3033 - Comparative Analysis of Matching Accuracy and Dose Discrepancies: Bone vs. Fiducial Alignment in Free-Breathing Ablative Pancreatic Cancer Radiotherapy

Y. Liu¹, P. Zhang², S. L. Berry³, Y. C. Hu³, L. I. Cervino², C. Gui⁴, C. H. Crane⁴, M. Reyngold⁴, and W. Lu³; ¹Memorial Sloan Kettering Cancer Center, MIDDLETOWN, NJ, ²Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, ³Memorial Sloan Kettering Cancer Center, New York, NY, ⁴Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY

Purpose/Objective(s): To quantify the misalignment of fiducial markers when aligning to bone during free-breathing ablative pancreatic cancer radiotherapy and to estimate the resulting dose discrepancies. Materials/

Methods: 24 patients with pancreatic cancer who underwent free-breathing ablative radiotherapy were included in this study. Fiducial markers were contoured on the free-breathing planning CT, and a structure called “iFiducials” was generated by combining fiducial contours from all 10 phases of the 4DCT, to be used for image registration. Fiducials on the pre-treatment on-board CBCT images were also contoured. A rigid registration based respectively on the bone (spine were used in our study) or on fiducials was performed to align one CBCT to the planning CT. Misalignments between fiducial and bone registrations were measured. The resulting dose discrepancies were calculated by shifting the isocenter accordingly on the planning CT. The impact on D95%, V100%, and Dmax of ITV and PTV was evaluated. For ablative treatment, usually two PTV dose levels are prescribed by physicians - ablative dose PTV (APTV) and microscopic disease PTV (MPTV), and they were investigated separately. In addition, the distance between the fiducial and bone was measured to evaluate its impact on the misalignment and dose discrepancy.
Results: The mean misalignments between fiducial and bone registrations were 1.7±1.3, 3.2±3.4, and 5.2±4.0 mm in the left-right, anteroposterior, and superoinferior direction, respectively. The observed shift may be caused by baseline drift of the target due to luminal OAR changes. Aligning to the iFiducial has a significant higher dose coverage to the tumor (t-test) compared to the bone. On average, D95% for APTV was 78.2±12.8% for bone matching and 91.4±8.1% for fiducial matching (p-value < 0.0001); for MPTV, it was 88.6±19.0% for bone matching and 99.7±4.4% for fiducial matching (p-value = 0.0092); and for ITV, it was 71.6±15.8% for bone matching and 75.6±15.6% for fiducial matching (p-value = 0.0182). V100% for APTV was 68.0±16.9% for bone matching and 81.8±10.4% for fiducial matching (p-value = 0.0001); for MPTV it was 87.5±7.4% for bone matching and 94.7±5.5% for fiducial matching (p-value < 0.0001); and for ITV it was 62.2±28.1% for bone matching and 67.9±26.5% for fiducial matching (p-value = 0.0027). No correlation was found between the misalignment of iFiducials and the distance to the spine.
Conclusion: This study reveals considerable interfraction soft tissue misalignment when aligning to the bone during free-breathing pancreatic cancer radiotherapy. The observed misalignment affected dose distribution, with significant dose discrepancies noted for target (tumor), raising concerns about treatment safety and efficacy. These findings highlight the critical need for refined alignment strategies to match to fiducial instead of bone for free-breathing ablative radiotherapy to ensure accurate dose delivery and effective treatment outcomes for pancreatic cancer.