Thermoradiotherapy Optimization Strategies Accounting for Hyperthermia Delivery Uncertainties

The combined effect of hyperthermia and radiation therapy can be quantified by an enhanced equivalent radiation dose (EQDRT). Uncertainties in hyperthermia treatment planning and adjustments during treatment can impact achieved EQDRT. We developed and compared strategies for EQDRT optimization of radiation therapy plans, focusing on robustness against common adjustments. Using Plan2Heat, we computed preplanning hyperthermia plans and treatment adjustment scenarios for 3 cervical cancer patients. We imported these scenarios into RayStation 12A for optimization with 4 different strategies: (1) conventional radiation therapy optimization prescribing 46 Gy to the planning target volume (PTV), (2) nominal EQDRT optimization using the preplanning scenario, targeting uniform 58 Gy in the gross tumor volume (GTV), keeping organs at risk doses as in plan 1, (3) robust EQDRT optimization, as plan 2 but adding adjusted scenarios for optimization, and (4) library of plans (4 plans) with strategy 2 criteria but optimizing on 1 adjusted scenario per plan. We calculated for each radiation therapy plan EQDRT distributions for preplanning and adjusted scenarios, evaluating each combination of GTV coverage and homogeneity objectives. EQDRT95% increased from 49.9 to 50.9 Gy in strategy 1 to 56.1 to 57.4 Gy in strategy 2 with the preplanning scenario, improving homogeneity by ∼10%. Strategy 2 demonstrated the best overall robustness, with 62% of all GTV objectives within tolerance. Strategy 3 had a higher percentage of coverage objectives within tolerance than strategy 2 (68% vs 54%) but a lower percentage for uniformity (44% vs 71%). Strategy 4 showed a similar EQDRT95% and homogeneity for adjusted scenarios than strategy 2 for a preplanning scenario. D0.1% (radiation dose received by the 0.1% most irradiated volume) for organs at risk was increased by strategies 2 to 4 by up to ∼6 Gy. EQDRT optimization enhances EQDRT levels and uniformity compared with conventional optimization. Better overall robustness is achieved by optimizing the preplanning hyperthermia plan. Robust optimization improves coverage but reduces homogeneity. A library of plans ensures coverage and uniformity when dealing with adjusted hyperthermia scenarios.