Diabetes and further risk of cancer: a nationwide population-based study.

BACKGROUND: Individuals with diabetes have a significantly higher risk of developing various forms of cancer, and the potential biological links between these two diseases are not completely understood. METHODS: This was a longitudinal retrospective nationwide cohort study, a study design that allows us to examine the natural course of cancer development over an extended period of time with a large sample size. Initially, 3,111,975 and 22,208,395 eligible patients aged ≥ 20 years with and without diabetes, respectively, were matched by age, sex, and the Charlson comorbidity index. Ultimately, 1,751,457 patients were selected from each group. Stratified populations for diabetic retinopathy (DR) (n = 380,822) and without DR (n = 380,822) as well as proliferative DR (PDR) (n = 141,150) and non-proliferative DR (NPDR) (n = 141,150) were analyzed in this study. The main outcome measure was the first-time diagnosis of cancer during the follow-up period. RESULTS: We observed a 20% higher risk of total cancer incidence [hazard ratios (HR), 1.20; p < 0.001] in the diabetes cohort compared to the non-diabetes cohort. The highest HR was observed for liver and pancreas cancers. Moderately increased risks were observed for oral, colon, gallbladder, reproductive (female), kidney, and brain cancer. Furthermore, there was a borderline significantly increased risk of stomach, skin, soft tissue, female breast, and urinary tract (except kidney) cancers and lymphatic and hematopoietic malignancies. The stratified analysis revealed that the total cancer incidence was significantly higher in the DR cohort compared to the non-DR cohort (HR, 1.31; p < 0.001), and there was a borderline increased risk in the PDR cohort compared to the NPDR cohort (HR, 1.13; p = 0.001). CONCLUSIONS: This study provides large-scale, nationwide, population-based evidence that diabetes is independently associated with an increased risk of subsequent development of total cancer and cancer at specific sites. Notably, this risk may further increase when DR develops.

Identifying Lymph Nodes and Their Statuses from Pretreatment Computer Tomography Images of Patients with Head and Neck Cancer Using a Clinical-Data-Driven Deep Learning Algorithm.

BACKGROUND: Head and neck cancer is highly prevalent in Taiwan. Its treatment mainly relies on clinical staging, usually diagnosed from images. A major part of the diagnosis is whether lymph nodes are involved in the tumor. We present an algorithm for analyzing clinical images that integrates a deep learning model with image processing and attempt to analyze the features it uses to classify lymph nodes. METHODS: We retrospectively collected pretreatment computed tomography images and surgery pathological reports for 271 patients diagnosed with, and subsequently treated for, naïve oral cavity, oropharynx, hypopharynx, and larynx cancer between 2008 and 2018. We chose a 3D UNet model trained for semantic segmentation, which was evaluated for inference in a test dataset of 29 patients. RESULTS: We annotated 2527 lymph nodes. The detection rate of all lymph nodes was 80%, and Dice score was 0.71. The model has a better detection rate at larger lymph nodes. For those identified lymph nodes, we found a trend where the shorter the short axis, the more negative the lymph nodes. This is consistent with clinical observations. CONCLUSIONS: The model showed a convincible lymph node detection on clinical images. We will evaluate and further improve the model in collaboration with clinical physicians.