ESTRO 38 Abstract book

S818 ESTRO 38

Indications for adaptation included OAR constraint violation (n=32/38 fractions) or DE alone (n=6). 100% of OAR violations were reversed through adaptation. In 16/32 cases where OAR violation reversal was required, target coverage was simultaneously increased. RECIST control of the treated lesion was 100% at 3 months in evaluable patients; one patient did not have imaging at the required time point but had progression on imaging at 4.5 months post-treatment, consistent with 94% short-term lesion control. One treatment-related acute grade 3+ toxicity was observed (duodenal ulcer) with 100% of patients reporting. Conclusion SMART for patients with ovarian oligometastases is feasible, allowing for dose escalation and/or OAR sparing with low toxicity and promising efficacy. Daily imaging revealed motion primarily of abdominal and pelvic OARs that required frequent re-planning to avoid constraint violations. A phase II trial is planned to further evaluate SMART in this patient subset. EP-1511 Investigation of prognostic factors of cervical squamous cell carcinoma using pre-treatment MRI C. Tonoiso 1 , A. Haga 2 , A. Kubo 1 , T. Kawanaka 1 , S. Furutani 1 , H. Ikushima 3 , M. Harada 1 1 Tokushima University, Radiology, Tokushima, Japan ; 2 Tokushima University, Medical Image Informatics, Tokushima, Japan ; 3 Tokushima University, Therapeutic Radiology, Tokushima, Japan Purpose or Objective Diagnosis and prognostic prediction using features extracted from images have been reported recently. We analyzed MRI images before treatment to investigate prognostic factors of cervical cancer. Material and Methods 100 consecutive patients with cervical squamous cell carcinoma who underwent chemoradiotherapy at Tokushima University Hospital from January 2005 to November 2013 were included in this study. The median age was 59 years old (range 30-79). The FIGO classification was I / II / III / IV in 16/42/34/8 cases, respectively. External irradiation and intracavitary brachytherapy was performed in all cases. The pretreatment T2WI images was transferred to the treatment planning device, and one radiologist defined gross tumor volume (GTV) for all cases. Then the MRI images was standardized with a minimum value of 0 and a maximum value of 1, 3-dimensional wavelet transformation was performed and 476 features were extracted for GTV region. Univariate analysis of prognostic factors was performed by Log-rank test using clinical information and features, and multivariate analysis was performed using Cox proportional hazard The median observation period was 55 months (range 2- 123), and the 5-year survival rate of all cases was 77.7%. In the univariate analysis, significant differences were observed in the performance state, M factor, 7 features of GTV, and one feature of GTV was a significant factor in multivariate analysis. Conclusion Analysis using pretreatment MRI may be useful for prognostic prediction of cervical squamous cell carcinoma. model. Results

1 Kasr Al-Ainy Hospitals- Cairo University, Clinical Oncology Dept., Cairo, Egypt ; 2 United Lincolnshire Hospitals- NHS Trust, Department of Dosimetry & Imaging- Radiotherapy physics, Glascow, United Kingdom ; 3 Kasr Al-Ainy Hospitals-cairo University, Diagnostic & Intervention Radiology Dept., Cairo, Egypt ; 4 Kasr Al- Ainy Hospitals-Cairo University, Clinical oncology Dept., Cairo, Egypt Purpose or Objective The study aims to investigate sources of errors in radiotherapy that cause uncertainty in the treatment delivery. Measurement of systemic & random errors introduced by the online marker match, the offline marker match & the offline bone match verification and consequently estimating the PTV needed if the verification was done guided by: (i) The fiducial markers match (PTV marker ), (ii) The bone structure match (PTV bone ) & (iii) The laser markers on the skin only (PTV laser) . Material and Methods Thirty localized prostate cancer patients treated with dose escalated VMAT giving 78 Gy/35# to the whole prostate & 87Gy boost to the focal lesion. Three fiducial markers were inserted US guided; two at the base & the 3rd at the apex. For each patient, two daily MV images were acquired; 1 st pre-online marker match was utilized for offline bone match, while the 2 nd post-online marker match used for offline marker match as well. Systematic & random errors were calculated for each verification method to estimate the PTV needed. To composite the PTV marker , the online & offline marker matches were compared to check the uncertainty of the marker match, while CT scans were repeated weekly to estimate the marker migration. The PTV marker was then added to the evaluated PTV from the online marker match to obtain the PTV laser . The offline bone match was used to estimate the PTV due to skin marks migration (PTV skin marks migration ), which was then deducted from the PTV laser to acquire the PTV bone . Results The comparison of the systemic & random errors for the online & offline marker match showed no major difference in the lateral & vertical direction (<0.5 mm [p>0.38]), while it was significant with a maximum of 3 mm [p=0] in the longitudinal direction. The outcome of the CT study revealed a mean of the maximum marker migration of 2 mm ±1.2, this was combined with the mean difference between the offline & the online marker match to produce the PTV markers of 2.7±0.6, 3.34±1.1 & 4.44±2.2 mm in lateral, longitudinal & vertical directions respectively. The resulted PTV margin from the systematic & random errors data of the online marker match was added to the PTV markers to yield the PTV laser of 13±0.6, 17.7±1.1 & 15.8±2.2 mm in the same directions respectively. On the other hand, the systemic & random errors equations for the offline bone match shifts revealed PTV skin marks migration of 7.1, 9.1 & 8.6 mm , which implies PTV bone of 5.9±0.6, 8.6±1.1 & 7.2±2.2 mm in the same directions respectively. Conclusion It was clarified that the PTV needed for fiducial marker image guided verification in dose escalated VMAT to the prostate ranged between 3 mm and 4 mm in the lateral, longitudinal & vertical directions; which is about 10 mm smaller than that is guided by the laser skin marks only and tighter by about 5 mm relative to that is based on the bone structure image matching only. EP-1513 Patient-reported adverse events following trans-rectal ultrasound-guided prostate marker insertion T. Rosewall 1 , A. Bayley 1 , P. Chung 1 , C. Catton 1 1 Princess Margaret Cancer Centre, Radiation Medicine Program, Toronto, Canada

Electronic Poster: Clinical track: Prostate

EP-1512 Evaluation of target volume margins in prostate dose escalated VMAT by fiducial markers’ technique R. Fawzy 1 , R. Abdel-Malek 1 , M. Metwaly 2 , O. Abdel Aziz 3 , S. Alsirafy 4 , A. Seleem 4