sign in sign up
<<
Home , Brain abscess, Ventriculitis

7oumnal ofNeurology, , and Psychiatry 1993;56:1243-1258 1 243 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from NEUROLOGICAL EMERGENCY

Management of cerebral

Milne Anderson

Acute is the most common of the echo, and less commonly by herpes simplex cerebral infection syndromes. Epidemiology, type 1, varicella zoster, mumps, lymphocytic pathogenesis, and evolution are complex, and choriomeningitis, and HIV. vary with geography and with age. The clinical onset is usually rapid over Presentation occurs within hours, or at most hours, with pyrexia, malaise, , neck a few days, of the onset of symptoms which, stiffness, photophobia, lethargy, myalgia and with the signs, are characteristic: headache, irritability. Most cases do not progress fur- , photophobia, irritability, neck stiffness, ther, and the subject can be roused easily and and changed mental state. The causal agent is remains coherent. If the conscious level either a or a bacterium. Fungi and para- reduces, or focal signs or occur, sites cause acute meningitis only in excep- is implied. Resolution begins tional circumstances. Viral meningitis is a within a few days and is complete within two much more benign illness than the bacterial weeks in most. A few will have persistent form. malaise and myalgia for some weeks. The is seldom identified clinically- parotitis and orchitis may point to mumps, Viral meningitis arthralgia and to HIV, The term "acute aseptic meningitis" was myalgia and myocarditis to coxsackie, and coined by Wallgren in 19251 to describe acute rashes to enteroviruses. The illness is not as meningeal irritation, benign and self-limiting, severe and prolonged as bacterial meningitis, with complete recovery and sterile pleocytic and the signs are not so marked. CSF. It has since become evident that at least 70% of such cases are caused by . DIFFERENTIAL DIAGNOSIS CNS viral are complications of sys- Conditions from which viral meningitis must temic viral infections, and the virus gains be differentiated are the early stages of bacte- access to the brain by the bloodstream or, less rial meningitis, some cases of subarachnoid

commonly, by travelling up peripheral haemorrhage and other causes of aseptic http://jnnp.bmj.com/ nerves.2 Viral meningitis is the result of meningitis, partially treated bacterial menin- haematogenous infection and, to enter the gitis, meningitis caused by fastidious bacteria, CNS, the virus must cross the endothelial cell fungi and parasites which do not readily grow junctions of the blood-brain barrier, the abil- in routine culture, parameningeal infection, ity to do this being dependent upon surface or neoplasia, and collagen dis- adhesion molecules on the cells, surface ease. To confirm the diagnosis, CSF exami- charges and cellular receptors of the virus, nation is mandatory, but not before and the property of entering infected cells.' dangerously raised on October 2, 2021 by guest. Protected copyright. Certain viruses preferentially infect the (ICP) or a space-occupying lesion has been , choroid plexus, and ependyma excluded by appropriate imaging. CSF pres- rather than cerebral parenchyma, causing sure is normal or slightly raised, and the fluid meningitis; others infect neurons and glia to is clear to the naked eye. Cell counts are in cause encephalitis. There is, however, consid- the range of up to 500-1000 white cells, erable overlap, and some viruses may cause mainly lymphocytes, although in some, poly- , incorporating signs of morphs may predominate. In such cases it is both. prudent to re-examine CSF 12-24 hours later Most cases of viral meningitis occur in chil- to identify a lymphocytosis and exclude a dren and young adults worldwide. Infections bacterial cause.7 CSF protein may be slightly occur throughout the year, with a preponder- raised, glucose is normal or only a little ance in summer and autumn in temperate cli- reduced. Numerous laboratory tests have Midland Centre for mates. been applied to CSF with the claim that they Neurosurgery and The annual reported incidence varies , Holly from 1 1 to 27 cases per 100 000,4 5 and over differentiate a bacterial from a viral aetiology, Lane, Smethwick, 7000 cases are reported annually in the but none is sufficiently discriminating to be Warley, West United States6: the actual number of infec- useful. These include lactate, lysozyme, C- Midlands B67 7JX Milne Anderson tions is almost certainly several multiples reactive protein, and creatine kinase estima- Correspondence to: higher, because of under-reporting. Over half tions. A similar lack of specificity applies to Dr Anderson are caused by enteroviruses-coxsackie-B or the occasional abnormalities that may be seen 1244 Anderson

in blood counts, blood biochemistry, and the mumps? Has there been travel to or from an J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from EEG. area that harbours known vectors? Is there In most cases it is not necessary to estab- past or present evidence of an animal or lish an exact aetiology for treatment purposes, insect bite? What is the season of the year? as the disease is benign and self-limiting and, Diagnostic tests may help, yet in perhaps as provided other similar treatable diseases have many as a third of all cases, no specific aetiol- been excluded, symptomatic treatment with ogy can be established. analgesics and anti-emetics is all that is is not rare, and it occurs required. should not be given. globally. In the UK and Europe, most cases To establish the aetiology, the virus may be are sporadic accompaniments of common isolated from CSF or by serological studies of infections such as mumps, measles, and her- acute and convalescent serum samples, iden- pes simplex. In the United States, sporadic tification of IgM antibody or viral antigen in and epidemic forms are caused by the CSF. The application of recently developed arboviruses (arthropod-borne), and Japanese immunological and DNA probe amplification B encephalitis causes most epidemic infec- techniques, including polymerase chain reac- tions elsewhere. As many as 20 000 cases per tion (PCR), to detect viral antigen, is promis- annum may occur in the United States." ing.8 In the UK, their use at present is Patients who develop viral encephalitis dictated more by local availability than by often have a prodrome lasting several days, clinical need. which may include myalgia, fever, malaise, mild upper respiratory infection, rash, or parotitis. The development of headache, Viral encephalitis mental change, and drowsiness implies Acute viral encephalitis is due to direct inva- encephalitis, which is usually associated with sion of brain parenchyma, and the clinical meningitic features. As the disease progresses, manifestations are caused by cell dysfunction disorientation and disturbance of behaviour from this invasion and associated inflam- and speech worsen, and drowsiness becomes matory change. At the bedside this may . Epileptic seizures are common, and be indistinguishable from post-infectious focal signs may appear appropriate to the area encephalitis, the of which is perive- of the brain taking the brunt of the infec- nous demyelination thought to be caused by tion-hallucinations and memory upset from allergic or immune reactions triggered after a the temporal lobes, , spasticity, latent period by viral infection.9 10 Viruses are sensory loss and speech upset, and cerebellar far and away the commonest cause of deficits. There may also be signs of raised encephalitis globally, but in certain locations intracranial pressure. The very young, the and seasons other organisms such as malaria very old, and those with compromised and other protozoans, rickettsiae and fungi immune systems often have more severe dis- may induce an encephalitic syndrome. ease. Signs should be sought in other organ As with meningitis, so with encephalitis, systems, which may point to a particular the virus reaches the brain by the blood- virus. Some forms of encephalitis have spe- stream. Entry may be through the skin fol- cific features which are briefly discussed later. lowing an insect bite, as with arbovirus infection, or via the respiratory or gastroin- testinal route. Local replication ensues, the DIFFERENTIAL DIAGNOSIS inoculum spills into blood and so to the retic- The list of other diseases that may cause a http://jnnp.bmj.com/ uloendothelial system whence, following fur- similar clinical picture is large, and includes ther replication, viraemia increases, and all forms of bacterial meningitis, malaria and spread takes place to other sites including the other protozoal and fungal infestations, CNS. Modification of this process by host intracranial suppuration, septicaemia and immune responses may occur, and if these endocarditis, metastatic disease and col- are compromised, disease progression may be lagenoses, and drug abuse.

fulminant. Most cases of encephalitis are Faced with a patient with rapid onset of on October 2, 2021 by guest. Protected copyright. caused in this way. The virus may also ascend pyrexia and stupor or coma, the above poten- neurons centripetally to lodge in brain cells, tially remediable conditions need to be as with herpes simplex encephalitis and excluded quickly. Blood counts and biochem- rabies. Fortunately, encephalitis is a rare ical tests are not diagnostic but may be complication of common viral infections, and abnormal as a result of inadequate hydration most patients with systemic viral infections do or inappropriate secretion of antidiuretic hor- not develop neurological signs. mone. Blood films should be examined for Certain viruses exhibit tropism towards malaria parasites, and blood cultures set up. specific cell types-the limbic system in As these results are awaited, the intracra- rabies, temporal lobes in herpes simplex nial contents should be imaged to determine encephalitis-and this may produce clinical if there is space-occupation, cerebral oedema signs that are diagnostically useful. In most or focal areas of infarction such as may occur cases of encephalitis, however, signs and in the temporal lobes with herpes simplex symptoms are common to most , encephalitis. Scans must be interpreted with and diagnostic clues must be sought else- close attention to their timing in the evolution where: is there evidence of infection else- of the disease: in the first two or three days where, such as the characteristic rash of no abnormality may be evident, and changes varicella or measles, or the parotitis of may not show for five or six days, so repeat Management ofcerebral infection 1245

scanning may be necessary. The EEG is over 50.'4 The clinical presentation is enor- J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from abnormal, most often showing non-specific mously variable. Onset is usually insidious, diffuse slow-wave activity, perhaps with with a prodrome of 4-10 days, with malaise, activity. Temporal lobe focal abnor- pyrexia, and irritability, followed by frontal mality with high-voltage spike and slow-wave and temporal lobe disturbance of personality complexes is highly suggestive of herpes sim- change, hallucinations, psychiatric upset, and plex encephalitis. CSF should be examined as increasing focal signs, seizures, and deterio- soon as possible when deemed safe; unfortu- rating conscious level. In perhaps as many as nately cranial scanning does not convey an 87%, focal signs appear."5 In a few, the onset accurate picture of intracranial pressure, and is cataclysmic, and the evolution is com- judgement of the potential safety of lumbar pressed into only a few days. , puncture requires not a little luck. CSF is tumour, , vascular disease, and often under pressure and the cell count is other forms of viral encephalitis are the other raised, from 10 to as many as several thou- diagnoses that have been mistaken for herpes sand white cells per m3', usually lympho- simplex encephalitis.'6 Focal changes on CT, cytes, but polymorphs may predominate in MRI, and EEG, as already described, are the early stages. Red cells may be found if helpful but not pathognomonic. there is a necrotising component, as in herpes Treatment must be given quickly with acy- simplex encephalitis. The glucose content is clovir, 10 mg/kg by i.v. infusion over one normal, protein is raised, and organisms are hour, every eight hours for 10 days. This has not seen. been shown to diminish mortality to 30%.'7 Most people use dexamethasone despite TREATMENT objections that interferon synthesis may be In the recent past, treatment of viral inhibited. encephalitis was symptomatic, and no specific drug would influence the outcome. This has VARICELLA ZOSTER changed with the advent of acyclovir and Encephalitis accounts for 90% of the neuro- other forms of antiviral chemotherapy, so it is logical complications of varicella, which are in now imperative to identify the organism as themselves rare, affecting 0-1% of cases.'8 In soon as possible, or at least to determine half, the encephalitis is of cerebellar type, whether it is herpes simplex. Early reports of with ataxia, dysarthria, headache, and drowsi- the application of PCR and other techniques ness coming on about a week after the rash to the rapid detection of viral antigen in CSF begins,9 but the neurological onset may pre- are encouraging. 1213 In my view, brain biopsy cede the rash. Convulsions are common,19 no longer has a place in the diagnosis of viral and progression to hemiplegia, cranial nerve encephalitis, except to determine if a lesion palsies, , and coma may ensue. demonstrated by scan is an abscess, granu- Patients with the cerebellar form usually loma, or tumour. The complication rate is recover completely, but 10% of those with over 3%, and there is no guarantee that the the general form die.9 Encephalitis may rarely area biopsied will contain virus. It is safer to follow shingles. Management is symptomatic, begin treatment with acyclovir if clinical sus- as already described. picion is high. With the exception of that caused by her- Cytomegalovirus pes simplex, the treatment of viral encephali- This is not an important cause of encephalitis tis is symptomatic, and requires the except in neonates and in immunocompro- http://jnnp.bmj.com/ maintenance of adequate nutrition, hydra- mised patients, recipients of organ trans- tion, and oxygenation. Seizures are controlled plants, and patients with AIDS. Treatment with anticonvulsants, and secondary infec- with antiviral agents has, to date, been nuga- tions are treated as necessary. Intracranial tory. hypertension and cerebral oedema may be a problem, and there is no consensus on the Epstein-Barr virus correct treatment. Intubation and hyperventi- Meningoencephalitis has been described as a on October 2, 2021 by guest. Protected copyright. lation, glycerol, mannitol, and dexametha- rare complication of Epstein-Barr virus infec- sone have all been used, and I continue to use tion.20 There are no specific features, and the dexamethasone in severely ill patients, despite prognosis is excellent. theoretical objections. Measles Acute encephalitis occurring in the course of SPECIFIC ENCEPHALITIDES measles infection is usually caused by a Herpes simplex postinfectious, perivenous, demyelinating, This is the most common cause of sporadic allergic phenomenon, although in some there encephalitis in Europe and North America, may be direct virus-induced cellular damage, with an incidence of up to 0-5 per 100 000 and virus has been isolated.2' Clinically the population each year-almost certainly an manifestations are the same, and occur in up underestimate because milder cases go to one in 1000 cases above the age of two, the unrecognised. It occurs throughout the year, frequency increasing with age. Typically, as and almost all cases are caused by HSV-1. It the exanthem begins to fade after seven or is no more common in those with compro- eight days, fever recurs and encephalitis mised immune systems, and it affects patients develops rapidly, with convulsions, focal of any age. A third are below 20, and half signs, , and coma.22 There are no 1246 Anderson

specific features on investigation, and treat- is an Aedes mosquito. Children are affected, J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from ment is symptomatic. Mortality is high at with a peak in August and September. 15%, and most are left with some disability.2' Clinical onset is sudden, and recovery takes place within 10 days. Mumps encephalitis This complication is rare, occurring in less Central European encephalitis than 1 % of cases, and has no specific fea- This encephalitis may be contracted from tures; it follows up to two weeks after devel- woodlands in Scandinavia through to north- opment of parotitis, though this may be ern Greece and the former Yugoslavia, and absent in up to half.2425 Treatment is sympto- more cases are being reported from Austria. matic and prognosis is excellent. The virus is maintained in small woodland mammals and transmitted to people who EPIDEMIC ENCEPHALITIS enter the forest for recreation or work and are Arboviruses cause epidemic encephalitis in bitten by an Ixodes tick. There is a biphasic various parts of the world. They are not an course with an initial flu-like illness, followed important cause of encephalitis in Britain, but within two weeks by mild meningitis or with the increasing ease and frequency of encephalitis, perhaps with muscle weakness. intercontinental travel, cases are now being Management is symptomatic and recovery is recognised, and the likelihood is that the fre- usually complete. Diagnosis is serological. quency will increase. It is imperative therefore that a full history of travel abroad be taken Russian spring-summer encephalitis from anyone suspected of neurological infec- Russian spring-summer encephalitis is a tion, and arbovirus encephalitis should be similar, more severe, condition. suspected if there has been exposure to insect bites in countries known to harbour these J7apanese encephalitis viruses. This is the most common arbovirus infection All of these arboviruses are perpetuated by worldwide, and is endemic in southeast Asia, zoonoses of insignificant infection in birds Japan, China, Philippines, Borneo, and large and smaller vertebrates. Transmission is by areas of the Indian subcontinent. Birds and an arthropod vector, mosquito or tick. domestic animals are the reservoir, mosquitos Replication and viraemia develop after being the vector. Encephalitis is seen most com- bitten. Most people suffer only a mild sys- monly in those under 15 and in the aged. temic illness, but some go on to develop After an incubation period of one to two encephalitis, the severity of which depends weeks there is abrupt onset of encephalitis, upon the strain of the virus. Clinical features often with myalgia. Extrapyramidal features are common to all, with some variation in have been noted.'728 Convalescence may be incubation, progression, and severity, and prolonged and sequelae are common. they cannot be diagnosed on clinical grounds Treatment is symptomatic and diagnosis is alone. There are no specific features to be serological.'9 found on imaging, on the EEG or in the CSF. Diagnosis is by demonstration of anti- body rise in paired serum samples. Treatment Rabies is symptomatic. In Britain, recognition of a case of clinical rabies is exceptionally rare and, numerically, Eastern equine encephalitis barely merits inclusion in a description of http://jnnp.bmj.com/ Birds of the Atlantic and Gulf coasts of neurological emergencies. However, rabies is America are the reservoir of this rare condi- endemic throughout the world except for the tion. Epidemics may occur in horses. Cases UK, Ireland, Antarctica, Australasia, Japan, occur at migration time, summer and and islands of the Caribbean, Mediterranean, autumn, and children are most affected, with and Scandinavia. Most of the cases now seen an acute encephalitis which may be abrupt in the United States have been contracted

and virulent, with a mortality rate of 70%.26 abroad and, where countries have introduced on October 2, 2021 by guest. Protected copyright. programmes to control domestic rabies in Western equine encephalitis cats and dogs, the incidence in humans has This is less severe, occurring in eastern, cen- dropped.30 tral and western United States, Canada, and eastern South America. DIAGNOSIS Diagnosis is straightforward if an appropriate St Louis encephalitis history can be obtained, but where that is This disease shares the same clinical features lacking, the onset of hydrophobia provides in the same areas and in central America. the clue. Rabies should be considered in the West of the Mississippi it is endemic, and differential diagnosis of all encephalitides cases occur in epidemics or sporadically from contracted abroad, and it is important to July to October. Most cases are benign and of remember that perhaps 15% of all cases pre- short duration. sent as a Guillain-Barre type of syndrome. Cranial imaging and EEG do not show spe- California and La Crosse viruses cific features, and CSF may show a pleocyto- These viruses are similar and cause the same sis and raised protein. Rapid diagnosis can be disease in central and midwest United States. reached by the demonstration of rabies anti- The reservoir is small rodents and the vector gen in neck skin-biopsy,3" which is said to be Management ofcerebral infection 1247

more accurate than corneal smears. Virus fungi appropriate to geography, J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from may be isolated from saliva and other fluids. and syphilis are all frequent pathogens. Progressive multifocal leukoencephalopathy TREATMENT may also complicate AIDS and, unfortu- Treatment does not alter the inevitably fatal nately, more than one syndrome may be pre- outcome but does relieve distress to patients sent at any given time, so the clinical picture and relatives. Sedation may be necessary in can be confused. the "furious" stage, anticonvulsants may pre- All patients with AIDS should be checked vent seizures but it may be necessary to para- for syphilis with a specific antitreponemal lyse and ventilate the patient to prevent the test. A negative tuberculin skin test is likely in severe generalised spasms of advanced these patients even if they have tuberculosis.'7 hydrophobia. There is no evidence that If a focal lesion is seen on CT or MRI, it may steroids, immune suppression, antibiotics, or be toxoplasmosis,38 39 tuberculoma, lym- antiviral agents help. All secretions from the phoma, or . Lymphoma usually patient should be regarded as potentially haz- shows solid, multicentric lesions, but can ardous, and any staff thought to have been show ring enhancement. Single or multiple inoculated by them should have a course of ring-enhancing, hyperdense or hypodense immunisation. lesions may be caused by any of these : no appearance is specific. Because 10-40% of patients with AIDS will HIV develop toxoplasmosis during the course of Cerebral infection is a frequent accompani- their illness, and serological tests are not suf- ment of HIV infection. The virus itself affects ficiently specific to confirm that toxoplasmo- the brain, and the systemic immune suppres- sis is the cause of the lesion, biopsy must be sion which accompanies AIDS renders the considered. In such circumstances accepted sufferer liable to superadded and opportunis- practice (which I adopt) is to treat for toxo- tic infection which frequently attacks the plasmosis with pyrimethamine and sulphadi- brain. The manifestations of HIV infection of azine, and with dexamethasone to reduce the brain coexist with opportunistic infection, brain swelling. If the lesion is due to toxoplas- and may result in a complex and confused mosis, improvement occurs within 10-14 syndrome constellation. Many of the features days and can be monitored by serial scan- are chronic in evolution and some are acute. ning.40 In practice, any change in neurological status If improvement does not occur, if there is in a patient with AIDS usually presents as an clinical deterioration, or if diagnostic doubt emergency because of the danger of oppor- remains, biopsy should take place, and treat- tunistic infection. Conversely, the finding of ment given as appropriate. Tuberculosis is an unusual cerebral infection in a patient not an increasing problem due both to known to be immune suppressed should raise Mycobacterium tuberculosis and to M avium the suspicion that he or she may have HIV complex disease, and the author's threshold infection. Many excellent accounts of the epi- for deciding to perform a biopsy is diminish- demiology and classification of the natural ing, in order to make this diagnosis early: progression of HIV infection are available, treatment of M avium complex disease and are not discussed further here."'-3 responds better to regimens that incorporate

clofazimine and amikacin.41 At present there http://jnnp.bmj.com/ is no satisfactory regimen for the treatment of AIDS COMPLEX established HIV infection, but appropriate AIDS dementia complex occurs during estab- and prompt treatment of these infective com- lished AIDS and, with more subtle neuropsy- plications can lead to prolonged remission. chological and neurophysiological techniques, it is possible to demonstrate mild cognitive defects in early infection.'5 It is now recog- Bacterial meningitis nised that 90% of AIDS patients will develop Any bacterium is capable of causing menin- on October 2, 2021 by guest. Protected copyright. this complex before death. The usual course gitis, given favourable circumstances. is for the subject to become withdrawn, apa- Meningitis may come about as part of a gen- thetic, and mentally slow, and progress inex- eralised illness or as a complication of disease orably to become increasingly demented.'6 elsewhere. In practice, most cases are caused The course may, however, be interrupted by by organisms with properties of virulence acute relapses with and psychosis, which allow access to the CNS. With the when the patient may present as a neurologi- exception of the neonatal period, Haemophilus cal emergency, and it is necessary to exclude influenzae, Neisseria meningitidis, and other CNS infections, neoplasia, and sys- account for 70% of temic illness as a cause. all cases. Neonatal meningitis may be caused by any organism, and the most frequently encountered pathogens are Gram-negative OPPORTUNISTIC INFECTIONS bacilli, particularly E coli Kl, other enteric Infection and lymphoma can complicate and bacilli, Pseudomonas and group B strepto- alter the progression of AIDS and precipitate cocci. Neonatal meningitis is not discussed a neurological emergency. Toxoplasma, further here. cytomegalovirus, herpes simplex, herpes The epidemiology of meningitis is com- zoster, , histoplasma and other plex, and varies with the age of the patient 1248 Anderson

and geography. About 70% of cases are seen yet ill understood, cerebral blood flow J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from in children. H influenzae and N meningitidis increases, inducing vasogenic cerebral are the main causes in this group; in adults, oedema and raised intracranial pressure. With S pneumoniae and N meningitidis predominate, the progression of inflammation, cytotoxic and there is an increasing number of cases oedema develops, cerebral blood flow falls, caused by Gram-negative bacilli and Listeria, vasculitis progresses and intracranial pressure particularly in the elderly. Streptococci and rises further. Interstitial oedema results from staphylococci also cause sporadic cases-the obstruction of CSF flow from subarachnoid meningococcus remains the only significant space to blood. Cerebrovascular autoregula- cause of epidemics. H influenzae is a more tion is disrupted5' and the brain becomes at common pathogen in the United States than risk from either hyper- or hypoperfusion. in the UK (48% v 29%) where the meningo- Evidence is accumulating that many of these coccus and pneumococcus predominate changes occur early in the course of meningi- (45% v 33%).,4243 Gram-negative bacilli tend tis, and treatment regimens that reduce to occur in association with head injuries, fol- intracranial pressure while maintaining lowing neurosurgery, in patients who have cerebral perfusion and blood flow may be compromised immune systems, and in the useful.50 elderly. One disturbing factor is the increase in the number of cases nosocomially DIAGNOSIS acquired." The streptococcus infects puer- The clinical features of a typical case of peral women, the elderly and those with meningitis are easy to recognise-fever, immunosuppression, as does Listeria. malaise, headache, neck stiffness, irritability, Staphylococcal infection may complicate neu- and confusion. Usually these features appear rosurgical procedures and head trauma, and over the space of a day or two but in some also affect those immunosuppression or endo- fulminant cases the signs may develop within 'carditis. Anaerobic organisms cause less than hours. In as many as 20% of cases these typi- 1% of cases. Infection with more than one cal features may be lacking: this is particularly organism also occurs in 1%, and in similar so with the very young, the very old, the very circumstances-paranasal sinus and ear ill, and those with compromised immune infections, skull fractures, neurosurgery, systems. The possibility of meningitis should and immunosuppression.45" Tuberculosis be entertained in every child or adult in the remains a significant cause of meningitis in above groups who has lethargy, altered the UK and the United States. Many other mental status, drowsiness, or pyrexia, organisms may also cause meningitis, com- especially if neck stiffness is present. It is monly of a more chronic nature, which may important to recognise that neck stiffness may rapidly develop into an acute emergency in not be present, and the physician must patients with compromised immune systems. include meningitis in the differential diag- nosis of a very wide range of diseases. PROGRESS IN DIAGNOSIS AND TREATMENT Epileptic convulsions occur with meningi- Advances in the investigation and manage- tis in any age group, are particularly common ment of bacterial meningitis over the past 40 in young children, affecting up to 40% of years or so, including the continuing develop- cases, and may be the presenting feature.52 ment of antibiotics, the introduction of CT Convulsions associated with pyrexia-"febrile

and MRI imaging, the application of convulsions"-commonly occur in the very http://jnnp.bmj.com/ improved techniques of bacterial culture, and young age group; each such patient may have now the widespread use of molecular meth- meningitis, which poses the question, does ods incorporating PCR,47 have permitted each require ? Evidence sug- treatment to be given on a more rational basis gests that the answer is no. In one study,5' to a wider range of patients. It is disappoint- four out of 328, and 15 of 304 in another,54 ing therefore that mortality and morbidity had signs of meningitis on lumbar puncture. from the main forms of bacterial meningitis If the convulsion is brief and not associated have failed to diminish to any substantial with continuing neurological deficit or focal on October 2, 2021 by guest. Protected copyright. degree over the same length of time. feature, if the child rapidly regains conscious- Fortunately, in recent years, new understand- ness, and there are no other signs of meningi- ing of the pathophysiology of the changes that tis, it is reasonable to observe the child closely occur in bacterial meningitis has become for some hours rather than perform a lumbar available,"-50 from which it will be possible to puncture straight away. develop new strategies for treatment. Skin rashes are evident in about a third of One of the major features of meningitis is a patients, and their presence should be sought, CSF neutrophil response. The exact mecha- for in some they may be sparse. A diffuse nism by which white blood cells cross the maculopapular eruption which progresses to blood-brain barrier is not yet known. The include petechiae or frank purpura accompa- release of some bacterial cell wall components nies 50-60% of cases of meningococcal into the CSF stimulates the release of inflam- meningitis. Other bacteria-Listeria, staphylo- matory , including interleukins 1 cocci, pneumococci, Haemophilus, and and 6, tumour necrosis factor, and viruses, particularly echo 9, may also cause into CSF, and there is strong rashes. Rarely, a rash may be due to a reac- evidence that their presence induces inflam- tion to an or other drug, but this mation and disruption of the blood-brain bar- seldom occurs early in the course of the dis- rier. By a complex series of interactions, as ease. Shock may accompany any fulminant or Management ofcerebral infection 1249

overwhelming meningitis, and often compli- the problem with the microbiological labora- J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from cates meningococcal infection. Focal neuro- tory before taking CSF usually more than logical signs complicate meningitis in at least recoups the time spent, in order to ensure 15% of cases.55 Hemiparesis implies arteritis that appropriate culture media are available, or phlebitis and resulting infarction, or an and precautions taken to look for anaerobic evolving brain abscess. Cranial nerve palsies, organisms. commonly of the sixth and third nerves, are caused by basal exudation and inflammation. Lumbarpuncture Sixth-nerve palsies may indicate that intracra- Bacterial meningitis raises intracranial pres- nial pressure is rising dangerously. The level sure. In most cases physiological mechanisms of consciousness must be closely observed cope, and no long-term, untoward effects and documented, for it correlates best of all result. It is now becoming recognised that the with the likely clinical outcome. Mortality intracranial pressure can rise to potentially rises dramatically as consciousness lethal levels, with cerebral oedema and brain decreases,5657 and approaches 55% for adults herniation. In such circumstances, to do a in coma. Cerebellar signs may be noted in lumbar puncture can lead to rapid deteriora- some cases, and have no aetiological signifi- tion and death. To reconcile this with the cance. necessity of reaching an early diagnosis is, in When assessing a case of meningitis, dili- some cases, difficult; the correct decision gent search must be made for a non-neuro- depends on experience and a degree of luck. logical locus of infection, and cultures made In my view, CT should be undertaken imme- from it, or for a pre-existing or contempora- diately if there is evidence of focal neurology neous illness that enables organisms to on examination or historically, if the con- become more rapidly pathogenic. scious level is reduced or the patient is con- There is some correlation between organ- fused, if headache is severe, or if there are ism and condition. Pneumococcal meningitis frequent seizures. A recent report57 suggested is associated with , skull fractures, that CT had little influence on the manage- alcoholism, and sickle cell disease, and up to ment of bacterial meningitis in children, but 50% will have pneumonia. Haemophilus did not address the problem of recognition of meningitis is uncommon after early childhood raised intracranial pressure and lumbar punc- and its occurrence raises the possibility of a ture. Cerebral herniation and death do occur parameningeal focus of infection, including as a result of injudicious CSF examination in ear, nose, and basal skull fracture. cases of suspected meningitis,58-0 and a Staphylococcal infection complicates the normal CT scan does not always mean that it implantation of neurosurgical devices such as is safe to proceed to lumbar puncture: until shunts. In those with immunosuppression, better procedures to monitor intracranial including AIDS, simultaneous infection with pressure become routine, this decision must more than one organism can occur. be left to clinical experience, reassured by Tuberculosis meningitis may rarely be com- brain imaging to exclude space occupation plicated by superadded infection. In the and cerebral oedema. author's experience, diabetic patients are If there is a diathesis or coagula- dangerously liable to misdiagnosis because tion defect, these should be corrected before their symptoms may be misinterpreted as lumbar puncture. If it is decided that lumbar

being due to hypoglycaemia or to uncon- puncture is too dangerous, or there will be a http://jnnp.bmj.com/ trolled diabetes, and meningitis can exacer- delay before scanning can be done, immediate bate diabetes. If there is a history of treatment with antibiotics should be given, meningitis in the past, there is probably a after blood cultures have been taken. The dural fistula from previous or, antibiotic regimen chosen will be a "best much more rarely, the patient suffers from an guess", based on the age, medical background, inherited immunodeficiency state. In all geographical location, and clinical findings of cases, blood cultures should be taken before the patient. CSF may be examined later when antibiotics are given.55 it is considered to be safe, and evidence of on October 2, 2021 by guest. Protected copyright. Once the possibility of meningitis is enter- bacterial infection can be found by demonstra- tained the diagnosis must be, confirmed tion of antigen with newer techniques.6162 straight away. Next to the conscious level on Anxiety that infection may be introduced to admission, the factor that has the most dele- CSF by lumbar puncture in patients with terious effect on prognosis is delay in begin- bacteraemia is largely unfounded.63 ning treatment. CSF examination is the only CSF findings are of a raised cell count with sure way to confirm the diagnosis, and this is polymorphs predominating, at least 60%. almost always done by lumbar puncture. It There is no agreed value for the absolute may, in circumstances such as the existence count but usually this falls within the hun- of , be necessary to puncture dreds or more. The protein content is the ventricle to obtain a specimen. In estab- increased, and the glucose is reduced. A sam- lished meningitis there is usually no difficulty ple of blood should be taken at the time of in reaching a diagnosis. Very early in the evo- lumbar puncture for comparison: a ratio of lution of the disease, however, or if antibiotics CSF:blood glucose of less than 0-3 is abnor- have already been administered, or if the mal and is found in about three-quarters of organism is fastidious in its requirements for patients."M Unfortunately this is not specific laboratory culture, it may not be possible to and may be found in other chronic forms of identify the causal bacterium. Discussion of meningitis. 1 250 Anderson

Identification of the causal organism under raised intracranial pressure and, given for a J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from the microscope or on culture should be possi- short time, any side effects are minimal. ble in up to 80% of patients. A plethora of Other anti-inflammatory agents which act tests has been devised over the years to differ- at different stages are currently undergoing entiate bacterial from other forms of meningi- trials. These include non-steroidal, anti- tis: unfortunately they all suffer from lack of inflammatory drugs, pentoxifylline and specificity or sensitivity,65 and are being monoclonal antibodies targeted specifically superseded by newer, molecular techniques.6' against human f62 integrin.48 There is a com- CT and MRI scans need to be interpreted plex inter-relationship between cerebral blood with due consideration given to the timing of flow, cerebral blood volume, cerebral perfu- the scan in relation to the evolution of menin- sion pressure, cerebral oedema, and intracra- gitis: early in the disease the scan may be nial pressure which becomes disturbed in normal, but such a finding does not exclude bacterial meningitis. Decreased cerebral meningitis. There may be meningeal and ven- blood flow,74 cerebral oedema,58 and reduced tricular ependymal enhancement, basal puru- cerebral perfusion pressure75 all adversely lent , small ventricles with cerebral affect the outcome. Intracranial pressure is oedema, or large ventricles from CSF flow raised maximally within the first 48 hours,7677 obstruction in the aqueduct or basal cisterns. and it should be possible to anticipate this Later in the disease, areas of infarction from and ameliorate the effects. Patients whose vasculitis, and subdural collections may be conscious level deteriorates should be consid- seen, particularly in children with ered for intracranial pressure monitoring; if Haemophilus infection.6667 Brain imaging also the level rises above 15 mmHg, treatment to permits exclusion of other conditions that lower the pressure should be given. The head may present like meningitis-subarachnoid of the bed should be raised, hyperventilation and intracerebral haemorrhage, infarction, induced, and dexamethasone and mannitol tumour, and brain abscess. MRI is better at administered.78 demonstrating more subtle changes than CT, but it is technically much more difficult to CHOICE OF ANTIBIOTIC carry out MRI on a confused and uncoopera- To treat meningitis effectively, it is necessary tive patient, or on one who requires artificial to achieve adequate bactericidal levels of ventilation, than it is to perform CT. antibiotic in CSF 10-20 times higher than A full blood count and sedimentation rate the minimal bactericidal concentration in should be carried out. These are frequently vitro for the particular organism. Ideally the abnormal but unfortunately do not point to a antibiotic should be lipid soluble to facilitate specific pathogen. Blood gases, electrolytes transfer across the blood-brain barrier but and serum osmolality must be monitored this is not usually a problem in practice carefully, especially in children. Inappropriate because meningeal inflammation disrupts the secretion of antidiuretic hormone can lead to barrier and allows sufficient penetration. The a rapid fall in sodium, particularly but not rate at which it is metabolised and cleared exclusively with Haemophilus infection.68 from CSF determines the frequency and Other tests should be carried out according to amount which must be given, and it should clinical need. be active within purulent and acidic CSF. The choice of antibiotic will be determined TREATMENT on cul- by the organism found microscopy, http://jnnp.bmj.com/ Once the diagnosis of bacterial meningitis has ture or antigen detection, and modified by been made, treatment must be begun imme- knowledge of local and nosocomial drug diately. Precedence is given to maintaining resistance patterns. cardiorespiratory function, adequate oxygena- As important as the choice of antibiotic, is tion, and tissue perfusion while the causal the identification and eradication of any para- organism is being identified and the appropri- meningeal focus of suppuration. ate antibiotic regimen determined. If no For H influenzae type B infections, ampi- organism can be found, antibiotic treatment cillin and chloramphenicol have been the on October 2, 2021 by guest. Protected copyright. is given on a "best guess" basis. mainstay of treatment, but an increasing As mentioned earlier, the release of bacter- number of fl-lactamase-producing strains are ial cell wall products into CSF stimulates resistant to ampicillin, and a smaller number inflammation, mediated in part by cytokines. of chloramphenicol acetyltransferase-produc- Studies to determine the efficacy of the anti- ing organisms are resistant to chlorampheni- inflammatory action of corticosteroids in col.79 Third-generation cephalosporins are to reducing this inflammation have produced be preferred, and or ceftriaxone encouraging results in children, and trials are have been most widely used and demon- proceeding with adults. There is now good strated to be effective.808' For meningococcal evidence that dexamethasone in a dose of meningitis, penicillin remains the preferred 15 mg/kg body weight every six hours for four drug. There have been reports of meningo- days reduces the incidence of neurological cocci resistant to penicillin,82 but to date this sequelae,6F7' and this should be begun just does not appear to cause problems if large before antibiotics are given. I believe the evi- doses are given. Chloramphenicol or a third- dence is sufficiently good to justify the adop- generation cephalosporin may be used in peo- tion of a similar policy in adult bacterial ple with adverse reactions to penicillin. meningitis until further trial results are avail- Pneumococcal meningitis is best treated with able. Dexamethasone also helps to reduce penicillin. Penicillin-resistant strains are being Management ofcerebral infection 1251

recognised,79 and in these circumstances, evolves is longer than purulent meningitis, up J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from a third-generation cephalosporin or van- to three weeks in children and rather longer comycin should be used. Cephalosporins in adults. In some cases, however, the onset have been shown to be more effective than may be acute and exactly similar to purulent aminoglycosides for the treatment of Gram- meningitis, so tuberculosis must be consid- negative bacillary meningitis,83 and third-gen- ered in each case. eration antibiotics should now be used. Details of treatment regimens may be Group B streptococcal infection should be found in references 87, 88 and 89. treated with penicillin or a combination of Chemotherapy is given for between six penicillin and ampicillin, which is said to be months and a year, dependent upon the synergistic, and ampicillin covers those rare severity of disease and the response to treat- cases where the organism is penicillin-resis- ment. Steroids are not given routinely. tant. Listerial infection responds to ampi- Dexamethasone in high dose- for a short time cillin, and for those with penicillin allergies, is useful to reduce cerebral oedema. trimethoprin-sulfamethoxazole is effective. Hydrocephalus commonly develops, some- Nafcillin, flucloxacillin or oxacillin in high times acutely, and should be treated early by dose is the recommended treatment for S surgical drainage. aureus meningitis, with vancomycin for those who cannot have penicillin.84 Lyme disease Shunt infections are commonly caused by Lyme disease, caused by the spirochaete coagulase-negative staphylococci, and van- Borrelia burgdorferi, can present as meningitis, comycin-perhaps with added rifampicin-is in most cases of chronic type but sometimes effective. In most cases it is necessary to in an acute fulminant manner. remove the device to eradicate the infection Meningitis may occur as part of the second completely. The combination of penicillin stage of the disease, one to six months after and chloramphenicol with metronidazole is being bitten by the vector, an Ixodes tick, and recommended for anaerobic infections. because the original bite and the skin lesion In cases where an organism cannot be of erythema chronicum migrans were not identified, children should be treated with a noticed or had been forgotten, presentation combination of ampicillin and chlorampheni- may rarely be with acute meningitis.9>92 col, or a third-generation cephalosporin Treatment should be given with parenteral alone85; adults with penicillin or a third-gen- third-generation cephalosporins92 for a mini- eration cephalosporin; and the elderly with a mum of two weeks. Intravenous penicillin in third-generation cephalosporin and ampi- high dose is also effective,93 but may induce cillin. For empirical treatment of those with temporary worsening of symptoms akin to the immunosuppression, the type of immunologi- which occurs cal abnormality governs the choice of antibi- Jarisch-Herxheimer reaction, otic.86 If the defect is of cell-mediated when treating neurosyphilis; so it would be immunity, ampicillin should be included to prudent to cover the first doses with steroids. treat Listeria; if humoral immunity is depressed, penicillin and a third-generation Syphilitic meningitis cephalosporin are recommended; and for the Another spirochaete, Treponema pallidum, patient with neutropenia, a third-generation may also cause acute meningitis. Admittedly cephalosporin, with ceftazidime and an rare, such cases are occurring more frequently aminoglycoside, if Pseudomonas infection is in association with HIV infection and clinical http://jnnp.bmj.com/ likely.49 AIDS.9495 It occurs within two years of infec- Parenteral treatment for all forms of bacte- tion, and presents acutely with headache, rial meningitis is recommended, and dosage vomiting, pyrexia and neck stiffness. Cranial schedules and a description of the adverse nerve palsies are not sainfrequent,ernenor arer effects of these drugs may be found in phar- epileptic seizures. 96 CT scan appearances are macological texts. The author's policy is to of non-specific meningeal enhancement, and

discuss the treatment of any case which is at CSF contains a predominantly mononuclear on October 2, 2021 by guest. Protected copyright. all complicated with a specialist in infectious pleocytosis, raised protein, and normal sugar. diseases. The length of time for which treat- The diagnosis is made by finding a positive ment is given must be tailored to each case CSF or serum VDRL test, and confirming it and should be at least 10 days. Failure to with the FTA-ABS test. The results of PCR respond prompts a review of the diagnosis tests for treponemal DNA are awaited with and treatment, search for a continuing source interest, in the hope that it may be possible to of infection, and repeat brain imaging to differentiate active syphillis from an acute demonstrate abscess or subdural fluid accu- neurological syndrome occurring in a patient mulation. Repeat CSF examination should-be who has had yaws in the past. considered in most cases where no explana- Acute syphilitic meningitis is managed by tion can be found. the parenteral application of large doses of penicillin for 10 days, with steroids for the first 24 hours to diminish the possibility of SPECIFIC BACTERIAL MENINGITIDES inducing a Jarisch-Herxheimer reaction. It is Tuberculous meningitis necessary to follow up all cases closely to Tuberculous meningitis does not commonly ensure that serological changes and CSF present as a neurological emergency. The revert to normal, and to treat any re-infection time course over which the clinical syndrome rapidly. 1252 Anderson

Rickettsial meningitis fakiparum and results in cerebral malaria. For J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from As a consequence of the proliferation of inter- practical purposes, P vivax, P ovale and P continental travel by air, the person arriving malariae do not cause cerebral malaria. from abroad may harbour an infection not Malaria is found in large parts of Asia, sub- usually seen in the UK and, as the incubation Saharan Africa, parts of Greece, Turkey, and period of some of these infections may be two the Middle East, and Central and South to three weeks, symptoms may not appear America. People who live in endemic areas until some time has been spent in the coun- may acquire a degree of immunity which try. It is therefore of the first importance to diminishes with time. Those most at risk of obtain a full history of recent travel, places developing severe malaria are the traveller visited, including stop-overs, and activities who has no immunity and is being exposed undertaken, from any patient who presents for the first time, the emigrant returning with an acute meningoencephalitis. Such an home from a long spell abroad, whose immu- illness may be caused by Rickettsial diseases, nity has lapsed, pregnant women and those which are transmitted to man by the bites of who have immunosuppression. ticks or mites, and are prevalent in every con- Cerebral malaria is an acute diffuse tinent except Antarctica. Rocky Mountain with fever, which will kill spotted fever caused by R rickettsii may be within 72 hours if not recognised and treated, contracted in the Americas; Mediterranean and even then carries a mortality rate of spotted fever (R conoriz) in Africa, Asia and 25-50%.101 Delay in making the diagnosis the Mediterranean basin; scrub typhus (R and initiating treatment is a major factor con- tsutsugamushz) in Asia, the Pacific and tributing to this. Australia; typhus (R prowazckiz) and Q fever (Coxiella burneti) are ubiquitous. Rickettsiae PATHOGENESIS invade small blood vessels throughout the The clinical manifestations of cerebral body, multiply, causing endothelial wall pro- malaria are the end result of a complex, and liferation, thrombosis, and perivascular not fully understood, series of interactions inflammation, and when this affects the brain, within the cerebrovascular tree mediated by meningoencephalitis results. humoral, vascular, and haematological fac- The incubation period, clinical features, tors. There is evidence of an immune-medi- and progression of the illness vary between ated inflammatory reaction that releases organisms, but all share the triad of high vasoactive products which produce endothe- fever, skin rash, and headache, with menin- lial damage.'02 103 The presence of tropho- goencephalitis developing during the second zoites within erythrocytes induces changes as week of the illness in a significant proportion they mature, which renders them capable of of cases. Rocky Mountain spotted fever may adhering to vascular endothelium, and these occur without a rash; Q fever is not usually red cells packed with mature trophozoites are accompanied by a rash; and Mediterranean to be found sequestrated in small cerebral spotted fever is characterised by a distinctive venules, resulting in microvascular congestion eschar at the site of the bite. and tissue hypoxia.'04 105 Neurological features are typically non- Other organ systems are vulnerable to sim- focal, with headache, neck stiffness, photo- ilar changes. Acute tubular necrosis and the phobia, and, in more severe cases, confusion adult respiratory distress syndrome may need and reduction in conscious level.9798 simultaneous treatment. Disseminated Convulsions may occur. CSF examination intravascular occurs in severe http://jnnp.bmj.com/ reveals no specific features: there may be a cases, but seems to be a complication and slight pleocytosis, the protein content may be does not affect most patients. Whether cere- raised, and glucose is normal. Diagnosis is bral oedema contributes significantly to the confirmed by serological tests, but treatment pathophysiology of cerebral malaria is debat- must be started on the basis of clinical suspi- able. It has been observed in pathological cion, and not delayed for test results to specimens and it is claimed that this may be

become available. Direct immunofluores- an agonal event. CT studies do not support on October 2, 2021 by guest. Protected copyright. cence of skin biopsies taken from a rash of the presence of cerebral oedema,'06 yet CSF Rocky Mountain spotted fever offered hope pressure has been found to be high at the as a rapid method of diagnosis. Unfortunately time of lumbar puncture in a group of African it is not very specific, and is not widely avail- children.'07 Routine administration of dexam- able. The treatment of choice is tetracycline ethasone does not improve the outcome.'08 500 mg four times per day for 10 days in Hypoglycaemia is an important and fre- adults. Chloramphenicol may also be used quent complication of cerebral malaria and but it is thought that tetracycline is better and undoubtedly contributes to depression of less likely to result in relapse.99 consciousness and the appearance of neuro- logical signs. It results from a complex series of interactions which include malabsorption Malaria of glucose from the gut, hyperinsulinaemia Malaria is the most important parasitic dis- which may be induced by treatment using ease in the world, affects as many as 300 mil- intravenous quinine, the metabolic demands lion people, and is being imported to the UK of infection, and increased metabolism of * more often-12 patients died of malaria in available glucose by a large biomass of para- 1991, 10 in 1992.1'0 Involvement of the brain sites.'09110 Pregnant women are particularly occurs in patients infected with Plasmodium susceptible. Management ofcerebral infection 1 253

DIAGNOSIS before lumber puncture and, if the brain J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from Definitions of the clinical criteria required to looks swollen, it may be better to treat for justify a diagnosis of cerebral malaria have meningitis empirically, with cover for Gram- been drawn up to standardise parameters for negative organisms, and to consider giving research.'08 For the purposes of this discus- mannitol. Blood cultures should be taken. sion, any person from an endemic area, with All such patients should be nursed in inten- cerebral symptoms or signs should be sus- sive care. pected of suffering from cerebral malaria, and Because of widespread resistance of demands immediate monitoring, investiga- P falciparum to chloroquine, this should not tion, and treatment. Such a patient will usu- be used: quinine is the drug of choice. In the ally have been ill for days and have a United States quinine is not available, and temperature. Exceptions occur: the onset of quinidine gluconate is used instead."' 112 neurological symptoms may be abrupt, and Quinine is given by slow, intravenous infu- very ill patients can be hypothermic. sion after an initial loading dose (provided Commonly there are convulsions, more so in none has been given in the preceding two the young and in 45% of adults; coma follows days). Dosage schedules vary but approxi- the convulsion. Young children have febrile mate to 5 mg/kg by intravenous infusion over convulsions associated with many varieties of two hours every 8-12 hours. In patients from infection, and with these, consciousness is southeast Asia, it may be necessary to double rapidly regained. Coma lasting more than an the loading dose. Quinine should not be hour or two in such circumstances should given by bolus injection. When the patient raise the suspicion of cerebral malaria. regains consciousness, oral treatment can be Symptoms before the onset of the cerebral given. Plasma concentrations should be mon- state are diverse and non-specific, and the itored, as should the ECG. It is recom- most common misdiagnosis is of flu. The mended that an antifolate metabolite, either level of consciousness is invariably impaired, pyrimethamine-sulfadoxine (3 tablets) or and there is evidence of an organic brain syn- doxycycline, be given during the course of drome which manifests as confusion, halluci- quinine. If the parasite load is very high and nations, delirium, and psychosis; in some, the patient is very ill, exchange transfusion is complicated by motor signs and movement useful. Anticonvulsants are given to counter disorders. Meningism and opisthotonus may any tendency to seizures. Intravenous glucose be found, and it is important to remember is given for hypoglycaemia, anaemia is cor- that bacterial and other forms of meningitis rected by blood transfusion, and renal failure may coexist with malaria. Furthermore, by dialysis. Mechanical ventilation is neces- severely ill patients are susceptible to Gram- sary for the adult respiratory distress syn- negative septicaemia. There may be signs of drome. Gram-negative septicaemia causing damage to other organ systems-anaemia is septic shock can be a life-threatening compli- common and may be severe, jaundice compli- cation which should be anticipated and cates severe infection in adults, treated. Chloroquine is just as effective as hepatosplenomegaly of mild to moderate quinine provided that the organism is not degree is frequent, and haemorrhages may be resistant. found in the retina. Prolonged coma and fre- Features that adversely affect the outcome quent convulsions should trigger a search for are failure to make the diagnosis, delay in hypoglycaemia and pyogenic meningitis. starting treatment, not recognising concur- rent infection, deep, and prolonged coma, and http://jnnp.bmj.com/ continuing seizures. TREATMENT Confirmation of the diagnosis of a case of suspected cerebral malaria must be obtained Primary amoebic meningoencephalitis forthwith, but if this cannot be done, and sus- This condition is caused by free living picion is high, treatment should be given amoebae of the species Naegleria fowleri. straight away. The best way to confirm the Acanthamoeba tends to cause a subacute gran- on October 2, 2021 by guest. Protected copyright. diagnosis is to have thick and thin smears of ulomatous meningoencephalitis. They live in blood examined by an experienced micro- moist soil, and most cases have been reported scopist after appropriate staining. Should no in children who have been swimming or play- parasites be seen, and the diagnosis seems ing in stagnant water."'1 Amoebae enter the likely, treatment should be given and the nasal cavity, cross the nasal epithelium and blood re-examined every six to eight hours for ascend to the brain along olfactory nerves and the next 48 hours. Blood sugar should be blood vessels to frontal and basal meninges, estimated and monitored frequently, and and spread, causing a florid necrotising hypoglycaemia corrected if found. A full inflammation. blood count, urea, electrolytes, and blood Clinically the presentation is of sudden gases must be checked and monitored as onset, with a severe meningitis indistinguish- treatment progresses. It is necessary to exam- able from bacterial meningitis-the clue to ine the CSF to exclude pyogenic or other diagnosis is the history of exposure to warm, forms of meningitis as a cause of coma: CSF stagnant water. CSF examination reveals examination is not otherwise helpful in the pleocytosis in which polymorphs predomi- diagnosis of malaria. In the UK it should be nate, raised protein and reduced glucose. No possible to obtain an urgent CT brain scan, organisms are seen on the , and to give some estimate of intracranial pressure special examination of fresh, warm specimens 1254 Anderson

of CSF will show motile trophozoites. Most focus and the bacterial composition of the J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from patients die rapidly despite treatment, but abscess. Otogenic infection from otitis media survival following therapy with amphotericin or a spawns in the B has been described,"14 and this should be temporal lobe or' cerebellum. Paranasal given in the highest tolerated dose parenter- leads to abscess formation in the ally, reinforced by intracisternal injection, and frontal lobes and the deeper parts of the tem- by rifampicin and tetracycline, which have poral lobes. Metastatic abscesses, borne by some activity against Naegleriae. blood, have a predilection for grey-white mat- ter junctions where the blood supply is rela- tively reduced, and they are often multiple. Brain abscess The primary source may be , , or Parenchymal brain abscess remains a diag- teeth, but a source of infection such as nostic and therapeutic challenge to the physi- osteomyelitis may be implicated. Chronic cian and surgeon, notwithstanding the pulmonary sepsis, congenital heart disease, considerable advances that have taken place particularly cyanotic, and pulmonary arterio- in the last decades in imaging techniques, venous fistulae are all fruitful sources, as are neurosurgical practice, bacteriological isola- dental sepsis and dental manipulation. Those tion of causal organisms, and the introduc- abscesses that are caused by direct implanta- tion of more potent antibiotics. Patterns of tion occur in proximity to the site of injury or infection are changing but the number of surgery, commonly a depressed fracture or cases seen is not diminishing, and the result gunshot wound. Bacterial meningitis may be of treatment is little better than it was years complicated by brain abscess but it is difficult ago. Brain abscesses frequently present as to know how often this happens. emergencies, and as with other forms of cere- Depression of the immune system from bral infection, delay in diagnosis and imple- drugs or a disease such as AIDS renders the mentation of treatment, and inappropriate neuraxis vulnerable to attack by a wide range investigation, adversely affect the end result. of organisms, including those that would not The incidence of brain abscess has normally be pathogenic. The relative fre- remained at one in 10 000 of general hospital quency with which these predisposing factors admissions, approximating to a prevalence of occur changes with time, age of the patient, 4-5 per million population,"5 which is almost population studied, and geographical loca- certainly an underestimate. tion. In any single patient, all potential sources should be considered and excluded in turn by appropriate investigation. PATHOGENESIS Bacteria reach brain parenchyma via the Causative organisms bloodstream, by direct extension from an A very wide spectrum of organisms has been adjacent focus of infection or by implantation isolated from brain abscesses, and is set to through wounds as a result of trauma or neu- widen further as immune suppression and rosurgery. In as many as 20% of cases the AIDS spread. The ability to identify an source of infection cannot be identified. organism from pus is dependent on several Experimental data indicate that bacteria can- factors: have antibiotics already been admin- not set up a nidus of infection in normal, istered? Has there been a delay between undamaged brain, and it may be that throm- collection and culture? Is the organism fastid- bophlebitis spreads from contiguous infec- ious, and have the correct methods of culture http://jnnp.bmj.com/ tion, or microinfarction occurs from emboli been used? When proper attention has been or hypoxaemia to produce a microscopic area given to those details it has become evident of necrosis where infection can become estab- that anaerobic bacteria play a larger role in lished. Thereafter the evolution of an abscess brain abscess than had been appreciated, and passes through four main stages, which have the yield of positive cultures can be signifi- been documented experimentally and corre- cantly increased."18 late well with both clinical and CT findings. Otogenic brain abscesses commonly have on October 2, 2021 by guest. Protected copyright. The first stage is of , with sur- a mixed flora, which include Entero- rounding oedema of white matter; next the bacteriaceae, streptococci and Bacillus fragilis. centre of cerebritis becomes necrotic, Those caused by sinus infection grow enlarges, capsule formation begins with the Streptococcus milleri and Bacteroides. Dental appearance of fibroblasts and neovascular sepsis causes mixed infection with strepto- change at the periphery, reactive astrocytosis, cocci, Bacteroides and Fusobacterium; pul- and surrounding oedema. The third stage is monary disease results in a very mixed bag of capsular development; and the fourth is of including Fusobacterium, other anaerobes, capsular maturity as it thickens and reactive streptococci and actinomycetes. Patients with astrocytosis increases."16 117 The time course is congenital heart disease are likely to have variable and probably related to the virulence S viridans, anaerobic and microaerophilic of the organism and to the immune status of streptococci. Staphylococci are found with the host, and may be as short as three weeks penetrating head trauma, as are streptococci to encapsulation and fibrosis, and the capsule and Clostaidium. In patients who are immune may thicken by 1 mm per month. compromised, and a wide range of The range of primary foci from which bac- fungi may be implicated. teria spread to produce an abscess is large. Published reports abound of single cases of There is a relationship between the primary abscesses from which unusual organisms have Management ofcerebral infection 1255

been grown. It therefore behoves the attend- ance as encapsulation takes place. Later still J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from ing physician to alert the bacteriological labo- this becomes more evident and the centre of ratory to the relevant circumstances of the the lesion or lesions becomes increasingly case under investigation so that appropriate hypodense. Adjacent white matter is swollen techniques can be employed to identify the with cerebral oedema. Location and multi- causal organism. plicity of abscesses, hydrocephalus, and cere- bral mass effect are readily identified.'25 126 DIAGNOSIS Unfortunately such appearances are not spe- The clinical features of brain abscess are cific and may be seen with brain tumours, many and varied. Typically there is a combi- , necrotising encephalitis, and nation of raised intracranial pressure, focal infarction. If is demonstrated by neurological signs, signs of infection-either ependymal enhancement, then infection is focal or systemic-and rapid progression. likely. Unfortunately such a combination exists in MRI is likely to prove more sensitive than only half of all patients, with the result that CT in demonstrating early cerebritis, cerebral many other disease states may be mimicked: oedema, and the contents of abscess cavi- "Any neurologist or neurosurgeon who can ties.'27128 As an emergency investigation it is claim never to have made an error in the clin- not widely available in the UK, however, and ical diagnosis of a brain abscess either has a there are problems accommodating severely conveniently short memory span or has not ill patients within the machine. Lumbar been in practice long enough."119 puncture should never be carried out on Headache is common and may be diffuse, patients suspected of harbouring a brain localised and intermittent. abscess. If the urgency of the patient's condi- Emergency presentation may be with focal tion allows, a chest radiograph should be car- signs pointing to the site of the lesion, with ried out. convulsions, which are usually generalised but may be focal, with raised intracranial TREATMENT pressure even to the point of coning, and with When the diagnosis of brain abscess is con- neck stiffness to suggest meningitis. There firmed, the therapeutic strategy will be influ- may be a pyrexia in 50% of patients and, in enced by several factors, and there is no some, symptoms relating to the source of overall agreement on when surgery should be infection, such as otitis, may predomi- undertaken, and if it is, which technique nate.'20'24 In young children, the presentation should be used. Indications for surgery are may be even more confusing-an enlarging relief of space occupation, confirmation of the head associated with vomiting and seizures diagnosis, and obtaining specimens of pus for can be mistaken for tumour or congenital culture. If the abscess is a consequence of anomaly. Cerebral abscess therefore must be head trauma, then surgery is mandatory for included in the differential diagnosis of those appropriate toilet, debridement, removal of who present acutely with raised intracranial fragments, and closure of dural defects. If pressure, epileptic seizures, , meningi- lesions are multiple or deeply situated, it is tis and encephalitis, and primary and sec- best to treat with antibiotics on the basis of ondary intracranial malignancy. organism identification from other sources, or Faced with a patient in whom intracranial if that is not possible, on a "best guess" prin-

abscess is a possible diagnosis, the priority is ciple governed by the likely source of infec- http://jnnp.bmj.com/ to confirm the diagnosis, identify the source tion. Close monitoring of the lesions with of infection and the responsible organism or serial CT or MRI is necessary; if they do not organisms, once vital functions have been sta- diminish in size, aspiration should be under- bilised. A full examination should be made taken. In people with immunosuppression, for a locus of infection, such as otitis media including those with AIDS, the threshold for or pelvic sepsis, and if found, cultures should aspiration of pus to identify the offending be made and steps taken to eradicate the organism is substantially lowered. It is seldom source. Blood cultures should be set up. necessary to resort to complete surgical exci- on October 2, 2021 by guest. Protected copyright. Baseline haematological and biochemical sion of an abscess: simple aspiration of the parameters should be established: they are contents of an abscess is the most frequently seldom useful for diagnosis but may change advocated technique-when carried out rapidly in the course of the disease. CT of the under CT stereotactic control, diagnostic head should be done as soon as possible. In material may be obtained in as many as 95% addition to visualising the intracranial con- of cases.'29'3' There is an increasing trend tents, note should be made of the state of the towards treating multiple abscesses, or those paranasal sinuses and the mastoid air cells. that are deeply situated, with antibiotics Skull fractures and cranial defects should be alone.'24 132 133 Results seem to justify this looked for. course of action. Interpretation of the CT scan should always be made with reference to timing in Choice ofantibiotics the evolution of the abscess. In the earliest The choice of antibiotic to be used is influ- stages of cerebritis, a low-density area may be enced by its capacity to penetrate not only evident, perhaps with surrounding oedema, into CSF but through the capsule wall and and no enhancement after contrast injection. into the pus, and to be active in the presence Later there may be patchy enhancement of pus. Little informnation is available on the which, with maturity, adopts a "ring" appear- penetration of newer antibiotics into brain. 1256 Anderson

Currently the favoured regimen for empirical Headache and signs of meningeal irritation J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from treatment, while awaiting the results of cul- are common, vomiting, papilloedema, and ture studies, is a combination of penicillin 24 diminution of conscious level imply raised mega units per day intravenously, chloram- intracranial pressure. Focal signs of hemi- phenicol 1 g intravenously every six hours, paresis, dysphasia and, less frequently, cere- and metronidazole. If staphylococci are sus- bellar disturbance may occur. Epileptic pected, nafcillin or flucoxacillin, vancomycin, seizures are very common and are often or fusidic acid should be used. The length of focal."37 A parafalcine collection may herald time for which treatment should be given its presence with signs of a paraparesis, must be determined for each case by clinical sphincter disturbance, or weakness of one response and improvement of CT scan leg."38 The diagnosis should be considered in appearances, and is usually no less than six any patient who presents acutely with neuro- weeks, often longer. In patients with AIDS, logical dysfunction on a background of infec- toxoplasmosis is the likeliest cause of a focal tion, especially if there are focal seizures and brain lesion, particularly if multiple. The if there is evidence of infection in the ear or range of possible pathogens is very wide, paranasal sinuses. however, and biopsy should be considered for To confirm the diagnosis, CT of the head every patient who has a negative toxoplasma with contrast enhancement. should be under- serology test or in whom CT scans are atypi- taken as an emergency. MRI may be more cal. In others, treatment should be given with sensitive;"39 but the same strictures apply as pyrimethamine and sulfadiazine, and progress for brain abscess. CT appearances may be monitored.- Clinical and CT improvement subtle in the early stages of development. The should take place within 14 days. classical appearance is of an extracerebral col- Dexamethasone is commonly used to treat lection of lower density than brain over the cerebral oedema, and in practice the benefit surface of the hemispheres and in the parafal- obtained in reducing intracranial pressure cine region. A rim of enhancement follows outweighs the potential hazard of diminishing contrast injection. If large enough, adjacent the host inflammatory response. Mannitol structures will be displaced, but bilateral effu- and hyperventilation may be necessary in sions may not cause evident shift. There may addition. The risk of developing epileptic be adjacent infarction and brain oedema.'40 seizures is not insubstantial, and prophylactic Pus or fluid may be seen in the paranasal anticonvulsants are recommended. sinuses or in the ear, and there may be evi- dence of osteomyelitis. A blood count should be taken: it often shows a polymorphonuclear Subdural empyema leucocytosis and raised ESR. Lumbar punc- This potentially lethal infection of the space ture should not be carried out even if menin- between the dura and arachnoid mater is gitis is suspected. most often due to spread of infection from ear or nose, 134135 and less commonly TREATMENT osteomyelitis. Rarely it may be caused by Most authorities agree that once the diagnosis haematogenous spread from a distant focus, is established, treatment should be started and it may follow penetrating head trauma. immediately with a combination of systemic In children, subdural effusions that compli- chemotherapy together with surgical explo- cate meningitis become infected.'36 Because ration and removal of pus, and the eradica- the subdural space is extensive, pus can read- tion of the primary source of infection. http://jnnp.bmj.com/ ily track over the brain surface from one side Unless culture sensitivities indicate otherwise, to another, and may accumulate along the a combination of penicillin, chloramphenicol, *falx, or it may remain localised and become and metronidazole is given, as for brain encapsulated in proximity to the primary abscess. Prophylactic anticonvulsants should focus. The organisms that are found reflect be prescribed and measures taken to combat the underlying cause. In adults organisms raised intracranial pressure.

from otitis and sinusitis predominate-aero- There are considerable differences of opin- on October 2, 2021 by guest. Protected copyright. bic and anaerobic streptococci, staphylococci ion concerning the relative merits of burrhole and Gram-negative organisms including aspiration and craniotomy in the treatment of Proteus and Pseudomonas. In infants and subdural empyema, and which method is young children H influenzae, Gram-negative employed depends upon the prejudices and bacteria and pneumococci are common. In as experience of the neurosurgeon on call. many as a quarter of all cases no organism Whichever technique is used it is important can be cultured from pus."34 to ensure that all loculi of pus are removed. Working in an institution where there are DIAGNOSIS vociferous proponents of each method, I con- Subdural empyema affects the young-more sider that craniotomy produces better results. than half of all cases are aged less than 20, and it is more common in boys. Presentation is frequently acute, as an emergency with 1 Wallgren A. Une nouvelle maladie infectieuse du sys- rapid deterioration. There is usually evidence teme nerveux central? (Meningite septique aigue). Acta Paediatr Scand 1925;4:158-82. of general infection: fever, tachycardia, 2 Johnson RT, Mims CA. Pathogenesis of viral infections malaise and rigors, and there may be signs of of the . N Engl J Med 1968;278:23-30, 84-92. local infection: face ache, and tenderness over 3 Dropulic B, Masters GL. Entry of neurotropic the sinuses or over the mastoid process. arboviruses into the central nervous system: an in vitro Management ofcerebral infection 1257

study using mouse brain epithelium. J Infect Dis ciency virus infection. N Engl J Med 1991;324: J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from 1990;161:685-91. 1644-50. 4 Ponka A, Pettersson T. The incidence and aetiology of 38 Navia BA, Petito CK, Gold JWM, Cho E, Hordan BD, central nervous system infections in Helsinki in 1980. Price RW. Cerebral toxoplasmosis complicating the Acta Neurol Scand 1982;66:529-35. acquired immune deficiency syndrome: clinical and 5 Beghi E, Nicolosi A, Kurland LT, Mulder DW, Hauser neuropathological findings in 27 patients. Ann Neurol WA, Shuster L. Encephalitis and aseptic meningitis, 1986;19:224-38. Olmsted County, Minnesota 1950-1981. 1. 39 Porter SB, Sande MA. Toxoplasmosis of the central ner- Epidemiology. Ann Neurol 1984;16:283-94. vous system in the acquired immunodeficiency 6 Centers for Disease Control. Sunmary of notifiable dis- syndrome. N Englj Med 1992;327: 1643-8. eases, United States 1988. MMWR 1989;37:1-57. 40 Cohn JA, McMeeking A, Cohen W, Jacobs J, Holzman 7 Feigin RD, Shackelford PG. Value of repeat lumbar RS. Evaluation of the policy of empiric treatment of puncture in the differential diagnosis of meningitis. N suspected Toxoplasma encephalitis in patients with the EnglJMed 1973;289:571-4. acquired immunodeficiency syndrome. Am J Med 8 Birkenmayer LG, Mushahwar IK. DNA probe amplifica- 1989;86:521-7. tion methods.J Virol Methods 1991;35:117-26. 41 Young IS. Mycobacterium avium complex infection. J 9 Miller HG, Stanton JB, Gibbons JL. Parainfectious Infect Dis 1988;157:863-7. and related syndromes: a critical 42 Noah ND. Epidemiology of bacterial meningitis: UK review of the neurological complications of specific and USA. In: Williams JD, Burnie J, eds, Bacterial . QJMed 1956;25:427-505. meningtis, London: Academic Press, 1987;93-115. 10 Johnson RT. The pathogenesis of acute viral encephalitis 43 Roos KL, Tunkel AR, Scheld WM. Acute bacterial and post-infectious encephalomyelitis. J Infect Dis meningitis in children and adults. In: Scheld WM, 1987;155:539-64. Whitley RJ, Durack DT, eds, Infections of the central 11 Whitley RJ. Viral encephalitis. N Engi J Med 1990; nervous system, New York: Raven Press, 1991:335-409. 323:242-9. 44 Durand ML, Calderwood SB, Weber DJ, et al. Acute 12 Klapper PE, Cleator GM, Dennett C, Lewis A. bacterial meningitis in adults. A review of 493 Diagnosis of herpes encephalitis via Southern blotting episodes. NEnglJ3Med 1993;328:21-8. of DNA amplified by polymerase 45 Downs NJ, Hodges GR, Taylor SA. Mixed bacterial chain reaction. JMed Virol, 1990;32:261-4. meningitis. Rev Infect Dis 1987;9:693-703. 13 Rowley AH, Whitley RJ, Lakeman FD, Wolinsky SM. 46 Heerema MS, Ein ME, Musher DM, Bradshaw MW, Rapid detection of herpes simplex virus DNA in cere- Williams TW. Anaerobic bacterial meningitis. AmJ brospinal fluid of patients with herpes simplex Med 1979;67:219-27. encephalitis. Lancet 1990;335:440-1. 47 Tompkins LS. The use of molecular methods in infec- 14 Whitley RJ, Tilles J, Linneman C, et al. Herpes simplex tious diseases. NEnglJ Med 1992;327:1290-7. encephalitis: clinical assessment. JAAM 1982;247: 48 Quagliarello V, Scheld WM. Bacterial meningitis: patho- 317-20. genesis, pathophysiology and progress. N Engl Y Med 15 Barza M, Pauker SG. The decision to biopsy, treat or 1992;327:864-72. wait in suspected herpes encephalitis. Ann Intern Med 49 Tunkel AR, Wispelwey B, Scheld WM. Bacterial menin- 1980;92:641-9. gitis: recent advances in pathophysiology and treat- 16 Whitley RJ, Cobbs CG, Alford CA, et al. Diseases which ment. Ann Intern Med 1990;112:610-23. mimic herpes simplex encephalitis: diagnosis, presenta- 50 Ashwal S, Tomasi L, Schneider S, Perkin R, Thompson tion and outcome. JAMA 1989;262:234-9. J. Bacterial meningitis in children: pathophysiology and 17 Whitley RJ, Alford CA, Hirsch MS, et al. Vidarabine ver- treatment. Neurology 1992;42:739-48. sus acyclovir therapy of herpes simplex encephalitis. N 51 Tureen JH, Dworkin RJ, Kennedy SL, Sachdeva M, EnglJMed 1986;314:144-9. Sande MA. Loss of cerebral autoregulation in experi- 18 Kennedy PGE. Neurological complications of varicella mental meningitis in rabbits. J Clin Invest 1990; zoster virus. In: PGE Kennedy, RT Johnson, eds, 85:577-81. Infections of the Nervous System. London: Butterworths, 52 Hambleton G, Davies PA. Bacterial meningitis. Some 1987;177-208. aspects of diagnosis and treatment. Arch Dis Child 19 Johnson R, Milbourn PE. Central nervous system mani- 1975;50:674-84. festations of chicken pox. Can Med Assoc J 1970; 53 Rutter N, Smales OR. Role of routine investigations in 102:831-4. children presenting with their first febrile convulsion. 20 Gautier-Smith PC. Neurological complications of glan- Arch Dis Child 1977;52:188-91. dular fever (infectious mononucleosis). Brain 1965; 54 Lorber J, Sunderland R. Lumbar puncture in children 88:323-34. with convulsions associated with fever. Lancet 1980; 21 Ter Meuien V, Muller D, Kackell Y, Katz M. Isolation i:785-6. of infectious measles virus in measles encephalitis. 55 Klein JO, Feigin RD, McCracken GH. Report on the Lancet 1972;2:1172-5. task force on diagnosis and management of meningitis. 22 Johnson RT, Griffin DE, Hirsch RL, et al. Measles Pediatrics 1986;78S:959-82. encephalomyeitis-clinical and immunologic studies. 56 Geiseler PJ, Nelson KE, Levin S, Reddy KT, Moses VK. NEngl3 Med 1984;310:137-41. Community acquired purulent meningitis: a review of 23 Aarli JA. Nervous complications of measles: clinical 1316 cases during the antibiotic era, 1954-1976. Rev manifestations and prognosis. Eur Neurol 1974;12: Infect Dis 1980;2:725-45. 79-93. 57 Friedland IR, Paris MM, Rinderknecht S, McCracken 24 Koskiniemi M, Donner M, Pettay 0. Clinical appear- GH. Cranial computed tomographic scans have little

ance and outcome in mumps encephalitis in children. impact on management of bacterial meningitis. Am J http://jnnp.bmj.com/ Acta Paediatr Scand 1983;72:603-9. Dis Child 1992;146:1484-7. 25 Azimi PH, Cramblett HG, Haynes RE. Mumps menin- 58 Horwitz SJ, Boxerbaum B, O'Bell J. Cerebral hemiation goencephalitis in children. JAMA 1969;207:509-12. in bacterial meningitis in childhood. Ann Neurol 26 Goldfield M, Sussman 0. The 1959 outbreak of Eastern 1980;7:524-8. encephalitis in New Jersey. 1. Introduction and 59 Addy DP. When not to do a lumbar puncture. Arch Dis description of outbreak. Am J Epidemiol 1968;87:1-10. ChiZd 1987;62:873-5. 27 Dickerson RB, Newton JR, Hausen JE. Diagnosis and 60 Rennick G, Shann F, de Campo J. Cerebral herniation immediate prognosis of Japanese B encephalitis. Am J during bacterial meningitis in children. BMJ 1993;306: Med 1952;12:277-90. 953-5. 28 Lincoln AF, Siverston SE. Acute phase of Japanese B 61 Desforges JF. The use of molecular methods in infec- encephalitis. Two hundred and one cases in American tious diseases. NEnglJMed 1992;327:1290-7. on October 2, 2021 by guest. Protected copyright. soldiers, Korea. JAMA 1950;150:268-73. 62 Eisenstein BI. The polymerase chain reaction: a new 29 Burke DS, Nisalak A, Ussery MA, Laorakpogese T, method of using molecular genetics for medical diag- Chantavibul S. Kinetics of IgM and IgG responses to nosis. NEnglJMed 1990;322:178-83. Japanese encephalitis virus in human serum and cere- 63 Eng RHK, Seligman SJ. Lumbar puncture induced brospinal fluid. J Infect Dis 1985;151:1093-9. meningitis. JAMA 1981;245:1456-9. 30 Acha PN. A review of rabies prevention and control sta- 64 Marton KI, Gean AD. The spinal tap: a new look at an tus in the Americas 1970-1980: overall status of rabies. old test. Ann Intn Med 1986;104:840-8. Bull World Health Organ nt Epiz, 1981;93:9-52. 65 Anderson M. Bacterial meningitis. In: Matthews WB, 31 Blenden DC, Creech W, Torres-Anjel MJ. Use of Glaser GH, eds, Recent advances in dinical neurology 4. immunofluorescence examination to detect rabies virus Edinburgh: Churchill Livingstone, 1984:87-121. antigen in skin of humans with clinical encephalitis. J 66 Cabral DA, Flodmark 0, Farrell K, Speert DP. Infect Dis 1986;154:698-701. Prospective study of computed tomography in acute 32 Heyward WL Curran The epidemiology of AIDS in bacterial meningitis. Pediatr 1987;111:201-5. JW. _ the US. SciAm 1988;259:72-81. 67 Sze G, Zimmerman RD. The magnetic resonance imag- 33 Selik RM, Jaffe HW, Solomon SL, Curran JW. CDC's ing of infections and inflammatory diseases. Radiol Clin definition of AIDS. N EnglJMed 1986;315:761. North Am 1988;26:839-59. 34 Redfield RR, Wright DC, Tramont EC. The Walter 68 Kaplan SL, Feigin RD. The syndrome of inappropriate Reed staging classification for HTLV-IIIfLAV infec- secretion of antidiuretic hormone in children with bac- tion. NEnglJ)rMed 1986;314:131-2. terial meningitis. J Paediatr 1978;92:758-61. 35 Goodin DS, Aminoff MJ, Chernoff DN, Hollander H. 69 Lebel MH, Freij BJ, Syrogiannopoulos GA, et al. Long latency event related potentials in patients Dexamethasone therapy for bacterial meningitis. N infected with Human Immunodeficiency Virus. Ann EngilJMed 1988;319:964-71. Neurol 1990;27:414-9. 70 Lebel MH, Hoyt MJ, Waagner DC, Rollins NK, Finitzo 36 Navia BA, Jordan BD, Price RW. The AIDS dementia T, McCracken GH. Magnetic resonance imaging and complex. I. Clinical features. Ann Neurol 1986;19: dexamethasone for bacterial meningitis. Am J Dis Child 517-24. 1989;143:301-6. 37 Barnes PF, Bloch AB, Davidson PT, Snider DE. 71 Havens PL, Wenderberger KJ, Hoffman GM, Lee MB, Tuberculosis in patients with human immunodefi- Chusid MJ. Corticosteroids as adjunctive therapy in 1258 Anderson

bacterial meningitis. A meta analysis of clinical trials. 106 Looareesuwan S, Warrell DA, White NJ, et al. Do J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.12.1243 on 1 December 1993. Downloaded from Am Y Dis Child 1989;143:1051-5. patients with cerebral malaria have cerebral oedema? 72 Odio CM, Faingezicht I, Paris M, et al. The beneficial A computed tomography study. Lancet 1983;i:434-7. effects of early dexamethasone administration in 107 Newton CRJC, Kirkham FJ, Winstanley PA, et al. infants and children with bacterial meningitis. N Engl J Intracranial pressure in African children with cerebral Med 1991;324:1525-1. malaria. Lancet 1991;337:573-6. 73 Schattner A. Should we add corticosteroids to the treat- 108 Warrell DA, Looareesuwan S, Warrell MJ, et al. ment of acute bacterial meningitis? Q J Med 1992; Dexamethasone proves deleterious in cerebral malaria. 82:181-3. A double blind trial in 100 consecutive patients. N Engl 74 Ashwal S, Stringer W, Tomasi L, Schneider S, JMed 1982;306:313-9. Thompson J, Perkin R. Cerebral blood flow and car- 109 White NJ, Looareesuwan S. Cerebral malaria. In: bon dioxide reactively in children with bacterial menin- Kennedy PGE, Johnson RT, eds. Infections of the ner- gitis. J Paediatr 1990;117:523-30. vous system, London: Butterworths, 1987:118-44. 75 Rebaud P, Berthier JC, Hartemann E, Floret D. 110 Phillips RE. Hypoglycaemia is an important complica- Intracranial pressure in childhood central nervous sys- tion of Falciparum malaria. QJ Med, 1989;71:477-83. tem infections. Intensive Care Med 1988;14:522-5. 111 Miller KD, Greenberg AE, Campbell CC. Treatment of 76 Minns RA, Engleman HM, Stirling H. Cerebrospinal severe malaria in the United States with a continuous fluid pressure in pyogenic meningitis. Arch Dis Child infusion of quinidine gluconate and exchange transfu- 1989;64:814-20. sion. NEnglJMed 1989;321:65-70. 77 McMenamin JB, Volpe JJ. Bacterial meningitis in 112 Wyler DJ. Malaria: overview and update. Clin Infect Dis infancy: effects on intracranial pressure and cerebral 1993;16:449-58. blood flow velocity. Neurology 1984;34:500-4. 113 Carter RF. Primary amoebic meningoencephalitis: clini- 78 Pickard JD, Czosnyka M. Management of raised cal, pathological and epidemiological features of six intracranial pressure. Y Neurol Neurosurg Psychiatry cases. J Pathol Bacteriol 1968;96: 1-25. 1993:56:845-58. 114 Seidel JS, Harmatz P, Visvesvara GS, Cohen A, Edwards 79 Kaplan SL, Fishman MA. Update on bacterial meningi- J, Turner J. Successful treatment of primary amoebic tis. Child Neurol 1988;3:82-93. meningo-encephalitis. N EnglJMed 1982;306:346-8. 80 Peltola H, Anttila M, Renkonen O-V, Finnish Study 115 McLelland CJ, Craig BF, Crockard HA. Brain abscesses Group. Randomised comparison of chloramphenicol, in Northern Ireland: a 30-year community review. J7 ampicillin, cefotaxime and ceftriaxone for childhood Neurol Neurosurg Psychiatry 1978;41:1043-7. bacterial meningitis. Lancet 1989:1:1281-7. 116 Britt RH, Enzmann DR, Yeagar AS. Neuropathological 81 Lebel MH, Hoyt MJ, McCracken GH. Comparative effi- and computerized tomographic findings in experimen- cacy of ceftriaxone and cefuroxime for treatment of tal brain abscess. J Neurosurg 1981;55:590-603. bacterial meningitis. J Pediatr 1989;1 14:1049-54. 117 Britt RH, Enzmann DR. Clinical stages of brain 82 Sutcliffe EM, Jones DM, El-Sheikh S, Percival A. abscesses on serial CT scans after contrast infusion. Penicillin-insensitive meningococci in the UK. Lancet Computerized tomographic, neuropathological and 1988;i:657-8. clinical correlations. J Neurosurg 1983;59:972-89. 83 Landesman SH, Corrado ML, Shah PM, Armengaud 118 de Louvois J, Gortvai P, Hurley R. Bacteriology of M, Barza M, Cherubin CE. Past and current roles for abscesses of the central nervous system: a multicentre cephalosporin antibiotics in the treatment of meningi- prospective study. BMJ 1977;ii:981-4. tis: emphasis on use in gram negative bacillary menin- 119 Ingham HR, Sisson PR, Mendelow AD, Kalbag RM, gitis. Am JMed 1981;71:693-703. McAllister VL. Brain abscess-clinical features. In: 84 Schlesinger LS, Ross SC, Schaberg DR. Staphylococcus Pyogenic neurosurgical infections. London: Edward aureus meningitis: a broad based epidemiological Arnold, 1991:39-50. study. Medicine (Baltimore), 1987;66: 148-56. 120 Chun CH, Johnson JD, Hofstetter M, Raff MJ. Brain 85 Klein NJ, Heyderman RS, Levin M. Antibiotic choices abscess. A study of 45 consecutive cases. Medicine for meningitis beyond the neonatal period. Arch Dis 1986;65:415-31. Child 1992;67:157-61. 121 Nielsen H, Gyldensted C, Harmsen A. Cerebral abscess: 86 Rubin RH, Hooper DC. Central nervous system infec- etiology and pathogenesis, symptoms, diagnosis and tions in the compromised host. Med Clin North Am treatment. Acta Neurol Scand 1982;65:609-22. 1985;69:281-93. 122 Shaw MDM, Russell JA. Cerebellar abscess-a review of 87 Wood M, Anderson M. CNS tuberculosis. In: Walton J, 47 cases. J Neurol Neurosurg Psychiatry 1975;38: ed. Neurological infections, London: Saunders, 1988: 429-35. 172-96. 123 Saez-1lorens XJ, Umann MA, Odio CM, McCracken 88 ShellerJR, Des Prez RM. CNS tuberculosis. Neurol Clin GH, Nelson JD. Brain abscess in infants and children. 1986;4: 143-58. Pediatr Infect DisJ 1989;8:449-58. 89 American Thoracic Society. Treatment of tuberculosis 124 Mampalam TJ, Rosenblum ML. Trends in the manage- and tuberculosis infections in adults and children. Am ment of bacterial brain abscess; a review of 102 cases Rev Respir Dis 1986;134:355-63. over 17 years. Neurosurgery 1988;23:451-8. 90 Reik L, Steere AC, Bartenhagen NH, Shope RE, 125 New PFJ, Davis KR, Ballantine HT. Computed tomog- Malawista SE. Neurologic abnormalities of Lyme dis- raphy in cerebral abscess. Radiology 1976;121:641-6. ease. Medicine (Baltimore), 1979;58:281-94. 126 Miller ES, Psrilal SD, Uttley D. CT scanning in the 91 Pachner AR, Steere AC. The triad of neurologic mani- management of intracranial abscess: a review of 100 festations of Lyme disease: meningitis, cranial neuritis cases. BrJ7Neurosurg 1988;2:439-46. and radiculoneuritis. Neurology 1985;35:47-53. 127 Davidson HD, Steiner RE. Magnetic resonance imaging 92 Finkel MJ, HalperinJJ. Nervous system Lyme borrelio- in infections of the central nervous system. Am J sis-revisited. Arch Neurol 1992;49:102-7. Neuroradiol 1985;6:499-504. http://jnnp.bmj.com/ 93 Steere AC, Pachner AR, Malawista SE. Neurologic 128 Schroth G, Kretzschmar K, Gawehn J, Voigt K. abnormalities of Lyme disease: successful treatment Advantage of magnetic resonance imaging in the diag- with high dose intravenous penicillin. Ann Intern Med nosis of cerebral infections. Neuroradiology 1987;29: 1983;99:767-72. 120-6. 94 Katz DA, Berger JR. Neurosyphilis in acquired immuno- 129 Apuzzo ML, Chandrasoma PR, Cohen D, Zee CS, deficiency syndrome. Arch Neurol 1989;46:895-8. Zelman V. Computed imaging stereotaxy: experience 95 Katz DA, Berger JR, Duncan RC. Neurosyphilis. A com- and perspective related to 500 procedures applied to parative study of the effects of infection with masses. Neurosurgery 1987;20:930-7. immunodeficiency virus. Arch Neurol 1993;50:243-9. 130 Dyste GN, Hitchon PW, Menezes AH, Vangilder JA, 96 Merritt HH, Moore M. Acute syphilitic meningitis. Greene GM. Stereotaxic surgery in the treatment of

multiple brain abscesses. J Neurosurg 1988;69: 188-94. on October 2, 2021 by guest. Protected copyright. Medicine 1935;14:119-83. 131 Young B. Role of stereotactic biopsy in the management 97 Kirk LJ, Fine PD, Sexton JD, Muchmore G. Rocky of transplant patients with intracranial lesions. Neurol mountain spotted fever: a clinical review based on 48 Clin 1988;6:639-44. confirmed cases 1943-1986. Medicine 1990;69:35-45. 132 BoomWH, Tauzon CV. Successful treatment of multi- 98 Raoult D, Zuchelli P, Weiller PJ, et al. Incidence, clinical ple brain abscesses with antibiotics alone. Rev Infect Dis observations and risk factors in the severe form of 1985;7: 189-99. Mediterranean spotted fever among patients admitted 133 Rousseaux M, Lesoin F, Destee A, Jomin M, Petit H. to hospital in Marseilles 1983-1984. Y Infect 1986; Developments in the treatment and prognosis of multi- 12:111-6. ple brain abscesses. Neurosurgery 1985;16:304-8. 99 Shaked Y. Rickettsial infection of the central nervous 134 Bannister G, Williams B, Smith S. Treatment of sub- system: the role of prompt antimicrobial therapy. QY dural empyema. J Neurosurg 198 1;55:82-8. Med, 1991;79:301-6. 135 Kaufman DM, Miller MH, Steigbigel NH. Subdural 100 Molyneux M, Fox R. Diagnosis and treatment of malaria empyema. Analysis of 17 recent cases and review of the in Britain. BrMedy 1993;306:1175-80. literature. Medicine 1975;54:485-98. 101 Romans GC. Cerebral malaria: the unsolved riddle. Y 136 Smith HP, Hendrick EB. Subdural empyema and Neurol Sci 1991;101:1-6. in children. Y Neurosurg 1983;58: 102 Maegraith B, Fletcher A. The pathogenesis of mam- 392-7. malian malaria. Adv Parasitol 1972;10:49-75. 137 Cowie R, Williams B. Late seizures and morbidity after 103 Molyneux M. Cerebral malaria in children; clinical subdural empyema. J Neurosurg 1983;58:569-73. implications of cytoadherence. Am J Trop Med Hyg 138 Hitchcock E, Andreadis A. Subdural empyema: a review 1990;43:38-41. of 29 cases. J Neurol Neurosurg Psychiatry 1964;27: 104 MacPherson GG, Warrell MJ, White NJ, Looareesuwan 422-34. S, Warrell DA. Human cerebral malaria: a quantitative 139 Weingarten K, Zimmerman RD, Becker RD, Heier LA, ultrastructural analysis of parasitized erythrocyte Haimes AB, Deck MDF. Subdural and epidural sequestration. Am J Pathol 1985;119:385-401. empyemas: MR imaging. Am J Neuroradiol 1987; 105 Warrell DA, White N, VeallS, et al. Cerebral anaerobic 10:81-7. glycolysis and reduced cerebral oxygen transport in 140 Zimmerman RD, Leeds NG, Danziger A. Subdural human cerebral malaria. Lancet 1988;2:534-8. empyema CT findings. Radiology 1984;150:417-22. Book reviews 877

The introductory chapter by R J Douglas students, postgraduate trainees, non-neu- £75-00). Oxford, Blackwell Scientific sets out the view that the various properties rologists, paramedicals and basic neuro- Publications Ltd. ISBN 0-07-001020-X. of the visual world such as form, orienta- scientists'. I would be pleasantly surprised if tion, colour and movement are processed American students, let alone British, would Stereotactic Radiosurgery has had an un- by separate cortical regions with the pri- read a book of this size as a basic text. Basic expectedly large influence on neurosurgical mary visual cortex acting as an elementary neuroscientists usually require highly practice as Charles Wilson points out in his detection and relay station. However there detailed information from reference books foreword to this excellent book on this sub- are differing views on how these regions are when considering the clinical context of ject. As Radiosurgery in its various forms is arranged. In the hierarchical model, based their own specific research endeavour. Dr used in many centres by doctors of widely mainly on morphological studies of neural Guberman's book might provide a suitable differing experiences and training it connexions, visual stimuli are analysed for neurology reference text for paramedicals, becomes more and more crucial that its use the various properties of form, colour and but surely the book's main niche must be is securely based on sound principles of orientation by successive relay to specific those training in general internal medicine, radiobiology and physics as well as of regions at increasing distance from the cal- such as MRCP candidates in the UK. Here stereotaxy and neurosurgical practice. Its carine cortex. A second view proposes two it already faces competition from estab- distinguished panel of authors are respected major pathways one concerned with what a lished books such as "Brain and Bannister's exponents of the technique and well placed visual object is and the other where it is. Clinical Neurology" and from the neuro- to give this volume authoritative accounts of The 'what' task is performed by the ven- logy chapters of general medicine text- the different characteristics of Gamma trally situated inferior temporal cortex and books. Knife, Linac, and Heavy Particle therapy, the 'where' task by the dorsally situated Dr Guberman is to be congratulated on the principles involved as well as the results occipito-parietal cortex. This view is derived the exhaustive breadth of his subject cover- that can be achieved and pitfalls to be from behaviour experiments in animals and age making him one of the most generally avoided. The range of conditions that have on the effects of brain lesions in man. A knowledgeable neurologists around. The been treated by radiosurgery is wide but it is third view based on physiological recording disease coverage is comprehensive, well-laid no panacea and its proper application from various points on the visual pathway out, with extensive referencing to core requires careful selection and disciplined proposes that visual information is divided papers. There is a most useful compendium execution. This book offers a valuable sur- from the retina onwards into two streams, a of the neurological side-effects of therapeu- vey of the present situation as well as some P (parvocellar) system thought to convey tic drugs. Inevitably reviewers can have provocative thoughts about what the future high resolution visual images at high con- minor quibbles. Colour fundus photographs may hold for the method. trast and an M (magnocellular) system con- would have been more helpful. D M C FORSTER veying low acuity visual information Unfortunately, the illustrations include too concerned with motion and depth percep- many CT scans and myelograms, some- The Surgical Management of . tion. The first chapter discusses these times reproduced with mediocre quality, in Edited by A R WYLER AND B P HERMANN. theories and also describes the precise ocu- an era where MRI now reigns supreme. I Publisher: Butterworth-Heinemann, Oxford lomotor mechanisms necessary to maintain personally don't believe that you can see 1994. (Pp 277; £60.00). ISBN 0-7506- high quality retinal images of moving fibrillations, rather than fasciculations, in 9416-5. objects. There follow a series of chapters on the tongue. How often is it necessary to various visual disorders in man including 'extend' the legs to measure the spinal fluid This is another book targeted at the grow- visual illusions and hallucinations, colour pressure when relaxed breathing and only ing market of clinical groups seeking educa- vision, prosopagnosia, Balint's syndrome, slight unbunching from the foetal position tion in the surgical treatment of epilepsy. object agnosia and residual vision and generally seem to be sufficient? It is a multi-author book divided into blindsight. The chapter by Milders and The extensive sections on clinical method four sections. The first deals with prognosis Perrett on facial recognition and by include topics such as probability, Baye's and is a helpful contribution; especially the Ruddock on acquired disorders of colour theorem and decision trees. I am never sure chapter devoted to epileptic syndromes. vision are particularly informative and com- about the decision tree model for the diag- The next section deals with preoperative prehensive. nostic process. Most neurologists of my evaluation and is reasonably comprehen- Considering the scope of the book and acquaintance seem to arrive at the right sive, although the separation of the chapters the different backgrounds and nationalities diagnosis intuitively, only resorting to the on basic neuropsychology and the Wada of the contributors the style is remarkably decision tree approach in desperation when test by a chapter on preoperative psychiatric even. Illustrations are used sparingly, per- otherwise baffled. Dr Guberman's advice evaluation is bizarre. Unfortunately the haps too sparingly, and the book compares about testing hypotheses during history tak- chapter on surgical procedures is weak. Ten unfavourably in this respect with the ing is sound enough in principle, but his pages on cortical resections including tem- superbly illustrated accounts of visual physi- example is somewhat abstruse. When faced poral lobe resections is quite inadequate. ology and psychology aimed at the general with a subacutely weak patient, how many The chapters on outcome are reasonable, reader. References are comprehensive and neurologists, let alone students, are suffi- although the one on reoperation presents up to date. Altogether this is a scholarly and ciently blessed with knowledge or time only the editors' own series. stimulating book for neurologists interested enough to enquire about 'making lead glass Much of the material in this book is from in how the brain analyses the visual world as a hobby or whether he has ever had wine- the editors' own writings, or those of their (and that should include all of us). coloured urine'? The advice about the colleagues, and although much of it is valu- R ROSS RUSSELL clinical methodology or neurology lacks suf- able, it does not constitute a good overview. ficient encouragement to focus the history As an introduction to the topic it is excel- An Introduction to Clinical Neurology on the problem in hand, and to develop the lent, well produced and at a reasonable (Essentials of Clinical Practice). By technique of performing a selective, five- price, and as such would be a good intro- ALAN GUBERMAN. (Pp 610; Price: £36.95). minute examination pertinent to this prob- duction for trainees in the appropriate 1993. A Little, Brown Medical title distrib. lem. Yet Dr Guberman's account of how to neurosciences disciplines, and nursing and in Europe by Churchill Livingstone, perform all the myriad components of neu- paramedical staff who wish to have a Edinburgh. ISBN 0-316-33073-6. rological examination is lucid and could be detailed account of these matters. used profitably by students wishing to learn CE POLKEY Contemporary medical books may have two "how to do it all". purposes. Either they can provide funda- Dr Guberman has undertaken the diffi- mental instruction about the principles and cult task of writing a comprehensive text- practice of clinical medicine, or they may book of general neurology by writing 26 out CORRECTION assemble and order factual knowledge for of the 29 chapters himself. I wonder how reference purposes. The huge growth in fac- many more detailed neurological treatises of Anderson, Milne. Management of tual detail has meant that fewer and fewer this quality, written by a single author, we cerebral infection. J Neurol Neurosurg books combine these roles. Dr Guberman's will see in the future. Psychiatry 1993;56: 1243-58 (Neurological book is one of a dying breed which tries MICHAEL DONAGHY emergency). The dosage of dexamethasone to serve both functions. I think it is pre- should be 0-15 mg/kg body weight every six dominantly intended as a textbook, being Stereotactic Radiosurgery. Edited by hours for four days. titled 'An Introduction...... ' and the preface E ALEXANDER III, J S LOEFFLER and This correction was wrongly attributed to states that it is intended for 'medical L D LUNSFORD. (Pp 254, Illustrated; Price: the editorial in the May issue. 1302 Book reviews

Who will read this review? Probably not dementia and principally in Alzheimer's dis- paediatric neurologists/epileptologists. They ease. Excellent though the late Professor will already have the first edition of the C E Wells book on dementia is, the recent book and will not need me to tell them how rapid advances in research methodology, Neuropsychiatric Disorders. By GARETH good the second edition is. Probably not principally in molecular biology and in W ROBERTS, P NIGEI LEIGH, DANIEL R paediatricians (not likely to take the jour- imaging techniques, mean that a replace- WEINBERGER. Published by Mosby Europe nal)-however any paediatrician who has ment is timely. The editors (two academic Ltd, London 1993. ISBN 1-56375-510-6. wandered into view should make sure that psychogeriatricians) "aim to encompass in a the local library has a copy. I hope adult single volume, all aspects of all types of The neurology-psychiatry distinction is neurologists and their colleagues will read dementia", a Herculean task. They admit being rendered obsolete by strident this review and will be persuaded to put the that they could not hope to succeed in such advances in neuroscience. Basic neuro- book on their own shelves. They will then an aim and have therefore concentrated on science and clinical neurology have much to understand why paediatricians have such a "providing a solid core of information offer each other, and this book, written by difficult (but interesting!) time. which is likely to remain part of the main distinguished neuroscientists and clinicians, CHRIS VERITY stream view of the field." Let us see if they aims to increase cross-fertilisation. have achieved this aim. The broad areas of study are the demen- The book is divided into two main sec- tias, motor disorders, epilepsy and the psy- tions, dementia of the Alzheimer type and choses. Within these areas, all of the major non-Alzheimer dementia. Considering the diseases are covered. large number of authors, there is little repe- Each main disease has a chapter in which Mitochondrial Disorders in Neurology. tition. There are some excellent sections in neuroscience and clinical aspects are Edited by A H V SCHAPIRA AND S DIMAURO. the Alzheimer section and I would recom- seamlessly integrated. Chapters follow the Published by Butterworth Heinemann, mend the overview by Absher and same overall structure of clinical features, Oxford 1994. Pp 254 £45-00 ISBN pathology, investigation and treatment. 0-7506-0585-5. Cumming on cognitive and non-cognitive aspects of dementia. The chapters on Pathological and imaging illustrations are of high--quality, while the accompanying text is Although there are many areas of neurology cholinergic and non-cholinergic neuro- transmitter systems as well as the careful a distillate of current thought, clearly where disease classification is in its infancy, explained for those new to the area. syndromes to and detailed chapter on the molecular attributable mitochondrial pathobiology of Alzheimer's disease are It is difficult to fault the book. If pushed abnormalities are the most neonatal in this to do so, there is little regarding the role of regard. This book aims, and largely suc- excellent. The chapters on the assessment of memory failure and dementia are thor- neuropsychology in diagnosis and disease ceeds, in assimilating a huge volume of data staging. The reference section is also some- which tries to correlate biochemical, clinical ough and well referenced but the section on imaging is disappointing. The chapters on what brief. Advances, such as the role of and molecular defects. From to superoxide dismutase in motor neurone dis- myopathy and peripheral neuropathy to computed tomography and neurophysiology exemplify the dictum that experts should ease, and the trinucleotide repeat sequences Parkinson's disease there are few areas for in the diagnosis of Huntington's disease, which a mitochondrial aetiology has not write on their own subject. However, the first two-thirds of the book, have occurred too recently for inclusion. been sought, and even fewer to which an The book is likely to succeed in its aim of acronym has not been assigned. MEPOP, the section on Alzheimer's disease, deals only with the elderly patient. Alzheimer's breaking down boundaries between neurol- MNGIE, MiMyCa and NARP are perhaps ogy and psychiatry, and between clinical less well known than LHON, MERRF and original patient was 51 and this is acknowl- edged but then the younger demented and basic neuroscience. Its main value is in MELAS to name but a few. The introducing the clinician and neuroscientist inescapable bottom line is well pointed out patient is ignored. This is particularly true in the section on services. The chapter on to the other's discipline. by Lewis Rowland in his excellent chapter JOHN GREENE which attempts to synthesise the field: the services in the U.K. does not mention the same mutation can cause a varied pheno- younger patient nor does it mention the ser- type and yet the same phenotype may be vices provided by voluntary organisations associated with a variety of mitochondrial and it does not reflect the situation in the DNA abnormalities. U.K. as I know it. In seeking to approach this problem from The last third of the book describes the different angles there is inevitably a degree non-Alzheimer . There is a lucid of repetition. Some of the chapters are more account of by O'Brien useful to molecular biologists and neuro- and a chapter on pathology of vascular chemists within the field and yet there is dementia by Brun. He regards Binswanger's much here for the interested clinician. disease as identical with etat lacunaire, a However, whilst occasional photographs of concept with which many British neurolo- gels may illustrate points, the absence of a gists would disagree. Towards the end of single ragged red fibre is rather an oversight. the book, almost as an appendage, are Perhaps we have all seen enough of these chapters on unusual causes of dementia. beasts but a chapter on the pathology of These are the chapters of most interest to would have been use- neurologists. Perhaps in a later edition these ful. Having said this, there are some excel- chapters together with the excellent chapter lent diagrams and tables which sum-up on prion diseases could be expanded and current research and allow the non-molecu- placed earlier in the book. As it stands, the lar biochemist to understand some of the balance seems wrong, too little attention is intricacies of these difficult but important given to younger patients and to non- disorders. Overall this is a unique and Alzheimer causes of dementia and yet it is invaluable reference book for all those inter- these patients who deserve accurate diag- ested in mitochondrial diseases. noses and for whom a different manage- JOHN ZAJICEK ment might be appropriate. The indexing and proof reading is careless-look up the nomenclature of Binswanger's disease and you would be directed to Page 624, which is blank. CORRECTION So have the authors succeeded in their Dementia. Editors: ALISTAIR BURNS and aim? For geriatricians and psychogeriatri- RAYMOND LEVY. Publisher: Chapman and cians this book will become the standard Anderson M. Management of cerebral Hall, London 1994. (Pp 874; £125-00.) reference book. For neurologists however, infection. J Neurol Neurosurg Psychiatry ISBN 0-412-54740-6. this book has serious flaws and does not 1993;56:1243-1258 (neurological emer- achieve its potential. Perhaps it will if there gency). The dosage of dexamethasone After decades of neglect, the last twenty is a second edition. should be 0-15 mg/kg body weight every six years has seen a blossoming of interest in MYFANWY THOMAS hours for four days.