Background: Painful spinal bone metastases (PSBMs) patients regularly receive palliative radiation therapy (RT) with response rates in about 2 of 3 patients. In this exploratory study, we evaluated the value of machine learning (ML) models based on radiomic, semantic and clinical features to predict complete pain response. Methods: Gross tumour volumes (GTV) and clinical target volumes (CTV) of 261 PSBMs were segmented on planning computed tomography (CT) scans. Radiomic, semantic and clinical features were collected for all patients. Random forest (RFC) and support vector machine (SVM) classifiers were compared using repeated nested cross-validation.Results: The best radiomic classifier was trained on CTV with an area under the receiver-operator curve (AUROC) of 0.62 ± 0.01 (RFC; 95% confidence interval). The semantic model achieved a comparable AUROC of 0.63 ± 0.01 (RFC), significantly below the clinical model (SVM, AUROC: 0.80 ± 0.01); and slightly lower than the spinal instability neoplastic score (SINS; LR, AUROC: 0.65 ± 0.01). A combined model did not improve performance (AUROC: 0,74 ± 0,01).Conclusions: We could demonstrate that radiomic and semantic analyses of planning CTs allowed for limited prediction of therapy response to palliative RT. ML predictions based on established clinical parameters achieved the best results.

Background: An aberrant cellular microenvironment characterized by pathological cells or inflammation represents an added risk factor across various cancer types. While the significance of chronic inflammation in the development of most diffuse tumors has been extensively studied, an exception to this analysis exists in the context of Chondrosarcomas. Chondrosarcomas represent 20–30% of all skeletal sarcomas and have an estimated incidence of 1 in 100,000 worldwide. The mean age of presentation is 50, with over 70% of patients over the age of 40 at the time of diagnosis. The aim of this retrospective study was to analyze the role of C-reactive protein (CRP) as a prognostic factor in relation to the histopathological findings for Chondrosarcoma. Methods: This retrospective study included 73 patients with confirmed Chondrosarcoma diagnosis treated between 2004 and 2019. Preoperative CRP determination was assessed in mg/dL with a non-pathological value below 0.5 mg/dL. Disease-free survival time was calculated as the time between the initial diagnosis and an event like local recurrence or metastasis. Follow-up status was described in death of disease, no evidence of disease and alive with disease. The exclusion criteria of this study included insufficient laboratory values, lack of information regarding the follow-up status and incomplete histopathological report. Results: The calculate risk estimation of a reduced follow-up time was 2,25 time higher in the patients with an CRP level >0.5 mg/dL (HR 2,25 and 95% CI 1,13-4,45) and 3 times higher in patients with a tumour size >pT2 (HR 3 and 95% CI 1,59-5,92). We can easily confirm that the most elevated risk for a reduce follow-up time relies in chondrosarcomas high grading, preoperative pathological CRP- level and a size >8 cm. Conclusions: In Chondrosarcoma cases, we can consider CRP pretreatment value >0.5 mg/dL a sensitive prognostic and sensitive risk factor for distant metastasis.