Purpose Multiple smaller studies have demonstrated an association between overall survival and lymph node (LN) count from neck dissection in patients with head and neck cancer. This is a large cohort study to examine these associations by using a national cancer database. Patients and Methods The National Cancer Database was used to identify patients who underwent upfront nodal dissection for mucosal head and neck squamous cell carcinoma between 2004 and 2013. Patients were stratified by LN count into those with < 18 nodes and those with ≥ 18 nodes on the basis of prior work. A multivariable Cox proportional hazards regression model was constructed to predict hazard of mortality. Stratified models predicted hazard of mortality both for patients who were both node negative and node positive. Results There were 45,113 patients with ≥ 18 LNs and 18,865 patients with < 18 LNs examined. The < 18 LN group, compared with the ≥ 18 LN group, had more favorable tumor characteristics, with a lower proportion of T3 and T4 lesions (27.9% v 39.8%), fewer patients with positive nodes (46.6% v 60.5%), and lower rates of extracapsular extension (9.3% v 15.1%). Risk-adjusted Cox models predicting hazard of mortality by LN count showed an 18% increased hazard of death for patients with < 18 nodes examined (hazard ratio [HR] 1.18; 95% CI, 1.13 to 1.22). When stratified by clinical nodal stage, there was an increased hazard of death in both groups (node negative: HR, 1.24; 95% CI, 1.17 to 1.32; node positive: HR, 1.12; 95% CI, 1.05 to 1.19). Conclusion The results of our study demonstrate a significant overall survival advantage in both patients who are clinically node negative and node positive when ≥ 18 LNs are examined after neck dissection, which suggests that LN count is a potential quality metric for neck dissection.

Purpose: In locally advanced p16+ oropharyngeal squamous cell carcinoma (OPSCC), (i) to investigate kinetics of human papillomavirus (HPV) circulating tumor DNA (ctDNA) and association with tumor progression after chemoradiation, and (ii) to compare the predictive value of ctDNA to imaging biomarkers of MRI and FDG-PET. Experimental Design: Serial blood samples were collected from patients with AJCC8 stage III OPSCC (n = 34) enrolled on a randomized trial: pretreatment; during chemoradiation at weeks 2, 4, and 7; and posttreatment. All patients also had dynamic-contrast-enhanced and diffusion-weighted MRI, as well as FDG-PET scans pre-chemoradiation and week 2 during chemoradiation. ctDNA values were analyzed for prediction of freedom from progression (FFP), and correlations with aggressive tumor subvolumes with low blood volume (TVLBV) and low apparent diffusion coefficient (TVLADC), and metabolic tumor volume (MTV) using Cox proportional hazards model and Spearman rank correlation. Results: Low pretreatment ctDNA and an early increase in ctDNA at week 2 compared with baseline were significantly associated with superior FFP (P &lt; 0.02 and P &lt; 0.05, respectively). At week 4 or 7, neither ctDNA counts nor clearance were significantly predictive of progression (P = 0.8). Pretreatment ctDNA values were significantly correlated with nodal TVLBV, TVLADC, and MTV pre-chemoradiation (P &lt; 0.03), while the ctDNA values at week 2 were correlated with these imaging metrics in primary tumor. Multivariate analysis showed that ctDNA and the imaging metrics performed comparably to predict FFP. Conclusions: Early ctDNA kinetics during definitive chemoradiation may predict therapy response in stage III OPSCC.

Lymph node metastasis (LNM) is a significant prognostic factor in patients with head and neck cancer, and the ability to predict it accurately is essential to optimizing treatment. Positron emission tomography (PET) and computed tomography (CT) imaging are routinely used to identify LNM. Although large or highly active lymph nodes (LNs) have a high probability of being positive, identifying small or less reactive LNs is challenging. The accuracy of LNM identification strongly depends on the physician's experience, so an automatic prediction model for LNM based on CT and PET images is warranted to assist LMN identification across care providers and facilities. Radiomics and deep learning are the two promising imaging-based strategies for node malignancy prediction. Radiomics models are built based on handcrafted features, while deep learning learns the features automatically. To build a more reliable model, we proposed a hybrid predictive model that takes advantages of both radiomics and deep learning based strategies. We designed a new many-objective radiomics (MO-radiomics) model and a 3-dimensional convolutional neural network (3D-CNN) that fully utilizes spatial contextual information, and we fused their outputs through an evidential reasoning (ER) approach. We evaluated the performance of the hybrid method for classifying normal, suspicious and involved LNs. The hybrid method achieves an accuracy (ACC) of 0.92 while XmasNet and Radiomics methods achieve 0.79 and 0.79, respectively. The hybrid method provides a more accurate way for predicting LNM using PET and CT.

Glioblastoma is the most common primary brain tumor in adults. Standard therapy depends on patient age and performance status but principally involves surgical resection followed by a 6-wk course of radiation therapy given concurrently with temozolomide chemotherapy. Despite such treatment, prognosis remains poor, with a median survival of 16 mo. Challenges in achieving local control, maintaining quality of life, and limiting toxicity plague treatment strategies for this disease. Radiotherapy dose intensification through hypofractionation and stereotactic radiosurgery is a promising strategy that has been explored to meet these challenges. We review the use of hypofractionated radiotherapy and stereotactic radiosurgery for patients with newly diagnosed and recurrent glioblastoma.

Lymph node metastasis (LNM) is a significant prognostic factor in patients with head and neck cancer, and the ability to predict it accurately is essential for treatment optimization. PET and CT imaging are routinely used for LNM identification. However, uncertainties of LNM always exist especially for small size or reactive nodes. Radiomics and deep learning are the two preferred imaging-based strategies for node malignancy prediction. Radiomics models are built based on handcrafted features, and deep learning can learn the features automatically. We proposed a hybrid predictive model that combines many-objective radiomics (MO-radiomics) and 3-dimensional convolutional neural network (3D-CNN) through evidential reasoning (ER) approach. To build a more reliable model, we proposed a new many-objective radiomics model. Meanwhile, we designed a 3D-CNN that fully utilizes spatial contextual information. Finally, the outputs were fused through the ER approach. To study the predictability of the two modalities, three models were built for PET, CT, and PET&CT. The results showed that the model performed best when the two modalities were combined. Moreover, we showed that the quantitative results obtained from the hybrid model were better than those obtained from MO-radiomics and 3D-CNN.