take place.

Regulation of the blood-brain barrier

A wide variety of receptors are expressed on brain microvessels suggesting that the barrier, transport and enzymatic function can be regulated by endogenous mechanisms. Probably the best example of this regulation is the demonstration that blood-brain barrier permeability to water can be altered by the noradrenervic innervation from the locus ceruleus. In fact, ablation of the locus cerules abolishes the Na+/K+ - ATPase activity in brain capillaries and thus active control of electrolyte balance may be lost. Other examples of regulation include the ability of atrial natriuretic peptide to increase the blood-brain barrier permeability to water. The tightness of the junction can also be influenced by hormones and neurotransmitters. Dexamethasone, progesterone and corticotrophin may all be responsible for increasing the tightness of the junction in normal animals. Conversely, activation of the pathways which result in an increase in calcium concentration within the endothelial cells could be the final common pathway for an increase in permeability deriving from action of 5-hydroxytryptamine or activation of the B2-bradykinin receptors. Obviously, the destruction of the endothelial cell membrane by any pathological process would lead to the alteration of the passive permeability of the barrier; the mechanism advocated for the alteration of the blood-barrier produced by the action of free radicals, such as that occurring after cerebral ischaemia.

In recent years, a growing importance in the regulation of the activity of the barrier has also been attributed to message coming directly from the brain and acting at the brain vascular interface. Among the possible substances which act as a messenger, cytokines have acquired a major role. For example, both tumour necrosis factor alpha and interleukin - 1ß are rapidly produced in response to injury. Both cytokines are able to induce the expression of intercellular adhesion molecule -1, a protein on the cell surface which mediates leucocyte adhesion to the endothelial cells. Following binding to ICAM-1, leucocytes will pass through the endothelial cell layer into the brain and instigate an inflammatory reaction. Parallel induction of nitric oxide production and free radical formation will enhance the inflammatory reaction whilst producing an alteration in the barrier properties of the endothelial cells. The complex interaction between cytokines and free radicles on blood-brain barrier permeability and function is an area of growing interest to the understanding of the pathophysiology of a number of conditions and may represent a potential site for therapeutic intervention.

Summary

The blood-brain barrier is formed by a complex interaction between the endothelial cells of most cerebral capillaries and specialized astrocytes within the brain parenchyma. The role of the barrier is not solely to isolate the brain from the circulating blood, but to provide a selective crossing point for those substances required in order to maintain normal brain function whilst excluding substances prejudicial to neuronal survival. Additionally, the barrier actively contributes to the maintenance of a constant extracellular environment within the brain even in the presence of fluctuating peripheral ionic concentrations. The properties of the barrier may be exploited for therapeutic benefit, but they present a significant difficulty in the design of centrally neuroactive drugs. An alteration of the barrier therefore results in serious consequences for the continued functioning of the central nervous system.

BACK

Brain Oedema

Brain oedema is one of the most frequent diagnoses based on computed tomography (CT) or magnetic resonance imaging (MRI) in neurosurgical patients, in particular in intensive care patients. In clinical practice this diagnosis is rarely questioned. Various existing treatment modalities directed against brain oedema formation such as application of steroids, mannitol or controlled hyperventilation are mostly symptomatic. However, their mode of action, treatment dosage and their therapeutic effects are controversial.

One reason for this is the early and uncritical use of the term 'brain oedema', leading to an overestimation of this potentially life-threatening complication. Some years ago, clinical deterioration of neurosurgical patients following head injury, tumour removal or aneurysm clipping was explained by the presence of brain oedema or brain swelling, since it was often difficult to diagnose brain oedema clearly by CT or MRI. The therapeutic consequence was mostly sterotype: steroids were administered, a high-dose osmotherapy or uncontrolled hyperventilation was performed. However, it has to be kept in mind that neurological deterioration can also be related to other pathological processes, for example, extension of an infarct, enlargement of a haematoma, cerebral vasospasm and so on. The diagnosis of brain oedema is based on complex pathophysiological changes in blood-brain barrier function, regulation of brain cell volume and release of toxic mediator substances.

The aim of this chapter is to contribute to a critical use of the term 'brain oedema', and to understand pathophysiological mechanisms of brain oedema formation, possibly leading to a more targeted, while still symptomatic, therapy of brain oedema.

There are several detailed reviews and monographs recommended for the study of cerebral oedema, brain cell volume regulation and the blood-brain barrier.

According to Pappius (1974) brain oedema is defined as an increase in brain water content leading to an increase in tissue volume. This simple definition clearly distinguishes brain oedema from hydrocephalus and cerebral atrophy as well as 'brain swelling' due to an increase in cerebral blood volume, a state which is rarely proven but often incriminated if intracranial pressure rises or is increased.

Both brain oedema and cerebral swelling by vascular engorgement concern the Monro-Kellie doctrine: the intracranial contents consists of four distinct compartments, the intracellular space, interstitial fluid, cerebrospinal fluid (CSF) and blood. Expansion of any of these compartments either leads to a compensatory decrease of another or produces an increase in intracranial pressure. Thus, brain oedema and brain swelling are closely connected to an increase in intracranial pressure, its physiology and pathophysiology.

Blood-brain barrier

The blood-brain barrier is an essential factor for the regulation of normal brain volume. The state of the blood-brain barrier is an important feature in distinguishing different types of brain oedema. Thus, the term blood-brain barrier shall briefly be explained anatomically and functionally.

The existence of a barrier between blood and brain was first demonstrated by Paul Ehrlich in the late nineteenth century. Ehrlich injected trypan blue, a vital dye, intravascularly and noted that the brain was not stained, in contrast to all other organs of the body. Thereafter, Goldmann injected trypan blue into the CSF and observed that only the brain was stained and no other organ. These simple experiments demonstrated the presence of a barrier between brain and blood and that there is no barrier between brain and CSF.

Anatomy

For many years the anatomical substrate for the blood-brain barrier was unclear. Although Spatz and Krogh had proposed that cerebral vasculature was responsible for the formation of this barrier, it had also been proposed that the close attachment of astrocytic foot processes around cerebral capillaries, or the basement membrane surrounding these capillaries, might form the anatomical basis for the blood-brain barrier. In 1967 Reese and Karnovsky, using electron microscopy, studied the distribution of horseradish peroxidase given intravenously. They proved that the vascular endothelial cells are tightly connected to each other by so-called 'tight junctions' forming the anatomical substrate of the blood-brain barrier. Moreover, there are no transendothelial channels and there is no pinocytotic activity in cerebral endothelial cells.

Brain endothelial cells are metabolically highly active. They are five times richer in mitochondria than other endothelial cells and contain various enzymes known to metabolizes neurotransmitters and their precursors. Thus, the endothelial cells might also be actively involved in removing substances from the interstitial space. These endothelial cells are equipped with various receptors for hormones and neurotransmitters suggesting that endothelial function is regulated in many ways.

There are only a few small areas within the brain void of any blood-brain barrier, as for example the pituitary gland and circumventicular organs. These areas are important for the regulation of systemic hormone systems. In these areas endothelial cells are fenestrated and tight junction are incomplete.

Permeability

The blood-brain barrier limits the passage or entrance of numerous substances into the brain, but vital metabolic substrates can pass the barrier easily and metabolic wastes can also be removed.

The amount of a substance which can enter the brain depends on its permeability, capillary surface area, the concentration of the substance and the time the substance is present in plasma.

Cerebral blood flow (CBF) is the determining variable of uptake for highly permeable substances. In contrast, uptake of low permeability substances is mainly diffusion limited.

Permeability is highly dependent on lipid solubility. Lipid-soluble compounds can enter the brain by simple diffusion, whereas polar compounds cannot. They require carrier systems, such as facilitated diffusion, cotransporters or active pumps. Proteins are large, polar molecules which do not move easily across the blood-brain barrier. Oxygen and carbon dioxide rapidly diffuse across the blood -brain barrier due to their lipid solubility. The glucose transport system is one of the most important transport systems of the blood-brain barrier, since the almost exclusive source of energy for the brain is glucose. There are other specific carrier systems for amino acids and nucleic acid precursors, referred to as facilitated diffusion transport system.

Free water quickly crosses the barrier by simple diffusion and there is constant exchange of water between blood and brain . About 50% of total brain water can be exchanged within seconds by bidirectional diffusion. On the other hand, the blood-barrier impedes the accumulation of water in the brain due to a high reflection coefficient for electrolytes.

Unlike other organs, flux of water within the brain is only limited by osmotic pressure and not by hydrostatic pressures, due to the tightness of the blood-brain barrier. Nevertheless, the osmotic driving force is rather powerful. Since a difference of 1 mmol is equivalent to 20 mmHg of hydrostatic pressure, a difference of only 20 mmol even surpasses the capillary hydrostatic pressure of about 30 mmHg. As a consequence, brain water content is primarily influenced by osmolar forces, providing the blood-brain barrier is intact. If, however, barrier function is disturbed, hydrostatic pressure gradients prevail.

Brain Oedema

At the turn of the century, Reichardt (1905) introduced the term 'brain oedema' in a first attempt to differentiate between oedema and brain swelling. It was not until the 1950s and 1960s that brain oedema was classified into two major types by the neuropathologist Igor Klatzo (1967). His fundamental experimental work enabled the classification of brain oedema on pathophysiological grounds. The state of the blood-brain barrier is one keystone of this differentiation. If the blood-brain barrier breaks, a so-called 'vasogenic' brain oedema develops. Cellular injury without blood-brain barrier disturbance is called 'cytotoxic'. Fishman (1975) added to these prototypes a third entity called 'interstitial' or 'hydrocephalic' oedema. A fourth type of oedema is due to osmotic imbalances where there is neither injury to endothelium nor to astrocytes and neurones.

Vasogenic Oedema

To induce and imitate the vasogenic type of oedema, Klatzo (1967) used a 'cold injury'. In this experimental model the cortex of the animal is frozen, a cortical necrosis develops surrounded by a border in which the endothelial lining is damaged and the blood-barrier breaks down. Thus, it can also be called 'open barrier oedema'. It is characterized by a protein-rich exudate derived from plasma due to increased permeability of vessels to albumin and other plasma proteins. In this context, it is interesting to note that mere opening of the blood-brain barrier does not induce the vasogenic type of oedema, for example during hypertonic crises there is distension of endothelial cells and blood-brain barrier opening but no, or only minimal, development of brain oedema in its strict sense. The same holds true for osmotic opening of the barrier in which the endothelial cell lining is shrunk by infusion of hypertonic solutions. Other factors must be associated with barrier opening leading to oedema. It has been hypothesized that mediator substances which are released or generated within the necrotic focus or its border rim may be responsible. For many years, the cold injury model has been used as an experimental model for traumatic contusion. Traumatic brain oedema is a complex phenomenon that is not well modelled by the cold injury, as discussed below. It has always been criticized that this type of lesion and the developing oedema cannot be transferred easily to the clinical situation following trauma, ischaemia, infection or metastases. This model, however, has provided insights into the development and resolution of clinical forms of oedema associated with blood-brain barrier damage, as seen in metastases and abscesses. Thus, it has been shown that the entry port for vasogenic oedema is within and around the focus, and the oedema front propagates through the white matter to the ventricles, probably because of less packing of cells within the white as compared to the gey matter. Excision of the necrotic focus, the cold lesioned area itself, resulted in cessation of oedema formation. Even if the excised tissue was reinserted, there was no oedema is vessels in an around the focus, and it could well be that there are mediators, either from the necrotic focus or the blood enhancing blood-brain barrier damage the oedema process.

Cellular/cytotoxic oedema

Klatzo defined the cytotoxic prototype of oedema as cellular swelling associated with a decreasing extracellular space while the endothelial cell lining, that is, the blood-brain barrier, is still intact. Swelling of cellular elements is a typical characteristic. Cytotoxic oedema is characterized by an increase in cellular volume, not simply due to a toxic state, and therefore is preferably called cellular oedema. Contrary to vasogenic oedema, the blood-brain barrier remains intact in this type of oedema.

In principle, there are three mechanisms responsible for neuronal and glial cell swelling.
1. Increased sodium permeability of the cell membrane with increase Na+ influx into the cell.
2. Dysfunction or failure of the sodium - potassium - adenosine triphophatase pump.
3. Membrane energy depletion followed by a failure of the active ion pumps.

These mechanisms, single or in combination, cause cellular swelling. A simplistic model is to describe cell swelling in the brain as 'pump leak equilibrium' distance. Under normal conditions, the influx of osmotically active solutes is equilibrated by their active elimination. This avoids an accumulation of solutes within the cell; immediate cell swelling would follow due to high membrane permeability to water molecules. The increased membrane permeability of Na+ ions in cellular oedema causes a shift of the 'pump leak equilibrium'. The Na+ influx along the electrochemical concentration gradient cannot be compensated by the active Na+ pump. Further pathological conditions, for example ischaemia with energy depletion, lead to a complete pump failure.

There are several causes for the increased Na+ permeability in cellular oedema. The excitotoxic amino acid glutamate plays an important role. Increased concentration of glutamate in the extracellular space, for example in ischaemia or head injury, interact with glutamate receptor of the cell membrane and stimulate the opening of Na+ channels. The efforts of the cell to maintain or re-establish the extracellularly low glutamate concentrations, finally leads to energy depletion followed by failure of cell volume control. The presence of high concentrations of glutamate and K+ ions, for example after terminal depolarization of brain cells in ischaemia, stimulates clearance mechanisms of glial cells. Astrocytes have high-affinity transporters for glutamate to guarantee fast clearance from the extracellular space. The active transport of one glutamate molecule is coupled to a simultaneous influx of two or three Na+ ions. This process is regarded as a compensatory function of glial cells to maintain and restore extracellular homeostasis may lead to cell swelling by itself.

It is not only glutamate but also increased concentrations of lactic acid that induces cell swelling. This process is a similar compensatory mechanism to normalize the intracellular pH. The elimination of H+ in exchange for Na+ leads to a net accumulation of Na+, thus causing cell swelling.

After failure of the active Na+/K+ - ATPase pump the accumulated Na+ ions cannot be eliminated but further increase the osmotic concentration.

Cellular oedema, present in cerebral ischaemia, hypoxia, after trauma or due to toxins, is an important manifestation of secondary brain damage. A mere shift of the interstitial fluid into cells of the brain parenchyma does not change brain volume, but further cell swelling increases brain volume and therefore intracranial pressure. Consequently, at least a minimal CBF is necessary to promote the process of oedema formation.

Hydrocephalic (interstitial oedema)

This term was introduced by Fishman. It occurs if CSF outflow is obstructed while there is still production of CSF leading to an increase in ventricular pressure and causing subsequent ventricular dilation. The increase in hydrostatic pressure in the ventricles drives CSF into the periventricular tissue and the increased hydrostatic pressure stops drainage of interstitial fluid. Thus, the periventricular areas released with CSF. CT scanning and MRI illustrate 'periventricular lucencies' or increased signal intensities for water.

Osmotic brain oedema

Since water diffuses easily across the blood-brain barrier, osmotic gradients between blood and tissue play an important role. It has been stated above that brain water content is primarily controlled by osmolar forces, providing the barrier is intact. Thus, it is easy to understand that all states with low plasma osmolarity can lead to water accumulation within the brain and osmotic brain oeema. Hyperosmolar states, for example, are seen if there is inappropriate secretion of antidiuretic hormone or a disequilibrium following dialysis. Osmotic brain oedema is neither vasogenic nor cytotoxic. There is no blood-brain barrier opening nor a cytotoxic mechanism. Of decisive importance for the development of osmotic brain oedema is the speed at which osmotic imbalances develop, since concentration or dilution of osmolarity within the brain can be compensated if osmolar gradients develop slowly. The hyperosomolar syndrome is of particular clinical importance, for example, following dehydration or systemic disturbances of the water and electrolyte system. Under these conditions, increases in osmolarity of cerebral tissue are associated with symptoms such as drowsiness. If such hyperosmolar states are corrected too rapidly by infusion of isotonic or even hypotonic solutions, the brain swells due to osmotic brain oedema. This can even cause a critical increase in intracranial pressure, herniation and finally brain death. Thus, it is recommended that plasma osmolarity is reduced slowly.

Following cerebral ischaemia with reperfusion there is also an osmotic component of brain oedema. With reperfusion the hypertonic tissue is perfused with isotonic blood causing an additional diffusion of water into the hypoxic/ischaemic tissue.

Ischaemic brain oedema

The classic experimental model for the study of ischaemic brain odema is stroke induced by clipping of the middle cerebral artery. With sensitive volume gauges allowing monitoring of the displacement of the cortex, incipient swelling can be detected after 1 - 2 min following vascular occlusion.

Ischaemic brain oedema is composed of various oedema prototypes, developing from cellular, vasogenic and osmotic components. Important factors determining the formation of ischaemic cerebral oedema are the duration and depth of the ischaemia, as well as the quality and quantity of reprefusion.

As long as flow remains above a critical threshold, water and electrolyte content of brain tissue remain normal even if electrophysiological disturbances are already present. Below this threshold, the water and electrolyte homeostasis is disturbances are already present. Below this threshold, the water and electrolyte homeostasis is disturbed. The threshold for oedema development is species dependent and oedema develops when blood flow decreases below a threshold of about 10ml/100g/min. At these low flow values, ion exchange pumps break down. Ischaemic oedema is characterized by the initial uptake of water and electrolytes from blood and CSF, followed by a breakdown of the blood-brain barrier, making it permeable to serum proteins in cases of irreversible tissue damage. Oedema, in consequence, is initially cytotoxic and later vasogenic.

The interval when the vasogenic type supervenes is not clearly defined. During the initial 3 - 6h of ischaemia the blood-brain barrier remain impermeable to the passage of conventional barrier tracers such as Evans blue.

If ischaemia is absolute and blood flow zero, for example during a cardiac arrest, cellular oedema, mostly derived from the extracellular space, develops after a few minutes. The endothelial cells swell by taking up sodium and water from the interstitial fluid due to increased membrane permeability. Up to this stage the process is potentially reversible, but with stepwise sequence of organelle failure the threshold for survival is reached.

There is no net increase in brain water as long as CBF is not re-established to serve as the water source. During ischaemia there is an increase in tissue osmolality. Following reperfusion there is a rapid increase in extracellular fluid and cerebral water content associated with a corresponding rise in intracranial pressure.

Severe ischaemia followed by reperfusion is characterized by a biphasic development of oedema. During the first hours after ischaemia the cellular component dominates and the blood-brain barrier remains intact, demonstrated by the absence of Evans blue leakage, a marker for the vasogenic component. The second phase starts several hours after ischaemia, when the blood-brain barrier becomes permeable to serum proteins and blood serum extravasates into the brain leading to oedema formation. This vasogenic phase lasts much longer and is associated with clinical deterioration.

In contrast to complete ischaemia there is no biphasic odema formation during continuous incomplete ischaemia. Instead, oedema develop e s progressively over a 24 - 48h period. During the first 4-12h of incomplete ischaemia, the blood-brain barrier remains intact but subsequently breaks down.

Development of ischaemic oedema, as described above, requires a certain amount of persisting blood flow, which functions as a reservoir of fluid and electrolytes taken up by the brain. During complete ischaemia, as induced by cardiac arrest, this reservoir is restricted to the CSF compartment, and ischaemic brain swelling, in consequence, is minor. However, as soon as recirculation is restored after a period of ischaemia the brain swells abruptly, and remains swollen. This postischaemic brain oedema is proposed as one of the factors responsible for postischaemic recirculation disturbances and hence for the irreversibility of ischaemic brain damage.

Traumatic brain oedema

For many years it has been known that intracranial hypertension is a frequent complication of traumatic brain injury. In more than 50% of patients dying from brain trauma there is a refractory increase in intracranial pressure. However, to what extent brain oedema contributes to this process has been debated. In the pre-CT era brain oedema was often held responsible for fatalities. This opinion has since been modified. The widely accepted assumption was that there is development of vasogenic brain oedema following traumatic brain injury, with a maximum on day 2 - 3 post-trauma. Since steroids were used very effectively to reduce peritumoral brain oedema which is of clear vasogenic origin, steroids were also tested to treat severe head injury in general, not simply pericontusional oedema. While there were initially two positive studies with steroids following severe head injury, three subsequent controlled trials were unable to show a beneficial effect on mortality or clinical outcome. Thus, steroids have been abandoned and are now not part of the standard protocols to treat severe head injury. Nevertheless, there is very recent evidence that steroids given in very high doses might ameliorate the clinical course of patients with cerebral contusions and perifocal oedema.

The assumption that vasogenic brain oedema is a major factor following trauma was one reason to use Klatzo's cold injury to model focal contusion experimentally. Unfortunately, some of the results obtained using this model were too easily transferred to the clinical situations. More recently, the contribution of vaosgenic oedema to severe head injury has again been intensely debated.

In a critical review, Miller and Corales questioned the significance of post-traumatic oedema. Based on experimental data obtained with the fluid percussion model of traumatic brain injury and CT observations made clinically in the early post-traumatic period, they concluded that the very early post-traumatic intracranial hypertension observed experimentally is related to a state of hypertension observed experimentally is related to a state of hyperaemic CBF. This corresponds to an increase in Hounsfield units in CT scanning and 'diffuse brain swelling' in patients. A few hours after fluid percussion injury there was no increase in cerebral water content. The authors then asked: brain oedema as a result of head injury - fact or fallacy? Meanwhile, numerous experimental investigations have clarified that there is brain oedema in various models of head injury including fluid percussion brain injury and that it takes time to develop brain oedema.

The question of diffuse brain swelling due to vascular engorgement, especially in the very early post-traumatic period, is still not solved. Very recent studies of Marmarous et al, using MRI techniques for non-invasive tissue water and cerebral blood volume measurements indicated that brain water was increased while blood volume decreased. These studies provide compelling evidence that the major contributor to early brain swelling is rain oedema and not blood volume. These studies, however, did not differentiate between intra- or extracellular water accumulation.

Moreover, positron emission tomography studies could not demonstrate increased intracranial blood volume. In addition, it is now widely accepted that the very early hyperaemic response to brain injury is followed by hypoperfusion on the first post-traumatic day.

There is no doubt that there is perifocal or pericontusional brain oedema following trauma. For a long time this pericontusional oedema has been regarded as vasogenic. There is, however, only limited blood-brain barrier damage and uptake of contrast media, both in CT and MRI patients. Thus, it has been speculated that pericontusional oedema is also cellular/cytotoxic. Indeed, recent experimental studies using the weight drop technique or the controlled cortical impact injury indicate that there is both opening of the blood-brain barrier and vasogenic as well as cellular/cytotoxic oedema. Following controlled cortical impact injury the extent of cellular/cytotoxic oedema is even more pronounced than the vasogenic component

Measurement of brain oedema

To judge and interpret clinical and experimental studies of brain oedema it is important to know the principles of various methods of measuring brain oedema and their potential errors. Tissue water content can be measured directly in experimental animals or small surgical tissue samples. Other techniques can be calibrated against actual tissue water determinations. Thus, they can also be used to give quantitative data in contrast to indirectly measuring techniques like blood-brain barrier permeability studies, electrical impedance and CT.

Wet weight/dry weight

This is the 'gold standard' technique to determine tissue water content. It is highly reliable and other techniques should be compared to this. A sample of brain tissue is taken, weighed immediately after removal and then desiccated in an oven for at least 24h. After equilibration to room air the sample is reweighed. The loss of weight indicates the total tissue water. This a very simple, although time-consuming, investigation that requires a reasonable amount of tissue. Since this technique measures total water content of a specimen it cannot distinguish between intra-or extracellular accumulation of water or oedema.

Specific Gravity

Nelson et al. described the technique of specific densitometry. Two liquids of different densities are put on top of each other in a cylinder and a gradient is established by slowly mixing these liquids. Thereby a linear density gradient is obtained which is calibrated by standards of known specific gravity. Thereafter, small specimens of tissue are dropped into the density column and will settle until they reach their isogravimetric point. Thus, many investigators simply give the specific gravity value as an index of tissue water content which can be calculated if the amount of solid is known.

The technique is erroneous if samples of different protein contents are compared. The technique is especially suitable and valid for measuring cellular/cytotoxic oedema.

Magnetic resonance imaging

MRI is the most useful clinical tool to image the brain . MRI uses the ability of protons to be magnetized. Since the majority of hydrogen atoms are within water molecules, MRI predominantly images water in various tissues. Although many attempts have been made to differentiate between extra- and intracellular water using MRI, this is still not feasible on a routine basis.

In normal brain with an intact blood-brain barrier, the capillaries are impermeable to intravascularly injected contrast agents. Dural vessels and structures in which capillaries are fenestrated allow diffusion of contrast material into the extracellular space. Also, tumor tissue lacking an intact blood-brain barrier enhances due to accumulation of the paramagnetic contrast medium in the interstitial space. Tumour capillaries in gliomas may have a near-normal structure with an intact blood-brain barrier, so that such tumours will not enhance with contrast. In more malignant gliomas, however, formation of capillaries is stimulated whose endothelia are fenestrated and therefore, have no blood-brain barrier- these tumours do enhance. Metastatic lesions equipped with non- CNS capillaries that are similar to their tissue of origin virtually always enhance. Extra-axial tumours arise from tissue whose capillaries lack tight junctions and, consequently, these tumours enhance. Usually, there is no correlation between MRI or CT enhancement and angiographic findings of hypervascularity, although it has been suggested that the vascular pooling in angioplastic neoplasm may represent 20 - 30% of the noted enhancement. It is important that a lack of enhancement does not necessarily signify lack of tumour and one cannot use enhancement to separate tumour from oedema in infiltrative gliomas or anaplastic astrocytomas. On MRI, oedema is seen as high intensity abnormality on T2-weighted images. This 'oedema' pattern is a reflection of a combination of tumour and oedema on histology.

Computed tomography

The understanding of the spread and resolution of brain oedema under clinical conditions is closely connected to CT. CT scanning is capable of visualizing even small changes in absorption of X-rays on a regional basis due to brain oedema. Comparative analyses between CT density and changes of brain water content demonstrated that it is possible to detect brain oedema if cerebral water content is elevated by 1.3%. This is equivalent to a decrease of 1 Hounsfield unit in CT density. Investigations of tissue water content and CT density in patients with severe head injury have confirmed these close correlations. Nevertheless, there are also obvious problems in differentiating between oedema and low-grade gliomas, as there are also attenuations in Hounsfield units. Moreover, it is not possible to distinguish between cellular/cytotoxic oedema and brain infarction.

An increase in permeability is visualized by contrast enhancement. CT scanning was shown to be useful in the investigation of the resolution of vasogenic oedema around metastases and abscesses by studying the rate of penetration of intravenous injected contrast media into the tissue and their spread towards the ventricles.

Blood-brain barrier permeability

Visualization of tissue uptake of dyes bound to proteins, contrast media and radiolabelled proteins indicate blood-brain barrier damage and therefore, vasogenic oedema. The easiest technique to demonstrate vasogenic oedema is intravenous administration of Evans blue dye. It binds to albumin and enters the brain only in areas where the blood-brain barrier is defective. Thereafter, it spreads within the tissue due to bulk flow and diffusion. Uptake of Evans blue is therefore a technique that may provide both a qualitative as well as quantitative assessment of vasogenic oedema. Quantitative assessment of Evans blue uptake can be achieved by extraction and colorimetric assays.

Radiolabelled substances like I-labelled albumin or C aminoisobutyric acid have also been used to determine breakdown of the blood-brain barrier experimentally, Clinically mTc- or Rb-labelled markers are used in CT or PET scanning.

Electrical impedance

Measuring the electrical impedance of tissue can be used for studying the development of cellular/cytotoxic swelling. This technique uses a low-frequency alternating current which is transmitted extracellularly, thus estimating the size of the extracellular space. This technique was used experimentally to study the development of cellular brain oedema in ischaemia and cytotoxic oedema following administration of glutamate. It is not applicable for clinical use.

Pathophysiology-mediatory of oedema

There are two important differences between brain oedema and oedema in other tissues, one being the state of the blood-brain barrier as described above and the fact that oedema development in the brain is limited because the intracranial compartment is a closed container with only minimal facilities for expansion. Thus, oedema development will soon cause an increase in intracranial pressure which, in turn, causes an increase in intracranial pressure which, in turn, causes a decrease in cerebral perfusion pressure and a decrease in cerebral perfusion pressure and a decrease in cerebrovasclar resistance followed by an increase in cerebral blood volume resting in increasing intracranial pressure. This vicious circle illustrates the fact that brain oedema can ultimately lead to cerebral herniation and finally brain death.

The intracranial situation is particularly dangerous if there are changes in vascular permeability leading to vasogenic oedema and loss of cerebral autoregulation as seen in very severe cases of traumatic brain injury. Under these circumstanses as evaluation in systemic blood pressure causes an increase in both vasogenic oedema and vascular volume. Taken together, brain oedema development may often affect CBF either locally or globally and influence intracranial pressure. Conversely, there are other important causes of an increase in intracranial pressure such as hematoma or an increases in cerebral blood volume due to hypercapnia amongst others. Thus, a significant rise in intracranial pressure should always be further evaluated by CT scanning.

Acute cerebral insults which are associated with cellular/cytotoxic and vasogenic oedema formation are also often associated with a release of substances which might promote or enhance the oedema process. Such compounds are called 'mediators'. The understanding of their pathophysiology is indispensable for the development of specific and more effective methods of treatment. Mediator substances may be formed from inert precursors or released in effective concentrations to the extracellular space where these compounds are usually not present or found only in minimal concentrations. Mediator formation is likely to occur in damaged brain tissue, for example contusions or infarcted tissue. The substances may then spread into primarily undamaged perifocal tissue where they can induce secondary processes like opening of the blood-brain barrier, alterations of the cerebral microcirculation or cytotoxic cell swelling. Extravasation of plasma-borne factors as active mediators or precursors can also occur.

To identify such factors, a number of criteria should be met (a) these substances should be found to induce cellular swelling or blood-brain barrier damage if administered to the brain ; (b) they should be formed or released under pathological conditions; and (c) inhibition of the release, formation or function of these mediator compounds should prevent or reduce brain oedema.

A considerable number of substances have been discussed such as glutamate, lactate, H+, K+, arachidonic acid and metabolites, oxygen free radicals, histamine and kinins.

The possible involvement of glutamate, H+ ions, potassium ions and lactate for the development of cellular/cytotoxic oedema are depicted. Ultimately, there is also an uptake of calcium by the cells associated with cellular swelling and rupture of cell membranes. Thus, the intracellular release of calcium has been invoked as the final common pathway to cell death. This process is also accompanied by a release of AA, polyunsaturated free fatty acids and oxygen free radicals. Experimentally, it has been shown that AA induces both cellular swelling and also leads to a breakdown of the blood-brain barrier. Thus, AA is proposed as a mediator substance of vasogenic as well as cellular/cytotoxic oedema. AA induces as unspecific opening of the blood-barrier for small and large tracers, but only moderate vasomotor responses. Since effective AA concentrations have been detected following brain injury within the oedema fluid, AA appears to be an important mediator of oedema formation. However, it is difficult specifically to inhibit the release of AA which might be considered as a potential therapeutic avenue. Nevertheless, inhibition of phsopholipase A2 by steroids and other compounds might explain some beneficial effects of these substances.

Free radicals are also believed to be involved in the generations of cellular/cytotoxic oedema. Their noxious effect on blood-brain barrier function has been debated. While it has been shown that there is certainly release or generation of free radicals following acute cerebral insults associated with oedema formation, their inhibition by various substances such as aminosteroids, superoxide dismutase, melatonin and others, moderately reduced brain oedema formation in various experimental models, but did not significantly affect the clinical course following severe head injury. The Kallikrein - kinin system, and its active polypeptide bradykinin, could also play an important role as mediator of vasogenic oedema. Bradykinin is capable of inducing oedema if administered to the brain. There is formation of kinins following brain injury and it is possible to reduce oedema formation by specifically inhibiting kinin release, at least experimentally. Oedema formation is supposed to be caused by an increase in blood-brain barrier permeability to small solutes, supported by an increase in blood pressure in the microcirculation due to arterial dilatation and venous construction. Therapeutic studies with synthetic kinin antagonists in severely head injured patients have recently exerted moderate beneficial effects.

Biogenic amines, such as serotonine, histamine and polyamines, could also be involved in vasogenic oedema formation. A potential role for leukotrienes has been controversial. They are potent constrictors of cerebral vessels if administered from the extravascular side and in most studies there was neither alteration in blood-brain barrier function due to leukotrienes, nor formation of oedema.

Taken together, glutamate, lactate, H+, K+, AA and free radicals are the most pertinent mediators of cellular/cytotoxic oedema. Kinins, AA and histamine seem to be the most intriguing mediators of vasogenic brain oedema. There are, however, a number of additional potential mediator compounds that might be considered to be like cytokines, for example tumour necrosis factor or p olyamines, thrombin and others.

Effect of oedema on brain function

When well localized or mild in degree, brain oedema is associated with little or no clinical evidence of brain dysfunction; however, when it is severe it causes focal or generalized signs of brain dysfunction.

Even following infusion oedema of the brainstem, electrophysiological alterations are moderate at best. In fact, it is not likely that oedematous tissue becomes hypotoxic, unless the intercapillary distance increase. Microscopic studies suggest that vasogenic oedema could alter neuronal function due to widening of the extracellular space separating adjacent foot processes from their neurones and synapses, described in the while matter as lamellar blebs. With resolution of oedema these changes reverse.

Resolution of oedema

Development and resolution of oedema are dynamic processes. Oedema production is estimated to be 0.09 - 1.63 ml/h in metastases and 0.42 - 3.39 ml/h in gliomas, based on CT scan analyses in patients. The speed of edema propagation ranged in this population from 0.2 to 2.2mm/h

There are three main mechanisms by which the pathologically accumulated oedema fluid is removed from the tissue, depending on oedema location and type.

1. It is widely accepted that the predominant portion of vasogenic oedema fluid is reabsorbed into the cerebral ventricles. Marmarou et al. found that the clearance of radioiodinated cat serum albumin leaving the brain occurred at 87.14% into the CSF and only 19.96% into the blood. Thus, 'vascular clearance' is minimal, in the acute state at least, and oedema resolution occurs mainly via bulk flow into the CSF.
2. The second mechanism, 'vascular' oedema resolution, involves uptake of proteins and ions by capillary cells, reducing the osmotic gradient between tissue and cerebral vessels and leading to a diffusion of water molecules back across the blood-brain barrier. This implies in intact blood-brain barrier.
3. The third pathway of oedema resolution, in cellular oedema, needs a normalized cellular metabolism with sufficient energy sources to rebalance the ionic imbalances responsible for intracellular water accumulation.
   In vasogenic and hydrocephalic oedema, oedema clearance into the CSF is studied in detail. The oedema fluid, of a composition similar to plasma, passes through extra-cellular channels along a pressure gradient from the site of blood-brain barrier breakdown from grey to white matter. The driving force for the spread of oedema fluid is a pressure gradient. Initially, a relatively high pressure is required to open the channels and gain access to performed pathways of low resistance, for instance the perivascular space, but once open for spreading of oedema only little hydrostatic pressure is necessary.

In metastases with a small perifocal oedema the amount of oedema resolution within the tissue averages 0.0086 ml/h/cm3, probably representing the reabsorption of oedema fluid into capillaries within the oedematous tissue. In large tumours with pronounced perifocal oedema the main fraction of oedema fluid is drained into the ventricular, and to a less extent into the subarachnoid CSF.

Brain swelling describes a physiological condition of the brain accompanied by high intracranial pressure that is potentially life-threatening. Frequently, the terms 'brain swelling' and 'oedema' are used similarly, probably because 'oedema' arises from the Greek word 'oedema' which means swelling. At the turn of the century, Reichardt tried to differentiate brain swelling from oedema. If the surface of a brain cut was wet, 'oedema' was diagnosed, whereas a dry brain cut was called 'brain swelling' or 'swollen brain'.

Since the late 1960s, brain swelling has been attributed to an increased blood volume secondary to vasoparalysis. There is, however, little evidence for this assumption. At least in trauma, the 'diffuse brain swelling' seen immediately after trauma is probably not due to vascular engorgement and an increase in cerebral blood volume but rather due to brain oedema, be it cellular or vasogenic. Studies combining non-invasive tissue water measurement by MRI and cerebral blood volume techniques are focusing on this issue.

Brain swelling due to vascular engorgement and increased cerebral blood volume has also been blamed for the often observed intracranial pressure rise seen in children after severe head injury. CT scans in these children are characterized by increased CT densities and signs of diffuse brain swelling.

Clinical conditions associated with brain oedema

A variety of clinical conditions are associated with oedema formation, the presence of which may be a major causes of clinical deterioration. Most often, the oedema in a particular clinical condition consists of various oedema proto-types. Vasogenic oedema is predominantly observed around brain tumours, abscesses and haemorrhages, and is seen in the later stages of infarction. There are several causes of cellular oedema, for example ischaemia and hypoxia. Osmotic oedema is observed for example, in acute hypoosmolar states, osmotic disequilibrium syndromes occurring with haemodialysis and diabetic ketoacidosis. Interstitial oedema is a typical feature of hydrocephalus.

Tumours

Peritumoral oedema is a common occurrence in patients with malignant gliomas, meningiomas and metastatic tumours. The predominant oedema type associated with brain tumours is classified as vasogenic. Responsible for this is the formation of tumour capillaries deficient of a functioning blood-brain barrier and often with fenestration, rather than the active destruction of intact cerebral capillaries by tumour invasion. Consequently, plasma proteins and other macromolecules pass freely into the intersitial extracellular space. The precise mechanism of peritumoral oedema formation remains poorly understood, but it is presumed to be related to the production of a vascular permeability factor known to be associated with gliomas. The grade of peritumoral oedema is often closely related to the degree of malignancy of the brain tumour, location of the tumour and extent of venous involvement.

Brain oedema develops in approximately 50% of meningiomas. It is more common with large lesions but may be extensive with small ones. Studies have indicated its presence is significantly correlated with either the meningioma blood supply coming in some degree from cerebral pial arteries, or with its venous drainage connecting to cortical veins. While varying amounts of oedema may be present with any of the meningoma cell types, fibroplastic and transitional cell tumours have been reported to have only mild to moderate degrees of oedema. Severe oedema tends to be associated with meningiomas of the syncytial or angioblastic cell types and tend to be hyperintensive on T2-weighted images.

In glioblastomas an extensive mass effect mainly due to fairly extensive oedema, usually apparent in the adjacent white matter, is often seen even in relatively small tumour masses. As with all infiltrative gliomas, there is no clear miscroscopic margin showing where tumour cells stop and reactive gliosis, oedema or normal brain begins. Therefore, what is termed 'oedema' is more accurately described as 'tumour plus oedema'.

Metastases are notoriously surrounded by massive amounts of oedema, often extending far from the site of a relatively small metastatic focus. The extent of associated oedema has no direct relationship to the size of the metastasis. Usually, metastatic lesions are distinguishable from their associated oedema on both CT and MRI and intravenous contrast clearly shows the metastasis to separate from the surrounding oedema. On MRI, a metastasis is typically a focus of variable intensity surrounded by high-intensity oedema. The oedema accompanying metastases does not usually cross the corpus callosum, nor does it involve cortex, features which often help to distinguish this lesion from primary infiltrative brain malignancies. Regardless of the appearance of the enhancement, there is in general a greater degree of oedema associated with a metastatic focus if compared to the oedema associated with most of the benign entities, as well as compared to edema associated with primary gliomas.

Abscesses

Another frequent clinical condition associated with oedema is the brain abscess. Oedema surrounding an abscess may be greater in volume than the abscess itself, and may cause much of the associated mass effect. With MRI, due to the sensitivity of T2-weighted images to alterations in tissue water, earlier detection of cerebritis and brain abscess is possible, compared to CT. on T1-weighted images of a bacterial abscess, oedema is moderately hypointense surrounding a marked hypointense central focus with a hyperintense rim. On T2-weighted images the signal intensities are quite variable.

Trauma

Brain trauma is a complex of a variety of cerebral lesions, including contusion, haematoma, subarachnoid haemorrhage and diffuse axonal injury. Development of oedema depends on the kind of primary lesions and concomitant conditions, such as hypoxia and/or ischaemia. Probably both oedema types, the vasogenic and cytotoxic component, are present around haemorrhagic contusions, independent of the primary and, eventually, secondary injury mechanisms. Most frequently, pericontusional oedema is present, responsible for the often extensive mass effect with midline shift in such patients.

Intracerebral haematomas

A cerebral haematoma causes compression of surrounding tissue, reading perfusion and therefore oxygen delivery from oxygenated blood to these regions. Therefore, similar processes as described in traumatic brain injury occur. Oedema is frequently revealed both on CT and MRI and is more prominent around acute haematomas compared to subacute or chronic haematomas. As in trauma, both cellular and vasogenic oedema components are prominent. Recently, evidence is growing that thrombin is an important mediator of oedema surrounding intracerebral haematomas.

Radiation

Asymptomatic focal oedema is commonly seen on CT and MRI following focal or large-volume irradiation. When radiation necrosis is present, mass effect and oedema are common findings with clinical evidence of focal neurological abnormality and raised intracranial pressure. Microscopically, the lesion shows characteristic vascular changes and white matter pathology. The delayed form, generally seen after treatment of malignant gliomas, has its onset usually from 6 months to 2 years after treatment and may - in the case of a significant mass effect and oedema - demand surgical decompression. Radiation -induced peritumoral necrosis and vascular changes can occur sooner than the typical 1-year interval, for example acute radiation encephalitis can show a disrupted blood-brain barrier and enhance dramatically on MRI. Again, radiation - induced brain oedema is a composite mixture of cellular and vasogenic oedema.

Cerebral ischaemia

Various situations with vascular dysfunction may ultimately lead to oedema formation, for example circulatory arrest, cerebral vasospasm, emboli and venous sinus obstruction. An example of a thromboembolic occlusion of the right middle cerebral artery with concomitant ischaemic oedema is demonstrated. Hypoxia after cardiac arrest or asphyxia may result in cerebral energy depletion, and therefore, cellular swelling with increased intracellular osmoles which induce rapid entry of water into cells. In clinical practice, however, hypoxia-induced oedema is rare and described only after longer periods of insufficient reanimation.

Cerebral vasospasm associated with subarachnoid haemorrhage may reduce local CBF and cause incomplete ischaemia. Venous sinus obstruction may occur as a result of infection, from direct occlusion by trauma or surgery, invasion by tumor such as a meningioma or pathological thrombosis usually seen with hypercoagulable states. If large or abrupt, the stasis will diminish or obstruct flow through the arterial capillary bed and produce cerebral ischaemia and infarction involving the venous territories. The resulting increase in hydrostatic pressure will produce a mixture of vasogenic and cellular oedema.

Miscellaneous

Brain oedema has been reported in pseudotumour cerebri, a disease often affecting obese younger women with typical symptoms of elevated intracranial pressure, such as headache, nausea, vomiting, diplopia, ataxia or altered consciousness and always with optic disc swelling. There is still discussion over whether or not brain oedema is present.

Reye's syndrome, a neurological disorder of children, is characterized by fulminant hepatic failure, a rapid progressive encephalitis and severe intracranial hypertension, with brain oedema as a major and often fatal complication. The classic cytotoxic cerebral oedema is present. Electron microscopic features are cellular swelling of astrocytic foot processes and intralamellar myelin blebs.

In multiple sclerosis, gadolinium enhancement, especially in the acute phase, indicates blood-brain barrier opening due to inflammation in the white matter and spinal cord.

Hypertensive encephalopathy is a syndrome consisting of headache, seizures, visual changes and other neurological disturbances in patients with markedly elevated systemic blood pressure. Acute hypertensive encephalopathy is probably caused by failure of autoregulatory vasoconstriction with focal for general dilation of small arteries and arterioles. This is associated with an increased CBF, dysfunction of the blood-brain barrier and formation of vasogenic oedema that is thought to cause clinical symptoms. Oedema generally resolves after reduction of blood pressure.

Treatment of brain oedema

Treatment of brain oedema is largely sympotmatic. Most of the treatment modalities used are directed towards a decrease in intracranial pressure.

To date, only steroids are used as a 'specific' therapy of peritumoral oedema, and perhaps the oedema around abscesses. All other modalities are either used to dehydrate the brain or decrease cerebral blood volume. Therapeutic option to reduce brain oedema or intracranial pressure, their mechanisms, advantage and disadvantage

Steroids

Since the early 1960s corticosteroids have been used to treat brain oedema associated with brain tumours, metastases and abscesses. A common feature of these processes is a blood-brain barrier dysfunction with consecutive vasogenic brain oedema formation. It has thus been postulated that corticosteroids are 'sealing' the endothelial lining in and around tumours and metastases. In line with this assumption, it has been shown that extravasation of plasmatic markers into brain tissue decreases following steroid treatment. The exact mechanism of 'endothelial sealing', however, is unknown. Certainly, there are other well known mechanisms of corticosteroids which might contribute to the antioedematous effect, like 'stabilization' of lysosomal membranes, induction of various enzymes, inhibition of CSF secretion, inhibition of release of AA as well as lipid hydroperoxides. It could also be that corticosteroids are interfering with granulocytes and other inflammatory mechanisms operative in and around abscesses.

Whereas corticosteroids are highly effective in reducing peritumoral oedema, they are less effective in abscesses; their effect on the perifocal oedema around contusions is still under debate and there is certainly no effect on post-ischaemic oedema. The fact that the perifocal oedema around contusions is of cytotoxic origin to a great extent, might explain that the efficacy of steroids to treat perifocal pericontusional oedema is limited at best.

Osmodiuretics

Intravenous osmodiuretics, for example 20% mannitol, increase plasma osmolarity and dehydrate the brain due to the osmotic gradient. This effect is not confined to oedematous tissue, but effective all over the brain. A second mechanism of mannitol is associated with the ensuing haemodilution and initially increased total blood volume. Thereby, mannitol causes an increase in CBF which is followed by an autoregulaotry constriction of cerebral vessels and a consecutive decrease in cerebral blood volume. By this, intracranial pressure is effectively reduced.

It has been argued that mannitol infusion might cause a 'rebound phenomenon'. It is postulated that mannitol molecules may enter the tissue in cases of a defective blood-brain barrier. Once in the extracellular space mannitol might bind water molecules after dissipation of the osmotic gradient and thus lead to an increase of water content after mannitol has been cleared from the plasmatic compartment. This 'rebound phenomenon' has, however, been controversial. In clinical practice, mannitol is given abundantly and the rebound phenomenon is not observed. Also, repeated doses of mannitol are usually given to treat intracranial hypertension following severe head injury where the blood-brain barrier defect has probably been overestimated.

Mannitol is potentially nephrotoxic so it should not be given if the plasma osmolarity exceeds 320 mmol/l.

THAM

THAM is buffering substance which is used to treat acidosis and when given intravenously leads to a significant decrease in intracranial pressure. Apart from the buffering of intracellular acidosis, a diuretic effect of the substance has been postulated. THAM has been demonstrated to be beneficial in reducing brain oedema following cerebral ischaemia. It could well be that this substance is in fact interfering with the mechanisms of cellular/cytotoxic oedema following ischaemia. Consequently, THAM has been used to treat traumatic brain oedema and post-traumatic intracranial hypertension. Similarly to steroids, THAM could not be shown to be clinically useful regarding the neurological outcome of severely head injured patients in general. Thus, treatment with THAM has been abandoned. Moreover, it has recently been reported that THAM administration may cause a decreased cerebral oxygenation due to vasoconstriction. Nevertheless, it may remain a therapeutic option to decrease intracranial pressure under certain circumstances.

Therapies decreasing CSF, cerebral blood volume and intracranial pressure.

The simplest method to reduce intracranial pressure is to drain CSF. This is certainly a very effective method as long as there is CSF in the ventricles to drain.

Reduction of CSF production by furosemide or the carbonic anhydrase inhibitor acetazolamide also decreases intracranial pressure. Their effects is time-limited and it is not advisable to use them on a long - term basis.

The following therapeutic modalities have in common that they decrease intracerebral blood volume, thereby reducing intracranial pressure. Thus, these therapeutic options do not directly pressure. Thus, these therapeutic options do not directly interfere with brain oedema.

Hypercapnia causes cerebral vasodilation and an increase of cerebral blood volume with brain swelling which should be strictly avoided. Conversely, hyperventilation causing hypocapnia reduces cerebral blood volume by vasoconstriction. Hyperventilation is called 'moderate' if the arterial Pco2 is between 30 and 35 mmHg, and called 'forced' if the arterial Pco2 drops below 30 mmHg. Since hypocapnia causes vasoconstriction, there is always a danger of inducing cerebral ischaemia by this manoeuvre. This is in fact imminent if arterial Pco2 is below 30 mmHg. It is now possible to control cerebral oxygenation either by monitoring of oxygen saturation in the jugular bulb or by direct measurement of the partial pressure of tissue oxygen. It has been suggested therefore, that such techniques should be used to monitor forced hyperventilation in order to prevent cerebral maloxygenation.

Recently, hypothermia has attracted enormous interest. Decades ago it had already been used to decrease cerebral metabolism and thus local tissue demands for substrate supply. Thereby hypothermia decreases CBF and cerebral blood volume as well as intracranial pressure. Hypothermia has been shown to be very effective in various experimental situations of acute brain injury, ischaemic and traumatic. Clinical trials have yet to show whether this procedure is also efficacious in a clinical setting. It should be mentioned that hypothermia is a considerable technical effort with numerous side-effects, for example interference with coagulation.

Barbiturates also decreases cerebral metabolism, CBF and cerebral blood volume. It has also been postulated that barbiturates might interfere with mediator mechanisms of brain oedema, like AA release, free radical generation and so on. Barbiturates should generally not be given to severely head injured patients but only in cases where intracranial hypertension cannot be controlled by CSF drainage, mannitol and hyperventilation. Barbiturates bear the risk of arterial hypotension, thus potentially decreasing cerebral perfusion pressure.

Surgical decompression of the brain must be mentioned. Obviously, tumour excision removes the source of oedema and decompression brain tissue.

There has also recently been a revival of interest in removing haemorrhagic contusions in severely head injured patients with intracranial hypertension.

Finally, a massive bony decompression combined with an enlargement of the dura might be useful to cope with brain swelling due to ischaemic brain oedema following MCA occlusion or diffuse brain swelling in traumatized patients, especially in children and adolescents. Again, this is a symptomatic therapy aimed at preventing the secondary deleterious effects of brain oedema.

Interference with mediators of brain oedema.

To date, there are various compounds known to interfere with different mediator mechanisms of brain oedema which are efficiently reducing brain edema under experimental conditions. However, clinically, their effect is either moderate or non-existent. As mentioned above, steroids and barbiturates may interfere with lipid peroxidation, while aminosteroids and superoxide dismutase inhibit oxygen free radical generation. Clinical trials with the latter agents have not been encouraging and these substances have thus been abandoned. Conversely, calcium antagonists, nimodipine in particular, were shown no improve neurological outcome following subarachnoid haemorrhage and may also be of benefit in patients with traumatic subarachnoid haemorrhage. This effect is probably not related to vasospasm, but rather is a 'neuroprotective' property of the agent. It might be hypothesized that nimodipine also interferes with cellular/cytotoxic oedema generation in ischaemia and trauma.

At the moment various glutamate receptor antagonists are being tested both experimentally and clinically following stroke and brain trauma. The crucial role of glutamate in cellular/cytotoxic oedema has been mentioned before. Needless to say, this is a fundamental rational approach which might prove the concept of excitotoxicity and cytotoxic brain oedema.

Interference with inflammatory processes, like inhibition of cytokines as, for example, by tumour necrosis factor and so on, might be another potential avenue to search for a more rational treatment under certain circumstances.

Conclusion

Brain oedema is an important factor determining the course of numerous cerebral diseases. The distinction between different prototypes of oedema is helpful for the understanding of oedema formation and resolution. Under clinical conditions, there is most often a combination of cellular/cytotoxic oedema with extracellular vasogenic oedema. Oedema itself is not necessarily harmful and is principally reversible. It may, however, lead to an increase in intracranial pressure and decreased cerebral perfusion pressure, and eventually to herniation. Under clinical conditions, brain oedema is diagnosed by CT and MRI supplemented by contrast enhancement to visualize a defect of the blood brain barrier, that is, the vasogenic oedema component. Treatment of cerebral oedema is still largely symptomatic. Steroids are clearly useful for the treatment of peritumoral oedema incluidng abscesses, while their predominant mode of action is still unclear. All other modalities to reduce brain oedema are primarily directed towards a generalized dehydration of the brain or to decrease cerebral blood volume and intracranial pressure. Further progress towards a more specific antioedematous treatment can only be made if our knowledge about oedema formation and inhibition of oedema-mediating substances and processes is explained.

BACK

The Paraneoplastic Syndromes

The clinical manifestation of cancer are usually non-specific - eg, anorexia, malaise, weight loss, fever - or are due to local effects of tumor growth either in the primary site or at a distant site. The term "paraneoplasia" has been coined to denote the remote effects of malignancy that cannot be attributed either to direct invasion of metastatic lesions. These syndromes may be the first sign of a malignancy and may affect up to 15% of patients with cancer.

The paraneoplastic syndromes are of considerable clinical importance for the following reasons :

The paraneoplastic syndromes are usually caused by the secretion of proteins not normally associated with a cancer's normal tissue equivalent. Clinical findings may resemble those of primary endocrine, metabolic, hematologic, or neuromuscular disorders. The mechanisms for such remote effects can be classified into three groups : (1) effects initiated by a tumor product, (2) effects due to the destruction of normal tissues by tumor products, and (3) effects due to unknown mechanisms such as unidentified tumor products or circulating immune complexes stimulated by the tumor. Even such nonspecific symptoms as fever and weight loss are truly paraneoplastic and are due tot he production of specific factors by tumor cells or by normal cells in response to the tumor.

Paraneoplastic syndromes associated with ectopic hormone production are the best-characterized entities. Tumor cells secrete a hormone or prohormone that may be of a higher or lower molecular weight that may be of a higher or lower molecular weight than hormones secreted by the more differentiated normal endocrine cell. This ectopic hormone production by cancer cells is believed to result form activation of genes in malignant cells that are normally suppressed in most somatic cells. A single syndrome such as hypercalcemia may be due to more than one of a variety of causes. Effective antitumor treatment usually results in return of the serum calcium to normal, though additional therapy may be required. Several neurologic paraneoplastic syndromes have been found to be caused by the production of antineuronal antibodies that circulate in the serum and spinal fluid. It is thought that the underlying tumor expresses a similar antigen, resulting in production of a cross-reactive antibody. Treatment of the underlying tumor usually results in only modest improvement of the neurologic deficit. Examples of antineuronal antibodies include the anti-Hu antibody causing sensory neuropathy or encephalitis, associated with small cell cancer of the lung; the anti-Yo antibody causing cerebellar degeneration, associated most often with breast or gynecologic malignancies; the stiff man syndrome, associated with breast cancer; and the anti-Purkinje cell antibodies causing cerebellar ataxia, associated with Hodgkin's disease as well as gynecologic breast, and lung cancers.

Other well-described paraneoplastic syndromes include those involving the skin with or without other organ involvement, hematologic syndromes and those involving the kidneys, the gastrointestinal tract, and the joints.

The most common cancer associated with paraneoplastic syndromes is small -cell cancer of the lung. This is thought to be due to its neuroectodermal origin.

BACK

ACUTE IDIOPATHIC FACIAL PARALYSIS (BELL'S PALSY)

DEFINITION
A unilateral, lower motor neuron facial paralysis that is probably due to active viral inflammatory demyelination of the facial nerve causing swelling and secondary nerve ischemia within the facial canal.

EPIDEMIOLOGY
The most common cause of facial paralysis.

• Incidence : 20 - 30 per 100 000 persons per year.
• Age : any age
• Gender : M = F

PATHOLOGY

• Inflammation of the facial nerve is present within the subarachnoid space, composed largely of lymphocytes with associated demyelination or axonal destruction or both.
• Studies of autopsy material have shown latent herpes simplex virus (HSV) in the geniculate ganglia of humans.

ETIOLOGY AND PATHOPHYSIOLOGY

• Uncertain
• Probably acute viral inflammatory demyelination within the facial canal, causing swelling of the nerve within the facial canal and secondary ischemia.
• Viral infection:

  - HSV is the strongest association: reactivation of HSV 1 genomes. HSV might be present in the geniculate ganglia where it could cause a facial nerve palsy when the virus travels down the nerve axon, perhaps infecting the Schwann cells.
  - Mumps
  - Epstein - Barr virus
  - Cytomegalovirus
  - Coxaskievirus
  - Influenza
  - HIV

• Autoimmune reaction.

Risk factors

• Diabetes : fivefold increased risk
• Pregnancy : threefold increased risk

CLINICAL FEATURES

• A viral prodrome is present in about 60% of patients.
• Pain behind the ear, evolving over about 48 hours before reaching a plateau.
• Unilateral facial weakness of lower motor neuron type follows the pain, and may be associated with excessive tearing due to weakness of the orbicularis oculi (which normally holds the lacrimal puncta against the conjunctiva)
• Ipsilateral facial numbness is a common symptom, but objective sensory testing is normal.
• Taste may be altered.
• Sensitivity to sound (hyperacusis) may be increased, such as when using a telephone.

DIFFERENTIAL DIAGNOSIS

Lower motor neuron facial weakness

• Isolated unilateral lower motor neuron facial weakness, without other neurologic deficits.
• Direct facial nerve infections : herpes viruses (e.g., reactivation of the varicella-zoster vitrus (Ramsay Hunt syndrome): vesicles in the ear from which varicella-zoster virus can be easily recovered; seroconversion often, with increase in specific antibody to VZV); Lyme disease.
• Ischemic facial neuropathy due to small vessel disease (e.g. Vasculitis)
• Sarcoidosis, Behcet's syndrome, Sjogren's syndrome, syphilis.
• Middle ear tumors
• Temporal bone fracture
• Temporal bone tumors: metastatic, invasive meningioma
• Parotid gland tumors or infections
• Facial lacerations.

Upper motor neuron facial weakness

• Stroke
• Brain tumor
• Brain abscess

INVESTIGATIONS

• Not usually necessary, unless there is doubt about the diagnosis.

Blood serology titres
A light preponderance of elevated titres against HSV, compared with controls but increased titres are the exception rather than the rule.

CSF
Inconsistently shows mildly elevated cell counts and protein levels.

MRI brain Scan

• Indicated when a brain lesion is suspected (e.g. when more than an isolated lower motor neuron facial palsy is present).
• MRI scans are more sensitive than CT brain scan of the posterior fossa, which is limited by bone artifacts.
• Reveals gadolinium-induced enhancement of the facial nerve in acute causes of Bell's palsy.

PCR techniques

• These are used to amplify viral genomic sequences from endoneurial fluid collected from the facial nerve or tissue of the posterior auricular muscle innervated by the facial nerve is patients undergoing decompressive surgery.
• In one study, HSV type 1 genomes were amplified from the nerve or muscle tissue of 11 of 14 patients with Bell's palsy. However, PCR does not distinguish between viral genomes that are present in a latent state and those that are present because of reactivated virus in the course of a productive infection.

Electrodiagnostic studies
Electroneurography may be used to prognosticate recovery, but not to make the diagnosis.

DIAGNOSIS

A clinical diagnosis of exclusion:

• Lower motor neuron facial weakness (peripheral facial nerve palsy)
• No other neurologic deficits.
• No apparent cause, other than a herpes simplex virus, after prolonged follow-up (until the condition has resolved or fresh clues to a specific cause appear).

TREATMENT

Corticosteroids
Controversial, party because of the good prognosis of the untreated condition and the failure of controlled trials to prove a beneficial effect on long term outcome: Nevertheless, steroids are used empirically by some neurologists to :

• Receive pain that is sometimes an early failure
• Prevent progression of incomplete to complete paralysis.
• Prevent denervation in cause of complete paralysis
• Shorten the time to recovery
• Retard development of abnormal synkinesias.

It is likely that any benefit obtained is due to steroids used early in the course. The usual regime is prednisolone, 25mg per day, orally, for 1 week, with patient review on completion. Corticosteroids should not be used when contraindications exist, such as diabetes, hypertensions, peptic ulcer disease, osteoporosis, glaucoma, or pregnancy

Acyclovir
May be effective if the underlying cause is herpes virus infection. More data are needed from a large, randomized controlled and blinded trial with at least 12 months' follow up.

Other treatments
There is no proven place for adjunctive therapies or surgical decompression of the facial nerve in Bell's palsy.

Avoid complications
Exposure keratitis

• Occurs if the cornea is not adequately protected.
• It can be avoided by using artificial tears, instilling lubricating paraffin ointment, and taping (rather than padding) the eye closed at night.
• Dark glasses should be worn outdoors.
• Ophthalmic advice should be sought if the patient reports eye discomfort or the eye becomes irritated despite the above measures.
• Botulinum toxin injection into the eyelid levator to weaken it may be considered if conservative measures fail.
• Tarsorrhaphy is rarely necessary in cooperative patients.

PROGNOSIS

60 - 80% of patients recover completely. In these cases, recovery usually begins within 8 weeks and is complete by 6 - 12 months. The most favorable prognostic sign is an incomplete rather than complete facial palsy. If weakness in severe or complete, recovery commencing within 3 weeks is a favorable sign. The longer the delay in return of movement the poorer the recovery.

Predictors of incomplete recovery are :

• Complete facial weakness
• Pain other than in or around the ear(i.e. back of head, cheek, other).
• Systemic hypertension, diabetes or pyschiatric illness.
• Older age
• Hyperacusis
• Decreased tearing.

Residual deficits include:

• Facial weakness
• 'Jaw winking' and other abnormal facial movements caused by aberrant reinnervation of the muscles of facial expression by regenerating facial nerve fibers. (i.e. a miswiring).
• 'Crocodile tears' (rare): an excessive flow of tears when eating, caused by aberrant reinnervation of the lacrimal gland by regenerating facial nerve fibers.

Recurrent facial palsy occurs in about 10% of patients. If this occurs, alternative causes should be excluded, such as diabetes, sarcoidosis, tumors, or infection.

BACK

CENTRAL NERVOUS SYSTEM VASCULITIS

DEFINITION

A heterogeneous group of disorders characterized by histologic evidence of inflammation, and often necrosis, of blood vessels and clinico-pathologic evidence of brain ischemia or, less commonly, hemorrhage.

EPIDEMIOLOGY

• Incidence : rare
• Age : any age
• Gender : either sex

CLASSIFICATION

Large arteries (aorta and its primary branches) Takayasu¡¦s arteritis.
Large and medium ¡V sized arteries Giant cell (temporal) arteritis.

Medium ¡V sized and small muscular arteries

Primary systemic necrotizing angiitides

• Polyarteritis nodosa
• Allergic angiitis and granulomatosis of Churg-Strauss.
• Polyangiitis overlap syndrome
• Wegener¡¦s granulomatosis
• Lymphomatoid granulomatosis.

Angiitis associated with other systemic diseases

• Sarcoidosis
• Behcet¡¦s disease
• Relapsing polychondritis
• Inflammatory bowel disease
• Kohlmeier ¡V Degos disease.

Hypersensitivity angiitis associated with connective tissue disease

• Systemic lupus erythematosus.
• Mixed connective tissue disease
• Sneddon¡¦s syndrome
• Rheumatoid arthritis
• Sjoren¡¦s syndrome
• Scleroderma.

Primary isolated angiitis of the CNS

Isolated granulomatous angiitis of the CNS.

Other angiitis syndromes

• Eale¡¦s disease
• Radiation angiitis.

Small vessels (arterioles, capillaries, venules)

Hypersensitivity angiitis

Exogenous stimuli proved or suspected:
- Drug-induced angiitides.
- Henoch ¡V Schonlein purpura.
- Serum sickness and serum sickness ¡V like reactions.
- Angiitis associated with infectious diseases.

Endogenous antigens likely to be involved:
- Angiitis associated with neoplasms (particulary lymphoid malignancies)
- Angiitis associated with other underlying diseases
- Angiitis associated with congenital deficiencies of the complement system (hypocomplementemic angitis).
- Mixed cryoglobulinemia.
- Cutaneous angiitides.

PATHOLOGY
Acute, subacute or chronic inflammation in the arterial and/or venous
wall with or without granuloma formation and necrosis.

Vascular lesions
Granulomatous angiitis
A distinctive chronic inflammatory reaction of blood vessels characterized by a predominance of modified macrophages which are aggregated into nodular clumps referred to as granulomas and which respond to foreign bodies by coalescing to form giant cells that often conglomerate around the
foreign body.

Necrotizing angiitis
Inflammation and necrosis (usually fibrinoid necrosis) of vessel walls.

Brain lesions

• Focal or multi-focal cerebral infarction
• Intracerebral hemorrhage
• Subarachnoid hemorrhage.

ETIOLOGY

A primary manifestation of disease or a secondary component of another disorder such as connective tissue disease, drug abuse, neoplasia or infection.

PATHOGENESIS

Immunopathogenic

• In situ formation or deposition of immune complexes in blood vessel wall leading to activation of the complement-mediated inflammatory response.
• Direct antibody-mediated damage via antibodies directed at endothelial cells or other tissue components.
• Antibody-dependent cellular cytotoxicity directed against blood vessels.
• Cytotoxic T lymphocytes directed at blood vessel components.
• Granuloma formation in blood vessel wall or adjacent to blood vessel.
• Cytokine-induced (i.e. interleukin 1, TNF-alpha) expression of adhesion vehicles for leukocytes on endothelial cells.

Non-immunopathogenic

• Infiltration of blood vessel wall or surrounding tissue by microbiologic agents.
• Direct invasion of blood vessel by neoplastic cells.
• Unidentified mechanisms.

Mechanisms of tissue dysfunction

• Angiitis (Causing ischemia or hemorrhage)
• Coagulopathy
• Emboli
• Compression from granulomas.
• Antineuronal antibody effects.

CLINICAL FEATURES

• Variable in onset, nature and duration.
• Features are determined partly by the size and location of the involved vessel(s).
• Most commonly vasculitis presents as an acute or subacute focal or diffuse encephalopathy or meningo-encephalopathy with headache, altered mentation, seizures and cognitive and behavioral abnormalities, with
  multifocal neurologic signs.
• Less commonly, patients present with a multiple sclerosis-like
  picture, features of a rapidly progressive space-occupying lesion, multiple
  cranial neuropathies and rarely, with spinal cord syndrome, extrapyramidal
  syndrome or stroke syndrome.
• Systemic symptoms and signs may be present such as fever, headache,
  malaise, weight loss, joint aches and pains, facial rash, livido
  reticularis.

CLINICAL HISTORY

Demographic data

• Age : young : Takayasu¡¦s arteritis, SLE; older: GCA.
• Gender : F: Takayasu¡¦s arteritis
• Race : oriental: Takayasu¡¦s arteritis.

Symptoms

• Headache : GCA.
• Scalp, face, temporal pain: GCA
• Blindness: GCA, WG.
• Diplopia: WG
• Syncope : Takayasu¡¦s arteritis.
• Jaw / tongue claudication: GCA. Takayasu¡¦s arteritis.
• Arm claudication: Takayasu¡¦s arteritis
• Sinus pain and drainage, nasal discharge : WG.
• Oral ulcers : SLE, Behcet¡¦s disease
• Dry eyes, dry mouth: SS
• Anxiety, depression: SLE.
• Fever, headache, malaise, fatigue, weight loss, arthralgia, sweats:
  non specific.
• Skin rash : PAN, AAG, LG, Sarcoid, Behcet¡¦s disease, SLE,
  Sneddon¡¦s hypersensitivity.
• Malar rash: SLE.
• Photosensitivity : SLE
• Joints aches and pains : SLE, RA.
• Stiffness/ aches/ pains in neck, shoulders, lower back, hips and legs:
  GCA.
• Chest pain: SLE, LG, Behcet¡¦s.
• Asthma: AAG.
• Cough, shortness of breath: LG, Sarcoid, Behcet¡¦s/
• Abdominal pain: PAN, inflammatory bowel disease, Kohlmeier ¡V Degos disease.

Past history

• Allergic : AAG
• Deep vein thrombosis / pulmonary emboli: SLE, APLAb, Behcet¡¦s.
• Recurrent spontaneous abortions: SLE, APLAb.
• Epileptic seizures : SLE.
• Anemia : non-specific
• Thrombocytopenia : SLE, APLAb.
• Positive CDRL: APLAb.
• Infection with toxoplasma, aspergillus, varicella-zoster, cytomegalovirus, herpes simplex virus, HIV : infectious arteritis.
• Illicit drug abuse: amphetamine, cocaine: drug induced arteritis.

Family history

• Collagen vascular diseases or angiitides.
• Multiple spontaneous abortions: APLAb.
• Deep venous thrombosis or pulmonary emboli: APLAb.
• Neonatal heart block in patient¡¦s child.

PHYSICAL EXAMINATION

• Malar rash: SLE
• Skin nodules, purpura: AAG, LG, Sarcoid, Kohlmeier-Degos disease, hypersensitivity.
• Skin macules, papules, plaque: LG Sneddon¡¦s syndrome.
• Skin infarcts
• Skin fibrosis : scleroderma.
• Oral ulcers: SLE, Behcet¡¦s disease.
• Ischemic necrosis of lips, palate, nasal septom: TA WG
• Sinus inflammation / nasal mucosal ulceration : WG.
• Chondritis (auricular, nasal, laryngo-traneal): relapsing polychondritis.
• Altered mental state, dementia, psychosis: SLE.
• Blindness or altitudinal visual field defect: GCA.
• Uveitis: Sarcoid, Behcet¡¦s disease.
• Optic nerve head swelling / pallor/ hemorrhage: GCA
• Retinal vein occlusion: Behcet¡¦s disease
• Mononeuritis mulplex: PAN, AAG, WG, LG.
• Temporal artery tenderness, thickening, nodularity: GCA.
• Absent peripheral pulses : TA.
• Carotid / chest bruit : TA
• Hypertension : TA, PAN
• Blood pressure difficult to record or different between the arms: TA.
• Aortic regurgitation: TA
• Pleuritic or pericardial friction rub: SLE.
• Chest Crackles / wheeze: AAG, LG, SLE.
• Arthritis: Sarcoid, SLE RA.

DIFFERENTIAL DIAGNOSIS

Non-vascular disorders

• Dementias.
• Meningo-encephalitides
• Multiple sclerosis.

Cardiac disorders

• Non-bacterial thrombotic endocarditis
• Cardiac tumor.

Hematologic disorders

• Coagulopathies (paraproteinemia, thrombotic thrombocytopenic purpura).
• Antiphospholipid antibody syndrome.

Angiopathies (non-inflammatory)

Isolated angiopathy of the CNS

• Arterial dissection.
• Vasoconstriction or due to drug use (methamphetamine or cocaine).
• Cholesterol embolization syndrome.
• Malignant angioendotheliomatosis (intravascular lymphomatosis).
• Neurofibromatosis.
• Atherosclerosis
• Fibromuscular dysplasia.
• Moyamoya disease.
• Angiitis.

INVESTIGATIONS

MRI brain*

For imaging brain parenchymal lesions. It may show:

• Normal brain
• Areas of increased signal on T2 and PDW images typically in peripheral white matter and gray matter. There may be evidence of hemorrhage in these lesions.

Cranial CT scan may be normal or show single or multiple non-enhancing areas of low density, which may involves both cerebral gray and white matter.

Intra-arterial angiography*

For imaging arterial lesions:

• May show areas of alternate narrowing and dilatation of intracranial arterial branches, or areas of extracranial arterial occlusion.
• In primary angiitis of the brain the small intracranial arteries and arterioles are involved, whereas in vasculitis complicating meningitis, tumors or other cause the major basal intracranial arteries may be
  involved. However, the angiogram may be normal even in biopsy-proven
  vasculitis, particularly if the affected vessels are <500 microns in diameter and
  too small to be seen, so biopsy is strongly recommended.
• Magnetic resonance angiography is not reliable enough to confirm or exclude vasculitis; it is too prone to flow-related artefacts and shows insufficient detail to image the peripheral small arteries
  satisfactorily. Spiral CT has not been widely evaluated.

Causes of segmental narrowing of cerebral arteries on cerebral angiography

• Arteritis (non-infective, infective).
• Arterial dissection
• Vasospasm (drugs, subarachnoid hemorrhage, severe hypertension, migraine).
• Arteriosclerosis.
• Fibromuscular dysplasia.
• Moyamoya disease
• Radiation
• Leptomeningitis (infective, carcinomatous, chemical).
• Multiple emboli / recanalizing embolism.
• Intracerebral hematoma
• Sickle cell disease.
• Neoplasia (vascular, malignant angioendotheliosis, meningeal, glial, metastatic: atrial myxoma).
• Trauma: closed head injury.
• Surgical manipulation
• Neuroectodermal dysplasia.
• Systemic inflammation
• Full blood count* :
  - Anemia: TA, GCA, PAN, WG, SLE.
  - Neutrophilia : PAN, WG.
  - Lymphocytopenia : LG, SLE.
  - Eosinophilia: AAG, hypersensitivity.
  - Thrombocytosis : GCA, PAN
  - Thombocytopenia: SLE, APLAb.
• ESR* : TA, GCA, PAN, WG, hypersensitivity.
• C- reactive protein*.
• Immunoglobulins A, G, M, E*:
  - Hypergammaglobulinemia: TA, WG, SLE, hypersensitivity.
  - Hypogammaglobulinemia: PAN
• Complement C3, C4, C9* :
  - Hypocomplementemia: SLE, hypersensitivity
• Circulating immunocomplexes

Underlying connective tissue disease Nucleic acids

• Antinuclear antibodies* : SLE

DNA

• ds DNA : SLE

Ribonucleoproteins

• UI ribonucleoprotein (RNP)*: MCTD, SLE, SS (sensitivity high, specificity low).
• P Sm* : SLE (Sensitivity 5 ¡V 30%, specific)
• P Ro (SS-A)*: SS/SLE overlap (high sensitivity and specificity for SS)
• La (SS-B)*: SS/SLE overlap (high sensitivity and specificity for SS).

DNA binding proteins

• Histones : drug induced SLE.
• Ku : Sclerodactyly
• PCNA: severe SLE.

Cell membrane antigens

• Cardiolipin: APLAb, SLE (Sensitivity 20 ¡V 30%, specificity low).

Other

• IgM rheumatoid factor-latex fixation test*: RA (Sensitivity 80%, specificity low).
• Antineutrophil cytoplasm antibodies*: WG (Sensitivity and specificity 90%).

Extent of visceral and other organ involvement

• Urine
  - Glomerular red cells: PAN, WG, LG, SLE.
  - Casts*
  - Protein (if positive, 24 hour urine protein)*.
  - Eosinophilis.
• Creatinine : PAN, WG, LG, SLE.
• Liver function tests: transaminases*: GCA
• Hepatitis B surface antigenemia : PAN.
• Chest X-ray* : WG, LG, SLE.
• Gallium scan: Sarcooid.
• Ophthalmologic examination using low dose fluorescin angiography with slit-lamp video microscopy of the anterior segment: slowing of flow, multifocal attenuation of arterioles, erythrocyte aggregates, area of
  small vessel infarction, and multifocal segments of intense leakage from post
  capillary and collecting venules.

Underlying infection

• Antibodies against borreliosis*, salmonellosis, yersiniosis, toxoplasmosis, aspergillus.
• Viral antibodies
• Hepatitis screening
• VDRL, TPHA
• Cryoglobulins*
• Paraproteins (Serum protein electrophoresis)*
• CSF*
  - Cell count, total protein, glucose, quantitative analysis of
  immunoglobulins, oligoclonal band analysis.
  - Moderate mononuclear CSF pleocytosis
  - CSF protein usually elevated to round 1 ¡V 1.5 g/1, but may be up to 5.8 g/1.
  - Xanthochromia, hypoglycorrhachia and an elevated CSF gammaglobulin with
  oligoclonal banding of IgG have been reported.
  - The primary value of CSF studies is to exclude other causes of a
  meningitis.

Associated coagulation abnormalities

• PT, APTT, dRVVT*: APLAb, SLE
• Anticardiolipin antibodies*: APLAb, SLE

Others

• Angiotensin converting enzyme: sarcoidosis.
• Lysozyme and beta 2-microgrobulin: sarcoidosis.
• Cell-mediated immunity : sarcoidosis, LG.
• Sinus x-ray/CT: WG.
• Echocardiography: atrial myxoma
• EEG : frequently abnormal, showing generalized non-specific slow wave
  activity of variable degree and location.
• Indium-labelled white-cell brain imaging: increased uptake and accumulation in the brain imaging: increased uptake and accumulation in the brain of indium-labelled leukocytes.

Brain Biopsy

• Cerebral or spinal cord biopsy establishes the diagnoses in about 70% of cases
• False-negative biopsies occur in about 30% of autopsy-proven patients.
• The relatively low diagnostic yield from cortical biopsy may be related
  to the segmental nature of the lesions and the lack of leptomeningeal tissue in the biopsy.
• The risk of serious morbidity from brain biopsy is about 0.5 ¡V 2.0%.
• The decision to biopsy is made on an individual basis; if an extended period of treatment with potentially hazardous immunosuppressant drugs is being considered, then the need to establish an unequivocal tissue
  diagnosis becomes of primary importance.

The result of blood and CSF laboratory tests, EEG, brain imaging with CT and MRI, and cerebral angiography are neither sensitive nor specific but are usually essential to role out infectious or malignant disease, which
can mimic CNS angiitis clinically.

DIAGNOSIS

The diagnosis is made histologically. A combined leptomeningeal and wedge cortical tissue biopsy, preferably of the temporal tip of the non-dominant hemisphere and including a longitudinally ¡V orientated surface vessel
is required. If organs other than the brain are affected, biopsy specimens should be obtained.

TREATMENT

The decision to treat and how will depend on the clinical condition and course of the patient and the philosophies of the attending clinician. If the patient is well then time is available to allow a period of observation
of the natural history of the disease. If the patient is very sick, however, empirical immunosuppressive therapy, may be commenced whilst awaiting biopsy confirmation.

Once the histologic diagnosis is confirmed then anti-inflammatory and immunosuppressive therapy should be considered, bearing in mind that the treatment of CNS angiitis is inferred from clinical experience with
systemic angiits or intravenous glucocorticoid in divided doses every 8 ¡V 12
hours.

After the disease is controlled, reduce to one morning dose, and thereafter, taper the daily dose as rapidly as clinical disease permits. Ideally, patients should be slowly converted to alternate day therapy with a single morning dosage of short active glucocorticoid so as to minimize adverse effects; prednisone doses of 15 mg daily given before noon usually do not suppress the hypothalamic pituitary axis. However, the disease may flare on the day off steroids, in which case use the lowest single daily dosage that suppresses disease.

Strategies to minimize adverse effects of steroid include:

• Calcium for most patients to minimize osteoporosis.
• Vitamin D 50 000 units one to three times weekly if 24-h urinary calcium excretion <120 mg,
• Estrogen replacement therapy if post-menopausal.
• Calcitonin and biphosphonates may also be useful.
• Hyperglycemia, hypertension, edema and hypokalemia should be treated.
• Infections should be identified and treated early.
• Immunizations with influenza and pneumococcal vaccines are safe and should be given if the disease is stable.

The addition of cyclophosphamine to prednisolone may be effective.

Strategies to minimize adverse effects of cyclophosphamide include:

• Monitor the leukocyte count closely, keeping it above 3000 per cubic millimeter and the neutrophil count about 1500 per microlitre.
• If severe neutropenia or bladder toxicity occurs despite low doses of cyclophosphamide and a fluid intake of >3 liters/day, azathioprine 1 ¡V 2 mg/kg per day or methotrexate can be used successfully with prednisone.
• A pulsed regimen of 10-15 mg/kg intravenously once every 4 weeks has less bladder toxicity that daily oral doses but bone marrow suppression can be severe.

When the disease has been controlled for a few months, taper immunosuppressive agents and attempt to discontinue them. The role of antiplatelet agents and anticoagulants is uncertain. Campath- 1H
humanized monoclonal antibody treatment remains experimental.

Clinical approach to the patient with suspected CNS vasculitis
1. Is it cerebrovascular
2. Where is the lesion neuroanatomically?
3. What is the pathologic nature of the lesion: ischemic or hemorrhagic?
4. What is the etiology of the lesion: have more common disorders of the
arteries, heart and blood been excluded?
5. Are there other clinical features or investigation results to indicate that this is part of a specific vasculitic syndrome.
6. If a syndrome is recognized and if it is associated with an underlying disease or an offending antigen, treat the underlying disease or remove the offending antigen where possible.
7. Establish the histologic diagnosis of angiitis by obtaining a tissue biopsy before committing the patient to a long course of immunosuppressive medication.
8. Determine the extent of disease activity
9. Start treatment with appropriate agents in disorders in which treatment is of proven benefit and is essential
10. In patient with systemic angiitis , start with glucocorticoid therapy.
Add a cytotoxic agent such as methotrexate or cyclophosphamide if an adequate response does not result of if the disorder is likely to respond only to cytotoxic agents, such as Wegener¡¦s granulomatosis.
11. Avoid immunosuppressive therapy in disorders which rarely result in irreversible organ system dysfunction and which usually do not respond to such agents.
12. Closely follow patients for development of toxic adverse effects of treatment.
13. Continually attempt to taper glucocorticoids to an alternate day regimen and discontinuation when possible, and to taper and discontinue cytotoxic drugs as soon as is feasible upon induction of remission.
14. In the event of unacceptable adverse effects or lack of efficacy, consider alternative agents such as azathioprine.

Prognosis
Variable.

BACK

Mannitol

Mannitol, a derivative of the carbohydrate mannose, was introduced as an osmotic diuretic over 30 years ago. Since then, it has been successfully and eclectically employed in many areas of medicine. These include transfusion medicine, where it prolongs red blood survival in stored blood, as a pharmaceutical excipient, e.g. with dantrolene, in ophthalmology to reduce intraocular pressure, and as a substitute glycine in urology. Mannitol has also been employed as an adjunct to chemotherapy for intracranial malignancies, where it is used to improve the penetration of cytotoxic drugs given selectively via the carotid artery. However, the areas in which it excites most controversy and confusion are as a protective and therapeutic agent in situations of neurological or renal compromise and as a potential scavenger of free oxygen radicals. This article attempts to review these roles and assess the value of mannitol for these indications.

MANNITOL AND NEURO CRITICAL CARE

HISTORY

The use of osmotic agents to reduce cerebral water content was suggested by the work of Weed and McKibben in 1919. They noted cerebral dehydration after hypertonic saline in contrast to cerebral oedema following intravenous distilled water. Amongst the earliest proponents of osmotherapy, as it was then called, were Fremont-Smith and Forbes, who used hyperosmolar urea of reduce intracranial pressure (ICP) during neurosurgery in the late 1920s. Mannitol was introduced in the early 1960s after urea was found to be associated with rebound intracranial hypertension. Since then, it has become the most widely used agent to control ICP in both elective neurosurgery and following traumatic brain injury (TBI).

POSSIBLE MECHANISMS OF NERUOPROTECTIVE EFFECTS

Osmotic action
Because mannitol is hyperosmolar relative to intracellular fluid, intravenous administration results in movement of 'free water' from the tissues into the plasma. There is a more pronounced effect in the brain due to the fact that the blood-brain barrier (BBB) differs from all other capillaries membranes in being relatively impermeable to mannitol, thus maintaining the osmotic gradient.

In support of the osmotic theory, the work of Cascino' has shown that even small changes in total brain water content can translate into relatively large alterations in brain volume. On the other hand, clinically significant changes in brain water content can take up to 30 min to develop, whereas the observed reduction in ICP following mannitol usually far sooner. There is also a discrepancy between the large amounts of mannitol needed to reduce brain water experimentally and the much smaller doses known to reduce ICP clinically. It is likely, therefore, that any osmotic effect is more responsible for the sustained, rather than immediate, reduction in ICP seen with mannitol administration.

Haemodynamic and Viscosity changes
Under normal physiological conditions, ICP is principally determined by cerebral blood volume (CBV). In turn, the CBV is controlled by a number of variables, the most important of which are those influencing oxygen delivery to the rain. These include cerebral perfusion pressure (CPP), cerebral vascular resistance and blood viscosity. When intracranial compliance is reduced, such as after a traumatic brain injury (TBI), any manoeuvre which improves oxygen delivery reduces CBV, and thus ICP, by inducing vasoconstriction of cerebral blood vessels.

Mannitol probably improves oxygen delivery both by increasing CPP and reducing blood viscosity. The increase in CPP comes from its plasma expanding effect, which causes an increased cardiac output. Factors favouring viscosity reduction, include haemodilution, decreased adhesiveness of red cells and increased red cell deformability, thereby reducing resistance to capillary blood flow. These changes are thought to be most responsible for the prompt reduction in ICP seen clinically with mannitol, given that they occur soon after administration.

Diuretic action of mannitol
Mannitol could decrease ICP through a fall in CVP as a result of the osmotic diuresis. However, this is unlikely given that ICP reduction occurs long before the diuresis, and a normal or elevated CVP is not incompatible with a good response to mannitol. Even patients with renal failure, who have a persistently elevated CVP, can have a brisk reduction in ICP with mannitol.

Mannitol and CSF production / resorption
Although never proven satisfactory and probably of little significance, mannitol may increase resorption and reduce production of CSF by increasing plasma osmolality. The site of action is postulated to occur at the level of the pial circulation or at the cerebral ventricles.

Mannitol as an oxygen free radical scavenger
An additional, although theoretical, beneficial effect of mannitol could be a reduction of ischemic damage and oedema by a free radical scavenging action. This may be as a result of specific interference with oedema promoting factors or a reduction in brain tissue necrosis from the primary injury.

MANNITOL IN TRAUMATIC BRAIN INJURY

It seems that regardless of the intervention strategy chosen, the outcome from a severe TBI is bleak. Nevertheless, it is accepted that reducing a raised ICP is a useful manoeuvre, the rationale being that a fall in ICP to near normal limits improves CBF and CPP, and increases oxygen delivery to the brain which, in turn, reduces anaerobic glycolysis and hence cerebral oedema. Indeed, one study showed that a persistently raised ICP resulted in a mortality of > 90%, while successful reduction of near normal limits decreased mortality of 26%.

Although mannitol has become the cornerstone of the pharmacological management of raised ICP after TBI, it has never been subjected to a controlled trial against placebo. There has been one study comparing its efficacy with barbiturates, where mannitol was shown to have a better outcome mortality of 41% compared to 77% for pentobarbital. Certain factors do indicate its ability to produce a prompt, persistent reduction of ICP to or below 15 mmHg, to reduce an initially high ICP of >50 mmHg or the ability to maintain CPP at or below limit of autoregulation. Diffuse widespread injuries respond more than focal injury and there is a greater benefit seen in patients < 40 years of age with an initial systolic blood pressure > 90 mmHg. There was a suggestion that mannitol was less effective after loss of cerebral blood flow autoregulation, but this is no longer thought to be the case.

It is becoming increasingly apparent, however, that the relationship between ICP, CPP and development of cerebral ischaemia is far clear-cut. Following TBI, certain factors adversely influence CBF or contribute directly to cerebral damage without having much effect on ICP. These include loss of autoregulation, traumatic vasospasm or neurochemical changes from elevated excitatory amino acids. This suggests that, in addition to measuring ICP, it is equally important to assess adequacy of the CBF, and thus oxygen delivery to the brain. The development of techniques such as somatosensory evoked potentials and jugular bulb oxygen monitoring now makes this possible. In the future, therefore, it is hoped that having a multimodal approach to monitoring cerebral function after a TBI will help direct therapeutic interventions, including the use of mannitol, more appropriately so as to optimise CPP further and ameliorate the consequences of secondary ischaemic damage to the brain.

PRACTICAL ISSUES FOR ADMINISTRATION OF MANNITOL

Effective dose
The initial dose of 2 g/kg was derived from a comparison with urea. The recommended dose has fallen in recent years to between 0.25 - 1 g/kg. Mannitol shows substantial inter-patient variability, making it difficult to predict an effective dose for every patient, but larger doses appear to prolong the effect of ICP. Mannitol should be used with caution in renal impairment as it is cleared exclusively by the kidney.

Rate and timing of administration
There is no firm consensus on this, although mannitol is usually given by slow bolus infusion over 15 - 30 min to limit significant haemodynamic changes. There is no fixed interval for repeat boluses, but it is usually every 3 - 4 h, as this is the time taken for mannitol to clear from the circulation if renal function is normal. Tachyphylaxis may occure with more than 3 - 4 doses per 24 h. Continuous infusion is best avoided as small amounts of mannitol may pass into the brain and accumulate. This could potentially result in reverse osmotic shift with water entering the brain to compensate for the increased osmolality with the consequence of increasing cerebral oedema and exacerbating ICP.

Criteria for administration
Clinical indications for the use of mannitol in the absence of ICP monitoring include signs of transtentorial herniation or progressive neurological deterioration not attributable to systemic pathology. Optimal criteria based on ICP have not been clearly established. Smith, for example, failed to show any improvement in outcome after a severe head injury regardless of whether mannitol was given only when ICP was > 25 mmHg or empirically every 2 h until serum osmolality was 310 mOsm/1. Because cerebral oedema occurs soon after the primary event, it has been suggested that mannitol should be given as soon as possible after injury, even if this is outside a hospital setting. There have been reservations concerning potential adverse haemodynamic effects, but these have recently been questioned.

Optimal concurrent fluid therapy
Moderate fluid restriction is traditionally advocated in the acute management of TBI. However, this has to be balanced against the need for adequate fluid replacement to prevent dehydration, particularly given the diuretic effects of mannitol and the consequent risks of hyperviscosity and renal impairment. The use of large volumes of hypotonic 0.45% saline solution over 4 --8 h have been advocated to counteract the hyperosmolar effect of mannitol and also promote its rapid clearance so limiting renal impairment. Good hydration also maintains intravascular volume and optimises blood pressure and CPP.

COMPLICATIONS OF MANNITOL

There is some controversy concerning the importance and relevance of potential complications, as outlined below.

Haemodynamic changes
Although mannitol usually increases blood pressure slightly, it can occasionally cause a fall in blood pressure due to a decrease in systemic vascular resistance. Possible explanations for this include a fall in plasma pH, increased release of atrial natriuretic factor (ANF), basophil histamine release and direct impairment of the contractile properties of vascular smooth muscle. Any hypotension is usually transient, but it could compromise cerebral perfusion to an extent that offsets any potential benefit of mannitol. This is particularly true in volume depleted patients or after multiple trauma with a closed head injury, in whom even brief periods of hypotension correlated with a poor outcome. In these cases, the benefits from reducing ICP must be balanced against the risk of reducing CPP. In really compromised patients, care must be taken not to precipitate cardiac failure secondary to sudden volume expansion.

Paradoxical increases in ICP
Bolus infusions of mannitol can double CBV. This is not usually clinically significant, however, as CBV comprises such a small percentage of the total intracranial space. Any increases in ICP are usually mild and transient and are usually offset by the reduction in brain water and CSF volume. Mild hyperventilation, often used in TBI, also tends to conunteract any increases in CBV.

Dehydration and electrolyte disturbances
Dehydration is often concealed as mannitol shifts fluid into the intravascular compartment, thus preserving circulating volume so that clinical signs of haemodynamic compromise may not develop until intracellular dehydration is severe. Generally, there is electrolyte depletion, but it is proportionally much less than fluid loss. Sodium may increase in the long term as a result of the hyperosmolar state, but is often compensated by an increase in ADH, a common finding in the critically ill. Hyperkalaemia may also occur, rarely, from haemolysis.

Hyperosmolality and osmotic compensation
Sustained use of mannitol results in a hyperosmolar state, which leads to movement of osmotically active particles and electrolytes intracellularly. This increase in osmotic activity within the cell counteracts the dehydrating effect of hyperosmolar plasma and thus places finite limits on the reduction of brain volume mannitol can achieve. The threshold for osmotic compensation is not precisely known, but is thought to be 25 mOsm/kg above normal osmolality. Because of this, mannitol should not be used when plasma osmolality exceeds 320 mOsm/1.

Cerebral oedema may occur if a hyperosmolar state is reversed too quickly, and the speed of return to normal osmolality should approximately match the duration of the hyperosmolar state. Cautious use of isotonic saline is safer than hypotonic fluids such as enteral feeds or dextrose solutions.

Renal failures
Mannitol in excess of 200 g/day may cause acute renal failure (ARF). This is more likely in the elderly, in pre-existing renal failure and where patients are having concomitant treatment with other diuretics. It is probably due to structural and functional charges in the nephron, but prompt haemodialysis usually results in complete resolution.

THE FUTURE ROLE OF MANNITOL FOR NEUROPROTECTION

Mannitol has a well-established role in the short-term control of ICP for intracranial surgery and after TBI and this is the basis for its continued use. More recently, it has been prompted as a 'small volume resuscitation fluid' for cerebral protection and volume expansion in multiple trauma patients. However, mannitol is being increasingly challenged - by two other osmotic agents, glycerol and hypertonic saline.

Glycerol's effect on ICP and CPP are similar to those of mannitol but appear to be more consistent and sustained. Its diuretic action is less pronounced, favoring normovolaemia. Glycerol crosses the BBB but there is minimal accumulation in the cerebral tissues as it is metbolised by neuronal cells. Reverse osmotic shift is not, therefore, thought to be a risk. It may also provide nutritional support to the damaged cells thus having a protein sparing action. Its main side effect is haemolysis which can be limited by using a low concentration, a slow infusion rate and adding glucose, fructose or chloride to the solution. It has been suggested that glycerol be used as the first line treatment in the management of brain injured patients with increased ICP and impaired CPP, while mannitol is reserved for sudden increases in ICP.

Hypertonic saline is also being promoted as superior to mannitol. It is as effective in reducing brain volume and ICP in animal and human studies, in elective neurosurgery and in brain injured children. More significantly, it has also been found to be effective in reducing a raised ICP refractory to treatment with mannitol. Hypertonic saline also has the advantage of not significantly reducing intravascular volume. The diuretic action of mannitol, once considered beneficial, may, infact, worsen outcome in TBI as the damaged brain appears to be exceptionally vulnerable to a fall in CPP caused by hypovolaemia-induced hypotension.

Thus, although mannitol is currently the most popular osmotic agent, it may well be that hypertonic saline, the original osmotic fluid first used by Weed and McKibben almost a century ago, will become the treatment of choice for raised intracranial pressure in the next millennium.

MANNITOL AND RENAL PROTECTION

In 1945, mannitol was experimentally found to induce diuresis in dogs subjected to an ischaemic insult. This led to the idea that it might be of value in preserving renal function in humans. Since then, mannitol has been widely, but controversially, used in prophylaxis and treatment of acute renal failure (ARF).

POSSIBEL MECHANISMS OF RENAL EFFECTS

Mannitol is effectively inert and is freely filtered at the glomerulus without being secreted or substantially re-absorbed. Diruesis is brisk, with total body water loss up to 30% if the filtered amount. The precise mechanisms by which the diuresis occurs have not been fully elucidated, but include changes of systemic and renalhaemodynamics as well as local effects at the glomerular and tubular levels.

Systemic haemodynamic changes
Following intravenous administration, mannitol is confined mainly to the intravascular compartment causing a water shift from the intracellular compartment. This leads to intravascular volume expansion and to decreased plasma oncotic pressure, blood viscosity and haematocrit, all of which may improve renal perfusion. Mannitol increases plasma levels of the vasodilatory hormone, atrial natriuertic factor (ANF), probably as a result of intravascular volume expansion and may have a syngergistic action with it to improve renal perfusion.

Renal Haemodynamic effects
Infusion of mannitol has been shown to improve renal blood flow (RBF) and glomerular filtration rate(GRF) by reducing renal vascular resistance (RVR). This effect is explained mainly by a reduction in plasma viscosity, although mannitol may also stimulate release of vasoactive agents such as prostacyclin. This increased renal perfusion is associated proportional increases in both cortical and medullary blood flow. The latter may be responsible for the loss of medullary hypertonicity observed during osmotic diuresis, which could explain the inhibition of salt and water resorption in the loop of Henle. Conversely, high doses of mannitol may cause renal artery vasospasm and increased RVR resulting in decreased renal perfusion, which may paradoxically impair renal function.

Glomerular effects
There are conflicting accounts of the glomerular effects of mannitol on GFR. Results vary according to the animal model being studied and whether single nephrons or whole kidneys are being assessed. In many in vitro studies, mannitol has been shown to restore GFR of a hypoperfused kidney to near normal values. Mechanisms proposed include preferential dilatation of the afferent glomerular arteriole due to suppression of renin release and intrarenal formation of angiotensin II. Unfortunately, these protective effects of mannitol on the GFR seen experimentally are not reflected in the results seen in intact animals or normal volunteers where, under circumstances of low renal perfusion, for example, glomerular filtration remains unchanged or even decreases with mannitol.

Tubular effects
Renal ischaemia of whatever origin leads to tubular cell oedema. This compresses the interstitial and vascular spaces, resulting in poor renal perfusion and impaired renal function. Mannitol has been shown to reduce tubular cell swelling, particularly in the proximal tubule and thick ascending limb of the loop of Henle and it is assumed that this helps to maintain both tubular flow and glomerular filtration. In addition, mannitol has been found to increase intratubular pressure by increasing the flow of urine which may also help maintain patency of the tubulular lumen.

The diuresis seen with mannitol leads to reduced tubular resorption of water and solutes. As resorption is an energy requiring process, this reduction may be protective in situations where impaired renal perfusion leads to poor oxygen delivery. The precise cause of this effect is unknown although it is probably unrelated to the hyperosmolality of the solution, given that a glucose infusion of similar osmolarity does not produce the same protective effect as mannitol.

MANNITOL AND EFFECTS ON RENAL FUNCTION

Although there are good theoretical reasons for mannitol to have a role in the prophylaxis and salvaging of renal function, they are not convincingly supported by the available evidence. There is a wide gap between the expectation of mannitol's value and its observed clinical benefit, although rigorous studies on its efficacy in the management of renal failure are lacking. Such studies are often difficult to conduct in view of the wide spectrum of aetiologies, clinical settings and severity of ARF. In addition, the criteria for successful treatment differ between studies, and mortality is often due to other causes.

Nevertheless, one area where mannitol has been found to be invaluable is in preserving renal function and reducing the incidence of ARF after cadaveric renal transplantation. A number of well controlled prospective studies have shown that mannitol in the perfusate of a cadaveric kidney prior to transplantation substantially reduced the incidence of both graft rejection and the incidence of post-transplant ART. The same effect can be seen if mannitol is given just prior to restoration of blood flow to the transplanted kidney.

MANNITOL FOR THE PROPHYLAXIS OF ARF

Other than in renal transplantation, prophylactic mannitol does not appear to have any significant protective effect on renal function when given prophylactically. For example, pre-dosing with mannitol prior to the use of radiocontrast agents in interventional radiology has shown no protective effect on renal function. Despite initial hopes, the results have also been equally disappointing in preserving renal function after surgical relief of obstructive jaundice. Patients who have infrarenal cross clamping during abdominal aortic aneurysm surgery are particularly at risk of ARF, but here too, most studies have found mannitol to be of little value with the exception of one group who found a reduction in subclinical renal injury as measured by albumin and N-acetylglucosamindase, more sensitive markers of renal dysfunction.

Mannitol has also been used to preserve renal function in rhabdomyolysis, which results in the release of nephrotoxic myoglobin form severe muscle necrosis following a crush injury. The renoprotective action of mannitol is thought to be mediated through the diuretic action which has the dual effect of increasing both urine flow and intratubular pressure to overcome tubular obstruction from haem pigment trapping and cast formation. These effects, however, are not exclusive to mannitol and, in addition to there being little evidence that mannitol improves outcome, it has also been shown that expanding plasma volume with other agents, such as glycerol and normal saline, has similar renoprotective actions. As such, nowadays the first line of treatment for prevention of myoglobinuric ARF is volume expansion with saline.

MANNITOL FOR THE MANAGEMENT OF ACUTE RENAL FAILURE

In established acute renal failure, mannitol is not only ineffective but can be hazardous due to the risk of fluid overload. Its role in early renal failure is less clear as it can be difficult to determine whether patients have pre-renal failure or have crossed the line into established failure given that biochemical measurements are too insensitive to be helpful in the early stages. Some clinicians advocate mannitol therapy within 24 h of onset of oliguria. However, conversion of oliguria to an non-oliguric state is no guarantee of prevention of ARF, and the risk of progression remains high regardless of whether mannitol has been given or not.

There are concerns that mannitol may acutally increases the risk of ARF by increasing renal artery oxygen consumption. Despite receiving approximately 20% of cardiac output, the kidney is very susceptible to ischaemia. There are two main reasons for this. Firstly, the kidney extracts relatively little oxygen for metabolic processes and, secondly, the medullary part of the kidney receives only 6% of renal blood flow despite being responsible for most of the energy requiring work of the kidney. Oxygen delivery to the medulla is thus critical. If it is reduced any means, renal function will be compromised and this may occur even when total renal blood flow appears to be adequate.

Mannitol increased both renal blood flow and glomerular filtration and thus there is a greater solute and water load to be resorbed by the tubules. This leads to energy expenditure and a concomitant rise in oxygen demand, so that mannitol may do more harm than good in situations where there is impaired oxygen delivery. This increase in solute load, however, needs to be balanced against mannitol's ability to inhibit resorptive processes in the loop of Henle by loss of medullary hypertonicity, and thus, presumably, decrease the potential for increased oxygen consumption. The relative importance of these two opposing forces is unknown.

SUMMARY OF THE ROLE OF MANNITOL IN RENAL PROTECTION

Mannitol's theoretical promise of effective protection against ischaemic injury of the kidney proves in practice as illusory as its role in neuroportection after acute brain injury. Clinical trials fail to adequately demonstrate convincing advantages for mannitol except in the case of cadaveric renal transplantation.

MANNITOL AS AN OXYGEN FREE RADICAL SCAVENGER

Oxygen free radicals are extremely toxic molecules constantly produced in all metabolically active cells during certain stages of many different types of biological processes. One common example is the formation of OFRs witht eh electron reduction of oxygen by cytochrome oxidase. Under normal circumstances, OFRs are inactivated immediately by intracellular enzymes such as catalase, superoxide dismutase and glutathione preoxidase; but, in certain situations, such as tissue ischaemia, they are produced in excess, which may overwhelm the normal limiting processes. This can lead to oxidation of cellular components such as mitochondrial and cellular membranes resulting in cell damage or death. The most reactive OFR is the hydroxyl radical but others include nitric oxide and superoxide radicals.

Mannitol has been proposed to act as an antioxidant OFR scavenger, specifically targeting the OH raidcal to cause its disproportionationor dimerization to non-toxic metabolites. Much of the research on mannitol's OFR scavenging capabilities has looked at reperfusion injury following the restoration of oxygen supply to an ischaemic organ. Models of this type are used because direct, consistent measurement of OFRs are not possible in view of their very short half-life of a few nanoseconds. Cardiac surgery provides one of the easiest models for study in this field. In spite of cardioplegia solutions and cooling for myocardial protection, there is often some postopeartive depression of myocardial function, which is thought largely to be due to an excess of OFRs produced after cardiopulmonary bypass. Cardioplegia solutions also provide a relatively simple and convenient method for introducing intervention therapies, such as mannitol.

One such study, for example, has demonstrated decreased OFR production following CPB. Others, both animal and human, have shown a slight decrease in the incidence and duration of artrial and ventricular arrhythmia as well as, evidence of histological preservation of the myocardial cell architecture and improvement in ventricular and renal function post CPB. However, given the problem to detecting and measuring OFR production accurately, it is difficult to know whether the results of these studies are directly related to mannitol's OFR scavenging capabilities or not and, as such, the results frequently remain open to interpretation. Moreover, these positive studies are increasingly being matched by others suggesting mannitol has little effect on reperfusion injury. For example, mannitol has equally been shown to have any effect on the incidence and severity of reperfusion atrial or ventricular arrhythmias, nor has it been shown to improve ventricular function, renal function or haemodynamic stability after CPB.

There are several valid explantations for mannitol's ineffectiveness in protecting against reperfusion injury. OFRs are produced intracellularly and need to be mopped up instantaneously in order to prevent molecular damage. Mannitol, on the other hand, tends to remain in the extracellular compartment, with minimal amounts crossing the cell membrane. It is thus unlikely that mannitol plays a significantly intracellular role in preventing any significant oxidative damage. Moreover, although mannitol reacts rapidly with hydroxyl radicals it fails to scavenge any other OFR to an extent that would offer protection in a biological system.

Although the evidence for any direct OFR scavenging effect of mannitol is poor, it may act indirectly to reduce OFR production. It is an effective chelating agent of ferric ions, which play a key role in converting poorly reactive oxygen species into highly reactive and damage OH-radicals. It may thus protect against such damage by removing iron from the cellular molecules and structures at risk of oxidative damage. This has been seen both in vitro and in vivo and it has been suggested that it may, under some conditions, be more important than direct scavenging. Mannitol also forms a stable compound with hydrogen peroxide, which is not a free radical itself but can generate OH radicals in the presence of ferric iron. This effect has been investigated in patients undergoing CPB, where mannitol appears to react directly with H2O2 to decrease its formation, although it has yet to be determined how this significantly improves clinical outcome.

Another potentially advantageous and indirect action of mannitol may be through a modifying action on complement formation. Complement fractions C3 and C5a are released during CPB and bind to plymorphonuclear leukocytes which are then stimulated to produce OFRs. Sequestered PMNL and complement fragments have been found in the lung after CPB and have been proposed to be responsible for post bypass lung changes such as pulmonary oedema. One beneficial effect of mannitol in CPB appears to be a reduction in the incidence to postoperative pulmonary oedema, and this may reflect an OFR scavenging reaction, although its diuretic action cannot be excluded.

One final area where mannitol has been successfully used is in the management of acute compartment syndrome. It is uncertain, however, whether the reduction in the perfusion oedema and injury are due to its hyperosmolar or OFR scavenging effects or both. It appears that the former is the more important, but that OFR scavenging may contribute to a reduce severity of muscle necrosis. Once again, the clinical evidence is inconclusive.

SUMMARY OF THE ROLE OF MANNITOL AS AN OXYGEN FREE RADICAL SCAVENGER

On balance, inconclusiveness is a recurring theme in any discussion of mannitol's role as an OFR scavenger. Nevertheless, although the evidence for any significant effect is scanty, it cannot be totally discounted are more conclusive research is required.

CONCLUSIONS

Although mannitol is still widely used in many areas of medicine and its worth in renal transplantation is unsurpassed, the evidence for its continued use for the management of head injuries and renal failure is limited, while its value as an OFR scavenging agent has yet to be finally decided.

References :

1. Weed L H, McKibben P S. Pressure chagnes in the cerebrospinal fluid following intravenous injection of solutions of various concentrations. Am J Physiol 1919;48:512 - 530.

2. Paczynski R P. Osmotherapy, basic concepts and controversies. Crit Care Clin 1997;13:111 - 129.

3. Cascino T, Baglivo J, Conti J et al. Quantitative assessment of furosemide and mannitol induced changes in brain water content. 1983;33:898 - 903.

4. Pollay M, Fullenwider C, Roberts P A et al. Effect of mannitol and furosemide on blood brain osmotic gradient and intracranial pressure. J Neurosurg 1983;59:945 - 950.

5. Burke A M, Quest D O, Chien S et al. The effects of mannitol on blood viscosity. J Neurosurg 1981;55:550 - 553.

6. Roberts J, Schierhout, Alderson P. Absence of evidence for the effectiveness of five interventions routinely used in the intensive care management of severe head injury. J Neurol Neurosurg Psychiatry 1998;65:729 - 733.

7. Anon. The use of mannitol in severe head injury. J Neurotrauma 1996;13:705 - 709.

8. Schwartz M L, Tator C H, Rowed D W. A prospective randomised comparison of pentobarbitol and mannitol. Can J Neurol Sci 1984;11:434 - 440.

9. Le Roux P D, Newell D W, Lam A M et al. Cerebral arteriovenous oxygen difference: a predictor of cerebral infarction and outcome in patients with severe head injury. J Neurosurg 1997;87: 1 - 8.

10. Smith H P, Kelley D L, McWorter J M et al. Comparison of mannitol regimens in patients with severe head injury undergoing intracranial monitoring. J Neurosurg 1986 65: 820 - 824.

11. Sayre M R, Daily S W, Sterns S A et al. Out of hospital administration of amnnitol to head injured patients. Acad Emerg Med 1996;3:840 - 847.

12. Cote C J, Greenhow D E, Marshall B E. The hypotensive response to rapid intravenous administration of hypertonic solutions in man and in the rabit. Anesthesiology 1979;50:30 - 35.

13. Gadallah M F, Lynn M, Work J. Case report: mannitol nephrotoxicity syndrome role of haemodialysis and postulate of mechanisms. Am J Med Sci 1995;309:212 - 222.

14. Biestro A, Alberti R, Galli R et al. Osmotherapy for increased intracranial pressure: comparison between mannitol and glycerol. Acta Neurochir 1997;139:725 - 733.

15. Freshmann S P, Battistella F D, Matteucci M et al. Hypertonic saline versus mannitol a comparison for treatment of acute head injuries. J Trauma 1993;35:344 - 348.

16. Suarez I, Qureshi AI, Bhardwaj A et al. Treatment of refractory intracranial hypertension with 23.4% saline. Crit Care Med 1998;26:1118 - 1122.

17. Prough D S. Mannitol: an old friend on the skids Crit Care Med 1998;26:997 - 998.

18. Selkurt E E. Changes in renal clearance following complete ischaemia of kidney. Am J Physiol 1945;144:231 - 250.

19. Better O S, Rubinstein I, Winaver J M et al. Mannitol therapy revisited. Kidney Int 1997;51:886 - 894.

20. Flores J, Dibona D R, Beck C H et al. The role of cell swelling in ischaemic renal disease and the protective effect of hypertonic solute. J Clin Invest 1972;51:118 - 126.

21. Conger J D. Interventions in clinical acute renal failure. What are the data? Am J Kidney Dis 199X;26:565 - 576.

22. Shilliday I, Allison M E. Diuretics in acute renal failure. Ren Fail 1994;16:3 - 17.

23. Grino J M, Miravitiles R, Castelao A M. Flush solution with mannitol in the prevention of post transplant renal failure. Transplant Proc 1987;19:4140 - 4142.

24. Stevens M A, McCullough P A, Tobin K J et al. A prospective randomised trial of prevention measures in patients at high risk of contrast nephropathy. J Am Coll Cardiol 1999, 32:403 - 411.

25. Dawson J L, Stirling G-A. Protective effect of mannitol on anoxic jaundiced kidneys. A histological study. Circulation 1963;27:254 - 259.

26. Kretzschmar M, Klein U, Palutke M et al. Reduction of ischaemia-reperfusion syndrome after aortic aneursymectomy by N-acetylcysteine but not mannitol. Acta Anaesthesiol Scand 1996;40:657 - 664.

27. Nicholson M L, Baker D M, Hopkinson B R et al. Randomized controlled trial of the effect of mannitol on renal reperfusion injury during aortic aneurysm surgery. J Surg 1996;83:123003.

28. Homsi E, Barreiro M F F L, Oriando J M C etal. Prophylaxis of acute renal failure in patients with rhabdomyolysis. Ren Fail 1997;19:283 - 288.

29. Gelman S. Does mannitol save the kidney Anesth Analg 1996;82:899 - 901

30. Macgovern JrG J, Bolling SF, Casale A S et al. The mechanisms of mannitol in reducing ischaemic injury: hyperosmolality or hydroxyl scavenger. Circulation 1984;70:191 - 195

31. Hearse D J, Humphrey S M, Chain E B. Abrupt reoxygenatoin of the anoxic pottasium arrested rat heart: a study of myocardial enzyme release. J Mol Cell Cardiol 1973;5:395 - 407.

32. Bernier M, Hearse D J. Re-perfusion induced arrhythmias and oxygen-derived free radicals. Am J Physiol 1988;254:H862 - H870.

33. Ouriel K, Ginsberg M E, Patti C S et al. Preservation of myocardial function with mannitol reperfusate. Circulation 1985;72:254 - 258.

34. Fisher A P, Jones P, Barlow P et al. The influence of mannitol on renal function during and after open-heart surgery. Perfusion 1998;13:181 - 186.

35. Woodward B, Zakaria M N M. Effect of some free radical scavengers on reperfusion induced arrhythmias in the isolated rat heart. J Cell Mol Cardiol 1985;17:484 - 493.

36. Menasche C, Grousset C, Gaudual C et al. Prevention of hydroxyl radical formation: a critical concept for improving cardioplegia: protective effect of desferrioxamine. Circulation 1987;XX: 108 - 118.

37. Gutteridge J M C. Ferrous salt promoted damage to deoxyribose and benzoate. The increased effectiveness of hydroxyl radical scavengers in the presence of EDTA. Biochem J 1987; 243 709 - 714.

38. Oredsson S, Plate G, Qvarfordt P. The effect of mannitol on reperfusion injury and postischaemic compartment pressure in skeletal muscle. Eur J Vasc Surg 1994;8:326 - 331

39. Gillibe C E, Sage F J, Gutteridege J M C. Mannitol: molecule magnifique or a case of radical misinterpretation. Free Radic Res 1996; 24: 1 - 7.