SYMPTOMS

Headaches are present in 50% of patients at hospital presentation, although less than 10% of all patients have the “classical” headache of raised intracranial pressure.¹ Patients with headaches are more likely to have larger tumors.² The brain parenchyma does not contain pain fibers. Blood vessels, dura, and choroid plexus do have sensory nerve endings, and pain from stretching of these structures causes vascular or referred pain. In general, patients with supratentorial tumors have headaches referred frontally (cranial nerve V), and those with tumors involving the posterior fossa have headaches referred to the occipitocervical region (cranial nerves IX and X).

Sadly, the headaches associated with tumors have very few good discriminating features. Papilledema is found at some stage in more than 50% of patients with headache; however, less than 15% are identified as having papilledema at first presentation.^1,3 Some patients with midline shift do not have any headache.² The headache of tonsillar herniation is maximal in the neck and occiput and may be associated with nuchal rigidity and painful extensor spasms of the spine and limbs. These may mimic a generalized tonic seizure, but are not associated with loss of consciousness. Seizures are the first symptom of an intracerebral tumor in 21% of patients.Their frequency increases to 26.5% by hospital presentation, and approximately 50% of patients seize at some stage during the illness.⁴

By the time of hospital presentation, 81% of patients have some neurological signs on examination. The most common signs are unilateral focal motor or sensory signs involving the limbs; dysphasia; or impairment of memory and/or cognition. Neurological impairments are more common with increasing age, and older patients are more likely to have multiple and more severe impairments.⁵ Some authors have identified subtle cognitive problems in as many as 91% of patients with brain tumors before surgery.⁶ Anxiety is present in 17% to 30% prior to surgical operation; this may be more common in patients with right hemisphere tumors.^7–9 Clinical depression reaching the DSM-IV criteria for major depressive disorder occurs in 19% to 38% of patients with a brain tumor and has been found to be the single most important independent predictor of quality of life.¹⁰ Depression is most common in women and those with a past or family history of psychiatric disorder.^7,11 Patients with left hemisphere tumors reported significantly more memory problems and depressive symptoms.¹² If the depression is severe, suicide is a real risk.

MANAGEMENT OF SYMPTOMS

Steroids have been shown to improve symptoms and neurological impairments and to reduce cerebral edema seen on imaging. In many patients with primary CNS lymphoma, and occasional patients with malignant glioma, the tumor can disappear on imaging. Steroids are not indicated in patients without significant symptoms or without cerebral edema seen on imaging, since the side effects of steroids are likely to outweigh any benefit.¹³ Patients with focal neurological deficits, and without symptoms or signs of raised intracranial pressure, respond well to dexamethasone 4 mg/day. This dose is as effective as 8 mg/day or 16 mg/day and has fewer side effects.¹⁴ If patients have symptomatic raised intracranial pressure or a significant shift seen on imaging studies, or in cases where herniation is suspected, an initial dose of 12 to 16 mg intravenously may be followed by 16 mg orally each day, until the tumor has been decompressed surgically or until a maximum benefit has been achieved. The dose should then be titrated downwards to the minimum effective dose. Steroids should be avoided after 6 pm because of the common side effect of insomnia. A clinical response to steroids was found in 37% of patients with a primary brain tumor at the time of diagnosis, 40% of patients with symptoms during radiotherapy, and 6% of patients after completion of radiotherapy.¹³ The frequency of improvement in symptoms was similar in patients with brain metastases.

In patients with symptoms suggesting impending clinical herniation, who do not respond to steroids, fluid restriction and mannitol 0.5 to 1.0 g/kg (with the higher doses being effective up to 6 hours) to produce an osmotic diuresis and decrease the formation of CSF by removal of sodium and water from the brain will commonly be effective. This is a temporary solution which should be followed by urgent surgical tumor resection or cyst drainage.

There is an uncommon phenomenon sometimes found in patients with large mass lesions, in which paroxysmal neurological symptoms are triggered by standing. The attacks may occur despite dexamethasone therapy, but the addition of acetazolomide will often promptly stop symptoms.¹⁵ Patients with dysphasia, dysarthria and dysphagia and dysphonia should be assessed by a speech therapist and a swallowing assessment should be performed where indicated. Patients with hemiparesis or gait problems should be assessed by a physiotherapist and given advice regarding gait, mobility, and prevention of deep vein thrombosis. Resective surgery will lead to further improvements in many patients, often after an early postoperative worsening. Postoperative neurological deterioration occurs in 30% of patients undergoing resection and 36% who were biopsied.⁵ The postoperative nonneurological complication rate is highest for complete (9.7%) or partial resection (8%), and lowest for stereotactic biopsy (3.8%).

Where imaging has revealed an intracranial tumor and the patient presents with seizures, the patient should receive the advice of the neuro-oncology multidisciplinary team about seizure control and treatment of the tumor. The seizure threshold will be lowered by poor sleep, irregular diet, anxiety, noncompliance with medications, and the use of certain medications (e.g., antipsychotics, antidepressants, or alcohol). It is essential to discuss risk-avoidance strategies with the patient, both for work (i.e., dangerous machinery, heights) and leisure (i.e., swimming or bathing unsupervised). Issues such as: driving, oral contraceptives, pregnancy, education, and management of tonic-clonic seizures must be discussed. Relatives must be informed about first aid measures. The legislation regarding work and driving will vary from country to country and in the US, from state to state. UK regulations can be found on www.dvla.gov.uk.¹⁶

Tonic-clonic seizures are most likely to become completely controlled with medication, whereas focal seizures are only completely controlled by a single anticonvulsant in 30% to 40%, with an additional 30% to 40% achieving a 50% reduction in seizure frequency. Any of several anticonvulsants can be used to control seizures. Anticonvulsants have not been compared head to head in tumor-associated epilepsy. The anticonvulsant used will depend on the urgency for treatment (e.g., status epilepticus), possible interactions with other drugs likely to be used in management of the tumor and cost-effectiveness issues. Carbamazepine, phenytoin, and phenobarbital are enzyme-inducing agents whereas valproate, lamotrigine, levetiracetam, and topiramate are not. The newer anticonvulsants may be better tolerated, have fewer drug interactions, and may have fewer cognitive side effects, but they are more expensive.^17,18 Guidelines are available on the efficacy and tolerability of the newer antiepileptic drugs.¹⁹ These recommend initial use of the older agents unless there is a satisfactory reason for not using them. Certain antiepileptic agents may alter the bioavailability of chemotherapy (see later). Whether this has any significant effect on survival or outcome is still uncertain. Acute seizure management is crucial, because prolonged seizures cause neurochemical changes occur that lead to neuronal damage. Evidence from a trial comparing four treatment regimes for the management of status epilepticus has demonstrated that intravenous injection of lorazepam 0.1 mg/kg, at a rate of less than 2 mg/min, is effective in the initial treatment in 65% of patients. Phenobarbital (15 mg/kg) was effective in 58%, diazepam (0.15 mg/kg) followed by phenytoin 18 mg/kg was effective in 56%, and phenytoin alone in 44% of cases. The American Academy of Neurology Guidelines for the management ofstatus epilepticus have been summarized in Table 12-1.

TABLE 12-1 Management of Status Epilepticus

Surgical resection of some low-grade neoplasms (gangliogliomas, pilocytic astrocytomas, dysembryoplastic neuroepithelial tumors) can cure epilepsy or reduce seizure frequency, particularly if the epileptogenic area, as determined by noninvasive recordings, is resected in addition to the tumor.²⁰ Fractionated radiotherapy and radiosurgery can reduce seizure frequency to Engel classification I or II in 54% of patients with brain tumors and intractable epilepsy.²¹ As with tumor-associated epilepsy surgery, temporal lobe tumors had a better success rate than extratemporal tumors. Temozolomide may reduce the seizure frequency in intractable epilepsy.²²

There does not seem to be a difference in how patients with brain tumors cope with their illness as compared to patients with other brain diseases such as stroke and Parkinson disease.²³ Anxiety or depression may occur due to uncertainty about symptoms, likelihood of control of symptoms, and worries about the immediate future, including death, or family and financial matters. All of these affect quality of life. Psychological morbidity is related to physical and neuropsychological functioning. Psychological morbidity is associated with high levels of physical disability and also with cognitive dysfunction, but is not related to the grade of tumor or the extent to which the patient was aware of the nature of his or her disease. In one prospective study, 5% had clinically significant levels of anxiety, and six (15%) had clinically significant levels of depression, yet 92% felt they had full or intermediate knowledge about their prognosis.

A variety of psychological approaches have been used, but cognitive-behavior therapy (CBT) has been most researched in people with cancer. Treatment with antidepressants has been shown to be more effective than either placebo or no treatment in patients with physical illness.²⁴ However, there are no randomized controlled trials of the use of antidepressants in patients with brain tumors. Management guidelines for the treatment of depression in those who are also physically ill have been published.²⁵ Short-term, highly focused forms of psychotherapy such as cognitive-behavioral, supportive, and group therapies are helpful but time-consuming. They may be the only forms of treatment available in patients disinclined to accept antidepressants or who are intolerant of them. CBT is as effective as antidepressants. By 3 to 8 months, 50% to 60% of patients are in remission, compared with only 27% treated with placebo.^26,27

Depression should be and can be successfully treated.²⁸ The importance of this is supported by the observation that the presence of depression impacts negatively on both psychological and physical quality-of-life outcomes in patients with brain tumors.²⁹ Neuropsychiatrists may be able to help identify whether sedation, confusion, and possible lowered seizure threshold are related to the disease or to antidepressant or antipsychotic medication.³⁰ There are no randomized controlled trials of different treatment approaches in the neuropsychiatric management of patients with brain or CNS tumors. The advice, therefore, is at the level of recommendations for best practice from a consensus group of experts. In the “Glioma Outcome” study (2004) that included 598 patients diagnosed with a glioma,³¹ concordance between physician recognition of depression and treatment of depression was initially low (33%), but increased at 3 and 6 months (51%and 60%, respectively). Antidepressants may have an adverse effect on seizure control, although this is rarely a serious complication in my experience, and it has not been found in recent prospective studies.³² Antidepressants should be prescribed when necessary, as depression is a major cause of poor quality of life in patients with epilepsy.³³ Selective serotonin-reuptake inhibitors have largely replaced tricyclic antidepressants in the management of somatic psychiatric conditions, but good comparative treatment studies in patients with brain tumors are not available.

PROPHYLACTIC ANTICONVULSANTS

Although it has been common practice in the USA and parts of Europe to prescribe prophylactic anticonvulsants prior to neurosurgery for malignant brain tumors, there is now evidence from a meta-analysis of randomized controlled trials showing that anticonvulsant prophylaxis does not reduce the seizure incidence (OR 1.09; 95% CI, 0.63 to 1.89; p=0.8), seizure-free survival (OR 1.03; 95% CI, 0.74 to 1.44; p=0.9), or overall survival (OR 0.93; 95% CI, 0.65 to 1.32; p=0.7) in patients with brain tumors who have not had an epileptic seizure.³⁴ More recent studies, in which patients were randomized either to no treatment or to treatment with carbamazepine or phenytoin for 6 or 24 months, showed no significant differences in any group. A high incidence of drug-related side effects was found in this and other studies.^35–37 Anticonvulsant prophylaxis in patients with newly diagnosed brain tumors is therefore not justified.

PROPHYLACTIC ANTICOAGULATION

Cancer, immobility, hemiplegia, and surgery are all risk factors for deep vein thrombosis and pulmonary embolus. Risk of DVT and/or pulmonary embolus is age-related, with an annual incidence of less than 1:3000 in patients under the age of 40 and 1:500 in those over the age of 80.³⁸ The frequency of DVT in malignant glioma is between 19% and 28%, and the risk that a patient with a brain tumor will display a DVT in follow-up studies is between 22% and 45%.³⁹ The frequency of DVT is highest in meningioma (72%), followed by glioma (60%) and metastasis (20%).^40,41 Despite the very high likelihood of thrombotic complications in patients with malignant glioma, there is no consensus about DVT prophylaxis.⁴² The risk of DVT is reduced by use of compression stockings, pneumatic intermittent compression boots, early mobilization, low-dose heparin or fractionated heparin. A meta-analysis of four randomized controlled trials of heparin prophylaxis in neurosurgical patients has shown that with placebo, 29% of people develop DVT, compared with 16% of those receiving low molecular weight heparin or unfractionated heparin.⁴³ A critical appraisal of the literature on prophylactic anticoagulation therapy in neurosurgery concluded that treatment with heparin resulted in a 45% relative risk reduction of venous thromboembolism (OR 0.48; 95% CI 0.35-0.66; p<0.001).⁴⁴ Major bleeding complications were infrequent, but heparin (both UFH and LMWH) resulted in a 71% increased relative risk of major bleeding events (OR 1.71; 95% CI 0.69-4.27; p=0.24). A randomized controlled trial of heparin 5000 U subcutaneously, starting 2 hours before surgery and continuing every 12 hours thereafter until full ambulation or for 7 days, in 103 patients undergoing surgery for removal of a supratentorial tumor, showed no increase in bleeding tendency in any of the parameters examined compared with the placebo arm.⁴⁵ There is no evidence to suggest that enoxaparin 40 mg/day is superior to unfractionated heparin. In the randomized controlled trial comparing these agents as prophylaxis, none of the 150 participants in the study suffered symptomatic DVT and less than 10% had asymptomatic DVT (mostly calf) using the combination of graduated compression stockings, intermittent pneumatic compression, and either enoxaparin or subcutaneous heparin 5000 U twice daily.⁴⁶ A multimodality approach to prophylaxis, using compression stockings, pneumatic intermittent compression boots, and prophylactic perioperative minidose heparin during surgery is safe and is recommended.

DEEP VEIN THROMBOSIS AND PULMONARY EMBOLUS

Deep vein thrombosis (DVT) may be difficult to diagnose. Although DVT may cause unilateral calf or thigh swelling, pain and tenderness along the line of the deep veins, low-grade pyrexia, distension of superficial veins, or color change, many cases are asymptomatic. Homan sign is poorly predictive. There is a 10% rate of symptomatic pulmonary embolus in untreated proximal DVT, and 18% to 30% mortality if this is left untreated.⁴⁷ Pulmonary embolus is rare if the DVT is treated by anticoagulation therapy.⁴⁸ DVT can be complicated by critical limb ischemia (less than 5%), recurrent DVT (20% in 5 years), or postthrombotic syndrome (50% to 75%). If clinically suspected, intravenous heparin should be started until the diagnosis is excluded by diagnostic imaging.⁴⁹ as the benefits of anticoagulation outweigh the risks of tumor hemorrhage. Compression ultrasound will reliably identify proximal DVT, but contrast venography is the gold standard if ultrasound is negative.