SEARCH

SEARCH BY CITATION

Keywords:

  • metastatic spinal cord compression;
  • best supportive care;
  • radiotherapy;
  • survival prognosis;
  • scoring system.

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

BACKGROUND:

The objective of the current study was to develop a scoring system that identifies those patients with metastatic spinal cord compression who may be candidates for best supportive care or single-fraction radiotherapy.

METHODS:

Ten potential prognostic factors were retrospectively analyzed in 2029 patients, including age, gender, Eastern Cooperative Oncology Group performance status, tumor type, number of involved vertebrae, further bone metastases, visceral metastases, interval from time of cancer diagnosis to the development of MSCC, time to the development of motor deficits, and ambulatory status.

RESULTS:

On multivariate analysis, Eastern Cooperative Oncology Group performance status, tumor type, bone metastases, visceral metastases, interval from cancer diagnosis to the development of metastatic spinal cord compression, time to the development of motor deficits, and ambulatory status were found to be significantly associated with survival. The risk score represented the sum of the scores for each of these factors, obtained from the probability of the patient dying within 2 months (shown as the percentage) divided by 10. Risk scores ranged between 6 and 25 points. At a cutoff value of ≥ 24 points, the specificity was 99.8% and the positive predictive value was 96.0%, which indicates that approximately 4% of the patients predicted to die within 2 months survived > 2 months.

CONCLUSIONS:

This score identifies patients who have a very poor survival with a high specificity and a high positive predictive value. Patients with a score of ≥ 24 points have a very high probability of dying within 2 months. Thus, overtreatment with intensive therapies can be avoided in these patients, who are very unlikely to benefit.Cancer 2013. © 2012 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

Metastatic spinal cord compression (MSCC) is considered an oncologic emergency, and occurs in 5% to 10% of adult patients with cancer during the course of their disease.1 The length of survival for patients with MSCC ranges from a few weeks to several years. For those patients with an extraordinarily limited life span of only a few weeks, decompressive surgery is generally not indicated.2 Longer courses of fractionated radiotherapy appear to be similarly ill advised because the daily transport to the radiation oncology department and positioning on the treatment couch can be associated with added distress for these often heavily debilitated patients.

Single-fraction radiotherapy with 8 grays (Gy) results in similar pain relief and improvement of motor deficits in patients who receive radiation for MSCC as multifraction, short-course programs (eg, 20 Gy in 5 fractions over 1 week) or longer-course programs (eg, 30 Gy in 10 fractions over 2 weeks or 40 Gy in 20 fractions over 4 weeks).3–7 Therefore, a single fraction of radiotherapy can be recommended for patients with a very limited survival prognosis. Depending on the patient's preferences and performance status, even a single fraction of radiotherapy may be more than a gravely ill patient is willing or able to withstand. The patient has to lie flat and be immobile during a single treatment of 8 Gy over 10 to 15 minutes. Best supportive care including corticosteroids and analgesic drugs may be considered reasonable for carefully selected patients.1 To avoid burdensome overtreatment in patients with a very limited remaining lifetime, it would be helpful to have an instrument that can assist in identifying those patients who are likely to die within a few weeks. Because a scoring system was requested at several oncologic meetings within the past year, we decided to develop a survival score from our large database of patients with MSCC that helps identify those patients who are likely to die within 2 months. This information can be used to avoid unnecessary overtreatment for carefully selected patients.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

Patients

Data derived from 2029 patients who were treated with radiotherapy for MSCC between 1992 and 2011 were retrospectively analyzed. Ten potential prognostic factors were investigated, including age (≤ 64 years vs ≥ 65 years), gender, Eastern Cooperative Oncology Group (ECOG) performance status (1–2 vs 3–4), tumor type (breast cancer vs prostate cancer vs myeloma/lymphoma vs lung cancer vs others), number of involved vertebrae (1–2 vs ≥ 3), further bone metastases at the time of radiotherapy (no vs yes), visceral metastases at the time of radiotherapy (no vs yes), interval from cancer diagnosis to the development of MSCC (≤ 15 months vs > 15 months), time to the development of motor deficits before radiotherapy (1–7 days vs > 7 days), and preradiotherapy ambulatory status (no vs yes).

Criteria for inclusion were weakness of the legs due to MSCC, no prior surgery or radiotherapy to the involved parts of the spinal cord, adequate imaging with spinal computed tomography or spinal magnetic resonance imaging, and corticosteroid (12 mg–32 mg of dexamethasone per day) treatment during radiotherapy for at least 1 week. The data were collected from the patients, treating physicians, and patient files. The irradiated volumes encompassed 1 normal vertebra above and below the involved vertebrae.

Because the scoring system may have been influenced by the radiotherapy regimen, separate subgroup analyses were performed for patients who had received short-course radiotherapy (8 Gy in 1 fraction or 20 Gy in 5 fractions over 1 week) and for those patients who had received longer-course radiotherapy (30 Gy in 10 fractions over 2 weeks, 35 Gy–37.5 Gy in 14–15 fractions over 3 weeks, or 40 Gy in 20 fractions over 4 weeks).

Statistical Analysis

The 10 potential prognostic factors being investigated were evaluated in a multivariate analysis performed with the Cox proportion hazards model. The Bonferroni correction for multiple comparisons was used to adjust the P values derived from the multivariate analysis. Because 10 potential prognostic factors were investigated, values < .005 were considered statistically significant, representing an α level of < .05. The prognostic factors found to be independent predictors for survival were included in the risk score. Diagnostic usefulness of the risk score (sensitivity, specificity, and positive and negative predictive values) was calculated using the BiAS 9.02 statistical program.8

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

Of the 2029 patients, 396 (20%) died within 2 months after radiotherapy. On multivariate analysis, improved survival was found to be significantly associated with an ECOG performance status of 1 to 2 (P < .001), favorable tumor type (P < .001), no further bone metastases (P = .002), no visceral metastases (P < .001), a longer interval from cancer diagnosis to the development of MSCC (> 15 months; P < .001), a slower development of motor deficits (> 7 days; P < .001), and the ability to walk before radiotherapy (P < .001). The risk ratios and 95% confidence intervals of the multivariate analysis are summarized in Table 1.

Table 1. Multivariate Analysis of Survival
CharacteristicRisk Ratio95% CIPa
  • Abbreviations: 95% CI, 95% confidence interval; ECOG, Eastern Cooperative Oncology Group; MSCC, metastatic spinal cord compression, RT, radiotherapy.

  • a

    According to the Bonferroni adjustment, P values < .005 were to be considered statistically significant.

Age ≤ 64 y vs ≥ 65 y1.020.91-1.14.78
Gender: female vs male1.151.01-1.300.03
ECOG performance status: 2 vs 3-41.461.22-1.74<.001
Tumor type: breast cancer vs prostate cancer vs myeloma/lymphoma vs lung cancer vs other tumors1.181.23-1.23<.001
No. of involved vertebrae: 1-2 vs ≥31.030.94-1.12.55
Further bone metastases: no vs yes1.341.12-1.61.002
Visceral metastases: no vs yes4.463.92-5.08<.001
Interval from cancer diagnosis to MSCC: ≤15 mo vs > 15 mo1.221.14-1.29<.001
Time to developing motor deficits: 1-7 d vs > 7 d1.731.53-1.95<.001
Ambulatory status prior to RT: ambulatory vs not ambulatory1.941.65-2.29<.001

The score for each of the 7 significant prognostic factors was obtained from the probability of the patient dying within 2 months (shown as the percentage) divided by 10. The corresponding scores for each of the prognostic factors are given in Table 2. Risk scores represented the sum of the scores for each factor and ranged between 6 and 25 points. A higher score was associated with a worse survival prognosis (ie, a higher probability of the patient dying within 2 months after radiotherapy).

Table 2. Percentage of Patients Who Died Within 2 Months After RT and Corresponding Scores (N = 2029)a
CharacteristicPercentage of Patients Who Died Within 2 MonthsScore, Points
  • Abbreviations: ECOG, Eastern Cooperative Oncology Group; MSCC, metastatic spinal cord compression, RT, radiotherapy.

  • a

    A higher score means a worse survival prognosis.

ECOG performance status  
 2 (N = 1009)2%0
 3–4 (N = 1020)37%4
Tumor type  
 Breast cancer (N = 447)9%1
 Prostate cancer (N = 382)15%2
 Myeloma/lymphoma (N = 200)6%1
 Lung cancer (N = 399)32%3
 Other (N = 601)27%3
Further bone metastases  
 No (N = 888)12%1
 Yes (N = 1141)26%3
Visceral metastases  
 No (N = 1221)6%1
 Yes (N = 808)40%4
Interval from cancer diagnosis to MSCC, mo  
 ≤15 (N = 1031)27%3
  >15 (N=998)11%1
Ambulatory status prior to RT  
 Not ambulatory (N = 776)39%4
 Ambulatory (N = 1253)8%1
Time to developing motor deficits, d  
 1–7 (N = 628)41%4
 >7 (N = 1401)10%1

To achieve both high sensitivity and high specificity, the cutoff was selected at ≥ 19 points, at which point the sensitivity and specificity were 81.3% and 83.5%, respectively. The negative predictive value (NPV), which described the probability of surviving > 2 months if a patient had a score lower than the cutoff score, was also quite high (94.9%). This finding indicated that only 5.1% of patients who had a score of < 19 points were incorrectly predicted and died within 2 months after radiotherapy. However, the positive predictive value (PPV), which described the probability of dying within 2 months if a patient had a score ≥ the cutoff score, was only 54.5%. This indicates that 45.5% of patients with a score of ≥ 19 points were incorrectly predicted and survived > 2 months (median, 4 months; range, 3 months-17 months). Approximately 11.5% of patients with a score of ≥ 19 points survived for ≥ 6 months. These patients are likely to benefit from a more intensive regimen than best supportive care or single-fraction radiotherapy, but would not receive it if the cutoff was set at ≥ 19 points. Therefore, we attempted to determine a cutoff with a very high PPV to avoid undertreatment in patients who would potentially benefit from intensive treatment including longer-course radiotherapy or decompressive surgery. At a cutoff of ≥ 24 points, the PPV was 96.0% (ie, 96% of these patients were correctly predicted to die within 2 months after radiotherapy). The NPV and specificity were also quite high (84.4% and 99.8%, respectively). In contrast, the sensitivity was only 24.0% but was not critical for the decision to withhold intensive treatment and assign patients to receive best supportive care or single-fraction radiotherapy. The results of the sensitivity, specificity, PPV, and NPV rates for cutoff scores of 6 to 25 points in the entire cohort are summarized in Table 3. The 4 patients, who survived for > 2 months and had scores of ≥ 24 points, died after 5, 6, 10, and 12 months, respectively.

Table 3. Test Statistics of the Risk Score for the Entire Cohort{TC}
Cutoff ScoreDeath Within 2 MonthsSurviving > 2 MonthsSensitivity (95% CI), %Specificity (95% CI), %PPV (95% CI), %NPV (95% CI), %
 n = 396)n = 1633)    
  1. Abbreviations: 95% CI, 95% confidence interval; NPV, negative predictive value (probability of surviving >2 months if a patient had a lower score than the cutoff score); PPV, positive predictive value (probability of death within 2 months if a patient had a score ≥ the cutoff score).

≥11396975100 (99.1-100)40.3 (39.9-42.7)28.9 (26.5-31.4)100 (99.4-100)
≥1239487899.5 (98.2-99.9)46.2 (43.8-48.7)31.0 (28.4-33.6)99.7 (99.1-100)
≥1339381599.2 (97.8-99.8)50.1 (47.6-52.6)32.5 (29.9-35.3)99.6 (98.9-99.9)
≥1438772197.7 (95.7-99.0)55.9 (53.4-58.3)34.9 (32.1-37.8)99.0 (98.2-99.6)
≥1538467297.0 (94.8-98.4)58.9 (56.4-61.3)36.4 (33.5-39.4)98.8 (97.9-99.4)
≥1637353694.2 (91.4-96.3)67.2 (64.8-69.5)41.0 (37.8-44.3)98.0 (96.9-98.7)
≥1736143191.2 (87.9-93.8)73.6 (71.4-75.7)45.6 (42.1-49.1)97.2 (96.1-98.0)
≥1834335286.6 (82.9-89.8)78.4 (76.4-80.4)49.4 (45.6-53.1)96.0 (94.8-97.0)
≥1932226981.3 (77.1-85.0)83.5 (81.6-85.3)54.5 (50.4)-58.6)94.9 (93.6-95.9)
≥2029319974.0 (69.4-78.2)87.8 (86.1-89.4)59.6 (55.1-63.9)93.3 (91.9-94.5)
≥2124512161.9 (56.9-66.7)92.6 (91.2-93.8)66.9 (61.9-71.7)90.9 (89.4-92.3)
≥222239156.3 (51.3-61.3)94.4 (93.2-95.5)71.0 (65.7-76.0)89.9 (88.4-91.3)
≥231651941.7 (36.8-46.7)98.8 (98.2-99.3)89.7 (84.3-93.7)87.5 (85.9-89.0)
≥2495424.0 (19.9-28.5)99.8 (99.4-99.9)96.0 (90.0-98.9)84.4 (82.7-86.0)
 2585321.5 (17.5-25.8)99.8 (99.5-100)96.6 (90.4-99.3)84.0 (82.3-85.6)

Of the 99 patients with scores ≥ 24 points, only 4 patients (4%) were able to walk after radiotherapy, whereas 1844 of those 1930 patients with scores of < 24 points were ambulatory after radiotherapy.

The results of the 2 subgroup analyses of the patients who received short-course radiotherapy or longer-course radiotherapy were similar to the corresponding figures in the entire cohort (Table 4). Thus, the fractionation pattern did not appear to influence the results when using this scoring system.

Table 4. Subgroup Analysis: Test Statistics of Score in Patients Receiving Short-Course or Longer-Course RT
 Cutoff ScoreDeath Within 2 MonthsSurviving >2 MonthsSensitivity (95% CI), %Specificity (95% CI), %PPV (95% CI), %NPV (95% CI), %
  1. Abbreviations: 95% CI, 95% confidence interval; NPV, negative predictive value (probability of surviving >2 months if a patient has a lower score than the cutoff score); PPV, positive predictive value (probability of death within 2 months if a patient has a score ≥ the cutoff score); RT, radiotherapy.

Short-course RT (181 patients died within 2 mo; 641 patients lived > 2 mo)≥1816114189.0 (83.5-93.1)78.0 (74.6-81.2)53.3 (47.5-59.1)96.2 (94.1-97.6)
≥1915511185.6 (79.7-90.4)82.7 (79.5-85.5)58.3 (52.1-64.3)95.3 (93.2-96.9)
≥201388676.2 (69.4-82.2)86.6 (83.7-89.1)61.6 (54.9-68.0)92.8 (90.4-94.8)
≥211175464.6 (57.2-71.6)91.6 (89.2-93.6)68.4 (60.9-75.3)90.2 (87.6-92.4)
≥221024256.4 (48.8-63.7)93.5 (91.3-95.2)70.8 (62.7-78.1)88.4 (85.7-87.6)
≥2376942.0 (34.7-49.5)98.6 (97.4-99.4)89.4 (80.9-95.0)85.8 (83.0-88.2)
≥2449127.1 (20.8-34.2)99.8 (99.1-100)98.0 (89.4-99.9)82.9 (80.1-85.5)
2541022.7 (16.8-29.5)100 (99.4-100)100 (91.4-100)82.1 (79.2-84.7)
Longer-course RT (215 patients died within 2 mo; 992 patients lived >2 mo)≥1818221184.7 (79.1-89.2)78.7 (76.1-81.2)46.3 (41.3-51.4)96.0 (94.4-97.2)
≥1916715877.7 (71.5-83.1)84.1 (81.6-86.3)51.4 (45.8-56.9)94.6 (92.9-96.0)
≥2015511372.1 (65.6-78.0)88.6 (86.5-90.5)57.8 (51.7-63.8)93.6 (91.9-95.1)
≥211286758.7 (51.9-65.3)93.3 (91.5-94.7)65.6 (58.5-72.3)91.1 (89.2-92.8)
≥221214956.3 (49.4-63.0)95.1 (93.5-96.3)71.2 (63.7-77.9)90.9 (89.0-92.6)
≥23891041.4 (34.7-48.3)99.0 (98.2-99.5)89.9 (82.2-95.1)88.6 (86.6-90.4)
≥2446321.4 (16.1-27.5)99.7 (99.2-99.9)93.9 (83.1-98.7)85.4 (83.2-87.4)
2544320.5 (15.3-26.5)99.7 (99.2-99.9)93.6 (82.5-98.7)85.3 (83.1-87.3)

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

Many patients with MSCC have a very poor survival prognosis and may survive for only a few weeks.1 These patients are not candidates for decompressive surgery.2 Radiotherapy alone is the most reasonable alternative option. However, because each treatment session may be associated with discomfort for these often debilitated patients who are experiencing severe pain and decreased mobility, the overall treatment time of radiotherapy should be as short as possible. A single fraction of 8 Gy has been shown to be as effective as multifraction regimens with regard to pain relief and the improvement of motor function.3–7 Depending on the patient's performance status and personal preferences, even a single fraction of radiotherapy may cause undo distress for some extremely ill patients. To avoid overtreatment (spinal surgery or longer-course radiotherapy) in these patients with very limited lifespans, it would be helpful to have an instrument that helps to identify these patients before selecting the treatment regimen.

In the current study, a score was developed based on the data derived from 2029 patients who underwent radiotherapy for MSCC. At a cutoff of ≥ 24 points, the PPV was 96.0%. This means that only 4% of patients were not correctly predicted. Using our score, these patients would not receive intensive treatments, although 4% of them would be expected to survive > 2 months after radiotherapy.

Highly selected patients with a favorable survival prognosis and a good performance status may be candidates for decompressive surgery before radiotherapy.2 To the best of our knowledge, the role of upfront surgery in addition to radiotherapy for patients with MSCC is still unclear. In 1980, a small randomized study of 29 patients that compared radiotherapy plus upfront laminectomy (which is considered an inappropriate surgical approach) with radiotherapy alone found no difference with regard to pain relief, improved ambulation, or improved sphincter function between the 2 groups.9 In 2005, another randomized trial that was closed after an interim analysis of 101 patients demonstrated better posttreatment ambulatory status and survival if appropriate surgery (decompression plus stabilization of the involved vertebrae) was performed before radiotherapy was administered.2

If decompressive surgery is not indicated, longer-course radiotherapy alone would be preferable to best supportive care or single-fraction radiotherapy in patients with a favorable survival prognosis. Two randomized trials from Italy have included a single fraction of 8 Gy for the treatment of MSCC.10,11 However, to our knowledge, neither trials has compared 1 × 8 Gy with the most common regimen of 10 × 3 Gy but instead have compared it with very uncommon regimens. In the first trial, 1 × 8 Gy was compared with a split-course regimen comprised of 3 × 5 Gy followed by a week of rest and 5 × 3 Gy.10 In the second trial, 1 × 8 Gy was compared with 1 × 8 Gy followed by a week of rest and another fraction of 1 × 8 Gy.11 It can be concluded from both trials that the 1 × 8 Gy regimen is feasible and effective to improve or maintain motor function. However, these trials do not appear to contribute to the important question of whether 1 × 8 Gy is as effective as 10 × 3 Gy. Another prospective study from Germany and the Netherlands has compared short-course radiotherapy with 1 × 8 Gy or 5 × 4 Gy with longer-course radiotherapy with 10 × 3 Gy, 15 × 2.5 Gy, or 20 × 2 Gy.7 This prospective study has confirmed the results of a very large retrospective study demonstrating that longer-course radiotherapy resulted in significantly better local control of MSCC than short-course radiotherapy.7,12 Because the risk of a local recurrence in patients with MSCC increases with life expectancy, short-course radiotherapy cannot be recommended for patients with a favorable survival prognosis.

In the current study, subgroup analyses were also performed for patients who had received short-course radiotherapy and for those patients who had received longer-course radiotherapy to exclude a potential bias caused by the radiation regimen. At a cutoff of ≥ 24 points, the PPV was also found to be very high in these subgroups (98.0% and 93.9%, respectively). Therefore, this scoring system can be used regardless of the radiation regimen administered.

In the past, several scoring systems have been developed to predict survival in patients being treated with palliative care. In 1999, the Palliative Prognostic Score (PaP) was presented based on the data from 519 patients with a median survival of 32 days.13 The PaP included 6 prognostic factors: clinical prediction of survival, Karnofsky performance status, anorexia, dyspnea, total white blood count, and lymphocyte percentage. According to the results of a multivariate analysis, each factor was given a numerical score. The PaP represented the sum of the 6 factor scores. Based on these scores, 3 prognostic groups were designed, taking into account the probability of the patients still being alive at 30 days. The probabilities were 70% for a score of ≤ 5.5 points, 30% to 70% for 5.6 to 11.0 points, and < 30% for a score of > 11.0 points. The PaP has been validated in an additional 451 patients from an Italian hospice program.14 In contrast to our new score, the PaP was developed in a cohort of hospitalized patients with a great variety of advanced oncologic diseases, and was not designed for a specific oncologic diagnosis such as MSCC. It appears justified to examine patients with MSCC separately. In some patients, MSCC is the event that led to the initial diagnosis of cancer. Moreover, MSCC is considered an oncologic emergency, which generally requires that treatment begin as soon as possible.15 Therefore, patients with MSCC are quite different from terminally ill cancer patients in a hospice setting, who receive no active therapy. In the current study, approximately 20% of patients with MSCC (N = 396) died within 2 months of undergoing radiotherapy. This was unknown at the time treatment was delivered because of a lack of an appropriate scoring system. Of these 396 patients, 215 (54%) received what we would consider to be overtreatment with long-course radiotherapy over 2 to 4 weeks. This means that some patients spent ≤ 4 weeks of their remaining life span without daily radiotherapy. The primary goal of this new scoring system was to avoid these situations by identifying those patients who have an extremely poor prognosis.

Our new scoring system has been designed for patients referred for palliative treatment, whereas the PaP was designed for patients who received palliative care alone. The finding that the PaP is not appropriate for patients receiving active anticancer treatment has been recently suggested in a series of 208 patients with a median survival of 19 weeks who had been admitted to a department of medical oncology and radiotherapy in Italy.16

Another prognostic index for palliative cancer patients, the Palliative Prognostic Index (PPI), has been developed based on a series of 150 terminally ill cancer patients in a Japanese hospice.17 The PPI was based on performance status, oral intake, edema, dyspnea at rest, and delirium. For a PPI of 6 points used as the cutoff, the probability of surviving 3 weeks was predicted with a sensitivity of 80% and a specificity of 85%, and for a PPI using 4 points as the cutoff, the probability of surviving 6 weeks was predicted with a sensitivity of 80% and a specificity of 77%. However, similar to the PaP, the PIP was designed for terminally ill patients receiving palliative care in a hospice setting. This is in contrast to the scoring system in the current study, which has been designed for patients who were referred for the treatment of MSCC, which is considered to be a palliative oncologic emergency. Therefore, the development of our new score is both new and distinct. Moreover, a study from Germany suggested that both the PaP and the PIP are not able to predict an individual patient's prognosis precisely and reliably.18

In addition to scoring systems for terminally ill patients, Chow et al developed and validated a predictive model for patients attending an outpatient palliative radiotherapy clinic.19,20 In a series of 395 patients, 6 prognostic factors were found to be significantly associated with survival including primary tumor type, site of metastases, Karnofsky performance score, fatigue, appetite, and shortness of breath. The Survival Prediction Score (SPS) ranged from 0 to 32 points. Survival probability was measured at 3 months, 6 months, and 12 months. The 3-month survival rates were 83% for patients with an SPS of 0 to 13 points, 67% for patients with an SPS of 14 to 19 points, and 36% for patients with an SPS of 20 to 32 points. The 6-month survival rates were 70% for patients with an SPS of 0 to 13 points, 41% for patients with an SPS of 14 to 19 points, and 18% for patients with an SPS of 20 to 32 points. However, in contrast to the current scoring system, the SPS included a great diversity of patients referred for palliative radiotherapy. Furthermore, the SPS did not allow for the identification of patients with an extremely poor prognosis because 36% of the patients survived 3 months, even in the group with the worst prognosis.

In addition, prognostic scores have been developed in patients with bone metastases in the vertebral column. In 1990, Tokuhashi et al presented a score developed in 64 patients with a metastatic spine tumor who underwent spinal surgery, which was last revised in 246 patients in 2005.21, 22 van der Linden et al developed a score in 342 patients who received radiotherapy for painful spinal metastases without neurologic impairment.23 These scores were not designed for patients with motor deficits due to MSCC. Our own previous score, which was developed in 1852 patients who underwent radiotherapy for MSCC, included 5 prognostic groups and was based on 6-month survival rates.24 That previous score allowed us to estimate the patient's survival prognosis. However, it was not designed to offer best supportive care or single-treatment radiotherapy instead of multifraction radiotherapy. In addition, in contrast to the score described in the current study, several statistical analyses, in particular the PPV, were not calculated in the previous scoring system.

Considering the available scoring systems for predicting survival in patients receiving palliative care, the present scoring system, which can identify patients with MSCC who are very likely to die within 2 months after treatment with a very high PPV, is unique. This score was derived from retrospective data that may include hidden selection biases. However, survival was a valid and easily determined endpoint, even in a retrospective study, as were the majority of the prognostic factors included in our scoring system (ie, age, gender, tumor type, number of involved vertebrae, further bone metastases, visceral metastases, interval from cancer diagnosis to MSCC, and ambulatory status). The ECOG performance status, although well defined, is always subject to interpretation by the physician examining the patient. The time to the development of motor deficits before radiotherapy may sometimes be difficult to determine. This appears to be more difficult in a retrospective study than in a randomized trial. We chose 1 week versus > 1 week, which distinguishes between “rapid” and “slow” development. According to our 15 years of experience concerning MSCC studies, 1 week versus a longer period of time can be easily distinguished. However, the score is based on retrospective data. Furthermore, data regarding chemotherapy after radiotherapy were not available in the majority of patients. Because of these aspects, the risk of a hidden selection bias cannot be completely excluded. Therefore, our score optimally should be validated in a large prospective study. This might not happen in the near future because it would likely require several years to perform just such a prospective trial, which would require the accrual of hundreds of patients treated with radiotherapy alone for MSCC. For example, a current randomized multicenter trial in Germany comparing 2 radiation regimens for patients with MSCC has taken 1 year to accrue 50 patients.

The new scoring system described in the current study was able to identify patients with MSCC who are likely to die within 2 months after radiotherapy. Because patients with a score ≥ 24 points have a very high probability of dying within 2 months (PPV, 96.0%), some degree of discomfort and overtreatment can be avoided by offering single-dose radiotherapy or best supportive care. This score should be validated in a large prospective study. However, such a prospective study would be unlikely for several years. Therefore, this new scoring system should be considered the most appropriate index to help avoid unnecessary overtreatment in patients with MSCC who have an extremely poor prognosis and will not benefit from aggressive therapy.

FUNDING SUPPORT

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

No specific funding was disclosed.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES
  • 1
    Prasad D, Schiff D. Malignant spinal-cord compression. Lancet Oncol. 2005; 6: 1524.
  • 2
    Patchell R, Tibbs PA, Regine WF, et al. Direct decompressive surgical resection in the treatment of spinal cord compression caused by metastatic cancer: a randomised trial. Lancet. 2005; 366: 643648.
  • 3
    Steenland E, Leer JW, van Houwelingen H, et al. The effect of a single fraction compared to multiple fractions on painful bone metastases: a global analysis of the Dutch Bone Metastasis Study. Radiother Oncol. 1999; 52: 101109.
  • 4
    8 Gy single fraction radiotherapy for the treatment of metastatic skeletal pain: randomised comparison with a multifraction schedule over 12 months of patient follow-up. Bone Pain Trial Working Party. Radiother Oncol. 1999; 52: 111121.
  • 5
    Chow E, Harris K, Fan G, Tsao M, Sze WM. Palliative radiotherapy trials for bone metastases: a systematic review. J Clin Oncol. 2007; 25: 14231436.
  • 6
    Rades D, Stalpers LJ, Veninga T, et al. Evaluation of five radiation schedules and prognostic factors for metastatic spinal cord compression. J Clin Oncol. 2005; 23: 33663375.
  • 7
    Rades D, Lange M, Veninga T, et al. Final results of a prospective study comparing the local control of short-course and long-course radiotherapy for metastatic spinal cord compression. Int J Radiat Oncol Biol Phys. 2011; 79: 524530.
  • 8
    Ackermann H. BiAS 9.02: Biometrische Analyse von Stichproben. Darmstadt, Germany: Epsilon Verlag; 2009.
  • 9
    Young RF, Post EM, King GA. Treatment of spinal epidural metastases. Randomized prospective comparison of laminectomy and radiotherapy. J Neurosurg. 1980; 53: 741748.
  • 10
    Maranzano E, Bellavita R, Rossi R, et al. Short-course versus split-course radiotherapy in metastatic spinal cord compression: results of a phase III, randomized, multicenter trial. J Clin Oncol. 2005; 23: 33583365.
  • 11
    Maranzano E, Trippa F, Casale M, et al. 8Gy single-dose radiotherapy is effective in metastatic spinal cord compression: results of a phase III randomized multicentre Italian trial. Radiother Oncol. 2009; 93: 174179.
  • 12
    Rades D, Fehlauer F, Schulte R, et al. Prognostic factors for local control and survival after radiotherapy of metastatic spinal cord compression. J Clin Oncol. 2006; 24: 33883393.
  • 13
    Pirovano M, Maltoni M, Nani O, et al. A new palliative prognostic score: a first step for the staging of terminally ill cancer patients. Italian Multicenter and Study Group on Palliative Care. J Pain Symptom Manage. 1999; 17: 231239.
  • 14
    Maltoni M, Nanni O, Pirovano M, et al. Successful validation of the palliative prognostic score in terminally ill cancer patients. Italian Multicenter Study Group on Palliative Care. J Pain Symptom Manage. 1999; 17: 240247.
  • 15
    Loblaw DA, Laperriere NJ. Emergency treatment of malignant extradural spinal cord compression: an evidence-based guideline. J Clin Oncol. 1998; 16: 16131624.
  • 16
    Numico G, Occelli M, Russi EG, et al. Survival prediction and frequency of anticancer treatment in cancer patients hospitalized due to acute conditions. Role of clinical parameters and PaP score. Support Care Cancer. 2011; 19: 18231830.
  • 17
    Morita T, Tsunoda J, Inoue S, Chihara S. The Palliative Prognostic Index: a scoring system for survival prediction of terminally ill cancer patients. Support Care Cancer. 1999; 7: 128133.
  • 18
    Stiel S, Bertram L, Neuhaus S, et al. Evaluation and comparison of two prognostic scores and the physicians' estimate of survival in terminally ill patients. Support Care Cancer. 2010; 18: 4349.
  • 19
    Chow E, Fung K, Panzarella T, Bezjak A, Danjoux C, Tannock I. A predictive model for survival in metastatic cancer patients attending an outpatient palliative radiotherapy clinic. Int J Radiat Oncol Biol Phys. 2002; 53: 12911302.
  • 20
    Chow E, Abdolell M, Panzarella T, et al. Validation of a predictive model for survival in metastatic cancer patients attending an outpatient palliative radiotherapy clinic. Int J Radiat Oncol Biol Phys. 2009; 73: 280287.
  • 21
    Tokuhashi Y, Matsuzaki H, Toriyama S, Kawano H, Ohsaka S. Scoring system for the preoperative evaluation of metastatic spine tumor prognosis. Spine (Phila Pa 1976). 1990; 15: 11101113.
  • 22
    Tokuhashi Y, Matsuzaki H, Oda H, Oshima M, Ryu J. A revised scoring system for preoperative evaluation of metastatic spine tumor prognosis. Spine (Phila Pa 1976). 2005; 30: 21862191.
  • 23
    van der Linden YM, Dijkstra SP, Vonk EJ, Marijnen CA, Leer JW; Dutch Bone Metastasis Study Group. Prediction of survival in patients with metastases in the spinal column: results based on a randomized trial of radiotherapy. Cancer. 2005; 103: 320328.
  • 24
    Rades D, Dunst J, Schild SE. The first score predicting overall survival in patients with metastatic spinal cord compression. Cancer. 2008; 112: 157161.