Biological optimization for mediastinal lymphoma radiotherapy – a preliminary study

Figures & data

Figure 1. Beam setup for preplanning before optimization with 16 beam angles surrounding the patient (A) and examples of open and partially closed subfields that were created for each beam angle (B–E). The CTV is shown in red.

Figure 2. Comparison of beams and dose distributions for clinical 3DCRT, VMAT, and outcome-optimized (O-OPT) plans for an example patient with a large benefit (patient 3 in Table S5). The CTV is shown in pink and the heart is shown in yellow. The PTV is shown in cyan for the VMAT plan.

Figure 3. Total risk for outcome-optimized (O-OPT) plans compared to clinical 3DCRT plans for all patients in two cardiac risk factor (CRF) scenarios (CRF = 0 and CRF > 0).

Figure 4. Dose-volume histograms (DVHs) showing the difference in O-OPT plans (CRF = 0 with hot-spot avoidance requirement) with respect to gEUD parameter choice in target model for O-OPT planning for one patient (patient 25). The DVHs from the clinical 3CDRT plan are shown in dotted lines.

Table 1. Recalculation of the term $p_{DR}$ (risk of disease recurrence) for various values of the gEUD parameter a for one patient (patient 25).

gEUD parameter a (for recalculation, not optimization)	1	−1	−5	−9	−13	−22
Recalculated $p_{\mathrm{DR}}$ for clinical 3DCRT plan (%)	0.04	0.05	0.05	0.06	0.06	0.07
Recalculated $p_{\mathrm{DR}}$ for O-OPT plan (%)	3.0	8.1	13.9	14.4	14.6	14.8
Increase in $p_{\mathrm{DR}}$ for the O-OPT plan relative to clinical 3DCRT plan for each value of a (O-OPTa-3DCRTa) (%)	3.0	8.0	13.9	14.4	14.5	14.7
Increase in $p_{\mathrm{DR}}$ for the O-OPT plan for each value of a relative to O-OPT plan with a =1 (O-OPTa-O-OPTa=1) (%)	0.0	5.0	10.9	11.4	11.6	11.7

For this recalculation, the O-OPT plan that was analyzed was created with the assumption of a = 1. Then, the plan was kept constant and $p_{DR}$ was recalculated to see the impact of a different “true” value of a if 1 was assumed during optimization but was incorrect.