Abstract Book

S839

ESTRO 37

EP-1555 Multiple re-irradiation for locally recurrent prostate cancer: proof of concept and clinical outcome S. Volpe 1,2 , B.A. Jereczek Fossa 1,2 , D. Zerini 1 , D.P. Rojas 1,2 , C. Fodor 1 , A. Vavassori 1 , P. Romanelli 1 , S. Vigorito 3 , E. Rondi 3 , S. Comi 3 , R. Cambria 3 , F. Cattani 3 , S. Di Cuonzo 1 , P. De Marco 3 , G. Beltramo 4 , G. Musi 5 , O. De Cobelli 5 , G. Marvaso 1 , R. Orecchia 1,6 1 European Institute of Oncology, Department of Radiation Oncology, Milan, Italy 2 University of Milan, Department of Oncology and Hemato-Oncology, Milan, Italy 3 European Institute of Oncology, Department of Medical Physics, Milan, Italy 4 CyberKnife Center CDI, Department of Radiation Oncology, Milan, Italy 5 European Institute of Oncology, Department of Urology, Milan, Italy 6 European Institute of Oncology, Scientific Directorate, Milan, Italy Purpose or Objective Although androgen deprivation therapy (ADT) is widely recognized as a mainstay in the treatment of recurrent prostate cancer (PCa), newer data suggest that focal approaches (e.g. radiotherapy, RT, or radiofrequency) may be considered in adequately selected patients. The aim of the current study is to present the technical feasibility and clinical outcomes of multiple re-RT for low-burden locally recurrent PCa on a series of patients (pts) treated at a tertiary care center. Material and Methods Retrospective analysis of an updated series of pts who received multiple (> 3 RT courses to prostate/prostate bed) re-RT with stereotactic image-guided technique and hypofractionated RT. Eight pts received three RT courses, of those two were candidate to subsequent forth RT following further local recurrence. No alternative focal approaches were considered. Two patients received surgery as a primary treatment modality; six were treated with RT±ADT. Local relapse was assessed by multiparametric magnetic resonance and/or choline positron emission tomography. The presence of local recurrence was confirmed histologically in 4 cases (in other cases with univocal PET, MRI and PSA findings all suggestive for recurrence, biopsy was not required). All pts had been evaluated for toxicity from previous RT per RTOG/EORTC. Biochemical control was assessed according to Phoenix definition. Dosimetric constraints were based on previously published institutional experience on PCa re-RT. Gross tumor volumes (GTVs) were limited to the site of relapse; planning target volume margins were achieved expanding the GTVs of 3 mm posteriorly, and of 5 mm in any other direction. Results Mean age at second re-EBRT was 68 (standard deviation, SD: 7.2) years. At diagnosis, four cases were classified as high risk PCa, three as intermediate and one as low per NCCN 2017. Median follow-up time from diagnosis was 168.5 (IQR: 144.2-203.1) months for the whole cohort. Median follow-up time from the third RT was 12 (IQR: 3.1-42.5) months; follow-up time from the forth RT course were 17 and 6 months. Biochemical progression free interval after first and second RT-course were 74 (interquartile range, IQR: 59.3-133.6) months and 33 (IQR: 20.8-53.1) months, respectively. Biochemical and radiological response was registered in all patients; at present, seven out of eight patients are free of disease. Overall toxicity profile was good; no severe acute or late genitourinary or gastrointestinal events were recorded.

Conclusion Multiple-course re-EBRT with high precision technology and image guidance can be proposed as a possible salvage therapy for locally recurrent, low-burden PCa recurrence in adequately selected patients. Some room for further dose escalation is suggested. Additionally, re-RT may be regarded as an option for avoiding/deferring ADT in carefully selected pts. Larger series and longer follow-up are warranted to assess the potential of multiple re-RT in the setting of local salvage therapies for PCa. EP-1556 Early deltaradiomics and PSA after radiotherapy for prostate cancer: a CKNOPRO trial ancillary study A. ESCANDE 1,2 , N. Betrouni 3 , E. Tresch 4 , B. Renard 5 , O. Colot 1 , N. Reynaert 6 , E.F. Lartigau 2,7 , D. Pasquier 1,2 1 Lille 1 University, CRIStAL laboratory- UMR 9189, Villeneuve d’Ascq, France 2 Oscar Lambret Comprehensive Cancer Center, Academic Department of Radiation Oncology, Lille, France 3 Lille 2 University, INSERM- U1171, Lille, France 4 Oscar Lambret Comprehensive Cancer Center, Department of biostatistics-, Lille, France 5 Oscar Lambret Comprehensive Cancer Center, Department of medical Imaging-, Lille, France 6 Oscar Lambret Comprehensive Cancer Center, 6Department of medical physics, Lille, France 7 Lille 2 University, Faculty H.Waremboug, Lille, France Purpose or Objective Intermediate-risk prostate cancer (PCa) is considered as a heterogenous population due to large variation of disease-free survival. Only scarce data are published about early variation of texture as biomarker of outcome for PCa whereas MRI is well-established as essential at the time of diagnosis. New textural analysis techniques currently available, such as radiomics, seem very promising as prediction tool in many cancers. Here, we investigated, early texture variation from pre-treatment and post-treatment MRI as early prognostic biomarker of prostatic specific antigen level (PSA) during the follow-up of intermediate risk PCa after hypofactionated radiotherapy. Material and Methods We examined clinical record from 48 patients treated for a histological proven intermediate risk PCa according to d’Amico risk classification within the frame of phase II CKNO PRO trial. Patients were exclusively treated by EBRT with intensity modulation without hormonotherapy for a total dose of 46 Gy (2Gy/fraction) followed by stereotactic boost to the prostate gland for a total dose of 18Gy (6Gy/fraction). MRI was performed before treatment (M0) and six months after the end of the treatment (M6). We used a homemade tool to extract 17 features (four first order features, 12 second order from grey-level co-occurrence matrix, and one from fractal dimension) from T2-weighted MRI. Delta-radiomics were calculated in percent variation from M0 to M6 on T2w MRI as follow. We investigated correlation between early textural variation and PSA (in ng/ml) at 12 (M12) and 18 months (M18) after the treatment. Test of nullity of the spearman correlation coefficient and Wilcoxon Mann- Whitney tests were used. Results Median follow-up was 18 months after the end of the treatment. Median prostatic specific antigen at M0, M12 and M18 were 8.2ng/ml (range from 2.91 – 22.40), 0.81 ng/ml (0.27 – 10.00) and 0.60 ng/ml (0.07 – 7.34), respectively. Delta-radiomics from variance, kurtosis and