MRI-Guided Brachytherapy PLAN Optimization with Three Versus FOUR Fraction Treatment for Locally Advanced Cervical Cancer

Abstract

MRI-guided brachytherapy (MRgBT) has become the standard treatment for locally advanced cervical cancer. The ability to optimize target and organ-at-risk (OAR) doses with three-dimensional image-based planning has improved toxicity profiles and local control. MRgBT, however is considerably more resource-intensive than the classical 2D planning and most of the available data on MRgBT has been based on a four fraction protocol of 7 Gy per fraction (7x4). A potential alternative to minimize resource utilization is a three fraction regimen of 8 Gy per fraction (8x3). However, there is limited data on MRgBT using the 8x3 protocol. The ability of this higher dose per fraction treatment to meet EMBRACE II planning aims has not been extensively studied. This study aims to compare optimized 8x3 vs 7x4 plan dosimetry. Ten patients with locally advanced cervical cancer (FIGO IB2 – IVA) treated at a single institution between July 2018 and January 2019 were included in this planning study. Clinical plans that were optimized and treated using a prescription 7x4 and the EMBRACE II planning aims (CTVHR D90% 90-95 Gy, rectum D2cm3 < 65 Gy and bladder D2cm3 < 80 Gy) were retrieved. Plans were then scaled to 8x3 and adjusted using graphical optimization. Three scenarios were evaluated: (1) The CTVHR D90% EQD2 dose was adjusted to be equivalent to the 7x4 plans; while limiting doses to organs at risk. Bladder and rectum D2cm3 doses and GTV D98% doses were recorded. (2) The bladder D2cc EQD2 dose was adjusted to be equivalent to the 7x4 plans (3) The rectal D2cc EQD2 dose was adjusted to be equivalent to the 7x4 plans. An a/b of 10 was used for the tumor bioequivalent dose calculations and an a/b of 3 for the OAR calculations. Doses were compared using Mann-Whitney-Wilcoxon. The median GTVres and CTVHR volumes were 13.42 cc (1.02 – 92.2) and 32.5 cc (10.99 – 120.4). For the 7x4 plans optimized for clinical treatment, median CTVHR D90%, rectum and bladder D2cm3 EQD2 doses were 93.7 Gy (87.4 – 104.4), 58 Gy (51.3 – 83.7) and 77.7 Gy (72.2 – 89.1). For scenario 1, where CTVHR D90% was made equivalent , median rectum and bladder D2cm3 were 58.5 Gy (51.3 – 84.6 ) and 77.1 Gy (73.1 – 90.0), respectively, with no significant differences compared to the 7x4 plans (P=1.0, P=0.85). Median GTVres D98% for 8x3 plans vs 7x4 was 105.1 Gy (84.5 – 150) vs 110.2 Gy (86.4 - 144.4), P = 0.63. When bladder and rectum in 8x3 were set equal to 7x4 plans (scenarios 1 & 2), there was no significant difference in CTVHR D90% doses (P=0.63, P=0.91) or GTVres D98% (P=0.53, P=0.53) . In this study, planning optimization with the 8x3 fraction protocol yields similar EQD2 doses for CTVHR D90%, bladder and rectum D2cm3 when compared with the 7x4 fraction protocol. The EQD2 GTV D98% is marginally lower for 8x3 when compared to 7x4; however, remains clinically acceptable. Further evaluation in a large cohort of patients is underway.