Abstract：
OBJECTIVE: Radiation myelitis (RM) is a rare complication of radiation therapy (RT). The Pediatric Normal Tissue Effects in the Clinic spinal cord task force aimed to identify RT dose effects and assess risk factors for RM in children. Through systematic review, we analyzed RT dose, fraction size, latency between completion of RT and toxicity, chemotherapy use, age when irradiated, and sex.
METHODS: We conducted literature searches of peer-reviewed manuscripts published from 1964 to June 2017 evaluating RM among children. Normality of variables was assessed with Kolmogorov-Smirnov or Shapiro-Wilk tests. Spearman\'s rank correlation coefficients were used to test correlations between RT dose/fraction size and latency between RT and development of toxicity.
RESULTS: Of 1329 identified and screened reports, 144 reports were fully reviewed and determined to have adequate data for analysis; 16 of these reports had a total of 33 cases of RM with a median age of 13 years (range, 0.2-18) at the time of RT. The most common primary tumor histologies were rhabdomyosarcoma (n = 9), medulloblastoma (n = 5), and Hodgkin lymphoma (n = 2); the most common chemotherapy agents given were vincristine (n = 15), intrathecal methotrexate (n = 12), and intrathecal cytarabine (n = 10). The median RT dose and fraction size were 40 Gy (range, 24-57.4 Gy) and 1.8 Gy (range, 1.3-2.6 Gy), respectively. RT dose resulting in RM in patients who also received chemotherapy was lower than in those not receiving chemotherapy (mean 39.6 vs 49.7 Gy; P = .04). There was no association of age with RT dose. The median latency period was 7 months (range, 1-29). Higher RT dose was correlated with longer latency periods (P = .03) to RM whereas sex, age, fraction size, and chemotherapy use were not. Two of 17 patients with adequate follow-up recovered from RM; unfortunately, it was fatal in 6 of 15 evaluable patients. Complication probability modeling was not possible because of the rarity of events.
CONCLUSIONS: This report demonstrates a relatively short latency from RT (with or without chemotherapy) to RM and a wide range of doses (including fraction sizes) associated with RM. No apparent association with age at the time of RT could be discerned. Chemotherapy appears to reduce spinal cord tolerance. Recovery from RM is rare, and it is often fatal.
METHODS: We conducted literature searches of peer-reviewed manuscripts published from 1964 to June 2017 evaluating RM among children. Normality of variables was assessed with Kolmogorov-Smirnov or Shapiro-Wilk tests. Spearman\'s rank correlation coefficients were used to test correlations between RT dose/fraction size and latency between RT and development of toxicity.
RESULTS: Of 1329 identified and screened reports, 144 reports were fully reviewed and determined to have adequate data for analysis; 16 of these reports had a total of 33 cases of RM with a median age of 13 years (range, 0.2-18) at the time of RT. The most common primary tumor histologies were rhabdomyosarcoma (n = 9), medulloblastoma (n = 5), and Hodgkin lymphoma (n = 2); the most common chemotherapy agents given were vincristine (n = 15), intrathecal methotrexate (n = 12), and intrathecal cytarabine (n = 10). The median RT dose and fraction size were 40 Gy (range, 24-57.4 Gy) and 1.8 Gy (range, 1.3-2.6 Gy), respectively. RT dose resulting in RM in patients who also received chemotherapy was lower than in those not receiving chemotherapy (mean 39.6 vs 49.7 Gy; P = .04). There was no association of age with RT dose. The median latency period was 7 months (range, 1-29). Higher RT dose was correlated with longer latency periods (P = .03) to RM whereas sex, age, fraction size, and chemotherapy use were not. Two of 17 patients with adequate follow-up recovered from RM; unfortunately, it was fatal in 6 of 15 evaluable patients. Complication probability modeling was not possible because of the rarity of events.
CONCLUSIONS: This report demonstrates a relatively short latency from RT (with or without chemotherapy) to RM and a wide range of doses (including fraction sizes) associated with RM. No apparent association with age at the time of RT could be discerned. Chemotherapy appears to reduce spinal cord tolerance. Recovery from RM is rare, and it is often fatal.
摘要：
目的：放射性脊髓炎（RM）是放射疗法（RT）的罕见并发症。临床脊髓任务组的小儿正常组织效应旨在确定RT剂量效应并评估儿童RM的危险因素。通过系统审查,我们分析了RT剂量，分数大小，RT完成和毒性之间的潜伏期，化疗使用，辐照时的年龄，和性爱。
方法：我们对1964年至2017年6月发表的评估儿童RM的同行评审手稿进行了文献检索。用Kolmogorov-Smirnov或Shapiro-Wilk检验评估变量的正式化。使用Spearman的等级相关系数来测试RT剂量/分数大小与RT和毒性发展之间的潜伏期之间的相关性。
结果：在1329份确定和筛选的报告中，全面审查了144份报告，并确定有足够的数据进行分析；其中16份报告共有33例RM，中位年龄为13岁（范围，0.2-18）在RT时。最常见的原发肿瘤组织学是横纹肌肉瘤（n=9），髓母细胞瘤（n=5），和霍奇金淋巴瘤（n=2）；最常见的化疗药物是长春新碱（n=15），鞘内注射甲氨蝶呤（n=12），鞘内注射阿糖胞苷（n=10）。中位RT剂量和分数大小为40Gy(范围，24-57.4Gy）和1.8Gy（范围，1.3-2.6Gy),分别。也接受化疗的患者的RT剂量导致RM低于未接受化疗的患者（平均39.6对49.7Gy；P=.04）。RT剂量与年龄无关。中位潜伏期为7个月(范围，1-29）。较高的RT剂量与RM的潜伏期较长(P=0.03)相关，而性别，年龄,分数大小，和化疗的使用没有。经过充分随访的17名患者中有2名从RM康复；不幸的是，15名可评估患者中有6名是致命的.由于事件的稀有性，不可能进行并发症概率建模。
结论：该报告显示了从RT（有或没有化疗）到RM的潜伏期相对较短，以及与RM相关的广泛剂量（包括分数）。在RT时与年龄没有明显的关联。化疗似乎会降低脊髓耐受性。从RM中恢复是罕见的，而且往往是致命的.
方法：我们对1964年至2017年6月发表的评估儿童RM的同行评审手稿进行了文献检索。用Kolmogorov-Smirnov或Shapiro-Wilk检验评估变量的正式化。使用Spearman的等级相关系数来测试RT剂量/分数大小与RT和毒性发展之间的潜伏期之间的相关性。
结果：在1329份确定和筛选的报告中，全面审查了144份报告，并确定有足够的数据进行分析；其中16份报告共有33例RM，中位年龄为13岁（范围，0.2-18）在RT时。最常见的原发肿瘤组织学是横纹肌肉瘤（n=9），髓母细胞瘤（n=5），和霍奇金淋巴瘤（n=2）；最常见的化疗药物是长春新碱（n=15），鞘内注射甲氨蝶呤（n=12），鞘内注射阿糖胞苷（n=10）。中位RT剂量和分数大小为40Gy(范围，24-57.4Gy）和1.8Gy（范围，1.3-2.6Gy),分别。也接受化疗的患者的RT剂量导致RM低于未接受化疗的患者（平均39.6对49.7Gy；P=.04）。RT剂量与年龄无关。中位潜伏期为7个月(范围，1-29）。较高的RT剂量与RM的潜伏期较长(P=0.03)相关，而性别，年龄,分数大小，和化疗的使用没有。经过充分随访的17名患者中有2名从RM康复；不幸的是，15名可评估患者中有6名是致命的.由于事件的稀有性，不可能进行并发症概率建模。
结论：该报告显示了从RT（有或没有化疗）到RM的潜伏期相对较短，以及与RM相关的广泛剂量（包括分数）。在RT时与年龄没有明显的关联。化疗似乎会降低脊髓耐受性。从RM中恢复是罕见的，而且往往是致命的.
