Room: Exhibit Hall
Purpose: Bolus fit in radiotherapy (RT) for extremity sarcomas can be challenging due to the complex anatomy curvature. Rigid 3D-printed bolus has shown benefit in terms of conformality for particular RT treatment sites. However, its rigid nature may prove challenging for sarcoma sites, where minor variability in anatomy position may compromise bolus fit. This study investigates 3D-printed bolus fit on a knee sarcoma phantom under different knee flexions.
Methods: A ballistic gel phantom of a leg was created from the CT of a previously treated patient. This phantom can simulate knee flexion, mimicking interfractional leg position changes that may compromise bolus fit. A 3D-printed bolus was also created from the CT data and compared to an adhesive, in-house gel bolus. The two sets of bolus were tested for fit on a range of knee flexions with maximum flexion based on matching tolerance criteria. Fit for each bolus was assessed based on calculated Haussdorff distance (HD) distributions from CT scans of the phantom in different flexion positions. Resource allocation for bolus creation was also assessed.
Results: Mean and max HD for the 3D-printed bolus were (5.4, 16.4) mm and (4.8, 31.6) mm for the normal and max knee flexions. In comparison, gel bolus yielded (2.7, 12.4) and (2.7, 26) mm. Cost of the 3D-printed bolus was 240 CAD, and required a 28-hour fabrication time. The in-house gel bolus cost 80 CAD for a 5mm thick, 30 cm² sheet. Fabrication time was 12 hours.
Conclusion: Fit was compromised for a rigid 3D-printed bolus under clinically relevant interfractional motion on a knee phantom. The gel bolus provided a more conformal fit as flexion changed and its production was less resource intensive. Under interfractional motion, bolus fit for knee sarcoma looks to benefit more from the in-house gel bolus than rigid 3D-printed bolus.
Funding Support, Disclosures, and Conflict of Interest: Project was funded through the Cameron Daye Sarcoma Fellowship. No discolures. No conflicts of interest.