Main Session
Sep 29
PQA 05 - Physics

3131 - Retrospective Margin and Micro-Boost Target Analysis in CBCT-Guided Adaptive Prostate SBRT

12:30pm - 01:45pm ET
Poster Hall - Exhibit Hall A
Screen: 22
POSTER

Presenter(s)

Domenic Sievert, MS - Washington University School of Medicine in St. Louis, St. Louis, MO

D. Sievert, E. Laugeman, E. Sigmund, R. Beckert, W. Liu, A. Willett, M. Mahmood, H. A. Gay, and A. K. Bhatt; WashU Medicine, Department of Radiation Oncology, St. Louis, MO

Purpose/Objective(s): This study assesses appropriate planning target volume (PTV) margins for the prostate and MRI-defined microboost volumes (GTVmb), based on inter- and intra-fractional motion quantified using a CBCT-guided adaptive platform.

Materials/Methods: Twenty-five prostate cancer patients were enrolled on a prospective CBCT-guided adaptive protocol to 5 fractions delivered once weekly. All patients were treated to 4 dose levels: 25Gy to the pelvic LNs, 36.25Gy to the prostate and proximal seminal vesicles (SV), 40Gy to the prostate, and 50Gy to the GTVmb. Prostate + SV PTV margins were 5mm (3mm posterior), while prostate-only PTV used a 2mm superior, inferior, and lateral margins. There was no margin for GTVmb. For each adaptive session, targets were rigidly aligned via advanced CBCT imaging using implanted fiducial markers. Verification CBCTs were acquired pre-treatment (CBCTpre) and mid-treatment (CBCTmid), and applied shifts were evaluated. In a subset of 10 patients, fraction 5 (>5 weeks post fiducial implantation) online propagated GTVmb coverage was compared to coverage of the pre-plan GTV determined by MRI registration to the online CBCT (GTVpre).

Results: Over 125 treatments, CBCTpre and CBCTmid shift amplitudes were assessed. Absolute median [IQR] CBCTpre shifts (vertical, longitudinal, lateral) were 0.06 [0.15] cm, 0.00 [0.12] cm, and 0.00 [0.04] cm, respectively; corresponding CBCTmid shifts were 0.00 [0.03] cm, 0.00 [0.00] cm, and 0.00 [0.00] cm. During delivery, 97.6% of cases remained within 3 mm shifts in all translational directions. Protocol aims for GTVmb included V45Gy > 95% (primary) and V50Gy > 85% (secondary). Mean [SD] V45Gy for GTVmb and GTVpre were 99.0% [2.0%] and 97.4% [3.2%]; corresponding V50Gy was 92.9% [6.7%] and 74.8% [14.7%]. Two out of ten cases did not meet the V45Gy > 95% for GTVpre, where it did meet for GTVmb, but still maintained Dmin > 40 Gy.

Conclusion: Analysis of intrafraction motion across 125 adaptive stereotactic fractions supports a 3 mm isotropic margin for prostate + SV. CBCTpre shifts were generally larger, with >10% of fractions exceeding 3mm in any direction, likely due to the adaptive process time, however CBCTmid shifts supported that a 3 mm intrafraction margin was adequate during delivery. GTVmb coverage prescription target coverage robustness declined by fraction 5 (> 5 weeks post-implant), however 90% of cases of the subset patient population still met protocol goals. These findings warrant further investigation of the fiducial robustness over extended treatment intervals.