Cazoulat G, de Crevoisier R, Simon A, Louvel G, Manens J-P, Lafond C, Haigron P
Inserm, U642, 35000 Rennes, France.
Cancer Radiother. 2009 Sep;13(5):365-74. doi: 10.1016/j.canrad.2009.06.002. Epub 2009 Jul 29.
To quantify the prostate and seminal vesicles (SV) anatomic variations in order to choose appropriate margins including intrapelvic anatomic variations. To quantify volumetric benefit of image-guided radiotherapy (IGRT).
Twenty patients, receiving a total dose of 70 Gy in the prostate, had a planning CT scan and eight weekly CT scans during treatment. Prostate and SV were manually contoured. Each weekly CT scan was registered to the planning CT scan according to three modalities: radiopaque skin marks, pelvis bone or prostate. For each patient, prostate and SV displacements were quantified. 3D maps of prostate and SV presence probability were established. Volumes including minimal presence probabilities were compared between the three modalities of registration.
For the prostate intrapelvic displacements, systematic and random variations and maximal displacements for the entire population were: 5mm, 2.7 mm and 16.5mm in anteroposterior axis; 2.7 mm, 2.4mm and 11.4mm in superoinferior axis and 0.5mm, 0.8mm and 3.3mm laterally. Margins according to van Herk recipe (to cover the prostate for 90% of the patients with the 95% isodose) were: 8mm, 8.3mm and 1.9 mm, respectively. The 100% prostate presence probability volumes correspond to 37%, 50% and 61% according to the registration modality. For the SV, these volumes correspond to 8%, 14% and 18% of the SV volume.
Without IGRT, 5mm prostate posterior margins are insufficient and should be at least 8mm, to account for intrapelvic anatomic variations. Prostate registration almost doubles the 100% presence probability volume compared to skin registration. Deformation of SV will require either to increase dramatically margins (simple) or new planning (not realistic).
量化前列腺和精囊（SV）的解剖变异，以便选择合适的边界，包括盆腔内的解剖变异。量化图像引导放射治疗（IGRT）的体积效益。
20例接受前列腺总剂量70 Gy的患者，在治疗期间进行了一次计划CT扫描和八次每周一次的CT扫描。手动勾勒前列腺和SV的轮廓。根据三种方式将每次每周的CT扫描与计划CT扫描进行配准：不透射线的皮肤标记、骨盆骨或前列腺。对每位患者的前列腺和SV位移进行量化。建立前列腺和SV存在概率的三维地图。比较三种配准方式下包括最小存在概率的体积。
对于前列腺盆腔内位移，整个人群的系统和随机变异以及最大位移分别为：前后轴5mm、2.7mm和16.5mm；上下轴2.7mm、2.4mm和11.4mm；横向0.5mm、0.8mm和3.3mm。根据van Herk公式（覆盖95%等剂量线的90%患者的前列腺）的边界分别为8mm、8.3mm和1.9mm。根据配准方式，100%前列腺存在概率体积分别对应37%、50%和61%。对于SV，这些体积分别对应SV体积的8%、14%和18%。
如果没有IGRT，5mm的前列腺后边界是不够的，应为至少8mm，以考虑盆腔内的解剖变异。与皮肤配准相比，前列腺配准几乎使100%存在概率体积增加一倍。SV的变形将需要大幅增加边界（简单）或重新规划（不现实）。
