|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Intractable back pain is a frequent symptom of disseminated cancer, and in patients who have spinal metastases the pain is related to spinal instability13. The spinal instability is mechanical and is related to extensive bone destruction due either to tumor-mediated osteolysis or to iatrogenic causes, and the pain is due to vertebral fracture. Cancer of the breast is the most common primary tumor associated with metastases to the spine (Table II)17. However, any tumor can metastasize to the spine and produce sufficient destruction to render the spine unstable, leading to excruciating pain that is induced by mechanical instability and to spinal-cord or nerve-root compression resulting in paresis. When the pain is excruciating, the patient is comfortable only when lying absolutely still. Any movement, including log-rolling (even by two or three trained nurses), is associated with agonizing pain, and the patient may not be able to sit, stand, or walk because of the pain, even with use of a spinal orthosis. When the spine is minimally involved, the patient may be relatively free of pain when wearing a rigid spinal orthosis, but any movement of the back (for example, turning in bed, sitting, or standing) may be impossible without the orthosis.
Plain radiographs show destruction of the vertebral body with minimal-to-complete vertebral collapse. It is unusual to see a discrete fracture. Nevertheless, the bone destruction in the vertebral body should be considered equivalent to a pathological fracture of an appendicular bone because the pain is due to a pathological fracture that creates mechanical instability and not to the metastases22. Neither radiation therapy nor chemotherapy, even if successful in controlling the tumor, will alleviate the pain. As in the treatment of pathological fractures of the long bones, stabilization of the vertebral segments is required for pain relief. The spine can be stabilized with either an anterior or a posterior approach5,7-9,12,15,20-22,31,32. Vascular metastatic lesions necessitate preoperative embolization, particularly if anterior stabilization is being considered18. This is particularly relevant with regard to metastases from renal carcinoma; however, other tumors also may be very hyperemic. Compression of the spinal cord or the cauda equina has been commonly treated with laminectomy and decompression. The tumor usually involves the anterior elements, and laminectomy may destabilize a spine that has been weakened as a result of tumor-induced bone destruction. Studies have shown that radiation therapy combined with steroids is at least as effective as laminectomy6,11,19. Because of the risk of subsequent instability, it is recommended that the spine be stabilized at the time of decompression14,23,24,33. The decompression can be carried out anteriorly or posteriorly, but it is essential that posterior decompression be combined with stabilization of the spinal segments to avoid the development of progressive kyphosis, which may induce additional compression of the spinal cord and neurological impairment. Laminectomy alone is contraindicated because it results in spinal instability. Removal and replacement of the lamina has been described10. Operative treatment should be combined with a short course of high-dose steroids to minimize any postoperative edema and by a course of postoperative radiation therapy. Initially, the Harrington rod was the only instrumentation available for spinal stabilization. It soon became apparent that the fixation provided by this device was inadequate despite the hooks being embedded in methylmethacrylate and that, if a posterior approach was to be used, some form of segmental fixation was also necessary. Several types of instrumentation were subsequently developed, all of which rely on fixation at multiple levels that include at least three vertebrae cephalad and three vertebrae caudad to the level of involvement. The Banks-Dervin rod, a one-centimeter square rod with multiple screw-holes, is contoured to the shape of the spine and is placed in the gutter between the spinous process and the lamina, often with use of methylmethacrylate as a grout; it then is fixed with screws that pass beneath the spinous process dorsal to the neural elements and into the contralateral lamina. Double Luque rods and the Hartshill rectangle are attached to the vertebrae with sublaminar wires. When the lumbar spine is stabilized the instrumentation is contoured to provide lumbar lordosis, and when the thoracic spine is stabilized the instrumentation is contoured to provide dorsal kyphosis. If posterior fixation is inadequate for stabilization of the spine in patients with metastatic lesions in the fourth and fifth lumbar vertebrae, an anterior stabilization procedure also is performed. Newer methods of instrumentation whereby the rods can be fixed with interpedicular screws, hooks, or wires have recently become available. We used one such instrumentation system, the Isola device (DePuy AcroMed, United Kingdom), in our patients. Patients with localized metastatic lesions have been treated with anterior stabilization with use of the Kaneda device, which initially was supplemented with methylmethacrylate and more recently has been augmented with a vertebral replacement cage.
In a series of eighty patients, we found an overall referral rate of approximately four patients per year. This is a much lower rate than would be expected from the prevalence of metastatic disease to the spine reported by one of us (C. S. B. G.) and Sylvester13. Thus, a study was undertaken to determine whether patients who had a metastatic lesion in the spine were being appropriately diagnosed and referred for treatment. Because cancer of the breast accounted for approximately half of the patients who had spinal instability secondary to metastases, we decided to review the records of all patients who had been treated for breast cancer in a single region in the United Kingdom. We examined the records of patients in one region rather than those of patients who had been treated in a particular hospital because referral patterns may differ from one hospital to another. Since it takes some time for skeletal metastases to become manifest, it also was decided that patients would be evaluated seven years after the diagnosis of breast cancer had been confirmed. As it was assumed that many patients had died from the breast cancer during this seven-year period, a retrospective analysis was conducted of the medical records. The vast majority of patients had been referred to the regional cancer hospital for radiation therapy after primary treatment had been obtained at their local hospital. Subsequent follow-up often had been performed at the regional cancer hospital and the local district general hospital, and the records from both hospitals were reviewed and matched for each patient. A total of 1412 women were diagnosed with and treated for breast cancer in thirty-one hospitals in the survey area. Of these patients, 963 had complete records, which were used for analysis. The remaining notes either had been destroyed or were incomplete. Analysis of the records showed that, at seven years, 499 (52 percent) of the 963 women had died from breast cancer. Overall, 420 (44 percent) of the 963 patients had complained of bone pain at some time during the seven years after the initial diagnosis of breast cancer. Of these 420 patients, 207 (21 percent of the 963) had proven skeletal metastases whereas 213 (22 percent) had a benign etiology (such as arthritis) for the bone pain. Of the 207 patients who had skeletal metastases, 89 percent died compared with 36 percent of the 213 patients who had nonmetastatic bone pain. Only forty-five (22 percent) of the 207 patients who had painful skeletal metastases were referred to an orthopaedic surgeon for evaluation and treatment. A large number of patients had complications as a consequence of structural lesions. Of the 207 patients who had bone metastases, eighty-eight had one to five complications and several had more than one complication (Table I). Many patients were treated nonoperatively, including those who were considered to have terminal breast cancer. The difference in the referral patterns between patients who had sustained a pathological fracture of the femur and those who had spinal instability was dramatic. All twenty-two patients who had a pathological femoral fracture were referred for an orthopaedic opinion. Nineteen were treated operatively, and three who were thought to have end-stage cancer were treated nonoperatively. Only six of fifty-one patients who had spinal instability were referred for an orthopaedic opinion. None of these patients had operative treatment.
The results of this study suggest that spinal instability due to metastases develops within seven years in 5 percent of patients who have breast cancer and that patients who have spinal instability secondary to metastatic disease are not being referred for treatment. Similar results were reported by O'Donoghue et al.29,30, who evaluated 269 consecutive patients with breast cancer who had been seen in an internationally known breast unit. Those authors found that patients who had a pathological fracture did better after operative intervention combined with adjunctive therapy than did those who had nonoperative treatment alone. Forty-seven women had a total of eighty-two instances of structural bone destruction, fifty-two (63 percent) of which involved the spine and eleven of which were associated with cord compression. The average duration of symptom-free survival was thirty-nine weeks in the group that had been treated operatively compared with eleven weeks in the group that had been treated nonoperatively, and the average duration of overall survival was twenty-two and nine months, respectively (p < 0.05 for both comparisons). O'Donoghue et al. stated that clinical review by an orthopaedic surgeon would have been appropriate in 89 percent of the instances but was sought in only 46 percent and that operative intervention was feasible in 65 percent but was performed in only 31 percent.
The failure to refer patients who have spinal instability secondary to metastatic disease of the spine suggests that the education of breast surgeons, oncologists, radiation therapists, and others involved in the treatment of patients who have advanced cancer has been incomplete. In order to determine the reasons for the lack of dissemination of information on the treatment of metastatic disease of the spine, we reviewed textbooks on oncology, breast cancer, general surgery, and orthopaedic surgery found in our university library, the library in the regional cancer hospital, and the library of our teaching hospital. Although most of the textbooks mentioned skeletal metastases, only two of nine oncological textbooks, one of twelve textbooks on breast cancer, one of eleven textbooks on general surgery, and, disappointingly, only five of eight textbooks on orthopaedic surgery even mentioned spinal instability secondary to metastatic disease of the spine. In 1995, the European Journal of Surgical Oncology published a supplement on the treatment of symptomatic breast disease4. The supplement was thirteen pages long, with a double-column format; however, in the objectives that were cited there was no mention of bone pain. Fifty-four centimeters of text was devoted to the management of metastatic disease. The only reference to bone metastases was included in the phrase "palliative measures, such as radiotherapy for bone metastases, bone stabilisation by orthopaedic surgery, pleurodesis, etc." Similar guidelines developed by the European Society of Surgical Oncology1, based on the guidelines developed by the British Association of Surgical Oncology and the Danish Breast Cancer Cooperation Group, contained no mention of skeletal metastases or associated complications. Recently, the British Association of Surgical Oncology updated its guidelines, which now include the phrase "radiotherapy for bone metastases, and bone stabilisation by orthopaedic surgery (see the forthcoming BASO Guidelines for the Management of Bony Metastases)."2 It was therefore extremely disappointing to note that the Association's guidelines on the management of metastatic bone disease in the United Kingdom, which are twenty-one pages long and include three pages devoted to the use of bisphosphonates, do not contain a subsection on metastases to the spine3. Spinal instability was mentioned as a complication of spinal metastatic disease, and patients most likely to benefit from operative intervention were described as those in whom "pain [was] exacerbated by movement and relieved on rest (spinal instability)." The guidelines also suggested that patients who have pain and destruction of more than 50 percent of the vertebral body should have an immediate neurological examination and, in the absence of neurological signs, should be evaluated by an orthopaedic surgeon prior to being considered for radiation therapy3. The British Medical Journal, in a review article on myeloma34, published a radiograph that was suggestive of spinal instability16,26. The caption to the radiograph was: "Bone pain from mechanical effects of myeloma damage (as in spine shown here) often necessitates long-term treatment with strong analgesia despite response to chemotherapy." The reader might ask why the editor accepted this article, which made no mention of operative stabilization of the spine despite the publication, for more than fifteen years, of reports on the need for stabilization and its efficacy5,7,8,12,14,15,17,20,22,25,31,32. In ABC of Breast Diseases, neither spinal instability nor spinal stabilization for metastatic disease of the spine was mentioned27,28. However, it was stated that "cord compression is not usually amenable to surgery" and that "patients with isolated metastases causing cord compression who are fit can be treated by emergency laminectomy," even though laminectomy should never be carried out in a patient who has metastatic disease of the spine unless the procedure is combined with stabilization. Laminectomy alone further destabilizes a spine that is already compromised by metastatic bone destruction. Finally, we reviewed the practices of consultant neurosurgeons, consultant orthopaedic surgeons who did not have a particular interest in spinal operations, and consultant orthopaedic surgeons who did have such an interest. These practitioners were asked whether they would treat cord compression with laminectomy alone and whether they would simultaneously stabilize the spine in a patient who had skeletal metastases and spinal instability. Fifty-nine (78 percent) of seventy-six neurosurgeons indicated that they would consider a laminectomy, although most said that they would perform a decompression with stabilization. Several neurosurgeons indicated that they would perform only a laminectomy in a patient who was unfit for a more major procedure such as stabilization. Eleven (15 percent) of seventy-five orthopaedic spinal surgeons said that they would consider only a laminectomy. Fifty-three (79 percent) of sixty-seven spinal surgeons said that they would stabilize the spine in the presence of spinal instability; three (4 percent), that they would refer the patient to a neurosurgeon; and eleven (14 percent), that they would not consider operative intervention but would refer the patient to an oncologist or a radiation therapist. Thirteen (6 percent) of 208 general orthopaedic surgeons indicated that they would perform a laminectomy for the treatment of cord compression, and fifty-four (26 percent) said that they would refer a patient who had spinal instability to an oncologist or a radiation therapist and not to a spinal surgeon or a neurosurgeon. In conclusion, studies show that spinal instability is a common complication of skeletal metastases, occurring in at least 5 percent of patients who have breast cancer. Breast cancer accounts for approximately half of patients who have spinal instability secondary to metastatic cancer, with multiple myeloma being the next most common cause. Spinal instability is associated with considerable morbidity. The symptoms can be relieved by decompression and spinal stabilization, and many patients remain pain-free for several years after operative intervention. The combination of operative intervention with radiation therapy (if feasible) and endocrine therapy or chemotherapy, or both, not only gives patients with these lesions a better quality of life but also is associated with a longer life expectancy than is associated with nonoperative treatment alone. However, our review suggests that only a very small number of patients whose quality of life could be improved by appropriate operative treatment are being referred for such treatment.
Principles of Evaluation and Treatment We reviewed our treatment of patients who had spinal instability secondary to metastatic disease of the spine. All of the patients required preoperative assessment to determine their fitness for operative intervention. This assessment included skeletal scintigraphy to evaluate the extent of skeletal dissemination of the disease. Preoperative magnetic resonance imaging was indicated to demonstrate not only the extent of skeletal involvement throughout the spine but also the presence of any extradural tumor. Whenever possible, the patients received postoperative irradiation. However, some patients had received the maximum tolerable dose preoperatively. The spinal cord is sensitive to radiation, and any additional radiation therapy increases the risk of transverse myelitis. The underlying tumor was treated with endocrine therapy or chemotherapy, depending on the type of primary tumor. If there are other small skeletal metastases, bisphosphonates may be indicated, particularly if the primary tumor is breast carcinoma or if the patient has multiple myeloma; however, bisphosphonates are not indicated for large lytic metastatic lesions.
Clinical Data Thirty-four of the eighty patients had clinical evidence of compression of the spinal cord or the cauda equina and weakness of the lower extremities severe enough to affect walking or standing; the symptoms sometimes were associated with loss of bladder function. These patients were treated with decompression at the time of stabilization. Twenty-three (68 percent) of the thirty-four patients recovered neurological function to the extent that they could walk without orthoses and bladder function, if it had been compromised, returned to normal. It must be emphasized, however, that these patients presented primarily with pain due to spinal instability. In two patients, cord compression recurred at ten and fourteen months after decompression. All eighty patients had decompression and spinal stabilization with use of a variety of spinal instrumentation systems (Table III)17. The most frequently used device was the Hartshill rectangle (thirty-one patients), followed by the Banks-Dervin rod (twenty-six patients).
Results Seventy-one (89 percent) of the eighty patients had complete relief of pain, and four patients (5 percent) had partial relief (Table IV)17.
Five patients had failure of the treatment. One of these patients died of septicemia after a wound infection. Another patient had an infection, requiring removal of the implant. Two patients, in whom we had used Harrington rods and hooks, had loosening of the implant. In the fifth patient, who had complete relief of pain, paraplegia developed as a result of an extradural bleed despite removal of the clot. He had previously received anticoagulants for the treatment of an intercurrent illness. Preoperative radiation therapy predisposed to infection. The first patient in our series had multiple myeloma and had been treated with chemotherapy and radiation therapy. Nevertheless, he was unable to sit up or walk and was confined to bed because of excruciating back pain. Even log-rolling by two trained nurses was painful. At the time that he was evaluated, the only device that was available to stabilize the spine was the Harrington rod. The thoracic spine was stabilized with two Harrington rods, and the hooks were cemented in situ with methylmethacrylate. Even though one hook loosened, the patient had total relief of pain to the extent that he could walk and could return home. Eight months later, he was readmitted to the hospital with clinical signs of cord compression. The instrumentation had to be removed at the time of decompression of the spinal cord. It is now a routine practice to evaluate the spinal canal and, if there is any tumor within the canal, the spinal cord or the cauda equina, depending on the level of involvement, is decompressed at the site of involvement at the time of stabilization. In the initial years of our series, the spinal canal was evaluated with myelography; later, it was evaluated with myelography combined with computerized tomography; and currently, magnetic resonance imaging is used. The decompression is carried out prior to stabilization; if the tumor is localized to one vertebral level and it is mainly anterior, anterior decompression and stabilization is performed, and if the tumor extends over several vertebral levels, posterior decompression and stabilization is carried out. Once the decompression is performed, the spine is stabilized irrespective of whether the patient has neurological signs or symptoms or whether there is tumor within the canal. Anterior decompression and stabilization is now more commonly carried out.
Complications The instrumentation failed in two patients. In one patient who had myeloma (the only patient in whom two Luque rods were inserted), one rod broke and the patient required anterior bone-grafting. Twelve years after the procedure, the patient was alive and active, without a recurrence of compression of the cauda equina or pain due to spinal instability. In another patient, the instrumentation (a Hartshill rectangle) fractured after five years of complete relief from pain. Several patients lived for many years after the spinal stabilization. One patient who had multiple metastases to the spine from breast cancer responded well to tamoxifen, which was given after the performance of spinal decompression and stabilization for the treatment of spinal instability associated with compression of the cauda equina. Five years after the procedure, she was in a motor-vehicle accident and sustained an undisplaced fracture of the odontoid process, which healed with nonoperative treatment. A year later, she had additional spread of the cancer and did not respond to chemotherapy. She died six and a half years after the spinal stabilization, without any recurrence of the back pain or neural compression.
Results similar to those in our series17 have been reported by other authors. Onimus et al.31 reported the results of operative intervention in 100 patients who had spinal metastases. Anterior stabilization was performed in fifty-eight patients; posterior stabilization, in thirty-three; and combined anterior and posterior stabilization, in nine. Preoperatively, ninety-six patients had vertebral pain, and forty-three had radicular pain. Patients who had metastases from lung cancer had intractable pain. Walking was impossible for fifty patients, either due to neural compression or to pain caused by the spinal instability. Thirty-five (70 percent) of the fifty patients were able to walk after decompression and stabilization. Of thirty-eight patients with a neurological deficit preoperatively, thirty had a decrease of the deficit postoperatively. Sixty-two patients stopped using analgesics after the operation. Overall, there was clinical improvement in eighty of the 100 patients. Our review shows that only a very small percentage of patients who have spinal instability secondary to metastatic disease of the spine are being referred to the appropriate specialists in a timely manner. We believe that oncological education in terms of the management of patients who have spinal instability secondary to metastatic cancer has failed, probably for many reasons: 1. Breast surgeons apparently do not have a specific awareness of how to manage bone pain in their patients. 2. Oncologists and radiation therapists are not kept up to date on the role of modern orthopaedic surgery and what it can offer their patients. 3. The editors of broad-spectrum journals such as the British Medical Journal fail to understand the subject of spinal instability, make poor selections with regard to authors and reviewers, and often refuse to publish articles aimed at the general medical profession because they consider the topic of spinal instability to be too specialized. 4. Courses on breast cancer do not cover the orthopaedic management of patients with this disease. 5. The vast majority of textbooks on breast cancer, oncology, and general surgery do not mention spinal instability secondary to metastatic disease of the spine, and even textbooks on orthopaedic surgery fail to include this complication. 6. Specialized journals, such as The Journal of Bone and Joint Surgery, are not read widely by physicians in other disciplines, who therefore remain unaware of recent advances in the treatment of spinal instability. 7. Some neurosurgeons still perform a laminectomy without stabilization in patients who have metastatic disease of the spine. 8. Some orthopaedic surgeons perform similar inappropriate procedures, and others fail to refer patients for decompression and stabilization. As orthopaedic surgeons, we have failed to educate our colleagues in other specialties about the principles of the management of spinal instability secondary to metastatic disease of the spine. What is the future for the management of metastatic disease of the spine? We believe that the only solution is for large orthopaedic departments to establish bone-pain clinics for patients who have metastatic cancer so that they can be evaluated adequately. Then, if a patient required operative intervention for a complication secondary to skeletal metastases, it would be carried out by the appropriate orthopaedic surgeon in that department. Treatment based on the site of involvement would include spinal stabilization with or without decompression by the spinal surgeon, proximal femoral replacement with or without acetabular reconstruction by the hip surgeon, fixation of pathological fractures by the trauma surgeon, and so on. The development of such clinics would do much to improve the quality of life of patients who have skeletal metastases and should also raise the profile of orthopaedic surgery among all doctors involved in the treatment of patients who have malignant disease. Note: The authors are grateful to the Centre for Cancer Epidemiology, Withington, Manchester, for help with the patient database.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||