ESTRO 2020 Abstract book

S408 ESTRO 2020

PD-0671 Matched pair analysis of salvage radiotherapy vs observation for post-prostatectomy PSA recurrence T. Wiegel 1 , D. Tilki 2 , F. Preisser 3 , R. Pompe 2 , F. Chun 3 , M. Graefen 4 , H. Huland 4 , A. Siegmann 5 , D. Böhmer 5 , V. Budach 5 , R. Thamm 1 , D. Bartkowiak 1 1 University Hospital Ulm, Department of Radiation Oncology, Ulm, Germany ; 2 University Hospital Hamburg- Eppendorf, Department of Urology, Hamburg, Germany ; 3 University Hospital Frankfurt, Department of Urology, Frankfurt, Germany ; 4 University Hospital Hamburg Eppendorf, Martini-Klinik Prostate Cancer Center, Hamburg, Germany ; 5 Charité University Hospital Berlin, Department of Radiation Oncology, Berlin, Germany Purpose or Objective Salvage radiotherapy (SRT) may lead to better oncologic outcomes in prostate cancer (PCa) patients who develop biochemical recurrence (BCR) after radical prostatectomy (RP). However, the benefit of SRT is still debatable. We compare long-term oncologic outcomes of SRT vs. no radiotherapy (noRT) in patients with BCR after RP. Material and Methods Within a two-institutional high-volume center database we identified patients with BCR after RP between 1989 and 2016 for PCa. Only patients without lymph node invasion were included. Patients with adjuvant radiotherapy were excluded. SRT was delivered to the prostatic bed in all cases. Propensity score matching (PSM) was performed to account for differences in pathologic tumor characteristics. Kaplan-Meier analyses and Cox regression models tested the effect of SRT vs. noRT on metastasis- free (MFS) and overall survival (OS) after PSM. Results Of 1,832 patients with BCR, 32.9% (n=603) received SRT. Patients managed with SRT more frequently had positive margins (41.6 vs. 24.8%, p<0.001), a higher median pre-RP PSA (9.5 vs. 8.4 ng/ml, p<0.001), more frequently had organ-confined disease (53.7 vs. 45.3, p<0.001) and Gleason Grade Group 4 (12.9 vs. 1.9%, p<0.001). Median SRT dose was 70.2 Gy, median PSA at SRT was 0.3 ng/ml and median time to RP was 25 months. After 1:1 PSM, at 15-years after RP (median follow-up 96 months), MFS and OS rates were 85.9 vs. 81.3% (p<0.001) and 86.4 vs. 68.9% (p<0.001) for SRT and noRT, respectively. In multivariable Cox regression models, SRT was an independent predictor for metastasis (hazard ratio: 0.37, 95%-confidence interval 0.25-0.56, p<0.001) and OS (hazard ratio: 0.52, 95%- confidence interval 0.34-0.78, p<0.01). Conclusion To our best knowledge, this is the first matched-pair analysis investigating the impact of SRT vs. observation in post-RP recurrent PCa. After compensating for established risk factors in our large cohort of patients, SRT was associated with better long-term MFS and OS in patients with BCR after RP. These results on clinical endpoints underline the curative potential of SRT which should therefore be offered to all patients with BCR after RP.

1 University Medical Center Utrecht, Department of Radiotherapy, Utrecht, The Netherlands

Purpose or Objective Limited data is available on the possible gain of using a vacuum mattress for immobilization during SBRT for pelvic or para-aortic lymph nodes. Furthermore, no data is available on the impact of longer session times with online MR-guided treatments (≈30 min. intra-fraction time from pre- to post-SBRT scan). In this pilot study, we investigated the effect of a vacuum mattress on intra-fraction motion for MR-guided lymph node SBRT. Material and Methods 40 patients with pelvic or para-aortic lymph node oligometastases received 5x7Gy on a 1.5 T MR-linac (Elekta AB, Stockholm, Sweden). 20 patients were treated with a BlueBAG vacuum mattress (Elekta AB) and 20 without. Intra-fraction motion of bony anatomy was investigated by rigid registration of pre-, position verification (PV) and post-treatment MRI scans using the ITK-based toolkit Elastix [1]. GTV motion was assessed by calculating the distance in center of gravity for the manual GTV delineations on these scans. The GTV intra-fraction motion data was used to calculate the difference in PTV margins, based on [2]. Translations were derived in left-right (LR), anterior-posterior (AP) and cranio-caudal (CC) directions, for bony anatomy also rotations around the LR, AP and CC axes. Means and absolute means were used per patient over 5 fractions. Statistics: Mann-Whitney U-test. Results The mean absolute translations of bony anatomy were only significantly different in AP direction: for patients without vacuum mattress, larger absolute translations were found in AP direction (median 0.4 mm with mattress (range 0.0 – 2.1) vs. 0.7 mm without mattress (0.0 – 5.7), p = 0.005). No significant differences were found for the GTV mean absolute translations. Mean, systematic and random components of bony anatomy intra-fraction motion are displayed in the table. With a vacuum mattress, the derived PTV margins were 0.4 and 0.8 mm smaller in LR and AP directions but comparable in CC direction (based on 30 min. time interval, taking 50% of the GTV motion into account as in [2]). For 15 min. (PV scan) the reductions were 0.3 and 0.6 mm for LR and AP directions. For both groups, a systematic intra-fraction motion of both GTV and bony anatomy was found towards posterior: bony anatomy mean 0.4 mm with mattress (SD 0.4) and 0.9 mm without mattress (SD 1.0).

Poster discussion: RTT 2

PD-0672 Effect of a vacuum mattress on bone & target intra-fraction motion during MR-guided lymph node SBRT A.M. Werensteijn-honingh 1 , P.S. Kroon 1 , D. Winkel 1 , C.G. Gadellaa-Van Hooijdonk 1 , G.G. Sikkes 1 , N.G.P.M. Vissers 1 , W.S.C. Eppinga 1 , M. Intven 1 , B.W. Raaymakers 1 , I.M. Jürgenliemk-Schulz 1