Assessing the Necessity of Fiducial Markers for Image‑Guided Prostate SBRT: Do Implanted Markers Add Clinical Value In the Era of CBCT‑Based Soft‑Tissue Registration?
Abstract
Purpose
Implanted gold seeds fiducials (FMs) have historically been the standard for patient setup. At our center, 5‑fraction prostate SBRT treatments are delivered with beams at 2400 MU/min, resulting in short treatment times and reduced susceptibility to intrafraction motion. With standardized bowel and bladder preparation and high‑quality CBCT imaging and registration algorithms, soft‑tissue-based setup can be performed reliably by experienced therapists. Because FMs do not account for deformation (Mutanga et al., IJROBP), and particularly in workflows where real‑time fiducial tracking isn't used, their added value is uncertain. Given a growing waitlist for seeds implantation, we conducted a retrospective evaluation to assess the clinical utility of FMs in our workflow.
Methods
Thirty patients (15 seeds, 15 non‑seeds) treated with SBRT (36.25 Gy/5 with seeds; 35 Gy/5 palliative, without seeds) using identical bowel and bladder preparation were reviewed. Clinically accepted CBCT registrations and resulting shifts/rotations were extracted. In the seeds-cohort, an automated seed‑based registration was performed, and the fiducial registration error (FRE) for both accepted clinical registrations and seeds-based registrations was calculated. Residual shifts/rotations, and the number of CBCTs per treatment day were compared between the seeds and no-seeds cohorts.
Results
A total of 147 CBCTs and 133 were acquired for seeds and no-seeds cohort respectively (approx. 2/day). Mean FRE was 0.6 mm (0.1–1.4 mm) for seed‑based registrations vs 2.2 mm (0.6–9.8 mm) for clinical registrations. Systematic errors were small in both cohorts (≤0.35 mm, ≤0.91°) with differences ≤0.1 mm and ≤0.2°, and random errors were similarly low (≤0.51 mm, ≤1.57°) with differences ≤0.15 mm and ≤0.4°, indicating equivalent setup accuracy.
Conclusion
Clinical registrations often diverge from seeds-based alignment, showing that therapists rely primarily on soft‑tissue anatomy rather than fiducials when making final setup decisions. Setup accuracy in the no‑seeds cohort is not inferior, indicating that high‑quality CBCT soft‑tissue-only setup is feasible.