Introduction: An increasing number of parameters can be considered when making decisions in oncology. Tumor characteristics can also be extracted from imaging through the use of radiomics and add to this wealth of clinical data. Machine learning can encompass these parameters and thus enhance clinical decision as well as radiotherapy workflow. Methods: We performed a description of machine learning applications at each step of treatment by radiotherapy in head and neck cancers. We then performed a systematic review on radiomics and machine learning outcome prediction models in head and neck cancers. Results: Machine Learning has several promising applications in treatment planning with automatic organ at risk delineation improvements and adaptative radiotherapy workflow automation. It may also provide new approaches for Normal Tissue Complication Probability models. Radiomics may provide additional data on tumors for improved machine learning powered predictive models, not only on survival, but also on risk of distant metastasis, in field recurrence, HPV status and extra nodal spread. However, most studies provide preliminary data requiring further validation. Conclusion: Promising perspectives arise from machine learning applications and radiomics based models, yet further data are necessary for their implementation in daily care.

Background: Activating mutations in the estrogen receptor 1 (ESR1) gene are recurrent mechanisms of acquired resistance to aromatase inhibitors (AI), and may be the target of other selective estrogen receptor down-regulators. To assess the clinical utility of monitoring ESR1resistant mutations, a droplet digital PCR (ddPCR)-based assay compatible with body fluids is ideal due to its cost-effectiveness and quick turnaround.Methods: We designed a multiplex ddPCR, which combines a drop-off assay, targeting the clustered hotspot mutations found in exon 8, with another pair of probes interrogating the E380Q mutation in exon 5. We assessed its sensitivity in vitro using synthetic oligonucleotides, harboring E380Q, L536R, Y537C, Y537N, Y537S or D538G mutations. Validation of the assay was performed on plasma samples from a prospective study and compared to next generation sequencing (NGS) data.Results: The multiplex ESR1-ddPCR showed a high sensitivity with a limit of detection ranging from 0.07 to 0.19% in mutant allele frequency depending on the mutation tested. The screening of plasma samples from patients with AI-resistant metastatic breast cancer identified ESR1 mutations in 29% of them with perfect concordance (and higher sensitivity) to NGS data obtained in parallel. Additionally, this test identifies patients harboring polyclonal alterations. Furthermore, the monitoring of ctDNA using this technique during treatment follow-up predicts the radiological response to palbociclib-fulvestrant. Conclusion:The multiplex ESR1-ddPCR detects, in a single reaction, the most frequent ESR1 activating mutations and is compatible with plasma samples. This method is thus suitable for real-time ESR1 mutation monitoring in large cohorts of patients..