ESTRO 2024 - Abstract Book

S4848

Physics - Quality assurance and auditing

ESTRO 2024

Impact of dose reporting modes on dosimetric accuracy in a national end-to-end spine SBRT audit

Rushil Patel 1 , Mohammed Hussein 2 , Jonny Lee 3 , Patricia Diez 1 , Ileana Silvestre Patallo 2 , Alexandros Douralis 2 , Catharine H Clark 2,4,5 1 Mount Vernon Cancer Centre, National Radiotherapy Trials Quality Assurance Group, Northwood, United Kingdom. 2 National Phyics Laboratory, Radiotherapy, Teddington, United Kingdom. 3 Dorset Cancer Centre, Radiotherapy, Poole, United Kingdom. 4 University College London Hospital, National Radiotherapy Trials Quality Assurance Group, London, United Kingdom. 5 University College London Hospital, Radiotherapy, London, United Kingdom

Purpose/Objective:

The delivery of spine SBRT is demanding of both the TPS and treatment delivery machines. The proximity of the spinal cord to the PTV necessitates a very steep dose gradient at the boundary with an acceptance of lower dose conformity elsewhere. These sharp gradients pose a challenge to accurate dose measurement and delivery. To ensure a high degree of geometric and dosimetric accuracy is achieved, an end-to-end audit was delivered as part of a national commissioning program of SBRT services. Additional challenges lie in centres choosing whether to report in dose to water or dose to medium, particularly in denser media such as bone [1].

Material/Methods:

The audit was initially developed as an onsite visit but, due to Covid restrictions, it was converted to a remote audit. An anthropomorphic phantom, made of materials suitable for MV dosimetry, was scanned, planned and treated at participating centres according to their local protocols, including on-treatment imaging for accurate positioning of the phantom, and following national guidelines. Dose measurements were made in the vertebral PTV using 3 alanine pellets, whilst dose distribution data were acquired using EBT3 Gafchromic film. Alanine and film doses were compared with the corresponding treatment planning system (TPS) calculated doses. All plans were also recalculated using all available algorithms and reporting modes: Dose to water (Dw,w), Dose to water in a medium (Dw,m) and Dose to medium in a medium (Dm,m). Film was analysed using VIGO, a custom MATLAB programme developed for national auditing. Gamma analysis using 5%/1mm and 3%/2mm global criteria and mean distance to agreement (DTA) were calculated. Kruskal Wallis tests were used to test statistical differences between calculated and measured doses and between dose reporting modes.

Results:

A total of 60 plans from 45 TPSs at 42 centres were analysed. Audited TPSs included Eclipse (22), Pinnacle (6), Precision (5), Raystation (8) and Monaco (4). Dose reporting modes were 39 Dw,w, 9 Dw,m and 12 Dm,m.

The mean dose deviation measured with alanine was 1.3% [-6.5 to 8.5%] for all algorithms. Ninety percent of measurements were within ±5% of the expected dose, with 75% within ±3%. When stratified by dose reporting mode, the mean dose deviations were 2.5% (Dm,m), -0.6% (Dw,m) and 1.3% (Dw,w); see Figure 1. One centre recorded a negative dose deviation > 3% (for both Dm,m and Dw,m). The dose difference between the alanine measurement and the expected dose, grouped by dose reporting mode, was statistically significant (chi-squared = 7.02, p = 0.029).