Toussaint André, Richter Anne, Mantel Frederick, Flickinger John C, Grills Inga Siiner, Tyagi Neelam, Sahgal Arjun, Letourneau Daniel, Sheehan Jason P, Schlesinger David J, Gerszten Peter Carlos, Guckenberger Matthias
Department of Radiation Oncology, University of Wuerzburg, Wuerzburg, Germany.
Departments of Neurological Surgery and Radiation Oncology, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA.
Radiat Oncol. 2016 Apr 18;11:57. doi: 10.1186/s13014-016-0631-9.
The aim of this study was to quantify the variability in spinal radiosurgery (SRS) planning practices between five international institutions, all member of the Elekta Spine Radiosurgery Research Consortium.
Four institutions provided one representative patient case each consisting of the medical history, CT and MR imaging. A step-wise planning approach was used where, after each planning step a consensus was generated that formed the basis for the next planning step. This allowed independent analysis of all planning steps of CT-MR image registration, GTV definition, CTV definition, PTV definition and SRS treatment planning. In addition, each institution generated one additional SRS plan for each case based on intra-institutional image registration and contouring, independent of consensus results.
Averaged over the four cases, image registration variability ranged between translational 1.1 mm and 2.4 mm and rotational 1.1° and 2.0° in all three directions. GTV delineation variability was 1.5 mm in axial and 1.6 mm in longitudinal direction averaged for the four cases. CTV delineation variability was 0.8 mm in axial and 1.2 mm in longitudinal direction. CTV-to-PTV margins ranged between 0 mm and 2 mm according to institutional protocol. Delineation variability was 1 mm in axial directions for the spinal cord. Average PTV coverage for a single fraction18 Gy prescription was 87 ± 5 %; Dmin to the PTV was 7.5 ± 1.8 Gy averaged over all cases and institutions. Average Dmax to the PRV_SC (spinal cord + 1 mm) was 10.5 ± 1.6 Gy and the average Paddick conformity index was 0.69 ± 0.06.
Results of this study reflect the variability in current practice of spine radiosurgery in large and highly experienced academic centers. Despite close methodical agreement in the daily workflow, clinically significant variability in all steps of the treatment planning process was demonstrated. This may translate into differences in patient clinical outcome and highlights the need for consensus and established delineation and planning criteria.
本研究的目的是量化五个国际机构（均为医科达脊柱放射外科研究联盟成员）之间脊柱放射外科（SRS）计划制定实践的差异。
四个机构各提供一个具有代表性的患者病例，包括病史、CT和MR成像。采用逐步计划制定方法，在每个计划制定步骤之后达成共识，该共识构成下一步计划制定的基础。这使得能够对CT-MR图像配准、GTV定义、CTV定义、PTV定义和SRS治疗计划的所有计划制定步骤进行独立分析。此外，每个机构基于机构内部的图像配准和轮廓勾画，为每个病例额外生成一个SRS计划，与共识结果无关。
在四个病例中，平均而言，图像配准差异在三个方向上的平移范围为1.1毫米至2.4毫米，旋转范围为1.1°至2.0°。四个病例的平均GTV勾画差异在轴向为1.5毫米，在纵向为1.6毫米。CTV勾画差异在轴向为0.8毫米，在纵向为1.2毫米。根据机构方案，CTV到PTV的边缘范围为0毫米至2毫米。脊髓在轴向的勾画差异为1毫米。单次分割18 Gy处方的平均PTV覆盖率为87±5%；所有病例和机构的平均PTV的Dmin为7.5±1.8 Gy。PRV_SC（脊髓+1毫米）的平均Dmax为10.5±1.6 Gy，平均帕迪克适形指数为0.69±0.06。
本研究结果反映了大型且经验丰富的学术中心当前脊柱放射外科实践中的差异。尽管在日常工作流程中有密切的方法学一致性，但在治疗计划过程的所有步骤中都显示出临床上显著的差异。这可能转化为患者临床结果的差异，并突出了达成共识以及确立勾画和计划标准的必要性。
