ESTRO 38 Abstract book

S432 ESTRO 38

warranted to identify novel immunometabolic targets to overcome radioresistance. PO-0826 On the value of a prognostic tumour score in locally advanced cervical cancer J.C. Lindegaard 1 , P. Petric 1 , A.M. Lindegaard 1 , K. Tanderup 1 , L.U. Fokdal 1 1 Aarhus University Hospital, Department of Oncology, Aarhus, Denmark Purpose or Objective Staging of cervical cancer is primarily based on clinical examination using the FIGO staging system. In early stage disease subjected to surgery pathology is included. However, imaging and particularly MRI may provide important additional prognostic information on local tumour extension not incorporated in FIGO. As these findings may be of significance in locally advanced cervical cancer (LACC) treated with definitive radiotherapy and brachytherapy, the aim of the present study was to investigate the feasibility of a simple but wide-ranging tumour score for methodical reporting and detailed prognostication. Material and Methods 400 pts with LACC FIGO stage IB-IVA treated between 2005-2018 with external beam radiotherapy +/- concomitant cisplatin and image guided adaptive brachytherapy were included. The diagnostic work-up included clinical examination, FDG PET-CT and T2 weighted MRI. FIGO stage distribution was IB-IIA 9%, IIB 61% and III-IV 30%. The degree of involvement of 8 anatomical locations (cervix, left parametrium, right parametrium, vagina, bladder, ureter, rectum and uterine corpus was scored according to a ranked ordinal scale with 0-3 points (Table 1). The total sum of points was calculated (T-score). Uni- and multi-variate analysis of the T-score in relation to age, performance status (WHO), comorbidity, histology, nodal-stage (PET-CT) and FIGO- stage was performed using Cox regression (SPSS). Endpoints were overall survival (OS), disease free survival (DFS), cancer specific survival (CSS) and local control (LC). Results The median T-score for all pts was 6 (range of 0- 20). Potential prognostic findings not included in FIGO were involvement of the uterine corpus 41%, bladder wall 10%, and mesorectum/rectal wall in 12% (Table 1). Vaginal involvement was unaccounted for in 50% (IIB, IIIB, IVA). In addition, bilateral parametrial invasion was found in 51% and bilateral hydronephrosis in 14%. Proximal only versus distal parametrial involvement was found in 41% and 27%, respectively. Based on the frequency distribution of the T- score, 4 equally sized risk groups were formed: 0-4, 5-6, 7-9 and >9 points (Table 1). The T-score was highly significant in both univariate and multivariate analysis and outperformed FIGO stage for all endpoints (OS, DFS, CSS and LC). When analysing the 245 pts with FIGO stage IIB (Figure 1) the T-score demonstrated a progressively worse OS with increasing score (p=0.021). Similar results were obtained in 36 pts in stage IB-IIA (p<0.001) and 119 pts in stage III-IV (p=0.041). Conclusion The use of a tumour score based on both clinical examination and conventional MRI provided significant information capable of intra FIGO-stage prognostication. Confirmation of the results and tuning of the T-score in relation to the relative prognostic importance of the individual anatomical locations require analysis of an independent and larger cohort including especially more pts in stage IB-IIA.

patients with combined of EBRT/VBT. In the group with VBT alone, 96% of patients received 5x6Gy HDR Ir-192, however, in the group with combined of EBRT/VBT, 52% of patients received 46Gy and 48% over 46Gy (up to 50.4Gy) by EBRT and the various fractionation regimens by VBT, the most of these patients received 3x6Gy (62%) and 2x6Gy (28%) HDR Ir-192. Results Ten-year risk of locoregional recurrence was 3.2% in the EBRT/VBT group and 2% in the group of patients with VBT alone (p=1, Fisher's exact test). However, the 10-year risk of distant metastases was 10.2% in combined EBRT/VBT group and 3% in VBT alone group of patients (p=0.2085). During the median follow-up time of 72 months, 89% of patients with VBT alone lived and 79% of patients with combined of EBRT/VBT (p=0.1827). The rate of late toxicity, in particular genitourinary was significantly lower in the VBT alone than in the combined of EBRT/VBT group (7.6% vs. 29%) (p=0.0057). Ten-year DFS was 76% for the VBT alone group and 70% for group of patients with combined of EBRT/VBT and 10-year OS was 80% for VBT alone versus 77% for combined of EBRT/BT group of patients (p=0.8686, p=0.7310, respectively). Conclusion Also, our update results showed that VBT alone is as effective as combined of EBRT/VBT in preventing the vaginal recurrences and achieving comparable survival rates as adjuvant settings in patients with early stage uterine carcinoma. PO-0825 Differential impact of GLUT1 overexpression between HPV16-positive and -negative cervical cancer B.H. Kim 1 , J.H. Chang 1 1 SMG-Seoul National University Boramae Medical Center, Radiation Oncology, Seoul, Korea Republic of Purpose or Objective Glucose transporter-1 (GLUT1) has been reported as a poor prognosticator associated with radioresistance and immune evasion through metabolic communication in various cancers, but little data is available on cervical cancer. Meanwhile, most cervical cancer is known to be caused by human papilloma virus (HPV), but treatment response and prognosis according to the HPV subtype are also unknown. This hypothesis-generating study was conducted to investigate the prognotic impact of GLUT1 in cervical cancer, in conjunction with HPV subtype. Material and Methods Clinicopathologic factors along with mRNA expression data were obtained using The Cancer Genome Atlas database. Tumor HPV status and immune cell abundance scores were extracted from previous publications. Total 298 patients with appropriate information were analyzed. Results High GLUT1 expression was associated with old age, squamous cell carcinoma, high tumor stage, pelvic lymph node metastases, and low hysterectomy rate. HPV16 positivity itself was not associated with other factors, except low tumor grade. Multivariate survival analysis revealed that high GLUT1 expression (HR 2.57, p = 0.002) and HPV16 subtype (HR 0.56, p = 0.033) were independent prognostic factors for OS. In the subgroup analysis, poor prognostic impact of high GLUT1 expression was maintained in the HPV16 (+) group (p < 0.001), but not in the HPV16 (-) group (p = 0.495). Decreased immune cell abundance scores of B cells, CD8 T cells, and Th1 cells by high GLUT1 expression were observed only in the HPV16 (+) group. Conclusion Despite of heterogenous treatments, our results suggested that GLUT1 expression and HPV16 subtype may have an independent prognostic value in cervical cancer. Especially in the HPV16 (+) group, GLUT1-mediated immunomodulation might be an important cause of treatment failure. Further proof-of-concept study is