ESTRO 2020 Abstract Book

S407 ESTRO 2020

Cancer Institute & Research Centre, Senior Resident Medical Oncologist, Rohini- Delhi, India ; 4 Rajiv Gandhi Cancer Institute & Research Centre, Consultant Radiologist, Rohini- Delhi, India ; 5 Rajiv Gandhi Cancer Institute & Research Centre, Consultant Radiation Oncologist, Rohini- Delhi, India ; 6 Rajiv Gandhi Cancer Institute & Research Centre, Attending Consultant Radiation Oncologist, Rohini- Delhi, India ; 7 Rajiv Gandhi Cancer Institute & Research Centre, Senior Resident Radiation Oncologist, Rohini- Delhi, India ; 8 Rajiv Gandhi Cancer Institute & Research Centre, Resident Radiation Oncologist, Rohini- Delhi, India ; 9 Rajiv Gandhi Cancer Institute & Research Centre, Research Assistant, Rohini- Delhi, India ; 10 Rajiv Gandhi Cancer Institute & Research Centre, Director Radiology, Rohini- Delhi, India ; 11 Rajiv Gandhi Cancer Institute & Research Centre, Chair Radiology, Rohini- Delhi, India Purpose or Objective Apparent Diffusion Coefficient (ADC) is proposed as a potential marker in response evaluation in many malignancies.The purpose of our study was to compare quantitative MRI (ADC values) and clinically correlate with outcomes in patients with cervical cancer. Material and Methods MR studies of histopathologically confirmed cervical Squamous Cell Carcinoma patients (n= 133) presenting to our institute, were retrospectively analyzed. Patients included for the study had locally advanced disease, stages IIB to IVA. ADC values were retrospectively measured from pretreatment MR images. The mean ADC value was measured in a region of interest (ROI by drawing an arbitrary circle of area 0.35 cm 2 ) chosen after imaging analysis and corresponding to the best area of diffusion restriction. Further treatment with brachytherapy(BT) was done as per the residual disease after EBRT. Brachytherapy (BT) was either intracavitatory (ICRT) or interstitial using the MUPIT template. Change of ADC values was correlated with the outcomes and disease related parameters. Median ADC was charted for all patients and subdivided into a low ADC change group <0.48 vs high ADC change group >0.48. Results 100 patients underwent ICRT and 33 underwent MUPIT. Baseline ADC mean value (range ±SD) was 0.93(0.55- 1.53±0.159)in ICRT vs 0.94( 0.53-1.55 ± 0.265) in MUPIT. Post EBRT mean ADC in ICRT was 1.50(0.68-1.94 ± 0.237) vs 1.42 (0.66-1.86 ± 0.329) in MUPIT. ADC change in ICRT was 0.66(0.51-2.35± 0.413) vs -0.46(-0.78 to -0.07 ± 0.173) in MUPIT. 4% (4/100) had nodal positivity in ICRT compared to 24%(8/33) in MUPIT. In ICRT 13% (13/100) had metastatic disease on follow up vs 10 % (3/30) in MUPIT arm. Stage distribution in ICRT arm (stage II, III, IV) is 49%, 45%, 6% vs 27%, 61%, 12% in MUPIT arm. In ICRT arm 4%(4/100) had recurrence compared to 8/33(24%) in MUPIT arm. Change in ADC was significant between 2 groups(p value <0.001). Median recurrence free survival for ICRT was 25 vs 9 months in MUPIT arm. Correlation of change in ADC with recurrence in the ICRT vs MUPIT group was statistically insignificant. High ADC change had better survival in both groups when compared to low ADC Group and more significant in the MUPIT group as shown in Figure. Change of ADC is significantly correlated with the stage(p value = 0.025) and correlation is more strong when stage III and IV were clubbed together compared to stage 2 (p value=0.012). Lymph node status was not significantly correlated with ADC change probably because of the low nodal burden in the study group.

Conclusion Change in ADC, a quantitative MRI parameter seems to be a potential prognostic marker in uterine cervix cancer with strong correlation in advanced stages. This study shows that low change in ADC in advanced stages is a poor prognostic marker and needs treatment intensification to improve clinical outcomes. PH-0719 18F-FDG-PET/CT parameters as predictors of survival and response to nCRT in esophageal cancer L. Marr 1 , T. Pyka 2 , J.C. Peeken 1 , B. Haller 3 , M. Jesinghaus 4 , W. Weichert 4 , W. Weber 5 , S.E. Combs 1 , S. Münch 1 1 Technical University Munich TUM, Department of Radiation Oncology, Munich, Germany ; 2 Technical University Munich TUM, Department of Diagnostic and Interventional Neuroradiology, Munich, Germany ; 3 Technical University Munich TUM, Institute of Medical Informatics, Munich, Germany ; 4 Technical University Munich TUM, Institute of Pathology, Munich, Germany ; 5 Technical University Munich TUM, Department of Nuclear Medicine, Munich, Germany Purpose or Objective In patients with advanced esophageal squamous cell carcinoma (ESCC) a multimodal therapy including neoadjuvant chemoradiation (nCRT) and subsequent surgery is the treatment of choice. The aim of this study was to validate the predictive potential of clinical parameters and 18F-FDG-PET/CT derived textural features in ESCC patients undergoing nCRT + surgery. Material and Methods Between 2005 and 2014, 38 ESCC patients (18 females and 20 males, median age 62 years) with pre-treatment 18F- FDG-PET/CT scans were treated with nCRT and esophagectomy at our department. All patients received neoadjuvant treatment including platinum-based chemotherapy and a total radiation dose of 45Gy. Response to nCRT was evaluated by histomorphological workup of the resected specimen as described by Becker et al. [1]. Good response was defined as <10% residual tumor (Becker 1a + 1b), while poor response was defined as ≥10% residual tumor (Becker 2 + 3). Regression analyses were used to identify parameters predicting overall survival (OS) and response to nCRT. The optimal cut-off value of significant parameters was determined by the maximally selected LogRank test and a test on association using the R package maxstat.test. [1] Becker et al., Histomorphology and grading of regression in gastric carcinoma treated with neoadjuvant chemotherapy. Cancer, 2003. 98(7): p. 1521-30. Results The histopathologic analysis of the resected specimen reported good response in 28 patients (71%) and poor response in 11 patients (29%). After a median follow-up of 68 months for surviving patients, median overall survival was 40 months. Using binary logistic regression analysis, we saw no significant predictive potential for any of the PET/CT parameters regularly described in the literature in terms of treatment response (e.g. maximum standardized uptake value (SUVmax), Metabolic tumor volume (MTV) and texture features). While clinical parameters T-stage, N-stage, tumor length, and sex were not associated with