GAMMA KNIFE NEUROSURGERY – A REVIEW
Michael Torrens
Department of Gamma Knife Radiosurgery and Neurosurgery Clinic,
Hygeia Hospital, Athens.
ABSTRACT
Gamma Knife radiosurgery has an established therapeutic role in neurosurgery with around
250,000 cases treated and a follow up over 20 years. The success and control rates have
improved as more experience of appropriate dosage has been gained and as more planning
accuracy has been achieved by new machines, new computer programs and better quality
control. The currently quoted control rates for benign lesions less than 3cm diameter are:
astrocytoma grade 1 – 92%, astrocytoma grade 2 – 87%, acoustic neuroma (vestibular
schwannoma) – 96%, meningioma – 96 to 99%, non-secretory pituitary adenoma – 92%, GH
and ACTH secreting adenomas – 80%, prolactin secreting adenomas – 40%, chordoma and
chondrosarcoma – 73%, glomus tumour 100%, epilepsy (mesial temporal sclerosis) – 81%,
trigeminal neuralgia – 83%, parkinsonian tremor 88%, parkinsonian rigidity/bradykinesia – 66%,
obsessive compulsive neurosis – 60%, arterio-venous malformation 90%, cavernoma – 86%,
uveal melanoma – 84%. In malignant glioblastoma, mean survival of 116 weeks has been
described. In cerebral metastases the lesions can be controlled in almost every case so that
survival then depends on the extracerebral disease. In many cases these various results have
been achieved after failure of surgery or radiotherapy treatments. Complications are
significantly less than with microsurgery, radiotherapy or LINAC based radiosurgery. The
method is more cost effective than traditional surgical treatment and is also more acceptable
for the patients.
Figure 1. Leksell Gamma Knife Perfexion
INTRODUCTION
The Leksell Gamma Knife (fig. 1) is a well known treatment method, especially to
neurosurgeons, but there exist some misconceptions which it is the intention of this paper to
clarify as well as to review the contemporary indications for Gamma Knife treatment.
Gamma Knife is not a form of radiotherapy. Radiotherapy relies for its effect on the differential
radiosensitivity of neoplastic tissue and the ability of normal tissue to repair itself. Gamma
Knife treatment is a form of surgical ablation that competes with microsurgery by producing
targeted cell death. This results in shrinkage of tumours (including ‘radioresistant’ ones),
control of hormonal and electrical dysfunction and thrombosis of dysplastic vessels, all without
the risks of open surgery.
Mechanism of action
Radiation from 201 separate cobalt sources is finely collimated. The rays all meet at one point.
The size of the point target is adjusted by changing collimator helmets. The guaranteed
machine accuracy is better than 0.5mm. A stereotactic frame is attached to the patient and
the lesion target is located, guided by MRI, CT, PET and DSA images that can be fused
together. Multiple target isocentres are used to create a treatment plan in three dimensions
that exactly fits the lesion shape with insignificant radiation to normal tissue. An even higher
accuracy and selectivity has been achieved with the new model ‘C’ Gamma Knife by using
many more small (4mm) isocentres combined with an automatic positioning system controlled
by computer which rapidly changes the patient’s position between each treatment isocentre.
Because of the fine collimation of the rays, the accuracy of engineering and automatic
positioning system, the absence of radiation source movement together with excellent
calibration and quality assurance, the Gamma Knife ‘C’ is by far the most accurate and
selective radiosurgery machine as yet developed. For this reason it is not always appropriate
to quote results from older reviews using the ‘B’ model or other systems.
Gamma Knife versus LINAC
All machines giving a similar defined radiation dose to a target should have the same effect at
the target. Those who claim that a LINAC systems are just as effective as a Gamma Knife are
correct. The problem is that the LINAC is also effective on the surrounding tissues outside the
target, causing a higher risk of complications due to its lower selectivity (fig 2). This risk, the
normal tissue complication probability (NTCP) has been shown to be 2-4 times higher using a
LINAC (1). While surrounding damage may not matter much in a frontal lobe metastasis, it
matters a great deal elsewhere.
Fig 2.1 Treatment plan with LINAC
Fig 2.2 Treatment plan with Gamma Knife
Radiosurgery is potentially dangerous if improperly used. Results depend on planning
accuracy as well as machine accuracy. Gamma Knife centers, by concentrating many patients
in one unit, develop more experience. All over the world LINAC radiosurgery centers treat
fewer patients per centre – and often without neurosurgical anatomical expertise.
Consequently they have less experience. This, combined with a less accurate machine,
produces not only more complications but also worse tumour control. Shaw et al. (2) have
documented a 2.84 times greater risk of local tumour progression after LINAC treatment as
compared with Gamma Knife.
As noted above, the guaranteed accuracy of isocentre location is to within 0.5mm for the
Gamma Knife. In reality it is better than this, usually 0.3mm. The equivalent quoted accuracy
of LINAC is 2.0mm. In reality it is often worse than this and deteriorates over time due to wear
in the rotation bearings as the machine gets older.
In consequence most established LINAC units are resorting to stereotactic fractionated
radiotherapy (SRT) to try to reduce the complication rate. This has the potential disadvantage
of a lower long-term control rate in benign lesions as well as being a longer and more complex
therapy.
Fractionation
The definition of radiosurgery, from the first use of the word by Lars Leksell in 1951, has always
been a single dose therapy. Fractionation in time is therefore not radiosurgery but
radiotherapy. The Gamma Knife has been criticized as being unable to deliver stereotactic
fractionated radiotherapy in situations where it may be appropriate such as malignant lesions in
the brain stem. In fact it is exactly the same in practice as SRT using LINAC. Treatment using
both systems can be repeated as often as required as long as the patient can tolerate the
frame on the head, certainly several days.
There is another form of fractionation that is really multiple single doses – fractionation in
space. Where a lesion is larger than 3-3.5cm in diameter and the risk of complications
becomes greater, then treatment can be given to two or more areas at intervals ranging from
one day to several months. This is especially appropriate for AVM’s.
Conditions treatable
The various diseases and disorders that have been effectively treated by Gamma Knife
radiosurgery are listed in Table 1 and the frequency of occurrence shown diagrammatically in
pie charts in Figure 3. The source of information is the Leksell Gamma Knife Society, which
monitors Gamma Knife activity and performance all over the world.
Figure 3
Functional 5%
Malignant tumour 36%
Vascular 20%
Benign tumour 39%
PRIMARY BRAIN TUMOURS
Low grade astrocytomas
Total resection of grade 1 and 2 gliomas is usually dangerous or impossible, even though
advanced imaging techniques often reveal a well-defined tumour margin. For lesions that
cannot be controlled by surgery (fig. 4), Gamma Knife has been used with good effect. We
have personal experience of a number of children with recurrent, and therefore growing,
pilocytic astrocytomas all of whom have been controlled by radiosurgery. In a recent series (3)
the control rate in grade 1 was 91.7% and in grade 2 was 87.2%.
Figure 4. Pineal region astrocytoma before and after radiosurgery
High grade astrocytomas
Survival in patients with malignant glioma increases as the dose of radiation increases but the
dose is limited by the risk of radionecrosis of surrounding brain. Supplementary local boost
radiation using brachytherapy has been used with good effect (4) but the open stereotactic
procedure has a risk of about 1% for each target. However one of the best median survival
rates in glioblastoma (110 weeks) was recorded in cases treated by thermo-brachytherapy (5).
Gamma Knife radiosurgery can be delivered with better conformity than brachytherapy and
without the risks of implantation. A review of this technique has reported median survival in
glioblastoma (RTOG grade 3) of 116 weeks (6). It may be possible to improve on this figure by
using PET to target the active tumour. L-methyl (11C) methionine PET has been shown to
correlate better to active tumour than CT or MRI enhancement (7) and we hope to report on the
value of this targeting method soon. As is well known, survival is strongly related to RTOG
grading but it seems that poor prognosis patients (RTOG grades III-V) have more to gain than
RTOG grades I-II (6).
Haemangioblastoma
Haemangioblastoma is relatively easy to excise when single and associated with a cyst. When
multiple (von Hippel Lindau disease), excision is often impossible. In such cases Gamma Knife
has been used and the result, in small series, has been control in almost every case.
METASTASES
Figure 5. Metastasis before, and 10 months after treatment.
Up to 50% 0f all patients with malignancies will develop brain metastases and, after
radiotherapy alone, more than half of these will die from progressive neurological disease. In
the past, surgical excision of accessible metastatic tumours combined with whole brain
radiotherapy (WBRT) was thought to be the treatment of choice for single metastases.
Several studies, however, have shown that Gamma Knife treatment of single metastases (fig.
5) can produce tumour control and survival as good as surgery plus WBRT (8). Moreover the
results for Gamma Knife are the same for multiple metastases as for a single lesion (9), the
chance of local recurrence is less than after surgical removal and control does not depend on
the histological diagnosis. Since radiosurgery is more cost effective, easier for the patient and
has a much lower complication rate, it would seem that the only indications for surgery are
rapid neurological deterioration due to mass effect and the large (>3.5cm) diameter single
lesion especially in the posterior fossa where any temporary oedema due to radiation would
have serious results. The remaining indications for WBRT are multiple tumours (>10-15) or
miliary lesions and CSF seeding.
Gamma Knife treatment can control >90% of metastatic brain tumours and complications,
usually transient swelling after about 9 months which is controllable by steroids, occur in 5-10%
(10). 80% 3 year survival has been achieved in cases where the extracranial disease has been
controlled (11).
In a comparison of Gamma Knife treatment with WBRT the quantity and quality of survival was
better with Gamma Knife (12). It is not clear how much benefit WBRT has in preventing new
metastases since these may develop by new spread after treatment has finished. Accordingly
the present recommendation is to repeat Gamma Knife surgery as necessary and avoid all
WBRT. In this way patients will not die from the brain metastases and prognosis depends only
on the extracranial disease.
ACOUSTIC AND OTHER NEUROMAS
Pre
6 months post
2 years post
Figure 6. The typical response of an acoustic neuroma to Gamma Knife treatment.
Due to great improvements in surgical technique and the introduction of cranial nerve
monitoring, microsurgeons have reduced the mortality and morbidity of acoustic neuroma
surgery. Samii (13) has reported 97% tumour control with 93% normal facial movement (Grade
I-II) and 47% hearing preservation. At the time of publication (1997) this was equal to the
results from the ‘B’ series Gamma Knife (90% facial function and 50% hearing preservation).
However, unfortunately, not all patients can be operated by a Samii. Average results, including
those from the American Acoustic Neuroma Registry, are 90% tumour control (10% recurrence
or death), 80% adequate facial nerve function and 6-30% hearing preservation (14,15). In
addition there is 1% or more mortality even in centers of excellence (14).
At the same time the results from using highly conformal and selective, multiple isodose
Gamma knife treatment with the ‘C’ series machine have improved. The results reported
contemporaneously (16,17), using a marginal dose of 12Gy, are of 100% normal facial function
and 59-70% hearing preservation (fig. 7) together with 96% control of tumour growth.
Except for having 0% mortality, the Gamma Knife treatment also has a much lower incidence
than microsurgery of the minor complications and disabilities that are usually not discussed but
are common after open operation (18). For example, minor neurological or functional
deterioration (GK 9%, microsurgery 39%), hospital stay (GK 3 days, microsurgery 23 days),
time away from work (GK 7 days, microsurgery 130 days), proportion keeping the same job
(GK 100%, microsurgery 56%) (17).
It is clear that, for smaller acoustic neuromas, Gamma Knife surgery (fig. 6) is the treatment of
choice. For larger lesions (>3cm) fractionation in space may be chosen, especially in the
elderly, or subtotal removal followed by interval radiosurgery for the remnants, which is safer
than attempts at total removal in all but the most experienced surgical hands.
The same comments apply in principle to all other intracranial neuromas.
Figure 7. The percentage risk of cranial neuropathy following Gamma Knife depends
on the size of the acoustic neuroma but in all cases is significantly less than the risk
after microsurgery
Risk of Neuropathy
70
60
50
40
VII
VIII
V
30
20
10
0
< 1 cm
1 - 2 cm
2 - 3 cm
3 - 4 cm
Surgery
MENINGIOMAS
Figure 8. Illustration of the dose plan for an infiltrating meningioma of the cavernous
sinus showing conformal treatment (yellow line) and that even low doses (10Gy green
line) can be designed to avoid the optic chiasm (blue line).
The mortality and morbidity of total surgical removal of meningiomas of the skull base and
venous sinuses is still unacceptably high. Many surgeons elect to perform a subtotal removal
in which case recurrence during the patient’s lifetime is at least 55%.
Gamma Knife therapy (fig.8) has been shown to control 96-99% of meningiomas less than 3cm
diameter, with two thirds of the tumours becoming smaller (19,20). Average morbidity (usually
transient cranial nerve palsies) is 2%. This control is maintained over the long term. A ten year
follow up study, which would have used the inferior, less conformal techniques of the ‘U’ and ‘B’
series gamma knives, showed 93.1% tumour control (21).
It is no longer acceptable to damage the patient by attempting complete removal of
meningiomas. Elective surgical policy should be primary gamma knife treatment of tumours
less than 3cm diameter (except those on the convexity dural surface) and partial removal
followed by Gamma Knife treatment for larger tumours. If the tumour is very close to the optic
nerve then a cushion of fat, Teflon or other material should be inserted to create a space
between the tumour and the nerve. This allows safer and more effective radiosurgery later (the
same applies to pituitary tumours). During a recent EANS debate (22) the notable experts Al
Mefty, Samii and Dolenc all agreed with this strategy. Al Mefty suggested waiting until the
residual tumour actually shows signs of growth since, as noted above, the tumour may remain
dormant.
PITUITARY TUMOURS
Adenomas
Figure 9. Pituitary adenoma before and 54 months after treatment
Surgical removal of pituitary adenomas is, at present, the treatment of first choice because it is
effective immediately. The new minimally invasive techniques, such as transnasal endoscopic
or endoscopically assisted approaches, are better tolerated by the patient and complications
should be minimal. However total removal of macroadenomas is almost impossible and
hormonal control of functioning microadenomas by surgery is reported to be 23-77% (23).
There is therefore a high incidence of failure or recurrence. Also some patients, especially
those with Cushings disease have a high operative risk. Attempts at second operation are
much more difficult and dangerous.
It follows that Gamma Knife treatment (fig. 9) is used most often when surgery has failed. Even
in these cases normalization of hormone function can be achieved in 35-95% depending on
diagnosis (23,24,26). Tumour growth control rate is 92%, shrinkage occurs in 65-80% (25).
Non-secretory adenomas respond best. Among secretory microadenomas, growth hormone
and ACTH secretion responds best (60-80%) with a lesser effect on prolactin (40%) (25).
Normalisation of growth hormone in 96% of cases at 2 years has recently been reported (26).
Better results have been obtained in recent years since it was realized that treatment with
bromocriptine and somatostatin protects against the effects of radiosurgery. Treatment with
such drugs must be stopped before using the Gamma Knife. Radiosurgery has a continuing
effect on hormonal control even beyond 4-5 years and long delayed success is now
recognized. There is also a late incidence of hypopituitarism of up to 20% (25). For these
reasons, patients need lifetime follow up.
Comparison of Gamma Knife with conventional radiotherapy (27) shows that Gamma Knife
works faster and more effectively with less risk of complications such as hypopituitarism or
optic nerve damage. With more experience of dose calculation to achieve more curative and
rapid effects without damage to functioning pituitary tissue, it is likely that Gamma Knife will
replace surgery as primary treatment for microadenomas. Chiasmal compression will continue
to require surgical intervention.
Craniopharyngiomas
Craniopharyngiomas continue to be a major surgical problem. The only hope for cure is total
surgical removal at the first operation. Such surgery has a high mortality and morbidity, is only
possible in about 10% of cases and the recurrence rate even after ‘total’ removal is 19% (28).
Conventional fractionated radiotherapy improves the ten year survival rates from 40% to 75%
(29), but in children often causes impaired growth with mental and sexual impairment.
Use of the Gamma Knife alone has produced an 87% tumour control rate in 31 patients
observed for 1-7 years (30), but larger groups and longer follow up are needed for evaluation of
results. Gamma Knife has more often been used for recurrence after failure of conventional
fractionated radiation. In these cases there is greater danger of radiation damage to the optic
chiasm and to any residual pituitary function. It is better therefore to proceed straight to
Gamma Knife and avoid radiotherapy.
Management of this condition is very complex. Radiosurgery often needs to be combined with
stereotactic cyst aspiration and/or installation of Yttrium (31) or bleomycin (32). Such
multimodality approach requires a team with experience of microsurgery, radiosurgery,
stereotaxy and radiotherapy.
OTHER SKULL BASE TUMOURS
Tumours that infiltrate the skull base are often even more difficult to manage surgically than
meningiomas. The spectrum includes chordoma, chondrosarcoma, glomus tumour,
haemangiopericytoma, paranasal sinus carcinoma, juvenile angiofibroma,
esthesioneuroblastoma and metastases.
En bloc excision is potentially curative and so must be considered where possible. The
majority of cases require management in other ways however and Gamma Knife is one of the
most effective.
Haemangiopericytoma responds very rapidly and usually disappears, however it has a
tendency for late recurrence and patients must be followed. Chordoma and chondrosarcoma,
although ‘radioresistant’, have been controlled in 73% of cases (33). Metastases respond as
expected and described above. Glomus jugulare tumours, where surgical resection is usually
complex with many complications and recurrence is common, seem to respond very well.
100% control with 64% improvement of symptoms has been reported (34). Other conditions
have been the subject of case reports and small series. Treatment is worth considering but
large series are not available yet to predict success.
FUNCTIONAL DISORDERS
Epilepsy
In the series of patients with AVM treated at Karolinska by Gamma Knife it was observed that
52 out of 59 cases with epilepsy were cured of their fits even though the AVM was not always
completely obliterated. This led to the concept that radiosurgery may modify epileptogenic
activity and attempts were made to ablate the epileptic focus in other patients where it could be
defined.
The largest experience is in mesial temporal sclerosis. Regis et al (35) performed
amydalohippocampectomy with the Gamma Knife and after this 81% of patients were seizure
free at 2 years or more follow up. Ablation of the whole area of amygdala and hippocampus
requires a large radiation dose and risks a local radiation reaction. Regis has concentrated
recently on lesioning the parahippocampal gyrus, with satisfactory results and reduced
complications. The results will be published shortly.
Research using magnetoencephalography, functional MRI and PET to localize epileptic activity
is likely to increase the proportion of patients found to have focal epilepsy. These imaging
methods allow stereotactic targeting of foci and the role of the Gamma Knife is likely to become
more important.
Trigeminal Neuralgia
Figure 10. The lesion produced in the fifth nerve (left) and the 4mm collimator
plan used to produce it (right).
A Gamma Knife lesion at the root entry zone of the fifth nerve (fig. 10) offers a high chance of
improvement in drug resistant cases and avoids most of the morbidity of invasive surgical
approaches. A multicentre study (36) showed 58% pain free, 36% improved and 6%
unchanged. The risk of facial dysaesthesia is 2-4%. If the radiation dose is increased or
repeated the cure rate is much higher but so is the risk of mild dysaesthesia. A recent study
with a 90Gy dose gave a cure rate of 73% but a 16% incidence of dysaesthesia (37). Giving a
second dose to failed cases resulted in an overall cure rate of 98% but all of those given a
second dose suffered dysaesthesia (38). An even more recent study has identified the plexus
triangularis, 5-8mm from the brain stem, as a more appropriate target with a cure rate of 83%
and dysaesthesia 4% (39).
Post-herpetic trigeminal neuralgia, which does not respond to other surgical treatment (except
brain stem DREZ lesions), has been shown to respond to the gamma knife in 44% of cases
(40). Since the overall results of Gamma Knife radiosurgery can be better than other
techniques it is likely to become progressively more popular.
Parkinson’s disease and other dyskinesias
There has been renewed interest in surgical intervention to treat Parkinson’s disease. This is
due to the realization that many patients become refractory to levodopa after long-term use,
combined with the occurrence of dyskinetic movements associated with Sinemet. The classical
approach to stereotactic neurosurgery for Parkinson’s disease has been to perform lesions by
placing an electrode in the appropriate target and heating it by radiofrequency current. The
main complications are intracerebral haemorrhage and neurological damage due to misplaced
lesions. The use of stereotactic deep brain stimulation has fewer cerebral complications but
more practical problems associated with the implanted electrode and stimulator.
The Gamma Knife was originally designed and intended as an instrument for thalamotomy, but
only recently have large series with adequate follow up been reported. In a series of 158
gamma thalamotomies for tremor the cure rate was 88% with results maintained at 4 years
follow up (41). From the same center the relief of Sinemet induced dyskinesia by pallidotomy
was 85% and pallidotomy also improved two thirds of patients with rigidity and bradykinesia.
Two patients (1.3%) had a mild lesion induced neurological deficit.
The effectiveness of Gamma Knife treatment is the same as a traditional open stereotactic
procedure but with much lower complications. The control of dyskinesia is delayed for 6-18
months from treatment.
Chronic pain
In general destructive lesions for the treatment of pain are to be discouraged, but it seems
unreasonable to withhold the benefit of medial thalamotomy in cases of cancer pain where all
else has failed. 65% of unilateral pain cases are effectively controlled in this way.
Hypophysectomy is also useful in bone pain from Ca breast metastases. Both thalamotomy
and anterior cingulotomy can benefit chronic pain of non-malignant aetiology but the indications
and results are not yet clarified.
Obsessive-compulsive neurosis
An unscientific backlash against stereotactic psychosurgery has effectively halted operations in
many countries but this is unjustified. In the subgroup of obsessive-compulsive anxiety it is
well known that limbic lesions can rehabilitate socially destroyed patients. Many studies have
shown that 60% of stereotactic anterior capsulotomy procedures are successful (42) and this
is in agreement with my personal experience of 60 cases of subcaudate tractotomy and
anterior cingulotomy performed in Bristol (43). It has been suggested that a single right-sided
lesion may be adequate (42). The procedures are greatly underrated and a placebo controlled
double blind study is currently underway in Sweden and USA using the Gamma Knife. The
results should cause a change in medical and social prejudice.
VASCULAR CONDITIONS
Arteriovenous malformations
Figure 10. A thalamic arteriovenous malformation before and two
years after Gamma Knife treatment.
There has been a very long experience of Gamma Knife treatment of AVM because accurate
stereotactic localization was possible from angiography before the advent of CT and MRI. The
first lesion was treated in 1972. This large experience has made it possible to develop
computer programs to predict the results of treatment (44) and in particular the probability of
total occlusion, the risk of radiation damage and the risk of haemorrhage before occlusion
occurs. It is therefore possible for all units to calculate the dose that has the greatest chance of
success and the lowest rate of complications.
Favourable factors for success of radiosurgery are lower flow, compact or plexiform lesions,
multiple shunts or fistulas and few draining veins. Gamma Knife is particularly indicated when
lesions are inaccessible to surgery and embolisation, or when a residual nidus remains after
other treatment. It is also important to stress that the risk of radiosurgery is less than other
methods and so it is appropriate as a primary treatment unless contraindicated, for example a
single high flow fistula is best managed endovascularly.
Many groups have published similar results (44-46). Total obliteration in lesions less than 3cm
diameter should be 40% at one year, 80% at 2 years and 90% at 3 years. Smaller lesions
have a higher success rate and lower complication rate. The risk of local oedema due to
radionecrosis is 2-4% and the risk of rebleeding within 2 years is also 2-4%. Associated
epilepsy is usually abolished. Success and risk are strongly related to total dose and lesion
size, rebleeding is inversely related to dose. For this reason smaller lesions, which can safely
tolerate a higher dose, are more effectively treated.
The indications for microsurgery are certainly very limited now; mainly compact lesions at or
near the convexity of the brain in non-eloquent areas. However it is not possible to say any
one treatment is dominant. It is often necessary, especially in large lesions, to rely on
multimodality treatment (46). In particular, embolisation followed after 3 months by Gamma
Knife (to allow time for collaterals to open) has been a productive combination. It is evident
that, for the best results, patients should be evaluated by a team experienced in microsurgery,
embolisation and Gamma Knife surgery.
Cavernous angiomas (cavernomas)
Cavernomas are vascular malformations with well-defined margins and a surrounding
haemosiderin deposit, but no angiographically demonstrable lesion. Often these are
asymptomatic lesions that require no therapy. Recurrent bleeding or other symptoms may be
cured by surgical excision if they are easily accessible but frequently they are not.
Two large series have established radiosurgery as the optimal therapy. Kondziolka et al. (48)
reported 47 patients with bleeding cavernomas and noted a significant reduction in risk of
bleeding from 32% per year to 9% per year. Kida et al. (49) analysed 51 cases presenting
either as haemorrhage or intractable epilepsy. Haemorrhage was controlled in 86% and
epilepsy was controlled in 64%. Permanent complications were 5%. The complication rate for
radiosurgical treatment of cavernomas is unexpectedly high and this has been attributed to an
amplification or radiosensitisation effect caused by the haemosiderin ring (50). For this reason
doses should be less than for AVM of the same size and the haemosiderin ring should be
excluded from the treatment plan.
ORBITAL LESIONS
Uveal melanoma
Enucleation of the eye does not prevent metastases and there has been a suggestion that it
may encourage them. This has led to a search for treatments that conserve the eye and vision,
control the tumour growth and preserve life. These include proton irradiation, Iodine-125
brachytherapy and Ruthenium plaque insertion all of which have practical problems and
complications.
Gamma knife therapy has been used in certain centers with an opthalmological interest for over
10 years. Initial high dose treatments were accompanied by severe side effects such as
neovascular glaucoma (51). Reducing the dose has allowed the same rate of control (84%)
and a reduced risk of glaucoma (16%). In another series there was 92% control (52). Over a 4
year follow up 13% developed metastases and 10% died of their disease.
Other orbital lesions
The feasibility of treating other conditions depends on the need to preserve vision by avoiding
significant doses to the lens and optic nerve. Tumours such as haemangiomas, optic sheath
meningiomas and optic nerve gliomas could be treated if the option of surgery is inappropriate.
Recent experience in the management of very severe glaucoma has been encouraging with
the pain being controlled in all cases and the mean intra-ocular pressure dropping from
40mmHg to 22.5mmHg (53).
COMPLICATIONS
It has been emphasized above in relation to particular lesions that the complication rate after
Gamma Knife treatment is significantly less than after microneurosurgery, radiotherapy or
LINAC based radiosurgery. However the Gamma Knife is a very powerful tool with potential
complications and needs careful and experienced handling. The complications that exist are
mainly minor and transient; they may be immediate or delayed.
Immediate side effects are moderate headache, skin hypersensitivity if the target is near the
surface and acute cranial nerve disturbances (e.g. dizziness and vomiting after vestibular
schwannoma treatment).
Delayed complications are mainly local swelling in or around the lesion 6-9 months after
treatment (ARE – adverse radiation effect). This is dose related and can usually be controlled
easily by steroids. The incidence is 2-10% depending on the target. Very occasionally
radionecrosis occurs and may require decompressive surgery. This is usually because a larger
than safely tolerated dose has been given in malignant disease and is therefore expected.
The complication that worries some observers is the possibility of radiation induced tumours.
Because it is well known that radiotherapy can cause brain tumours it has been assumed that
Gamma Knife can do the same. Radiotherapy for tinea capitis is well known to predispose to
meningiomas (54) and fractionated radiotherapy for pituitary adenomas increases the risk of
other local tumour formation by nine times compared with the normal population (55). There
are reasons to believe that radiosurgery is less likely to induce tumours than radiotherapy but,
up to now, there have been four cases in the over 200,000 patients treated by Gamma Knife
(56). Because the delay to presentation may be as much as 30 years it is likely that this
proportion will rise. However this miniscule risk, similar to frequent air travel, is unlikely to
affect the indications for Gamma Knife treatment.
CONCLUSION
Assuming that the target is less than about 3.5 cm, it has been shown that the Gamma Knife
can control a large proportion of the neurosurgical pathologies that occur in the brain. As in
every branch of medicine, correct selection of patients and judicious, quality assured, use of
equipment are of paramount importance.
In addition Gamma Knife is more economic than both LINAC radiosurgery and
microneurosurgery. The overall costs of Gamma Knife in a European environment (Austria)
are about 5% less than those of a dedicated LINAC (57). Gamma Knife is also significantly
more cost effective than microsurgery in both Europe and USA. For acoustic neuroma
treatment Gamma Knife costs less than half (41%) of surgical treatment (58). For the
management of metastases the cost is three quarters (74%) of surgical treatment (59).
Better control of malignant tumour growth is achieved using Gamma Knife. There is a 2.84
times greater chance of local progression of primary brain tumours and metastatic tumours
after LINAC radiosurgery as compared with Gamma Knife (2). For benign tumours the growth
control rate is similar in all radiosurgery treatments.
The most important advantage is that there are fewer complications than with microsurgery.
The treatment mortality is zero. No equivalent surgical operation can claim this. All studies
that have compared the incidence of more minor complications between Gamma Knife and
microsurgery have shown a huge reduction of complications by using Gamma Knife
radiosurgery (17,59).
There are also fewer complications than when using LINAC for radiosurgery. Because the
planning is more accurate, the risk is less when using the Gamma Knife, especially in areas
near the brain stem, skull base, cranial nerves and critical functional areas. Studies have
shown that the normal tissue complication probability (NTCP) is 2-4 times higher if a typical
LINAC is used (1). In practice the risk is even greater if the target is near the brain stem or
skull base.
Gamma Knife treatment has been subjected to a more intense and comprehensive follow up
than most contemporary treatment methods. Almost all of the results summarized above are
better than those of alternative treatments, despite the fact that the majority of cases were
managed using the earlier ‘U’ and ‘B’ models and with dose planning systems that were
primitive by today’s standards. When long term follow up of future series using the ‘C’ model,
the highly conformal ‘Leksell Gamma Plan’ and multi isodose treatment using the APS
(automatic positioning system) come to be analysed in a few years time, the results will
inevitably be even better.
References
1.
2.
3.
4.
5.
6.
Smith V, Verhey L, Serago CF. Comparison of radiosurgery treatment modalities based on complication and
control probabilities. Int J Radiat Oncol Biol Phys 1998; 40: 507-513.
Shaw E, Scott C, Souhami L, Dinapoli R, Kline R, Loeffler J, Farnan N. Single dose radiosurgical treatment
of recurrent previously irradiated primary brain tumours and metastases: Final report of RTOG protocol 9095. Int J Radiat Oncol Biol Phys 2000; 47: 291-298.
Kida Y, Kobayashi T, Mori I et al. Gamma knife radiosurgery for low-grade astrocytomas: results of long term
follow up. J Neurosurg 2000; 93(suppl 3): 42-46.
Stea B, Rossman K, Kittleson J et al. Interstitial irradiation vs interstitial thermo-radiotherapy for
supratentorial malignant glioma. A comparative survival analysis. Int J Radiat Oncol Biol Phys 1994; 30: 591600.
Sneed PK, Gutin PH. Interstitial radiation and hyperthermia. In Berger MS, Wilson CB (eds), The Gliomas.
WB Saunders, Philadelphia, 2000, pp499-510.
Kondziolka D, Flickinger JC, Lunsford L Dade. Stereotactic radiosurgery for glial neoplasms. Prog Neurol
Surg, Basel, Karger, 1998; 14: 160-174.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
Tovi M, Lilja A, Bergstrom M et al. Deliniation of gliomas with magnetic resonance imaging using Gd-DTPA
in comparison with computed tomography and positron emission tomography. Acta Radiol 1990; 31: 417429.
Muacevic A, Kreth FW, Horstmann GA et al. Surgery and radiotherapy compared with gamma knife
radiosurgery in the treatment of solitary cerebral metastases of small diameter. J Neurosurg 1999; 91: 35-43.
Kondziolka D, Patel A, Lunsford LD et al. Stereotactic radiosurgery plus whole brain radiotherapy versus
radiotherapy alone for patients with multiple brain metastases. Int J Radiat Oncol Biol Phys 1999; 45: 427434.
Young RF, Jaques DB, Duma C et al. Gamma knife radiosurgery for treatment of multiple brain metastases.
In Alexander E, Kondziolka D, Loeffler J (eds) Radiosurgery, Basel, Karger, 1996; pp92-101.
Yamanaka K, Iwai Y, Nakajima H et al. Gamma knife radiosurgery for metastatic brain tumour: the usefulness
of repeated gamma knife radiosurgery for recurrent cases. Stereotact Funct Neurosurg 1999; 72 (suppl 1):
73-80.
Serizawa T, Iuchi T, Ono J et al. Gamma knife treatment for multiple metastatic brain tumours compared with
whole brain radiation therapy. J Neurosurg 2000; 93 (suppl 3): 32-36.
Samii M, Matthies C. Management of 1000 vestibular schwannomas (acoustic neuromas): surgical
management and results with an emphasis on complications and how to avoid them. Neurosurgery 1997; 40:
11-23.
Pollock BE, Lunsford LD, Noren G. Vestibular schwannoma management in the next century: a radiosurgical
perspective. Neurosurgery 2000; 43: 475-483.
Torrens M, Maw R, Coakham H, Butler S. Facial and acoustic nerve preservation during excision of
extracanalicular acoustic neuromas using the suboccipital approach. Br J Neurosurg 1994; 8: 655-665.
Iwai Y, Yamaka K, Shiotani M, Vyama T. Radiosurgery for acoustic neuromas: results of low dose treatment.
Neurosurgery 2003; 53: 282-287.
Regis J, Pellet W, Delsanti C et al. Functional outcome after gamma knife surgery or microsurgery for
vestibular schwannomas. J Neurosurg 2002; 97: 1091-1100.
Kane NM, Torrens MJ, Kazanas S et al. Functional outcome in patients after excision of intracanalicular
acoustic neuromas using the suboccipital route. Ann Roy Coll Surg Eng 1995; 77: 210-216.
Kondziolka D, Lunsford LD, Flickinger JC. Current concepts in gamma knife radiosurgery. Neurosurg Quart
1993; 3: 253-271.
Prasad D, Phillips CD, Steiner M, Steiner L. Outcomes for gamma knife surgery in cavernous sinus
meningiomas. Proc 8th Leksell Gamma Knife Society Meeting 1997, p22.
Lunsford LD. Long term outcomes after gamma knife radiosurgery for cavernous sinus meningiomas. Proc
11th Leksell Gamma Knife Society Meeting 2002, p73.
Advances in skull base surgery – debate. EANS meeting, Lisbon 2003.
Inoue HK, Kohga H, Hirato M et al. Pituitary adenomas treated by microsurgery with or without gamma knife
surgery: experience in 122 cases. Stereotact Funct Neurosurg 1999; 72 (suppl 1): 125-131.
Kim SH, Huh R, Chang VW et al. Gamma knife radiosurgery for functioning pituitary adenomas. Stereotact
Funct Neurosurg 1999; 72 (suppl 1): 101-110.
Liscak KR, Vladyka V, Marek J. Pituitary radiosurgery. Proc EANS, Lisbon 2003, p305.
Zhang N, Pan L, Wang EM et al. Radiosurgery for growth hormone producing pituitary adenomas. J
Neurosurg 2000; 93 (suppl 3): 6-9.
Landolt AM, Haller D, Lomax N et al. Stereotactic radiosurgery for recurrent surgically treated acromegaly: a
comparison with fractionated radiotherapy. J Neurosurg 1998; 88: 1002-1008.
Samii M, Tatagiba M. Surgery of craniopharyngiomas. In Kaye AH, Black PMcL (eds) Operative
Neurosurgery, London, Churchill Livingstone, 2000:741-749.
Manaka S, Treamoto A, Takakura K. The efficacy of radiotherapy for craniopharyngioma. J Neurosurg 1985;
62: 648-656.
Chung W, Pan D, Shiau C et al. Gamma knife radiosurgery for craniopharyngiomas. J Neurosurg 2000; 93
(suppl 3): 47-56.
Backlund E-O. Treatment of craniopharyngiomas: the multimodality approach. Paediat Neurosurg 1994; 21
(suppl 1): 82-89.
Makry M. Craniopharyngiomas: a six year experience with gamma knife radiosurgery. Stereotact Funct
Neurosurg 1999; 72 (suppl 1): 140-149.
Muthukumar N, Kondziolka D, Lunsford LD. Sterotactic radiosurgery for other skull base lesions. Prog
Neurol Surg, Basel, Karger 1998; 14: 128-144.
Leber KA, Eustacchio S, Pendl G. Radiosurgery of glomus tumours: midterm results. Stereotact Funct
Neurosurg 1999; 72 (suppl 1): 53-59.
Regis J, Bartolomel F, Rey M et al. Gamma knife surgery for mesial temporal lobe epilepsy. J Neurosurg
2000; 93 (suppl 3): 141-146.
36. Kondziolka D, Lunsford LD, Flickinger DC. Stereotactic radiosurgery for trigeminal neuralgia: a multiinstitutional study using the gamma unit. J Neurosurg 1996; 84 940-945.
37. Nicol B, Regine WF, Courtney C et al. Gamma knife radiosurgery using 90Gy for trigeminal neuralgia. J
Neurosurg 2000; 93(suppl3): 152-154.
38. Pollock BE, Foote RL, Stafford SL et al. Results of repeated gamma knife radiosurgery for medically
unresponsive trigeminal neuralgia. J Neurosurg 2000; 93 (suppl 3): 162-164.
39. Massager N, Lorenzoni J, Devriendt D et al. Gamma knife radiosurgery for idiopathic trigeminal neuralgia:
importance of the irradiation dose received by the brainstem. Proceedings EANS (Lisbon) 2003, p127.
40. Urgosik D, Vymazal J, Vladyka V, Liscak R. Treatment of post herpetic trigeminal neuralgia with the gamma
knife. J Neurosurg 2000; 93 (suppl 3): 165-168.
41. Vouna RF, Jaques S, Mark R et al. Gamma knife thalamotomy for treatment of tremor: long term results. J
Neurosurg 2000; 93 (suppl 3): 128-135.
42. Lippitz B, Mindus P, Meyerson B, Kihlstrom L, Lindqist C. Lesion topography and outcome after
thermocapsulotomy or gamma knife capsulotomy for obsessive compulsive disorder: relevance of the right
hemisphere. Neurosurgery 1999; 44: 452-460.
43. Crow H, Phillips D, Torrens M. Unpublished observations.
44. Karlsson B. Gamma knife surgery of cerebral arteriovenous malformations. Thesis, Karolinska Institute,
Stockholm 1996.
45. Pollock BE. Patient outcomes after arteriovenous malformation radiosurgery. In Lunsford LD, Kondziolka D,
Flickinger JC (eds) Gamma knife brain surgery. Prog Neurol Surg, Basel, Karger 1998; 14: 51-59.
46. Chang JH, Chang JW, Park YG et al. Factors related to complete occlusion of arteriovenous malformations
after gamma knife surgery. J Neurosurg 2000; 93 (suppl 3): 96-101.
47. Chang SD, Marcellus ML, Marks MP et al. Multimodality treatment of giant intracranial arteriovenous
malformations. Neurosurgery 2003; 53: 1-13.
48. Kondziolka D, Lunsford LD, Flickinger JC et al. Reduction of haemorrhage risk after stereotactic radiosurgery
for cavernous malformations. J Neurosurg 1995; 83: 825-831.
49. Kida Y, Kobayashi T, Tanaka T. Radiosurgery of symptomatic angiographically occult vascular
malformations with gamma knife. In Kondziolka D (ed), Radiosurgery, Basel, Karger 1995; 1:207-217.
50. St George EJ, Perks J, Plowman PN. Stereotactic radiosurgery XIV: the role of the haemosiderin ‘ring’ in the
development of adverse reactions following radiosurgery for intracranial cavernous malformations: a
sustainable hypothesis. Br J Neurosurg 2002; 16: 385-391.
51. Langmann G, Pendl G, Papaefthymiou G, Feichtinger KH. High versus low(er) dose in radiosurgery for uveal
melanomas. Proc Leksell Gamma Knife Soc, 11th Internat Meeting (Prague) 2002; p107.
52. Marchini G, Bonadimani M, Rigotti M et al. Gamma knife radiosurgical treatment of uveal melanomas:
analysis of mid- to long-term results. Proc Leksell Gamma Knife Soc, 11th Internat Meeting (Prague) 2002;
p110.
53. Vladyka V, Liscak R, Subrt O et al. Initial experience with gamma knife radiosurgery for advanced glaucoma.
J Neurosurg 2000; 93 (suppl 3): 180-183.
54. Sadetski S, Flint-Richter P, Ben-Tai T, Nass T. Radiation induced meningioma: a descriptive study of 253
cases. J Neurosurg 2002; 97: 1978-1082.
55. Brada M, Ford D, Ashley S et al. Risk of second brain tumour after conservative surgery and radiotherapy for
pituitary adenoma. BMJ 1992; 304: 1343-1346.
56. Loeffler JS, Niemierko A, Chapman PH. Second tumours after radiosurgery: tip of the iceberg or a bump in
the road? Neurosurg 2003; 52: 1436-1442.
57. Konigsmaier H, de Pauli-Ferch B, Hackl A, Pendl G. The costs of radiosurgical treatment: a comparison
between gamma knife and linear accelerator. Acta Neurochir (Wien) 1988; 140: 1101-1111.
58. van Roigen L, Nijs HG, Avezaat CJ Lindquist C, Pauw KH, Rutten FF. Costs and effects of microsurgery
versus radiosurgery in treating acoustic neuroma. Acta Neurochir (Wien) 1997; 139:942-948.
59. Rutigliano MJ, Lunsford LD, Kondzoilka D, Strauss MJ, Khanna V, Green M. The cost effectiveness of
stereotactic radiosurgery versus surgical resection in the treatment of solitary metastatic brain tumours.
Neurosurgery 1995; 37: 445-453.
TABLE 1.
CONDITIONS TREATABLE BY GAMMA KNIFE RADIOSURGERY
PRIMARY BRAIN TUMOURS
Low grade astrocytomas, high grade astrocytomas, haemangioblastomas, central
neurocytoma, choroid plexus papilloma, germinoma, pineal tumours.
METASTASES
NEUROMAS
Acoustic, facial, trigeminal, jugular fossa.
MENINGIOMAS
PITUITARY TUMOURS
Non-secretory adenomas, hormone secreting adenomas, craniopharyngiomas, lymphocytic
hypophysitis.
OTHER SKULL BASE TUMOURS
Chordoma, chondrosarcoma, glomus tumours, haemangiopericytoma,
haemangioendothelioma, paranasal sinus carcinoma, juvenile angiofibroma,
esthesioneuroblastoma, metastases.
FUNCTIONAL DISORDERS
Epilepsy, trigeminal neuralgia, Parkinson’s disease and other dyskinesias, chronic pain,
obsessive compulsive neurosis.
VASCULAR LESIONS
Arteriovenous malformations, dural A-V fistula, cavernous angiomas.
ORBITAL LESIONS
Uveal melanoma, glaucoma, orbital haemangioma, optic sheath meningioma, optic nerve
glioma, sub-foveal neovascularisation.