The relationship between cardiovascular disease and survival in patients having cardiac avoidance radiotherapy for lung cancer
Lung cancer and cardiovascular disease (CVD) share common aetiologies and about a quarter of patients treated for lung cancer have CVD. Thoracic radiotherapy is the mainstay of treatment for patients with inoperable lung cancer. However, incidental exposure of the heart to radiation can lead to adverse events and death. People with pre-existing CVD may be more sensitive to heart irradiation. The RAPID-RT study uses ‘real-world’ data and a ‘rapid-learning’ methodology to evaluate the impact of reducing the radiation dose to a defined cardiac avoidance area (CAA), comprising the right atrium, aortic valve root and coronary arteries. To examine the potentially modifying effects of pre-existing CVD and level of CAA dose delivered on survival. Patients with stage I–III lung cancer treated with curative-intent radiotherapy (≥40Gy in >10 fractions, excluding stereotactic ablative radiotherapy), 01/2021-08/2025. From 04/2023, patients were enrolled in RAPID-RT, applying a CAA dose constraint of 19.5Gy in 20 fractions (or equivalent) unless conflicting with tumour coverage. Patients were categorised according to CAA dose above (>19.5Gy) or below (≤19.5Gy). Clinical, treatment, comorbidity, and survival data were extracted from the electronic patient record, with comorbidities recorded using the Adult Comorbidity Evaluation-27. Missing data were addressed using multiple imputation. Survival was analysed using parametric Weibull proportional hazards models, including an interaction between pre-existing CVD and CAA dose, with adjustment for key prognostic clinical and treatment-related covariates. Adjusted 12-month survival probabilities and absolute survival differences were estimated for a typical patient. 1,880 patients were included, of whom 788 received higher CAA dose. 411 (21.9%) had pre-existing CVD. Among patients with CVD, lower CAA dose was associated with a 4.2 percentage point (pp) improvement in 12-month survival, compared with 1.5 pps among those without cardiac comorbidity. Overall survival was poorer in patients with cardiac comorbidity than in those without, particularly among those receiving a higher CAA dose (−5.7 pps), with a less pronounced difference at lower dose (−3.0 pps). However, CVD was not significantly associated with mortality among patients receiving either a higher CAA dose (HR 1.26 95%CI 0.98–1.64), or a lower dose (HR 1.17 95%CI 0.91–1.50). There was no evidence of interaction on the hazard ratio scale (HR for interaction 0.93, p=0.67). Although no statistically significant interaction was observed on the HR scale, the absence of statistical significance should be interpreted cautiously given the limited power to detect effect modification. In contrast, absolute survival estimates suggested a clinically meaningful benefit of lower CAA dose among patients with pre-existing CVD, supporting the potential importance of CAA minimisation in this higher-risk group.