OVERVIEW: BRAIN TUMOUR DIAGNOSIS AND MANAGEMENT/ROYAL COLLEGE OF PHYSICIANS GUIDELINES

Transcription

1 ii18 OVERVIEW: BRAIN TUMOUR DIAGNOSIS AND MANAGEMENT/ROYAL COLLEGE OF PHYSICIANS GUIDELINES PRIMARY Correspondene to: Dr Robert Grant, Department of Clinial Neurosienes, Western General Hospital, Crewe Road, Edinburgh EH4 2XU, UK; a.uk T RGrant J Neurol Neurosurg Psyhiatry 2004; 75(Suppl II):ii18 ii23. doi: /jnnp his artile is intended to reflet the urrent level of servies for patients with brain tumours in the UK, warts and all. I have taken the evidene from first hand experienes auditing pratie and implementation of the Royal College of Physiians guidelines for good pratie in three of the four neurosiene entres in Sotland. I antiipate the findings are no better or no worse than what you will find in your loal neurosiene entre. I hope it enourages you to work losely with neuropathology, neuro-radiology, neurosurgial and linial onologial olleagues to improve the standard of are for patients with brain tumours. Hopefully this review will provide some evidene and support for submissions to improve the servie in your loal entre. AND SECONDARY BRAIN TUMOURS Primary brain tumours, although numerially fairly unommon (inidene 8/ per year), have a major impat on family and working life as they are the most ommon solid tumour in hildren and the eighth most ommon aner in people of working age. 1 Cerebral gliomas aount for. 90% of primary brain tumours and are the fifth most ommon ause of death from aner under the age of 65 years. They have a five year survival of only 18%. Patients with brain tumours are rare in general pratie (four or five new ases in one general pratitioner s (GP s) working lifetime), yet they remain a ommon onern of patients and GPs. Primary and seondary brain tumours present with similar symptoms and an be diffiult to distinguish either linially or by imaging. Usually, patients are referred to a physiian or neurologist at their loal major hospital. Brain imaging demonstrates an abnormality that ould be a tumour. The patient may then be referred to a neurosurgeon. Care pathways an be depited as a timeline (fig 1). At eah deision time point patients are streamed into more or less ative forms of management depending on age, performane sore, and presumptive or known histology. Following the presumptive diagnosis of malignant glioma, management may involve: Steroids Dexamethasone is used to redue swelling around the tumour and help symptoms in up to two thirds of patients. Surgery The first deision (fig 1, D1) is whether surgery should be performed or not and if the deision is to operate, the seond deision is whether diagnosti biopsy or removal of the tumour should be performed (fig 1, D2). Radiotherapy The next deision is whether the patient should have radiation therapy or not (fig 1, D3). Rarely, patients with suspeted glioma reeive radiation without surgial onfirmation. Chemotherapy On reurrene the next deision is whether hemotherapy is advisable (fig1, D4). We will look at the mode of presentation, and fators influening speed of referral, and the various treatment options, while trying to emphasise ontinuity of are, good ommuniation, and the Royal College of Physiians guidelines for good pratie in the management of malignant glioma. 2 3 The Royal College of Physiians guidelines emphasise that quality of are adopts an inreasingly important role when ure is not possible. The linial areas overed are summarised in hronologial fashion in the patient s journey of are (surgery/radiotherapy) in tables 1 and 2. The Royal College of Physiians guidelines have been appraised, on behalf of the National Health Servie Exeutive, by St George s Hospital Medial Shool Health Care Evaluation Unit ( CLINICAL PRESENTATION In 2000, we ompleted a Sottish audit of 324 patients with imaging diagnosis of single intraerebral lesion possible glioma. Information regarding presentation was gathered from all ases (tables 3 and 4). Headahe was the most ommon first symptom in patients with single intraerebral tumours (23.5%) and was present in 46.5% at hospital presentation. However, by the time of hospital presentation 86% of patients had other symptoms spread fairly evenly between foal symptoms J Neurol Neurosurg Psyhiatry: first published as /jnnp on 14 May Downloaded from on 19 September 2018 by guest. Proteted by opyright.

2 NEUROLOGY IN PRACTICE Figure 1 Table 1 Metastasis The diagnosti phase Metastasis Time line of patient journey. LGG, low grade glioma; MG, malignant glioma; RT, radiotherapy. Neurosurgial: Royal College of Physiians guidelines Image diagnosis should be made by CT san with ontrast or MRI san (CT or MRI are not aurate preditors of tumour type therefore histologial onfirmation must be atively onsidered and usually undertaken) Performane status should be reorded before surgery for example, Karnofsky Breaking the news of diagnosis News of the diagnosis and information about the prognosis should be imparted in an appropriate environment (a quiet, private plae rather than a publi setting; ideally no more than two health professionals should be present. Most patients want a partner/lose relative there when being told of the diagnosis) After eah onsultation and treatment a brief written summary of speifi rather than general details of what has been dislosed should be inluded in the notes An opportunity should be made for relatives to ask further questions when the full impat of the diagnosis has been absorbed Ensure the patient has a lear enough idea of prognosis to be able to weigh up the merits and drawbaks of treatment Cortiosteroid treatment Treatment entres should develop written guidelines for the use of steroids Patients and their relatives must be given written guidane about steroid mediation and instrutions about the manner in whih the dosage may be altered Steroid dosage should be redued to the lowest amount ompatible with the ontrol of neurologial symptoms and disontinued, if possible, after radiotherapy treatment Neurosurgery Resetion of a malignant glioma is justified to give rapid relief from distressing symptoms (for example, raised ICP) or when the tumour is in a non-eloquent area and the risks of resetion are onsidered aeptable When debulking is arried out omplete marosopi resetion is preferable to partial removal Where biopsy is onsidered in preferene to more extensive resetion, image direted biopsy should be used rather than freehand needle biopsy (histologial reporting should be available if neessary at the time of operation) Histologial reporting should use one of the aepted grading systems (WHO or Daumas Duport) Provide lear written information explaining basi fats about the disease and its treatment, to bak up onsultations and redue the hane of patients reeiving onfliting information about their disease Early postoperative imaging (48272 hours) should determine the extent of tumour resetion Performane status should be reorded in ase notes before disharge Patients should be informed of legal requirement to inform the driving liene authority (DVLA) and this doumented in the notes Support and follow up of patients Treatment/are should be planned in a oordinated fashion Care should be ideally supervised by a single speialist and team, and a key oordinator (for example, nurse) All patients should have aess to onologial advie Patients and relatives must know a main point of ontat for advie/support at all stages of illness The patient s GP should always be informed of the diagnosis before the patient needs to see them The disharge summary should ontain details of arrangements for follow up and management Patient information leaflets should be available in the department, aessible and well displayed Information should be given about further soures of support as provided by national organisations. CT, omputed tomography; GP, general pratitioner; ICP, intraranial pressure; MRI, magneti resonane imaging; WHO, World Health Organization. ii19 J Neurol Neurosurg Psyhiatry: first published as /jnnp on 14 May Downloaded from on 19 September 2018 by guest. Proteted by opyright.

3 ii20 Table 2 Radiation onology: Royal College of Physiians guidelines Radiotherapy treatment Performane status should be reorded before treatment Patients and relatives should take part in the deision about whether or not to have radiotherapy All patients should be offered a suitable wig to ope with hair loss The aeptable and effetive total tumour dose for patients who are fit is 60 Gy in 30 frations Loalised irradiation is preferable to whole brain irradiation Volumes should enompass the presenting enhaning radiologial abnormality with margins 2 3 m The optimum radiation shedule for poor prognosis patients is unknown. A palliative ourse of 30 Gy in 2 weeks and similar palliative regimens are aeptable The time from surgery to start of radiotherapy should be kept to a minimum ideally,4 weeks Supportive management inludes weekly neurologial assessment (this should be doumented) and the development of a are plan for the patient and arer Follow up and advie Care should ideally be supervised by a single speialist and team and a key oordinator (for example, nurse) Patients and relatives must know a main point of ontat for advie/support at all stages of their illness and easy aess to designated linis for neuro-onology patients Rehabilitation, support, and ounselling to patients and their relatives must be offered early in the ourse of disease and treatment Information should be given to patients and relatives about further soures of support Performane status (for example, Karnofsky) and simple measures of quality of life should be reorded at follow up Patients who are treated with steroids alone should be onsidered for referral to a palliative are Information leaflets should be kept available in the department aessible and well displayed The patient s GP should always be informed of the diagnosis, arrangements for follow up and management if possible before the patient has to see them again Patients should be offered the hane to speak with a soial worker about any benefits for whih they may be eligible, and be offered physiotherapy, oupational therapy, and speeh therapy servies Chemotherapy treatment Routine adjuvant hemotherapy at primary diagnosis is not reommended; however, if the patient and/or liniian determine that hemotherapy should be administered the appropriate treatment is either single agent nitrosourea (BCNU or CCNU) or a ombination (for example, PCV therapy) Patients whose disease has relapsed after a reasonable period, and who are not severely disabled, should be onsidered for palliative hemotherapy (for example, BCNU or PCV) with experimental protools Treatment options following reurrene Further treatment after failure of hemotherapy should be onsidered. (Nevertheless it is essential not to raise unrealisti hopes for patients who initially responded well to primary treatment and whose performane sore and pereived quality of life is still high) Seleted patients may be onsidered for further surgery if it will relieve distressing symptoms BCNU, bishloroethylnitrosourea; CCNU, hloroethylylohexlnitrosourea; PCV, proarbazine, CCNU, and vinristine. (hemiparesis, dysphasia, hemisensory symptoms, or diplopia) and more non-foal symptoms (onfusion/memory problems, personality hange, visual symptoms, or unsteadiness). Ten patients presented with headahe as the sole presenting symptom and six of these (60%) had no signs at hospital presentation that is, only six of the 310 patients (1.9%) with intraranial tumours presented with headahe without any other pointers to an intraranial tumour. It is possible that this ould be a higher perentage if patients were seen sooner. Seizure was the next most ommon presenting symptom (21%). The time between first presenting seizure and diagnosis of a tumour was, however,. 6 months in 26% of ases. There are several potential reasons that this ould have happened: patients may not have presented to their GP early Table 3 First presenting symptom (omplete data on 310) First symptom Number (%) (n = 310) Median duration (weeks) beause seizures were omplex partial rather than toniloni and they may not have reognised the importane; GPs may not have reognised the symptoms as being epilepti; 23% of patients presenting with epilepsy had an initial negative omputed tomographi (CT) san before the eventual diagnosti san; of those who remained undiagnosed. 6 months after the first seizure 44% had a negative CT san for tumour (some were normal, some suggested infart, haemorrhage or other pathology); 44% did not have any abnormal neurologial signs at hospital presentation. Foal symptoms (for example, unilateral weakness, numbness or dysphasia or diplopia) were rare presenting symptoms (, 15%) but were frequently present by the time of hospital diagnosis (approximately 80%). When patients presented Interquartile range (weeks) At hospital presentation Headahe 73 (23.5%) (46.5%) Seizure 66 (21.3%) (26.5%) Confusion 14 (4.5%) (30.6%) Personality problem 5 (1.6%) (21.6%) Visual problem 10 (3.2%) (26.1%) Language 18 (5.8%) (35.5%) Unilateral weakness 22 (7.1%) (35.8%) Unilateral numbness 7 (2.3%) (17.1%) Unsteadiness 19 (6.1%) (41.6%) Diplopia 1 (0.3%) (10.0%) Other 75 (24.2%) NEUROLOGY IN PRACTICE J Neurol Neurosurg Psyhiatry: first published as /jnnp on 14 May Downloaded from on 19 September 2018 by guest. Proteted by opyright.

4 NEUROLOGY IN PRACTICE Table 4 Frequeny of signs found at hospital presentation and the frequeny with whih the patient had noted symptoms related to that sign Sign Number (% of total n = 324) with foal symptoms median time to CT sanning was short. The time between first symptom and CT diagnosis was. 6 months in only 6% of patients. Non-foal symptoms (for example, headahe, personality problems, visual problems, and unsteadiness) were rarely the first symptom or were not omplained of early. Patients took longer to present to hospital. At hospital presentation all patients with onfusion, personality problems, visual problems, or unsteadiness had at least one other neurologial symptom (between 40 75% of these had a foal symptom). Time from first symptom to hospital presentation was. 6 months in 20% of patients who presented with visual problems or unsteadiness. Non-foal neurologial symptoms (headahe, personality problems, visual problems, and unsteadiness) may not eliit the same degree of urgeny for referral. Referral usually ourred when additional foal neurologial symptoms developed or a ombination of symptoms (for example, all patients with personality problems as a presenting symptom had headahe in addition at hospital presentation plus other neurologial symptoms in most ases). SIGNS ON CLINICAL PRESENTATION (COMPLETE DATA ON 324 PATIENTS) No signs Sixty two patients (19%) had no signs at the time of linial presentation. In these ases the most ommon first symptom was: seizures (47%), headahe (23%), and a foal symptom (either a language problem, or unilateral numbness or weakness) (11%). Complained of symptom No omplaint of symptom (%) Memory/onfusion 100 (31%) 60 (66%) 31 (34%) 9 Personality problem 33 (10%) 20 (62.5%) 12 (37.5%) 1 Papilloedema 40 (12%) 20 (50%) 20 (50%) 0 Homonymous field defet 37 (11%) 19 (56%) 15 (44%) 3 Unilateral visual 10 (3%) 6 (60%) 4 (40%) 0 Dysphasia 69 (21%) 50 (79%) 13 (21%) 6 Hemiparesis 114 (35%) 83 (81%) 19 (19%) 12 Hemisensory loss 43 (13%) 28 (72%) 11 (28%) 4 Unsteadiness 40 (12%) 30 (83%) 6 (17%) 4 Diplopia 13 (4%) 10 (77%) 3 (23%) 0 Missing data Signs A total of 262 patients (81%) had neurologial signs. The most ommon signs at hospital presentation were: hemiparesis/hemi-sensory loss 82% of these patients had omplained of their unilateral motor or sensory symptoms and 18% were asymptomati but found on examination memory/onfusion 66% of these patients had omplained of these symptoms (memory problems (59%) or onfusion (78%)) a visual sign papilloedema, homonymous field defet and unilateral visual signs; just over 50% were symptomati dysphasia 79% of these patients omplained of this symptom. ACCURACY OF CRANIAL IMAGING Neuroradiologial interpretation of intraerebral tumour is usually aurate. The diagnosti error rate is, 5% (normal, stroke, absess, enephalitis, multiple slerosis). In the Sottish audit, 10% of patients had an initial negative CT san for intraerebral tumour, before the eventual diagnosti san. Of these patients, 22/32 had no abnormality, 6/32 had erebral infartions, 2/32 had erebral haemorrhage, and 2/32 had meningioma. Almost two thirds of negative sans (62%) were done without ontrast. Forty four per ent of patients with late onset epilepsy, whose tumour remained undiagnosed. 6 months after the first seizure, had a previous negative non-ontrast CT san. This suggested that a previous negative CT san probably led to delay in eventual diagnosis. Negative CT sans were more ommon with tumours of the temporal lobe (20% v 10% other lobes). ii21 J Neurol Neurosurg Psyhiatry: first published as /jnnp on 14 May Downloaded from Clinial presentation: key points Patients with malignant glioma ommonly present with headahe or epilepsy as the first symptom Patients with non-foal symptoms ommonly have delayed presentation Patients with non-foal symptoms ommonly had developed foal symptoms or multiple symptoms by the time of hospital appointment Almost 20% of patients with a brain tumour had no abnormal signs at hospital presentation; 50% of these patients had epilepsy Patients with foal symptoms ommonly had foal signs (80%) Visual field defets and papilloedema are ommonly asymptomati; these should be arefully heked for in suspeted ases Cranial imaging: key points A negative CT san does not exlude the possibility of a erebral glioma 10% of patients with presumed glioma will have an initial negative CT san for tumour Negative CT san is more likely if ontrast is not given and there is a temporal lobe tumour All CT sans for presumed brain tumour should be given with ontrast All patients with a definite linial diagnosis of adult onset partial epilepsy who have a negative CT san should have an MRI san with ontrast Imaging is not an aurate preditor of tumour type, supporting the Royal College of Physiians guideline that surgery should always be onsidered and usually performed on 19 September 2018 by guest. Proteted by opyright.

5 ii22 Table 5 Comparison of the positive preditive value (PPV) and auray of best guess diagnosis of malignant glioma, low grade glioma, and metastasis (n = 222) Tumour type PPV Auray Likelihood ratio Malignant glioma Low grade Metastasis No patient with a posterior fossa tumour had a negative CT san for intraerebral tumour. More errors start appearing when the neuroradiologist is asked whih type of tumour it is. It an be very diffiult distinguishing malignant glioma from metastasis (or metastases) and from primary entral nervous system (CNS) lymphoma. In one study of patients with known systemi aner with a single intraranial lesion thought, by experiened neuroradiologists to be a metastasis, 11% turned out to be inorret. 4 An important linial question is whether radiologists are aurate enough in prediting the diagnosis of malignant glioma to obviate the need for a biopsy. Radiotherapists are unlikely to irradiate patients without histologial onfirmation. To examine the preditive value and diagnosti auray of CT sans the best linial guess was ompared with the eventual histology. San reports were prospetively ategorised malignant glioma, low grade glioma, and metastasis ; 222 patients fell into these ategories (table 5). Positive preditive values and auray are shown in table 5. Others have found that approximately 45% of patients with imaging diagnosis of a low grade glioma atually have a higher grade of malignany. 5 Interpretation of CT/magneti resonane image (MRI) sans is not an aurate preditor of the histologial diagnosis. The treatment for malignant glioma, low grade glioma, and metastasis are different. This supports the Royal College of Physiians guideline that histologial onfirmation should be sought in most ases. PRE-SURGICAL WORK UP A good history, general examination, and hest x ray/ct are indiated in ases where the neuroradiologist may suspet single metastasis from unknown primary. A slit lamp examination of the eyes to look for vitreous ells, lumbar punture (where there is no signifiant intraerebral mass), hest and abdomen CT, and HIV test are usually all that are required as a work up when primary CNS lymphoma is suggested. Where germ ell tumours are in the differential diagnosis (espeially pineal region tumours), erebrospinal fluid (CSF) tumour markers an be helpful. Malignant teratoma, germinoma with synytioblasti ells, embryonal arinoma, and endodermal sinus tumours may have elevated levels of alphafetoprotein. Chorioarinoma and embryonal arinoma may have elevated levels of b-hcg, whereas only germinomas and germinomas with synytiotrophoblasti ells have elevated levels of plaental alkaline phosphatase. Non-germ ell tumours are negative for all these markers. If the initial investigation is a CT san, many surgeons prefer a preoperative MRI to exlude undislosed multiple NEUROLOGY IN PRACTICE lesions whih might suggest metastases (or multifoal glioma) if this would influene the deision about surgery. SURGICAL COMPLICATIONS Postoperative haematomas at the operative site our in approximately 5% of patients. A transient neurologial defiit will our in approximately 10% of patients postoperatively. Morbidity is less with stereotati operations. 6 There is reovery of the neurologial defiit in approximately 50% of ases. There is also a risk of seizures as a result of operation in those who have no prior history of seizures and sometimes a flurry of seizures in the postoperative period. Intraoperative stroke will our in, 1% (depending on seletion). Perioperative mortality from raniotomies for malignant glioma or metastasis approah 5%, and for stereotati neurosurgery are approximately 1% (depending on the seletion of patients and the experiene of the surgeon). 7 DELAYS IN PROCESS OF CARE Referral guidelines for GPs highlighting red flags to attempt to identify patients at an earlier stage may be helpful, but are likely to be non-speifi. Open aess imaging for GPs is very likely to speed up referral in many ases. One the tumour has been identified there is little delay until surgial review. In ases seleted for operation, there is usually little in the way of delay. Patients with headahe or foal neurologial signs are usually started on steroids and planned surgery performed eletively within a few days. At some stage, either as an inpatient or outpatient, the patient is seen by the radiation onologist with the formal results of the pathology and imaging. We found that 87.5% of patients had not started radiation therapy within four weeks of their initial CT san. The Royal College of Physiians guidelines state that patients suitable for treatment should have started radiation therapy within four weeks of diagnosis. The Joint Counil for Clinial Onology reommend that waiting time targets from the date of first onology onsultation to the start of radiotherapy be two weeks (with a maximum aeptable of four weeks) for radial radiotherapy involving omplex treatment planning. 8 It is lear that these targets are not being ahieved. The failure of lear lines of responsibility/transfer of are is an important reason for delays (doumentation in the notes, ommuniation with patients, retrieving investigations, booking proedures for example, mould room, radiotherapy and appointments). Hospital and surgery-radiotherapy delays ould be improved by redesigning and improving loal pathways of are/proess planning. The seond part of the delay will depend on the number of patients with different forms of aner waiting to use the same radiation therapy mahine in the treatment entre. Improvements in this delay an only be ahieved by inreasing the number of mahines for treatment with the assoiated resoure impliations for hardware and maintenane or by redesigning the sheduling of radiotherapy to maximise the utilisation of the existing treatment mahines. In Edinburgh (Edinburgh Centre for Neuro-onology (ECNO)), a new model of are has been developed where liniians from different speialties (neurology, neurosurgery, and onology) use the same lerial, seretarial, and administrative resoure. The lynhpin is the neuro-onology speialist nurse. In the past speialist nurses were attahed to departments (for example, onology) and have only been involved in are for patients one referred to that department. J Neurol Neurosurg Psyhiatry: first published as /jnnp on 14 May Downloaded from on 19 September 2018 by guest. Proteted by opyright.

6 NEUROLOGY IN PRACTICE From our audit it was lear that only about 50% of patients with suspeted primary brain tumour reeive radiation therapy. The neuro-onology speialist nurse sees patients whom they are admitted to the neuro-sienes ward and bridges the speialties and follows the patient through their pathway of are and ensures horizontal integration of are and onsistent ommuniation and advie. She is the foal point for patients to ontat and triages problems that she is unertain about. This redues the risk of poor ommuniation and speeds up the proess of are. In addition there are joint neuro-onology linis and neuropathology review meetings; a single neuro-onology reord instead of neurosienes and onology ase reords; and a entral point for referral, results, information, advie, audit, and researh. COMPLICATIONS OF TREATMENT Radiation therapy The toxi side effets of ranial radiation an be divided into loal effets and CNS effets. The side effets of radiation therapy depend on the dose frationation shedule used, the natural history of the underlying disease, and the likelihood of obtaining a radiotherapeuti response. Some people will feel nauseated about 30 minutes to one hour after treatments. Hair loss starts about 2 3 weeks into treatment and maximum re-growth has ourred by six months. Most people feel tired and sleepy at the end of a ourse of radiation and some feel sik. Some years after ranial irradiation where the pituitary/hypothalamus is in the radiation field, there may be further loal neuro-endorine ompliations. Radiation usually affets the prolatin and sex hormones first (prolatin rises, leuteinising hormone (LH) and follile stimulating hormone (FSH) fall) and auses problems with periods or infertility, then the thyroid stimulating hormone (TSH) falls and produes seondary hypothyroidism. If the opti nerve is in the treatment field, one ommonly finds an afferent pupillary defet with opti neuropathy whih is usually asymptomati or only produes mild visual auity disturbane. Years after temporal lobe or posterior fossa irradiation one may find mild sensorineural hearing loss. The most serious CNS ompliations of radiation to the nervous system are: aute enephalopathy; subaute (early delayed) demyelination; delayed radiation nerosis; and hroni leuoenephalopathy. Aute enephalopathy is rare, but omes on usually within 24 hours of ranial irradiation. Early delayed reation is ommon and is seen four weeks to three months after ompletion of ranial radiation. In patients with erebral tumours the symptoms are indistinguishable from those of tumour progression. The treatment is to re-institute steroids for a period of 4 8 weeks and then gradually redue them. Patients should be made aware of possible early delayed effets, to allay the fear of early reurrene, and should have a target direted plan for reovery whih inludes their own programme for rehabilitation and for periods of rest. Delayed erebral radiation nerosis is infrequent and an start months or years after erebral irradiation. The linial and radiologial appearanes mimi tumour reurrene. MRI annot adequately distinguish ative tumour from radiation nerosis. Positron emission tomography/single photon emission omputed tomography (PET/SPECT) an help differentiate radiation nerosis from ative tumour. Chroni leuoenephalopathy is usually only found in long term survivors of ranial irradiation. 9 Relatives notie that the patient may lak motivation, there is psyhomotor retardation, memory impairment, and ataxia or apraxia of gait. As time passes there may be urinary inontinene, pronouned dementia, inability to walk due to apraxia or ataxia, and ortial myolonus. High dose and large frationation shedules are thought to be assoiated with a higher inidene of radiation indued leuoenephalopathy. The CT/MRI san shows diffuse white matter hanges in the ortial white matter and ventriular dilatation and ortial atrophy. Chemotherapy Nitrosoureas (for example, BCNU) ause nausea that starts about two hours after starting an infusion that may persist for hours. Faial flushing or dizziness may our during infusion whih is rate dependent and resolves on stopping the drug. Bone marrow suppression ours with almost all agents and is maximal 4 6 weeks after reeiving the drug and usually settles by eight weeks. Risks of infetion, bleeding, and tiredness are greatest around four weeks after treatment. Lung toxiity usually starts after a total dose of 1 g. A restritive ventilatory defet is found and it is valuable to monitor vital apaity regularly in patients who reeive. 1 g total dose. Renal funtion should also be monitored. Proarbazine auses haematologial and gastrointestinal symptoms, but in addition an ause flu like symptoms, rash, and neurologial symptoms (ataxia, headahes, paraesthesia, dizziness). Proarbazine an ause hypertensive risis and severe gastrointestinal symptoms if given with antidepressants, alohol, or tyramine rih foods (for example, heese, bananas). Vinristine auses neurotoxiity (neuropathy, myopathy, and autonomi disturbane), gastrointestinal symptoms, and sometimes alopeia. Haematologial toxiity is mild. Cisplatinum derivatives ause neurotoxiity (peripheral neuropathy and deafness), renal toxiity, and bone marrow suppression. Temozolomide an ause marrow suppression but the side effet profile is better than oral proarbazine. High dose ytosine arabinoside and 5 fluorourail an ause reversible erebellar ataxia, in addition to bone marrow suppression and gastrointestinal and liver toxiity. Alopeia and infertility an our with any drugs. Coming to terms with diagnosis an be eased by aurate, understandable, medial information about the disease and its treatment options. This is best done by a dotor/speialist nurse experiened in managing patients with brain tumours. REFERENCES 1 Counsell CE, Grant R. Inidene studies of primary and seondary intraranial tumours: a systemati review of their methodology and results. J Neuro-Onol 1998;37: Davies E, Hopkins A on behalf of a working group. Good pratie in the management of adults with malignant erebral glioma: linial guidelines. Br J Neurosurg 1997;11: Davies E, Hopkins A, eds. Improving are for patients with malignant erebral glioma. London: RCP Publiations, Pathell RA, Tibbs PA, Walsh JW, et al. A randomized trial of surgery in the treatment of single metastases to the brain. N Engl J Med 1990;322: Kondziolka D, Lunsford LD, Martinez AJ. Unreliability of ontemporary neurodiagnosti imaging in evaluating suspeted adult supratentorial (low grade) astroytoma. J Neurosurg 1993;79: Clyde Z, Chataway J, Signorini D, et al. Signifiant hange in tests of neurologial impairment in patients with brain tumours. J Neuro-Onol 1998;39: Cabantog AM, Bernstein M. Compliations of first raniotomy for intra-axial brain tumour. Can J Neurol Si 1994;21: Royal College of Physiians of London. Reduing delays in aner treatment. London: Royal College of Physiians, DeAngelis LM, Delattre JY, Posner JB. Radiation indued dementia in patients ured of brain metastases. Neurology 1989;39: ii23 J Neurol Neurosurg Psyhiatry: first published as /jnnp on 14 May Downloaded from on 19 September 2018 by guest. Proteted by opyright.