ESTRO 2021 Abstract Book

S185

ESTRO 2021

the fraction of events that would not have occurred if there had been no exposure (i.e. death due to cardiac dose). PAF depends on both the increased risk associated with the exposure and prevalence of the exposure. To account for confounding, time-to-event, multivariable models are required to provide adjusted hazard ratios. PAF is plotted as a function of time with associated 95% confidence intervals. Previous work analysed 938 patients with NSCLC treated with curative-intent radiotherapy, treated with 55Gy in 20 fractions. A combined region in the base of the heart, consisting of the right atrium, origin of the coronary arteries and aortic valve, was presented as a region cardiac avoidance area (CAA) with a max dose limit of 19.5Gy (23Gy EQD2). Planning studies identified the proportion of patients where this constraint can be achieved. A multivariable model, including common clinical confounding, alongside dose to CAA, was defined. The adjusted hazard ratio for the CAA was used to calculate the PAF, which represents the percentage reduction in mortality due to reduction of CAA dose. Results 558 out of 938 available patients (with complete case data) were estimated to benefit from reduction in dose to CAA, figure 1a. In multivariable analysis, max dose to the avoidance region was associated with worse overall survival, Hazard ratio: 1.01Gy -1 , p=0.01. Log tumour volume (p<0.001), age (p=0.001), gender (p=0.05) and performance status (p<0.01) were also significant. Median dose to the avoidance region was 37.4Gy, with 622 patients receiving higher than proposed 19.5Gy dose limit. The PAF was fitted assuming the best-case scenario, with dose to CAA reduced to below 19.5Gy. Figure 1b shows the PAF plotted up to 2-years post-radiotherapy. Potential reductions in mortality are 12.3% at 3 months, 9.6% at 12-months and 7.3% at 24-months. The point-wise 95% confidence intervals remain non-zero across the time-range, showing a statistically significance reduction in mortality due to reduction to CAA. However, the confidence intervals remain large, suggesting further unaccounted confounding variables influencing outcomes.

Conclusion The casual effect of dose to CAA has been modelled for a population of patients with lung cancer. Results demonstrate potential reductions in early mortality from implementing dose reduction, with 9.6% reduction at 12 months, which is highly clinically relevant. PH-0276 Prospective study exploring early cardiac toxicity in radiotherapy for LA-NSCLC:Preliminary Analysis M. Miele 1 , M. Fiore 1 , E. Ippolito 1 , S. Mega 2 , M. Carpenito 3 , S. Ramella 1 1 Campus Bio-Medico University , Radiation Oncology, Rome, Italy; 2 Campus Bio-Medico University, Unit of Cardiology, Rome, Italy; 3 Campus Bio-Medico University , Unit of Cardiology, Rome, Italy Purpose or Objective To evaluate early cardiac function variation after chemoradiotherapy (CRT) in LA-NSCLC through the multimodal use of advanced imaging methods such as 2D speckle tracking echocardiography (STE) to explore the associations between early cardiac effects risk factors and late adverse events.