Evidence Tables

Tokuhashi, Y., Matsuzaki, H., Toriyama, S., Kawano, H. & Ohsaka, S. (1990). Scoring system for the preoperative evaluation of metastatic spine tumor prognosis. Spine, 15 (11), 1110–1112.
Design: Test evaluation/epidemiological study, evidence level 3
Country: Japan; Setting: Department of Orthopaedic Surgery
Population N=64 patients with spinal metastasis
Test items
General condition (performance status), number of extraspinal bone metastases foci, number of metastases in the vertebral body, metastases to the major internal organs, primary site of the cancer, spinal cord palsy
Scoring
0, 1 or 2 scores for 6 areas (primary cite of cancer: lung, stomach = 0; kidney, liver, uterus = 1; others, unidentified, thyroid, prostate, breast, rectum = 2), maximum total score of 12
Life expectancy prognosis/Treatment consequences
0–5 scores: palliative operation (securing support),
9–12: excisional surgery (securing support and prolonging life)
Received treatment 53 cases treated palliatively, 11 excisionally treated (the score did not decide upon treatment)
Follow up at least 12 months
Results
OUTCOMERESULT
SurvivalScores 0–5: average survival: 3 months or less
Scores 6–8: average survival: 12 months or less
Scores 9–12: average survival: 12 months or more
Correlation total score and survivalr = 0.65 (p<0.01)
Author’s conclusion. An excisional operation should be performed on those cased who scored more than 9 points, while a palliative operation is indicated for patients scoring less than 5 points.
General comments Test score and received treatment may be intercorrelated, and treatment and outcome (survival) may not be independent.
Kluger, P., Korge, A. & Scharf, H.-P. (1997). Strategy for the treatment of patients with spinal neoplasms. Spinal Cord, 35, 429–436.
Design: Algorithm evaluation study, evidence level 3
Country: Germany; Setting: Orthopaedic Clinic
Population N=154 patients with tumourous osteolyses of the spine
Test items
Narrowed spinal canal, unstable vs stable free spinal canal, survival prognosis from enhanced diagnostic results (MRI, bone scan, myelography)
Scoring
algorithm
Life expectancy prognosis/Treatment consequences
Group 1: < 6 months: no further operation; temporary aids, nursing at home
Group 2: 6–24 months: vertebrectomy, alloplastic vertebral body replacement; temporary aids, nursing at home
Group 3: > 24 months: vertebrectomy, autogeneic vertebral body replacement; individual aids, rehabilitation in hospital
Received treatment according to algorithm; exclusive posterior approach (n=104), secondary anterior procedure with alloplastic replacement of the vertebral body (n=20), additional anterior approach with autogeneic vertebral body replacement (n=12), posterior transpedicular biopsy alone (n=18)
Follow up at 30 days; up to 80 months
Results
OUTCOMERESULT
SurvivalGroup 1: mean survival 13.9 months, range: 0.5–56
Group 2: mean survival 18.2, range: 0.25–48
Group 3: mean survival 41.7 months, range: 11.8–80.4
Author’s conclusion. The treatment of patients with tumourous osteolyses of the thoracic and lumbar spine treated according to the proposed algorithm underline the unequivocal advantages of initially posterior procedures..
General comments The received treatment depended on the preliminary diagnosis, the treatment and outcome (survival) may not be independent.
Enkaoua, E. A., Doursounian, L., Chatellier, G., Mabesoone, F., Aimard, T. & Saillant, G. (1997). Vertebral metastases: A critical appreciation fo the preoperative prognostic Tokuhashi Score in a series of 71 cases. Spine, 22(19), 2293–2298.
Design: Test evaluation study, evidence level 3
Country: France; Setting: Department of Orthopaedic Surgery and Traumatology
Population N=85 patients with vertebral metastases
Test items
Tokuhashi (1990)
Scoring
Tokuhashi (1990), median as cut-off: scores 0–7 vs 8–12
Life expectancy prognosis/Treatment consequences
Tokuhashi (1990)
Received treatment patients with single metastases received excisional surgery, multiple metastases were treated palliatively to restore stability, to reverse neurological compromise and/or pain relief
Follow up up to 62 months for single patients
Results
OUTCOMERESULT
SurvivalScores 0–7: median time to death: 5.3 (+ − 1.2) months
Scores 8–12: median time to death: 23.6 (+ − 5.8) months
Survival curveComparison of length of survival and preoperative score of both groups is significant (p=0.0063)
Multivariate analysisThe modified Tokuhashi score was independently associated with a poorer survival rate
OtherThe survival of patients with tumours of unknown primary sites was 2 months, the survival of those with renal tumours 8.6 months
Author’s conclusion. The Tokuhashi score is a successful prognostic tool; the rating of unknown primary tumours should be reduced to 0.
General comments Test score and received treatment may be correlated, and treatment and outcome (survival) may not be independent. The authors state that Tokuhashi (1990) give 1 point to renal cancers and unknown primary tumour, but the original states 2 points.
Day, G.A., McPhee, I.B., Swanson, C., Tomlinson, F.H., McCombe, P. & Coyne, T. (1998). Outcomes following treatment of metastatic spine tumors. Bulletin – Hospital for Joint Diseases, 57, 11–15.
Design: Case series (scoring system applied to all), evidence level 3
Country: Australia; Setting: Spine Unit
Population N=57 consecutive patients with metastatic spine tumours
Test items
Modified Tokuhashi (1990): Involvement of the major organs, Disseminated skeletal metastases, Severe cord compromise, Contiguous vertebral body disease more than 2 bodies
Scoring
1 point for each prognostic factor
Life expectancy prognosis/Treatment consequences
Scores 2–4: Radiotherapy/chemotherapy
Scores 0–2: Surgery
Received treatment as outlined above; 29 received surgery
Follow up 60 weeks
Results
OUTCOMERESULT
SurvivalSurgery: mean survival: 30 weeks
No surgery: mean survival: 16 weeks
Survival curveThe difference between the groups was not statistically significant
Author’s conclusion Patients with severe cord compromise at presentation are instructed that surgery will not significantly improve outcomes on average; operating on patients who have a low serum albumin and lymphocyte count as well as preoperative radiotherapy and steroids is not indicated due to the substantial complication rate.
General comments Test score and received treatment are interdependent, and treatment and outcome (survival) may not be independent. The authors present also a detailed algorithm that starts with the diagnosis.
Bűnger, C. Laursen, M., Hansen, E.S., Neumann, P., Bjarke, F., Høy, K. & Helmig, P. (1999). A new algorithm for the surgical treatment of spinal metastases. Current Opinion in Orthopedics, 10, 101–105.
Design: Test validation/epidemiological study, evidence level 3
Country: Denmark; Setting: Department of Orthopaedic Surgery
Population N=unclear patients with spinal metastasis
Test items
Tokuhashi (1990);
Tomita (1984)
Scoring
Tokuhashi (1990);
Tomita (1984);
Combination of Tokuhashi (1990) and Tomita (1984)
Life expectancy prognosis/Treatment consequences
Combination of Tokuhashi and Tomita system; treatment modification: patients with Tokuhashi 0–4 (Tomita 1–7): are treated with laminectomy;
Tokuhashi 5–8 (Tomita 1–7): posterior decompression, stabilisation and reconstruction; Tokuhashi 9–12 (Tomita 1–3): En bloc resection with vertebrectomy and 360° reconstruction
Tomita 4–6: Intralesional vertebrectomy and 360° reconstruction
Tomita 7: Posterior decompression and stabilisation
all groups: treated with radiation therapy
Received treatment as suggested by score
Follow up at least 644 days
Results
OUTCOMERESULT
SurvivalTokuhashi scores 0–4: mean survival rate 99 days;
Tokuhashi scores 6–8: mean survival rate 175 days
Tokuhashi scores 9–12: 644 days
Patients who died within 3 months had a mean score of 5.2;
patients dying after 3 to 6 months 6.8;
patients dying after 6 months had a mean score of 8.2
Author’s conclusion The use of the Tokuhashi score facilitates good clinical practice. Tokuhashi system is a reliable tool for estimating life expectancy. For patients with long life expectancy the Tokuhashi (1990) and the Tomita (1984) systems should be combined.
General comments Reports 1998 conference data. Test score and treatment, and treatment and outcome (survival) may not be independent. The study does not report data to validate Tomita (1984) or their modification.
Tomita, K., Kawahara, N., Kobayashi, T., Yoshida, A., Murakami, H. & Akamaru, T. (2001). Surgical strategy for spinal metastases. Spine, 26 (3), 298–306.
Design: Test evaluation/prognostic study, evidence level 3
Country: Japan; Setting: Department of Orthopaedic Surgery
Population N=67 patients with spinal metastasis
Test items
Primary tumour (slow, moderate, rapid growth), visceral metastases (no, treatable, untreatable), bone metastases (solitary/isolated, multiple)
Scoring
Derived from hazard ratios; Primary tumour: 1, 2, 4; Visceral metastases: 0, 2, 4; Bone metastases: 1, 2; total score range: 2–10
Life expectancy prognosis/Treatment consequences
2–3: long-term local control; wide or marginal excision;
4–5: middle-term local control; marginal or intralesional excision
6–7: short-term palliation; palliative surgery
8–10: terminal care; supportive care
Received treatment not reported
Follow up maximum 5 years
Results
OUTCOMERESULT
SurvivalScores 2–3: average survival: 49.9 months, range: 18–84
Scores 4–5: average survival: 23.5 months, range: 7–57
Scores 6–7: average survival: 15 months, range: 5–33
Scores 8–10: average survival: 5.9 months, range: 1–14
Correlation total score and survival−0.69 (p<0.0001)
OtherThe correlations of the 3 prognostic factors and survival time were for malignancy of primary organs: −0.492 (p<0.0001), for visceral metastases to vital organs: −0.536 (p<0.0001) and −0.250 (p<0.05) for bone metastases
Author’s conclusion The surgical strategy system provides appropriate guidelines for treatment in all patients with spinal metastases.
General comments Test score and treatment, and treatment and outcome (survival) may be interrelated. Further data in the publication show that the treatment consequences were not strictly adhered to (the variation of scores was higher than suggested in the system).
Tokuhashi, Y. (2002). Surgery for metastatic spine tumor. Jpn J Neurosurg (Tokyo), 11, 446–452.
Design: Test evaluation/epidemiological study, evidence level 3
Country: Japan; Setting: Department of Orthopaedic Surgery
Population N=117 Patients with metastatic spine tumour
Test items general condition (performance status), number of extraspinal bone metastases foci, number of metastases in the vertebral body, metastases to the major internal organs, primary site of the cancer, severity of the spinal cord palsy
Scoring 0, 1 or 2 scores for 5 areas, primary site of the cancer is scored 0 to 5 (0: lung, stomach, esophagus, bladder, osteosarcoma, pancreas, 1: gallbladder, unidentified, 2: others, 3: kidney, uterus, 4: rectum, 5: thyroid, prostate, breast, carcinoid tumour); maximum total score 15
Life expectancy prognosis/Treatment consequences
0–8 score: predicted prognosis up to 6 months
9–11: 6 months or more
12–15: 1 year or more
Received treatment 110 patients treated with posterior decompression and stabilisation as palliative procedure for thoraco-lumbar metastases, 7 treated with en bloc spondylectomy as excisional procedure
Follow up 10 years, 6 months
Results
OUTCOMERESULT
SurvivalSurvival periods ranged from 10 days to 10 years, 6 months.
Patients with scores 0–5: 96% survived up to 6 months; 4% survived between 6 months and 1 year.
6–8: 61% survived up to 6 months, 27% survived up to 1 year, 12.2% survived 1 year or more.
9–11: 29% patients survived only up to 6 months, 33% survived between 6 months and 1 year and 38% survived 1 year or more.
12–15: 1 patient survived only up to 6 months, 16 patients (94%) survived 1 year or more.
Consistency predicted survival and survivalThe consistency was 78.7%.
Score 0–8: 79.5%, score 9–11: 71.1%, 12–15: 94.1%
Correlation total score and survivalr = 0.55 (p<0.0001)
Author’s conclusion Proper operative indications should be applied and surgical procedures have to be selected in consideration of the patients’ goals and within the limits of their life expectancy.
General comments Japanese language publication, only abstract, tables and figures were used. It is possible that there is overlap in patients described in Tokuhashi et al. (1990). Test score and received treatment may be correlated, and treatment and outcome (survival) may not be independent
Bartanusz, V. & Porchet, F. (2003). Current strategies in the management of spinal metastatic disease. Swiss Surgery, 9, 55–62.
Design: Non-systematic review; evidence level 4
Country: Switzerland; Setting: Department of Neurosurgery (author affiliation)
Population N=- Patients with spinal metastatic disease
Test items
  1. Has the spinal metastasis already been irradiated?;
  2. Is the spinal segment stable;
  3. What is the motor performance status?;
  4. Is the primary tumour known?
Scoring
Decision tree;
2 categories: patients with uncontrolled neoplastic disease presenting visceral dissemination and/or multiples spinal metastases vs patients with unknown tumour or controlled primary disease presenting with a solitary spinal metastasis; ultimately this should differentiate patients with a life expectancy of more vs less than 3 months
Life expectancy prognosis/Treatment consequences -
Received treatment -
Follow up -
Results
OUTCOMERESULT
No data presented
Author’s conclusion It is important to clearly identify those patients who can benefit from surgery.
General comments The publication does not cite the Tokuhashi or Tomita or scoring systems in general.
Clar, H. E. (2004). Operatives oder konservatives Behandlungskonzept bei spinalen Metastasen. Klinikarzt, 33(5), 149–152.
Design: Non-systematic review, evidence level 4
Country: Germany; Setting: Neurosurgery Clinic (author affiliation)
Population N=- patients with spinal metastasis
Test items
simplified Tokuhashi (1990): 5 areas (Karnowsky general condition, tumour progress, risk factors, spinal compression, neurological symptoms);
in addition patient preferences
Scoring
0, 1 or 2 scores for 5 areas with a maximum total score of 10
Life expectancy prognosis/Treatment consequences
0–4 scores: conservative therapy,
7–10: surgery;
use as continuous scale (the higher the score, the more aiming at curative (surgery) rather than palliative care;
intermediate scores: patient preferences should be discussed with the patient
Received treatment -
Follow up -
Results
OUTCOMERESULT
No data presented
Author’s conclusion A score system is helpful; the primary question is from which therapy does the individual patient profit the most.
General comments German language publication.
Gasbarrini, A., Cappuccio, M., Bandiera, M.S., Terzi, S., Barbanti Brodano, G. & Boriani, S. (2004). Spinal metastases: treatment evaluation algorithm. European Review for Medical and Pharmacological Sciences, 8, 265–274.
Design: Case series (all undergoing same algorithm), evidence level 3
Country: Italy; Setting: Orthopaedic and traumatology department
Population N=182 patients with spinal metastases from a solid tumour (patients with plasmacytoma and lymphoma excluded)
Test items
Is the patient operable, Does the tumour respond to adjuvant therapies (assessed at multiple time points), Frankel score, Is there acute and ingravescent spinal cord damage, Is there a single metastasis only, Pathological fracture evaluation, Are there hypervascularised tumours/metastases from renal cell carcinoma and from sarcoma, Is en bloc removal easy to perform
Scoring
algorithm, sequential decision process; multidisciplinary input (anaesthetist, radiotherapist, oncologist)
Life expectancy prognosis/Treatment consequences
Adjuvant therapies (CHT, RXT . . .), pain relief, emergency surgery, resection of the lesion, decompression and stabilisation, en bloc resection or debulking
Received treatment 79 were treated with decompression and stabilisation, 64 with intralesional resection (debulking), 27 had an en bloc resection (vertebrectomy, corporectomy, sagittal resection, posterior resection)
Follow up 81 months
Results
OUTCOMERESULT
No data on comparison to other algorithm/different procedure
Author’s conclusion The treatment options for metastatic spinal disease has increased, it has become clear that effective implementation of these treatments can only be achieved by a multi-disciplinary approach.
General comments
Katagiri, H., Takahashi, M., Wakai, K., Sugiura, H., Kataoka, T. & Nakanishi, K. (2005). Prognostic factors and a scoring system for patients with skeletal metastasis. The Journal of Bone and Joint Surgery – British Volume, 87B.5, 698–703.
Design: Prognostic study, evidence level 3 (case series level)
Country: Japan; Setting: Division of Orthopaedic Oncology
Population N=350 consecutive patients with skeletal metastases
Test items
Site of primary lesion, performance status, presence of visceral or cerebral metastases, any previous chemotherapy, multiple skeletal metastases
Scoring
Primary lesion 0, 2 or 3 according to speed of growth, visceral or cerebral metastases 2; all other areas 1 point; possible range 0–8
Life expectancy prognosis/Treatment consequences
-
Received treatment 71% non-surgical (palliative alone, chemotherapy, radiotherapy alone, combined chemotherapy and radiotherapy), 29% surgery (internal fixation or endoprosthetic replacement, posterior decompression and instrumentation, endoprosthetic replacement and spinal instrumentation, resection without reconstruction, amputation)
Follow up at least 2 years, mean: 13 months for deceased, 39 for survivors
Results
OUTCOMERESULT
SurvivalScores 0–2: survival rate of more than 80% at 12 months; survival rate 0.98 at 6 months; 0.75 at 24 months
Scores 3–5: survival rate of 30–70% at 12 months, survival rate 0.71 at 6 months; 0.28 at 24 months
Scores 6–8: survival rate less than 20% at 12 months, survival rate 0.31 at 6 months; 0.02 at 24 months
Survival curvesThe survival curves of the three groups are significantly different (p<0.0001)
Author’s conclusion The scoring system can be used to determine the optimal treatment for patients with pathological fractures or epidural compression.
General comments The number of patients with spinal metastases is unclear. Test score and received treatment may be intercorrelated, and treatment and outcome (survival) may not be independent
Huch, K., Cakir, B., Ulmar, B., Schmidt, R., Puhl, W. & Richter, M. (2005). Prognose, operative Therapie und Verlauf bei zervikalen und hochthorakalen Tumorosteolysen. Z.Orthop.Ihre Grenzgeb., 143 (2), 213–218.
Design: Test validation study and case series, evidence level 3
Country: Germany; Setting: Department of Orthopaedic surgery and spinal cord injury
Population N=14 patients with tumour osteolysis
Test items
Tokuhashi (1990)
Tomita (2001)
Karnofsky-Index
Scoring
Tokuhashi (1990)
Tomita (2001)
Karnofsky Performance status 10–40% = 0, 50–70% = 1, 80–100% = 2
Life expectancy prognosis/Treatment consequences
Tokuhashi (1990), Tomita (2001)
Received treatment surgery with modular rod-screw implant system for the posterior instrumentation of the occipito-cervical, cervical and cervicotheracic spine; comotherapy or radiotherapy
Follow up
up to 3 years possible
Results
OUTCOMERESULT
Survival6 survivors: Tokuhashi score: 6.8 (+ − 1.6, CI: 6.0–7.7), mean survival 9.8 months
8 deceased: Tokuhashi score 8.5 (+ − 1.1, CI: 7.9–9.1), mean survival 11.5 months
Tokuhashi score 0–5: both patients still alive after approximately 3 and 7 months
Tokuhashi score 6–7: one patient died around 3, one just before 6 months, two around 9, one at 32 months, and three patients were still alive after 3/15 months
Tokuhashi score 8–12: two patients died at 3 or 4 months, one at 27 months
Tomita score 2–3: one patient died at 4, one at 8 months and two were still alive at 12 or 17 months
Tomita score 4–5: one patient died at 3, one at 27 and one at 32 months, one was still alive at 4 months
Tomita score 6–7: one patient died at 5 and one at 8 months, two patients were still alive 3 months or 15 months
Tomita score 8–10: one patient still alive at 1 month
Karnofsky index 1: two patient died at 5 or 5 months, two at 8 or 9 months, and one at 32 months, the survivors were still alive at 4, 12, 16 or 17 months
Karnofsky index 2: one patient died at 3 months, one at 22 months, the survivors were still alive at 2 and 3 months
Correlation total score and survivalIn the patients who died the correlation was 0.11 with the Tokuhashi score
Author’s conclusion The time of survival appeared to be difficult to estimate by a scoring system.
General comments German language publication. The scores were not used to select patients for surgery. The publication is at best an intermediate report, with only few patients and too short observation periods.
Ulmar, B., Richter, M., Cakir, B., Muche, R., Puhl, W. & Huch, K. (2005). The Tokuhashi score: Significant predictive value for the life expectancy of patients with breast cancer with spinal metastases. Spine, 30 (19), 2222–2226.
Design: Test validation study, evidence level 3
Country: Germany; Setting: Department of Orthopaedic surgery and spinal cord injury
Population N=55 consecutive patients with vertebral metastases secondary to breast cancer
Test items
Tokuhashi (2005)
Scoring
Tokuhashi (2005)
Life expectancy prognosis/Treatment consequences
Scores 0–5: predicted survival ≤ 3 months; palliative group
Scores 6–8: predicted survival ≤ 12 months
Scores 9–15: predicted survival > 12 months
Alternatively: Scores 0–4: group 1 (predicted survival < 3 months)
Alternatively: Scores 5–8: group 2 (≤ 12 months)
Received treatment surgery for spine fusion for reduction of pain or for neurological deficits
Follow up maximum 132.2 months, median 16.2
Results
OUTCOMERESULT
SurvivalMean survival 27.2 (+ − 28.6), range: 0.8–132.2; median: 16.2
Scores 0–5: mean survival: 11.5 months (+ − 17.2); range: 0.8–49.4; median: 5.0
Scores 6–8: mean survival: 21.5 months (+ − 18.4); range 1.9–64.9, median: 14.8
Scores 9–15: mean survival: 38.9 months (+ − 36.7); range: 2.7–132.2; median: 25.6
Scores 0–4: mean survival: 2.9 months (+ − 2.0)
Scores 5–8: mean survival: 21.7 months (+ − 18.4)
Prognostication/survival curveThe prognostication of the survival interval was significant (p<0.0103) for the original system; the modification showed a higher significance (p<0.0001)
Author’s conclusion The modified Tokuhashi score supports decision making based on reliable estimators of life expectancy in patients with breast cancer and spinal metastases.
General comments Test score and received treatment may be intercorrelated, and treatment and outcome (survival) may not be independent
Tokuhashi, Y., Matsuzaki, H., Oda, H., Oshima, M. & Junnosuke, R. (2005). A revised scoring system for preoperative evaluation of metastatic spine tumor prognosis. Spine, 30 (19), 2186–2191.
Design: Test evaluation/epidemiological study, evidence level 3
Country: Japan; Setting: Department of Orthopaedic Surgery
Population N=246 patients with metastatic spinal tumours
Test items General condition (performance status), number of extraspinal bone metastases foci, number of metastases in the vertebral body, metastases to the major internal organs, primary site of the cancer, palsy (Frankel)
Scoring
0, 1 or 2 scores for 5 areas, primary site of the cancer is scored 0 to 5 (see Tokuhashi 2002) with a maximum total score of 15
Life expectancy prognosis/Treatment consequences
0–8 scores: less than 6 months life expectancy; conservative or palliative surgery
9–11: up to 1 year life expectancy; palliative surgery or in case of single lesions only and no metastases to the major internal organ excisional surgery
12–15: 1 year or more; excisional surgery
Received treatment decisions based on oncologist opinion and preoperative prognostic score; 142 patients treated with palliative surgery (posterior decompression and stabilisation, posterior stabilisation alone, laminectomy), 22 treated with excisional surgery (anterior curettage and stabilisation, combined curettage and stabilisation, en bloc resection and stabilisation), 82 treated conservatively (radiation, chemotherapy, hormonal therapy, only analgesics). Surgery was not suggested for patients with predicted survival of <= six months, with a poor general condition, who responded well to oral narcotic analgesics, showed marked effects of radiotherapy, had ultra-rapid palsy progression (complete motor paralysis 2–3 days after onset) or had lost the inclination to live. Conservative therapy was given priority in patients with multiple metastases.
Follow up 115 months
Results
OUTCOMERESULT
SurvivalSurvival periods ranged from 10 days to 115 months (mean survival after treatment = 8.7 months, SD=12.3)
Scores 0–8: 85% survived up to 6 months;
Score 9–11: 73% survived 6 months or more;
Score 12–15: 95% survived 1 year or more
Respective numbers for a prospectively followed subgroup (N = 118) were 89%, 79% and 88%
Correlation total score and survivalr = 0.57 (p<0.0001); the correlations within the treatment group were 0.61, 0.53 and 0.62
Kaplan-Meier curves of score groupsThe mean survival amongst the three score cut-of groups differed significantly (p<0.01)
Author’s conclusion. The total scores from the revised scoring system were useful for the pre-treatment evaluation of metastatic spinal tumour prognosis irrespective of treatment modality or local extension of the lesion.
General comments Received treatment depended in parts on the score, and treatment and outcome (survival) may not be independent. It is possible that there is overlap in patients described in Tokuhashi (2002) and Tokuhashi et al. (1990), only some results for a subgroup of non-overlapping patients are presented.
Jansson, K-A & Bauer, H.C.F. (2006). Survival, complications and outcome in 282 patients operated for neurological deficit due to thoracic or lumbar spinal metastases. Eur Spine J, 15, 196–202.
Design: Case series, evidence level 3
Country: Sweden; Setting: Department of Orthopaedics
Population N=282 consecutive patients with thoracic or lumbar spinal metastasis
Test items
Frankel classification: A: complete paraplegia, B: no motor function, C: motor function useless, D: slight motor function deficit, E: no motor deficit
Scoring
A–C: non-walkers
D–E: walkers
Life expectancy prognosis/Treatment consequences
-
Received treatment 212 patients received posterior decompression and stabilisation, posterior stabilisation with rods (hooks, pedicle screws, mixed; CD, Isola, Synergy or USS implant; augmentation with methyl methacrylate), bone cement for anterior reconstruction, Z-plates or Synergy rods for instrumentation
Follow up 3 months postoperatively, later once yearly
Results
OUTCOMERESULT
SurvivalScores A–C: 0.26 survival rate at 1 year
Scores D–E: 0.33 survival rate at 1 year
Neurological functionScore A–B: 38% walked during follow up
Score A–D: 5% worsened postoperatively, 70% improved
Score C: 36% walked at discharge (new Frankel: D–E)
Score E: 100% retained normal motor function (still E)
Author’s conclusion. The post-operative survival was not related to the preoperative neurological function.
General comments Received treatment depended in parts on the score, and treatment and outcome (survival) may not be independent. No attempts were made to predict survival and compare it with actual survival, only the actual survival was recorded; no treatment consequences were formulated either.
Chow, E., Harris, K. & Fung, K. (2006). Successful validation of a survival prediction model in patients with metastases in the spinal column. Int. J. Radiation Oncology Biol. Phys., 65 (5), 1522–1527.
Design: Test evaluation study, evidence level 3
Country: Canada; Setting: Cancer Centre
Population N=231patients with metastases in the spinal column
Test items
Dutch model: Karnofsky Performance Score (KPS), primary tumour site, visceral involvement;
RRRP model: Primary cancer site, site of metastases, KPS, Edmonton Symptom Assessment Scale (ESAS) fatigue score, ESAS appetite score, ESAS shortness of breath score;
NRF method: Number of risk factors (non breast cancer, sites of metastases other than bone only, KPS≤50, fatigue score 4–10, appetite score 8–10, shortness of breath score 1–10
Scoring
Dutch model: KPS 10–40=0, 50–70=1, 80–100=2; Other cancer=1, lung=1, prostate=2, breast=3; present visceral involvement= 0, absent=1; total score range: 0–6
RRRP model: 0–7 partial scores (items were weighted according to survival prediction scores), total score range: 0–32
NRF method: 0–3 risk factors vs 4 vs 5–6
Life expectancy prognosis/Treatment consequences -
Received treatment not reported
Follow up 3, 6, 12 months, median: 44.9 months
Results
OUTCOMERESULT
SurvivalThe median survival was 7 months, range: 0–70
Dutch model score 0–3: median survival: 4.4 months (95% CI: 3.5–6.2)
Dutch model score 4–5: median survival: 12.2 months (95% CI: 9.6–16.0)
Dutch model score 6: median survival: 51.3 months (95% CI: 19.7–62.3)
RRRP model score 0–13: median survival: 3 months (95% CI: 2.3–3.8)
RRRP model score 14–19: median survival: 1.2 months (95% CI: 0.8–2.2)
RRRP model score 20–32: median survival: 0.6 months (95% CI: 0.3–0.9)
NRF method score 0–3: median survival: 12.2 months (95% CI: 10.2–16)
NRF method score 4–5: median survival: 3.3 months (95% CI: 2.3–4.5)
NRF method score 6: median survival: 1.6 months (95% CI: 1.1–4.3)
Survival probabilityDutch model score 0–3: 63% at 3 months, 42% at 6 months, 30% at 12 months
Dutch model score 4–5: 88% at 3 months, 74% at 6 months, 62% at 12 months
Dutch model score 6: 100% at 3, 6 and 12 months
RRRP model score 0–13: 91% at 3, 78% at 6, 61% at 12 months
RRRP model score 14–19: 67% at 3, 46% at 6 months, 42% at 12 months
RRRP model score 20–32: 44% at 3, 22% at 6 months, 9% at 12 months
NRF method score 0–3: 92% at 3 months, 77% at 6 months, 62% at 12 months
NRF method score 4–5: 53% at 3 months, 31% at 6 months, 25% at 12 months
NRF method score 6: 40% at 3 months, 18% at 6 months, 9% at 12 months
Survival curvesThe Dutch model led to significant separation of survival curves (p<0.0001).
The RRRP model led to significant separation of survival curves (p<0.0001).
The NRF method led to significant separation of survival curves (p<0.0001).
CalibrationR2 was 0.90 for the Dutch model and 0.86 for the RRRP model (predicted survival probability plotted against fraction of patients surviving past 3, 6 and 12 months)
Author’s conclusion The Dutch and the RRRP model were successfully validated, the Dutch model is easier to administer.
General comments The publication gave little information about the performed treatments. The test score and received treatment may be correlated, and treatment and outcome (survival) may not be independent. The publication does not cite the Tokuhashi or Tomita system.
Ulmar, B., Naumann, U., Catalkaya, S., Muche, R., Cakir, B., Schmidt, R., Reichel, H. & Huch, K. (2007a) Prognosis scores of Tokuhashi and Tomita for patients with spinal metastastases of renal cancer. Annals of Surgical Oncology, 14(2), 998–1004.
Design: Test evaluation study, evidence level 3
Country: Germany; Setting: Department of Orthopaedics
Population N=37 consecutive patients with spinal metastases of renal cancer
Test items
Tokuhashi (1990); Tomita (2001);
Scoring
Tokuhashi (1990); Tomita (2001)
Life expectancy prognosis/Treatment consequences
Tokuhashi (1990); Tomita (2001)
Received treatment Surgery for spine fusion, reduction of pain and/or neurological deficits
Follow up 83.3 months
Results
OUTCOMERESULT
SurvivalThe mean survival was 13.7 (+ − 18.9) months, range: 0–83.3, median: 6.2, 1-year survival: 13 patients, 2-year survival: 4 patients, 5-year survival: 2 patients
Tokuhashi scores 0–5: mean survival: 4.7 (+− 5.8) months, range: 0.1-2-.4, median: 2.6
Tokuhashi scores 6–8: mean survival: 9.5 (+− 10) months, range: 0–34.9, median: 6.2.
Patients with Tomita scores 2–3: mean survival: 8.7 (+− 7.6) months, range: 2.4–19.6, median: 6.4
Tomita scores 4–5: mean survival 23.3 (+− 30.9) months, range: 0–75.9, median: 17.2
Tomita scores 6–7: mean survival 21.6 (+− 34.6) months, range: 1.6–83.3, median: 7.7
Tomita scores 8–10: mean survival 10.6 (+− 11.1) months, range: 0.1–35.5, median: 5.7.
Percentage of correctly predicted survivalTokuhashi scores 0–5: 46% of patients died within the predicted (≤3 months)
Tokuhashi scores 6–8: 78% of patients died as predicted
Tokuhashi scores 9–12: 100% of patients survived as predicted (≥12 months)
Kaplan-Meier life- table analysisThe prediction with the Tokuhashi system was significant (p<0.0010).
The prediction with the Tomita system did not predict survival significantly (p=0.69)
OtherWhen analysing single predictors, only ‘general condition (Karnofsky)’ was significant (p<0.0001) for the prognosticated survival
Author’s conclusion The Tokuhashi system seems to be much more valuable than the Tomita score for surgical decisions in renal cancer patients with spinal metastases.
General comments The publication gave little information about the performed treatments, the Test score and received treatment may be correlated, and treatment and outcome (survival) may not be independent

From: Chapter 6, Treatment Selection and Strategies

Cover of Metastatic Spinal Cord Compression
Metastatic Spinal Cord Compression: Diagnosis and Management of Patients at Risk of or with Metastatic Spinal Cord Compression.
NICE Clinical Guidelines, No. 75.
National Collaborating Centre for Cancer (UK).
Copyright © 2008, National Collaborating Centre for Cancer.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licenses issued by the Copyright Licensing Agency in the UK. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.