Control of brain metastases using = frameless image-guided radiosurgery; Neurosurgical FOCUS

Home =20

Activate=20 Account

Quick Links=20

=E2=80=A2	E= mail=20 this article
=E2=80=A2	Add=20 to Favorites
=E2=80=A2	View=20 related articles
=E2=80=A2	Download=20 to Citation Manager
=E2=80=A2	= Most-downloaded=20 articles

December 2009 = Volume 27, Number = 6

PDF
TOC=20 Alert
RSS=20
Table of=20 Contents=20
Call for=20 Papers=20
Archives=20 =

<= /TD>

Control of brain metastases using = frameless=20 image-guided radiosurgery

Joseph C.=20 T. Chen, M.D., Ph.D.¹, Darlene M. Bugoci,=20 M.D.², Michael=20 R. Girvigian, M.D.², Michael J. Miller,=20 M.D.², Alonso Arellano,=20 M.S.², and Javad Rahimian,=20 Ph.D.²

¹Departments of = Neurological=20 Surgery, and = ²Radiation=20 Oncology, Southern California Permanente Medical = Group, Los=20 Angeles, California

Abbreviations used in this = paper:=20 IGRS =3D image-guided radiosurgery; LINAC =3D linear = accelerator;=20 RPA =3D recursive positioning analysis; RTOG =3D = Radiation Therapy=20 Oncology Group; SRS =3D stereotactic radiosurgery; = WBRT =3D=20 whole-brain radiotherapy.

Address correspondence = to:=20 Joseph C. T. Chen, M.D., Ph.D., Department of = Neurological=20 Surgery, 1505 North Edgemont Street, 4th Floor, Los = Angeles,=20 California 90027. email: jctchen@yahoo.com.

DOI: 10.3171/2009.8.FOCUS09131

Figures = and Tables=20

Abstract	IntroductionMethodsResultsDiscussionConclusionsDisclaimerReferences

Object=20

Radiosurgery is an important and = well-accepted=20 method in the management of brain metastases. Using=20 conventional frame-based techniques, high lesional = control=20 rates are expected. The introduction of image-guided=20 techniques allows for improved patient comfort and = workflow.=20 Some controversy exists as to the accuracy of = imageguided=20 techniques and consequently the impact they might have = on=20 control of brain metastases (as opposed to the level = of=20 control achieved with frame-based methods). The = authors=20 describe their initial 15-month experience with = image-guided=20 radiosurgery (IGRS) using Novalis with ExacTrac for = management=20 of brain metastases.

Methods=20

The authors reviewed the cases of = brain=20 metastasis treated by means of IGRS in their tertiary = regional=20 radiation oncology service over a 15-month period. = During the=20 study period 54 patients (median age 57.9 years) = harboring 108=20 metastases were treated with IGRS. The median time = from cancer=20 diagnosis to development of brain metastasis was 12 = months=20 (range 0=E2=80=93144 months). The median tumor volume = was 0.98=20 cm³ (range 0.03=E2=80=9319.07 = cm³). The median=20 prescribed dose was 18 Gy to the 80% isodose line = (range 14=E2=80=9320=20 Gy). Lesions were followed with postradiosurgery MR = imaging=20 every 2=E2=80=933 months following = treatment.

Results=20

The median follow-up period was 9 = months (range=20 0=E2=80=9320 months). Median actuarial survival was = 8.6 months=20 following IGRS. Eight patients with 18 lesions died = within the=20 first 2 months after the procedure, before scheduled = follow-up=20 imaging. Thus 90 lesions (in 46 patients) were = followed up=20 with imaging studies. Lesions that were unchanged or = reduced=20 in size were considered to be under control. The = 6-month=20 actuarial lesion control rate was 88%. Smaller lesions = (< 1=20 cm³) had a statistically improved = likelihood of=20 complete imaging response (loss of all = contrast-enhancement p=20 =3D 0.01).

Conclusions=20

Image-guided radiosurgical treatment = of brain=20 metastases resulted in high rates of tumor control = comparable=20 to control rates reported for frame-based methods. = High=20 control rates were seen for small lesions in which = spatial=20 precision in dose delivery is critical. These data = suggests=20 that in regard to lesion control, IGRS using Novalis = with=20 ExacTrac is equivalent to frame-based radiosurgery=20 methods.

KEYWORDS:linear accelerator; = brain=20 metastasis; image guidance; radiosurgery.

Introduction	AbstractIntroduction = <<MethodsResultsDiscussionConclusionsDisclaimerReferences

Brain metastases are = a common=20 complication of some of the most prevalent cancer = types=20 including those of the lung and breast. The management = of=20 metastatic disease to the brain has included surgery, = WBRT,=20 and chemotherapy, as well as stereotactic = radiosurgery.=20 Stereotactic radiosurgery, by virtue of its = noninvasive nature=20 and high lesion control rates, has emerged as a key = method for=20 treatment of brain metastases.^2,27=20 Radiosurgery has been demonstrated to result in = superior local=20 control compared with WBRT alone. Furthermore, in = patients=20 with a single brain metastasis, the use of = radiosurgery in=20 conjunction with WBRT has shown survival benefit. = Patients=20 with oligometastases also may have benefit to quality = of life=20 from the addition of radiosurgery to WBRT. Frame-based = methods=20 of radiosurgery using either LINAC or gamma unit = devices are=20 well established.

Frameless image guidance as applied to radiosurgery = describes a method whereby high-resolution imaging is = obtained=20 at the time of treatment for patient positioning = purposes and=20 implies that patient immobilization is not obtained = with rigid=20 skeletal fixation, but rather with the noninvasive use = of a=20 mask. A number of solutions for IGRS have been devised = and are=20 currently available for clinical use. These systems = have used=20 stereoscopic x-ray imaging as well as on-board CT = imaging.

Frameless image-guided methods in the setting of=20 single-fraction radiosurgery have as their primary = advantage=20 the potential for improved patient comfort. As there = no=20 sedation or anesthesia is used, no vital monitoring is = required. The use of noninvasive patient = immobilization and=20 image-guided fiducialization also allows temporal = decoupling=20 of the various phases of treatment set-up, allowing = for=20 imaging, fiducialization, treatment planning, quality=20 assurance, and dose delivery to be performed in = separate=20 sessions, if needed. This can improve convenience for = both the=20 patient and the treating team.

Frame-based radiosurgery methods have a long = history, and=20 the reliability of these methods is not in dispute. In = contrast, since image-guided methods are relatively = new, few=20 reports are available detailing clinical results for = common=20 applications of this technology. Significant = controversy=20 exists as to the relative merits of image-guided = techniques in=20 comparison with frame-based methods.

The geometrical accuracy of frameless image-guided = methods=20 in phantom testing has been well described by a number = of=20 groups using different IGRS implementations.^3,10,22,30 Reports=20 have been presented regarding the accuracy of Novalis=20 frameless IGRS with ExacTrac (BrainLAB) in = particular.^18,38=20 Ultimately, however, the usability of frameless = methods in=20 comparison with traditional frame-based methods must = be=20 determined on the basis of clinical outcomes.

The relative accuracy of stereotactic methods = depends on=20 large part on the size of the target. Small target = volumes are=20 less forgiving with respect to geometrical error. = Small=20 departures from ideal localization may result in = significant=20 underdosing of the targeted lesion that may result in=20 treatment failure. Such reduced accuracy in brain = metastasis=20 radiosurgery may result in an anomalous increase in = failures=20 for very small=E2=80=93volume lesions.

We report our initial results using = frameless=20 IGRS for the control of brain metastases and in = particular=20 examine the role of size in overall lesion control = using these=20 techniques.

Methods	AbstractIntroductionMethods <<ResultsDiscussionConclusionsDisclaimerReferences

This study was reviewed and approved by the Kaiser=20 Permanente regional research committee.

Patient Population=20

Patients were referred to our regional radiosurgery = service=20 from 12 affiliated medical centers. Patients were = offered=20 treatment for metastatic disease to the brain with 4 = or fewer=20 metastases and a Karnofsky Performance Scale score of = 70 or=20 greater at time of initial presentation to our section = clinic.=20 Over a 15-month period, 54 patients harboring 108 = lesions were=20 treated. There were 4 patients who underwent a total = of 6=20 repeated radiosurgery sessions for treatment of new = lesions.=20 The median age at time of treatment was 58.9 years. = There were=20 32 female and 22 male patients in the study = population. The=20 distribution of primary tumor types is presented in Table=20 1.

<= /TBODY>

TABLE 1: = Distribution of=20 tumor types in 54 patients*

The median time from cancer diagnosis to = development of=20 brain metastasis was 12 months (range 0=E2=80=93144 = months). The=20 median time from diagnosis of brain metastasis to = initial=20 radiosurgery was 108 days (range 10=E2=80=93869 days). = The median=20 tumor volume was 0.98 cm³ (range = 0.03=E2=80=9319.07=20 cm³). The median prescription dose was 18 = Gy to the=20 80% isodose line (range 14=E2=80=9320 Gy). Patients = were followed-up=20 with postradiosurgery MR imaging every 2=E2=80=933 = months.

The RTOG RPA classification is useful in describing = the=20 overall condition of patients undergoing treatment. In = this=20 series, 10 patients were classified as RPA Class 1, 42 = patients Class 2, and 2 patients Class 3.

In this series, most patients = underwent WBRT at=20 some point during their course of treatment. = Thirty-two=20 patients underwent WBRT more than 2 months prior to = SRS, and=20 were treated with SRS for either lesion progression or = new=20 lesions. Eleven patients received WBRT for boost = purposes=20 during the first 2 months after SRS. Three patients = had WBRT=20 more than 2 months following SRS for treatment of = additional=20 brain lesions. Eight patients did not have WBRT during = the=20 study period.

Radiosurgery Technique=20

Prior to radiosurgery, patients underwent = high-resolution=20 MR imaging using 1-mm-thick contrast-enhanced = 3D-spoiled=20 gradient recalled sequences (GE Excite 1.5 T, GE = Medical=20 Systems).

Patient immobilization was accomplished by = customfitting=20 patients with a BrainLAB bivalve-style thermoplastic = mask=20 incorporating a bite bar. A high-resolution CT scan = using=20 1.25-mm slices was then used for fiducialization (GE=20 LightSpeed RT 16, GE Medical Systems).

Imaging data were then transferred via a local area = network=20 to a planning workstation running BrainScan v.5 = software=20 (BrainLAB). Structures at risk as well as target = lesions were=20 then contoured. Target lesions were contoured with no=20 additional margin.

Treatment plans were created using forward-planning = methods=20 incorporating either dynamic multileaf collimated = noncoplanar=20 arcs or circular collimated noncoplanar arcs. The = median=20 prescription dose was 18 Gy to the 80% isodose line = (range=20 14=E2=80=9320 Gy to the 80% isodose line). A single = isocenter per=20 lesion was used for all treatments. The median = conformity=20 index (RTOG method: prescription isodose volume/target = volume)=20 was 1.6. A typical treatment plan utilizing dynamic=20 noncoplanar arcs and a follow-up MR image are = presented in Fig.=20 1.

View = larger=20 version (43K)

Fig. 1. A = and B:=20 Magnetic resonance images with (A) and = without (B)=20 50 and 80% isodose lines showing a typical = treatment plan for brain metastases. The = patient=20 in this case was a 56-year-old woman with = 2=20 posterior fossa lesions secondary to = metastatic=20 breast cancer. C: Follow-up MR image = obtained 6=20 months after treatment showing complete = response=20 as evidenced by complete loss of contrast=20 enhancement within the treatment volume. = The=20 lesions were treated with 20 Gy to the 80% = isodose=20 line. The more lateral lesion measured = 0.53=20 cm³ and had a treatment = conformity=20 index of 1.7, while the more medial lesion = had a=20 volume of 0.33 cm³ and a = treatment=20 conformity index of 2.0.

Treatment data were transferred by local area = network to=20 the LINAC control station and the ExacTrac patient = positioning=20 system. The patient was then placed in a supine = position on=20 the treatment table and the head was placed into the = mask=20 immobilizer. An optical tracking array was then = applied to the=20 head mounting system (Fig. = 2).

View = larger=20 version (76K)

Fig. 2. = Photograph=20 showing the bivalve-style mask used in = conjunction=20 with an optical tracking = array.

For each lesion, the optical tracking system was = used for=20 initial patient positioning. Stereoscopic x-ray = imaging was=20 then performed and the results fused with digitally=20 reconstructed radiographs based on CT imaging via the = ExacTrac=20 system. A positioning offset was then calculated. = Patient=20 positioning corrections were made for each of 6 = degrees of=20 freedom using a robotically actuated couch in concert = with the=20 real-time optical tracking system. Treatment was = initiated=20 when linear offsets were no more than 0.5 mm and = angular=20 offsets were 1=C2=B0 or less. Treatment time was = typically under 20=20 minutes per lesion.

Following completion of dose delivery, patients = were=20 released from the mask and discharged home with = anticonvulsant=20 and corticosteroid medications where appropriate.

Patients were followed up with = contrast-enhanced=20 MR imaging at 6=E2=80=938 weeks following SRS = treatment and then every=20 2=E2=80=933 months until the end period of data = collection or patient=20 demise. Patients demonstrating evidence of lesion = progression=20 underwent dedicated brain FDGPET to differentiate = between=20 necrosis and tumor progression.^6,15

Data Analysis=20

Survival and lesion control were calculated using=20 Kaplan-Meier actuarial curves. Univariate analysis was = conducted using the Fisher exact test and the = Mann-Whitney=20 U-test.

The measured end points were freedom from targeted = lesion=20 progression, freedom from new metastases and overall = survival.=20 Lesion progression was determined if there was = evidence of=20 tumor enlargement following treatment regardless of = whether=20 surgery was undertaken. New metastases were dated from = the=20 time of the first imaging study demonstrating their = presence.=20 A separate end point of complete response was = determined by=20 surveillance imaging demonstrating complete resolution = of=20 contrast-enhancement within a previously treated = volume.

Survival functions were calculated = using=20 Kaplan-Meier survival curves. Univariate analyses were = calculated using the Fisher exact test for contingency = tables=20 and the Mann-Whitney U-test for continuous variables. = Survival=20 comparisons were calculated using the log-rank=20 method.

Results	AbstractIntroductionMethodsDiscussionConclusionsDisclaimerReferences

The median duration of follow-up at the time of = data=20 acquisition was 9 months (range 0=E2=80=9320 = months).

Overall Survival=20

At the time of data analysis, 20 of the 54 patients = in our=20 study group were still alive, 33 had died during the = reporting=20 period, and 1 had been lost to follow-up (with = survival data=20 unavailable). The actuarial median survival was = reached at 8.6=20 months (Fig. 3). = A=20 univariate analysis of survival demonstrated no = significant=20 effect for sex, time from diagnosis, primary disease = site, RPA=20 (Class 1 vs Class 2 or 3), number of lesions (1 vs = > 1), or=20 use of WBRT. A multivariate analysis was conducted = using the=20 same criteria and demonstrated a significant effect = only for=20 RPA (p =3D 0.04).

View = larger=20 version (9K)

Fig. 3. = Kaplan-Meier=20 curve demonstrating actuarial overall = survival.=20 Median survival was reached at 8.6=20 months.

Lesion Control=20

There were 108 individual brain metastases treated = during=20 this study. Nine patients, who had a total of 18 = lesions, died=20 or were lost to follow-up before follow-up imaging = studies=20 could be performed, leaving a population of 90 lesions = in 45=20 patients for analysis. The median lesion volume was = 0.98=20 cm³ (range 0.03=E2=80=9319.0 = cm³). Progression=20 was observed in a total of 9 lesions in 8 patients = during=20 follow-up.

Lesions were divided into 2 groups on the basis of = volume:=20 =E2=89=A5 1 cm³ and < 1 cm³. = These groups=20 were found to be similar by univariate analysis in = regard to=20 patient age, primary disease site, and RPA class. = Lesions of=20 < 1 cm³ were less likely to be solitary = in=20 nature (Table 2). = Three of=20 46 lesions demonstrated progression in the group of = < 1=20 cm³ volume and 6 of 44 lesions demonstrated = progression in the group of lesions of =E2=89=A5 1 = cm³ in=20 volume. Of the 9 lesions that showed progression, 5 = were=20 associated with breast cancer and 1 each with lung, = melanoma,=20 gastric, and renal cancers. Local control survival = curves are=20 presented in Figs. 4 = and 5 for all=20 of the lesions (pooled data) as well as for the = lesions=20 stratified by size (< 1 cm³ vs =E2=89=A5 = 1=20 cm³). The 6-month actuarial control rate = for all=20 lesions was 88 =C2=B14% (=C2=B1 SE). While there was a = trend toward=20 better control in lesions < 1 cm³ in = volume, it=20 did not reach statistical significance in this series = (p =3D=20 0.16, log-rank test). Univariate analysis and = multivariate=20 analysis of local control failed to show significant = effects=20 for age, RPA, lesion volume, number of coexistent = lesions, or=20 sex.

View = larger=20 version (8K)

Fig. 4. = Kaplan-Meier=20 curve demonstrating actuarial local = control for=20 all lesions for which there was imaging = follow-up=20 in this series.

View = larger=20 version (10K)

Fig. 5. = Kaplan-Meier=20 curve demonstrating actuarial freedom from = progression for lesions stratified by = volume (<=20 1 cm³ vs > 1=20 cm³).

TABLE 2: = Univariate=20 analysis of patient demographic and = clinical=20 characteristics versus lesion=20 = volume

A separate end point of complete response (complete = loss of=20 contrast enhancement) was evaluated (Table 3). = This was=20 seen in 9 lesions in 7 patients. Complete responses = were more=20 likely in lesions of < 1 cm³ in volume = (p =3D=20 0.01, Mann-Whitney U-test).

TABLE 3: = Imaging=20 response stratified by lesion size*

Radiation necrosis was noted involving = 6 lesions=20 in 5 patients with late progression of contrast = enhancement=20 but with negative FDG-PET scans. These 5 patients = included 3=20 with lung cancer, 1 with breast cancer, and 1 with = melanoma.=20 The median time to formation of necrosis was 28 weeks = (range=20 3=E2=80=939.9 months). In none of these cases was the = radiation=20 necrosis symptomatic and none of the patients = underwent=20 surgery or required steroid therapy. At last scheduled = follow-up, 2 of the 5 patients had died (at 11.3 and = 21.3=20 months following treatment). The 3 other patients were = alive,=20 with an average of 15.6 months of follow-up.

Freedom From Brain = Progression=20

Eleven patients (20%) developed new = metastases=20 following radiosurgery treatment. The median time from = treatment to development of new metastases was 6=20 months.

Complications=20

Complications were unusual in this = series. There=20 were no occurrences of postprocedure seizures or new=20 progressive neurological deficits in the absence of = tumor=20 progression. There were no apparent nontarget dose=20 deliveries.

Discussion	AbstractIntroductionMethodsResultsConclusionsDisclaimerReferences

Role of Radiosurgery in Management = of Brain=20 Metastases=20

Brain metastases are a common complication of some = of the=20 most common forms of cancer, with an overall incidence = estimated to be 8=E2=80=9311 per 100,000.^26,37=20 Radiosurgery has emerged as a key method of providing=20 definitive local control for brain metastases in = addition to=20 surgery and WBRT.¹³=20 Radiosurgery with WBRT has been consistently shown to = improve=20 overall local control in comparison with WBRT = alone.^2,20 In=20 addition, in comparison to WBRT alone, the combination = of WBRT=20 and SRS appears to improve survival in patients with a = single=20 brain metastasis and improve functional independence = in=20 patients with 3 or fewer metastases.² = Surgery=20 continues to play an essential role in the management = of=20 lesions complicated by mass effect or after failure of = less-invasive treatment methods.¹

Routine use of WBRT in the setting of = brain=20 metastases has become controversial.^4,8,9 = A number=20 of centers have advocated routine use of WBRT, while = others=20 typically will defer or withhold WBRT until there is = either=20 failure of local control or development of new = metastases.=20 While WBRT has been shown to reduce the likelihood of=20 formation of new brain metastases, addition of WBRT to = SRS has=20 not been shown to improve overall survival.^4,33=20 Prevention of new metastases with WBRT, however, may = result in=20 initial improved preservation of cognitive function. = This,=20 however, may be counterbalanced by the increasing = incidence of=20 radiation leukoencephalopathy in longer term = survivors,=20 particularly those surviving beyond 2 years following=20 WBRT.⁵ = Survival=20 to 24 months is approximately 15=E2=80=9320% in = patients with 3 or=20 fewer brain metastases on presentation.^2,27 While we=20 have in the past relied on WBRT as an important = component of=20 the initial treatment of brain metastases, it is = likely that=20 this modality will be used in a more limited fashion = as=20 overall survival from metastatic cancer improves. Of = interest,=20 new methods of WBRT delivery with sparing of temporal = lobe=20 structures may mitigate to some degree late-term = cognitive=20 decline.¹⁴ Lastly,=20 the role of radiosurgery in the adjuvant therapy of = surgically=20 treated brain metastases has been a subject of recent = interest=20 and may further expand the role of radiosurgery in the = management of cerebral metastases.^{12,16,23,29,34}

Radiosurgical Instrumentation =

Gamma-unit devices have traditionally had = significant=20 advantages over LINAC devices in regard to simplicity = of=20 quality control and work flow speed. Linear = accelerator=20 devices employing cone-based collimation required more = complex=20 treatment planning and work flow. The introduction of=20 multileaf collimation has allowed for development of = more=20 conformal and homogeneous treatment plans using single = isocenter methods,^7,17,32 thereby=20 simplifying the planning and delivery process = resulting in=20 more rapid workflow. There appears to be no difference = in=20 either overall survival or lesional control of brain=20 metastases using either LINAC or gamma-unit = technology.^2,36

The use of frame-based skeletal immobilization for=20 stereotactic procedures has a long history dating back = to the=20 1950s with the introduction of stereotactic systems = designed=20 by Leksell, Talaraich, Reichert and Mundinger, Todd = and Wells,=20 and others.¹¹ Studies=20 have demonstrated that the application accuracy of = these=20 devices is on the order of 1 mm.²¹ In=20 addition to potential errors due to frame application, = other=20 aspects of stereotactic treatment can introduce = additional=20 errors. These would include errors associated with = imaging=20 resolution and distortion as well as fiducialization, = image=20 fusion, device isocentricity, and radiation output. = Using=20 frame-based methods in the setting of the Novalis = LINAC, we=20 have previously demonstrated high accuracy and = precision with=20 an overall system error of 0.6 mm.²⁸

While simple and expedient, the use of the = stereotactic=20 frame does have some disadvantages including the = procedural=20 discomfort for most patients, with awake placement = being=20 typically performed with local anesthetic only. = Alternatively,=20 frame placement can be performed with sedation, = although at=20 our institution this entails mandated close patient=20 monitoring, with consequent labor demands on clinic = staff. The=20 use of image-guided techniques can potentially improve = the=20 patient experience while at the same time further = improve=20 workflow and other logistical issues by making it = unnecessary=20 to accomplish all phases of the = procedure=E2=80=94including frame=20 placement, imaging, and treatment=E2=80=94at the same = location and=20 time.

Noninvasive patient immobilization had been = accomplished in=20 the past by means of stereotactic devices, such as the = Spiegel-Wycis device, which used a plaster cast = applied to the=20 head.³⁵ In the=20 case of the Spiegel-Wycis device and similar devices = of the=20 era, intraoperative image guidance is achieved by = using=20 ventriculography. The modern implementation of = image-guided=20 radiosurgery procedures is therefore analogous to this = earlier=20 technique. Rather than relying on ventriculographic = methods,=20 however, the ExacTrac and similar systems use image = fusion of=20 digitally reconstructed preoperative radiographs with=20 intraoperative skull radiographs, allowing for precise = stereotactic positioning.

A number of reports regarding the = accuracy of=20 image-guided methods have demonstrated that = submillimeter=20 accuracy can be accomplished in bench testing for a = variety of=20 IGRS implementations, including the device used in = this=20 report.^10,24,38

Treatment of Brain Metastases With = IGRS=20

While bench testing suggests that image-guided = methods are=20 equivalent in accuracy to frame-based methods, = clinical data=20 are needed to confirm that outcomes are similar. From = the=20 clinical standpoint, the greatest concern is spatial = accuracy=20 of dose delivery. This is particularly critical for = small=20 lesion targets where geometrical imprecision may have = greater=20 effects on the resulting dose delivered to the target = volume.=20 One may hypothesize that an anomalous increase in = failures=20 involving smaller lesions may result if the technique = for=20 patient positioning is inadequate.

In this series, there appeared to be no increase in = the=20 risk of failure for patients with lesions of < 1=20 cm³ volume.

We are aware of 3 reports in the literature = utilizing IGRS=20 techniques for brain metastasis treatments (Table = 4).^19,25,31 These=20 reports demonstrate that lesion control rates are = similar to=20 those obtained with frame-based methods. We are, = however,=20 unaware of any other reports specifically addressing = lesion=20 control as a function of lesion size with the = frameless=20 methods used in the present study.

TABLE 4: = Recent case=20 series using LINAC radiosurgery treatment = for=20 brain metastases employing frame-based and = frameless techniques*

Conclusions	AbstractIntroductionMethodsResultsDiscussionConclusions = <<DisclaimerReferences

We present our early data in regard to = control=20 of brain metastases using IGRS methods. The overall = rate of=20 freedom from local lesion progression was 88% at 6 = months.=20 These data are comparable to those reported by others=20 employing conventional frame-based techniques and = alternative=20 IGRS methods. Lesions < 1 cm³ in volume = did not=20 have higher failure rates in comparison to larger = lesions,=20 demonstrating that frameless image-guided radiosurgery = can be=20 used for reliable and precise targeting of small = lesions. It=20 would appear that IGRS using Novalis with ExacTrac is = adequate=20 for treatment of the gamut of indicated brain=20 metastases.

Disclaimer	AbstractIntroductionMethodsResultsDiscussionConclusionsDisclaimer <<

The authors report no conflict of = interest=20 concerning the materials or methods used in this study = or the=20 findings specified in this paper.

References	AbstractIntroductionMethodsResultsDiscussionConclusionsDisclaimerReferences = <<

1.	Al-Shamy G, Sawaya R: = Management of brain = metastases:=20 the indispensable role of surgery. J = Neurooncol=20 92:275=E2=80=93282, 2009
2.	Andrews DW, Scott CB,=20 Sperduto PW, Flanders AE, = Gaspar LE,=20 Schell MC, et al.: Whole brain radiation = therapy=20 with or without stereotactic radiosurgery boost = for=20 patients with one to three brain metastases: = phase III=20 results of the RTOG 9508 randomised = trial. Lancet = 363:1665=E2=80=931672, 2004
3.	Antypas C, Pantelis E: = Performance evaluation = of a=20 CyberKnife G4 image-guided robotic stereotactic=20 radiosurgery system. Phys Med = Biol=20 53:4697=E2=80=934718, 2008
4.	Aoyama H, Shirato H,=20 Tago M, Nakagawa K, Toyoda T,=20 Hatano K, et al.: Stereotactic = radiosurgery plus=20 whole-brain radiation therapy vs stereotactic=20 radiosurgery alone for treatment of brain = metastases: a=20 randomized controlled trial. JAMA = 295:2483=E2=80=932491, 2006
5.	Aoyama H, Tago M, = Kato N,=20 Toyoda T, Kenjyo M, Hirota S, et = al.:=20 Neurocognitive = function of=20 patients with brain metastasis who received = either whole=20 brain radiotherapy plus stereotactic = radiosurgery or=20 radiosurgery alone. Int J Radiat = Oncol Biol=20 Phys 68:1388=E2=80=931395, 2007
6.	Belohlavek O, = Simonova G,=20 Kantorova I, Novotny J Jr, = Liscak R:=20 Brain metastases = after=20 stereotactic radiosurgery using the Leksell = gamma knife:=20 can FDG PET help to differentiate radionecrosis = from=20 tumour progression?. Eur J Nucl Med = Mol=20 Imaging 30:96=E2=80=93100, 2003
7.	Brahme A: Design principles and = clinical=20 possibilities with a new generation of radiation = therapy=20 equipment. A review. Acta = Oncol 26:403=E2=80=93412, 1987
8.	Brown PD, Asher AL,=20 Farace E: Adjuvant=20 whole brain radiotherapy: strong emotions decide = but=20 rational studies are needed. Int J Radiat = Oncol Biol=20 Phys 70:1305=E2=80=931309, 2008
9.	Chang EL, Wefel JS,=20 Maor MH, Hassenbusch SJ III,=20 Mahajan A, Lang FF, et al.: A pilot study of = neurocognitive=20 function in patients with one to three new brain = metastases initially treated with stereotactic=20 radiosurgery alone. Neurosurgery=20 60:277=E2=80=93284, 2007
10.	Chang SD, Main W,=20 Martin DP, Gibbs IC, Heilbrun MP: = An analysis of the = accuracy of=20 the CyberKnife: a robotic frameless stereotactic = radiosurgical system. Neurosurgery=20 52:140=E2=80=93147, 2003
11.	Chen JC, Apuzzo ML: = Localizing the point: = evolving=20 principles of surgical navigation. Clin = Neurosurg=20 46:44=E2=80=9369, 2000
12.	Do L, Pezner R, = Radany E,=20 Liu A, Staud C, Badie B: Resection followed by=20 stereotactic radiosurgery to resection cavity = for=20 intracranial metastases. Int J Radiat = Oncol Biol=20 Phys 73:486=E2=80=93491, 2009
13.	Flickinger JC, = Lunsford LD,=20 Somaza S, Kondziolka D: Radiosurgery: its role = in brain=20 metastasis management. Neurosurg Clin N = Am=20 7:497=E2=80=93504, 1996
14.	Gutierrez AN, = Westerly DC,=20 Tome WA, Jaradat HA, Mackie TR,=20 Bentzen SM, et al.: Whole brain = radiotherapy with=20 hippocampal avoidance and simultaneously = integrated=20 brain metastases boost: a planning study. = Int J Radiat = Oncol Biol=20 Phys 69:589=E2=80=93597, 2007
15.	Hustinx R, = Pourdehnad M,=20 Kaschten B, Alavi A: PET imaging for = differentiating=20 recurrent brain tumor from radiation = necrosis.=20 Radiol = Clin North=20 Am 43:35=E2=80=9347, 2005
16.	Jagannathan J, Yen CP,=20 Ray DK, Schlesinger D, = Oskouian RJ,=20 Pouratian N, et al.: Gamma Knife = radiosurgery to the=20 surgical cavity following resection of brain=20 metastases. J=20 Neurosurg 111:431=E2=80=93438, 2009 [Abstract]
17.	Jin JY, Yin FF, = Ryu S,=20 Ajlouni M, Kim JH: Dosimetric study using = different=20 leaf-width MLCs for treatment planning of = dynamic=20 conformal arcs and intensity-modulated=20 radiosurgery. Med Phys = 32:405=E2=80=93411, 2005
18.	Jin JY, Yin FF, = Tenn SE,=20 Medin PM, Solberg TD: Use of the BrainLAB = ExacTrac=20 X-Ray 6D system in image-guided = radiotherapy.=20 Med = Dosim=20 33:124=E2=80=93134, 2008
19.	Kamath R, Ryken TC,=20 Meeks SL, Pennington EC, = Ritchie J,=20 Buatti JM: Initial=20 clinical experience with frameless radiosurgery = for=20 patients with intracranial metastases. = Int J Radiat = Oncol Biol=20 Phys 61:1467=E2=80=931472, 2005
20.	Kondziolka D, Patel A,=20 Lunsford LD, Kassam A, = Flickinger JC:=20 Stereotactic = radiosurgery=20 plus whole brain radiotherapy versus = radiotherapy alone=20 for patients with multiple brain = metastases.=20 Int J = Radiat Oncol=20 Biol Phys 45:427=E2=80=93434, 1999
21.	Maciunas RJ,=20 Galloway RL Jr, Latimer JW: The application = accuracy of=20 stereotactic frames. Neurosurgery=20 35:682=E2=80=93685, 1994
22.	Masi L, Casamassima F,=20 Polli C, Menichelli C, Bonucci I, = Cavedon C: Cone beam=20 CT image guidance for intracranial stereotactic=20 treatments: comparison with a frame guided = setup.=20 Int J = Radiat Oncol=20 Biol Phys 71:926=E2=80=93933, 2008
23.	Mathieu D, = Kondziolka D,=20 Flickinger JC, Fortin D, Kenny B, = Michaud K, et al.: Tumor bed radiosurgery = after=20 resection of cerebral metastases. Neurosurgery=20 62:817=E2=80=93824, 2008
24.	Meeks SL, Bova FJ,=20 Wagner TH, Buatti JM, = Friedman WA,=20 Foote KD: Image=20 localization for frameless stereotactic=20 radiotherapy. Int J Radiat = Oncol Biol=20 Phys 46:1291=E2=80=931299, 2000
25.	Nishizaki T, Saito K,=20 Jimi Y, Harada N, Kajiwara K,=20 Nomura S, et al.: The=20 role of cyberknife radiosurgery/radiotherapy for = brain=20 metastases of multiple or large-size = tumors.=20 Minim = Invasive=20 Neurosurg 49:203=E2=80=93209, 2006
26.	Percy AK, Elveback LR,=20 Okazaki H, Kurland LT: Neoplasms of the = central nervous=20 system. Epidemiologic considerations. = Neurology = 22:40=E2=80=9348, 1972
27.	Rades D, Pluemer A,=20 Veninga T, Hanssens P, Dunst J,=20 Schild SE: Whole-brain = radiotherapy versus=20 stereotactic radiosurgery for patients in = recursive=20 partitioning analysis classes 1 and 2 with 1 to = 3 brain=20 metastases. Cancer = 110:2285=E2=80=932292, 2007
28.	Rahimian J, Chen JC,=20 Rao AA, Girvigian MR, Miller MJ,=20 Greathouse HE: Geometrical accuracy = of the=20 Novalis stereotactic radiosurgery system for = trigeminal=20 neuralgia. J=20 Neurosurg 101:3 Suppl351=E2=80=93355, 2004
29.	Roberge D, Petrecca K, = El=20 Refae M, Souhami L: Whole-brain = radiotherapy and=20 tumor bed radiosurgery following resection of = solitary=20 brain metastases. J = Neurooncol=20 95:95=E2=80=9399, 2009
30.	Ryken TC, Meeks SL,=20 Pennington EC, Hitchon P, = Traynelis V,=20 Mayr NA, et al.: Initial clinical = experience with=20 frameless stereotactic radiosurgery: analysis of = accuracy and feasibility. Int J Radiat = Oncol Biol=20 Phys 51:1152=E2=80=931158, 2001
31.	Shimamoto S, Inoue T,=20 Shiomi H, Sumida I, Yamada Y,=20 Tanaka E: CyberKnife=20 stereotactic irradiation for metastatic brain=20 tumors. Radiat = Med 20:299=E2=80=93304, 2002
32.	Shiu AS, Kooy HM,=20 Ewton JR, Tung SS, Wong J, = Antes K,=20 et al.: Comparison of=20 miniature multileaf collimation (MMLC) with = circular=20 collimation for stereotactic treatment. = Int J Radiat = Oncol Biol=20 Phys 37:679=E2=80=93688, 1997
33.	Sneed PK, Suh JH,=20 Goetsch SJ, Sanghavi SN, = Chappell R,=20 Buatti JM, et al.: A=20 multi-institutional review of radiosurgery alone = vs.=20 radiosurgery with whole brain radiotherapy as = the=20 initial management of brain metastases. = Int J Radiat = Oncol Biol=20 Phys 53:519=E2=80=93526, 2002
34.	Soltys SG, Adler JR,=20 Lipani JD, Jackson PS, Choi CY,=20 Puataweepong P, et al.: Stereotactic = radiosurgery of the=20 postoperative resection cavity for brain=20 metastases. Int J Radiat = Oncol Biol=20 Phys 70:187=E2=80=93193, 2008
35.	Spiegel EA, Wycis HT,=20 Marks M, Lee AJ: Stereotaxic apparatus = for=20 operations on the human brain. Science = 106:349=E2=80=93350, 1947
36.	Swinson BM, = Friedman WA: Linear accelerator = stereotactic=20 radiosurgery for metastatic brain tumors: 17 = years of=20 experience at the University of Florida. = Neurosurgery=20 62:1018=E2=80=931032, 2008
37.	Walker AE, Robins M,=20 Weinfeld FD: Epidemiology of brain = tumors:=20 the national survey of intracranial = neoplasms.=20 Neurology=20 35:219=E2=80=93226, 1985
38.	Wurm RE, Erbel S,=20 Schwenkert I, Gum F, Agaoglu D,=20 Schild R, et al.: Novalis frameless = image-guided=20 noninvasive radiosurgery: initial = experience.=20 Neurosurgery=20 62:A11=E2=80=93A18, 2008