A novel nomogram provides improved accuracy for predicting biochemical recurrence after radical prostatectomy / 中华医学杂志(英文版)

BACKGROUND@#Various prediction tools have been developed to predict biochemical recurrence (BCR) after radical prostatectomy (RP); however, few of the previous prediction tools used serum prostate-specific antigen (PSA) nadir after RP and maximum tumor diameter (MTD) at the same time. In this study, a nomogram incorporating MTD and PSA nadir was developed to predict BCR-free survival (BCRFS).@*METHODS@#A total of 337 patients who underwent RP between January 2010 and March 2017 were retrospectively enrolled in this study. The maximum diameter of the index lesion was measured on magnetic resonance imaging (MRI). Cox regression analysis was performed to evaluate independent predictors of BCR. A nomogram was subsequently developed for the prediction of BCRFS at 3 and 5 years after RP. Time-dependent receiver operating characteristic (ROC) curve and decision curve analyses were performed to identify the advantage of the new nomogram in comparison with the cancer of the prostate risk assessment post-surgical (CAPRA-S) score.@*RESULTS@#A novel nomogram was developed to predict BCR by including PSA nadir, MTD, Gleason score, surgical margin (SM), and seminal vesicle invasion (SVI), considering these variables were significantly associated with BCR in both univariate and multivariate analyses (P < 0.05). In addition, a basic model including Gleason score, SM, and SVI was developed and used as a control to assess the incremental predictive power of the new model. The concordance index of our model was slightly higher than CAPRA-S model (0.76 vs. 0.70, P = 0.02) and it was significantly higher than that of the basic model (0.76 vs. 0.66, P = 0.001). Time-dependent ROC curve and decision curve analyses also demonstrated the advantages of the new nomogram.@*CONCLUSIONS@#PSA nadir after RP and MTD based on MRI before surgery are independent predictors of BCR. By incorporating PSA nadir and MTD into the conventional predictive model, our newly developed nomogram significantly improved the accuracy in predicting BCRFS after RP.

Application of machine learning models in predicting early stone-free rate after flexible ureteroscopic lithotripsy for renal stones / 北京大学学报(医学版)

OBJECTIVE@#To establish predictive models based on random forest and XGBoost machine learning algorithm and to investigate their value in predicting early stone-free rate (SFR) after flexible ureteroscopic lithotripsy (fURL) in patients with renal stones.@*METHODS@#The clinical data of 201 patients with renal stones who underwent fURL were retrospectively investigated. According to the stone-free standard, the patients were divided into stone-free group (SF group) and stone-residual group (SR group). We compared a number of factors including patient age, body mass index (BMI), stone number, stone volume, stone density and hydronephrosis between the two groups. For low calyceal calculi, renal anatomic parameters including infundibular angle (IPA), infundibular width (IW), infundibular length (IL) and pelvic calyceal height (PCH), would be measured. We brought above potential predictive factors into random forest and XGBoost machine learning algorithm respectively to develop two predictive models. The receiver operating characteristic curve (ROC curve) was established in order to test the predictive ability of the model. Clinical data of 71 patients were collected prospectively to validate the predictive models externally.@*RESULTS@#In this study, 201 fURL operations were successfully completed. The one-phase early SFR was 61.2%. We built two predictive models based on random forest and XGBoost machine learning algorithm. The predictive variables' importance scores were obtained. The area under the ROC curve (AUROC) of the two predictive models for early stone clearance status prediction was 0.77. In the study, 71 test samples were used for external validation. The results showed that the total predictive accuracy, predictive specificity and predictive sensitivity of the random forest and XGBoost models were 75.7%, 82.6%, 60.0%, and 81.4%, 87.0%, 68.0%, respectively. The first four predictive variables in importance were stone volume, mean stone density, maximal stone density and BMI in both random forest and XGBoost predictive models.@*CONCLUSION@#The predictive models based on random forest and XGBoost machine learning algorithm can predict postoperative early stone status after fURL for renal stones accurately, which will facilitate preoperative evaluation and clinical decision-making. Stone volume, mean stone density, maximal stone density and BMI may be the important predictive factors affecting early SFR after fURL for renal stones.

Application of U-shaped convolutional neural network in auto segmentation and reconstruction of 3D prostate model in laparoscopic prostatectomy navigation / 北京大学学报(医学版)

OBJECTIVE@#To investigate the efficacy of intraoperative cognitive navigation on laparoscopic radical prostatectomy using 3D prostatic models created by U-shaped convolutional neural network (U-net) and reconstructed through Medical Image Interaction Tool Kit (MITK) platform.@*METHODS@#A total of 5 000 pieces of prostate cancer magnetic resonance (MR) imaging discovery sets with manual annotations were used to train a modified U-net, and a set of clinically demand-oriented, stable and efficient full convolutional neural network algorithm was constructed. The MR images were cropped and segmented automatically by using modified U-net, and the segmentation data were automatically reconstructed using MITK platform according to our own protocols. The modeling data were output as STL format, and the prostate models were simultaneously displayed on an android tablet during the operation to help achieving cognitive navigation.@*RESULTS@#Based on original U-net architecture, we established a modified U-net from a 201-case MR imaging training set. The network performance was tested and compared with human segmentations and other segmentation networks by using one certain testing data set. Auto segmentation of multi-structures (such as prostate, prostate tumors, seminal vesicles, rectus, neurovascular bundles and dorsal venous complex) were successfully achieved. Secondary automatic 3D reconstruction had been carried out through MITK platform. During the surgery, 3D models of prostatic area were simultaneously displayed on an android tablet, and the cognitive navigation was successfully achieved. Intra-operation organ visualization demonstrated the structural relationships among the key structures in great detail and the degree of tumor invasion was visualized directly.@*CONCLUSION@#The modified U-net was able to achieve automatic segmentations of important structures of prostate area. Secondary 3D model reconstruction and demonstration could provide intraoperative visualization of vital structures of prostate area, which could help achieve cognitive fusion navigation for surgeons. The application of these techniques could finally reduce positive surgical margin rates, and may improve the efficacy and oncological outcomes of laparoscopic prostatectomy.

Transcriptional regulation of BHLHB2 induced by PML-RARα / 中国实验血液学杂志

Objective of this study was to investigate the transcriptional regulation of BHLHB2 gene by the PML-RARα fusion protein in APL cells and reveal the pathogenesis of APL. RT-PCR was performed to detect the expression change of BHLHB2 before and after the induction of PML-RARα in PR9 cells, and its expression level after the treatment of ATRA in PR9 and APL patient derived NB4 cells. Chromatin immunoprecipitation (ChIP)-based PCR was used to analyze whether the BHLHB2 promoter could be bound by PML-RARα in vivo. A large-scale gene expression profile dataset was used to observe the expression pattern of BHLHB2 in AML. The results showed that the expression level of BHLHB2 was significantly reduced with the induction of PML-RARα and ATRA could reverse this inhibition in both PR9 and NB4 cells and increase the expression of BHLHB2. However, the expression of BHLHB2 could not be induced by ATRA in U937 cells which do not express PML-RARα. Mechanism study revealed that PML-RARα could bound to the promoter of BHLHB2 in vivo to regulate the the expression of BHLHB2. It was found that the expression of BHLHB2 was relatively lower in APL as compared with other subtypes of AML and normal bone marrow cells. It is concluded that BHLHB2 is the target of PML-RARα, and the expression of BHLHB2 is inhibited by PML-RARα through binding to its promoter in APL.

Genomic distribution of AML1-ETO fusion protein binding sites / 中国实验血液学杂志

This study was purposed to characterize the genomic distribution of the binding sites for AML1-ETO fusion protein on chromosome 2, 9 and 19, and to further gain insights into the characteristics of transcriptional regulation by AML1-ETO in acute myeloid leukemia so as to provide theoretical basis for the development of targeted therapy and optimization for treatment. Chromatin immunoprecipitation (ChIP) coupled with high density tiling arrays (chip), also known as ChIP-chip, was utilized in this study. ChIP-DNA enriched by an anti-ETO antibody and total genomic DNA of Kasumi cells were hybridized to tiling arrays, tiled through chromosome 2, 9 and 19. The ChIP enriched regions were identified using a model based analytical tool (MAT). Genomic distribution of the ChIP regions was analyzed using publicly available CEAS web server. The Gene Ontology (GO) enrichment analysis was performed to excavated the biological significance. The results indicated that a total of 588 enriched regions were identified on chromosome 2, 9 and 19 by the anti-ETO antibody. A number of the identified regions were located within enhancers (48.86%) or introns (37.35%), much smaller fractions were within proximal promoters (5.96%) or exons (5.49%). Functional enrichment analysis showed that cell proliferation and signal transduction biological pathways were enriched in potential genes of AML-ETO. It is concluded that half of the AML1-ETO binding sites are located within known transcriptional regulatory regions (promoter, 5' UTR and enhancer), while almost another half were within the sequences which were not previously reported as regulatory regions. The potential target molecular network of AML1-ETO is involved in several essential biological processes.