Impact of Immobilization on Interfractional Errors for Upper Extremity Soft Tissue Sarcoma Radiation Therapy

Introduction/Background: Owing to the rare nature and presentation of upper extremity soft tissue sarcomas (STSs) and the high mobility of associated anatomy, various patient positioning strategies are used for radiation therapy. The purpose of this study is to measure the interfractional setup errors associated with upper extremity sarcomas using different immobilization methods through conebeam computed tomography (CBCT) images. Methods: All patients treated with daily CBCT guidance for upper extremity STSs during 2014-2015 were identified and triaged based on type of immobilization. After defining an optimal region of interest for image registration, daily CBCT images were automatically registered to reference CT images to quantify positional discrepancies. Means and standard deviations were calculated, and one-way analysis of variance was calculated to determine significance of data. Results: Seventeen patients with upper extremity sarcoma met inclusion criteria: 13 were treated to the shoulder/axilla/upper arm and 4 to the arm/elbow/forearm. Three main types of immobilization were identified: vacuum cradle with custom thermoplastic shell, vacuum cradle alone, and no immobilization accessory used. Patient repositioning occurred if translational and rotational displacements were larger than 1 mm and 5 degrees, respectively, as per institutional guidelines. Patient repositioning rates were 18% for vacuum cradle with thermoplastic shells, 15% for vacuum cradles only, and 6% for no immobilization accessories. Mean translational displacements in right/left (R/L), superior/inferior (S/I), and anterior/posterior (A/P) directions were -0.04 +/- 0.33 cm, 0.32 +/- 0.33 cm, and 0.12 +/- 0.25 cm for vacuum cradle with thermoplastic shell; 0.25 +/- 0.10 cm, -0.07 +/- 0.22 cm, and 0.00 +/- 0.17 cm for vacuum cradle alone; and 0.14 +/- 0.15 cm, 0.08 +/- 0.45 cm, and -0.01 +/- 0.24 cm for no immobilization. For all patients, rotational displacements in the pitch, roll, and yaw were 0.15 +/- 1.99 degrees, 0.31 +/- 2.11 degrees, and -0.21 +/- 1.76 degrees, respectively. There were significant (P <.05) differences in systematic error values for all translational and rotational axes between immobilization methods. Conclusion: Large interfractional errors, especially in the rotational axes, were observed, regardless of immobilization strategy. Small study population and unequal representation of different parts of the upper extremity are identified limitations. The need for better immobilization techniques for upper extremity STS treatments is clearly demonstrated.