Uncertainty of inter-application D2cc locations represented in DIR of 4-fraction HDR IGABT of cervix

Abstract

Purpose or Objective Traditional summation of D2cc with worst case scenario, by means of total D2cc overlapping assumption, may overestimate dose to OARs, caveating opportunities to optimize dose to CTV in cervical cancer IGABT. Exploration of inter-application rectal and bladder D2cc locations by using deformable image registration (DIR) was performed. Materials and Methods Subpopulation of cervix IGABT in Chokvanich W, et al. (DOI: 10.1016/j.ijrobp.2020.07.1549) were explored, including patient with 2-Gy/F 50 Gy 3D-CRT without pelvic boost or midline block, and 4-5 fractions of CT-based dose calculation. 1st BT fraction CT was targeted as a primary image for both VOI based rigid registration and DIR from EBRT and subsequent BT fractions. EBRT+BT dose summation was performed. Bladder and rectal D2cc of each fractionation and of deformable image registered dose summation (DIR-D2cc,sum) were identified. Inter-application D2cc spatial variability were explored in two sets, expressed in terms of DSC, MDA, and HD. First comparison was between inter-fractional and 1st fraction (iFx-1st) D2cc, using 1st fractional D2cc as a reference. Second comparison was between inter-fractional and summed (iFx-Sum) D2cc, using DIR-D2cc,sum as a reference. Friedman test was performed to evaluate correlation of vaginal applicator type and DSC of D2cc spatial variety. Results Of 14 patients, whole rectal and bladder showed mean DSC of 0.54 and 0.75 from EBRT registration, while yielding 0.70 and 0.91 in among BT fractions. DIR-D2cc,sum were -0.82 ± 3.2 (SD) Gy for rectum, and -1.77 ± 6.2 Gy for bladder, comparing to traditional summation. iFx-1st D2cc of rectum showed mean DSC of 0.32 ± 0.17, and significantly correlated with vaginal applicator type (Friedman p=0.025), MDA of 3.8 ± 2.2 mm, and HD of 17.2 ± 6.2 mm. Whereas others did not show significant Friedman test. iFx-1st D2cc of bladder showed mean DSC of 0.32 ± 0.17, MDA of 2.9 ± 1.5 mm, and HD of 15.1 ± 5 mm. iFx-Sum D2cc of rectum showed mean DSC of 0.45 ± 0.14, MDA of 2.6 ± 1.4 mm, and HD of 14.9 ± 6.9 mm. iFx-Sum D2cc of bladder showed mean DSC of 0.44 ± 0.15, MDA of 1.9 ± 0.9 mm, and HD of 12.9 ± 4.7 mm. Conclusion Inter-application D2cc location versatility was observed higher in rectum than bladder, and iFx-1st than iFx-Sum D2cc. Rectal iFx-1st D2cc expressed highest spatial variability with DSC of 0.32 and significantly correlated with vaginal applicator type. By exploitation of DIR for cumulative D2cc spatial summation, voxel-based analysis could be applied to optimize higher dose to CTV.