ESTRO 2020 Abstract book

S86 ESTRO 2020

PH-0162 Is there a radiation dose-response relationship for non-operative management of rectal cancer? A. Appelt 1,2 , A. Jakobsen 2 , J. Gerard 3 , D. Sebag- Montefiore 1 1 Radiotherapy Research Group- Leeds Institute of Medical Research at St James's, University of Leeds, Leeds, United Kingdom ; 2 Danish Colorectal Cancer Center South, University of Southern Denmark, Vejle, Denmark ; 3 Department of Radiation Oncology, Centre Antoine Lacassagne, Nice, France Purpose or Objective The last decade has seen growing interest in organ preservation and non-operative management (NOM) for rectal cancer. Published literature primarily focuses on NOM following ‘opportunistic’ clinical complete response (cCR) after neoadjuvant (chemo)radiotherapy. If we are to increase the proportion of patients who avoid surgical intervention, dedicated up-front strategies are needed, using local treatment intensification. We conducted a systematic review of published studies to assess the role of radiation dose in NOM. Material and Methods We searched MEDLINE to identify studies reporting on NOM for rectal cancer (Table 1 legend for details). We included English language papers reporting on patient cohorts treated with (chemo)radiotherapy, with: 2) data for all patients treated on clinical response and local control at 2 years without surgery; and 2) information allowing estimation of tumour dose in EQD2(α/β=10Gy). We excluded studies with routine use of local excision; with extended chemotherapy regimens; focusing on palliative management (incl stage IV disease); solely based on cancer registries; or reporting on ≤10 patients with cCR. We extracted total number of patients treated, local control without surgery at 2 years, radiation dose to the primary tumour, and (if available) proportion of cT1-2 versus cT3-4 tumours treated. We assessed whether the report represented a ‘planned’ NOM strategy, i.e. with all patients systematically assessed for cCR and NOM, or an ‘opportunistic’ cohort. For the former, we estimated the relationship between EQD2 and local control at 2 years, using bounded logistic regression. Finally, we adjusted local control rates for T stage mix, using data from Maas et al (Lancet Oncol 2010; 11: 835–44), and fitted dose- response separately for cT1-2 and cT3-4. Results We screened 745 papers, and identified 15 non- overlapping cohorts for inclusion (Table 1). EQD2 to the primary tumour varied from 40.7Gy to >160Gy, with higher dose levels typically delivered using endorectal brachytherapy or contact X-ray therapy. Nine papers reported on dedicated NOM cohorts, with local control at 2 years at 12%-72%. A significant (p<0.001) dose-response relationship was estimated from these reports (Figure 1), characterised by D 50 = 70.6Gy (95% CI 61.5-79.7Gy) and γ 50 = 1.09 (0.95-1.23). The dose-response appeared to taper off for >100Gy, with upper bound on local control at 72%. The published cohorts had substantial heterogeneity in cT stage distribution, but this alone did not explain the variation in local control (p<0.001). Correcting for stage mix, the EQD2 needed for 50% local control at 2 years differed by nearly 20Gy for cT1-2 versus T3-4 tumours; D 50 = 66Gy for cT1-2 and D 50 = 85Gy for T3-4.

Conclusion This exploratory analysis of the rectal cancer organ preservation literature indicates a dose-response relationship for NOM. Future trials are needed for validation. The difference in dose response for T1-2 versus