目的：对比3D打印个体化非共面模板辅助CT引导125I粒子植入治疗锁骨上复发转移癌的术前计划与术后实际验证剂量，评价3D打印个体化非共面模板的剂量准确性。方法：研究纳入2016年1—9月于北京大学第三医院肿瘤放疗科行3D打印个体化非共面模板辅助CT引导125I粒子植入治疗锁骨上复发转移癌的患者14例（15个病灶），男性8例，女性6例，年龄46~79岁（平均年龄59.9岁），Karnofsky评分（KPS，百分法）60~90（中位数80）。患者既往未接受过放射治疗者1例，曾接受放射治疗但剂量不详者1例，其余12例曾接受放射治疗，中位数60 Gy (20~70 Gy)。所有患者术前均行增强CT扫描，术前计划设计，3D打印模板制作，穿刺及125I粒子植入，处方剂量110~150 Gy。对比患者术前、术后剂量学参数，包括90%靶体积的最小吸收剂量（D90）、100%处方剂量覆盖的体积占靶体积的百分比（V100）、150%处方剂量覆盖的体积占靶体积的百分比（V150）、200%处方剂量覆盖的体积占靶体积的百分比（V200）、匹配周缘剂量（matched peripheral dose, mPD）、适形指数（conformal index，CI）、靶区外体积指数（external index，EI）。利用配对t检验对术后实际剂量参数结果与术前计划所对应的参数进行比较，通过BlandAltman法分析术前、术后剂量参数的一致性。结果：术前和术后D90、V100、V150、V200、mPD、CI等差异均无统计学意义（P>0.05），EI差异有统计学意义（P<0.05）。经BlandAltman法分析，D90、V100、V150、V200、mPD一致性欠佳，CI、EI一致性较好。结论：3D打印个体化非共面模板辅助CT引导125I粒子植入治疗锁骨上复发转移癌方法剂量精准，可达到术前剂量设计要求。

Objective： To compare the dosimetric data between the preoperative plans and postoperative verification in personalized 3D printed template-assisted and computed tomography (CT) image-guided 125-iodine seed implantation for supraclavicular metastatic tumor, and to evaluate the accuracy and feasibility at the dosimetry level. Methods: A total of 14 patients with supraclavicular metastatic tumor (including 15 lesions) who received 3D printed template assisted and CT image-guided 125-iodine seed implantation in Department of Radiation Oncology of Peking University Third Hospital from January to September 2016 were enrolled. There were 8 males and 6 females, aged from 46 to 79 years (mean age: 59.9 years). The Karnofsky performance score (KPS) was from 60 to 90 (median of 80). There were one patient who had not received radiotherapy before, and one patient had received radiotherapy, but the dose was unknown. The remaining 12 patients had received radiotherapy, with the median of 60 Gy (20-70 Gy), and one of them with the dose unknown. All the patients underwent preoperative enhanced CT scan, preoperative planning design, 3D printing template, puncture and particle implantation, the prescription dose from 110 Gy to 150 Gy. Comparisons of the dosimetric parameters between the postoperative verification and preoperative plans were made by paired t-test. The dose parameters inclu-ded D90, V100, V150, V200, matched peripheral dose (mPD), conformal index (CI), and external index (EI). The agreement was evaluated between the preoperative planning and postoperative actual dose parameters using BlandAltman analysis. BlandAltman plot showed the difference against the average of preoperative planning and postoperative actual dose parameters with limits of agreement (LoA) (broken lines). Results: The difference of all the data between pre-and postoperation, included D90, V100, V150, V200, mPD, and CI, which was not statistically significant (P>0.05). EI was statistically significant (P<0.05). There was poor consistency of D90, V100, V150, V200, and mPD, but better consistency of CI and EI. Conclusion: Personalized 3D printed template-assisted and CT image-guided 125-iodine seed implantation for supraclavicular metastatic tumor is accurate and feasible.