Cazoulat G, Lesaunier M, Simon A, Haigron P, Acosta O, Louvel G, Lafond C, Chajon E, Leseur J, de Crevoisier R
Inserm, U642, 35000 Rennes, France.
Cancer Radiother. 2011 Dec;15(8):691-8. doi: 10.1016/j.canrad.2011.05.011. Epub 2011 Nov 16.
In case of tumour displacement, image-guided radiotherapy (IGRT) based on the use of cone beam CT (tomographie conique) allows replacing the tumour under the accelerator by rigid registration. Anatomical deformations require however replanning, involving an estimation of the cumulative dose, session after session. This is the objective of this study.
Two examples of arc-intensity modulated radiotherapy are presented: a case of prostate cancer (total dose=80 Gy) with tomographie conique (daily prostate registration) and one head and neck cancer (70 Gy). For the head and neck cancer, the patient had a weekly scanner allowing a dose distribution calculation. The cumulative dose was calculated per voxel on the planning CT after deformation of the dose distribution (with trilinear interpolation) following the transformation given by a non-rigid registration step (Demons registration method) from: either the tomographie conique (prostate), or the weekly CT. The cumulative dose was eventually compared with the planned dose.
In cases of prostate irradiation, the "cumulative" dose corresponded to the planned dose to the prostate. At the last week of irradiation, it was above the planned dose for the rectum and bladder. The volume of rectal wall receiving more than 50 Gy (V50) was 20% at the planning and 26% at the end of treatment, increasing the risk of rectal toxicity (NTCP) of 14%. For the bladder wall, V50 were 73% and 82%, respectively. In head and neck, the "cumulative" dose to the parotid exceeded the planned dose (mean dose increasing from 46 Gy to 54 Gy) from the 5th week of irradiation on, suggesting the need for replanning within the first 5 weeks of radiotherapy.
The deformable registration estimates the cumulative dose delivered in the different anatomical structures. Validation on digital and physical phantoms is however required before clinical evaluation.
在肿瘤发生位移的情况下,基于锥形束CT(断层扫描圆锥)的图像引导放射治疗(IGRT)可通过刚性配准将加速器下的肿瘤复位。然而,解剖结构变形需要重新规划,这涉及逐次估计累积剂量。这就是本研究的目的。
给出了两个弧形调强放射治疗的例子:一例前列腺癌(总剂量 = 80 Gy)采用断层扫描圆锥(每日前列腺配准),另一例头颈部癌(70 Gy)。对于头颈部癌患者,每周进行一次扫描以计算剂量分布。在剂量分布变形(采用三线性插值)后,根据非刚性配准步骤(恶魔配准法)给出的变换,在计划CT上按体素计算累积剂量,该变换来自:要么是断层扫描圆锥(前列腺),要么是每周的CT。最终将累积剂量与计划剂量进行比较。
在前列腺照射的情况下,“累积”剂量与前列腺的计划剂量相对应。在照射的最后一周,直肠和膀胱的剂量高于计划剂量。接受超过50 Gy(V50)的直肠壁体积在计划时为20%,治疗结束时为26%,直肠毒性(NTCP)风险增加了14%。对于膀胱壁,V50分别为73%和82%。在头颈部,从照射第5周起,腮腺的“累积”剂量超过计划剂量(平均剂量从46 Gy增加到54 Gy),这表明在放疗的前5周内需要重新规划。
可变形配准可估计不同解剖结构中递送的累积剂量。然而,在临床评估之前,需要在数字和物理模型上进行验证。