Feasibility study for marker-based VMAT plan optimization toward tumor tracking

This work investigates the incorporation of fiducial marker-based visibility parameters into the optimization of volumetric modulated arc therapy (VMAT) plans. We propose that via this incorporation, one may produce treatment plans that aid real-time tumor tracking approaches employing exit imaging of the therapeutic beam (e.g., via EPID), in addition to satisfying purely dosimetric requirements. We investigated the feasibility of this approach for a thorax and prostate site using optimization software (MonArc). For a thorax phantom and a lung patient, three fiducial markers were inserted around the tumor and VMAT plans were created with two partial arcs and prescription dose of 48 Gy (4 fractions). For a prostate patient with three markers in the prostate organ, a VMAT plan was created with two partial arcs and prescription dose 72.8 Gy (28 fractions). We modified MonArc to include marker-based visibility constraints ("hard"and "soft"). A hard constraint (HC) imposes full visibility for all markers, while a soft constraint (SC) penalizes visibility for specific markers in the beams-eye-view. Dose distributions from constrained plans (HC and SC) were compared to the reference nonconstrained (NC) plan using metrics including conformity index (CI), homogeneity index (HI), gradient measure (GM), and dose to 95% of planning target volume (PTV) and organs at risk (OARs). The NC plan produced the best target conformity and the least doses to the OARs for the entire dataset, followed by the SC and HC plans. Using SC plans provided acceptable dosimetric tolerances for both the target and OARs. However, OAR doses may be increased or decreased based on the constrained marker location and number of trackable markers. In conclusion, we demonstrate that visibility constraints can be incorporated into the optimization together with dosimetric objectives to produce treatment plans satisfying both objectives. This approach should ensure greater clinical success when applying real-time tracking algorithms, using VMAT delivery.