ESTRO 36 Abstract Book

S167 ESTRO 36 2017 _______________________________________________________________________________________________

Conclusion MRI-guided SBRT for pancreatic cancer with individually evaluated margins is technical feasible and safe, with no treatment related grade ≥3 toxicity. New strategies are applied, including an individual corset to reduce breathing motion, MRI based contouring and simulation of motion- integrated dose distributions. PV-0322 Rapid Early Response of Gastroesophageal Junction Tumors During Real-time MRI-Guided Radiotherapy H. Musunuru 1 , S. Rosenberg 1 , J. Bayouth 1 , K. Mitteur 1 , M. Ritter 1 , B. Paliwal 1 , M. Witek 1 , A. Baschnagel 1 , N. Uboha 2 , S. Lubner 2 , N. Loconte 2 , P. Harari 1 , M. Bassetti 1 1 University of Wisconsin Hospital and Clinics, Radiation Oncology, madison, USA 2 University of Wisconsin Hospital and Clinics, Medical Oncology, Madison, USA Purpose or Objective Multimodality comprehensive therapy has become the standard of care for locally advanced esophageal and gastroesophageal junction tumors. Tumor response to chemoradiation correlates with outcomes, however full response information generally awaits esophagectomy. Intra treatment predictors of response may allow improved personalization of therapy. Daily MRI allows direct quantification of GEJ tumor size. The aim of this study is to evaluate volumetric changes in gross tumor volume (GTV) for gastroesophageal junction (GEJ) cancer patients undergoing MRI-guided radiation therapy, as part of neoadjuvant chemoradiotherapy Material and Methods Five GEJ adenocarcinoma patients underwent MRI during simulation and with each treatment fraction immediately prior to radiation delivery. The GTV primary was contoured on MRI scans at fractions 5, 10, 15, 20 and 23 and compared to the baseline GTV (Fig 1). Change in GTV across time was expressed as percentage difference (between baseline and different fractions and between individual fractions) and in terms of absolute volume(cc). Results Median age and follow-up period for this cohort were 68 years and 46.5 days. The treatment regimen consisted of weekly carboplatin (AUC 2mg/ml/min) and paclitaxel (50mg/m 2 ) with concurrent radiotherapy, 50.4Gy in 28 fractions in three patients and 41.4Gy in 23 fractions in the remaining two patients. The earliest decrease in GTV (% change) was noted at fraction 10 when compared to baseline (Mean -52%, SD 4.6%; Fig 2A). Evaluation of percentage change in the GTV between different fractions (i.e. fraction 5 and fraction 10 etc.) also showed that the earliest change occurred between fractions 5-10 (Table 1). Mean (SD) of GTV at baseline and at fractions 5, 10, 15, 20 and 23 were 94.7cc(15.4), 93.9cc(14.8), 46.3cc(7.8), 39.7cc(6), 33.2cc(4.7) and 33.2cc(4.7), respectively.

Mean(SD) of change in GTV expressed as percentage

Time Points

Baseline and different fractions Fractions 0 and 5 Fractions 0 and 10

-11.5%(7.8) -52.0%(4.6) -57.0%(6.0) -64.0%(9.0)

Fractions 0 and 15 Fractions 0 and 20

Fractions 0 and last fraction

-66.0%(.0)

Baseline

different

fractions Fractions 0 and 5 Fractions 5 and 10

-11.5%(7.8)

-48.0%(3.9) Fractions 10 and 15 -14.5%(5.5) Fractions 15 and 20 -16.0%(12.2) Fractions 20 and the last fraction -6.0%(5.9) Conclusion

Real-time MRI-guided radiation provides previously unavailable data on tumor response during neoadjuvant chemoradiation. In this study, the most significant volumetric change in the GTV was observed earlier than expected, between fractions 5 and 10. Correlation of early volumetric response changes with clinical and or pathological outcomes may prove highly valuable. Daily MRI during radiation provides a unique opportunity to tailor individual treatment based on early response to chemoradiation, and suggests that functional imaging correlates are likely best undertaken early during chemoradiation. Additional patients are being recruited into this study to correlate imaging response with clinical and pathological outcomes. PV-0323 Development of a prognostic model incorporating PET texture analysis in oesophageal cancer patients K. Foley 1 , R. Hills 1 , B. Berthon 2 , C. Marshall 2 , W. Lewis 3 , T. Crosby 4 , E. Spezi 5 , A. Roberts 6 1 Cardiff University, Division of Cancer & Genetics, Cardiff, United Kingdom 2 Cardiff University, Wales Research & Diagnostic PET Imaging Centre, Cardiff, United Kingdom 3 University Hospital of Wales, Upper GI Surgery, Cardiff, United Kingdom 4 Velindre Cancer Centre, Oncology, Cardiff, United Kingdom 5 Cardiff University, School of Engineering, Cardiff, United Kingdom 6 University Hospital of Wales, Clinical Radiology, Cardiff, United Kingdom Purpose or Objective Texture analysis provides additional quantitative data extracted from radiological staging investigations. This exploratory study investigates the prognostic significance of PET texture variables when incorporated into a model predicting overall survival (OS) in patients with oesophageal cancer (OC). Material and Methods