ABSTRACT
BACKGROUND: Muscle depletion negatively impacts treatment efficacy and survival rates in cancer. Prevention and timely treatment of muscle loss require prediction of patients at risk. We aimed to investigate the potential of skeletal muscle radiomic features to predict future muscle loss. METHODS: A total of 116 patients with stage IV non-small cell lung cancer included in a randomised controlled trial (NCT01171170) studying the effect of nitroglycerin added to paclitaxel-carboplatin-bevacizumab were enrolled. In this post hoc analysis, muscle cross-sectional area and radiomic features were extracted from computed tomography images obtained before initiation of chemotherapy and shortly after administration of the second cycle. For internal cross-validation, the cohort was randomly split in a training set and validation set 100 times. We used least absolute shrinkage and selection operator method to select features that were most significantly associated with muscle loss and an area under the curve (AUC) for model performance. RESULTS: Sixty-nine patients (59%) exhibited loss of skeletal muscle. One hundred ninety-three features were used to construct a prediction model for muscle loss. The average AUC was 0.49 (95% confidence interval [CI]: 0.36, 0.62). Differences in intensity and texture radiomic features over time were seen between patients with and without muscle loss. CONCLUSIONS: The present study shows that skeletal muscle radiomics did not predict future muscle loss during chemotherapy in non-small cell lung cancer. Differences in radiomic features over time might reflect myosteatosis. Future imaging analysis combined with muscle tissue analysis in patients and in experimental models is needed to unravel the biological processes linked to the radiomic features.
Subject(s)
Antineoplastic Combined Chemotherapy Protocols/therapeutic use , Carcinoma, Non-Small-Cell Lung/drug therapy , Lung Neoplasms/drug therapy , Muscle, Skeletal/pathology , Tomography, X-Ray Computed/methods , Area Under Curve , Bevacizumab/administration & dosage , Carboplatin/administration & dosage , Carcinoma, Non-Small-Cell Lung/diagnostic imaging , Carcinoma, Non-Small-Cell Lung/pathology , Cohort Studies , Cross-Sectional Studies , Female , Follow-Up Studies , Humans , Lung Neoplasms/diagnostic imaging , Lung Neoplasms/pathology , Male , Middle Aged , Muscle, Skeletal/diagnostic imaging , Muscle, Skeletal/drug effects , Neoplasm Staging , Nitroglycerin/administration & dosage , Paclitaxel/administration & dosage , Survival RateABSTRACT
High-dose-rate brachytherapy is a tumor treatment method where a highly radioactive source is brought in close proximity to the tumor. In this paper we develop a simulated annealing algorithm to optimize the dwell times at preselected dwell positions to maximize tumor coverage under dose-volume constraints on the organs at risk. Compared to existing algorithms, our algorithm has advantages in terms of speed and objective value and does not require an expensive general purpose solver. Its success mainly depends on exploiting the efficiency of matrix multiplication and a careful selection of the neighboring states. In this paper we outline its details and make an in-depth comparison with existing methods using real patient data.
