Purpose: The superior soft-tissue contrast of MRI imaging improves dramatically the delineation accuracy of the target and organs at risk (OAR), compared to CT. However, the reconstruction of the titanium tandem and ring (T&R) applicators on an MRI scan is a challenge that has not been adequately addressed. Treatment planning on a complementary CT scan, utilizing radiopaque markers for accurate applicator reconstruction, is currently the standard of care. In this study we evaluate the viability of a MRI-only workflow from a dosimetric perspective.
Methods: We retrospectively re-planned 19 HDR T&R fractions from 4 recently treated patients, who each had a CT scan (with an accurately reconstructed applicator) and a T2 weighted 3T MRI scan. An experienced planner reconstructed the T&R applicator on the MRI scan and planned each treatment without any input from the clinically treated plan. The original CT-based plans and the new MRI-based plans all met the GEC-ESTRO guidelines for target coverage and OAR sparing. The MRI-based plan was then recalculated on the CT scan, in the CT-reconstructed applicator geometry, for comparison with the clinically treated plan.
Results: The uncertainties of the MRI-based applicator reconstruction resulted in 47% (9 of 19) of the re-planned fractions failing the GEC-ESTRO criteria when recalculated in the “gold standard” CT geometry. The failures were more prominent in the bladder (26%), followed by the sigmoid and bowel (21%), and rectum (11%). The absolute D2cc dose differences ranged between -83cGy and +154cGy, while the mean differences were below 40cGy per fraction. When the fractional doses for each course of treatment were added, one of the 4 patients had bladder dose exceeding the 80Gy (EQD2) limit.
Conclusion: The titanium T&R applicator reconstruction uncertainties from a 3T MRI scan can lead to appreciable dosimetric differences between the MRI-based and the standard (MRI and CT) clinical workflows.
Brachytherapy, MRI, Treatment Planning