ESTRO 2021 Abstract Book

S710

ESTRO 2021

MRI and the target volumes (TV) with surrounding organs-at-risk (OAR) were recontoured. TV and OAR doses were then predicted by applying the baseline radiotherapy plan to the anatomy of the day. On-table plan adaption was performed at the treating physician’s discretion. Results 6 patients presented with early-stage NSCLC, 1 patient with local-relapsed NSCLC and 14 patients with pulmonary oligometastases. 10 tumors were located peripheral, 2 central (according to the RTOG 0813 trial) and 11 ultracentral (PTV contact to central airways, esophagus or pulmonary artery). In total, 165 fractions were delivered. Plan adaption was performed in 154 fractions (93.3%), mostly due to planning target violations in the predicted plans (144 fractions, 87.3%). The majority of all plan violations (total N predicted = 187) occurred due to insufficient coverage of the PTV (N = 136) or hotspots inside the PTV (N = 33). Moreover, several overdoses inside OAR (N = 18) were detected. Plan adaption significantly reduced the amount of plan violations (total N adaption = 48, p < 0.001, Chi-squared test). Residual violations in the adapted plans mainly consisted of compromised PTV coverage to keep OAR constraints (N = 38). Figure 1 shows the percentage of violated planning targets for the following categories: planning target volume (PTV) coverage (cover), PTV overdose (over) and organ-at-risk (OAR) overdose.

After a median follow-up of 5.4 months (95%-CI 3.3 – 9.8 months), clinical outcomes were excellent with no case of local relapse and only one case of CTCAE grade ≥ 3° toxicity (grade 3 esophageal ulcer). Mean total treatment duration for adapted fractions was 55.8 ± 12.8 min, with detailed description of procedure times given in Figure 2 .

Only one patient had to cancel online-adapted treatment after 7/10 fractions due to increasing bone pain when lying on the couch.

Conclusion Online adapted MR-guided SBRT of lung tumors is clinically feasible and could avoid inadequate dose application to TV and OAR, but requires high treatment times and personnel expenditure. Hence, future clinical research should aim to refine patient selection. PD-0874 Impact of dose and prescription method in SBRT of NSCLC. A pooled analysis of 2 prospective trials. E. Gkika 1 , S. Adebahr 1 , G. Radicioni 1 , J. Exner 1 , E. Haehl 2 , S. Spohn 1 , A. Rühle 1,1 , C. Zamboglou 3 , N. Nicolay 1 , T. Sprave 1 , R. Wiehle 4 , D. Baltas 5 , U. Nestle 1 , A. Grosu 1