Urgent Magnetic Resonance Imaging for Malignant Spinal Cord Compression

Full Article

SSM Lo, KY Kwok, CW Siu, WK Wong, CM Chan, KW Tang, YL Cheung, SCH Chan

Hong Kong J Radiol 2010;13:111-9

Objectives: Malignant spinal cord compression is an uncommon complication in patients with cancer. It is frequently diagnosed late and causes irreversible damage to spinal cord. Malignant spinal cord compression is an oncological emergency, requiring a prompt imaging diagnosis and immediate treatment. Magnetic resonance imaging has a vital role in the management of such patients. Our audit was to assess whether patients with malignant spinal cord compression were recognised, investigated, and treated appropriately in our hospital. Large international audits of malignant spinal cord compression, such as clinical resource and audit group, were used as our reference standards. We also compared the level of spinal cord compression inferred clinically and based on magnetic resonance imaging findings, and evaluated whether localised spinal magnetic resonance imaging based on clinical assessment as adopted in our routine magnetic resonance imaging protocol was appropriate.

Methods: From January to December 2008, 1087 patients underwent magnetic resonance imaging scans of the spine in our hospital. Of these, 106 patients presented with clinical features suspicious of malignant spinal cord compression and urgent magnetic resonance imaging spine was performed. The demographic factors, primary tumour pathology, Tokuhashi score, clinical symptoms, clinical suspicion of the level of cord compression, magnetic resonance imaging findings, treatment response, and prognosis were assessed retrospectively from the medical records, the electronic Patient Record, and imaging reports from the Radiology Information System.

Results: Among the patients with a clinical suspicion of malignant spinal cord compression, prevalence rates were higher in males (61%) and older persons (mean age, 63 years; standard deviation, 14 years). Lung cancer was the most common responsible primary tumour in males (38%), whereas breast cancer was the most common primary tumour in females (32%). Spinal cord compression was the initial presentation of malignancy in 6 (8%) patients. Most patients showed evidence of bone metastasis (53%) and metastasis in major internal organs (51%) prior to the clinical presentation of malignant spinal cord compression. The most common clinical symptoms of malignant spinal cord compression were back pain (77%), limb weakness (94%) and numbness (90%). Of the 106 patients, 43% showed a low Tokuhashi score (<5) during admission. 77% of patients with a clinical suspicion of malignant spinal cord compression had magnetic resonance imaging spine in our hospital within 24 hours; the upper thoracic spine was the most common level of cord compression among these patients, and multilevel compression occurred in 8% of them. There was considerable discrepancy between the level of spinal cord compression inferred clinically and that determined by magnetic resonance imaging (57%); the average level of discrepancy being 5 vertebral bodies (95% confidence interval, 4-6). A greater degree of discrepancy was evident in patients with multiple spinal metastases (p < 0.05). 89% of patients with a clinical suspicion of malignant spinal cord compression had received steroids. The mean time required for starting definitive treatment after confirmation of the diagnosis by magnetic resonance imaging was 1.3 (standard deviation, 0.5) days; 74% were treated within 24 hours of the imaging. Patients with malignant spinal cord compression had an unsatisfactory survival rate of 75% at hospital discharge; the median survival time from first clinical presentation was 37 days (range, 9-884 days). The majority (57%) of patients showed clinical improvement after radiotherapy.

Conclusion: Patients with malignant spinal cord compression in our hospital were recognised, investigated, and treated appropriately. Our results were comparable to those reported in the literature. We found a great discrepancy between the level of spinal cord compression inferred clinically and by magnetic resonance imaging, particularly when multiple spinal metastases were present. We concluded that localised magnetic resonance imaging of the spine based on clinical findings was inadequate. Our magnetic resonance imaging protocol for malignant spinal cord compression patient was modified based on the results of this study.

 

中文摘要

惡性脊髓壓迫綜合症的緊急磁共振成像

羅尚銘、郭啟欣、蕭志偉、王旺根、陳智明、鄧國穎、張毓靈、陳慈欽

目的:惡性脊髓壓迫綜合症(malignant spinal cord compression,MSCC)是癌症病人一種少見的併 發症。由於確診經常較遲,對患者的脊髓會造成永久性損傷。MSCC屬腫瘤急症,必須要有及時的 影像診斷及即時治理,而磁共振成像(MRI)在其中扮演很重要的角色。本研究的目的是評估本院 在檢測、診斷及治療MSCC患者時是否恰當。我們把大型國際MSCC審核(如Clinical Resource and Audit Group;CRAG)作為參考標準,並把分別從臨床及MRI診斷得出的脊髓壓迫位置作比較,來判 斷我們慣常使用的MRI設定模式(即按醫生的臨床判斷而作MRI局部脊髓掃描)是否恰當。

方法:本院在2008年內共有1087名病人進行脊髓MRI掃描,其中106名病人因懷疑出現MSCC症狀而 進行緊急脊髓MRI。我們在這些病人的病歷紀錄、電子病歷及放射科信息系統內的影像報告中找出 有關患者的人口學資料、原發性腫瘤病理、Tokuhashi分數、臨床症狀、臨床估計的脊髓壓迫位置、 MRI結果、治療反應及預後。

結果:臨床疑診MSCC患者中,男性(61%)及年齡較大的病人(平均63歲;標準差14歲)有較高患 病率。原發腫瘤方面,男性病人最常見的是肺癌(38%),而女性病人則是乳癌(32%)。6名病人 (8%)脊髓壓迫為首發的腫瘤病症。大部分病人在出現MSCC臨床症狀前有骨轉移瘤(53%)及主 要內臟器官的轉移瘤(51%)。最常見的MSCC臨床症狀為背部疼痛(77%)、四肢無力(94%)和 麻木(90%)。有43%患者入院時都有低Tokuhashi分數(即少於5)。疑診MSCC患者中有77%於24 小時內進行脊髓MRI。上胸椎是最常見的脊髓壓迫位置,而8%患者出現多重壓迫位置。從臨床診斷 及MRI診斷得出的脊髓壓迫位置有很大的差距(57%),平均相差5個椎體(95%可信區間:4-6)。 多重脊髓壓迫患者的臨床及MRI診斷脊髓壓迫位置差距較大(p < 0.05)。疑診MSCC患者中,有89% 服用類固醇。被MRI確診的MSCC患者中,從確診至開始治療的時間平均1.3天(標準差0.5天);有 74%病人於MRI檢查後24小時內得到治療。MSCC患者直至出院為止的生存率並不理想,只有75%。 從首次病發至死亡的生存中位數為37天(介乎9-884天)。大部分病人(57%)接受放射治療後臨床 情況有好轉。

結論:本研究的結果與其他文獻相似,本院恰當地檢測、診斷及治療MSCC患者。研究結果亦顯示 從臨床及MRI診斷得出的脊髓壓迫位置有較大的差距,尤其在脊髓多發轉移瘤的患者。所以只按臨 床診斷而進行MRI脊髓局部掃描涵蓋範圍未必足夠。根據本研究的結果,我們醫院對MSCC患者所 進行的MRI檢查設定有所更改。