Metastatic Disease in the Thoracic and Lumbar Spine: Evaluation and Management

Review Article Metastatic Disease in the Thoracic and Lumbar Spine: Evaluation and Management

Abstract Peter S. Rose, MD Spinal metastases are found in most patients who die of cancer. Jacob M. Buchowski, MD, MS The number of patients with symptomatic spinal metastases likely will increase as therapy for the primary disease improves and as cardiovascular mortality decreases. Understanding the epidemiology of metastatic spine disease and its presentation is essential to developing a diagnostic strategy. Treatment may involve chemotherapy, corticosteroids, radiotherapy, surgery, and/or percutaneous procedures (eg, vertebroplasty, kyphoplasty). A rational treatment plan can help improve quality of life, preserve neurologic function, and prolong survival.

pinal metastases can present as a with metastatic disease as the result Sprogression of known cancer or as of new treatment modalities will a primary malignancy. They are a likely result in increased incidence of From the Department of source of considerable morbidity in pa- spinal metastases. Orthopaedic Surgery, Mayo Clinic, Rochester, MN (Dr. Rose), and the tients with disseminated malignancies. The skeletal system is the third Department of Orthopaedic Surgery, Early diagnosis is essential in reducing most common organ affected by Washington University, St. Louis, pain, improving or preserving neuro- metastatic cancer, after the lungs and MO (Dr. Buchowski). logic function, and maximizing quality the liver. As many as 70% of patients Dr. Rose or an immediate family of life.1 Spinal metastases are catego- who die of cancer have been shown member serves as a board member, rized as intradural or extradural. Ex- on autopsy to have spinal metasta- owner, officer, or committee member ≤ of the Scoliosis Research Society, tradural lesions account for 90% to ses, and 14% exhibit clinically 2,3 Musculoskeletal Tumor Society, 95% of spinal metastases. Careful symptomatic disease before death. In American Academy of Orthopaedic physical examination and imaging the United States, >20,000 patients Surgeons, and Minnesota studies (eg, radiography, CT, MRI) present with metastatic epidural spi- Orthopaedic Society. Dr. Buchowski 5,6 or an immediate family member aid in making the diagnosis. Treat- nal cord compression annually. serves as a board member, owner, ment must be individualized to each Breast, prostate, lung, renal, and officer, or committee member of the patient and may include nonsurgical hematopoietic tumors are most likely American Academy of Orthopaedic or palliative measures (eg, cortico- to metastasize to the spine. The tho- Surgeons, the North American Spine Society, the Scoliosis Research steroids, radiotherapy) or interven- racic spine is most commonly in- Society, and the Spine Arthroplasty tional treatments (eg, surgery, verte- volved in metastatic disease, possibly Society; and is a member of a broplasty, kyphoplasty). because it contains the greatest vol- speakers’ bureau or has made paid presentations on behalf of and ume of bone marrow for receiving 7 serves as a paid consultant to or is metastatic deposits. an employee of Stryker. Epidemiology

J Am Acad Orthop Surg 2011;19: Cancer is the second leading cause of 37-48 Presentation death in the United States, with Copyright 2011 by the American >550,000 cancer-related deaths each Pain is the most common symptom Academy of Orthopaedic Surgeons. year.4 Prolonged survival in patients in approximately 90% of patients

January 2011, Vol 19, No 1 37 Metastatic Disease in the Thoracic and Lumbar Spine: Evaluation and Management

Table 1 metastatic cancer who presents with spinal metastases requires further Neurologic Impairment Scales9,10 evaluation prior to treatment. Repeat Scale Grade Deﬁcit Below the Level of the Lesion oncologic staging studies are needed, ASIA A Complete: no motor or sensory function including CT scan of the chest, abdo- B Incomplete: sensory function but no motor function men, and pelvis as well as a total C Incomplete: some motor function; most muscle groups body scan (eg, bone, positron emis- grade <3 sion tomography). Typically, biopsy D Incomplete: some motor function; most muscle groups is done to confirm the initial metas- grade ≥3 tasis. Staging studies often reveal E Normal motor and sensory function other lesions that are technically eas- Frankel A Complete motor and sensory loss ier and safer to biopsy than vertebral B Complete motor loss; incomplete sensory loss lesions. For the patient with cancer, C Sensory function useless; some motor function; no the initial diagnosis of metastatic dis- functional strength ease carries significant treatment, D Sensory function useful; weak but useful motor function prognostic, and emotional weight. E Normal motor and sensory function Although restaging and biopsy may seem tedious to the patient, they are ASIA = American Spinal Injury Association essential in determining a treatment plan. with metastatic spine disease.8 Pain is recognition and treatment. Patients with no prior diagnosis of often poorly characterized, and it Compression fractures are com- cancer may present with vertebral le- may be difficult to distinguish from mon in patients with metastatic dis- sions. Spinal metastases are the ini- more typical causes of back pain. ease of the spine, and patient history tial manifestation of malignancy in However, progressive and unrelent- and/or imaging studies typically are approximately 20% of patients who 12 ing pain that is nonmechanical in na- indicative of the presence of a malig- present with vertebral metastases. ture and is present at night is nant lesion. For example, unrecog- The most common histologies that strongly indicative of malignant eti- nized cancer is rare in patients who present in this manner are carcinoma ology. undergo kyphoplasty with biopsy for of unknown origin, carcinoma of the Neurologic signs and symptoms presumed osteoporotic fracture. In lung, multiple myeloma, and lym- are frequently present, but they our experience, metastatic disease in- phoma. An established diagnostic rarely precede axial pain. Eccentri- volving the thoracic and lumbar strategy is required for these pa- cally located tumors may cause rapid spine rarely presents with frankly tients. In a prospective study, biopsy onset of radiculopathy, and neural displaced fractures, whereas lesions alone failed to identify the primary compression resulting from epidural in the cervical spine are more likely tumor in two thirds of patients with 13 extension or fracture can produce to present with such fracture. these histologies. Using a protocol myelopathy or cauda equina syn- that included history and physical drome. Few patients initially present examination, routine laboratory with frank quadriplegia or paraple- Diagnostic Approach studies, technetium Tc-99m phos- gia, but many demonstrate subtle ob- phonate whole-body bone scintigra- jective neurologic deficit that can be Patients who present with spinal me- phy, and plain radiography as well as measured with the American Spinal tastases have either known meta- CT of the chest, abdomen, and pel- Injury Association (ASIA) impair- static disease, history of malignancy vis, Rougraff et al13 identified the ment scale or the Frankel scale9,10 without prior metastases, or no prior primary site of disease in 85% of pa- (Table 1). The median Frankel score diagnosis of cancer. In patients with tients who presented with skeletal in patients with metastatic spine le- established metastatic disease who metastatic disease of unknown ori- sions is D (ie, decreased sensory present with typical imaging find- gin. The addition of biopsy identified function and decreased but useful ings, treatment generally proceeds the diagnosis in an additional 8% of motor function).11 These early neuro- without biopsy unless histologic patients. Many patients without a logic deficits are often mistakenly at- evaluation is expected to affect treat- prior cancer diagnosis who present tributed to other causes (eg, medica- ment decisions. in this manner do not require biopsy. tion side effects), which delays The patient with a history of non- For those who do require biopsy, it

38 Journal of the American Academy of Orthopaedic Surgeons Peter S. Rose, MD, and Jacob M. Buchowski, MD, MS may be possible to obtain a specimen nificant noncontiguous metastases,15 logic signs and symptoms—that in- at a site that is safer for the patient and an unknown number may have creases with axial or rotational load- than the spine. Proper evaluation is undiagnosed intracranial lesions. ing is suggestive of mechanical needed before performing biopsy of Careful assessment of motor insufficiency and local instability. In- any lesion. strength, sensory levels, propriocep- tractable mechanical pain is a strong Image-guided biopsy has sup- tion, and reflexes is critical in guid- indication for surgery in many pa- planted open surgical biopsy as the ing initial management. These find- tients with cancer. first-line invasive tool for the evalua- ings are used later in the evaluation Assessment of overall health status tion of patients with spinal lesions. of the level of response or deteriora- and burden of disease is important in Image-guided biopsy using a large- tion during treatment. New or ab- selecting a treatment modality. A bore needle can be safely performed normal clinical findings (eg, weak- careful approach to treatment is re- in an outpatient setting with sedation ness, sensory loss, hyperreflexia, quired for persons with cachexia, de- or light anesthesia in conjunction pathologic reflexes) may reveal pre- cubitus ulcers or compromised skin with sparing use of local anesthesia. viously undiagnosed lesions at other resulting from radiation therapy, and Overall accuracy with CT-guided bi- levels. impaired pulmonary status resulting opsy of spinal lesions has been re- Careful examination also yields from lung metastases, hepatomegaly, ported to approach 89%, with better valuable information regarding clini- or malignant ascites. accuracy in lytic lesions than in scle- cally significant instability. This is The last step in the physical exami- rotic lesions (93% versus 76%, re- important because the patient with nation involves objective assessment 14 spectively). Antibiotics should be spinal instability may require surgi- of pain, neurologic status, and per- withheld before biopsy, and a tissue cal stabilization. In contrast to trau- formance status. These assessment sample should be tested for standard matic injuries, in which spinal insta- scores inform research, treatment de- and atypical cultures because of the bility generally can be assessed using cisions, and subsequent follow-up. A increased risk of osteomyelitis in radiographic studies alone, oncologic multidisciplinary approach is re- cancer patients on immunosuppres- spinal instability is assessed based on quired in the assessment of perfor- sive chemotherapy. clinical and radiographic findings.7,16 mance status. Pain can be quickly as- The radiographic criteria for trau- sessed using the visual analog scale. Physical Examination matic injury (eg, degree of focal ky- Neurologic status is quantified with phosis, fracture pattern, degree of either the ASIA or the Frankel scale A detailed examination of the spine subluxation) are not necessarily ap- (Table 1). Performance status is a and neurologic system is required, plicable to pathologic fracture; how- measure of disease burden on overall along with a global assessment of the ever, the extent and location of bony patient activities. This can be as- patient’s health status. A standard- destruction demonstrated radio- sessed with the Eastern Cooperative ized evaluation of pain, neurologic graphically contribute to the assess- Oncology Group scale or the Kar- impairment, and performance status ment of spinal stability. nofsky Performance Scale17,18 (Table 7 is needed, as well. The surgeon Taneichi et al reported that pa- 2). Many palliative chemotherapy tients with thoracic spinal lesions in- should focus first on the presence of regimens require a minimum perfor- volving 50% to 60% of the vertebral clinical deformity. Kyphosis is the mance status. Aggressive treatment body and patients with costoverte- most common. It is typically the re- of spinal metastases often improves bral joint destruction and involve- sult of mild compression fractures performance status scores to a level ment of 25% to 30% of the verte- and decreased ambulation. Coronal at which the patient qualifies for fur- bral body were at risk for impending plane deformities and fixed deformi- ther adjuvant treatment. ties are rare. The presence of defor- collapse. Patients with lumbar spinal mity influences the assessment of spi- lesions involving 35% to 40% of the nal stability and the levels of vertebral body and those with poste- Imaging instrumentation should surgical in- rior element and/or pedicle destruc- tervention be required. tion as well as involvement of 20% Standing and/or weight-bearing ra- A detailed neurologic examination to 25% of the vertebral body were diographs of the spine provide an ac- and radiologic evaluation of the en- also at risk of impending collapse. In curate assessment of spinal align- tire neural axis is required. Nearly addition to radiographic findings, ment and stability as well as 15% of patients have clinically sig- pain—or, more importantly, neuro- pathologic fracture and surgical

January 2011, Vol 19, No 1 39 Metastatic Disease in the Thoracic and Lumbar Spine: Evaluation and Management

Table 2 Figure 1 Measurement of Performance Status Scale Grade Description

ECOG 0 Fully active: able to carry on all predisease activities without restriction 1 Restricted in strenuous activity; ambulatory; able to perform light work 2 Ambulatory; able to perform self care; unable to work; bedridden ≤50% of the time 3 Limited self care; bedridden ≥50% of the time 4 Completely disabled; incapable of self care; bedridden Karnofsky 100% Normal, with no complaints or signs of disease 90% Capable of normal activity with few signs or symptoms of disease 80% Normal activity with some difficulty, some signs or symptoms of disease 70% Self care; incapable of normal activity and work 60% Requires some help but can fulﬁll most personal AP radiograph demonstrating the requirements so-called winking owl sign (arrow), 50% Requires frequent help and medical care which is indicative of pediculolysis 40% Disabled; specialized care needed resulting from metastatic carcinoma. 30% Severely disabled; hospital admission indicated; death is not imminent 20% Very ill; urgent hospital admission and treatment required contiguous primary tumors is associ- 10% Moribund with rapidly progressive fatal disease ated with some malignancies (eg, processes Pancoast, apical lung). The Batson vertebral venous plexus, a valveless ECOG = Eastern Cooperative Oncology Group network with contributions from the pelvic and intercostal veins, may per- anatomy. Serial radiographs are used CT provides the greatest bony ana- mit the spread of cells from primary to assess disease progression (eg, tomic detail. In conjunction with my- tumors and contribute to the propen- changes in alignment, progression of elography, it allows visualization of sity of metastases in the spine. osteolysis). Radiographs are not suf- neural compression. CT is a valuable A magnetic resonance image of the ficiently sensitive to detect metastatic adjunct study, particularly in patients entire spine should be obtained. disease, however. Overlying visceral unable to undergo MRI. Nearly 15% of patients have clini- structures impair interpretation, and MRI provides the greatest sensitiv- cally significant lesions at noncontig- lesions cannot reliably be detected on ity and specificity in the detection of uous sites15 (Figure 2). Although fo- lateral radiographs until ≥30% to spinal metastases (98.5% versus cused radiographic studies offer cost 50% of trabecular bone has been de- 98.9%, respectively), with an overall and time savings, the risk of missed stroyed.19 Pediculolysis (ie, winking accuracy of 98.7%.20 MRI provides pathology is substantial. Sagittal and owl sign) is highly suggestive of tu- excellent detail regarding extraosse- axial T1- and T2-weighted fat- mor (Figure 1). However, most ra- ous extent of disease, neural com- saturated or short tau inversion re- diographic findings are nonspecific pression, leptomeningeal disease, covery (STIR) sequences should be in determining malignancy. and involvement of adjacent levels. obtained. Tumors present as hypoin- Although advances in CT technol- Tumors are most commonly lo- tense to normal marrow on T1- ogy have led to increased speed and cated in the vertebral bodies. Seventy weighted images and hyperintense greater detail, CT lacks the sensitiv- percent of tumors are located in the on T2-weighted images. Adjacent ity required to detect metastatic le- thoracic spine, a number that is edema and cerebrospinal fluid can be sions in the spine. A recent study roughly equal to the volume of the seen on T2-weighted images. Tumor showed sensitivity to be only 66% thoracic vertebrae as a percentage of may not be adequately visualized on and diagnostic accuracy to be 89%.20 the total spine.7 Direct invasion from T2-weighted fast spin-echo MRI.

40 Journal of the American Academy of Orthopaedic Surgeons Peter S. Rose, MD, and Jacob M. Buchowski, MD, MS

Figure 2 Table 3 Mean Survival by Tumor Histology in Patients Treated for Metastatic Spine Disease23 Histology Survival (mo)

Thyroid 26 Breast 19 Prostate 18 Rectal 18 Renal 10 Lung 6 Unidentiﬁed 5 carcinoma

Sagittal T2-weighted magnetic resonance image demonstrating metastatic dated. However, we have found the epidural spinal cord compression in the thoracic (A) and lumbar (B) spine. revised system to be helpful in guid- The arrows indicate discontiguous sites of epidural compression. ing treatment decisions, and we be- lieve that it may be efficiently ap- STIR sequences show enhanced con- many factors, including primary tu- plied in the clinical setting. Other trast between hypointense lipid mar- mor histology, overall burden of dis- scoring systems have been devel- row and hyperintense tumor, which ease, neurologic status, and perfor- oped. However, a recent analysis reported inaccuracies in all seven pre- compensates for decreased tumor vi- mance status. Mean survival rates in operative scoring systems studied.27 sualization on fast spin-echo MRI. patients who are treated for meta- Although STIR sequences are sensi- static spine tumors that arise from tive, they provide poor anatomic de- common histologies are shown in 21 Management tail. Gadolinium enhancement is Table 3.23 Considerable variation ex- not required to define intraosseous ists between patients with the same Spinal metastases may be managed tumors, but it is helpful in the evalu- histology. Clinician assessment alone with medical therapy, radiation ther- ation of soft tissue, epidural exten- is inadequate in predicting survival apy, and/or surgical treatment. The sion, and the spinal cord. Contrast in patients with osseous metastases.24 approach must be individualized should be given only after standard Several scoring systems have been based on tumor histology, overall fit- images have been obtained. developed to guide treatment of pa- ness or performance status, tumor tients with spinal metastases. burden, and neurologic status (Fig- 25 ure 4). Rapid progression is common Prognosis In 1990, Tokuhashi et al presented a scoring algorithm to guide in patients who present with neuro- Traditional teaching has been that surgical management of spinal me- logic deficits. Thirty percent of pa- patients undergoing surgery for spine tastases. Tomita et al26 later pro- tients who present with weakness metastases should have an antici- posed a similar scoring system as progress to paraplegia within 1 28 pated life expectancy of ≥3 months part of a new surgical management week. The likelihood of regaining beyond surgery and those undergo- strategy. Tokuhashi et al23 recently neurologic function is poor when ≥ ing radiation treatment should have revised their scoring system to incor- paraplegia has been present for 24 an anticipated survival of >1 month porate performance status, number hours, particularly in patients with beyond radiotherapy.22 Although of extraspinal bone metastases and rapid progression of symptoms. more liberal criteria are often applied vertebral metastases, presence of me- in clinical settings, these time frames tastases in major organs, tumor his- Medical Therapy are chosen because patient benefit tology, and neurologic deficit (Table Chemotherapy is the mainstay of from surgical procedures is not im- 4, Figure 3, Table 5). This system treatment for persons with meta- mediate. was based on a retrospective review static disease. However, because of Patient survival is influenced by and has not been externally vali- its delayed efficacy, chemotherapy

January 2011, Vol 19, No 1 41 Metastatic Disease in the Thoracic and Lumbar Spine: Evaluation and Management

Table 4 the event of a new diagnosis of malignancy, steroid treatment should be Scoring System for the Prognosis of Metastatic Spine Tumors withheld until a biopsy is obtained. Characteristic Score The tumorolytic effect of steroids General condition (performance status) can produce a false-negative biopsy, Poor (PS 10%–40%) 0 particularly in the case of hemato- Moderate (PS 50%–70%) 1 logic malignancies. Good (PS 80%–100%) 2 No. of extraspinal bone metastases foci Radiotherapy ≥30Although recent data have sparked 1–2 1 renewed enthusiasm for aggressive 02surgical management of spinal me- No. of metastases in the vertebral body tastases, radiation therapy is admin- ≥30istered to all patients except for the 21few with exquisitely chemosensitive 12tumors and those who are undergo- Metastases to the major internal organs ing en bloc excision.16,30 The degree Unremovable 0 of radiosensitivity varies by tumor Removable 1 type. Lymphoma, myeloma, and No metastases 2 seminoma are highly radiosensitive; Primary site of the cancer breast and prostate tumors have in- Lung, osteosarcoma, stomach, bladder, 0 termediate sensitivity; and most esophagus, pancreas other solid organ malignancies are Liver, gallbladder, unidentiﬁed 1 relatively radioresistant. Spinal cord Other 2 tolerance varies depending on the Kidney, uterus 3 dosing regimen, for example, 45 to Rectum 4 50 Gy for 1.80-Gy fractions but only Thyroid, breast, prostate, carcinoid tumor 5 30 to 33 Gy for 3.0-Gy fractions. Palsy Complete (Frankel A, B) 0 The efficacy of radiotherapy for Incomplete (Frankel C, D) 1 the management of metastatic dis- None (Frankel E) 2 ease was established in a prospective trial in which 71% of patients expe- Criteria of predicted prognosis: Total score (TS) 0–8 = <6 mo, TS 9–11 = 6 mo–1 yr, rienced pain relief and 76% of pa- TS 12–15 = ≥1yr PS = Karnofsky Performance Status tients preserved or regained ambula- Adapted with permission from Tokuhashi Y, Matsuzaki B, Oda H, Oshima M, Ryu J: A tory status.15 However, these results revised scoring system for preoperative evaluation of metastatic spine tumor prognosis. Spine 2005;30:2186-2191. were highly dependent on tumor histology. Patients with favorable histologies (eg, multiple myeloma, should be used as an adjuvant rather decrease local inflammation, and it is breast carcinoma, prostate carcino- than a primary treatment in most known to have a direct tumorolytic mas) had a durable clinical response ≤ persons with symptomatic vertebral effect in some histologies (eg, multi- ( 10 to 16 months), whereas those metastases. Exceptions to this in- ple myeloma, lymphoma). Published with unfavorable histologies (eg, re- clude highly chemosensitive tumors, evidence on corticosteroid use is lim- nal cell carcinoma, hepatocellular such as lymphoma, neuroblastoma, ited and suggests that only limited carcinoma) experienced rapid failure and seminoma. short-term improvement in ambula- (≤1 to 3 months). Other authors Corticosteroids are commonly used tory status is achieved.29 Patients are have reported similar results.31 The in the management of spinal metas- often continued on steroids long- ideal radiation treatment protocol is tases. Corticosteroid treatment is be- term for pain control; however, neu- not clearly defined,32 although 30 Gy lieved to reduce vasogenic edema, rologic benefit is typically seen in the in 10 fractions is a common treat- stabilize liposomal membranes, and first 10 to 14 days of treatment. In ment algorithm.

42 Journal of the American Academy of Orthopaedic Surgeons Peter S. Rose, MD, and Jacob M. Buchowski, MD, MS

8 Figure 3 patients. However, advances in imaging, surgical technique, and seg- mental instrumentation systems en- abled the development of direct decompressive surgery with concur- rent spinal stabilization. Outcomes with direct decompressive surgery are equal or superior to those with traditional radiotherapy in properly selected patients.33 Patchell et al11 sparked renewed enthusiasm in surgery as first-line treatment of metastatic disease. In their prospective randomized multicenter trial, outcomes with decompressive surgery followed by radiotherapy were shown to be superior to those following radiotherapy alone. In this study, 50 patients were treated with surgery followed by radiotherapy, and 51 were treated with radiotherapy alone. Average survival Kaplan-Meier curves representing estimated survival following treatment for was 126 days with surgical treatment spinal metastases. TS = total score. (Adapted with permission from plus radiotherapy versus 100 days Tokuhashi Y, Matsuzaki B, Oda H, Oshima M, Ryu J: A revised scoring with radiotherapy alone (P = 0.033). system for preoperative evaluation of metastatic spine tumor prognosis. Spine 2005;30:2186-2191.) Neurologic function, which was assessed using the ASIA and Frankel scales, was maintained for an aver- Table 5 age of 566 days in the surgical group Treatment Strategy for Spinal Metastases23 compared with an average of 72 Predicted Prognosis days in the radiotherapy-only group Total Score (mo) Treatment (P = 0.001 and P = 0.0006, respectively). Continence was maintained 0–8 <6 Nonsurgical or palliative surgery for a significantly longer period in 9–11 6–12 Palliative surgery; excisional surgery is surgically treated patients compared rarely indicated for a single lesion with no metastases to the major with patients treated with radiother- internal organs apy alone (average, 156 versus 17 12–15 ≥12 Excisional surgery days, respectively; P = 0.016). Surgi- cally treated patients retained the ability to walk for a significantly longer period than did those treated Surgery riorly based, however, so at best with radiotherapy alone (median, Surgical management of spinal me- these procedures provided indirect 122 days versus 13 days, respective- tastases is considered for four pri- neural decompression. Additionally, ly; P = 0.003). Significantly more mary indications: neurologic com- they caused iatrogenic instability as a surgical patients than radiotherapy- pression; spinal instability, including result of the removal of the only only patients regained the ability to pathologic fracture; unrelenting healthy bony structures at the in- walk (62% versus 19%, respectively; pain; and in cases in which a histo- volved level or levels. Because of the P = 0.01). logic diagnosis must be established. poor results and clinical deteriora- Although this study excluded pa- Historically, posterior laminectomy tion frequently associated with these tients with highly radiosensitive tu- was used to manage spinal metasta- procedures, radiation therapy be- mors and cauda equina lesions, it ses. Most metastatic lesions are ante- came a first-line treatment for most provides excellent evidence in sup-

January 2011, Vol 19, No 1 43 Metastatic Disease in the Thoracic and Lumbar Spine: Evaluation and Management

Figure 4

Treatment algorithm for metastatic lesions of the spine. In decreasing order of importance, the indications for surgical management of metastatic spine disease include neurologic deﬁcit secondary to neural compression, spinal instability, intractable pain, impending fracture and/or instability, and progressive tumor refractory to nonsurgical treatment (ie, radiotherapy ± chemotherapy). (Adapted with permission from Walker MP, Yaszemski MJ, Kim CW, Talac R, Currier BL: Metastatic disease of the spine: Evaluation and treatment. Clin Orthop Relat Res 2003;[415S]:S165-S175.) port of early surgical intervention in who then crossed over to the surgical should not be confused with onco- patients with neurologic signs or treatment arm had inferior clinical logic staging systems, which are used symptoms of metastatic disease. outcomes. Thus, if surgery is consid- to determine the total metastatic bur- These results were recently replicated ered, it is best performed before ra- den. In the Weinstein-Boriani-Biagini in a large multicenter observational diotherapy treatment is adminis- classification, an axial view of the study.34 Notably, patients in the tered. spine in a clock-face projection is study by Patchell et al11 who were Spinal surgical staging systems used to identify local tumor extent, first treated with radiotherapy and map the extent of local disease. They areas to be resected, and fixation

44 Journal of the American Academy of Orthopaedic Surgeons Peter S. Rose, MD, and Jacob M. Buchowski, MD, MS

35 points. This system is excellent for cient instrumented stabilization. We Figure 5 planning en bloc excision of primary avoid extensive decortication and malignancies and solitary metastases; bone grafting because it weakens the however, it is less applicable to com- remaining bone and increases blood mon metastatic scenarios with multi- loss in an environment in which ple diseased levels. The staging sys- bony fusion is unlikely. tem developed by Tomita et al26 En bloc resection in carefully se- accommodates lesions at multiple lected patients has been pro- levels and is more applicable to the posed.26,37 These procedures are tech- typical patient with spinal metasta- nically demanding and are associated ses. with a high rate of morbidity. En Anterior tumor location does not bloc excision obviates the need necessarily dictate an anterior ap- for supplemental radiotherapy to proach for surgical decompression. achieve local control in the patient Resection must be carefully tailored with insensitive tumors or in whom to each patient. Pulmonary compro- no further radiotherapy treatment Intraoperative photograph mise resulting from lung metastases options exist. Additionally, en bloc demonstrating circumferential or general deconditioning often resection may result in disease-free decompression of the spinal cord precludes the use of anterior thora- status in the patient with a solitary through a posterior approach. A stabilizing rod must remain in place cotomy or a thoracoabdominal ap- metastasis. We reserve these aggres- on one side of the spinal cord at all proach. The posterolateral trans- sive procedures for patients with sol- times during decompression to pedicular approach has been shown itary metastases following a long guard against spinal subluxation. to be safe and effective in spinal cord disease-free interval. decompression and stabilization of Preoperative medical optimization diseased segments, and it avoids the and attention to preoperative nutri- morbidity (particularly pulmonary) tional status may reduce surgical relief following osteoporotic verte- associated with thoracotomy36 (Fig- complications. Perioperative and bral compression fracture. These ure 5). Although a tumor often postoperative nutritional supplemen- closely related procedures also can presses dorsally from the vertebral tation should be used in patients be used to manage spinal metastases, body to the spinal cord, the posterior who are malnourished preopera- particularly spinal plasmacytoma longitudinal ligament provides an tively (eg, low albumin level, low and multiple myeloma.40 Pain relief excellent anatomic barrier to clear prealbumin serum level). We often is the main objective of vertebro- local disease and decompress neural add nutritional shakes to a patient’s plasty and kyphoplasty in this set- elements. diet when he or she is able to tolerate ting. These procedures are not indi- Management of metastatic disease oral food intake perioperatively, and cated as a primary treatment in is fundamentally palliative. Long- we recommend either total or partial patients with neurologic dysfunction term survival is well documented in parenteral nutrition immediately caused by epidural compression of carefully selected patients who un- postoperatively, especially in the pa- the neural elements or gross instabil- dergo aggressive en bloc resection; tient who may not be able to tolerate ity.40,41 however, patients are rarely cured. oral intake within 1 to 2 days of sur- Fourney et al42 reported marked or Most patients who are treated surgi- gery. Preoperative embolization may complete pain relief in 84% of pa- cally for metastatic disease undergo be liberally employed to minimize tients following vertebroplasty or ky- intralesional tumor resection to pro- bleeding.38 Preoperative radiotherapy phoplasty and no symptomatic im- vide neurologic decompression, local and/or neurologic deficit place pa- provement in 9%. No patient was stabilization, and gross total resec- tients at higher risk of surgical com- worse off following the procedure, tion. Subsequent radiotherapy is em- plications.39 Careful surgical tech- and neither symptomatic complica- ployed to minimize the risk of local nique, attention to nutritional status, tions nor death was reported. recurrence. Fixation is selected to and the liberal use of local wound Asymptomatic cement extravasation provide immediate stability and to flaps (eg, trapezial, gluteal) can de- was noted in 9.2% of cases. avoid the morbidity associated with crease wound complications. Dudeney et al43 reported significant orthoses. Bony fusion is unlikely in Vertebroplasty and kyphoplasty improvement in functional outcome these patients; thus, we favor effi- are commonly used to achieve pain scores following kyphoplasty. No

January 2011, Vol 19, No 1 45 Metastatic Disease in the Thoracic and Lumbar Spine: Evaluation and Management major complications were reported. fracture. Rapid treatment with corti- metastatic spinal disease. Although Asymptomatic cement leakage oc- costeroids and radiotherapy can the prognosis of these patients re- curred in 4% of patients. Others yield excellent results in the manage- mains guarded at best, careful surgi- have reported similar findings.44,45 ment of exquisitely radiosensitive cal management in conjunction with Vertebroplasty following radiother- tumors (eg, lymphoma, multiple my- medical and radiation oncology care apy appears to be effective in manag- eloma). Vertebroplasty or kypho- has great potential to improve qual- ing persistent pain, with no compli- plasty can be used to supplement ity of life and prolong survival in this cations.46 these modalities. However, many pa- challenging patient population. Re- Vertebroplasty and kyphoplasty tients present with symptoms caused cent studies highlight the benefits of are contraindicated in certain cir- by neurologic compression or im- carefully considered surgical man- cumstances. Poor candidates for pending or completed pathologic agement. these procedures include patients fracture related to solid organ metas- ≥ with 75% loss of vertebral body tases that are poorly responsive to References height, ≥20% spinal canal compro- radiotherapy. These patients should mise due to epidural disease, poste- be considered for urgent surgical in- Evidence-based Medicine: Levels of rior vertebral body cortex violation, tervention. In the patient who is ex- evidence are described in the table of more than three levels requiring pected to survive for >3 months and contents. In this article, reference 11 is treatment, radiculopathy, and/or un- who is able to tolerate surgery, we a level I study. References 13, 15, corrected coagulopathy.47-50 In per- recommend surgical decompression and 29 are level II studies. Refer- sons with these contraindications, and stabilization in the setting of ences 14, 20, 21, 32, and 33 are level vertebroplasty or kyphoplasty will neurologic deficit secondary to neu- III studies. References 7, 12, 16, 19, likely result in a higher rate of com- ral compression. We also recommend 23-27, 31, 34, 36, and 38-49 are plications than is seen in patients surgery for spinal instability, espe- level IV studies. References 1-3, 5, 6, without these contraindications cially in patients with concomitant 8, 28, 35, and 50-52 are level V ex- (39% versus 11%, respectively).47 neurologic deficit secondary to neu- pert opinion. ral compression. Finally, we strongly Citation numbers printed in bold consider surgery in patients with Treatment Strategy type indicate references published progressive symptoms, impending within the past 5 years. Treatment must be individualized. fracture, or inadequate pain relief following radiotherapy and/or che- 1. Riley LH III, Frassica DA, Kostuik JP, Patients who present with asympto- Frassica FJ: Metastatic disease to the matic lesions are good candidates for motherapy. A treatment algorithm is spine: Diagnosis and treatment. Instr 51 local disease control with systemic presented in Figure 4. According to Course Lect 2000;49:471-477. 52 treatment and/or radiotherapy, as are Bilsky and Smith, patients with 2. Jacobs WB, Perrin RG: Evaluation and treatment of spinal metastases: An those who present with pain but high-grade epidural spinal cord overview. Neurosurg Focus 2001;11(6): without extended extraosseous dis- e10. compression (N) from radiore- ease and who are not at risk of im- 3. Perrin RG, Laxton AW: Metastatic spine sistant disease (O) or [who] pending fracture. Bracing may be disease: Epidemiology, pathophysiology, demonstrate mechanical insta- and evaluation of patients. Neurosurg used to alleviate pain and prevent bility (M) are offered surgery if Clin N Am 2004;15(4):365-373. fracture during the course of radio- they can tolerate it from a sys- 4. American Cancer Society: Cancer Facts therapy. Vertebroplasty and kypho- and Figures 2007. Available at: http:// temic (S) standpoint. plasty are also used to palliate local www.cancer.org/acs/groups/content/ @nho/documents/document/ symptoms. Close observation for lo- The NOMS algorithm assesses neu- caff2007pwsecuredpdf.pdf. Accessed cal progression is required. Surgery November 16, 2010. rologic, oncologic, mechanical, and may be necessary to protect against systemic indications for surgery.52 5. Byrne TN, Benzel EC, Waxman SG: fracture. These patients should be Epidural tumors, in Byrne TN, Benzel EC, Waxman SG, eds: Diseases of the followed closely for the appearance Spine and Spinal Cord. New York, NY, of new metastases elsewhere in the Summary Oxford University Press, 2000, pp 166- spine. 205. Many patients present with pain Decreased cardiovascular mortality 6. Posner JB: Spinal metastases, in Neurologic Complications of Cancer. and evidence of compressive neuro- and improved cancer therapy may Philadelphia, PA, Davis Company, 1995, logic deficit or impending pathologic result in an increased incidence of pp 111-114.

46 Journal of the American Academy of Orthopaedic Surgeons Peter S. Rose, MD, and Jacob M. Buchowski, MD, MS

7. Taneichi H, Kaneda K, Takeda N, clinical evaluation of chemotherapeutic 2246. Abumi K, Satoh S: Risk factors and agents in cancer, in MacLeod CM, ed: probability of vertebral body collapse in Evaluation of Chemotherapeutic Agents. 31. Katagiri H, Takahashi M, Inagaki J, metastases of the thoracic and lumbar New York, NY, Columbia University et al: Clinical results of nonsurgical treatment for spinal metastases. Int J spine. Spine (Phila Pa 1976) 1997;22(3): Press, 1949, p 196. Radiat Oncol Biol Phys 1998;42(5): 239-245. 19. Edelstyn GA, Gillespie PJ, Grebbell FS: 1127-1132. 8. Gilbert RW, Kim JH, Posner JB: The radiological demonstration of osseous metastases: Experimental 32. Rades D, Stalpers LJ, Veninga T, et al: Epidural spinal cord compression from Evaluation of five radiation schedules metastatic tumor: Diagnosis and observations. Clin Radiol 1967;18(2): 158-162. and prognostic factors for metastatic treatment. Ann Neurol 1978;3(1):40-51. spinal cord compression. J Clin Oncol 20. Buhmann Kirchhoff S, Becker C, Duerr 2005;23(15):3366-3375. 9. American Spinal Injury Association: HR, Reiser M, Baur-Melnyk A: Available at: http://www.asiaspinal Detection of osseous metastases of the 33. Klimo P Jr, Thompson CJ, Kestle JR, injury.org/publications/2006_Classif_ spine: Comparison of high resolution Schmidt MH: A meta-analysis of surgery worksheet.pdf. Accessed November 16, multi-detector-CT with MRI. Eur J versus conventional radiotherapy for the 2010. Radiol 2009;69(2):567-573. treatment of metastatic spinal epidural disease. Neuro Oncol 2005;7(1):64-76. 10. Frankel HL, Hancock DO, Hyslop G, 21. Mehta RC, Marks MP, Hinks RS, et al: The value of postural reduction in Glover GH, Enzmann DR: MR 34. Ibrahim A, Crockard A, Antonietti P, the initial management of closed injuries evaluation of vertebral metastases: T1- et al: Does spinal surgery improve the of the spine with paraplegia and weighted, short-inversion-time inversion quality of life for those with extradural tetraplegia: I. Paraplegia 1969;7(3):179- recovery, fast spin-echo, and inversion- (spinal) osseous metastases? An 192. recovery fast spin-echo sequences. AJNR international multicenter prospective Am J Neuroradiol 1995;16(2):281-288. observational study of 223 patients. 11. Patchell RA, Tibbs PA, Regine WF, et al: Invited submission from the Joint Section Direct decompressive surgical resection 22. White A, Kwon B, Lindskog D, Meeting on Disorders of the Spine and in the treatment of spinal cord Friedlaender GE, Grauer JN: Metastatic Peripheral Nerves, March 2007. compression caused by metastatic disease of the spine. J Am Acad Orthop J Neurosurg Spine 2008;8(3):271-278. cancer: A randomised trial. Lancet 2005; Surg 2006;14(11):587-598. 366(9486):643-648. 35. Boriani S, Weinstein JN, Biagini R: 23. Tokuhashi Y, Matsuzaki H, Oda H, Primary bone tumors of the spine: 12. Schiff D, O’Neill BP, Suman VJ: Spinal Oshima M, Ryu J: A revised scoring Terminology and surgical staging. Spine epidural metastasis as the initial system for preoperative evaluation of (Phila Pa 1976) 1997;22(9):1036-1044. manifestation of malignancy: Clinical metastatic spine tumor prognosis. Spine features and diagnostic approach. (Phila Pa 1976) 2005;30(19):2186-2191. 36. Bilsky MH, Boland P, Lis E, Raizer JJ, Neurology 1997;49(2):452-456. Healey JH: Single-stage posterolateral 24. Nathan SS, Healey JH, Mellano D, et al: transpedicle approach for spondylec- 13. Rougraff BT, Kneisl JS, Simon MA: Survival in patients operated on for tomy, epidural decompression, and Skeletal metastases of unknown origin: A pathologic fracture: Implications for end- circumferential fusion of spinal prospective study of a diagnostic of-life orthopedic care. J Clin Oncol metastases. Spine (Phila Pa 1976) 2000; strategy. J Bone Joint Surg Am 1993; 2005;23(25):6072-6082. 25(17):2240-2249. 75(9):1276-1281. 25. Tokuhashi Y, Matsuzaki H, Toriyama S, 37. Yao KC, Boriani S, Gokaslan ZL, 14. Lis E, Bilsky MH, Pisinski L, et al: Kawano H, Ohsaka S: Scoring system Sundaresan N: En bloc spondylectomy Percutaneous CT-guided biopsy of for the preoperative evaluation of for spinal metastases: A review of osseous lesion of the spine in patients metastatic spine tumor prognosis. Spine techniques. Neurosurg Focus 2003; with known or suspected malignancy. (Phila Pa 1976) 1990;15(11):1110-1113. 15(5):E6. AJNR Am J Neuroradiol 2004;25(9): 1583-1588. 26. Tomita K, Kawahara N, Kobayashi T, 38. Prabhu VC, Bilsky MH, Jambhekar K, Yoshida A, Murakami H, Akamaru T: et al: Results of preoperative 15. Maranzano E, Latini P: Effectiveness of Surgical strategy for spinal metastases. embolization for metastatic spinal radiation therapy without surgery in Spine (Phila Pa 1976) 2001;26(3):298- neoplasms. J Neurosurg 2003;98(2 metastatic spinal cord compression: Final 306. suppl):156-164. results from a prospective trial. Int J Radiat Oncol Biol Phys 1995;32(4):959- 27. Leithner A, Radl R, Gruber G, et al: 39. Wise JJ, Fischgrund JS, Herkowitz HN, 967. Predictive value of seven prognostic Montgomery D, Kurz LT: Complication, scoring systems for spinal metastases. survival rates, and risk factors of surgery 16. Wang JC, Boland P, Mitra N, et al: Eur Spine J 2008;17(11):1488-1495. for metastatic disease of the spine. Spine Single-stage posterolateral transpedicular (Phila Pa 1976) 1999;24(18):1943-1951. approach for resection of epidural 28. Yalamanchili M, Lesser GJ: Malignant metastatic spine tumors involving the spinal cord compression. Curr Treat 40. Cortet B, Cotten A, Boutry N, et al: vertebral body with circumferential Options Oncol 2003;4(6):509-516. Percutaneous vertebroplasty in patients reconstruction: Results in 140 patients. with osteolytic metastases or multiple Invited submission from the Joint Section 29. Sørensen S, Helweg-Larsen S, Mouridsen myelomas. Rev Rhum Engl Ed 1997; Meeting on Disorders of the Spine and H, Hansen HH: Effect of high-dose 64(3):177-183. Peripheral Nerves, March 2004. dexamethasone in carcinomatous J Neurosurg Spine 2004;1(3):287-298. metastatic spinal cord compression 41. Cotten A, Dewatre F, Cortet B, et al: treated with radiotherapy: A randomised Percutaneous vertebroplasty for 17. Oken MM, Creech RH, Tormey DC, trial. Eur J Cancer 1994;30A(1):22-27. osteolytic metastases and myeloma: et al: Toxicity and response criteria of Effects of the percentage of lesion filling the Eastern Cooperative Oncology 30. Weigel B, Maghsudi M, Neumann C, and the leakage of methylmethacrylate at Group. Am J Clin Oncol 1982;5(6):649- Kretschmer R, Müller FJ, Nerlich M: clinical follow up. Radiology 1996; 655. Surgical management of symptomatic 200(2):525-530. spinal metastases: Postoperative outcome 18. Karnofsky DA, Burchenal JH: The and quality of life. Spine 1999;24:2240- 42. Fourney DR, Schomer DF, Nader R,

January 2011, Vol 19, No 1 47 Metastatic Disease in the Thoracic and Lumbar Spine: Evaluation and Management

et al: Percutaneous vertebroplasty and efficacy of percutaneous vertebroplasty. 49. Patel AA, Vaccaro AR, Martyak GG, kyphoplasty for painful vertebral body AJNR Am J Neuroradiol 2007;28(3): et al: Neurologic deficit following fractures in cancer patients. J Neurosurg 570-574. percutaneous vertebral stabilization. 2003;98(1 suppl):21-30. Spine (Phila Pa 1976) 2007;32(16): 46. Cheung G, Chow E, Holden L, et al: 1728-1734. 43. Dudeney S, Lieberman IH, Reinhardt Percutaneous vertebroplasty in patients MK, Hussein M: Kyphoplasty and the with intractable pain from osteoporotic 50. Wong W, Mathis JM: Vertebroplasty and treatment of osteolytic vertebral or metastatic fractures: A prospective kyphoplasty: Techniques for avoiding compression fractures as the result of study using quality-of-life assessment. complications and pitfalls. Neurosurg multiple myeloma. J Clin Oncol 2002; Can Assoc Radiol J 2006;57(1):13-21. Focus 2005;18(3):e2. 20(9):2382-2387. 47. Hentschel SJ, Burton AW, Fourney DR, 51. Walker MP, Yaszemski MJ, Kim CW, 44. Pflugmacher R, Kandziora F, Schroeder Rhines LD, Mendel E: Percutaneous Talac R, Currier BL: Metastatic disease RJ, Melcher I, Haas NP, Klostermann vertebroplasty and kyphoplasty of the spine: Evaluation and treatment. CK: Percutaneous balloon kyphoplasty performed at a cancer center: Refuting Clin Orthop Relat Res 2003;(415 suppl): in the treatment of pathological vertebral proposed contraindications. J Neurosurg S165-S175. body fracture and deformity in multiple Spine 2005;2(4):436-440. myeloma: A one-year follow-up. Acta 52. Bilsky M, Smith M: Surgical approach to Radiol 2006;47(4):369-376. 48. Masala S, Lunardi P, Fiori R, et al: epidural spinal cord compression. Vertebroplasty and kyphoplasty in the Hematol Oncol Clin North Am 2006; 45. Calmels V, Vallée JN, Rose M, Chiras J: treatment of malignant vertebral 20(6):1307-1317. Osteoblastic and mixed spinal fractures. J Chemother 2004;16(suppl metastases: Evaluation of the analgesic 5):30-33.

48 Journal of the American Academy of Orthopaedic Surgeons