ESTRO 2021 Abstract Book

S553

ESTRO 2021

United Kingdom

Abstract not received

SP-0713 Proving the clinical benefit of 15 years of IGRT: modelling of toxicity for prostate radiotherapy C. Fiorino 1 1 San Raffaele Institute, Medical Physics, Milano, Italy Abstract Text The gradual implementation of IGRT for prostate cancer in the last two decades translated into the possibility of comfortably delivering higher doses, mostly due to margin reduction. The process is still ongoing toward widespread use of daily IGRT, prostate tracking and MRIgRT, making also easier the implementation of SBRT. IGRT also clearly improved the delivery of pelvic node irradiation (PNI) in high-risk patients (pts). Despite the delivery of more aggressive RT in terms of higher dose/daily dose and/or larger volumes (with PNI), IGRT indubitably contributed to reduce (at least, not increase) most moderate/severe toxicities (tox). In general, the delivery of more precise treatments with less volumes of adjacent OARs included in the high-dose region also helped to better model dose/dose-volume effects and to enhance the impact of non-dosimetry factors, incorporated into multi-variable NTCP models. On the other hand, their level of evidence/validation is variable and, in some cases, still limited. Predictive models of rectal tox (mostly referred to bleeding and fecal incontinence) have robust evidence. IGRT translated into less rectal tox, also in the extreme hypo- fractionation case, and into identifying few clinical factors (such as previous surgery, diabetes and cardio- vascular pathologies) as strong modulators of the risk. Instead, the evidence of IGRT in reducing GU tox is still weak and limited to few end-points (such as hematuria), showing some volume effect. Instead, for most of the late impairing symptoms, the prevalent dose-effect seems to limit the benefit of IGRT, due to the unavoidable inclusion of bladder neck and urethra in the high-dose region. However, the identification of more sensitive bladder regions and the strong impact of several clinical factors is expected to translate into new approaches to better tailor plan optimization (and IGRT) to the individual, with the aim to reduce GU tox. Despite the limited evidence, IGRT was associated with a reduction of erectile dysfunctions, likely due to the “incidental” avoidance of penile bulb and other structures involved in erection: this is more evident in pts with MRI-based planning and using prostate tracking. Finally, looking to bowel tox in PNI, IGRT reduced the bowel irradiated at high dose while IMRT drastically reduced toxicity compared to 3DCRT, especially in the acute phase. However, NTCP models of bowel symptoms remain lacking: importantly, selected moderate bowel/GI symptoms (for instance, urgency) were reported to heavily impact QoL: non-dosimetry factors related to individual bowel sensitivity are highly involved. Bowel motion is also a challenging issue, largely influencing model’s reliability: more research is warranted, including innovative “motion-inclusive” models, IGRT and plan strategies aimed to spare “stable” bowel portions receiving high dose during PNI in pts at higher risk, such as the one with worse baseline bowel functionality. Abstract Text In the FLAME trial, we demonstrated that focal boosting of intra-prostatic lesions improves biochemical disease-free survival without increasing toxicity compared to the standard treatment. In the subsequent hypoFLAME trial the same concept was translated to an ultra-hypofractionated schedule of 5 fractions of 7 Gy, with an integrated boost to the intra-prostatic lesion up to 10 Gy. Toxicity levels in this trial were in line with those reported from studies testing ultra-hypofractionated radiotherapy without focal boosting. In both the FLAME and hypoFLAME trial, position verification was based on fiducial markers, implanted in the prostate. An on-line position verification and correction procedure was applied in all participating centers. Although various measures could be taken to minimize intra-fraction motion, there was no real-time monitoring or correction. PTV margins were 4-5 mm posteriorly and 5-8 mm in other directions in both trials. Before starting inclusion in the FLAME trial, it was demonstrated that the impact of intra-fraction motion is quite limited in standard fractionated radiotherapy schedules. As the direction of intra-fraction motion is largely random, the systematic error S over 35 fractions was found 0.6 mm. However, in ultra- hypofractionated schedules, this reasoning doesn’t hold as the duration of each fraction is longer and less averaging happens in 5 fractions. In both the FLAME and hypoFLAME trials, extreme care was taken to avoid the risk of toxicity. During treatment planning the dose constraints to organs at risk took priority over the boost dose. In the hypoFLAME trial in extreme cases where intra-fraction motion, observed on a CBCT acquired at the end of the fraction, shifted the focal boost towards the rectal wall in more than one fraction, no further boosting would be done in the final fractions. Thus, while nominally a boost dose of 95 Gy was prescribed in the FLAME trial, the median D98% to the GTV was only 84.7 Gy. For the hypoFLAME trial the boost dose of 50 Gy in practice resulted in a median dose of 40.3 Gy. Now that the benefit of focal boosting has been demonstrated, the route towards improved outcome is clear. The challenge has become a technical one: how to reach the high boost dose safely. The recent improvements in radiotherapy techniques offer a solution. Here modern (MR) image guidance and on-line adaptive treatment techniques may improve the capacity to boost the tumor by external beam radiotherapy without increasing the dose to the surrounding organs at risk. SP-0714 Focal boosting in prostate cancer: The FLAME/hypoFLAME experience U. van der Heide 1 1 the Netherlands Cancer Institute, Radiation Oncology, Amsterdam, The Netherlands

SP-0715 IGRT for head and neck M. Tomsej 1 1 CHU Charleroi, Radiation Medical Physics Department, Charleroi, Belgium