Dr Robert Smee, Staff Specialist, Radiation Oncology,
Prince of Wales Hospital, Sydney.
The therapeutic use of X-rays began 1 year after its discovery by Roentgen in 1896. The use of radiotherapy in the treatment of pituitary tumours goes back at least 70 years with obviously very primitive treatment methods available during those early years. Imaging at that time-frame relied upon plain x-rays with an expanded sella / pituitary fossa as the indicator, and the treatment approaches using what are now referred to as orthovoltage machines delivering quite a high dose to adjacent tissues so the pituitary received an adequate dose. In the 1960s Cobalt machines were available for treatment. In the 1970s the first versions of CT scans were developed for diagnosis, and linear accelerators (LinAc) for treatment.
The late 1980s – 90s saw MRIs become widespread and for treatment LinAcs became more computer controlled for accuracy, and the Gammaknife more refined for radiosurgery. Cyclotron produced Protons in Boston were used to treat pituitary tumours from the late 1970s.
It is thus from this embryonic beginning that we have a much better understanding of what can and can’t be achieved by any form of radiotherapy in the treatment of pituitary tumours. The features that we now have include:
- better understanding of the behaviour of pituitary tumours and the likelihood of response to treatment,
- better imaging methods with exact demonstration of the extent of tumour spread,
- better physiological support of the patients who have these tumours and have treatments,
- better surgical approaches with endoscopic methods being used increasingly with extended resections appropriate,
- better and more accurate delivery of radiotherapy by multiple means, defining adjacent dose limiting structures with subsequent avoidance / minimisation techniques,
- better follow-up so that audit of the process and outcome of treatment can occur, and
- better participation of the patient and family in the disease journey.
Despite the persistent use of the simple division of pituitary tumours into functioning and non-functioning, there is now much greater knowledge of the pathology of these tumours. However this has not changed the fact that for the majority of patients the time-frame from the onset of features to diagnosis is measured in years not months, the functioning tumours probably a little less, although with acromegalics using prior photos 10 years is still the average time-frame from first change to diagnosis.
Both functioning and non-functioning conditions have a tumour that needs to be addressed. This was the initial role of radiotherapy. It was felt that these tumours if they couldn’t be removed would behave as do many cancers treated by radiotherapy and disappear after treatment. With prolonged follow-up it was apparent that as with all benign lesions these tumours didn’t behave like cancers, the tumour remained stable, may decrease, but didn’t increase in size in the majority of patients (early on as imaged on CT, and over the last 15 years with MR.) In the context of treating the tumour it was noted that many patients had reduction in elevated hormone levels. Given that stable tumour bulk was the most likely outcome it was very early on recognised that radiotherapy was not a means of consistently improving vision if there is pressure on the optic chiasm.
Thus benefit was being gained in the 2 aspects of pituitary tumour presentation: tumour control and excess secretion. However 2 important features were recognised: the development of side effects and the prolonged time-frame to benefit. The former was addressed by means of improving the treatment delivery, as well as recognising the virtues of fractionated treatment when appropriate. The latter was approached by trying to increase the intensity of treatment.
However progress doesn’t just take place in 1 isolated field of medicine. Many publications in the 1980s - 90s regarded cure of a functioning tumour as merely reducing the excess secretion to a certain level. To a great extent this was real for Cushing’s Disease and Prolactinomas but certainly not for Acromegalics. Better understanding of the secretion of Growth Hormone has resulted in significant modification of the concept of cure to an extent that we have to regard any form of radiotherapy as being very helpful in the management of referred Acromegalic patients, but cures only the minority. Cushing’s patients if benefited early feel better, and the majority do. Patients with Prolactin producing tumours may have control of the tumour but major reduction in hormone excess secretion by radiotherapy is infrequent.
It is very clearly recognised now that the patient, adult or paediatric, with any pituitary tumour will see many clinicians. The radiologist may not be seen but is an important component of the team. Appropriate radiological investigations are the backbone of management. Is the tumour invading other structures, where is the normal pituitary, closeness to the chiasm, is the sphenoid sinus suitable for surgery, etc are important questions the surgeon will want answered. The endocrinologist evaluates excess secretion of the tumour and the status of overall pituitary function, and the necessity for pre and / or post-surgery hormone replacement. If the tumour is a Prolactinoma then treatment will be instituted at this point. The neurosurgeon will be involved in the majority of these tumours, importantly if there are pressure features on the chiasm, the exception again being Prolactinomas.
The surgical approach needs to be determined and another team member, may be unseen, is included, the ENT Surgeon to gain endoscopic access to the pituitary fossa. Further radiological imaging may be necessary immediately pre-surgery, and certainly post-surgery to provide comfort as to the extent of tumour removal. The 24hr post-surgery serum cortisol and growth hormone levels provide a very good, though not absolute, guide to the endocrinological cure of the excess secretion. The Anatomical Pathologist, also unseen, provides confirmation of the histological nature of the tumour. The Radiation Oncologist is usually seen later in the disease course, except in 2 circumstances, both infrequent: where the patient despite all discussion declines surgery, and where (if optic pathway compression is not present) surgery is going to be knowingly incomplete and radiotherapy required at some stage – usually relates to the non-functioning tumours. Thus referral may be in a short time-frame post-surgery to determine if / when radiotherapy is appropriate, or at a time-frame of clinical / radiological hormone / tumour progression.
All practising Radiation Oncologists will have been exposed at some time in their career to the so-called 3 field technique to treat pituitary tumours. This involves directing the radiotherapy beam from 3 different directions at the tumour – a superior or vertical field, and right and left lateral (for each side of the head) fields. The aim of this approach as with all radiotherapy is to deliver adequate dose to the tumour, and minimising the dose to adjacent structures. However in current times this is not considered a reasonable standard of care. This approach delivers a high dose to the hypothalamus, and can also result in a higher than necessary dose to the temporal lobes in patients likely to live many years. Thus acceptable current approaches are:
Stereotactic Radiosurgery (SRS)
This involves the delivery of a modest to high dose of irradiation to the pituitary tumour (very occasionally it may mean treating the whole pituitary gland.) The problem may however be that of hyperplasia – where there is a diffuse increase in the ACTH producing cells, rather than a discrete tumour mass. In both situations a single episode. Appropriate imaging, nowadays with MR (CTs if a contra-indication to MR exists) is mandatory, a means of defining a specific point to aim at, and an accurate and precise means of delivering the multiple irradiation beams to that defined tumour, in a single fraction. There are a number of different devices that can be used: the Gammaknife, the various Linear Accelerator modes (Brainlab, Cyberknife or modified general purpose LinAc,) and charged particles (typically protons.) Whilst the means of producing this energised beam is different between the various methods, the interactions within tissue are the same. It is not surprising then that generally the outcomes are the same. There is a degree of variability between each approach but also this exists within each device mode. Common across all different approaches is the likelihood of achieving tumour control (lack of progression) at > than 90% with long-term follow-up. More variable is the hormone excess secretion control. This has been influenced in recent years by the definition of control / cure. Typically there is a greater likelihood of hormone reduction in patients with Cushing’s Disease then with Acromegalics, than with Prolactinomas. Somewhere between 30 - 70% of patients will have major reduction in elevated hormone levels, with the time-frame to benefit being 15 – 18 months. A higher dose of irradiation needs to be given to functioning tumours for hormonal control than to control the tumour aspect. Noted above is the difference between tumour control likelihood (>90%) compared to hormone control (30–70%). Given that this treatment is delivered in a single event (hence the term surgery in the name) it is much more convenient.
Most SRS approaches use a fixed head ring for the duration of treatment, some facilities use a form of a non-fixed device for a single fraction. The former is the more accurate. Two of the Linac procedures (Cyberknife and Brainlab) use diagnostic x-ray based procedures to evaluate accuracy of treatment set-up. The Radionic approach is to use mega voltage x-rays to evaluate set-up accuracy. The Gammaknife relies upon the engineering construction of the delivery device.
SRS is not appropriate in 2 circumstances:
- When the tumour is within 3mm of the optic chiasm. This is because the optic chiasm does not tolerate, without risk of visual deficit, the same dose required to control the tumour and all the excess secretion
- When the tumour is very wide extending into each cavernous sinus. This is then amenable to stereotactic radiotherapy. The other cranial nerves adjacent do tolerate the appropriate dose and thus are not considered in determining suitability for treatment.
Treatment times are influenced by the delivery parameters of the treatment device. Typically treatment times for all devices would be 15-25mins for a non-functioning tumour and 20-40 mins for a functioning tumour. It should be noted however that there are many hours of preparation time prior to each procedure.
Stereotactic Radiotherapy (often referred to as Fractionated Stereotactic Radiosurgery)
This relates to use of multiple fractions to treat a tumour, the number ranging from 5-30 fractions. The determinant in using a number of fractions is definitely close proximity to the optic chiasm and occasionally the width of the tumour if it extends into both cavernous sinuses. Cyberknife as a standard tend to utilise up to 5 fractions where there is concern re contact with the optic chiasm. If the treatment is delivered over 25-30 fractions then the likelihood of visual deficit related to treatments is <1%. Delivering a low dose determines a lower likelihood of benefit. Thus low dose radiotherapy may lead to greater future risk of damage to the chiasm as the tumour enlarges.
To use a fractionated approach some form of re-attachable (non fixed) head fixation device is required. Various devices are available ranging from dental with a mouth piece (typically shaped like a denture / football mouth guard and based on the upper jaw) to plastic masks that cover the face enabling fixation of the chin and forehead whilst treatment is taking place. Frequent QA assessments determine the reliability of daily set-up.
A critical feature of all stereotactic approaches is to minimise the dose to adjacent normal structures. There are limits though to how this can be achieved. Thus if the tumour is in direct contact then this structure will receive the same dose as the tumour (as described above) and the appropriate fractionated method utilised. Similarly if the tumour extends into the cavernous sinus then normal tissues within / near this will also receive close to if not the same dose as the tumour. Usually a lower dose will be given to the hypothalamus and temporal lobes. The impact upon the pituitary gland will be described below.
Conventional (non-stereotactic) radiotherapy
In this circumstance it is a specific volume that is targeted, not a specified point. The standard is the 3-field approach with the x-ray beam directed at the pituitary tumour from both sides of the head (referred to as lateral – right and left – fields) and a vertical field. Further beam directions should be now considered the norm if this approach is to be used, thus use 4-6 beam entry directions instead of the 3-field approach. Whilst there is equivalent (or close to) coverage of the tumour, these approaches usually result in somewhat higher dose to adjacent normal structures compared to the stereotactic approaches.
Intensity Modulated Radiotherapy
This is a more modern means of radiotherapy delivery whereby the intensity of the radiotherapy beam is modulated (or varied) dependent upon the normal structures that the beam is passing through. This approach can be delivered stereotactically (ie aiming at a specific point.) However it is not possible to use this approach for small tumours such as many of the pituitary tumours that need to be treated.
For both functioning and non-functioning tumours it is important to control the tumour, by all techniques with appropriate doses this should be able to be achieved in >90% of patients treated. As indicated above this represents lack of progression, a much smaller proportion of patients will have major regression of the tumour over time. There is no real difference between single or multiple fraction methods in this outcome. Only long term follow-up however can demonstrate this situation.
The functioning tumours however have an added outcome and that is reduction of the abnormal hormone levels. This occurs in a lesser proportion and is influenced by the type of excess secretion. Although there is some disagreement it is generally accepted that the difference between the SRS and SRT techniques is a shorter time to benefit with the single dose approach. It is the preferred approach if it can be delivered safely.
Side Effects: Best divided into short and long term
This is defined as those events occurring within a few days to a few months after treatment. If there is a fixed headring then bruising to the forehead will be experienced. Minor headaches and nausea can occur usually settling with simple medication. There is however a variable amount of fatigue both in severity and duration. The great majority recover without many problems.
Concern re the optic chiasm has already been expressed. The risks of any deficit to vision should be <1%. A similar risk should apply to the impact upon blood vessels and other cranial nerves adjacent to the pituitary gland. Modern stereotactic approaches pose very low risk to the hypothalamus and the temporal lobes. Two specific risk events should be considered:
Damage to normal pituitary tissue – unless the tumour is small and located on the outer aspect of the pituitary fossa, it is not possible to deliver adequate dose to the pituitary tumour and a much lower dose to normal pituitary tissue. Thus over the ensuing years (typically 10-20 years) there is at least a 20-50% chance that pituitary hormone replacement of some form is required. The chance of this event occurring with a functioning tumour given the necessity of higher dose is thus greater.
Risk of another tumour developing – there are mixed views on this. Some centres report a 1-3% risk of another tumour (even a malignancy) developing within the brain over the next 10 – 20 years. However a large British study of 5000 patients treated by stereotactic irradiation noted no increased incidence of further tumours. This needs to be put in context in that the life time risk of any person in the community developing any form of cancer currently is 1 in 3.
There is now a lot more information about on the appropriate ways to treat pituitary tumours by radiotherapy whilst there are many different methods available. If all the stereotactic approaches are considered, there is little difference over time in the outcomes for both functioning and non-functioning tumours for the different modes of treatment. The standard 3-field means of radiotherapy delivery should no longer be considered an appropriate means of treatment. All patients referred for treatment of a pituitary tumour should be managed in a multi-disciplinary fashion. This includes all procedures and follow-up.
Should I take time off work while having treatment?
The great majority of patients tolerate treatment well, obviously there can be logistic circumstances such as travel time which can impact upon that. There is typically fatigue towards the end of treatment but many patients are able to continue work.
When can I drive again?
This is not influenced by having radiotherapy, or indeed radiosurgery
Will I lose hair, where and how long till it grows back?
In the great majority of situations in which stereotactic irradiation is used there is no hair loss. The only circumstance in which small patches of hair loss occurs is if Stereotactic IMRT is used, if the hair is long these are not evident. If there is any hair loss after about 6-8 weeks following conclusion of radiotherapy the hair will begin to grow again.
After treatment, if the tumour is still active, can I have radiotherapy?
Presumably in this circumstance initial treatment was surgery in which case it would be very appropriate to have radiotherapy.
After treatment, if the tumour is still active, can I revert to surgery?
Presumption is that the patient has had previous radiotherapy, in this circumstance surgery can usually be used. Radiotherapy can have an impact upon normal tissue in that very major surgical procedures can sometimes be more complicated. Certainly a repeat transphenoidal surgical procedure should be possible and reasonably safe.
If I’m seeing my Endocrinologist on a regular basis will I need to have regular follow-up with my Radiation Oncologist post-treatment?
Shared follow-up can be worthwhile both for the patient and the respective clinicians in this situation. As we are addressing different follow-up circumstances, ie the Endocrinologist will evaluate the pituitary function, where as the Radiation Oncologist may be addressing what is happening to the tumour.
What are my options if remaining cells are producing excess hormones after radiotherapy?
Depending upon the time frame after radiotherapy it may be possible to give repeat radiotherapy, if surgery is not possible, or there is no medical option to treatment the excess hormone production. A determinant will be the extent of benefit gained with the initial radiotherapy course. This is a situation however which requires a lot of consideration.
If residual hormone excess is present after surgery is it true that radiotherapy is more successful than if done close to surgery time?
Not necessarily, immediately following any surgery there is the presence of blood products which can distort the appearance of the residual tumour on MR scan. Thus the minimum time frame to radiotherapy post-surgery should be 3 months to allow for the blood products to be reabsorbed and gain best demonstration of tumour size and shape. However if the excess hormone secretion continues after surgery then it may make control by radiotherapy more difficult.
Prince of Wales Hospital, Sydney: Radiosurgery/Radiotherapy Clinical Program