Consensus Guidelines for Target Volume Definition of the Sacrum in Spinal Stereotactic Body Radiation Therapy (SBRT)

Abstract

To develop consensus contouring guidelines for spinal stereotactic body radiation therapy (SBRT) for metastatic disease to the sacrum to improve uniformity in clinical target volume (CTV) delineation. Nine radiation oncologists with spinal stereotactic radiotherapy expertise representing 9 international centres in 3 countries independently contoured gross tumor volume (GTV), CTV and neural elements (cauda equina, thecal sac or sacral canal) for 10 representative clinical scenarios in metastatic disease to the sacrum. Contours were imported into an in-house software program, and agreement between physicians calculated with an expectation minimization algorithm using simultaneous truth and performance level estimation (STAPLE) and with kappa statistics (<0, poor agreement; 0.01-0.20, slight agreement; 0.21-0.40, fair agreement; 0.41-0.60, moderate agreement; 0.61-0.80, substantial agreement; 0.81-1.00, almost perfect agreement). Optimized confidence level consensus contours were obtained using a voxel-wise maximum likelihood approach. Clinicians also completed an 18-question survey about sacral metastatic disease SBRT practice including dose/fractionation schedule, margin expansion for GTV to CTV and CTV to planning target volume (PTV), the practice of contouring the sacral plexus and/or peripheral nerve roots and acceptable dose constraints for neural elements and small/large bowel. The mean STAPLE agreement sensitivity and specificity was 0.73 (range, 0.59-0.86) and 1.00 respectively for GTV and 0.59 (range, 0.47-0.69) and 1.00, respectively for CTV. The mean kappa agreement was 0.76 (range, 0.60-0.89) for GTV and 0.60 (range, 0.45-0.72) for CTV (P<.001 for GTV and CTV in all cases). Different dose/fractionation schedules used included 16-24 Gy/1 fraction, 24 Gy/2 fractions, 24-27 Gy/3 fractions, 30-36 Gy/4 fractions and 35 Gy/5 fractions, primary histology of the tumor was a factor in certain cases when choosing a schedule. Six experts applied an anatomic approach to the CTV, in some cases adding an extra margin for soft tissue extension/extra-osseous disease. Median PTV expansion was 1 mm (range, 0 – 3 mm). Six experts contoured the thecal sac as a surrogate for the cauda equina. In the majority of cases the sacral plexus and peripheral nerves were contoured however the lack of consensus contouring guidelines and low rates of clinical plexopathy observed in practice were reasons why contouring these critical normal structures wasn’t universal practice. There was substantial agreement for GTV and moderate agreement for CTV contours with the majority of international experts applying an anatomic approach when contouring the CTV. Final consensus guidelines will be demonstrated in illustrations at the meeting.