Dosimetric evaluation of dose calculation performed on the average-4DCT imaging for lung cancer patients with intensity-modulated radiation therapy

Abstract

Introduction: One of the major challenges of delivering intensity modulated radiation therapy (IMRT) in non-small cell lung cancer (NSCLC) is managing tumour motion. Acquiring 4-dimensional Computer Tomography (4DCT) images can depict the position of both lung tumours and normal structures with respiration. By utilizing these 4DCT images, the internal target volume (ITV) is delineated to encompass the full extent of the target volume’s motions throughout all breathing phases. An average CT (Ave-CT) data set is generated from the pixel-by-pixel average of the ten phases of the 4DCT for dose calculations. When delivering radiation therapy to lung cancer patients, intrafractional motions arising from normal respiration and organ motion may cause some disparity in the patient’s anatomy compared to the Ave-CT data. The actual absorbed dose distribution is effectively influenced by the magnitude of tumour motion during respiration. Purpose: To assess the dose calculation performed on the Ave-CT dataset by comparing the tumour volume coverage and organ at risk (OAR) dose between the treatment plans that were used for treatment with their respective 4-dimensional (4D) plans. Material and Methods: Hybrid-RapidArc (H-RA) plans were retrieved for 20 stage III NSCLC patients (UICC (Union of International Cancer Control) 7th Edition), which used the Ave-CT data set for dose calculations. The reference plans were denoted as the “Ave-CT plans.” 4D dose distributions were generated using deformable image registration to account for the effects of respiration induced density variations and to represent the actual dose to NSCLC patients receiving high dose RT (60 Gray (Gy) in 30 fractions). Firstly, the Ave-CT plans were duplicated onto the 10 phases of the patient’s respiration cycle. With each phase contributing to one-tenth of the cumulative dose, the dose distributions were deformed and summated onto the Ave-CT data set. The accumulated dose distribution was labelled the “4D plan.” Cumulative dose volume histograms (DVH) were generated for both the Ave-CT plans and their respective 4D plans. The planning target volume (PTV) and OAR parameters of the 4D plan were measured and compared to that of their initial Ave-CT plan. Results: The 4D plans yielded lesser conformality than their respective Ave-CT plans (p < 0.001). The volume of the prescription dose (p < 0.001), the 98 % isodose volume (V98%) (p = 0.001) and the 95 % isodose volume (V95%) (p = 0.002) were also demonstrated to be significantly larger in the 4D plans. The 4D plans were more heterogeneous than their Ave-CT plans, which were supported by the generation of hot spots or cold spots in the PTV of the 4D plans. The PTV mean dose (p < 0.001), mean lung-CTV dose (p = 0.020), spinal cord maximum dose (p < 0.001), mean esophagus dose (p < 0.001) and esophagus maximum dose (p = 0.004) were also significantly higher in the 4D plans than their Ave-CT counterparts. Conclusion: The coverage of the PTV by the prescription isodose volume was greater in the 4D plans at the expense of decreased conformity. In cases where the OAR doses were close to their dose constraint on the Ave-CT plans, performing a 4D dose calculation may be considered since the larger high dose volumes around the PTV may increase the dose to nearby OARs.