ESTRO 36 Abstract Book

S913 ESTRO 36 2017 _______________________________________________________________________________________________

individualized PTV margins around the SV based on MRI information. Material and Methods We have obtained CT, T2 weighted (T2w) MRI, dynamic contrast enhanced (DCE) MRI and diffusion weighted imaging (DWI) MRI for 21 high-risk prostate cancer patients with an intact prostate. All patients completed their radiotherapy treatment course of 78Gy in 39 fractions. Our clinical standard margins of 5mm, 5mm and 7mm in the anterior-posterior, left-right and cranio-caudal direction, respectively, were applied to the SV delineation for preparing the clinical plans. MRI scans were then further examined retrospectively. All three MRI sequences were delineated by a radiation oncologist and a suspicious volume on T2W confirmed by an overlap with one of the functional scans (DCR and/or DWI) was assumed to be macroscopically involved, see Figure 1. For patients with confirmed suspicious volume in the SV, we applied a PTV with a margin of 11mm (PTV11mm) to the clinical delineation of SV to account for the mobility of the SV when registering on implanted markers within the prostate as suggested by [2]. Coverage of the PTV used in the clinical dose plan (PTV5mm) and the PTV required for reliable SV coverage (PTV11mm) were extracted from the dose-volume histogram (DVH) of the clinical dose plan.

radiotherapy despite daily match on fiducials in the prostate. This is a particular concern as studies found a higher rate of failure with highly conformal radiotherapy when comparing to rectangular fields [3], yet uniform wide margins to SV will substantially increase the dose to the rectum. With the current study we used multi- sequence MRI to yield risk-adapted margins to increase coverage of SV for selected patients, while maintaining standard, tight margins for patients without suspicious MRI findings. [1] Morlacco A, et al. Eur Urol. http://dx.doi.org/10.1016/j.eururo.2016.08.015, 2016. [2] Boer J, et al,Int J Radiat Oncol Biol Phys, vol 86, No 1, pp 177-182, 2013. [3] Heemsbergen W, et al. Radiother Oncol, 107, 134-139, 2013. EP-1694 DW-MRI as a predictor of tumor response to hypofractionated stereotactic boost for prostate cancer D. Pasquier 1 , A. Hadj Henni 2 , E. Tresch 3 , N. Reynaert 4 , E. Lartigau 1 , O. Colot 2 , N. Betrouni 5 1 CENTRE OSCAR LAMBRET- CRISTAL UMR CNRS 9189, Academic Department of Radiation Oncology, Lille, France 2 CRISTAL UMR CNRS 9189, LILLE University, Villeneuve d'Ascq, France 3 CENTRE OSCAR LAMBRET, Department of Biostatistics, Lille, France 4 CENTRE OSCAR LAMBRET, Department of Medical Physics, Lille, France 5 INSERM 1189 ONCO-THAI, LILLE University, Lille, France Purpose or Objective To evaluate the feasibility of diffusion-weighted MRI (DWI) as an early biomarker in patients receiving hypofractionated stereotactic boost for intermediate risk prostate cancer Material and Methods Patients with intermediate risk prostate cancer were included in a multicenter CKNO phase II trial. During the first part of the treatment, 23 fractions (2 Gy/session) were delivered over 42 days maximum using 3DCRT. During the second treatment part, hypofractionated stereotactic boost (3 fractions of 6 Gy) was delivered each other day. Median follow up was 26.4 months (range:13.6 - 29.9). Multiparametric (mp) MRI was realized before (M0) and at regular follow-up intervals after radiotherapy (6, 9, 12, 24, 30, 36 and 48 months). This ancillary study included the 24 patients treated in our center (3T MRI). None of them presented with recurrence so apparent diffusion coefficient (ADC) and ktrans variations were correlated to PSA kinetics. GTV was delineated by experienced radiologist and radiation oncologist on baseline and follow-up mpMRI. Mean tumor ADC values (b= 0, 1000 and 2000 s/mm2) were normalized to obturator muscle (nADC). Spearman’s coefficient correlation and non-parametric Wilcoxon Mann-Whitney tests were used for continuous and categorical variables respectively. Results GTV was visualized on MRI in 12 patients. Mean nADC improved from 1.14x10 -3 mm 2 /s (min 0.49 max 1.53) to 1.59x10 -3 mm 2 /s (min 1.15 max 1.94) between M0 and M24. The nADC variation between M0 and M12 was correlated with PSA at M18. The nADC increase between M0 and M12 was larger in patients with PSA < 1ng/mL at M18 (p=0.034) (Table). No other correlation was found. Ktrans variation was not associated with PSA kinetics.

Results For 14 out of 21 patients had a T2w suspicious volume confirmed either by DWI or DCE, with 6 cases having suspicious volumes in all three MRIs. The 14 patients had median (range) of dose to 98% of the PTV volume (D98%) of 73.1Gy (57.6-75.6) and 45.9Gy (39.3-61.7) for PTV5mm and PTV11mm, respectively. In four cases D98% of PTV5mm was below the clinical recommendation of 70.2Gy while all 14 patients had D98% of PTV11mm below 70.2Gy, see Figure 2.

Conclusion Coverage of the SV to the prescription dose is not guaranteed when using tight margins and highly conformal