Search

Advances in radiotherapy (RT) for lung cancer treatment have enhanced survival rates but also heightened the concern for cardiotoxicity following RT. While previous studies predominantly focused on ischemic heart diseases in RT, our prior study investigated the risk of atrial fibrillation (AF) using clinical, dosimetry, and coronary artery calcification (CAC) data. The current study aims to develop and validate learning-based predictive models, including machine and deep learning techniques, for predicting AF following chemoradiotherapy (CRT) in non-small cell lung cancer (NSCLC) patients. In this study, we analyzed 321 and 187 datasets from internal and external NSCLC patient cohorts, noting 17 and 6 cases of AF incidence respectively. The analysis incorporated 159 patient-specific features, including clinical data, dosimetry (dose-volume histograms), and CAC data. We addressed the imbalance between AF and non-AF cases using the Synthetic Minority Oversampling Technique (SMOTE). Our machine learning model used a feature intervention technique for optimal feature representation, selecting the feature with the highest coefficient in each cardiac sub-structure after the feature selection process. On the other hand, the deep learning model was specifically designed with hybrid architectures, enabling it to handle diverse input data types by pairing them with corresponding network architectures. The effectiveness of these models was evaluated using the area under the curve (AUC), with important feature identification using coefficients in classifiers and an open-source library, Captum, for machine and deep learning models, respectively. In this study, the hybrid deep learning model demonstrated superior performance in the internal cohort, achieving an AUC of 0.823, which surpassed the machine learning model with intervention, which had an AUC of 0.801. In external validation, this deep learning model also showed more consistent results, with an AUC of 0.806, compared to the machine learning model's AUC of 0.776. Notably, in both internal and external validations, the identified key features were the maximum dose to the heart and the sinoatrial node (SAN). In this study, we developed and validated machine and deep learning models for predicting AF in NSCLC patients following CRT, finding the deep learning model to be more accurate and identifying the maximum heart dose and SAN as important feature. Table1. Performance of predictive models. [ABSTRACT FROM AUTHOR]

The standard treatment approach for locally invasive thymoma primarily involves surgical excision followed by postoperative radiation therapy (PORT). Currently, the recommended radiation therapy dose after surgery is 45-50 Gy administered over a period of 5 weeks. However, there is a lack of sufficient research to determine the appropriateness of this dose. Consequently, our study aimed to investigate the feasibility of reducing the radiation dose while maintaining the disease specific survival rate (DSS) and progression free survival (PFS), with the goal of shortening the treatment duration and minimizing potential side effects. A total of 150 patients diagnosed with stage II-III thymoma underwent radiation therapy (RT) following surgical intervention between January 2016 and June 2022. The patients received intensity-modulated radiation therapy (IMRT), and the median follow-up duration was 40.8 months. The standard RT treatment regimen consisted of 45-50 Gy administered over 25 fractions (median BED 60 Gy), while the de-escalation RT group was given 30-35 Gy administered over 10 fractions (median BED 47.25 Gy). The primary endpoint of the study was progression-free survival (PFS), with secondary endpoints encompassing overall survival (OS), disease free survival (DSS) and treatment-related toxicities. No statistically significant difference was observed between the standard group and the de-escalation group in 3-year PFS(p = 0.406) and the 3-year DSS was 100% in both groups. 2 patients in de-escalation group were dead during follow-up duration, but the cause of death was double primary cancer. Analysis of the locoregional relapse distribution revealed that all recurrences occurred outside the radiation therapy (RT) field. There were no significant correlations found between PFS and factors such as age, initial tumor size, the presence of myasthenia gravis, or pathological types. These factors also exhibited no correlation with OS. In terms of toxicity, no instances of grade II toxicity were observed in the de-escalation group, while the standard group experienced three cases of grade II toxicity, specifically radiation pneumonitis. In stage II-III thymoma patients undergoing postoperative radiotherapy (PORT), the comparison between the de-escalation group and the standard group revealed no significant differences in terms of progression-free survival (PFS), disease-free survival (DSS), and showed lower treatment-related toxicity in the de-escalation group. The implementation of a radiation dose de-escalation strategy, coupled with a reduction in treatment duration by less than half, appears to be a viable approach. Nevertheless, to firmly establish the efficacy and safety of this de-escalation dose regimen, it is imperative to extend the follow-up period. [ABSTRACT FROM AUTHOR]