Session Item

Radiation-induced brainstem/spinal injury is of great concern in the treatment of posterior fossa (PF) tumors. However, there are currently no recommendations on treatment planning ‘best-practice’ for pencil beam scanning (PBS) proton therapy of pediatric PF tumors, even though it is the most common tumor location in these patients. This study therefore aimed at understanding the current PBS standard of care across European proton centers for the treatment of pediatric PF tumors, with special considerations for brainstem and spinal cord management.

Anonymized data (CT-scan, structure set) for two pediatric cases (an ependymoma (EP) and an atypical teratoid rhabdoid tumor (ATRT)) were distributed to nineteen European proton therapy centers treating pediatric patients. The EP tumor was wrapped around the brainstem, with overlap from the pons and caudally thereafter. The ATRT case was located in the caudal part of the PF, with full overlap with the brainstem at the medulla oblongata level.

The centers were asked to plan based on their own clinical practice, i.e. no instructions were given on the treatment planning process (e.g. planning technique, number of beams, dose constraints, robustness evaluation parameters etc.). For both cases the prescription dose (for RBE=1.1) was 54 Gy in 30 fractions to the clinical target volume (CTV), with an additional boost CTV2 to 59.4 Gy for the EP case. For all plans, dose-volume histograms were calculated and relevant metrics compared, as well as general planning parameters.

To date, 14 EP (of which 2 were simultaneously integrated boost plans) and 13 ATRT plans were received (Fig. 1). For the EP, thirteen of the plans used three beams and one had two beams. Nine plans were coplanar, and four used range shifters. All nominal plans achieved CTV2 coverage with a V95% > 95% (median 98.8%). The spread in organs at risk dose was large, especially for the brainstem where V54Gy ranged from 26% to 61% (median 46%), and D1% from 57.1 Gy to 61.1 Gy (median 57.8 Gy). Spinal cord V54Gy was 0% for most centers, while D1% ranged from 46.1 Gy to 59.3 Gy (median 53.2 Gy) (Fig. 2).

For the ATRT case, ten centers used three beams, two centers had two beams and one center had four. Seven plans were coplanar, and four used range shifters. All nominal plans achieved good CTV coverage (V95% > 99%), and spread in organs at risk doses was more limited with brainstem D1% ranging from 52.8 to 55.9 Gy (median 54.4 Gy) and V54Gy ranging from 0% to 25% (median 4%). Spinal cord D1% ranged from 51 to 54.7 Gy (median 52.8 Gy), with V54Gy being 0% for almost all centers.

Considerable quantitative differences were seen in clinical practice across European proton therapy centers for the treatment of pediatric PF tumors. While most centers used three coplanar beams, the balance between target vs. spinal cord/brainstem dose was largely center-dependent. This was especially seen for the EP case, which had a higher prescription dose.