Background: Bone marrow aspiration and biopsy remain the gold standard for the diagnosis of hematological diseases despite the development of flow cytometry (FCM) and molecular and gene analyses. However, the interpretation of the results is laborious and operator dependent. Furthermore, the obtained results exhibit inter- and intravariations among specialists. Therefore, it is important to develop a more objective and automated analysis system. Several deep learning models have been developed and applied in medical image analysis but not in the field of hematological histology, especially for bone marrow smear applications. Objective: The aim of this study was to develop a deep learning model (BMSNet) for assisting hematologists in the interpretation of bone marrow smears for faster diagnosis and disease monitoring. Methods: From January 1, 2016, to December 31, 2018, 122 bone marrow smears were photographed and divided into a development cohort (N=42), a validation cohort (N=70), and a competition cohort (N=10). The development cohort included 17,319 annotated cells from 291 high-resolution photos. In total, 20 photos were taken for each patient in the validation cohort and the competition cohort. This study included eight annotation categories: erythroid, blasts, myeloid, lymphoid, plasma cells, monocyte, megakaryocyte, and unable to identify. BMSNet is a convolutional neural network with the YOLO v3 architecture, which detects and classifies single cells in a single model. Six visiting staff members participated in a human-machine competition, and the results from the FCM were regarded as the ground truth. Results: In the development cohort, according to 6-fold cross-validation, the average precision of the bounding box prediction without consideration of the classification is 67.4%. After removing the bounding box prediction error, the precision and recall of BMSNet were similar to those of the hematologists in most categories. In detecting more than 5% of blasts in the validation cohort, the area under the curve (AUC) of BMSNet (0.948) was higher than the AUC of the hematologists (0.929) but lower than the AUC of the pathologists (0.985). In detecting more than 20% of blasts, the AUCs of the hematologists (0.981) and pathologists (0.980) were similar and were higher than the AUC of BMSNet (0.942). Further analysis showed that the performance difference could be attributed to the myelodysplastic syndrome cases. In the competition cohort, the mean value of the correlations between BMSNet and FCM was 0.960, and the mean values of the correlations between the visiting staff and FCM ranged between 0.952 and 0.990. Conclusions: Our deep learning model can assist hematologists in interpreting bone marrow smears by facilitating and accelerating the detection of hematopoietic cells. However, a detailed morphological interpretation still requires trained hematologists.

Background. Treating advanced colon cancer remains challenging in clinical settings because of the development of drug resistance and distant metastasis. Mechanisms underlying the metastasis of colon cancer are complex and unclear. Methods. Computational analysis was performed to determine genes associated with the exosomal long noncoding (lncRNA) plasmacytoma variant translocation 1 (PVT1)/vascular endothelial growth factor A (VEGFA) axis in patients with colon cancer. The biological importance of the exosomal lncRNA PVT1/VEGFA axis was examined in vitro by using HCT116 and LoVo cell lines and in vivo by using a patient‐derived xenograft (PDX) mouse model through knockdown (by silencing of PVT1) and overexpression (by adding serum exosomes isolated from patients with distant metastasis (M‐exo)). Results. The in silico analysis demonstrated that PVT1 overexpression was associated with poor prognosis and increased expression of metastatic markers such as VEGFA and epidermal growth factor receptor (EGFR). This finding was further validated in a small cohort of patients with colon cancer in whom increased PVT1 expression was correlated with colon cancer incidence, disease recurrence, and distant metastasis. M‐exo were enriched with PVT1 and VEGFA, and both migratory and invasive abilities of colon cancer cell lines increased when they were cocultured with M‐exo. The metastasis‐promoting effect was accompanied by increased expression of Twist1, vimentin, and MMP2. M‐exo promoted metastasis in PDX mice. In vitro silencing of PVT1 reduced colon tumorigenic properties including migratory, invasive, colony forming, and tumorsphere generation abilities. Further analysis revealed that PVT1, VEGFA, and EGFR interact with and are regulated by miR‐152‐3p. Increased miR‐152‐3p expression reduced tumorigenesis, where increased tumorigenesis was observed when miR‐152‐3p expression was downregulated. Conclusion. Exosomal PVT1 promotes colon cancer metastasis through its association with EGFR and VEGFA expression. miR‐152‐3p targets both PVT1 and VEGFA, and this regulatory pathway can be explored for drug development and as a prognostic biomarker.

Backgrounds: Severe microcytic anemia has been associated with BP changes. Aims and Objectives: Whether the erythrocyte indices are associated with long-term BPV is unknown. This study aimed to investigate the association of hemoglobin levels and erythrocyte size with long-term blood pressure variability (BPV) in young males. Methods: This study included 1,112 healthy military males, averaging 32 years of age, in Taiwan. All participants took a measurement of systolic and diastolic BP (SBP and DBP) every two-year from 2012 to 2018 (2012-14, 2014-15, 2015-16, 2016-18). Levels of hemoglobin and mean corpuscular volume (MCV) of erythrocytes were obtained at the first visit. Long-term BPV was assessed by the standard deviation (SD) and average real variability (ARV). Multivariate linear regression analysis with adjustment for the baseline BP levels and other covariates was used to elucidate the association. Results: Hemoglobin levels were borderline positively correlated with SD _DBP (β and standard errors = 0.016 (0.009), P =0.06). In those with hemoglobin levels of 10.0-13.9 g/dL, hemoglobin was negatively correlated with SD_SBP (β= -0.039 (0.018), P =0.03). In contrast, MCV levels were borderline positively correlated with SD_SBP (β =0.085 (0.052), P =0.09). In those with MCV levels <80 fL, MCV was positively correlated with SD_SBP and ARV_SBP (β= 0.445 (0.210) and 0.286 (0.149), p = 0.03 and 0.05, respectively). Conclusion: There were inconsistent patterns for the associations of erythrocyte indices with longterm BPV. We found a U-shaped relationship of hemoglobin levels with systolic BPV, whereas there was a positive linear relationship of hemoglobin and MCV levels with diastolic BPV, respectively.

We conducted a population-based cohort study enrolling patients with Stage II and III colon cancer receiving postoperative adjuvant chemotherapy with uracil and tegafur (UFT) or fluorouracil (5-FU) from the Taiwan National Health Insurance Research Database from 2000 to 2015. The outcomes of the current study were disease-free survival (DFS) and overall survival (OS). Hazard ratios (HRs) were calculated by multivariate Cox proportional hazard regression models. We compared our effectiveness results from the literature by meta-analysis, which provided the best evidence. Severe adverse events were compared in meta-analysis of reported clinical trials. In the nationwide cohort study, UFT (14,486 patients) showed DFS similar to postoperative adjuvant chemotherapy (adjusted HR 1.037; 95% confidence interval [CI] 0.954–1.126; P = .397) and OS (adjusted HR 0.964; 95% CI 0.891–1.041; P = .349) compared with the 5-FU (866 patients). Our meta-analysis confirmed the similarity of effectiveness and found the incidence of leucopaenia was statistically significantly reduced in UFT (risk ratio 0.12; 95% CI 0.02–0.67; I ² = 0%). Through our analysis, we have confirmed that UFT is a well-tolerated adjuvant therapy choice, and has similar treatment efficacy as 5-FU in terms of DFS and OS in patients with Stage II and III colon cancer.