ESTRO 38 Abstract book

S797 ESTRO 38

Dose to sphincter complex (Dmean, V50Gy and D90%) differed significantly between patients with no and major LARS (p= 0.049, 0.035 and 0.024 respectively). There was a significant correlation between D90% to the sphincter complex and the LARS sub-unit “Do you ever have accidental leakage of liquid stool?” (1: yes, 2: no), p= 0.047, with higher risk of incontinence with higher dose. The same applied for V45Gy to the bowel bag and question “How often do you open your bowels?” (1: <1 time per day, 2: 1-3 times per day, 3: 4-7 times per day, 4: >7 times per day), p=0.029, with higher risk of frequent bowel movements with higher dose. Conclusion Specific doses to the sphincter complex could become a useful predictor of anal incontinence and high LARS, and dose to bowl bag a predictor of frequent bowel movements after radiotherapy for anal cancer EP-1470 Response assessment in rectal cancer patients treated with MRI-guided RT: preliminary results A. Re 1 , G. Chiloiro 2 , L. Boldrini 2 , F. Cellini 2 , M. Giraffa 1 , P. De Franco 1 , B. Corvari 2 , E. Meldolesi 2 , B. Fionda 2 , D. Cusumano 3 , S. Teodoli 3 , V. Valentini 1 , M.A. Gambacorta 1 1 Università Cattolica del Sacro Cuore, Diagnostica per immagini- Radioterapia oncologica ed Ematologia, Roma, Italy ; 2 Fondazione Policlinico Universitario A. Gemelli IRCCS, Diagnostica per immagini- Radioterapia oncologica ed Ematologia, Roma, Italy ; 3 Fondazione Policlinico Universitario A. Gemelli IRCCS, Diagnostica per immagini- Radioterapia oncologica ed Ematologia - U.O.C. Fisica Sanitaria, Roma, Italy Purpose or Objective To evaluate the efficacy in terms of complete response (CR) of neoadjuvant magnetic resonance-guided radiotherapy (MRgRT) delivered with a hybrid 60Co-MRI unit in locally advanced rectal cancer (LARC) patients (pts). Material and Methods MRgRT includes daily MRI for setup and real-time cine MRI scans for target tracking. Pts affected by LARC and clinically eligible for MRgRT were consecutively enrolled. The clinical target volume (CTV)1 included the tumor volume and the corresponding mesorectum received 55 Gy/2.2 Gy per fraction and the CTV2, including total mesorectum and draining lymph nodes, 45 Gy/1.8 Gy with a simultaneous integrated boost. CTV to PTV margins were 0.5 cm in all directions. Fluoropyrimidine-based concomitant chemotherapy (CT) was planned, Oxalyplatin was added in selected cases. Acute adverse events were registered according to CTCAE v.4.0 scale. Surgery was planned at least 8 weeks after the end of chemoradiotherapy (CRT) and pathological complete response (pCR) was defined as ypT0N0. Results From February 2017 to July 2018 20 LARC pts were enrolled. Massive lymph nodal involvement (cN1 or cN2) or mesorectal fascia involment were observed in 30%, 55% and 60% of the cases, respectively. Extra-mesorectal pathological nodes were detected in 30% of pts. Primary lesion was located in low rectum in 40% of the pts. Compliance to CRT was overall good: all pts completed the prescribed treatment. Fifteen patients (75%) required at least one day of suspension from CRT(average 2,65 days, range 1-10). No patient reported grade 3 or 4 hematological, genitourinary or skin toxicity. Overall GI adverse events were more commonly reported: 5 pts (25%) showed grade 3 diarrhea during CRT. Two pts (10%) did not undergo surgery as complete clinical response (cCR) was described at restaging exams and confirmed at early follow up assessments and were

Results 16 pts (13 male and 3 female) were enrolled, out of which 5 pts (31%) showed cCR. The most predictive radiomics feature was the “grey level non-uniformity” ( glnu ), whose variation resulted to be always significant in discriminating cCR pts from not responding ones ( p = 0,0375 at t5; 0,0005 at t10; 0,0133 at t15; 0,0032 at t20; 0,0380 at t25). Volume variation resulted to be significant in 4 of the considered time fractions ( p = 0,0490 at t5; 0,0032 at t10; 0,0636 at t15; 0,0055 at t20; 0,0275 at t25). The trends of glnu and volume are reported in figure 1 and 2 respectively for all the pts. Conclusion This hypothesis generating study suggests that hybrid 0.35 T MRgRT images could be suitable for radiomics analyses despite the low spatial resolution and that radiomics parameters may perform better than standard tumor shrinkage measurements in the prediction of cCR. EP-1469 Radiation dose to pelvic floor muscles and functional outcome after treatment for anal cancer C. Kronborg 1 , P. Christensen 2 , B.G. Pedersen 3 , K.G. Spindler 1 1 Aarhus University Hospital, Oncology, Aarhus C, Denmark ; 2 Aarhus University Hospital, Surgery, Aarhus C, Denmark ; 3 Aarhus University Hospital, Radiology, Aarhus C, Denmark Purpose or Objective Chemoradiotherapy is the standard treatment for localized anal cancer. However, high dose pelvic radiation is associated with significant both acute and late toxicity. Late and persistent gastrointestinal dysfunction is among the most frequent, debilitating toxicities, with anal incontinence affecting up to 50 %. Radiation dose to the sphincter apparatus, but also to muscles of pelvic floor has been associated to functional gastrointestinal outcome in these patients. The LARS (lower anterior resection syndrome) questionnaire is a patient reported outcome measure, which evaluates bowel dysfunction after rectal surgery. It consists of five items, all relevant to the late toxicities seen after pelvic radiotherapy: frequent bowel movements, gas and fecal incontinence, fragmentation, and urgency. We aimed to: i: compare radiation dose to sphincters and pelvic floor muscles in patients with no LARS and major LARS. ii: correlate radiation dose to sphincters and pelvic floor muscles and the 5 individual items in LARS score. Material and Methods From 2009 to 2015, anal cancer patients were routinely asked to fill out LARS questionnaires at follow-up. For this project all patients with no LARS (score 0-19) were selected from the cohort and an equal number of patients with major LARS (score 30-42). The Internal and external sphincter (contoured as sphincter complex), levator ani and puborectal muscles were delineated on planning MRI or CT-scans. Dose volume parameters: V50Gy, V60Gy, D90%, Dmean, and Dmax were obtained for each structure and for bowel bag V45Gy and V30Gy. Differences between patients with no or major LARS were tested by Mann-Whitney U-test, correlations by Spearman´s rank correlation. Results A total of 36 patients were included, 18 patients with no LARS (median 11, range 0-20) and 18 with major LARS (median 37 range 31-39). Gender, age, TNM stage, PTV, chemotherapy, time to LARS score (mean 692 and 749 days) were not statistically different in the two groups. Thirty-three patients were treated with 64 Gy/ 51,2 Gy in 32 fractions, (one with 46 Gy, one 45Gy, and one 54Gy). Thirty patients received therapy with 2 arc VMAT and six with 6-field IMRT.