BACKGROUND: Head and Neck Sarcomas (HNSs) are rare malignant tumors derived from mesenchymal cells characterized by rapid growth, local aggressive behavior, and multiple re-lapses. The Rhabdomyosarcoma (RMS) and the Osteosarcoma (OS) are the more frequent HNS. The present study aims to evaluate the clinical impact of multimodal treatment in patients diagnosed with HNS. METHODS: A retrospective analysis of medical records from patients diagnosed with HNS multimodal treated, four groups were formed: a) surgery and chemotherapy (CTX); b) surgery, CTX, and RT; c) surgery and RT; d) CTX and RT exclusively. The overall survival (OSS), disease-free survival (DFS) and progression-free survival (PFS) rates were calculated using the Kaplan-Meier test. RESULTS: Among 39 patients, 23.1% were diagnosed with RMS , 66.7% with OS and 10.3% with other subtypes of HNS. Moreover, 59% showed initial clinical staging (T1/T2) and of those 35.9% were T1. Surgery was performed in 87.2% patients, RT in 35.9% and CTX in 84.7%. RT performed showed not statistically significant clinical role regarding OSS (p=0.74), DFS (p=0.09) and PFS (p=0.90) rates independent of histological subtype. Finally, OSS (p=0.89), DFS (p=0.13) and PFS (p=0.50) rates were not statistically different for patients treated with monotherapy or multimodal therapy. CONCLUSION: RT has not showed a clinically important role referred to improve OSS, DFS and PFS rates in patients with HNS, regardless of the histological subtype of the primary tumor. Multimodal treatment could influence in survival rates and locoregional control of the disease; however, longitudinal studies such as randomized clinical trials must be conducted.

Tumor entities of the Central Nervous System (CNS) are defined by the WHO classification and range from benign neoplasms to highly malignant tumors, making histopathological classification challenging. This diverse spectrum complicates clinical decision-making and the interpretation and validation of clinical trial results. The methodology based on DNA methylation profiling of CNS tumors offers greater diagnostic precision compared to traditional morphological methods. In this study, we analyzed 16 DNA samples from medulloblastomas and ependymomas, obtained from paraffin-embedded blocks from the Pathology Division archives of HCFMUSP, and with prior histological analysis. The samples were processed using the Infinium MethylationEPIC BeadChip (Illumina®) and methylation data analysis tools in R Studio. The data were submitted to the MolecularNeuropathology.org platform for methylation based classification and CNV chart generation. The classifier was consistent with the previous diagnosis in all cases, refining the diagnosis in 54% of cases to subclass, 38% to subtype, and 8% to family. The platform also provided CNV estimates for 87% of cases, revealing crucial information on genetic alterations such as gains, losses, amplifications, and gene fusions. Despite the challenges of implementation within the Brazilian Public Health System (SUS), the results demonstrate the method's robustness and reproducibility, promising invaluable support for precise diagnoses and the development of more effective therapeutic strategies.