ESTRO 38 Abstract book

S321 ESTRO 38

OC-0609 Urethra-sparing SBRT for prostate cancer: quality assurance of a randomized phase II trial M. Jaccard 1 , T. Zilli 1 , A. Dubouloz 1 , L. Tsang 2 , S. Zvi 2 , N. Linthout 3 , S. Bral 3 , W. Verbakel 4 , A. Bruynzeel 4 , M. Björkqvist 5 , H. Minn 5 , L. Escude 6 , S. Jorcano 6 , J. Lencart 7 , A. Oliveira 7 , Z. Ozen 8 , U. Abacioglu 8 , J.M. Pérez-Moreno 9 , C. Rubio 9 , M. Rouzaud 1 , R. Miralbell 1 1 Geneva University Hospital, Radiation Oncology, Geneva, Switzerland ; 2 Sheba Medical Center, Radiation Oncology, Ramat Gan, Israel; 3 Onze-Lieve- Vrouwziekenhuis, Radiation Oncology, Aalst, Belgium; 4 VU University Medical Center, Radiation Oncology, Amsterdam, The Netherlands; 5 University Hospital Turku, Radiation Oncology, Turku, Finland ; 6 Teknon Oncologic Institute, Radiation Oncology, Barcelona, Spain ; 7 Portuguese Institut of Oncology, Radiation Oncology, Porto, Portugal ; 8 Neolife Medical Center, Radiation Oncology, Istanbul, Turkey ; 9 Hospital Universitario Sanchinarro, Radiation Oncology, Madrid, Spain Purpose or Objective To present the radiotherapy quality assurance (RTQA) results from a prospective multicenter phase II randomized trial of short vs. protracted urethra-sparing stereotactic body radiotherapy (SBRT) for localized From 08/2012 through 12/2015, 165 patients (pts) from nine institutions with localized PCa were randomized and treated according to two different overall treatment time schedules: either 9 days (arm A, 82 pts), or 28 days, once- a-week, the same week-day (arm B, 83 pts). The prescribed dose was 36.25 Gy in 5 fractions of 7.25 Gy to the prostate planning target volume (PTV) with ( n =92) or without ( n =73) inclusion of the seminal vesicles (SV) (5- mm isotropic expansion, with a 3-mm posterior margin). The urethra planning risk volume (uPRV=prostatic urethra+3-mm margin) was planned to receive 32.5 Gy. Patients were treated either with a volumetric modulated arc therapy (VMAT) ( n =112, 72 with SV) or intensity modulated RT (IMRT) ( n =53, 20 with SV) technique under stereotactic conditions using Novalis linacs and ExacTrac image-guided technology. Dosimetric results, PTV homogeneity index (HI), Dice similarity coefficient (DSC), number of monitor units (MU), as well as deviations from protocol dose constraints were retrospectively analyzed and compared between pts treated with a VMAT vs . IMRT technique. Results Major protocol deviations occurred for 52 pts (32%), mostly for uPRV (12%) and PTV (5%) coverage, and for the rectal (RW) and bladder wall (BW) V 90% (7%, each) (Table 1), while minor deviations occurred for 67 pts (41%). For PTV and uPRV coverage, deviations were more frequent in the IMRT group ( p =0.001 and p <0.0001, resp.). As compared to IMRT, treatment planning with VMAT resulted in a slightly better uPRV coverage (D 98% =31.1 vs. 30.6 Gy, p<0.0001), less MU (2335 vs . 3454, p <0.0001), and better HI (0.09 vs. 0.11, p<0.0001) and DSC (0.85 vs. 0.75, p <0.0001) values. Moreover, VMAT yielded better dosimetric parameters for the BW V 50% (26.1% vs. 33.3%, p =0.0001) and the penile bulb (PB) D mean (30.7 vs. 31.1 Gy, p =0.008), while other dosimetric parameters were similar. The inclusion of the SV in the PTV impacted negatively on the RW V 90% (9.1% without vs . 10.4% with, p =0.0003) and RW V 80% (13.2% without vs . 15.7% with, p =0.0003). The PTV volume was found to mildly linearly correlate with the BW V 50% ( r =0.523, p <0.0001) and the femoral heads (FH) D 2% ( r =0.572, p <0.0001). prostate cancer (PCa). Material and Methods

analysed for dosimetric and positional accuracy: up to 5% global scaling was allowed, followed by gamma analysis at 2%/2mm/10% threshold. Positional accuracy at target- cord interface and target-cord dose gradient (80%-20% width) was required to be within 1 mm of planned. Results A total of 18 centres have been credentialed; eight centres had previous site visits and submitted film and 10 centres had an on-site SABR process review and measurement. Delivery devices included linear accelerator, TomoTherapy and Cyberkife systems. Five centres were not able to achieve 90% gamma passing rate (Figure 1). Of these, three failed in low (< 5 Gy) dose regions and were > 80% passing rate thus were deemed acceptable. Two failed over the full dose range; one resubmitted a new plan, with the original fail attributed to inappropriate MLC constraints. The remaining centre elected not to have spine remeasured. One centre had positional discrepancy greater than 1 mm, and all centres were able to deliver the planned dose fall off from 80% of prescription dose to 20% of prescription dose within 1 mm at the target-cord interface. Figure 2 shows consistency in the dose gradient between planned and measured dose at the target-cord interface.

Figure 1: Gamma passing rate at 2%/2mm for all centres

Figure 2: Dose gradient between the 80% and 20% of prescription dose at the target-cord interface Conclusion Credentialing measurements for spine SABR in a multi- centre trial have shown the majority of centres, independent of treatment modality can deliver spine SABR according to the treatment plan. The measurement process detected one centre with inappropriate MLC constraint setting that was rectified upon a repeat visit.