ESTRO 2024 - Abstract Book

S1346

Clinical - Head & neck

ESTRO 2024

Results:

The bootstrap estimate of the control limit was determined to be 1.015 with a standard error of 8.99 × 10 -8 between the 5000 bootstrap samples. Six patients were identified as OC (Figure 1). Patients 6, 59, 73, and 76 received lower brainstem doses, while Patients 29 and 51 received higher brainstem doses (Figure 2). Consequently, the treatment plans of Patients 29 and 51 were reoptimized with the aim of reducing the brainstem dose without compromising on coverage to all the planning target volumes (PTVs) and increasing dose to other OARs. Through reoptimization, Patient 29’s brainstem D 2 , D 20 , D 40 , D 60 , D 80 , and D 98 decreased by 2.12 Gy, 2.50 Gy, 2.84 Gy, 3.02 Gy, and 3.24 Gy, respectively. After reoptimizing Patient 51’s plan, the brainstem D 2 , D 20 , D 40 , D 60 , D 80 , and D 98 decreased by 1.17 Gy, 0.79 Gy, 1.93 Gy, 2.35 Gy, and 4.17 Gy, respectively.

Figure 1. Patients detected as OC by the SVDD-based control chart.

Figure 2. Brainstem DVHs of the original (solid curve) and reoptimized (dashed curve) patients.

Conclusion:

The SVDD-based control chart identifies patients receiving extreme OAR doses, prompting treatment plan reoptimization for those receiving higher OAR doses. These results demonstrate that OC plans can be reoptimized