Evaluation of verification accuracy of two different immobilization methods in stereotactic body radiotherapy of early stage non-small cell lung carcinoma and pulmonary oligometastases

Abstract

Purpose: The aim of the study is to compare the positioning accuracy of two immobilization systems commonly used in stereotactic body radiation therapy (SBRT) of non–small cell lung carcinoma (NSCLC) and lung oligometastases, Polyurethane Foam Cradles (PFC) and the BodyFIX System (BFS) with 2D and 3D image guidance. Both the interfraction and intrafractional positional errors were analyzed.

Methods and Materials: 189 CBCT scans from 44 patients with NSCLC or lung oligometastases who received SBRT between August 2008 and April 2014 were analyzed retrospectively. Of these, 20 and 24 patients were immobilized with a Polyurethane Foam Cradle (PFC) and the BodyFIX System (BFS) respectively. The results of on board imaging (OBI) and pre-treatment cone-beam computed tomography (CBCT) at initial setup and after correction were registered to planning CT for online matching. The positional errors in anteroposterior (AP), superior-inferior (SI) and medial-lateral (ML) directions were analyzed. Post-treatment CBCT were used to assess intrafraction tumour displacement for 19 patients. The planning target volume margins were calculated using the van Herk’s formula. Other possible factors contributing to setup uncertainty were also analyzed.

Result: By using skin mark as a reference, the mean setup errors were 0.09, -0.10 and 0.02 cm for PFC and 0.04, -0.19 and -0.10 cm for BFS in AP, SI and ML directions respectively. The mean setup errors were decreased to 0.04, 0.02 and 0.04 cm for PFC; and -0.04, -0.04 and -0.02 cm for BFS after the application of OBI. The errors were further decreased to below 0.02 cm in all directions after the application of first pre-treatment CBCT in both immobilization methods. Statistically significant difference (p < 0.05 ) was only found in the comparison of AP error between the two devices, when OBI was used as the verification method. For PFC, the 3D vector errors of skin mark, OBI and first pre-treatment CBCT were 6.4 mm, 2.9 mm and 0.5 mm, respectively cases. For BFS, the errors were 7.1 mm, 3.0 mm and 0.4 mm, respectively. Smaller PTV margins in various directions were needed in BFS when using CBCT as the verification method. Positioning errors of skin mark setup in AP and SI directions had major contributions to all the setup errors; gender and tumour location might significantly affect the setup uncertainties. Comparatively large intrafractional errors were found in the post-treatment CBCT results of PFC.

Conclusion: When employing the CBCT-based final couch position as the benchmark, the setup errors of skin mark, OBI and first CBCT results were compared relatively. The positioning accuracies of PFC and BFS were similar. Apart from the vertical error (AP direction) found in the OBI verification, there was no significant difference between the positioning accuracy of both immobilization devices. Both imaging guided RT techniques were superior to skin mark. OBI and CBCT online correction improved the positioning accuracy of lung SBRT and substantially reduces required target margins and normal tissue irradiation.