Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Look up keyword
Like this
1Activity
0 of .
Results for:
No results containing your search query
P. 1
Dodd Spine Neurosurg 2006

Dodd Spine Neurosurg 2006

Ratings: (0)|Views: 3|Likes:
Published by Jacob Repko

More info:

Published by: Jacob Repko on May 31, 2013
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

11/21/2014

pdf

text

original

 
CLINICAL STUDIES
C
YBER
K
NIFE
R
ADIOSURGERY FOR
B
ENIGN
I
NTRADURAL
E
XTRAMEDULLARY
S
PINAL
T
UMORS
Robert L. Dodd, M.D., Ph.D.
Department of Neurosurgery,Stanford UniversitySchool of Medicine,Stanford, California
Mi-Ryeong Ryu, M.D., Ph.D.
Department of Radiation Oncology,Kangnam St. Mary’s Hospital,Seoul, Korea
Pimkhuan Kamnerdsupaphon,M.D.
Division of Therapeutic Radiologyand Oncology,Faculty of Medicine,Chiang Mai University,Chiang Mai, Thailand
Iris C. Gibbs, M.D.
Department of Radiation Oncology,Stanford UniversitySchool of Medicine,Stanford, California
Steven D. Chang, Jr., M.D.
Department of Neurosurgery,Stanford UniversitySchool of Medicine,Stanford, California
John R. Adler, Jr., M.D.
Department of Neurosurgery,Stanford UniversitySchool of Medicine,Stanford, California
Reprint requests:
 John R. Adler, Jr., M.D.,Department of Neurosurgery,Stanford UniversitySchool of Medicine,Stanford, CA 94305.Email: jra@stanford.edu
Received,
July 20, 2005.
Accepted,
December 7, 2005.
OBJECTIVE:
Microsurgical resection of benign intradural extramedullary spinal tumorsis generally safe and successful, but patients with neurofibromatosis, recurrent tumors,multiple lesions, or medical problems that place them at higher surgical risk maybenefit from alternatives to surgery. In this prospective study, we analyzed our pre-liminary experience with image-guided radiosurgical ablation of selected benignspinal neoplasms.
METHODS:
Since 1999, CyberKnife (Accuray, Inc., Sunnyvale, CA) radiosurgery wasused to manage 51 patients (median age, 46 yr; range, 12–86 yr) with 55 benign spinaltumors (30 schwannomas, nine neurofibromas, 16 meningiomas) at Stanford Univer-sity Medical Center. Total treatment doses ranged from 1600 to 3000 cGy delivered inconsecutive daily sessions (1–5) to tumor volumes that varied from 0.136 to 24.6 cm
3
.
RESULTS:
Less than 1 year postradiosurgery, three of the 51 patients in this series (onemeningioma, one schwannoma, and one neurofibroma) required surgical resection of their tumor because of persistent or worsening symptoms; only one of these lesionswas larger radiographically. However, 28 of the 51 patients now have greater than 24months clinical and radiographic follow-up. After a mean follow-up of 36 months, allof these later lesions were either stable (61%) or smaller (39%). Two patients died fromunrelated causes. Radiation-induced myelopathy appeared 8 months postradiosurgeryin one patient.
CONCLUSION:
Although more patients studied over an even longer follow-up periodare needed to determine the long-term efficacy of spinal radiosurgery for benignextra-axial neoplasms, short-term clinical benefits were observed in this prospectiveanalysis. The present study demonstrates that CyberKnife radiosurgical ablation of such tumors is technically feasible and associated with low morbidity.
KEY WORDS:
CyberKnife, Image guidance, Meningioma, Neurofibroma, Radiosurgery, Schwannoma,Spinal tumors
Neurosurgery 58:674-685, 2006 
DOI: 10.1227/01.NEU.0000204128.84742.8F
www.neurosurgery-online.com
S
tereotactic radiosurgery (SRS) has be-come an important tool for managing arange of benign intracranial and cranial base tumors. Although similar tumor histolo-gies can be found in and around the spine, thefirst generation of radiosurgical instrumentswas, by virtue of being frame-based, unable totreat such extracranial targets. The more re-cent emergence of image-guidance technologynow makes it possible to use radiosurgicalmethods to ablate lesions throughout the bodyincluding in the spine (1, 45, 56). There arenow several published studies that describeoutcome in series of patients treated with spi-nal radiosurgery, primarily for metastases (9,10, 14, 41, 45). Although these preliminarystudies demonstrate safety and efficacy,follow-up is relatively short because these pa-tients tend to die early from their underlyingmalignancies.The safety and effectiveness of microsurgi-cal resection of most benign spinal neoplasmsis well documented (8, 15, 24, 37, 48, 49). Nev-ertheless, certain patients are less than idealcandidates for standard surgical resection be-cause of age, medical comorbidities, the recur-rent nature of their tumor, or because multiplelesions occur in the setting of one of the famil-ial phakomatoses. It is in such clinical circum-stances that radiosurgery could be an attrac-
674
| VOLUME 58 | NUMBER 4 | APRIL 2006
www.neurosurgery-online.com
 
tive clinical option. Despite this theoretical attraction, theliterature on malignant spinal metastases provides an insuffi-cient basis for judging the possible benefits of radiosurgicalablation of benign spine tumors. Because patients with suchlesions have prolonged life expectancies, the potential fordelayed and possibly catastrophic radiation myelopathy is aspecial concern. In addition, benign spine tumors have theirown unique presentation, spatial relationship to the spinalcord, and radiobiologic response to radiosurgery, any of which could present unique challenges to the safe and effec-tive application of radiosurgical ablation.Image-guided robotic radiosurgery, the CyberKnife (Ac-curay, Inc., Sunnyvale, CA), was introduced at Stanford Uni-versity Medical Center in 1994 (1), and in 1997, the first patientwith a benign spinal lesion (a hemangioblastoma) was treated.Since that time, 101 patients with a variety of benign spinaltumors and vascular malformations have undergone radiosur-gical ablation at our institution. Among this group were 51patients with meningioma, neurofibroma, or schwannomawho now have at least 6 months follow-up. In this report, weprovide a prospective analysis of this cohort. Our primaryobjectives were to 1) document the relative safety of thisprocedure and 2) develop a preliminary understanding of therange of potential treatment doses and outcome measures thatare appropriate for patients with benign spine tumors.
PATIENTS AND METHODS
Fifty-one patients with 55 benign intradural extramedullaryspinal tumors (16 meningiomas, 30 schwannomas, nine neu-rofibromas) were treated at Stanford University Medical Cen-ter from 1999 to 2005 as part of a protocol for spinal tumorsapproved by our institutional review board. A multidisci-plinary team of specialists that included neurosurgeons, radi-ation oncologists, and neuroradiologists evaluated all patients.Spinal radiosurgery was only offered to patients for whommicrosurgical resection was contraindicated because of med-ical comorbidities, underlying neurofibromatosis (NF) thatresulted in multiple lesions developing over time, or, occa-sionally, strong patient preference. Selected cases also hadwell-circumscribed lesions, no evidence of overt spinal insta- bility, and generally minimal compromise of spinal cord func-tion (i.e., myelopathy). Several of the NF patients in this serieshad an aggressive form of this illness that resulted at the timeof presentation in significant neurological compromise. Alltumors were known or presumed to be spinal meningioma,neurofibroma, or schwannoma on the basis of prior patholog-ical confirmation, characteristic appearance on contrast mag-netic resonance imaging (MRI) scans, or history of NF.The general clinical characteristics for the 51 patients in thisseries are summarized in
Tables 1–3
. After signing an institu-tional review board-approved consent form, the image-guidedCyberKnife radiosurgical system was used to administer spi-nal radiosurgery in every case. This instrument has been de-scribed previously in detail by Adler et al. (1). The deviceconsists of a 6 MV linear accelerator mounted on a computer-controlled robotic arm, which is coupled to an x-ray trackingsystem that monitors and adjusts in near real time on the basisof changes in the target’s position. Image-guidance eliminatesthe need for skeletal rigid immobilization.The process of CyberKnife spinal radiosurgery begins withan outpatient procedure that inserts small stainless steel mark-ers percutaneously into vertebral segments above and belowthe radiosurgical target (45). Next, a custom alpha cradle moldis fabricated for the supine patient (Smithers Medical Prod-ucts, Inc., Akron, OH). This device is used to noninvasivelyimmobilize the spine during computerized tomographic (CT)imaging (used later in treatment planning) and also to restrictpatient movement during the radiosurgery itself. For thosetumors that enhanced poorly with iodinated contrast or whensevere allergy precluded the acquisition of an enhanced CTimaging, a contrast MRI scan of the relevant spine was alsoacquired and fused to the pretreatment CT scan.Treatment planning begins with the treating neurosurgeondefining the target volume and critical structures within theCT/MRI scans using software tools provided on the Cy- berKnife workstation (
Fig. 1, D–F
). On imaging studies, thespinal cord volume was delineated as the primary “criticalstructure” beginning 1 cm cephalad to and ending 1 cm belowthe targeted lesion. A proprietary inverse planning computeralgorithm uses the above inputs to determine the number,direction, and duration of treatment beamlets so as to opti-mize dose conformality and minimize irradiation of criticalstructures (
Fig. 1G
). Visual inspection and analysis of dosevolume histograms for the target region and adjacent criticalanatomy are used to of find the best radiosurgical solution. Just before the administration of radiosurgery, a library of digitally reconstructed radiographs (i.e., computer simulatedx-rays) are calculated from the perspective of a pair of x-raysources and cameras used throughout the operation. Thisarray of images encompasses those vertebral elements, alongwith embedded fiducials, that are in close proximity to theradiosurgical target. During radiosurgery, the patient lies su-
TABLE 1. Characteristics of patients
Age (yr)
Mean
46.5
Range 
12.6–86.5Sex, n (%)
Female 
23 (45)
Male 
28 (55)Previously resected (%)
Subtotal 
24 (47)
Gross total 
2 (4)NF1 7 (14)NF2 10 (20)Previously radiated (%) 4 (8)
a
NF, neurofibromatosis.
C
YBER
K
NIFE
R
ADIOSURGERY FOR
B
ENIGN
S
PINAL
T
UMORS
N
EUROSURGERY 
VOLUME 58 | NUMBER 4 | APRIL 2006 |
675
 
pine on the operating table in the alpha cradle mold. Sequen-tial paired digital radiographs of the target region are thenobtained by ceiling-mounted, orthogonally directed, rigidlyfixed x-ray tubes. A computer workstation performs rapidimage-to-image correlation between the acquired images andthe previously calculated digitally reconstructed radiographs.Before each beamlet of radiation is administered, the x-rayimaging system determines target location and communicatesthe answer to the robot. The robot adjusts for small patientmovements by automatically realigning the treatment beamwith submillimeter accuracy (5).Postradiosurgical follow-up, typically performed at 3months, 6 months, 1 year, and then annually thereafter, in-cluded clinical evaluation, physical examination, and radio-graphic imaging. The formula for an idealized ellipsoid, Vol
ϭ
4/3
␲ 
(length
ϫ
width
ϫ
height) was used to estimate tumorvolume on contrast MRI scans. Pain was qualitatively assessed by patient report as either improved, stable, or worse andsemiquantitatively by recording a patient’s analgesia require-ment. This information was placed into a prospectively main-tained computer database. The median follow-up after radio-surgery for the entire series was 23 months (mean, 25 mo;range, 6–73 mo). Similarly, the median follow-up for eachhistological subtype was 25, 23, and 21.5 months in meningi-oma, schwannoma, and neurofibroma, respectively (mean,27.2, 26, and 19.9 mo, respectively).
RESULTS
Fifty-one patients (28 men, 23 women; median age, 46 yr;age range, 12–86 yr) with 55 intradural extramedullary benignspinal tumors were treated with multisession radiosurgeryusing the CyberKnife radiosurgical system (
Table 1
). A femalepredominance was observed among spinal meningiomas,whereas the male to female ratio in schwannomas and neuro-fibromas was 1.6:1 and 2:1, respectively. Twenty-six (51%)patients had undergone a previous surgical resection andwere being treated for residual or recurrent tumor. Four pa-tients developed tumors in radiation fields for other cancers(e.g., Hodgkin’s lymphoma, breast adenocarcinoma). One pa-tient developed a traumatic schwannoma (22, 43, 55) aftersurgery for removal of a synovial cyst. Seventeen patientscarried a diagnosis of either NF Type 1 (NF1) or NF2. Tumorswere observed throughout the entire spinal axis (
Table 2
) andvaried in configuration from entirely intraspinal to dumbbellshaped to predominantly foraminal. Presenting symptoms(pain, radiculopathy, and myelopathy) varied depending onspinal location and the precise relationship between the tumorand adjoining nerves/spinal cord (
Table 3
). Very few patientshad significant spinal cord compression and myelopathy be-fore radiosurgery. Furthermore, most patients presenting withmyelopathy developed this neurological deficit as a conse-quence of either previous tumor-related cord compression orspinal surgery. Eight (15%) asymptomatic patients underwentpreemptive radiosurgical ablation because of the size, loca-tion, or growth of their tumor on serial MRI scans.
Radiosurgical Doses and Fractionation
The specific fractionation schedule (median of two sessions;range, 1–5) was based on the size and volume of the treatedtumor as well as the length and total dose administered to thespinal cord. For intracranial meningiomas and nerve sheathtumors, it is widely accepted that an effective single dose canrange from 12 to 18 Gy depending on the size of the lesion (6).Because with spinal lesions there is added concern about thedose tolerance of the spinal cord, and because of our previouslack of experience with paraspinal radiosurgery, we incorpo-rated staging in our dose selection for a substantial number of these tumors. The choice of dose and staging schedule wasselected to minimize the risk of spinal cord injury. With use of the concepts of the linear quadratic model, the biologicalequivalent dose (BED) is estimated by the following formula:BED
ϭ
nd(1
ϩ
d/
/
). Multisession radiosurgical regimenswere devised that achieved a BED of 53 to 180 Gy, dependingon the proximity to the spinal cord, history of prior spinal cordirradiation, and total tumor volume. These BED values com-pare favorably with the single fraction guidelines. (60–126 Gy)for benign brain tumors. In most patients, radiosurgery wasdelivered in one (37%) or two sessions (42%). However, addi-tional daily sessions were administered in three (eight pa-tients), four (two patients), or five sessions (1 patient). Acutetoxicity was rare and limited to short-lived nausea.
Table 4
summarizes the radiosurgical dosimetry used in thisseries. Target volumes ranged from 0.136 to 24.6 cm
3
(mean4.29 cm
3
; median 2.18 cm
3
). Treatment plans were designed to
TABLE 2. Characteristics of lesionsNo. of lesions (%)
Level
Cervical 
38 (69)
Thoracic 
7 (13)
Lumbar 
8 (15)
Sacral 
2 (4)Histology
Schwannoma
16 (29)
Neurofibroma
9 (16)
Meningioma
30 (54)
TABLE 3. Presenting symptoms
Local or radicular pain 34 (63%)Radicular sensory loss 26 (48%)Radicular weakness 22 (41%)Myelopathic weakness 12 (22%)Axial sensory loss 4 (7%)Bladder paresis 3 (6%)Asymptomatic 8 (15%)
D
ODD ET AL
.
676
| VOLUME 58 | NUMBER 4 | APRIL 2006
www.neurosurgery-online.com

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->