Contouring compliance and variability of targets and organs at risk in spine stereotactic body Radiotherapy: A 34-Institution study

Abstract

Background and purpose

To ensure the proper implementation of stereotactic body radiotherapy (SBRT) for spinal metastases in a multicenter clinical trial, this study evaluated the compliance and variability in contouring for spine SBRT.

Materials and Methods

A dummy run was conducted at 34 institutions. Computed tomography (CT) and T1/T2-weighted magnetic resonance imaging (MRI) of a patient with fifth thoracic spinal metastasis involving the vertebral body and right pedicle were provided. The observers performed CT-MRI registration and delineated the gross tumor volume (GTV), clinical target volume (CTV), and spinal cord. A radiation oncologist from the study office performed a central review of 17 checklist items. Agreement was assessed using the Dice similarity coefficient.

Results

An average of 3.0 protocol deviations per institution was observed. Deviations related to CTV setting in sectors adjacent to the GTV, the addition of a CTV margin to the paraspinal disease, and accurate delineation of the spinal cord were noted in 11, 13, and 12 institutions, respectively. The mean ± standard deviation Dice similarity coefficients for the GTV, CTV, and spinal cord were 0.84 ± 0.06, 0.85 ± 0.04, and 0.76 ± 0.09, respectively. Although nine institutions had clinically unacceptable errors, all met the criteria on the first resubmission. The mean Dice similarity coefficients for the spinal cord improved from 0.66 to 0.82 after resubmission.

Conclusion

This study revealed frequent deviations in CTV setting and spinal cord delineation. This dummy run clarified critical points in contouring for spine SBRT, contributing to the standardization of treatment planning.