3D打印模板联合CT引导下放射性粒子植入治疗椎旁／腹膜后恶性肿瘤的剂量学验证观察

Dosimetry verification of radioactive seed implantation with 3D printing template and CT guidance for paravertebral/retroperitoneal malignant tumor

目的对比3D打印模板辅助放射性粒子植入治疗椎旁/腹膜后恶性肿瘤的术前、术后计划的剂量学结果，在计划层面探讨该技术对放射性粒子植入治疗精确性的影响，为优化、规范化植入技术提供数据支持。 方法纳入2015年12月—2016年7月于北京大学第三医院接受3D打印模板辅助CT引导放射性125I粒子植入的椎旁/腹膜后肿瘤患者共10例（12个病灶），病种包括宫颈癌、肾癌、腹部间质瘤、肾平滑肌肉瘤、食管癌、输尿管癌。处方剂量给予110～150 Gy。所有患者行术前计划设计、个体化模板设计及制作、穿刺及植入粒子，术后验证的剂量学结果与术前计划相对比，剂量学参数包括90%靶区体积接受剂量（D90）、靶区最小周边剂量（MPD）、接受100%和150%处方剂量的靶区体积百分比（V100和V150）、适形指数（CI）、靶区外体积指数（EI）以及2 cm3体积正常组织（脊髓、腹主动脉、肾）所接受到的剂量（D2cc）。统计软件采用SPSS 19.0，统计学方法采用非参数相关样本Wilcoxon检验，P〈0.05为差异有统计学意义。 结果设计并制作3D打印个体化模板10块，共12个治疗区。术后靶区（GTV）平均D90 131.1（97.8～167.4）Gy。模板引导下，穿刺针分布良好。术后验证D90、MPD、V100、V150的均值分别为131.1 Gy、69.3Gy、90.2%、65.2%，术前则分别为140.2 Gy、65.6 Gy、91.7%、26.8%，表明术后实际靶区接受处方剂量较术前预计划略低、靶区内部分高剂量范围较术前要大，但除V150外（P＝0.004），两组间各参数差异均无统计学意义（均P〉0.05）。术后CI均值较术前小（分别为0.58、0.62），表明术后实际靶区剂量适形度较术前差，差异有统计学意义（P＝0.019）。术后EI均值较术前大（分别为55%、45.9%），表明术后实际靶区外接受处方剂量的体积略大，但差异无统计学意义（P〉0.05）。正常组织剂量方面，术后脊髓、主动脉和肾的平均剂量分别为24.7、54.4、29.7 Gy，均较术前略有增加（分别为20.6、51.6、28.6 Gy），但两组差异无统计学意义。 结论3D打印模板辅助CT引导放射性粒子植入椎旁/腹膜后病灶的术后验证实际剂量分布与术前计划一致性良好，应用3D打印模板有良好的治疗准确性。

ObjectiveTo compare the dose distributions of postoperative plans with preoperative plans for seeds implantations of paravertebral/retroperitoneal tumors assisted by 3D printing guide template and CT guidance, explore the effects of the technology for seeds implantations in dosimetry level and provide data support for the optimization and standardization in seeds implantation. MethodsBetween December 2015 and July 2016, a total of 10 patients with paravertebral/retroperitoneal tumors （12 lesions） received 3D printing template assist radioactive seeds implantations in department of radiation oncology of Peking University Third Hospital, and included in the study. The diseases included cervical cancer, kidney cancer, abdominal stromal tumor, leiomyosarcoma of kidney, esophageal cancer and carcinoma of ureter. The prescribed doses was 110-150 Gy. All patients received preoperative planning design, individual template design and production, and the dose distribution of postoperative plan was compared with preoperative plan. Dose parameters including D90, MPD, V100, V150,conformal index（CI）, EI of target volume and D2cc of organs at risk （spinal cord, aorta, kidney）. Statistical software was SPSS 19.0 and statistical method was non-parameters Wilcoxon symbols test. ResultsA total of 10 3D printing templates were designed and produced which were including 12 treatment areas.The mean D90 of postoperative target area （GTV） was 131.1 （97.8-167.4 Gy） Gy. The actual seeds number of post operation increased by 3 to 12 in 5 cases （42.0%）. The needle was well distributed. For postoperative plans, the mean D90,MPD, V100,V150 was 131.1 Gy, 69.3 Gy, 90.2% and 65.2%, respectively, and which was 140.2 Gy, 65.6 Gy, 91.7% and 26.8%, respectively, in preoperative plans. This meant that the actual dose of target volume was slightly lower than preplanned dose, and the high dose area of target volume was larger than preplanned range, but there was no statistical difference in P value between the two groups except V150（P=0.004）. The actual dose conformity of target volume was worse than preplanned （CI was 0.58 and 0.62, respectively） and the difference was statistically significant（P=0.019）. The actual dose of external target volume was higher than preplanned （EI was 55% and 45.9%, respectively） and the difference had no significance. For organs at risk, the actual mean D2cc of spinal cord, aorta and kidney was 24.7, 54.4 and 29.7 Gy, respectively, which was higher than preplanned（20.6, 51.6 and 28.6 Gy, respectively）, and there was no significant difference in two groups. ConclusionsMost parameters of postoperative validations for 3D printing template assisted seeds implantation in paravertebral/retroperitoneal are closed to the expectations of preoperative plans which means the improvement of accuracy in treatment.