What I have tried to do in this section is to provide general information about brain tumours and their management as well as outlining my own experience in this field. As will become obvious there is no consensus regarding the management of many brain tumours so much depends on the philosophy of the clinicians involved. I hope my own philosophy about this will become apparent.
Neuro-oncology is the field of brain tumour treatment. The diagnosis of a brain tumour is always serious from a patient’s view. Whatever the type of tumour the diagnosis of a brain tumour will have life changing effects on a patient and their families. The field includes the management of entirely benign tumours to some extremely malignant tumours, some of which require surgery while others do not. The devil in this subject is in the detail which makes writing meaningful general comments about brain tumours and their management extremely difficult. They are not common tumours, so people’s experience of them is usually not great. If a patient or their family or indeed their GP does know of other people who have a brain tumour, the chances are that it will be different from the tumour that they have been diagnosed as having. Even two tumours with the same name may behave completely differently in different people and require completely different treatment.
Classification of brain tumours
Every first year medical student should be able to classify tumour into: benign, malignant, primary or secondary. They should be able to tell you that benign tumours are those that grow slowly without invading or destroying tissue locally and which do not spread around the body, while malignant tumours do the opposite. They do invade locally and they do spread around the body i.e. they metastasise. Brain tumours are further classified by their tissue of origin, either from the brain itself, the gliomas, or from the tissues surrounding the brain, meningiomas, acoustic neuromas, pituitary tumours etc. In simple terms tumours arising from the tissues surrounding the brain concur with the medical student classification outlined above. However the situation regarding primary brain tumours is more complex and often leads to confusion, worry and uncertainty amongst patients, their families and their doctors!
Are primary brain tumours benign or malignant?
Primary brain tumours arise from the support cells of the brain, the glial cells. They are therefore called gliomas or are named after the cells from which they arise, astrocytomas, oligodendrogliomas or ependymomas. Some of these tumours behave in an entirely benign fashion while others behave very aggressively. The thing that distinguishes one type from another is their rate of growth. They all infiltrate (invade) into normal brain and even the most malignant do not metastasise (except in the rarest circumstances). The simple definitions of benign / malignant cannot therefore be applied to brain tumours. We talk about low grade (grade 2), intermediate grade (grade 3) and high grade tumours (grade 4) to determine the rate of growth of a tumour and therefore its degree of malignancy. (There is a grade 1 tumour which is entirely benign but this only occurs in children) The situation is further complicated by the fact a low grade tumour may remain so for many years but then ‘change its spots’ and become more malignant. It used to be thought that all gliomas eventually did this which led to the traditional definition of all these tumours being classified as malignant This might still be considered to be the case by some clinicians. However the modern view is that this is not the case and that some low grade tumours do behave in an entirely benign fashion. (This is particularly the case amongst the population of patients that have chronic epilepsy). However this does produce a great deal of uncertainty and worry for patients who have been diagnosed as having a low grade glioma and who have been told about the likelihood of it behaving benignly, who then read in a text book or indeed on the internet that their tumour is actually classified as malignant.
Treatment of brain tumours
While there is uncertainty in the profession about the classification of brain tumours, there is certainly no consensus about how best to treat these tumours. Do you operate or don’t you? If so, what do you aim to achieve with surgery, just a biopsy for tissue diagnosis, a partial or more radical resection? When do you offer adjuvant therapy such as radiotherapy or chemotherapy? What is the role of newer therapies such as stereotactic radiosurgery? Do alternative therapies have a role? How a neurologist, neurosurgeon or oncologist approaches these questions depends a great deal on their experience and personal philosophy. What we are only just accepting as a profession is that the patient has a say in all this! There is probably no area of cancer care where a patient centered approach is more important or where an environment of ‘fully informed consent’ more difficult to achieve.
This is an outline of my experience and my philosophy. My first informative experience of the management of glioma patients was as a registrar in Liverpool in the early 1980’s. At that time neurologists were extremely conservative about the management of glioma patients especially those presenting with epilepsy. It was the norm not to scan patients unless there was a focal element to their seizures. The only type of scan available was a CT scan which in those days was not good quality. Patients presenting with progressive neurological deficit, cognitive impairment or symptoms of raised intracranial pressure were scanned more regularly and referred for surgery.
Worldwide the conventional neurosurgical approach was to do a craniotomy remove some of the tumour, then offer patients radiotherapy, as treatment for high grade tumours and as prevention in low grade tumours. This was the time when all gliomas were thought to be malignant. The results of this approach were pretty awful. Median survival times, i.e. the time within which 50% of patients died was 9 months with treatment and no more than 2 months without. It was recognized that patients with low grade tumours did better with median survivals of two years or so.
Liverpool proved to be a unique working environment in which to study the management of gliomas for two reasons. Firstly there was only one CT scanner in the whole region serving a population of 3.5 million. All patients with a scan diagnosis of a glioma were therefore managed at one hospital, the neurological center at Walton. Secondly everyone in the neuroscience department, surgeons, neurologists and oncologists took a very conservative approach to gliomas. Routinely the most aggressive surgery done was a biopsy, while radiotherapy was only offered to patients of in a good neurological state. This gave me the opportunity to study a unique database consisting of a complete cohort of glioma patients.
The results produced were very controversial at the time. Firstly there was no apparent difference in outcome between those patients treated by surgeons and those treated by neurologists – uniformly bad, except obviously the incidence of surgical complications. Secondly the median survival of patients undergoing a biopsy followed by radiotherapy was just the same as published series of patients undergoing the conventional craniotomy and internal decompression. Thirdly it was possible using multivariate analysis statistical techniques to demonstrate that what really influenced outcome in glioma patients was not the treatment they received but to certain presenting features that were associated with better survival, namely young age, presentation with epilepsy and good neurological condition at presentation. The philosophy in the department at the time was that there was no point in ‘wasting’ scarce neurosurgical resources on glioma patients when these resources were better spent on the treatment of other tumours such as meningiomas.
This situation influenced me in two fundamental ways. Firstly my studies did demonstrate a benefit from decompressive surgery and this was that the long term use of steroids in patients who had their tumours decompressed was considerably less than in those patients who had just had a biopsy. This had a considerable effect on the quality of a patient’s life. Secondly I was made aware that all conventional adjuvant therapy for gliomas was never going to be curative because it was not selective. In other words, the damage done to the normal brain always limited its use and therefore its effectiveness.
I spent a year in the mid 1980’s studying a form of laser treatment for brain tumours, namely photodynamic therapy which I hoped be able to produce selective tumour cell death. This we were able to demonstrate in the lab (perhaps for the first time) but the depth within the brain to which the selective photodynamic effect could be produced was only a few millimeters which was not enough to be clinically significant so I never developed the technique clinically although it was used in Australia for a while. I therefore left Liverpool with the conviction that thorough surgical decompression did have a role in glioma management after all.
Following my senior registrar years in Manchester, I was appointed as neurosurgical consultant in Bristol in 1991. My specific remit was to develop stereotaxy in the department in addition to taking a role in the neuro-oncology program. I was fortunate to be joining an enlightened neuro-oncology team led by my surgical colleague, Brian Cummins and his neuro-oncology colleague Gill Bullimore. They were both ahead of their time in recognizing the need for multidisciplinary team working, setting up multidisciplinary clinics as early as the late 1980’s. In Bristol at the time the surgeons took as radical approach to gliomas as they could within the limitations of the technology available to them. Reoperating on recurrent tumours was recognized as useful palliative treatment in selected cases. In general the system was much more supportive to patients than what I had been used to in the north.
Two developments in the last decade have improved our ability to resect intrinsic brain tumours, image guidance and improvements in anesthesia that have allowed us to operate under local anesthetic. Where it is appropriate it is now possible to remove the radiological extent of a tumour even in some cases where the tumour is arising from eloquent cortex (i.e. areas of the brain known to have important function as opposed to silent cortex that apparently we can do without!) Developing image guided surgery and local anaesthetic techniques have been one of my major activities as a consultant, allowing me to take an aggressive approach to the surgery of gliomas where appropriate. So although, in general, I have a very conservative approach to surgery including tumour surgery where radical surgery is indicated I am fortunate to have the tools to do this as safely as possible.
My prime indications for surgery for brain tumours and those of the clinicians with whom I work, are:
to establish a tissue diagnosis,
to reverse or prevent symptoms from a tumour,
to reverse or prevent life threatening problems from a tumour and where possible to cure patients.
Some tumours do not require immediate surgery although all need to be watched carefully with regular follow up and scanning. A common group of patients in this category are patients that present with an epileptic seizure out of the blue who have a lesion on an MRI scan that looks abnormal but not sinister and who is symptom free on anticonvulsant drugs. It is perfectly reasonable to follow patients like this up with regular scans. If, however, a patient presents with progressive symptoms, the commonest being neurological deficit like a stroke, progressive mental deterioration or symptoms of raised pressure inside the head i.e. headache, then surgery to remove some or all of a tumour is usually required. Sometimes I will just biopsy a tumour in order to plan further treatment. This is particularly important where the mainstay of treatment might be therapy other than surgery e.g. cerebral lymphoma. Otherwise I will offer surgical resection of the tumour.
It is not appropriate to opt for radical resection of all tumours as in many instances partial resection is all that is required in terms of establishing a tissue diagnosis, treating the presenting symptoms and preventing problems in the future. In these cases I want to avoid the extra risk associated with radical tumour resection where this is not in a patient’s interest. However radical resection is beneficial I would expect to be able to offer patients this service in an appropriate timescale with the resources to treat them safely. (This may sound like an obvious statement but this is not a foregone conclusion in the modern NHS) The definition of who requires radical surgery and who does not is too complex to define in a general article like this and needs to be discussed on an individual case basis. I would expect that this is discussed thoroughly with all my patients, all the options for treatment presented and a consensus reached as to the best way forward for each individual patient. However this more holistic, patient centered approach is not easy to achieve in the conveyor belt, target driven environment in which we now work.
Current Neuro-oncology practice – MDT working
I wrote the above in 2004. The situation has now changed regarding brain tumour management. In 2009, The National Institute of Clinical Excellence (NICE) published their guidelines for brain tumour management, the principles of which were laudable. The key recommendations are outlined here:
- All patients' care should be co-ordinated through a designated multi-disciplinary team (MDT)
- All patients should have face-to-face contact with healthcare professionals to discuss their care at critical points in their care pathway, and be provided with high quality written information to support this
- All patients should have a clearly defined key worker
- Patients should have ready access to specialist care services as appropriate
- Palliative care specialists should be core members of the neuroscience MDT and of the cancer network MDT
- Cancer networks should ensure that clinical trials on brain tumours carried out by the National Cancer Research Institute (NCRI) are supported and patient entry into these studies actively monitored.
Taking its lead from the organisation of breast cancer services, the government decreed that all tumours should be managed through a multidisciplinary team (MDT). Rigid structures for oncology MDT’s were laid down and the process of working. In particular the standard was set that all patients should have a management plan discussed by the MDT prior to the commencement of any treatment.
I had reservations about this, despite the principle sounding good in theory. These were simply that by applying the new process to brain tumour management we would lose some of the important benefits of our, then current service. Before the introduction of the MDT process, patients presented to our referring hospitals and a provisional diagnosis made with scanning. They were then referred to Frenchay as an emergency and their case taken on the on call neurosurgeon. Glioma surgery was then defined as part of general neurosurgery. Patients would be operated on, the diagnosis confirmed histologically and their cases then discussed at the equivalent of the MDT and a further management plan made, with the certain knowledge of the diagnosis by all the specialists with potential involvement in that patient’s care. Sometimes more surgery was required, this time taken on by a specialist oncology surgeon, like me. Combination chemotherapy and radiotherapy was planned from that meeting. Further follow up was then organised by the oncologists with input from neurosurgeons where necessary. Most importantly to the process, the specialist nurse practitioners were involved with the patients and their families from the outset and provided a vital link role in guided patients through their treatment journey.
The new process changed all that. First of all it was decreed that all glioma surgery must be done by surgeons with a specialist interest in neuro-oncology and that to be a specialist oncology surgeon this activity must represent 50% of a surgeon’s activity. Having worked as a consultant in this field for over 15 years with only 25% of my surgical activity being oncology based, I knew from practical experience that the psychological strain of focussing on oncology more than I had would prove very difficult for the new generation of oncology surgeons. Secondly the need to establish the diagnosis definitively for patients is paramount in neuro-oncology. Scans cannot be relied on to make this definitive diagnosis – tissue must be taken by some form of surgery for this. One of the worst disasters in neuro-oncology is to label a patient as having malignant pathology when they turn out to have benign pathology like abscess or stroke. Thirdly diagnostic surgery in the form of biopsy or partial resection of a tumour is not difficult to perform safely and does not require specialist oncology skills. There is no advantage in a process that delays diagnostic surgery in a favour of deferring to a specialist oncology surgeon. Fourthly a decision regarding surgery cannot be made by committee. Essential to the process is the personal interaction between surgeon, patient and their families where the surgical options are discussed along with the relative risks of each option. Applying a formulaic approach to brain tumour treatment flies in the face of the principles of patient centric care. Without a prompt, personal approach to each patient, it is very difficult to provide the basis for trust that is needed in neuro-oncology more than anything. Finally there is no point in having a team of specialist nurses to interact with patients and their families if they do not get involved in the patient’s care from the start.
What now happens is that a patient will have their provisional diagnosis made in the referring hospital by scanning as before. However instead of being transferred to Frenchay straightaway, their case is deferred to the MDT. The patient’s themselves are usually sent home having been told they have a brain tumour and that ‘someone will contact them’ from Frenchay. They then remain in limbo until that contact is made, a process that can take two weeks or more. They will then be sent a letter for an outpatient appointment to see a surgeon who will have the ‘benefit’ of an MDT decision regarding the surgery required. This decision can only act as a guide as any final decision can only be made at this meeting, when hopefully the specialist nurses are on hand to take the patient on to their books. I say ‘hopefully’ because oftentimes patients don’t meet the specialist nurses until they are admitted for surgery, up to a month later!
You will sense my disapproval of this system which undoes all the benefits of the patient centric system that we had in place prior to these changes. Having said this I am aware of the political benefit of setting up this working practice. Tumour incidence is normally defined using histological data, in other words after a tumour has been removed or biopsied. With brain tumours at least half are diagnosed with imaging without biopsy, for example low grade glioma patients, presenting with epilepsy and treated medically with anticonvulsant drugs and serial scanning without surgery. As a result official figures of brain tumour incidence (4-9/100,000 population) underestimate the true incidence (20/100,000) by a factor of at least two. Brain tumours are as common as cancer of the cervix in women and as such deserve as much attention in terms of resources for research etc. Political acknowledgement of this was made conditional on the adoption of the MDT system, fully appropriate for breast cancer but not for brain tumours without important modification in my view.
The modification that I recommended was for patients to be transferred to the on call teams as before and for diagnostic surgery to be performed straightaway with immediate input form the specialist nurse teams as before. Informed discussion regarding further management could then be had at the MDT. In those cases where further surgery was required this could then be done by surgeons with the necessary specialist skills as part of the MDT plan.
That suggestion was turned down and this was one reason why I am no longer part of the formal brain tumour team. I continue to operate on patients referred as an emergency where I think there is the clinical imperative to do so – e.g. uncertain radiological diagnosis, very rapid tumour growth. I also continue to manage those patients ‘on my books’ historically and all the tumour patients on the epilepsy surgery program of which there are many. I will also take on patients whose tumours require specialist surgical skills allied to epilepsy surgery, specifically tumours in the temporal lobe. I continue to hope that the pendulum will swing back and that we will progress to a process that returns to being patient centric again
The future of neuro-oncology surgery
The main reason why I have found being a brain tumour surgeon in the last 20 years so difficult is that conventional treatment doesn’t work! My own data showed that radical tumour resection alone combined with radiotherapy made no difference to outcome when the statistics were compared to my 1980’s data on biopsy and radiotherapy! However there is evidence that introduction of concomitant chemotherapy in the form of Temozolamide has improved survival slightly. However this approach has not produced cure. Survival still depends on the all important predisposing factors of young age at presentation, presentation with epilepsy and good neurological function at presentation. There is therefore a need for radical lateral thinking In this field with the rapid development of novel therapies to the clinical setting.
Much is being done to co-ordinate and facilitate this in the research community (www.sebta.org) with valuable input from the charitable organisations (www.hammerout.co.uk) but until recently there was no sign of a break through. There is some early evidence that a deeper understanding of diffusion within the brain may be changing this.
Chemotherapy delivered by conventional means cannot be anything other than palliative because of its systemic toxicity and the efficiency of the blood brain barrier in preventing diffusion into brain cells. As an academic exercise when I was a research registrar in the 1980’s I calculated that to achieve diffusion of a conventional chemotherapy agent a few millimetres from a brain blood vessel fatal concentration of the drug would have had to be maintained in the blood stream for 24 hours or more! This was not popular when presented in a lab whose work was based around chemosensitivity of brain tumours! Direct infusion into the brain from a drug impregnated source like a Gliadel wafer is effective over a few millimetres but tumour recurs beyond this. The simple injection of drug into the brain substance isn’t any more effective because it unless the injection parameters are carefully controlled the drug simply flushes back up around the injection catheter and doesn’t difuse into the brain at all.
I am proud to be associated with a colleague, Professor Steven Gill whose work has tackled these issues and when applied to the treatment of previously untreatable tumours is showing some promising results. By addressing the technical issue of diffusion through the brain and by optimising catheter size, flow rate and precise positioning within the brain, the professor and his team have defined a means of delivering a uniform flow rate of drug throughout the brain using multiple microcatheters robotically placed. The technique is applicable to a number of applications including brain tumours. His preliminary results using microinfusion of conventional chemotherapy agents in cases with otherwise untreatable pathology have been encouraging – remission for a few months in patients with diffuse brain stem glioblastoma for example. This is a technique that if proven to provide palliation in rarer untreatable and / or recurrent tumours should be incorporated rapidly into the treatment the commoner primary parenchymal tumours by the establishment of multicentre randomised controlled trials comparing Convexion Enhamced Diffusion (CED) with conventional adjuvant therapy. However organising this is likely to proven an uphill struggle for the first patients treated by the professor were treated outside the MDT structure – indeed were not authorised by the MDT. Another criticism I have of management by committee is that this structure discourages innovation and lateral thinking.