Clinical Uses of Radiosurgery
Clinical Uses of Radiosurgery
The term "radiosurgery" describes several related
techniques that combine stereotactic approaches to achieve precise three-dimensional
localization of highly focused, high-energy radiation. This
procedure allows for the delivery of a very large dose of radiation
to a small intracranial lesion while achieving rapid fall-off in dose
outside the target volume. Since Leksell pioneered the technique in
1951, radiosurgery has proven to be an effective alternative to
conventional neurosurgery or brachytherapy in the management of
selected small intracranial tumors. It is also an alternative to
surgery and/or intravascular embolization for the treatment of
patients who have arteriovenous malformations (AVMs).
Currently, there are three broad types of radiosurgical devices,
which are distinguishable from one another by the type and source of
high-energy radiation utilized. The gamma knife uses a hemispherical
array of radioactive cobalt-60 sources that emit gamma-ray photons.
With this device, 201 fixed, separate beams precisely converge on an
isocenter. Selection of a subset of these beams modifies the shape of
the irradiated volume to more closely match a specific target.
Linear accelerator-based radiosurgery uses x-ray photons that are
produced in a collision between accelerated electrons and a metal
target. Conventional linear accelerators are mounted on a gantry that
has a fixed axis of rotation. By rotating both the linear accelerator
and treatment couch, multiple intersecting arc fields can be produced
that concentrate the radiation dose at a precise isocenter.
The third form of radiosurgery uses heavy particles, such as protons
or heliumions, generated by massive accelerators, such as cyclotrons
or synchrotrons. These positively charged particles deliver most of
their radiation dose at a set depth in tissue that is determined by
energy. This allows for the delivery of high doses to small targets
with little radiation dose to surrounding normal tissues.
All three radiosurgical instruments utilize an external frame of
reference, which is attached to the patients head. This device
is needed for spatial localization of the target and cranial
immobilization during treatment.
This article surveys the clinical results and complications of
radiosurgery for the treatment of metastatic tumors, primary gliomas,
AVMs, and benign conditions, including pituitary tumors, acoustic
neuromas, and meningiomas. These results are derived from a
cross-section of published data and our experience treating patients
with these tumors at Stanford University Medical Center.
The treatment of brain metastases is the most common application of
radiosurgery due to the high incidence of this disease. The American
Cancer Society estimates that 170,000 cancer patients develop
cerebral metastases each year in the United States. Because most
patients with metastases eventually succumb to their underlying
malignancy, the primary benefits of radiosurgery are palliation of
symptoms and some modest prolongation of survival.
Radiosurgery produces median survival (6.4 to 10 months)[3-10]
comparable to that after surgical resection followed by conventional
fractionated radiotherapy (4 to 13 months),[11,12] with good local
control (Table 1 and
Figure 1).[3-9,13,14] Two studies have shown that patients
treated with radiosurgery for either one or two brain metastases have
prolonged survival equivalent to that achieved with surgical resection.[15,16]
Selection criteria for radiosurgery generally include: (1) limited
number of brain metastasis (typically three or less), (2) maximal
tumor diameter < 3.5 cm for each lesion, (3) suitable target
shape, and (4) absent or stable disease at the primary and other
extracranial sites. Surgical resection may be preferable to
radiosurgery for the treatment of brain metastases in two clinical
situations: when the primary source is unknown, and when there are
significant symptoms attributable to mass effect and edema unrelieved
by corticosteroids. Surgical resection typically relieves symptoms
more promptly than does radiosurgery.
Radiosurgery is preferable to surgery in patients whose medical
condition precludes craniotomy, those with tumors located in eloquent
cerebral regions, and those who refuse surgery. Radiosurgery avoids
the perioperative morbidity and mortality that craniotomy entails.
Furthermore, several tumors can be radiated during one outpatient
In general, radiosurgical ablation of brain metastases is followed by
whole-brain fractionated radiotherapy to minimize the likelihood of
regional relapse. Although radiosurgery can be used without
subsequent whole-brain fractionated radiotherapy, this approach to
brain metastases remains investigational. Unlike conventional brain
irradiation, radiosurgery can be repeated for meta-chronous lesions
months to years after initial treatment.
Before the advent of radiosurgery, treatment of pituitary tumors
consisted of surgical resection, medical management with
pharmacologic suppression using such agents as bromocriptine
(Parlodel) and somatostatin, or fractionated external-beam-radiotherapy,
which is often the preferred treatment for bulky tumors with
suprasellar extension. Most pituitary tumors are resectable with a
transsphenoidal approach. The recurrence rate after this procedure
ranges from 8% to 15%, with mortality varying from 0.3% to
1.8%.[17-20] Among hormone-secreting pituitary tumors, hormonal
control rates range from 42% to 86%, depending on the type of
Morbidity and mortality following surgical resection of pituitary
tumors are higher among patients who have recurrent disease or
significant intercurrent medical conditions. Radiosurgery is a
particularly attractive alternative to resection in these settings.
There is considerable experience with fractionated radiotherapy in
the management of pituitary tumors. When this treatment is used
alone, 10-year tumor control rates for pituitary tumors range from
77% to 87%.[21,22] However, the rates of pituitary dysfunction in one
study were 35% for thyroxine, 32% for glucocorticoids, and 33% for
sex hormones. Overall complication rates from fractionated
radiotherapy range from 3% to 7%.[22,23] Complications include
injuries to the visual tracts, late secondary malignancies, and rare
Although several small series report high control rates after
radiosurgery for small primary or recurrent pituitary tumors (Table
2),[3,24-26] follow-up has been limited to 2 to 5 years.
Reported rates of recurrence vary from 0% to 2%. Hormonal cure rates
after radiosurgery have been reported to range from 48% to 100%
depending on the subtype of tumor.[3,24-27] Treatment is most
successful for patients with prolactinoma, acromegaly, or
Cushings disease and least effective in those rare patients
with Nelsons syndrome.
Large pituitary tumors (> 3.5 cm) or tumors closer than 3 mm to
the optic nerve or chiasm cannot be treated by radiosurgery without a
significant risk of radiation injury to the optic apparatus.
Alternative strategies for such patients include: (1) combined
surgical resection and radiosurgery (with radiosurgery reserved for
treatment of any residual sellar and cavernous sinus tumor once it
has been debulked away from the optic apparatus), or (2) conventional
Regardless of the treatment approach employed, all patients with
pituitary tumors require serial endocrinologic studies to document
changes in hormonal function and allow for institution of hormonal
replacement if hypopituitarism develops. Patients who develop
recurrent pituitary adenoma after fractionated radiotherapy to the
pituitary region may be treatable with radiosurgery, although the
risks of radiation injury are not well defined in this setting.
Because small- to moderate-sized acoustic neuromas typically produce
hearing loss leading to diagnosis, these tumors are ideal targets for
radiosurgery. Acoustic neuromas occur as sporadic unilateral tumors
or as part of the genetic syndrome, type II neurofibromatosis
(bilateral tumors). Although these lesions are surgically
approachable, surgery entails a high probability of hearing loss and
a lower, but moderate, rate of facial nerve dysfunction.
Among 749 patients with acoustic neuroma from three surgical series,
only 15% to 37% retained hearing and approximately 20% had injury to
the facial nerve resulting in partial or complete facial
palsy.[28-30] Complete resection was achieved in 90% of
patients,[28-30] and only 1% of these tumors recurred over 5-year
follow-up. However, 10% recurred within 5 years if resection was incomplete.
A 1992 National Institutes of Health consensus conference recognized
the promising role of radiosurgery in the treatment of acoustic
neuroma. Several large series of patients treated with
radiosurgery for acoustic neuroma have been compiled (Table
3).[32-35] Across all of these series, 51% of patients had
hearing preserved and more than 90% retained normal facial nerve function.[32-35]
Although the results of radiosurgical treatment of acoustic neuroma
described above represent an improvement in cranial nerve
preservation compared to conventional surgery, these patients were
treated with single large fractions. Recent trends toward the use of
fractionated radiosurgery have further increased rates of hearing preservation.
In an ongoing series of patients at Stanford treated with three
fractions delivered over 36 hours, hearing was preserved in 88% of
patients at 1 year. Hearing was maintained more often in patients
with unilateral tumors than in those with neurofibromatosis type II
(100% vs 71%; P = .09). Two patients experienced temporary partial
facial numbness, but there have been no facial nerve injuries.
The radiosurgical end point of treatment is reduction or
stabilization in tumor size over several years. This result has been
achieved in 93% of patients reported in the radiosurgical literature
(55% to 63% of tumors decreased and 33% to 37% stabilized over 1
year; 78% of tumors decreased over 3 years).[32-35] Our experience at
Stanford parallels these results (Figure
Patients with recurrent acoustic tumors after resection and those in
whom craniotomy is associated with significant risk also represent
ideal candidates for radiosurgery. Despite the encouraging studies
reported in the literature, follow-up, both in terms of tumor control
and complications, is relatively short, given the long natural
history of acoustic neuroma. In addition, use of radiosurgery is
limited to the treatment of small tumors; large acoustic neuromas
must still be treated with surgical resection.