ESTRO 2020 Abstract book

S52 ESTRO 2020

developing second primary cancer (SPC) after RT. Modern intensity-modulated RT (IMRT), with increased low-dose baths and out-of-field scattering, may increase such risks even more. Here we report on the impact of IMRT and smoking status in a single-center retrospective cohort study, using an internal reference group of 3-dimensional conformal RT (3DCRT) patients. Material and Methods The study cohort comprised 1,468 prostate cancer survivors (mean age 70.0 years, 6.7 1SD) treated with either 3DCRT or IMRT in the period 2006-2012 ( Table 1 ). IMRT was gradually introduced during the period 2007- 2010. All patients received standard treatment at that time (72-78 Gy in 2Gy fractions) and had no previous or simultaneous RT to other pelvic areas. Data on SPC incidences (solid, non-skin) were retrieved from the Netherlands Cancer Registry. Only first SPCs were included in the analysis. The Fine and Gray model was applied with death and non-solid SPC as competing risk, to estimate relative risks (Subhazard Ratios [sHRs]) for IMRT vs 3DCRT within smoking categories. Models were adjusted for age at RT, and calendar year of RT. Time was calculated from start RT and was maximized at 10 years to adjust for differences between both groups in maximum follow-up. A latency period of 6 months was considered, resulting in the exclusion of 3 SPCs. Results Median follow-up was 7.5 (IMRT) and 9.5 years (3DCRT). We observed 236 SPCs in the period 0.5-10 years. Most frequent SPCs were bladder (n=44), lung (n=41), colon (n=35), and rectum (n=23). Vital status at the end of follow-up was: 62% alive, 37% died (13% after SPC), 1% emigrated. Distribution of smoking status was: 17% current, 20% previous, 31% never, 32% unknown. For the total cohort (regardless smoking status), the adjusted sHR with 95% CI to develop SPC (IMRT vs 3DCRT) was 1.4 (1.0- 2.0). For never smokers, 1.0 (0.5-2.0), for previous smokers 1.4 (0.6-3.0), for current smokers 2.8 (1.3-5.8), Figure 1 , and for unknown smoker status 1.2 (0.5-2.6). Interaction between smoking status (current y/n) and technique was significant ( p <0.01). Within the “current smokers” subgroup, the adjusted sHR was 1.9 (0.5-7.6) for SPCs in the pelvis, 4.6 (1.0-20.9) for SPCs in the abdomen, and 2.7 (1.0-7.3) for SPCs in the remaining anatomical regions. Conclusion Our study suggests a complex relationship between IMRT, smoking status, and SPC risks. IMRT in current smokers was significantly associated with increased SPC risks, which potentially means that exposure of healthy tissue to low- dose baths and scatter should be minimized in current smokers needing cancer treatment. This observation is biologically plausible, since nicotine is known to promote the growth of cancer cells. Further validation in a multicenter setting with prolonged follow-up is currently ongoing.

Proffered Papers: Proffered papers 6: Novel treatment planning strategies

OC-0102 Reducing secondary lung cancer risk by optimized planning of accelerated partial breast irradiation N. Hoekstra 1 , S. Habraken 1 , A. Swaak - Kragten 1 , S. Breedveld 1 , J. Pignol 2 , M. Hoogeman 1 1 Erasmus MC Cancer Institute, Radiation Oncology, Rotterdam, The Netherlands ; 2 Dalhousie University, Radiotherapy, Halifax NS, Canada Purpose or Objective The survival benefit of adjuvant radiotherapy for low-risk breast cancer patients might be partially offset by the risk of radiation-induced lung cancer. Reducing dose to the lungs during treatment planning for external beam APBI mitigates this risk, but also results in higher doses to the ipsilateral breast if target coverage and dose to the contralateral breast are kept constant. This could lead to a higher risk of breast fibrosis. Our purpose is to quantify the trade-off between secondary lung cancer and fibrosis risks. Material and Methods We conducted a treatment planning study on 20 female patients eligible for APBI, comparing coplanar and non- coplanar techniques, namely VMAT and CyberKnife robotic radiotherapy. We created 11 Pareto-optimal plans per patient per technique using automated treatment planning with the same constraints but gradually shifting priority from maximum sparing of the lungs to maximum sparing of the breast tissue. The excess absolute risk of developing lung cancer was based on lung dose and calculated with the model described by Schneider et al. (2011) accounting for fractionation, repair and repopulation. The risk of breast fibrosis was calculated using the model from Avanzo et al. (2012) for APBI with complete repair. Results The dose parameters, secondary lung cancer risks and fibrosis risks are summarized in Table 1. Prioritizing lung sparing resulted in a substantial reduction of the mean lung dose to as low as 0.3 Gy, and a five-fold median reduction of the secondary lung cancer risk compared to prioritizing sparing of breast tissue. The associated increase in breast dose resulted in a very small absolute increase in fibrosis risk of 0.4%-point. Figure 1 shows the Pareto-fronts of all patients of the trade-off between mean lung dose and mean ipsilateral breast dose. The thick lines depict the means per technique. The use of non- coplanar beams created more planning flexibility compared to coplanar techniques, as shown by the wider