ESTRO 38 Abstract book

S263 ESTRO 38

after IGABT in LACC patients. The risk of bleeding and cystitis was considerably higher in smokers. Finally, age was a predictor for cystitis, with younger patients at higher risk. OC-0508 MRI guided chemoradiation and brachytherapy for postsurgical vaginal recurrences: A phase II study S. Chopra 1 , R. Engineer 2 , U. Mahantshetty 2 , S. Mechanery 3 , T. Dora 3 , R. Shukla 4 , P. Popat 5 , J. Swamidas 3 , J. Ghosh 3 , S. Gupta 3 , S. Shrivastava 2 1 Advanced Centre for Treatment- Research and Education in Cancer- Mumbai, Radiation Oncology, Mumbai, India ; 2 Tata Memorial Hospital, Radiation Oncology, Mumbai, India ; 3 Advanced Centre for Treatment Research and Education in Cancer, Radiation Oncology, Navi Mumbai, India ; 4 Advanced Centre for Treatment- Research and Education in Cancer, Radiation Oncology, Mumbai, India ; 5 Tata Memorial Hospital, Radiodiagnosis, Mumbai, India Purpose or Objective To report outcomes of prospective phase II study that integrated MRI for planning external radiation and brachytherapy for postsurgical recurrences of cervical cancer. Material and Methods The prospective phase II study (NCT 01391065) recruited patients from January, 2011 until December, 2016. Patients with local residual or recurrent tumour after hysterectomy without any evidence of distant metastasis were included. All patients underwent baseline T2 MR imaging to define the extent and volume of disease and were planned for external radiation (50 Gy/25#/5weeks (IMRT or 3DCRT)) with concurrent weekly cisplatin (40mg/m2). Response assessment MRI was performed in all patients at completion of brachytherapy. While patients with baseline parametrial disease received interstitial high dose rate brachytherapy including vagina and parametrium (16-20Gy/4-5#) those with only central disease at presentation received only central vaginal brachytherapy (12-14 Gy/2-4#). Information from MR was utilized to delineate clinical target volume on CT obtained for planning interstitial or intracavitary brachytherapy. MRI was done at first follow up followed by clinical examination with additional imaging as and when indicated. Kaplan Meier Analysis was performed to evaluate Locoregional Relapse, Disease Free and Overall Survival. Univariate and multivariate analysis was performed to evaluate impact of known prognostic A total of 60 patients were included of which 51 were identified to have nonmetastatic disease. The median age of the cohort was 48 years (35-65 yrs). The median and mean tumour volume on T2W MRI was 25 cc (IQR 8.4-145 cc) and 118 cc ( IQR8.4-145) cc[A1] ). A vast majority of patients had squamous carcinoma (82.4%). Overall 16/50 (31.4%) patients had radiologically positive lymph nodes prior to treatment initiation. Overall 40 patients (78%) had complete response after external radiation, 8 had partial response and 3 patients had progressive disease. Planned treatment was completed in 47/50 patients. The median EQD2 was 70.8Gy (50-73 Gy). The median D 2cc for rectum, bladder and sigmoid was 65.8 Gy, 68 Gy and 62 Gy respectively. At a median follow up of 60 months (5- 93) local control was 88%. Of the 6 patients (12%) with local relapse 4 were local only and another 2 also had distant metastasis. Five patients (10%) had distant relapse in the absence of local relapse leading to 5 and 7 year disease free and overall survival of 72% and 69% and 88% and 77% respectively. Grade III proctitis and cystitis was observed in 4% and 2% patients respectively. Of the known prognostic, predictive and treatment related factors only baseline tumour volume >20 cc adversely impacted 7 year factors. Results

Purpose or Objective To identify risk factors for bladder fistula, bleeding and cystitis within the prospective, multi-institutional, observational EMBRACE I study (An international study on MRI-guided brachytherapy in locally advanced cervical cancer) that enrolled 1416 patients treated from 2008 to 2015. Material and Methods Bladder fistula, bleeding and cystitis (CTCAEv.3) were analysed in Locally Advanced Cervical Cancer (LACC) patients treated with radiochemotherapy and Image- Guided Adaptive Brachytherapy (IGABT). Adverse events arising at any time in the course of follow-up were considered. Patient, disease and treatment characteristics were tested as risk factors for moderate grade (G) ≥2 with univariate (UVA) and multivariable (MVA) analyses (Cox proportional hazards model) in patients without bladder involvement. UVA and MVA were also performed for severe G≥3 incidence pooled over the three endpoints. The cumulative (EBRT+IGABT) minimal dose to the most exposed 2cm 3 of the bladder (D2cm 3 ) and ICRU Bladder Point dose (EQD2) were considered as continuous variables in UVA and MVA. Urinary frequency and incontinence were not included in the analysis. Results In 1146 patients without bladder involvement, the crude incidences for G≥2 fistula, bleeding and cystitis were 0,7% (n=8), 2,0% (n=23) and 7,4% (n=84), respectively. The pooled incidences for G≥2 and G≥3 were 8,5% (n=97) and 1,4% (n=16), respectively, with 15% of symptomatic patients experiencing more than one endpoint. In 67 patients with bladder involvement, the incidences for G≥2 fistula, bleeding and cystitis were 2,5% (n=8), 12% (n=2) and 17% (n=11), respectively, and G≥3 in 15% (n=10) (pooled incidence). Mean (SD) bladder D2cm 3 was 75,8±9,7Gy and 83±11,5Gy in patients without and with bladder involvement, respectively. Median follow-up was 35 (range:1-97) months, median age was 49 (range:22-91) years, and 31% were smokers. Table 1 shows the covariates tested in UVA and MVA for G≥2 fistula, bleeding and cystitis, and G≥3 for pooled symptoms in patients without bladder involvement. The Hazard Ratios (HR) are shown for significant variables in UVA (p≤0.10) and MVA (p≤0.05). Bladder D2cm 3 was significant on MVA for all individual and pooled endpoints. ICRU Bladder Point dose was significant in UVA for fistula and pooled incidence, but was not included in the MVA because of correlation with bladder D2cm 3 and a lower HR. Smoking status was predictive for bleeding, cystitis and pooled incidence, with smokers at higher risk. Younger patients had a higher risk for cystitis compared to older patients.

Conclusion In the present study, bladder D2cm 3 was a dominant risk factor for developing bladder fistula, bleeding or cystitis