Radiosurgical Techniques

Noninvasive radiosurgical treatment modalities consist of percutaneous or external irradiation using photon radiation technology (GK and LINACs). Current GK units use 201 cobalt-60 sources enclosed within a hemispheric vault, leading to the emission of gamma ray energy converged at the isocenter and allowing high-precision treatment of an intracranial target. The aperture of collimators differs from 4 to 18 mm, so the diameter of the radiation beam can be adapted to the size of the lesion. Due to high radiation activity (up to 300 TBq) the propagation of the dose reaches 2 Gy/min by 80 cm from the radiation source. However, during the therapy of irregular lesions, multiple spherical isocenters have to be superimposed to allow conformal treatment, resulting in anatomic regions that receive much higher radiation dosage than the marginal dose. GK technology is limited exclusively to the treatment of intracranial lesions and high cervical lesions (when GK perfexion is available).

LINAC radiosurgery was introduced in the early 1980s following modifications of standard LINACs used for external beam radiotherapy to obtain high conformity of the radiation. The frequency of the electromagnetic field is about 3 GHz, and the accelerating energy is 4 to 25 MeV. X-ray photons are bundled to achieve a collimated beam of radiation. The advantages of LINAC devices compared to GK units are better penetration and homogeneity of radiation, steeper falloff of the dose at the margins of the irradiated lesion, and no problems with radioactive waste disposal. Moreover, application of the multileaf collimators enabled safe treatment of irregular intracranial lesions with an accuracy of less than 1 mm. Originally, the LINAC is mounted to a gantry that rotates over the patient’s head fixed in a stereotactic apparatus. However, recent LINAC devices (i.e., CyberKnife) applying dynamic position adjustment of the rotating arm do not require head fixation to preserve high accuracy of the radiation.

Particle beam radiation therapy with protons or heavy helium nuclei based on the so-called Bragg peak phenomenon, applied in only few centers worldwide, are not discussed further due to the paucity of reliable clinical data.

Another important technique of radiosurgical brain metastases therapy is interstitial radiosurgery (brachytherapy), allowing direct intratumoral application of low-energy ionizing radiation (0.5-10 cGy/min). Following stereotactic serial biopsy that provides the confirmation of the histopathologic diagnosis, single or multiple radioactive sources are temporarily implanted, allowing optimal dose distribution (e.g., I-125, iridium-192, and gold-198).¹² I-125 seeds (0.8 × 4.5 mm titanium cylinder) are fixed in a Teflon catheter, sterilized, and inserted into the center of the lesion, where they are left for 3 to 4 weeks (delivering a dose of 60 Gy to the tumor margins). The steep radial dose falloff is inversely proportional to the square of the distance from the radiation source, which provides minimal exposure to radiation of the healthy surrounding brain tissue.

At the beginning of 1990s, Photoelectron (Lexington, Waltham, MA) developed the Photon Radiosurgery System. It is a battery-powered, miniature x-ray generator capable of delivering low-energy radiation (soft x-rays) directly to small brain lesions in a single therapeutic session. Thus, it combines the direct dose application of brachytherapy with the advantages of short-time exposure typical for external radiosurgical techniques (GK and LINAC).¹³ Brief characteristics of the physical parameters of the radiosurgical procedures discussed in this chapter are summarized in Table 98-2.

The rest of the chapter is devoted to the presentation of the clinical results of radiosurgical treatment of brain metastatic diseases, based on the literature data and the experience of the authors. Furthermore, algorithms of clinical decision making concerning treatment devices for brain metastases developed in University Medical Centre in Freiburg, Germany, are described.

Linac-Based Stereotactic Radiosurgery

Due to the pioneer work of Betti and Derechinsky,¹⁴ Colombo et al.,¹⁵ Winston and Lutz,¹⁶ and others in the 1980s, precision and stability of LINAC radiosurgery systems was dramatically enhanced and became comparable to GK units. Nowadays, several techniques exist for LINAC systems that can be used to achieve an exact and highly conformal dose distribution. Target dose distribution can be adjusted by varying the collimator size, eliminating undesirable arcs, manipulating arc angles, using multiple isocenters, and differentially weighting the isocenters.¹⁷ A highly conformal dose distribution can be achieved by generating nonspherical beam shapes that are conformal to the beam’s eye view of the tumor with a multileaf collimator (Fig. 98-1A-C). Comparison of a modern LINAC radiosurgery system with GK units showed no differences in efficiency and safety.^18,19 The most important difference between these two systems is the number of metastases that can be irradiated in a single session. In GK units, up to 25 brain metastases can be treated simultaneously when the lesions are diffusely scattered and the total tumor volume is less than 15 to 30 cm³²⁰ In contrast, LINAC radiosurgery systems can irradiate a maximum of 3 to 4 metastases as the multiple intersecting radiation arcs may cause a hot spot outside the target volume within normal brain tissue.²¹

FIGURE 98-1 LINAC-based SRS with micromultileaf collimators (Virtuoso 3.4 planning software, Stryker Leibinger, Freiburg). A, Conformal application of 20 Gy to an irregularly shaped target volume (blue line). B, Beam’s eye view. C, 3D reconstruction of target volume (blue) and rise structures (optic system and brain stem).

Brain metastases are ideal targets for SRS, because these tumors are usually pseudospherical and show, in contrast to glioma, a noninfiltrative growth pattern with a sharp delineation to normal brain tissue. In addition, brain metastases occur often multifocally and have a diameter of less than 3.5 cm at the time of diagnosis. In Freiburg, we usually treat a maximum of three metastases during one session, and the prescribed dose for metastases with a diameter below 2.5 cm is 20 Gy calculated on the 80% isodose surrounding the outer tumor margin. Metastases with a diameter greater than 2.5 cm are treated with 18 Gy to prevent radiation necrosis. With nonspherical, irregularly shaped tumor margins, the application of one isocenter is not adequate to spare healthy brain tissue, especially when the metastasis is located within eloquent areas (e.g., the central region or brain stem). In these cases, we apply techniques like the use of multiple isocenters or a micromultileaf collimator to achieve a highly conformal target volume. Nataf et al.²² showed the importance of an exact dose application to the target volume by comparison of patients with a 2-mm margin to patients without this additional radiation volume. In contrast to Nöel et al.,²³ who achieved improved local control with a 1-mm margin, the addition of a 2-mm margin resulted in more severe parenchymal complications, no increase in the 1- and 2-year local control rates, and no statistical significant difference in median overall survival (classic radiosurgery of 11.3 months vs. 2-mm margin radiosurgery of 19 months, p = 0.34).

Taking into account the usually limited life span of patients with brain metastases, SRS is a less invasive treatment method with the potential to treat several targets at one time and with the possibility to be repeated during the further course of the disease. In comparison to open surgery, the disadvantages of this method are that the diagnosis of a metastasis is not histologically proven, the mass effect is not relieved at once but over weeks or months, and the surrounding edema may be aggravated and become symptomatic.

For a solitary metastasis, one prospective²⁴ and four retrospective studies^25–28 compared stereotactic radiosurgery (SRS) with open tumor resection. Because they generally found no difference in overall survival, open surgery and radiosurgery are not concurrent but complementary methods in the treatment of brain metastases. Therefore, indication for SRS versus open surgery is based on individual aspects of the patients.

Singular or Solitary Metastasis

Radiosurgical treatment of brain lesions is usually based on a histopathologic diagnosis. For a solitary brain lesion, we recommend confirming diagnosis by either open surgery or stereotactic biopsy. When the location of the lesion or the general condition of the patient requires stereotactic biopsy, an intraoperative smear preparation may confirm the diagnosis of a metastasis. In this case, interstitial radiosurgery with I-125 seeds can be initiated during the same surgical procedure. This technique is discussed in detail later in this chapter. For a known tumor manifestation in other organ systems, histologic confirmation of a brain metastasis is not mandatory before SRS.

Number and Size of Metastases

Up to four metastases can be treated by SRS at once. However, a diameter of 3.5 cm is the upper limit for SRS, and metastases larger than 3.5 cm, especially with a pronounced perilesional edema, should be surgically resected. A combination of both techniques can also be performed by resecting the largest lesion, thereby relieving the mass effect, reducing the peritumoral edema, and treating afterward the remaining metastases by SRS.

Clinical Condition of the Patient

Patients with a pronounced comorbidity and a reduced general condition are usually treated by SRS as long as the patient is not threatened by the mass effect or surrounding edema of the lesion. In these patients, the possibility of not interrupting chemotherapy while treating brain metastases adds another reason in favor of SRS.

The risks for postoperative wound infections (leucopenia), healing problems, or bleeding complications (thrombocytopenia) are therefore eliminated without jeopardizing their optimal care.

Location of the Metastases

In general, deeply located or small metastases are ideal candidates for SRS, whereas superficially located lesions with a pronounced mass effect or peritumoral edema are preferentially treated by microsurgical resection. However, the clinical condition of the patient, as well as the expected life span, has to be considered.

Neurologic Symptoms Caused by Metastases

A symptomatic metastasis, when surgically accessible, should be resected to relieve the mass effect at once, reduce the surrounding edema, and thereby improve neurologic condition of the patients faster.

Prognosis

According to Lutterbach et al.,⁸ prognosis of patients with up to three cerebral metastases treated by LINAC radiosurgery can be predicted accurately with the RPA classification. Median overall survival for patients is 13.4, 9.3, and 1.5 months in RPA classes 1, 2, and 3, respectively. Due to the poor prognosis of RPA class 3, patients with a KPS score below 70 should only be treated with SRS when their age is under 65 and have no signs of extracerebral tumor progression. These two factors are independent risk factors associated with an improved prognosis in RPA class 3 patients.

In summary, SRS achieves local tumor control rates at 1 year between 80% and 90%. Brain metastases from radioresistant tumors (renal cell carcinoma, melanoma, and sarcoma) may also benefit from SRS.²⁹ Complication rates associated with SRS are low. However, the risk of radiation necrosis increases with larger tumor volumes and prior radiation therapy. Other risk factors for neurologic complications after SRS are the location in eloquent brain areas, SRS doses higher than 15 Gy, and progression of the primary cancer.³⁰

Interstitial Radiosurgery with I-125 Seeds

Interstitial radiosurgery, a form of brachytherapy (from the Greek word brachy, meaning “short”), is an internal radiotherapy where a radiation source is placed inside or near a radiosensitive lesion. Brachytherapy can be used for the treatment of primary cancers of the prostate, cervix, breast, and skin. For intracranial lesions, brachytherapy is useful in the treatment of metastases and low-grade glioma with good clinical results.

In early procedures, permanent implants made from thin flexible wires of iridium-192 that could be cut to any length were implanted in a stereotactic procedure. Iridium-192 has a half-life of 73.83 days and a mean energy of 380 keV. Nowadays, mostly I-125 seeds are used, which are sized 4.5 × 0.8 mm with a half-life of 59.43 days and deliver lower photon energy of 27.4, 31.4, and 35.5 keV (Fig. 98-2A-D). These seeds contain I-125 adsorbed onto a radiopaque silver rod hermetically encapsulated in a welded titanium capsule. Typical radioactivities range from 1 to 25 mCi.

FIGURE 98-2 I-125 seed–based interstitial radiosurgery. A, Planning MRI showing the posterior fossa, the segmentation of the tumor (pink), the brain stem (turquoise), and the implantation tract (yellow). B, 3D model of these structures, including the 60-Gy isodose (red) surrounding the tumor. C, Postoperative CT for position control with the metallic artifact of the I-125 seed. D, MRI follow-up 7 months later. (STP-Workstation, Radio System 1.1-5, Stryker Leibinger, Freiburg.)

Temporary Versus Permanent Interstitial Radiosurgery

Interstitial radiosurgery can be done by temporary or permanent implantation of radiation sources, which is typically decided intraoperatively during the planning phase. A permanent seed has the advantage that no second surgical procedure is necessary, whereas the final explantation of the temporary seeds facilitates the macrophage activity and thereby accelerates the degradation of necrosis and repair processes.

Interstitial Radiosurgery Versus SRS

In contrast to external radiotherapy where a high-energy radiation beam (4-15 MeV) passes through brain tissue and the tumor, the much lower photon radiation energy (27.4-35.5 keV) of the I-125 seed is partially absorbed inside the tumor. For I-125, the half-value layer of tissue required to reduce the intensity of radiation by 50% is approximately 2 cm. This attenuation in tissue is superimposed by the inverse square law stating that dose delivery is inversely proportional to the square of the distance from the source, resulting in a steep dose decline on the surface in the treatment of small lesions.

Absorbed Dose Rates

In SRS, 18 to 20 Gy are administered in only a few minutes corresponding to a high absorbed dose rate of up to 10 Gy/min. In interstitial radiosurgery, 60 Gy are continuously delivered within 25 days, corresponding to mean absorbed dose rates of 0.1 Gy/hr or 1.67 mGy/min. This continuous dose delivery best corresponds biologically to a hyperfractionated radiation. Due to their physical properties with a sharp dose decline, I-125 seeds are well suited for the treatment of small (10-30 mm) intracranial lesions also near eloquent and radiosensitive areas, especially if the histopathologic diagnosis is going to be verified by a stereotactic biopsy during the same surgical procedure.

In our center, patients are proposed a stereotactic biopsy prior to radiosurgery to confirm the histopathologic diagnosis and to exclude misinterpretations of benign or inflammatory lesions if one of the following conditions is given: (1) missing extracranial proof of a primary tumor, (2) low probability for brain metastases of a known primary tumor (e.g., prostate carcinoma), (3) an interval longer than 3 years between the initial diagnosis of a primary tumor and a late cerebral lesion that is suspected to be a metastatic recurrence of this tumor, or (4) the presence of multiple and different primary tumors.

Treatment Planning

In our center, the patient is admitted to the ward the day before surgery. Frameless, nonreformatted magnetic resonance tomography imaging (T1 weighted 1 mm postgadolinium and T2 weighted 1 mm) and, if applicable, positron emission tomography (PET) are performed and transferred to the planning workstation. Next, the stereotactic base ring is fixed in local anesthesia, and a stereotactic CT scan postcontrast is acquired with fiducials fixed to the base ring. A stereotactic transformation with three-dimensional (3D) reconstruction is calculated, followed by fusions of the CT scan and the T1- and T2-weighted MRI using either anatomic landmark coregistration or fully automatic fusion. In the planning workstation, a spatial segmentation of the external contour, the tumor, and the organs of risk (e.g., eyes, chiasm, optical nerves and tracts, brain stem, and pituitary stalk) is done, and the dose at the tumor surface and the dose limitations at organs of risk are calculated. For brain metastases, a dose of 60 Gy to the surface of the lesion is applied. Then, the treatment using one or more seeds is planned. The seeds delivering the calculated dose in about 25 days are selected. The actual implantation time and stereotactic target coordinates are optimized using isodose diagrams and 3D surface modeling, respecting eloquent areas and blood vessels for the tracts. The appropriate I-125 seeds are taken from the radiation-protected vault, and the radiation activity is measured and documented. Each seed is heat shrunk into a Teflon catheter and autoclaved in a radiation-shielded container.

The stereotactic biopsy is done by a small skin incision and a 5-mm bur hole and typically following the tract of the seed catheter. The planned target position is retained by fluoroscopy, preferably in two plains, using a probe. The sterile seed within the catheter is exactly implanted at the planned target position under fluoroscopy control. The seed catheter is fixed in the bur hole using either fibrin foam or bone cement and a radiopaque hemoclip. The skin is closed by a suture. If high precision is required, the exact positioning of the seed can be controlled by a native CT scan and a fusion with the treatment plan. The radiation emission is measured and documented at distances of 0, 20, and 100 cm to the body’s surface according to legal rules. If applicable, the patient is instructed to avoid long-lasting and close contact (less than 1 m) with pregnant women and children. The patient is dismissed the second day after seed implantation. We propose autoclaving of the seeds and using a single-shot antibiosis during the seed implantation. A possible reactive brain edema during the implantation time can be treated best with oral administration of a low-dose steroid such as dexamethasone.

In most patients, one seed is inserted; however, two or more seeds are applied when the size or shape of the metastases makes it necessary (9% in our series). The radiation at the surface of the head and at distances of 20 and 100 cm is measured and documented. Rules for surrounding people are given to ensure radiation exposure far below 1 mSv during total implantation time in compliance with legislation and regulations. Isolation of the implanted patients was never necessary.

At the end of the implantation period, preferably after an MRI control, the seed with the catheter and the hemoclip are removed by a small skin incision in local anesthesia; this takes only a few minutes. The explantation of the seed increases macrophage activity, inducing further shrinking and degradation of the tumor over the following months. This has to be taken into account in the interpretation of PET, CT, and MRI during this time.

Radiation Parameters for Temporary I-125 Seeds

The radiation parameters of the last consecutive 250 patients treated with I-125 seeds in our center are as follows: The mean diameter of the metastases was 20.72 ± 5.8 mm (median 20.86 mm); the mean volume of the lesions was 5.77 ml. The median dosage at the tumor margin was 60 Gy (mean 60.28 ± 2.74 Gy, range 60-65 Gy), biologically corresponding to hyperfractionated radiotherapy. The continuous dose delivery of 60.28 Gy in 26 ± 7.28 days resulted in a mean energy dose rate of 10.56 ± 2.48 cGy/hr, as compared to SRS delivering around 18 to 20 Gy in about 30 minutes to the metastasis and resulting in a much higher absorbed dose rate. At the time of implantation, the mean total activity of the seeds was 12.05 ± 7.19 mCi (median 11.4 mCi).

In summary, interstitial radiosurgery based on I-125 seeds is a safe, minimally invasive, well-tolerated, and efficient neurosurgical procedure with a low rate of morbidity and adverse events. It can be done in local anesthesia and offers the additional possibility of verifying the histopathologic diagnosis by a stereotactic biopsy with minimal additional effort.

Discussion

The annual incidence of intracranial metastases is about 8.3 per 100,000 population³¹ and is present in up to 24% of cancer patients based on autopsy studies.³² Approximately one third of the patients with cerebral metastases are candidates for microsurgical resection. The other patients are amenable to nonsurgical treatment options due to their medical comorbidities precluding surgery, multiple metastases (roughly 50%-63% of the patients have multiple lesions at initial presentation),^1,6 or location in surgically inaccessible regions.³³ Brain metastases usually are almost spherical and small; unlike the glioma, they have a clear border to the surrounding brain tissue. Therefore, they are regarded as perfect targets for SRS, provided they are causing only minimal mass effect (less than 1 cm of midline shift).

The number of brain metastases that can be treated in one session is still an issue of debate. In most centers, including ours, it is common to treat up to three lesions in one radiosurgery session. In 2006, the American Society for Therapeutic Radiology and Oncology, the American Association of Neurological Surgeons, and the Congress of Neurological Surgeons jointly agreed to define SRS in a way that includes both traditional single-dose SRS and multidose SRS for up to five doses (two to five doses).^34,35

Both LINAC and GK radiosurgery have been used for the treatment of these lesions. There were no survival differences using either LINAC or GK radiosurgery, according to the RTOG 95-08 trial.¹⁹ The CyberKnife, which is a recently commercially available LINAC-based system, combines a robotic arm with an image guidance system to track the patient’s position, thereby obviating the need for application of a stereotactic frame.

SRS can be used alone or with surgical resection or WBRT.

Conclusions

Radiosurgery is effective in the treatment of single or multiple brain metastases that are smaller than 3 cm in diameter. The local control rates range from 85% to 100%. The incidence of side effects is low, from 2% to 5%. Radiation-induced edema and/or radiation necrosis was observed in 4% to 13% of the patients. Median survival ranged from 6 to 11 months. The results are comparable to those achieved with conventional therapy (microsurgical resection followed by whole brain radiotherapy). The advantages of radiosurgery are short hospitalization time (1-3 days), low morbidity, no delay of adjuvant therapy due to healing of surgical wounds, and favorable local control rates.