ESTRO 2020 Abstract Book

S176 ESTRO 2020

Purpose or Objective Patients (pts) with locally advanced rectal cancer (LARC) exhibit heterogeneous responses to treatment. Patients who achieved a pathological complete response (pCR) are considered to be a more favorable subpopulation with less local recurrence (LR) and distance metastases (DM) as well as better overall survival (OS). As reported in a previous large database analysis, comparing with pCR, 2 years disease free survival (2yDFS) can be considered a stronger predictor of OS. The aim of this study was to update the previous analysis on a more copious pooled set of LARC pts. Material and Methods Pooled and treatment subgroup analysis were performed on 11 large international rectal cancer trials: Accord 12/0405, EORTC 22921, FFCD 9203, CAO/ARO/AIO-94, CAO-ARO-AIO-04, INTERACT, I-CNR-RT, XELOX , two Polish trials and TROG 01.04. All the selected LARC pts received long-course radiotherapy (LC-RT) with or without concomitant and/or adjuvant chemotherapy (CT) and surgical procedure. Pts with no information regarding pCR were excluded. We calculated 2yDFS as no distance metastasis or local recurrence within 2 years. Furthermore we divided pts, depending of treatment, in 4 subgroup: only LC-RT, LC-RT+ concomitant CT, LC-RT + adjuvant CT and adjuvant LC-RT. Pearson’s Chi-squared, Kaplan-Meier curve and Logrank test were used for data analysis. A p-value less than 0.01 was considered as a statistical significant value. Results Overall 5492 pts out of 7936 LARC pts satisfied the inclusion criteria in this pooled dataset and were analyzed in this study. According to the patterns of recurrences, 3 pts groups were identified: 11% had pCR and 2yDFS (excellent prognosis), 66% had no pCR but 2yDFS (good prognosis) and 20% had neither pCR and 2yDFS (poor prognosis). Pathological CR was confirmed as an early good predictor of OS, independently of clinical tumor stage (cT3 p< 0.01, cT4 p= 0.02) while pCRs obtained after the only LC-RT seem to not predict OS (p= 0.21). Furthermore, the results confirmed that 2yDFS was a strong predictor of OS compared to pCR for LARC pts (Figure 1) independently of clinical tumor stage (p<0.01) and treatment type (p<0.01) (Figure 2). The landmark analysis shows that achieving a pCR is significantly beneficial when there is 2yDFS. In the case of early recurrence or metastasis, pCR status appears to be statistically irrelevant for OS. Conclusion This update analysis confirmed that 2yDFS is a stronger predictor of OS compared with pCR in more than 5000 LARC pts. These results support that accurate prediction models should be developed for 2yDFS to select LARC pts for more intensified treatment strategies. OC-0335 Circulating free DNA in squamous cell carcinoma of the anus: Relation to risk factors and recurrence A.C. Lefèvre 1 , C. Kronborg 2,3 , B.S. Sørensen 1 , E. Serup- Hansen 4 , K.G. Spindler 1,2 1 Aarhus University Hospital, Department of Experimental Clinical Oncology, Aarhus, Denmark ; 2 Aarhus University Hospital, Department of Oncology, Aarhus, Denmark ; 3 DCPT, Danish Center for Particle Therapy, Aarhus, Denmark ; 4 Herlev Hospital, Department of Oncology, Herlev, Denmark

Purpose or Objective Analyses of circulating free DNA (cfDNA) has set new frames for developing personalized treatment in oncology. We developed a simple method to quantify total cfDNA from a simple blood sample, and found promising results in rectal cancer. Here, we present the first study on total cfDNA in squamous cell carcinoma of the anus (SCCA). The aim of this study was to measure the total cfDNA level in SCCA during treatment and follow up (FU), and relate the results to risk factors and treatment outcome. Material and Methods Between April 2016 and September 2019, patients treated with standard radiotherapy (RT) (60-64Gy/49.5- 51.2Gy/30-32Fx) at the Department of Oncology, Aarhus, DK, were included in the prospective PLAN-A study. P16 status was measured by immune histochemical staining, and serum samples were analyzed for cfDNA using direct fluorescent assay (DFA) in quadruplets, and outliers removed if the standard division >15%. Results Eighty patients were included, and serum samples were available at baseline (n=73), mid-therapy (n=74), end of therapy (n=67) and at one-year FU (n=30). P16 status was positive in 87% of 55 tested cases. T1 tumors comprised 19%, T2=59%, T3=10% and T4=12% respectively. Eight patients received neo-adjuvant chemotherapy, and 57 patients concomitant cisplatin/5FU-based chemotherapy. A Mann-Whitney test indicated that the level of cfDNA was higher for P16 negative tumors (median 1.48) than for P16 positive tumors (median 0.9, P=0.04). Data revealed a weak association between baseline cfDNA levels and numeric tumor size (R 2 =0.07, P=0.05), and higher level with increasing T-stage (median values of T1= 0.80, T2=0.94, T3=1.11, T4=1.3). There was a significantly higher cfDNA level with increasing performance status (PS); the median cfDNA was 0.89 in patients with PS 0 compared to 1.15 in patients with PS>0, P<0.001. After a mean FU period of 21 months, eight patients experienced local recurrence and one distant failure. The median levels of cfDNA dropped significantly from baseline to mid-therapy (0.92-0.78, P=0.01), and from end-therapy to one-year FU (0.89-0.69, P=0.03), but fluctuations in cfDNA levels during RT showed no significant correlation to outcome. However, all patients with relapse had a baseline cfDNA level above the 25th quartile (p=0.05), which translated into difference in recurrence free survival, when using the 25% quartile as cut off for Kaplan- Meyer curves. However, the low number of events and short FU, does not yet allow for reliable statistical analysis. Conclusion The analyses showed an association between cfDNA levels and the risk factors P16 status, T-stage and PS, and low risk of relapse for patients with the lowest baseline levels. Results thus indicate a relation between cfDNA level and treatment outcome, but with few events in total. This is the first study describing total cfDNA in SCCA and further studies and validation of data are needed. OC-0336 Evaluation of bowel for online adaptive stereotactic radiotherapy of abdominal oligometastases J. Nuyttens 1 , L. Visani 2 , P. Granton 1 , M. Milder 1 1 Erasmus MC Cancer Institute, Radiation oncology, Rotterdam, The Netherlands ; 2 Azienda Ospedaliero- Universitaria Careggi, Radiation Oncology, Florence, Italy Purpose or Objective To investigate the deformable image registration (DIR) of the bowel compared to manual delineation and the