Search

This study retrospectively analyzed the performance of artificial neural networks (ANN) to predict overall survival (OS) or locoregional failure (LRF) in HNSCC patients undergoing radiotherapy, based on 2-[18F]FDG PET/CT and clinical covariates. We compared predictions relying on three different sets of features, extracted from 230 patients. Specifically, (i) an automated feature selection method independent of expert rating was compared with (ii) clinical variables with proven influence on OS or LRF and (iii) clinical data plus expert-selected SUV metrics. The three sets were given as input to an artificial neural network for outcome prediction, evaluated by Harrell's concordance index (HCI) and by testing stratification capability. For OS and LRF, the best performance was achieved with expert-based PET-features (0.71 HCI) and clinical variables (0.70 HCI), respectively. For OS stratification, all three feature sets were significant, whereas for LRF only expert-based PET-features successfully classified low vs. high-risk patients. Based on 2-[18F]FDG PET/CT features, stratification into risk groups using ANN for OS and LRF is possible. Differences in the results for different feature sets confirm the relevance of feature selection, and the key importance of expert knowledge vs. automated selection. [ABSTRACT FROM AUTHOR]

Simple Summary: Stereotactic body radiotherapy is used in the treatment of liver tumors. However, adjacent organs at risk (OAR) frequently limit the applicable dose to the target volume. The introduction of hybrid magnetic resonance imaging (MRI)-guided radiotherapy systems may allow dose escalation strategies with better OAR sparing due to improved soft tissue visualization, adaptive treatment planning and real-time motion management. Here we report the feasibility and early results of online adaptive MR-guided radiotherapy of primary and secondary liver tumors in eleven patients. The treatment was feasible and successfully completed in all patients. After a median follow-up of five months, no local failure occurred and no ≥ grade 2 toxicity was observed. The technique should be compared to conventional SBRT in further studies to assess the advantages of the technique. Purpose: To assess the feasibility and early results of online adaptive MR-guided radiotherapy (oMRgRT) of liver tumors. Methods: We retrospectively examined consecutive patients with primary or secondary liver lesions treated at our institution using a 0.35T hybrid MR-Linac (Viewray Inc., Mountain View, CA, USA). Online-adaptive treatment planning was used to account for interfractional anatomical changes, and real-time intrafractional motion management using online 2D cine MRI was performed using a respiratory gating approach. Treatment response and toxicity were assessed during follow-up. Results: Eleven patients and a total of 15 lesions were evaluated. Histologies included cholangiocarcinomas and metastases of neuroendocrine tumors, colorectal carcinomas, sarcomas and a gastrointestinal stroma tumor. The median BED10 of the PTV prescription doses was 84.4 Gy (range 59.5–112.5 Gy) applied in 3–5 fractions and the mean GTV BED10 was in median 147.9 Gy (range 71.7–200.5 Gy). Online plan adaptation was performed in 98% of fractions. The median overall treatment duration was 53 min. The treatment was feasible and successfully completed in all patients. After a median follow-up of five months, no local failure occurred and no ≥ grade two toxicity was observed. OMRgRT resulted in better PTV coverage and fewer OAR constraint violations. Conclusion: Early results of MR-linac based oMRgRT for the primary and secondary liver tumors are promising. The treatment was feasible in all cases and well tolerated with minimal toxicity. The technique should be compared to conventional SBRT in further studies to assess the advantages of the technique. [ABSTRACT FROM AUTHOR]