Gamma Knife Radiosurgery for Acoustic Neuromas

by Georg Noren M.D., Ph.D.

Director, New England Gamma Knife Center
Rhode Island Hospital, Providence, RI USA

This article appeared in The Connection, a quarterly publication of the Acoustic Neuroma Association of Canada (Vol. 8 Issue 4 December 15, 1995).  It was contributed to the Acoustic Neuroma Archive by Lisa Probst.


Gorgen.gif (22252 bytes)A little bit of history.

The first successful removal of an acoustic neuroma was performed in 1894. Excision of the tumor was the standard treatment and, for many years, the only available option. The results improved gradually but were still far from satisfactory in the early 60s when microsurgical techniques were gradually introduced into this field by otosurgeons and neurosurgeons. In 1951, the Swedish neurosurgeon Lars Leksell presented the idea of having a large number of converging beams of ionizing radiation (like X-ray) cross-fire targets in the brain. He coined the term radiosurgery to describe this concept, since the way radiation was used differed greatly from conventional radiotherapy. He suggested radiosurgery for the treatment of deep-seated, circumscribed intracranial tumors. This represented a path of development parallel to that of microsurgery. The first device for routine clinical use, based on this idea, was the prototype Gamma Knife constructed in 1967-68. The first acoustic tumor was treated in June 1969 at the Karolinska Hospital in Stockholm, Sweden. Since then, almost 600 acoustic tumor patients have been treated there and close to 4000 worldwide. One of the most characteristic features of the Gamma Knife is the high degree of accuracy down to half a millimeter or even less.

The Gamma Knife

This is an 18-ton machine with 201 permanently mounted Cobalt 60 sources arranged spherically around the patient's head. These sources emit gamma radiation which is the same as X-ray (not laser as frequently assumed) but more powerful, with higher energy. These beams are precisely shaped through two consecutive sets of tungsten channels (collimators) and they all aim at one point, the focus. Here, at the point of intersection, the radiation is potentially very powerful whereas each beam, on its way through the skull, is weak and will not cause any detectable biological effects. The gamma radiation does not act by inducing heat but by deranging molecules in the tumor cells so that they can no longer duplicate and eventually will die. One of the most characteristic features of the Gamma Knife is the high degree of technical accuracy, down to half a millimeter or even less. Another is the steep gradient (quick fall off) at the border of each shot of radiation. This means that a high dose of radiation can be delivered to targets with the minimum of harm to important sensitive structures just a fraction of a millimeter away or even adjacent to the surface.

Stereotactic radiosurgery is performed by a team composed of neurosurgeons, radiation oncologists, medical physicists and a nursing staff. Specialists in neuroimaging join the team when required.

The Gamma Knife Procedure

1. The ring is attached
The patient is fully awake throughout the procedure. Following light sedation, the patient's head is scrubbed. No hair is removed. Two spots in the forehead and two in the back of the head are made numb by injecting local anaesthetic. An aluminum ring is subsequently attached to the head by tightening four pointed screws which will penetrate through the skin and stop against the outer surface of the skull bone. There is consequently no drilling and no holes are created in the bone. The holes in the skin are tiny, heal quickly and become invisible very soon after the treatment. The ring (stereotactic frame) is now firmly attached to the head. This first step takes 15-20 minutes.

2. Magnetic resonance imaging (MRI)
Next follows an MRI with contrast injection. Several series of images are obtained with different projections and techniques (sequences).  Of importance is to get as much and detailed information as possible about the tumor and the surrounding structures. This stereotactic MRI will usually last for one hour.

3. A plan for the treatment
The third step involves the transfer of these images to a dedicated computer system so that a plan (dose plan) for the treatment can be created. In fact, the configuration of this plan, and consequently of the radiation, will be an upmost exact copy of the tumor itself. This is achieved by adding a number of shots of radiation of the same or different sizes to each other, thereby, step by step, completing the coverage of the tumor. It normally takes 1-2 hours to create a plan but sometimes more if the tumor is large or has a very irregular shape. Before going ahead with the treatment, the patient is offered a demonstration of the final plan on the computer screen.

4. The treatment
The final step is the Gamma Knife treatment. The numbers (x, y, and z coordinate values) for the different positions shots are set consecutively according to the final plan, and the patient's head is positioned in the collimator helmet. This means that the patient has to go in and out of the unit as many times as there are shots prescribed in the plan. For acoustic tumors the patient will always be positioned lying supine (face up). The total time of irradiation is frequently 20 to 40 minutes. The total treatment time is usually 1-2 hours when including the preparations between the different shots. The irradiation is painless but sometimes the pressure of the frame may give rise to some headache or nausea. This is easily relieved with medication if required.

The frame is removed as soon as the treatment is finished.  Dressing is applied temporarily and the patient stays in the hospital for a couple of hours or overnight. There is frequently some temporary local swelling around and below the pin sites in the forehead. This does not require any special care.  In most instances, the patient will be able to go back to work within one, two, or three days.

The follow_up

The patient will be followed clinically after the treatment with special emphasis on cranial nerve function (facial sensory and motor function, hearing, balance).  MRI is necessary to detect shrinkage and other changes of the tumor. Detailed information about hearing is obtained through tone and speech audiometries. All examinations are normally repeated 1/2, 1, 2, 3, and 4 years after the treatment.

What result could be expected

1. The tumor itself
Radiosurgery has a less direct effect than microsurgery. The tumor is not removed, the main goal of the treatment is rather to achieve arrest of growth. Since this is a very benign type of tumor, total eradication does not have to be attained. Growth control (meaning shrinkage or unchanged size) is actually found in approximately 95% of the unilateral cases and in 85% of the patients with the bilateral, inherited type of the disease usually referred to as neurofibromatosis 2 (NF2).  Shrinkage, a more evident result of the radiosurgical treatment than unchanged size, is found in about 1/3 of the unilateral tumors one year after the treatment, in 2/3 after four years, and in more than 90% after 10 years following the procedure.

Unilateral acoustic neuromas showing shrinkage and tumors with no change of size over a period of five years or more do not recur. The NF2 tumors seem to be more elusive and can show signs of re-growth at any time, even 10 years or more after the treatment. These NF2 neuromas also show a higher recurrence rate after microsurgery.  Five percent of the unilateral and 15% of the NF2 tumors do not respond to the treatment and have to be treated a second time (radiosurgery or microsurgery whichever is preferred).  In another 5-7% of the patients in either group the tumor shows signs of temporary swelling as an early response to the treatment. The peak is reached in the period 6-18 months after the treatment. In these cases, with increase of the tumor size, a definite assessment, whether the reason is temporary swelling or lack of response, can normally be made within two years of the treatment.

2. Cranial nerves
There is today, at some Gamma Knife Centers, a total incidence of temporary facial and trigeminal nerve dysfunction as low as less than 2%.  Hearing remains unchanged, or almost unchanged, in 75% of the patients over the first several years. The long term preservation (>5 years), where the treatment was based on currently used high resolution imaging and an extremely accurate dose-planning system, is still largely unknown. The balance nerve function is less well preserved but clinically important dizziness and other symptoms of vestibular nerve dysfunction are rare.

3. CSF circulation
Impairment of the cerebrospinal fluid (CSK) circulation, leading to hydrocephalus necessitating a shunt operation, may occur as a result of mechanical blockage or changed quality of the CSF induced by the tumor itself. Hydrocephalus may also develop following temporary swelling around the tumor induced by the treatment.  Hydrocephalus directly caused by the tumor is seen in 10% of the patients, almost half of them requiring a shunt before radiosurgical procedure. Peritumoral swelling giving rise to a blockage of the CSF circulation, requiring a shunt, is found in less than 0.5% of the patients. It is well-known that some acoustic tumors, primarily those with a size of 2 cm and up and, according to our experiences, in patients more than 60 years of age, tend to induce hydrocephalus spontaneously before they have been treated. The exact reason for this is unclear but some of  these tumors cause a marked increase of the CSF protein concentration, tenfold or more, which is known to be of importance for the development of hydrocephalus.

Radiosurgery or microsurgery?

Stereotactic radiosurgery has a number of evident advantages over microsurgery such as no mortality, no risk of intracranial bleeds or infections, short inpatient time, and almost no recovery period. These features are in (to be continued).

Georg Noren M.D., Ph.D.

Director, New England Gamma Knife Center
Rhode Island Hospital
593 Eddy Street, Grosvenor Bldg 123
Providence, Rhode Island 02903 USA

Associate Professor, Department of Neurosciences
Brown University School of Medicine

Last Edited: Wednesday, October 30, 2002