P168] Volumetric modulated arc therapy using multi-criteria dose plan optimization for anal cancer – Feasibility of selective sparing of organs at risk

Abstract

Purpose Radiotherapy treatment plan optimization often requires trade-offs between different organs at risk (OARs). This may facilitate individualized plan optimization, where sparing of specific OARs can be incorporated into plan selection. We previously demonstrated this for IMRT using multi-criteria dose plan optimization (MCO). We here present initial results for volumetric modulated arc therapy (VMAT) MCO. Methods VMAT dose plans were generated for ten representative anal cancer patients. Prescribed dose was 53.75 Gy to tumour (simultaneous integrated boost) and 45 Gy to elective nodes in 25 fractions. Plan optimization was done in RayStation® (v5.99.0.16 research build). Dual arc non-MCO VMAT plans were optimized with physician-defined organ-sparing priorities. Three additional VMAT dose distributions were optimized using MCO, representing a span of clinically relevant objectives: 1) minimum acceptable target coverage with physician-defined organ-sparing priorities; 2) maximum bowel sparing without losing target coverage (if needed at expense of the bladder); 3) maximum bladder sparing without losing target coverage (if needed at expense of the bowel). VMAT-MCO dosimetry metrics were compared to non-MCO VMAT plans; and feasibility of selective OAR sparing (bladder or bowel) with VMAT-MCO was compared to previously published findings. Results All plans satisfied constraints for target coverage. Non-MCO and VMAT-MCO plans showed similar conformity indices (CI) and dose volumes to bladder, although with considerable variation between patients. MCO plans had smaller hotspots in elective volumes (ΔV107%: 3.0%-points, interquartile range (IQR) [1.7–3.9]), but higher dose volumes to bowel (median bowel ΔV45 Gy: 57.5 cm[c]; IQR [21.3–298.0]). Compared to IMRT-MCO, VMAT-MCO demonstrated smaller hotspots and similar CI. VMAT-MCO allowed for redistribution of dose between bladder and bowel, but the optimization space was considerably smaller than for IMRT. To illustrate, median bowel ΔV45 Gy: 19.4 cm3; IQR [2.2–31.1] for bowel sparing VMAT-MCO compared to standard VMAT-MCO; while for IMRT-MCO ΔV45 Gy: 47.6 cm3; IQR [18.6–66.4]. Time spent on optimization was longer for VMAT-MCO due to considerable increase in the calculation and conversion time compared to IMRT-MCO. Conclusions Selective sparing of OARs in VMAT-MCO facilitates the same trade-offs as IMRT-MCO, but with significantly less room for dose redistribution. While VMAT-MCO plans are clinically acceptable, further improvement of the research implementation is warranted.