ESTRO 2020 Abstract Book

S636 ESTRO 2020

per institutional protocol (NCT01343121). Forty-eight tumour tissue sample blocks were available for analysis. Next generation sequencing of transcriptomic RNA was performed evaluating the expression of 395 immune response genes. Gene expression was categorised relative to validated house-keeping genes and analysed with respect to patients who had relapsed with distant metastases. Results After 3 years of follow up, 6 patients had developed distant metastases. In these patients, CD3 and IL10 expression was reduced indicating a low T-cell and monocyte infiltrate. M- TOR, PIK3CA and PTPN11 were elevated suggesting activation of signalling pathways often found in aggressive cancers. Conclusion This study is hypothesis generating, but provides data supporting a relationship between the immune tumour microenvironment and poor response to treatment and development of distant metastases. With the PIK3CA- MTOR-AKT and RAS-MAPK pathways implicated with an increased aggressive phenotype and resistance to anti- PD1/PDL1 therapy, immune response gene signatures should be investigated as both prognostic and predictive biomarkers in bladder cancer. PO-1207 Exploring molecular subtype as a biomarker of radiation response in muscle-invasive bladder cancer M. Tan 1,2 , G. Nyamundanda 3 , E. Fontana 3 , S. Hazell 4 , C. Ragulan 3 , K. Jones 5 , B. Abah 5 , T. Jacobs 5 , J. Bowes 5 , A. Sadanandam 3 , R. Huddart 1,2 1 The Institute of Cancer Research, Radiotherapy and Imaging, Sutton, United Kingdom ; 2 The Royal Marsden NHS Foundation Trust, Academic Urology Unit, Sutton, United Kingdom ; 3 The Institute of Cancer Research, Systems and Precision Cancer Medicine, Sutton, United Kingdom ; 4 The Royal Marsden NHS Foundation Trust, Department of Histopathology, London, United Kingdom ; 5 The Royal Marsden NHS Foundation Trust, Bob Champion Unit, Sutton, United Kingdom Purpose or Objective Muscle-invasive bladder cancer (MIBC) can be divided into molecular subtypes based on RNA expression profiles. Data suggests that subtype may be a biomarker of response to neoadjuvant chemotherapy in MIBC, but its role as a biomarker of radiotherapy response has been less explored. We therefore aimed to investigate whether molecular subtype might be associated with response to In this retrospective pilot study, RNA was extracted from diagnostic formalin-fixed paraffin-embedded tumour tissue, from patients with MIBC treated in a single UK centre with radical radiotherapy +/- chemotherapy. Patient tumour samples and data were obtained from our institution’s MIBC biobank. Samples were interrogated by a custom-designed Nanostring panel which included TCGA MIBC subtype genes and radiosensitivity genes. Expression data underwent non-negative matrix factorisation (NMF) to determine subtypes present. The resulting classifier was also applied to publically available data sets. Kaplan-Meier analysis was performed to examine clinical endpoints including invasive locoregional relapse-free survival (invLR_RFS) and overall survival (OS). Data was analysed using R and SPSS. Results Expression data was successfully generated for 43/44 (98%) patients tested. The mean age was 71.9 years and 79% were male. At presentation, 29/43 (67%) had T2-4 N0 M0 disease, 11/43 (26%) had locoregional nodal involvement and 3/43 (7%) had para-aortic nodal involvement. 37/43 (86%) had predominantly transitional cell carcinoma histology. 32/43 (74%) had neoadjuvant chemotherapy and radiotherapy in MIBC. Material and Methods

42/43 (98%) had concurrent chemotherapy. All patients completed radiotherapy. 39/43 (91%) patients had a post- radiotherapy cystoscopy at 3-4 months post-treatment; 26/39 (67%) had a biopsy result available from this procedure. NMF analysis identified 5 molecular subtypes. At a median follow-up of 3.8 years, 1/20 (5%) patients within subtypes 4-5 had a documented invasive locoregional recurrence compared to 8/23 (35%) of patients in subtypes 1-3 (p = 0.0243). 2-year invLR_RFS was 100% and 71% respectively (p=0.028). There was no significant difference in OS. With regards to post-radiotherapy biopsy results, the pathological complete response rate, defined as pT0, was 100% for the 11 patients in subtypes 4-5 compared to 60% for the 15 patients in subtypes 1-3 (p = 0.0237). On applying our classifier to publically available data, the 2-year OS for subtype 4 in the primarily surgically managed TCGA cohort was <50%. In our radiotherapy cohort, the 2- year OS for subtype 4 was 85%. This difference may suggest that patients in subtype 4, which displayed basal features, may derive greater benefit from radiotherapy +/- chemotherapy than surgery alone. Conclusion Data from this pilot study suggests that molecular subtype could be associated with response to radiotherapy in MIBC. Cohort numbers were small and further work is warranted. PO-1208 High-dose rate brachytherapy in the treatment of early stages of penile carcinoma D. Pohanková 1 , I. Sirak 1 , L. Kašaová 1 , J. Grepl 1 , P. Paluska 1 , M. Louda 2 , L. Holub 2 , J. Špaček 2 , P. Prošvic 3 , J. Petera 1 1 University Hospital Hradec Kralove, Oncology and Radiotherapy Department, Hradec Kralove, Czech Republic ; 2 University Hospital Hradec Kralove, Department of Urology, Hradec Kralove, Czech Republic ; 3 Hospital Náchod, Department of Urology, Náchod, Czech Republic Purpose or Objective Interstitial low-dose rate (LDR) brachytherapy (BT) allows a conservative treatment of T1-T2 penile carcinoma. High- dose rate (HDR) BT is often considered as a dangerous method for interstitial implants because of higher risk of complications. However, numerous reports suggest that results of HDR BT may be comparable to LDR. We present our first experience with HDR brachytherapy in the treatment of penile cancer. Material and Methods Twenty six patients with early penile carcinoma (T1- 2N0M0) were treated by HDR BT at dose 18 times 3Gy per fraction twice daily between years 2002-2018. The target volume encompassed tumor with 0.5 – 1.0 cm margin. In 7 patients the stainless hollow needles were inserted in 1 plane, in 12 patients in 2 planes and in 7 patients in 3 planes. The square pattern geometry and separation of 10 mm between needles were used. The distance of urethra was kept at least 5 mm from the plane of needles. Breast interstitial brachytherapy template was used for fixation and precise geometry reconstruction. The distance between both plates, between needle tip and plate, needle tip and mucosa exit point and needle tip and mucosa entry point were measured directly. The dose distribution was calculated with program for template with square pattern geometry on Brachyvision planning system (Varian, USA). The irradiation was performed by GammaMed afterloading device (GammaMed, Germany). Results The median number of implanted catheters was 4 (2; 12). The median volume covered by 3 Gy isodose was 7 cm 3 (2.1;11). The median V150 was 2.1 cm 3 (0.9; 5.0) and the mean dose homogeneity index was 0.62 (0.46; 0.87). Median follow up was 95 months (17; 210). Acute reaction consisted of grade II mucositis that dissolved during 8