Gliomas arise from the glial (supporting) cells of the brain. They are not derived from the neurones (nerve cells) themselves. These are the commonest primary brain tumours.
There are three types of glioma: the commonest is the astrocytoma, followed by the ependymoma and the oligodendroglioma, each one gaining its name from the cell type from which it arose.
Gliomas are graded by their histological appearances (appearances down the microscope) and nowadays also by their genomics ( a term which is used to describe the genetic mutations exhibited by the tumour – vide infra). For example, the oligodendroglioma usually has the 1p19q deletion in its genome and this has therapeutic implications – vide infra.
Grade 1 and 2 gliomas are relatively slow growing and may have a long natural history (often presenting to the neurologist with epileptic fits), so much so that sometimes they are watched carefully by neurologists without proceeding to active therapy.
High grade tumours: grade 3 and grade 4 gliomas (grade 4 also known as glioblastoma multiforme) are much more malevolent in the short term and require active management. The genomics of these is also now an aspect of importance to the Oncologist.
The diagnosis of a higher grade glioma may have been suggested on the MR scan as these faster growing tumours may outstrip their blood supply and this leads to a highly typical necrotic (dead) centre to the tumour as seen on scan, with an enhancig rim on the MRI. The biopsy or decompression specimen confirms the diagnosis. Gliomas are dangerous because of uncontrolled growth within the brain; they do not (or hardly ever) metastasise outside the brain.
Meningiomas arise from the coverings of the brain; the cell of origin is the arachnoid fibroblast. They are usually benign tumours that are cured by a radical surgical excision (or stereotactic radiation therapy e.g. gamma knife technology) in most instances.
Pituitary tumours are of interest in that they often secrete a hormone product (the pituitary being the master controller of the endocrine/hormone system in the body): prolactin by prolactinomas, growth hormone by the acidophil adenoma that causes acromegaly and gigantism, and adrenocorticotrophin by the basophil adenoma that causes Cushing’s disease. Furthermore, they arise just under the visual pathways, where the optic nerves pass backwards from the eyes into the brain , and by pressure exerted by their growth upwards can disturb vision (particularly vision out to the sides), and this can be the presenting symptom, particularly in the cases where there is no hormonal over-secretion.
Like meningiomas, pituitary tumours are almost invariably benign adenomas.
Craniopharyngioma is a benign overgrowth of embryonic remnants that occurs in the pituitary area, usually just above the pituitary itself; it declares itself by pressure effects e.g. on the visual pathways and with visual loss, or, in children particularly with growth disturbances due to pituitary dysfunction.
Acoustic/vestibular neuroma is a benign tumour of the sheath of usually the vestibular nerve. The vestibular nerve subserves the function of balance and runs from the brainstem to the inner ear accompanied by the highly sensitive auditory nerve which subserves hearing. The pressure of the tumour against the acoustic nerve is enough to cause gradual hearing loss and this is the method of presentation in the majority of cases.
Other cranial nerves may develop neuromas but there is a poorly understood predilection for the disease to affect the vestibular nerve and patients who have the condition neurofibromatosis are at high risk. The MR image is usually fairly characteristic of an acoustic/vestibular neuroma, although it may occasionally be mistaken on scan for a meningioma or other growths.
Other brain tumour types
This is a tumour that occurs in the cerebellum of younger patients (children more than adults) and infiltrates rather like a glioma. Additionally, it has a predisposition to shed off cancer cells into the fluid (cerebrospinal fluid or CSF) which surrounds the brain and this may lead to spread within the CSF and seed implantation and growth of metastatic disease at other sites within the brain or spinal cord. Similar tumours called: Primitive NeurEctodermal Tumours (PNET) arise in other sites in the brain and are treated as are medulloblastomas.
The treatment is surgical resection followed by radiotherapy to the whole neuraxis (the brain and all its coverings, together with the spinal cord and all its coverings). Furthermore, this tumour has been shown to be sensitive to chemotherapy and particularly to platinum based therapy. Therefore most current protocols have a chemotherapy component as well.
Primitive neuroectodermal tumour (PNET)
This is a homologous tumour to the cerebellar medulloblastoma down the microscope, but occurs in the cerebral hemispheres. It also has a predisposition to spread via the CSF to other neuraxis sites and it is basically treated as for medulloblastoma viz. surgical resection followed by neuraxis radiotherapy and chemotherapy. It should be noted that ependymomas (one type of glioma vide supra) can also spread via the CSF and neuraxis radiotherapy may also be indicated here in some cases (for example high grade ependymomas of the posterior fossa).
Intracranial germ cell tumours
These are homologous with the germ cell tumours of the testis or ovary, but are malignant and need to be treated seriously (in the modern era they are usually curable). The most common site of origin is in the pineal region followed by the region above the pituitary gland (the supra-sellar region). Others can occur at other midline sites e.g. the fourth ventricle.
These tumours present due to expansive growth and compressive neurological symptoms. The diagnosis is made by biopsy or tumour markers (blood or CSF: HCG or AFP – see testicular teratoma section for explanation of these marker terms).The therapy of intracranial germ cell tumours is by chemotherapy and neuraxis radiotherapy for they too have a high tendency to spread via the CSF to other sites within the neuraxis but rarely further afield.
These tumours are exquisitely sensitive to chemotherapy based on cis-platinum and shrink quickly and dramatically with this therapy. Treatment usually commences with chemotherapy and radiotherapy follows completion of the chemotherapy programme. Surgery is usually restricted to a biopsy (if there is doubt over the diagnosis) or ‘shunt’ to overcome hydrocephalus at presentation; there may be a role for removal of a residual lump at the end of therapy, but this necessity is unusual.
The effects of chemotherapy are so profound that many are now reducing the subsequent radiotherapy dosages, although most have not yet been convinced that the cure of these tumours can be routine without radiotherapy. Tumours which are not secreting markers HCG or AFP tend to have the best outlook for cure, perhaps in excess of 80%, whereas those secreting large quantities of AFP certainly have a lesser chance of cure.
These are high grade B cell lymphomas of brain and occur spontaneously but are predisposed to in immunosuppressed patients (e.g. renal transplant recipients, AID patients etc) more frequently than the rest of the population. They present as do gliomas with which they are often confused on brain MRI scans.
The diagnosis is made by biopsy and therapy is then commenced with chemotherapy followed by radiotherapy – although the role of radiotherapy is being questioned in the age of high dose methotrexate based chemotherapy. The tumour is highly responsive to this therapy and survivals in excess of 50% are achieved in patients without spread beyond the primary site at presentation.
These are vascular malformations within the brain and are not really brain tumours at all. However, they are of some importance as they tend to bleed and are the commonest cause of strokes in young patients. Furthermore the component blood vessels supply no blood to normal brain (i.e. they are redundant) and so surgical removal or obliteration by stereotactic radio surgery (vide supra) effects cure without depriving normal brain of oxygenated blood.
Up to 20% of all ostensibly primary brain tumours turn out to be metastases from a primary cancer outside the central nervous system. In particular primary breast, lung, renal and adrenal carcinoma and melanoma have a predisposition to spread to the brain; other primary cancers may also do so.
Exposure to ionising radiation predisposes to the later development of brain tumours in a small percentage of cases.
Whilst there are some rare genetic syndromes that predispose to brain tumours, the overwhelming majority of primary brain tumours are of a sporadic nature and, at present, the underlying cause is unknown.
The much publicised suggested risks of mobile phone use and habitat in proximity to ‘power lines’ have neither of them been shown (in well conducted epidemiological studies) to be related to an increased incidence of brain tumours.
For the vast majority of brain tumour cases, there is no predisposing cause known and no known familial risk.
The primary tumours of the brain/central nervous system (CNS) i.e. those that originate within the brain/spinal cord, are a mixed group of tumours, some of which are benign and some are malignant, the latter carrying a high risk of death. Following some general remarks, each group will be discussed in turn.
The incidence of tumours of the central nervous system is approximately 15 per 100,000 and they account for approximately 2% of all deaths in Western society. Curiously, there is one age peak of incidence in youth (in the age range 5-10 years) and a second one in the sixth decade of life. In general, the incidence of brain tumours occurs throughout life rather than being concentrated in the elderly as for many carcinomas. For some tumours (such as germ cell tumours and medulloblastoma) there is a male preponderance of incidence, whereas for some others (e.g. meningioma) there is a female predisposition.
Of tumours presenting to a neurosurgical service in the U.K., and apparently as primary brain tumours, the breakdown by tumour types is as follows: glioma 45%, meningioma 15%, acoustic neuroma 8%, pituitary adenoma and craniopharyngioma 8%.
Metastases (i.e. secondary spread of cancer which started outside the brain – but here masquerading as primary brain tumours) appear as high as 15%-20% of ‘apparently’ primary brain tumours (i.e. at time of presentation: no evidence of a primary outside the brain even though this will ultimately be found).
Other rarer forms of primary brain tumours account for up to 10% of the total.
In the community at large, the incidence of metastatic brain disease is higher (the point here being that many cases of brain metastases are managed outside a neurosurgical unit).
Symptoms & diagnosis: Brain tumours
The commonest clinical presenting features of a patient presenting with a brain tumour are:
- Symptoms or signs of raised pressure inside the head: these are headache, vomiting and pressure signs at the backs of the eyes (papilloedema) when the doctor looks with his ophthalmoscope.
- Progressive loss of neurological function attributable to a progressive defect in one brain area e.g. the progressive loss of strength in one limb, progressive difficulty in speaking etc.
- Late onset epilepsy – either focal fits (e.g. fitting of one limb) or generalised fitting (e.g. full blown fitting with loss of consciousness, called grand mal fitting).
The reasons for these three manifestations being the common mode of presentation of brain tumours is easy to understand when we think of what is going on inside the head. As a tumour grows, the pressure goes up inside the head as this is a closed compartment – the skull is rigid and cannot allow any expansion of brain containing a swelling tumour; therefore the pressure goes up and headache and vomiting are exhibited by the patient.
When a tumour arises and grows in a particular brain area, it destroys the function of that brain area and this is why progressive loss of neurological function (which is subserved by that brain area) is noted as the condition progresses.
Lastly, any tumour growing within the brain acts as an irritable focus, disturbing the smoothly established electrical neuronal circuits; it is for this reason that epileptic fits may be manifest and be the first symptom to draw attention to the tumour.
Other special types of brain tumour manifest in their own individual ways: Thus an acoustic neuroma will manifest with progressive hearing loss on the affected side and a functioning pituitary tumour may manifest because of the over-secretion of its hormone product e.g. acromegaly when it is an acidophil adenoma (vide infra).
The first test ordered by the doctor is an MRI (magnetic resonance image) of the brain. This is the best form of brain imaging available today and demonstrates the tumour in the majority of cases, as well as often giving a clue as to the tumour type.
If there is a high chance that the tumour is actually a secondary tumour that has spread from a cancer that started outside the central nervous system, the doctor will search for possible primary cancer in the lungs (chest x-ray), breasts (palpation and mammography) etc. before recommending surgical attack or biopsy of the brain tumour itself. Even so, sometimes the answer on pathological exam of the surgical specimen still demonstrates that the tumour started outside the brain – ‘occult primary’.
The certain diagnosis of a brain tumour is established for sure by surgical biopsy.
Sometimes, the biopsy material is obtained and the initial treatment given all at the same operation, when the surgeon operates to remove as much tumour as possible and also sends this resected material to the laboratory for microscopic analysis. It is the microscopic of material from the tumour that gives the certain diagnosis as to exactly what kind of tumour it is.
In pituitary tumours that over-secrete hormone products and intracranial germ cell tumours that over-secrete HCG and/or AFP (see testicular cancer section for explanation of these terms) then the diagnosis may be established from serum testing together with the MR image of a brain tumour that fits the picture.
Malignant brain tumours differ from cancers at most other sites in that their malignant potential is almost exclusively due to their local infiltrative growth within the brain around the region in which they arose. Staging systems comparable to those for other cancers is therefore inappropriate.
There is only rarely metastatic disease to distant sites, although some tumours such as the medulloblastoma (germ cell tumour, primary brain lymphoma and ependymoma), may seed/spread throughout the CSF space due to the ‘shedding-off’ of tumour cells into the CSF.
The extent of the growth is assessed by MRI scanning of the brain/spine at the time of diagnosis.
Treatment and outcomes: Brain tumours
The treatment of gliomas is common to all three subtypes and depends more on grade than whether the disease is an astrocytoma, ependymoma or oligodendroglioma.
The first and best therapy is surgical debulking of as much tumour as possible, without causing irrevocable neurological damage. Where there is raised intracranial pressure at presentation, surgical debunking is very important as the first measure.
The problem of normal and critical surrounding functioning brain is the reason that, complete resection of gliomas is usually impossible. Before taking the patient to theatre the surgeon may well wish the patient to have an initial period on potent steroid therapy. This serves to reduce the oedema (‘water-logging’) that surrounds every glioma (more so in higher grade tumours) and renders the patient in better neurological condition to undergo operation. The patient must be off any form of blood thinning therapy such as aspirin or anticoagulants.
The extent of glioma surgery will depend on the situation of the growth. For example, in the non-dominant frontal lobe (the right frontal lobe in right handed people), it is often safe to perform a major debulking operation, whereas in the middle of the dominant hemisphere such a radical debulking procedure could render the patient paralysed down his dominant right side and to the loss of speech (aphasia). The surgeon aims to remove as much of the tumour as he safely can without major risks to the patient’s recovery to status quo ante.
Following surgery the need for other therapy is discussed.
The Oncologist will take great interest in the histopathology report as well as the genomics. The isocitrate dehydrogenas (IDH1) mutaton is characteristic of the low grade gliomas which then carry this mutation when they de-differentiate into higher grade gliomas – a point of significance as these gliomas tend to respond better to therapy (radiotherapy and chemotherapy) than those high grade gliomas that arise as high grade ab initio (which are IDH wild type).
Similarly, those tumours that carry the epigenetic silencing of the MGMT gene (O6-methylguanine-DNA-mehtyltransferase – a DNA repair enzyme that inhibits the killing of cells by alkylating agents) are also more sensitive to these therapies; so the Oncologist needs to know if the glioma carries methylation of MGMT.
Other genomic mutations also have relevance and include TERT promoter, ATRX and recently gene fusion mutations all are having increasing relevance to therapy – e.g. we now often recommend adding chemotherapy to radiotherapy in low grade gliomas that are IDH mutant as this will prolong remissions and hopefully reduce the relapse rates.
In most gliomas this is a routine course of post-operative radiotherapy carefully delivered in small daily doses (as this is the kindest way to deliver radiation to high dose without harming the normal nervous system) over a period of six weeks. High grade gliomas are also treated with chemotherapy, usually by employing the orally active tablet temozolomide chemotherapy (for other chemotherapy, see further down this page).
At relapse, further surgical debulking has a selected place for patients with relatively superficially situated tumours that are causing pressure and in patients who are in good enough shape to withstand further surgery. At the time of such an operation, the surgeon may place some chemotherapy infiltrated wafers (Gliadel wafers) in the surgical cavity, specifically against the cavity walls. However, the advantages of this are small and the practice is dwindling. Subsequent chemotherapy is more complex and the drugs include BCNU/CCNU (carmustine/lomustine), irinotecan and bevacizumab (avastin). In the future, genomically targeted therapy (and perhaps immunotherapy) may have a role.
Occasionally, low grade gliomas may be watched for a time before radiation therapy is recommended: at the time when the scan shows progression. Chemotherapy is selectively recommended after radiotherapy and more certainty in relapsing patients.
The outlook (prognosis) in gliomas is predictable to some extent. Younger patient in good clinical condition (i.e. no severe neurological damage due to the tumour) and lower grade tumours do well; conversely elderly patients presenting with high grade tumours and marked neurological deficit due to their disease do badly. Indeed, most grade 4 (glioblastoma multiforme) patients are dead by two years.
In younger patients of all grades, the good effects of radiotherapy are more pronounced and there is a definite group of grade 3 patients who are long term survivors and an increasing proportion of the low grade patients.
In older patients, with higher grade tumours the decision for post-operative radiotherapy is taken in conjunction with the patient/family. Following radiotherapy, the patients lose some/a lot of hair and feel tired towards the end of the course.
Metastases to brain.
Occasionally the surgeon will have a radical attempt to resect a known single brain metastasis, but will not risk much neurological damage to do so, as radiation therapy can take over where the surgery has stopped.
Nevertheless, where the patient is in good clinical condition (i.e. not about to succumb to cancer progression in the rest of the body) then the results of surgical removal of an ostensibly isolated brain metastasis followed by brain radiotherapy are better than radiotherapy alone. Stereotactic radiation boost *(e.g. by gamma knife) may be used if a metastasis remains on the first follow up scan after broad field radiotherapy and the patient’s general condition remains otherwise good. For patients with good systemic (extra cranial/body) control of the cancer, then focal radiation techniques (stereotactic radiosurgery (SRS) – Gamma Knife, Cyber Knife) may be used to obliterate oligometastatic brain disease (and current practice is to include more than just a few metastases). The higher radiation dose that can be safely achieved by SRS is more likely to obliterate such metastases, but SRS does not “forestall” the development of other, nascent, metastases.
For the more usual situation of multiple metastases to the brain, then wide field brain radiotherapy is recommended and stands a reasonable chance of durable control of the brain – other methods such as focal stereotactic radiosurgery* being reserved for later relapses in one or two sites and only in patients in otherwise good condition.
Brain metastases that have resisted surgery and radiation therapy options:
It is now being recognised that multiple brain metastases can occur (ahead of metastases elsewhere) and where they are no longer operable or treatable by radiation therapy then systemic therapy must be considered. Indeed, as we become better at controlling the systemic metastases, brain metastases have become a bigger problem in Oncology.
The brain was previously thought to be a nested/sanctuary site for systemic therapy because of the Blood-Brain Barrer. Whilst this is partially true, nevertheless, it has become clear in recent years that systemic therapy can be used to help control brain metastatic disease.
It should be noted that both HER-2 ampliciation positive and triple negative breast cancer, bot have a predilection to metastasise to brain. In HER-2 mutated breast cancer, the combination of the small molecular HER-2 directed inhibitor: Lapatinib with the chemotherapy agent: capecitabine can cause remissions. A new generation of small molecular HER-2 directed therapies (e.g. Neratanib, Tucatanib) are now in increasing usage and have brain access.
In other cancers that are driven by mutated oncogenes that can be inhibited by drugs (e.g. BRAF mutated Melanoma, for which the drugs dabrafenib and trametanib can cause remission in brain metastases) and the genomically targeted option for therapy must be considered for all cancers. Osimertinib is a good example of this – being active in EGFR mutant lung cancer and having brain access.
In triple negative breast cancer. which can respond to immunotherapy, then immunotherapy can cause remissions. Immunotherapy more certainly can cause remission in brain metastases from melanoma.
Immunotherapy has a role in treating other brain metastases and particularly in melanomas ( which frequently metastasises to brain) and checkpoint inhibitors such as pembrolizumab- indeed the responses to this type of immunotherapy can be durable. Leptomeningeal metastatic disease is a life threatening condition but can respond durably to “smart” drugs of immunotherapy.
Meningioma and pituitary tumour
Surgery is the therapy of choice for meningiomas and for pituitary tumours (the surgical access to the latter being often via the nasal cavity (the trans-sphenoidal route) with the intention of cure. However, once again here too the operation, particularly for deep seated and small tumours is often being replaced by the non-invasive focal radiation* methods such as the gamma knife (not a knife at all but a form of non-invasive focal radiation therapy).
Similarly surgery is the first therapy for craniopharyngiomas but cure without post-operative radiotherapy is less readily achieved.
Vestibular (acoustic) neuroma
Surgery used to be recommended for acoustic neuromas with the intention of complete removal and cure, but in recent years the surgery has largely been replaced by focal stereotactic radiation therapy* (techniques such as the Gamma Knife). Such focal radiation therapy has the ability to focus a high dose of radiation on the tumour, usually resulting in cessation of any further growth and sparing hearing, without the risks of surgery. Only where the tumour is very large do we now contemplate surgery.
Stereotactic radiation therapy/ radiosurgery*:
*In stereotactic focal radiation therapy/radiosurgery (SRS), the patient’s head is immobilised in a frame which encircles the head as does an equator encircle the globe (the head in this case). The patient then has a scan and this allows three dimensional co-ordinates, longitudes and latitudes continuing the former analogy or x, y, z co-ordinates in stereotactic language, to be obtained and for the x or gamma ray beams to then be concentrated on the growth. Skull tracking may be an alternative mapping system.
Such is the concentration of dose on the target region and the speed of dose ‘fall-off’ at the edge of the targeted tumour that it is possible to deliver obliteratively high single radiation doses to the tumour without over-irradiating the surrounding normal brain. The technique of stereotactic radiation therapy is becoming an increasingly important tool in the neuro-oncologist’s armamentarium.
Chemotherapy has a limited place in the treatment of most brain tumours. Following surgical debulking and radiotherapy, the two most powerful therapies have been given and the addition of chemotherapy may help somewhat but it is not as strong as radiotherapy.
Temozolomide has been shown to be active in high grade gliomas, and has largely replaced nitrosomes in first line therapy.
Irinotecan and bevacizumab/avastin is a couple of drugs showing promise, and are offered to some patients in relapse after first therapy.
Other drugs that are active against gliomas are: nitrosoureas (BCNU, CCNU and Methyl-CCNU), procarbazine and vincristine.
Some brain tumours are driven by actionable mutations – e.g. BRAF mutation in pleomorphic xanthoastrocytoma and pilocyticastrocytoma and occasional EGFR mutation in glioma represent driving oncogene mutations that can be “drugged”.
Whilst the outcome for many patients with high grade gliomas remains very uncertain, progress is being made using new surgical, radiotherapeutic and chemotherapy treatments.
For some tumours, cure is now routinely possible. Meningiomas, pituitary tumours, germ cell tumours, medulloblastomas and lymphomas are good examples of this.
Primary brain tumours are rare and a population based screening programme is not considered indicated. It would involve population based MRI brain scanning of people at intervals and the health economics and the costs and the level of anticipated co-operation from the public would not make this a viable proposition.