Quantitative benefit of centralised quality assurance for the implementation of Lung SABR

ABSTRACT

Purpose or Objective SABR has become standard of care for early stage lung cancer where surgery is contraindicated. As a result of the COVID- 19 pandemic access to surgery was limited and demand for SABR as primary treatment has increased. A national program to implement lung SABR in all radiotherapy centres was commissioned and an associated QA program was developed. One of its components was a planning benchmark case to ensure optimal planning of target volumes whilst sparing organs at risk following the SABR Consortium Guidelines. Results of the benefits of the QA process are presented here. Materials and Methods A dual-lesion planning benchmark DICOM dataset was circulated amongst 24 participating centres, including a planning CT and a structure set. Centres had to plan the lesions to 55Gy in 5# and meet dose constraints, coverage and conformity criteria outlined in the Guidelines. All plans were reviewed on Velocity v4.1 (Varian Medical Systems) and PTV coverage, dose distribution, plan conformity and OAR dose constraints were assessed. Prescription Dose Spillage (PDS) was used to define conformity in the high dose area as Body V100(cc)/PTV V100(cc) and Modified Gradient Index (MGI) was defined for conformity in low dose area as Body V50(cc)/PTV V100(cc). The Mann Whitney test was used to evaluate differences in conformity across plans, with statistical significance set at 5%. Results Mean V100% for first submissions for lesions 1 and 2 were 97.26% (S.D. 1.86) and 98.19 % (S.D.1.61), respectively. All mandatory OARs were well within tolerance. The largest variation across centres was plan conformity, which is summarised in Table 1. (Table Presented) Ten plans failed their first attempt and centres were asked to resubmit following detailed feedback. Mean PDS for these plans changed from 1.19 (S.D. 0.09) to 1.13 (S.D. 0.05), although this was not statistically significant (p=0.12). Mean MGI was significantly improved on resubmission, decreasing from 7.08 (S.D. 0.8) to 6.16 (S.D.0.84), (p=0.03). Figure 1 shows increase in consistency and improvement in conformity across centres after resubmission. On completion of the QA process, the final set of accepted plans had improved conformity indices from initial PDS and MGI, however these were not statistically significant (p=0.31 and p=0.13, respectively). (Figure Presented) Conclusion A national QA program for lung SABR is critical for the safe implementation of this technique and to ensure standards are consistently high across multiple centres. The planning benchmark has highlighted differences in plan conformity and technique, in particular for MGI, however feedback within the QA process has allowed for increased consistency across departments through improved quality.