Neurosurg Focus 15 (2):Article 8, 2003, Click here to return to Table of Contents

Radiosurgery and radiotherapy for sacral tumors

IRIS C. GIBBS, M.D., AND STEVEN D. CHANG, M.D. Departments of Radiation Oncology and , Stanford University School of Medicine, Stanford, California

Sacral tumors represent a small subset of spinal and typically include chordomas, metastases, other prima- ry tumors, and benign schwannomas. Resection is the standard treatment for many sacral tumors, but many types of sacral lesions have the potential for recurrence after excision. In these cases, adjuvant radiotherapy is often benefi- cial. Although conventional radiotherapy plays an important role in the management of spinal lesions, the radiation doses required for adequate local control of many sacral lesions generally exceed the tolerance doses of normal tissues, thus limiting its definitive role in the management of sacral tumors. Recent advances in the field of stereotactic radio- have allowed precise targeting of the sacrum. In this report the authors review the use of these two forms of radiation treatment and their role in managing sacral tumors.

KEY WORDS • chordoma • metastases • • radiotherapy • sacrum • stereotaxy

Sacral neoplasms are relatively rare.10 Those involving the effects on normal tissue. The optimal therapeutic ratio the sacrum fall into three major categories: primary be- is determined by maximizing tumor obliteration while nign or malignant lesions, metastatic lesions, or lesions minimizing normal tissue complications. By allowing for that secondarily involve the sacrum by local extension. repair of sublethal damage in normal tissues, radiation The majority are benign, with giant cell tumor of the sa- fractionation enhances the therapeutic ratio. For example, crum representing 70% of all sacral tumors.21 Although it is estimated that a conventionally fractionated course of the majority of malignant sacral tumors involve metasta- radiotherapy of 45 Gy in 22 to 25 fractions minimizes the tic spread from nearby pelvic sites or distant sites, prima- risk of radiation-induced myelopathy to an acceptable ry malignant tumors comprise a wide range of histolo- level of less than 0.2%, and higher doses significantly in- 38 gies including chordoma, osteosarcoma, chondrosarcoma, crease the risk. For sacral metastases, radiotherapy may myeloma, lymphoma, and Ewing sarcoma. Chordoma is be the initial treatment of choice, whereas in some cases the most common primary malignant sacral tumor. Tu- of primary sacral tumors, conventional radiotherapy is mors secondarily involving the sacrum by local extension used in conjuction with surgery for adequate treatment of are predominantly of neural precursor, connective tissue, the . or supportive tissue origin. This group includes neurofi- Radiotherapy for Sacral Metastases broma, schwannoma, ependymoma, epidermoid, and ter- atoma. We limit our discussion to the role of radiotherapy Spinal metastases will develop in 5 to 10% of all and radiosurgery in the management of the most common patients during the course of their illness.21 Metastatic sacral tumors. tumors represent the majority of malignant sacral tumors. These metastases generally arise by hematogenic dissem- ination of tumor cells from primary breast, lung, renal, RADIOTHERAPY thyroid, and prostate carcinomas.21 In addition to hemato- Initial complete removal of sacral tumors is rare be- genic spread, it is not unusual for locally recurrent pelvic cause of their proximity to neural and vascular structures. tumors such as rectal adenocarcinoma or advanced pro- Therefore, radiotherapy plays a central role in their man- static tumors to invade the sacrum by local extension. Al- agement. The goal of any radiation-based treatment is to though the most common presenting symptom is pain, the allow maximal destruction of the tumor while minimizing most catastrophic outcome is neurological compromise causing loss of bowel control, urinary incontinence, and lower-extremity weakness.36 Abbreviations used in this paper: GTC = giant cell tumor; Palliative radiotherapy as a first-line intervention for IMRT = intensity-modulated radiotherapy; LINAC = linear accel- spinal has been the mainstay of management erator; SRS = stereotactic radiosurgery. since the late 1960s after several authors compared radio-

Neurosurg. Focus / Volume 15 / August, 2003 1

Unauthenticated | Downloaded 09/24/21 11:15 PM UTC I. C. Gibbs and S. D. Chang therapy with laminectomy and found no significant differ- tion and the risks are minimal. Cummings, et al.,10 report- ence in patient outcome.13,20,28,42,44 More recently, a pro- ed no differences in survival, duration of symptomatic re- spective trial of 209 patients with metastatic spinal cord sponse, and progression-free survival in a heterogeneous compression were evaluated after radiotherapy and steroid group of patients who underwent 50- to 60-Gy radiother- treatment of the spinal metastases.27 The authors reported apy compared with conventional fractions to less than 50 that pain was alleviated in 82% of patients, ambulatory Gy or by a hyperfractionated regimen of 44 Gy delivered ability was regained or maintained in 76%, and sphincter in 1-Gy fractions, four times daily for 14 days. Much dysfunction was improved in 44%. Based on a thorough higher doses of 60 to 70 Gy are likely required for poten- review of patient outcomes presented in the literature, tially curative intent.3 Loblaw and Laperriere26 concluded that radiotherapy is an To achieve these higher doses, specialists at the Massa- appropriate first-line therapy unless osseous instability or chusetts General Hospital and Harvard Cyclotron Labora- acute paraplegia necessitate surgical intervention. tory use a combined treatment of x-ray and proton-beam irradiation.19 Although the biological properties of protons Radiotherapy for Benign Sacral Tumors are not significantly different from x-ray beams, their The most common benign tumors that affect the spine physical characteristics exhibited by a finite range of ener- include hemangioma, osteoid osteoma, osteoblastoma, os- gy deposition with only minimal exit dose beyond the tar- teochondroma, aneurysmal bone cyst, and GCT.36 Al- get region are significantly advantageous. These advan- though excision and curettage are the mainstay of treat- tages are particularly important in achieving high-dose ment in most patients with benign lesions, radiotherapy delivery to sacral chordomas. Using this approach of x-ray has been performed for aggressive osteoblastoma, and proton beam therapy combined with optimal surgery, 2,34,35 improved local control was achieved at doses greater than aneurysmal bone cyst, and GCTs. Because of the con- 19 cern for malignant transformation, radiotherapy is gener- 77 CGE (cobalt Gray equivalent). ally not recommended in the initial management of these lesions. The incidence of radiation-induced malignancy Radiotherapy Techniques after treatment of benign bone tumors has been reported to Conventional radiotherapy for spinal metastasis is be as high as 27%.11 In modern series in which improved delivered through simple portals to doses of 30 to 40 Gy radiation techniques are used, the reported incidence has in 2- to 3-Gy fractions. For isolated sacral lesions, an op- been lower.39,41 Even though less aggressive surgical tech- posed anterior–posterior field is typically used, encom- niques such as intralesional excision and cementation passing one or two vertebral segments above and below have evolved as preferred treatment of GCT, it may be the lesion. Alternatively, a three-field arrangement with impractical to expect a good functional outcome for some two lateral portals and a single posterior field is used to aggressive sacral lesions. In such cases, radiotherapy is spare anterior pelvic structures. A single posterior field recommended. Local control rates with radiotherapy for arrangement is generally avoided for sacral lesions, par- aggressive recurrent GCT have been reported to range ticularly if lower energy beams are used, because the treat- from 70 to 90% in contemporary series involving conven- ment depth will vary along the sacral hollow. More com- tional doses of 35 to 68 Gy. There are data that support the plex radiation-beam arrangements can be achieved using contention that local control of GCT is improved when three dimensional conformal radiotherapy. Three-dimen- administering doses greater than 40 Gy. Consequently, the sional conformal therapy is particularly useful when the recommended dose for conventional radiotherapy of these extraosseous component of tumor is irregularly shaped or lesions is 40 to 45 Gy.11 extends into the region of normal sensitive tissues of the rectum, small bowel, and bladder. This modality involves Radiotherapy for Primary Malignant Sacral Tumors an array of radiation beam arrangements with each field Although multiple myeloma is the most common bone shaped according to the beam’s eye view around the tar- tumor in adults, chordoma is the most common primary get volume. In this way, normal tissue is optimally shield- neoplasm to involve the sacrum.3,36 Sacral chordoma com- ed form the primary radiation beam. The use of IMRT can prises 50% of all spinal chordomas. Chordoma is a histo- potentially spare even more of normal structures by incor- logically originating from remnants of the porating inhomogeneous beam intensities directed at tar- notochord. Although it is a slow-growing tumor, it is get lesions from multiple beam directions. locally aggressive with a high propensity for local recur- rence and rare potential for cure.40 Radical resection is the initial treatment of choice. When this is possible, it ex- STEREOTACTIC RADIOSURGERY tends the time to local recurrence compared with subtotal For most sacral tumors, radiotherapy is a treatment op- excision. A recent retrospective review of 27 patients with tion; however, concern regarding radiation dose to sur- sacral chordomas treated at M.D. Anderson Cancer rounding normal structures may limit treatment doses to Center, York, et al.,43 demonstrated a statistically signifi- suboptimal levels.22 The incidence of radiation-induced cant difference in the disease-free interval after radical injuries to the lumbosacral plexus appears to increase with resection of sacral chordomas compared with subtotal ex- an escalated dose,38 and relapses within the radiation cision (2.27 years and 8 months, respectively) Further- fields are not uncommon. Sacral lesions also often have more, the addition of radiotherapy in cases of subtotally complex topology, making conformal dose delivery diffi- resected lesions prolonged the disease-free interval to 2.12 cult,14 even with the development of IMRT. years compared with 8 months without radiotherapy. Confining radiation doses strictly to a target volume Radiation doses of 30 to 50 Gy are indicated for pallia- within or around the sacrum should, theoretically, enhance

2 Neurosurg. Focus / Volume 15 / August, 2003

Unauthenticated | Downloaded 09/24/21 11:15 PM UTC Radiosurgery and radiotherapy for sacral tumors lesion control while decreasing the risk of injury to the The System adjacent normal structures. Delivering radiation with a The CyberKnife (Accuray Inc., Sunnyvale, CA) is a steep dose gradient is possible if accurately targeted. Such 6-MV therapeutic LINAC mounted on a robotic arm that accuracy is difficult when using conventional radiothera- directs the treatment beam via computer control and incor- py because manual positioning of the patient, which porates an x-ray imaging feedback system. During treat- involves attaching surface markers to the skin, limits pre- ment, the imaging system, which consists of two ortho- cision and is not reproducible. With radiosurgery, single- gonally aligned x-ray cameras, acquires radiographs of stage sacral treatments are possible in many cases. In the target landmarks. The image guidance process analyzes event that lesions size or radiation dose to critical struc- these real-time radiographs and can register landmarks to tures prevents single-stage treatment, multiple stereotactic the treatment planning study to determine the target’s treatments (staged radiosurgery) are feasible and have ad- position in six dimensions.29 The process then sends con- vantages over conventional radiotherapy because each trol signals to the robot to adjust the originally planned stage allows irradiation of the target volume in a highly 23 treatment beam directions to accommodate any changes in precise and reproducible manner. Frameless image-guid- target position. This feedback eliminates the need for rigid ed radiosurgery systems such as the CyberKnife, which immobilization. The LINAC then delivers radiation treat- allows both spatially precise treatment delivery and re- ment beams from nearly anywhere around the patient producible multiple doses, allow treatment of lesions 1,4–9 according to the previously generated plan. This process is throughout the spine, even at the level of the sacrum. completely automatic and repeated at defined intervals during treatment. Near real-time target updating aligns the Technological Basis of Spinal Radiosurgery dose to the treatment volume with millimeter precision. Conventional intracranial SRS involves using a rigid Sacral lesions can be treated with minimal modification of frame to reference the target lesion to a known location in standard hardware and software.30,31,37 space. In stereotactic targeting in the it is assumed that the target has a fixed relationship with respect to the Spinal Procedure cranium. Because spinal lesions generally have a fixed re- Simple Immobilization. At our center patients with lationship to the osseous spine, stereotactic methods in- sacral tumors are fitted with an AlphaCradle (Smithers volving spinal fixation are analogous to frame-based ste- Medical Products, Inc., Akron, OH), a custom body mold reotactic localization of intracranial lesions. filled with a plastic similar to styrofoam. This Alpha- The first spinal stereotactic surgery is attributed to Wo- 25 Cradle maintains the general orientation of patients while roschiloff in 1874. He used a spinal localization device simultaneously restricting motion during pretreatment attached to spinous processes to make precise cord lesions scanning and during treatment. for in vivo electrophysiological studies. In 1972, Nad- vornik33 used a similar method of open spinal fixation to Fiducial Placement. The current CyberKnife system re- perform radiofrequency of the cord. Clark is said quires implanted fiducial markers for all spinal lesion reg- to have modified the attachment to the lamina for more istrations. Fiducials are placed during outpatient proce- 18 32 dures, typically lasting 15 minutes, and can be placed days accurate localization later that year. Nadvornik, et al., later modified the Woroschiloff system to perform stereo- to weeks before treatment. Four 2 6–mm surgical stain- tactic lesioning of the lumbar spinal cord. Lax, et al.,24 less-steel self-retaining tacks or self-tapping screws were proposed an external stereotactic frame that enclosed the placed through stab skin incisions under fluoroscopy in a body from the head to the middle femoral area with foam noncoplanar pattern within the bone tissue of the sacrum. pad stabilization, which was not rigid and permitted pa- Treatment Planning. Following AlphaCradle immobi- tient movement. lization and fiducial placement, contrast-enhanced com- Hamilton with Lulu15 and colleagues16,17 developed a puterized scans of the sacrum were acquired spinal SRS system that involved a modified LINAC and a for treatment planning and to provide a basis for generat- skeletal fixation frame. In the Hamilton–Lulu extracranial ing the digitally reconstructed radiographs used during stereotactic frame the patient was placed in a rigid box. position tracking. The treatment system uses an inverse Immobilization was achieved by fixing the frame to spi- planning solution. The plans are designed in the following nous processes of the patient after induction of general manner: the volume of the lesion to be treated is defined; anesthesia.14 The results were promising in many patients; the desired prescription dose is specified; volumes of sen- however, the fixation was cumbersome, resulted in long sitive structures such as the spinal cord are defined and the procedures, and was not feasible for fractionated delivery. maximum radiation doses to these critical regions are Image-guided SRS was developed to deliver nonframe- specified; and computed optimization then produces the based treatment with the precision of frame-based sys- best-fit treatment plans that conforms to the defined con- tems. This technology has three fundamental differences straints. from conventional frame-based radiosurgical systems. 1) Treatment Delivery. Patients with sacral lesions are Referencing is based on internal radiographic features typically treated in the supine position. The robotic arm such as skeletal anatomy or implanted fiducials. 2) Near moves the LINAC sequentially through the planned beam real-time radiographs are acquired to evaluate changes in positions delivering the planned dose. At each position, target location and to adapt the treatment delivery to ac- the imaging system can check the target position and send commodate these motions. 3) Fixed isocenters are not re- corrective pointing directions to the robot. These position quired, which allows for irregular dose shapes and homo- checks are performed approximately once a minute, al- geneity. though checks are made more f requently in patients with

Neurosurg. Focus / Volume 15 / August, 2003 3

Unauthenticated | Downloaded 09/24/21 11:15 PM UTC I. C. Gibbs and S. D. Chang more frequent movements. All patients are treated on an cral and received single-stage doses up to 18 Gy. No radi- out-patient basis, with staged treatments given on consec- ation-induced side effects were noted. utive days. CONCLUSIONS Clinical Experience With Radiosurgery Although sacral neoplasms represent only a small sub- At Stanford University School of Medical, we reviewed set of spinal tumors, their treatment can be challenging our experience of 53 patients with spinal lesions treated because of their involvement with adjacent neural and with the CyberKnife. Three of these patients harbored vascular structures. Conventional radiotherapy has been lesions involving the sacrum. The first patient harbored a performed for decades as either primary or adjunctive schwannoma of the L5–S1 nerve root; the neural foramen treatments for sacral tumors. More conformal forms of was widened secondary to local bone erosion along the radiotherapy, including IMRT, have the potential to reduce superior edge of the sacrum. This patient underwent 18- complications by including lower volumes of normal tis- Gy SRS to the 75% isodose line in two stages. The second sue within treatment volumes. Stereotactic radiosurgery patient had previously undergone surgery and radiothera- represents the most conformal treatment to date, and early py for an L-3 ependymoma when follow-up magnetic res- experiences at several institutions have shown this modal- onance imaging revealed a new S1–2 ependymoma (drop ity to be effective for both primary and secondary sacral metastasis). The lesion was treated with 18 Gy to the 80% lesions. isodose line in a single stage. A third patient underwent SRS for a large renal cell metastasis of the sacrum for Acknowledgment which previous 40-Gy radiotherapy had failed (Fig. 1). This tumor received a dose of 18 Gy delivered in two We thank Beth Hoyte for assistance with production of the stages to the 76% isodose line. There were no SRS-relat- figures. ed side effects in any of the three patients. Gerszten, et References al.,12 reviewed their series of 56 spinal tumors treated in 46 patients. Four of the tumors (all metastases) were sa- 1. Adler JR Jr, Murphy MJ, Chang SD, et al: Image-guided robot- ic radiosurgery. Neurosurgery 44:1299–1307, 1999 2. Akbarnia BA: Osteoid osteoma and osteoblastoma of the spine, in Akbarnia BA (ed): Spine: State of the Art Reviews. Phila- delphia: Hanley & Belfus, 1988, Vol 2, pp 249–263 3. Aprin H: Primary Malignant Tumors of the Spine, in Akbarnia BA (ed): Spine: State of the Art Reviews. Philadelphia: Han- ley & Belfus, 1988, Vol 2, pp 289–299 4. Chang SD, Adler JR Jr, Murphy MJ: Stereotactic radiosurgery of spinal lesions, in Maciunas RJ (ed): Advanced Techniques in Central Nervous System Metastases. Park Ridge, IL: American Association of Neurological , 1998, pp 269–276 5. Chang SD, Murphy M, Geis P, et al: Clinical experience with image-guided robotic radiosurgery (the Cyberknife) in the treatment of brain and spinal cord tumors. Neurol Med Chir 38:780–783, 1998 6. Chang SD, Murphy MJ, Doty JR, et al: Stereotactic radiosur- gery: new innovations, in Fisher WS III (ed): Perspectives in Neurological Surgery. Baltimore: Williams and Wilkins, 1999, pp 145–153 7. Chang SD, Murphy MJ, Martin DP, et al: Frameless stereotac- tic radiosurgery, in Petrovich Z, Brady LW, Apuzzo ML, et al (eds): Combined Modality Therapy of Central Nervous Sys- tem Tumors. New York: Springer, 2000, pp 387–396 8. Chang SD, Murphy MJ, Martin DP, et al: Image-guided robot- ic radiosurgery: clinical and radiographic results with the Cy- berKnife, in Kondziolka D (ed): Radiosurgery 1999. Basel: Karger, 2000, Vol 3, pp 23–33 9. Chang SD, Murphy MJ, Tombropoulos R, et al: Robotic ra- diosurgery, in Alexander E III, Maciunas RJ (eds): Advanced Fig. 1. Treatment plan obtained in a 69-year-old woman previ- Neurosurgical Navigation. New York: Thieme, 1998, pp ously treated with 40-Gy conventional radiotherapy for a large 443–449 renal metastasis involving the sacrum. Worsening sacral pain 10. Cummings BJ, Hobson DI, Bush RS: Chordoma: the results of developed, however, and she was referred for radiosurgery. megavoltage . Int J Radiat Oncol Biol Phys Radiographic evaluation revealed a large 7.8 7.1 7.3–cm 9:633–642, 1983 metastasis. The patient underwent CyberKnife-based SRS with 18 11. Feigenberg SJ, Marcus Jr RB, Zlotecki RA, et al: Radiation Gy delivered to the 76% isodose line in two 9-Gy stages. therapy for giant cell tumors of bone. Clin Orthop 41: Radiosurgical treatment plan shows the tumor and adjacent isodose 207–216, 2003 lines in axial (A), sagittal (B), and coronal views (C). D: A three- 12. Gerszten PC, Ozhasoglu C, Burton SA, et al: Feasibility of dimensional representation of the patient’s pelvis shows the vari- frameless single-fraction stereotactic radiosurgery for spinal ous trajectories of the multiple radiosurgical beams. lesions. Neurosurg Focus 13 (4):Article 2, 2002

4 Neurosurg. Focus / Volume 15 / August, 2003

Unauthenticated | Downloaded 09/24/21 11:15 PM UTC Radiosurgery and radiotherapy for sacral tumors

13. Gilbert RW, Kim JH, Posner JB: Epidural spinal cord compres- 30. Murphy MJ, Adler JR Jr, Bodduluri M, et al: Image-guided sion from metastatic tumor: diagnosis and treatment. Ann Neu- radiosurgery for the spine and pancreas. Comput Aided Surg rol 3:40–51, 1978 5:278–288, 2000 14. Hamilton AJ: Stereotactic radiosurgery for spinal metastases, in 31. Murphy MJ, Chang SD, Gibbs IC, et al: Image-guided radio- Maciunas RJ (ed): Advanced Techniques in Central Nervous surgery in the treatment of spinal metastases. Neurosurg Focus System Metastases. Park Ridge, IL: American Association of 11 (6):Article 6, 2001 Neurological Surgeons, 1998, pp 255–268 32. Nadvornik P: Woroschiloff’s locating device for interventions 15. Hamilton AJ, Lulu BA: A prototype design for linear accelera- on the spinal cord and its influence on spinal stereotaxis. Appl tor-based extracranial radiosurgery. Acta Neurochir Suppl 63: Neurophysiol 48:247–251, 1985 40–43, 1995 33. Nadvornik P, Frohlich J, Jezek V, et al: New apparatus for spi- 16. Hamilton AJ, Lulu BA, Fosmire H, et al: LINAC-based spi- nal cord stereotaxis and its use in the microsurgery of lumbar nal stereotactic radiosurgery. Stereotact Funct Neurosurg 66: enlargement. Confin Neurol 34:311–314, 1972 1–9, 1996 34. Papagelopoulos PJ, Choudhury SN, Frassica FJ, et al: Treat- 17. Hamilton AJ, Lulu BA, Fosmire H, et al: Preliminary clinical ment of aneurysmal bone cysts of the pelvis and sacrum. J experience with linear accelerator-based spinal stereotactic ra- Bone Joint Surg Am 83:1674–1681, 2001 diosurgery. Neurosurgery 36:311–319, 1995 35. Papagelopoulos PJ, Currier BL, Shaughnessy WJ, et al: Aneu- 18. Hitchcock ER: Stereotaxis of the spinal cord. Confin Neurol rysmal bone cyst of the spine. Management and outcome. 34:299–310, 1972 Spine 23:621–628, 1998 19. Hug EB, Fitzek MM, Liebsch NJ, et al: Locally challenging 36. Raque GH Jr, Vitaz TW, Shields CB: Treatment of neoplastic osteo- and chondrogenic tumors of the axial skeleton: results of diseases of the sacrum. J Surg Oncol 76:301–307, 2001 combined proton and photon radiation therapy using three- 37. Ryu SI, Chang SD, Kim DH, et al: Image-guided hypo-fraction- dimensional treatment planning. Int J Radiat Oncol Biol Phys ated stereotactic radiosurgery to spinal lesions. Neurosurgery 31:467–476, 1995 49:838–846, 2001 20. Khan FR, Glicksman AS, Chu FCH, et al: Treatment by radio- 38. Schultheiss TE, Kun LE, Ang KK, et al: Radiation response of therapy of spinal cord compression due to extradural metasta- the central nervous system. Int J Radiat Oncol Biol Phys 31: ses. Radiology 89:495–500, 1967 1093–1112, 1995 21. Kollender Y, Meller I, Bickels J, et al: Role of adjuvant cryo- 39. Schwartz LH, Okunieff PG, Rosenberg A, et al: Radiation ther- surgery in intralesional treatment of sacral tumors. Cancer 97: apy in the treatment of difficult giant cell tumors. Int J Radiat 2830–2838, 2003 Oncol Biol Phys 17:1085–1088, 1989 22. Kopelson G, Linggood R, Kleinman G: Management of intra- 40. Sundaresan N, Huvos AG, Krol G, et al: Surgical treatment of medullary spinal cord tumors. Radiology 135:473–479, 1980 spinal chordomas. Arch Surg 122:1479–1482, 1987 23. Larson DA, Flickinger JC, Loeffler JS: The of 41. Turcotte RE, Sim FH, Unni KK: Giant cell tumor of the sacrum. radiosurgery. Int J Radiat Oncol Biol Phys 25:557–561, 1993 Clin Orthop:215–221, 1993 24. Lax I, Blomgren H, Naslund I, et al: Stereotactic radiotherapy 42. Vieth RG, Odom GL: Extradural spinal metastases and their of malignancies in the abdomen. Methodological aspects. Acta neurosurgical treatment. J Neurosurg 23:501–508, 1965 Oncol 33:677–683, 1994 43. York JE, Kaczaraj A, Abi-Said D, et al: Sacral chordoma: 40- 25. Leksell L: The stereotactic method and radiosurgery of the year experience at a major cancer center. Neurosurgery 44: brain. Acta Chir Scand 102:316–319, 1951 74–80, 1999 26. Loblaw DA, Laperriere NJ: Emergency treatment of malignant 44. Young RF, Post EM, King GA: Treatment of spinal epidural extradural spinal cord compression: an evidence-based guide- metastases. Randomized prospective comparison of laminecto- line. J Clin Oncol 16:1613–1624, 1998 my and radiotherapy. J Neurosurg 53:741–748, 1980 27. Maranzano E, Latini P: Effectiveness of radiation therapy with- out surgery in metastatic spinal cord compression: final results from a prospective trial. Int J Radiat Oncol Biol Phys 32: 959–967, 1995 Manuscript received June 16, 2003. 28. Marshall LF, Langfitt TW: Combined therapy for metastatic ex- Accepted in final form July 10, 2003. tradural tumors of the spine. Cancer 40:2067–2070, 1977 Address reprint requests to: Iris C. Gibbs, M.D., Stanford Uni- 29. Murphy MJ: An automatic six-degree-of-freedom image regis- versity Medical Center, Department of Radiation Oncology, Room tration algorithm for image-guided frameless stereotaxic radio- A-095, Stanford, California 94305-5302. email: iris.gibbs@ surgery. Med Phys 24:857–866, 1997 stanford.edu.

Neurosurg. Focus / Volume 15 / August, 2003 5

Unauthenticated | Downloaded 09/24/21 11:15 PM UTC