Identification of nodal metastasis and tumor extranodal extension (ENE) is crucial for head and neck cancer management, but currently only can be diagnosed via postoperative pathology. Pretreatment, radiographic identification of ENE, in particular, has proven extremely difficult for clinicians, but would be greatly influential in guiding patient management. Here, we show that a deep learning convolutional neural network can be trained to identify nodal metastasis and ENE with excellent performance that surpasses what human clinicians have historically achieved. We trained a 3-dimensional convolutional neural network using a dataset of 2,875 CT-segmented lymph node samples with correlating pathology labels, cross-validated and fine-tuned on 124 samples, and conducted testing on a blinded test set of 131 samples. On the blinded test set, the model predicted ENE and nodal metastasis each with area under the receiver operating characteristic curve (AUC) of 0.91 (95%CI: 0.85–0.97). The model has the potential for use as a clinical decision-making tool to help guide head and neck cancer patient management.

No abstract

As the prognosis of metastatic non-small cell lung cancer (NSCLC) patients is constantly improving with advances in systemic therapies (immune checkpoint blockers and new generation of targeted molecular compounds), more attention should be paid to the diagnosis and management of treatments-related long-term secondary effects. Brain metastases (BM) occur frequently in the natural history of NSCLC and stereotactic radiation therapy (SRT) is one of the main efficient local non-invasive therapeutic methods. However, SRT may have some disabling side effects. Brain radiation necrosis (RN) represents one of the main limiting toxicities, generally occurring from 6 months to several years after treatment. The diagnosis of RN itself may be quite challenging, as conventional imaging is frequently not able to differentiate RN from BM recurrence. Retrospective studies have suggested increased incidence rates of RN in NSCLC patients with oncogenic driver mutations [epidermal growth factor receptor (EGFR) mutated or anaplastic lymphoma kinase (ALK) positive] or receiving tyrosine kinase inhibitors. The risk of immune checkpoint inhibitors in contributing to RN remains controversial. Treatment modalities for RN have not been prospectively compared. Those include surveillance, corticosteroids, bevacizumab and local interventions (minimally invasive laser interstitial thermal ablation or surgery). The aim of this review is to describe and discuss possible RN management options in the light of the newly available literature, with a particular focus on NSCLC patients.