Development and validation of a machine-learning model for prediction of shoulder dystocia.

OBJECTIVES: To develop a machine-learning (ML) model for prediction of shoulder dystocia (ShD) and to externally validate the model's predictive accuracy and potential clinical efficacy in optimizing the use of Cesarean delivery in the context of suspected macrosomia. METHODS: We used electronic health records (EHR) from the Sheba Medical Center in Israel to develop the model (derivation cohort) and EHR from the University of California San Francisco Medical Center to validate the model's accuracy and clinical efficacy (validation cohort). Subsequent to application of inclusion and exclusion criteria, the derivation cohort included 686 singleton vaginal deliveries, of which 131 were complicated by ShD, and the validation cohort included 2584 deliveries, of which 31 were complicated by ShD. For each of these deliveries, we collected maternal and neonatal delivery outcomes coupled with maternal demographics, obstetric clinical data and sonographic fetal biometry. Biometric measurements and their derived estimated fetal weight were adjusted (aEFW) according to gestational age at delivery. A ML pipeline was utilized to develop the model. RESULTS: In the derivation cohort, the ML model provided significantly better prediction than did the current clinical paradigm based on fetal weight and maternal diabetes: using nested cross-validation, the area under the receiver-operating-characteristics curve (AUC) of the model was 0.793 ± 0.041, outperforming aEFW combined with diabetes (AUC = 0.745 ± 0.044, P = 1e-16 ). The following risk modifiers had a positive beta that was > 0.02, i.e. they increased the risk of ShD: aEFW (beta = 0.164), pregestational diabetes (beta = 0.047), prior ShD (beta = 0.04), female fetal sex (beta = 0.04) and adjusted abdominal circumference (beta = 0.03). The following risk modifiers had a negative beta that was < -0.02, i.e. they were protective of ShD: adjusted biparietal diameter (beta = -0.08) and maternal height (beta = -0.03). In the validation cohort, the model outperformed aEFW combined with diabetes (AUC = 0.866 vs 0.784, P = 0.00007). Additionally, in the validation cohort, among the subgroup of 273 women carrying a fetus with aEFW ≥ 4000 g, the aEFW had no predictive power (AUC = 0.548), and the model performed significantly better (0.775, P = 0.0002). A risk-score threshold of 0.5 stratified 42.9% of deliveries to the high-risk group, which included 90.9% of ShD cases and all cases accompanied by maternal or newborn complications. A more specific threshold of 0.7 stratified only 27.5% of the deliveries to the high-risk group, which included 63.6% of ShD cases and all those accompanied by newborn complications. CONCLUSION: We developed a ML model for prediction of ShD and, in a different cohort, externally validated its performance. The model predicted ShD better than did estimated fetal weight either alone or combined with maternal diabetes, and was able to stratify the risk of ShD and neonatal injury in the context of suspected macrosomia. Copyright © 2019 ISUOG. Published by John Wiley & Sons Ltd.

Comprehensive transcriptomic analysis of cell lines as models of primary tumors across 22 tumor types.

Cancer cell lines are a cornerstone of cancer research but previous studies have shown that not all cell lines are equal in their ability to model primary tumors. Here we present a comprehensive pan-cancer analysis utilizing transcriptomic profiles from The Cancer Genome Atlas and the Cancer Cell Line Encyclopedia to evaluate cell lines as models of primary tumors across 22 tumor types. We perform correlation analysis and gene set enrichment analysis to understand the differences between cell lines and primary tumors. Additionally, we classify cell lines into tumor subtypes in 9 tumor types. We present our pancreatic cancer results as a case study and find that the commonly used cell line MIA PaCa-2 is transcriptionally unrepresentative of primary pancreatic adenocarcinomas. Lastly, we propose a new cell line panel, the TCGA-110-CL, for pan-cancer studies. This study provides a resource to help researchers select more representative cell line models.

Targeted repositioning identifies drugs that increase fibroblast growth factor 20 production and protect against 6-hydroxydopamine-induced nigral cell loss in rats.

Endogenous fibroblast growth factor 20 (FGF20) supports maintenance of dopaminergic neurones within the nigrostriatal pathway. Moreover, direct intracerebral infusion of FGF20 protects against nigrostriatal tract loss in the 6-hydroxydopamine lesion rat model of Parkinson's disease. Increasing endogenous FGF20 production might provide a less-invasive, more translational way of providing such protection. Accordingly, we adopted a targeted repositioning approach to screen for candidate FDA-approved drugs with potential to enhance endogenous FGF20 production in brain. In silico interrogation of the Broad Institute's Connectivity Map database (CMap), revealed 50 candidate drugs predicted to increase FGF20 transcription, 16 of which had profiles favourable for use in Parkinson's disease. Of these, 11 drugs were found to significantly elevate FGF20 protein production in MCF-7 cells, between two- and four-fold. Four drugs were selected for examination in vivo. Following oral dosing in rats for 7 days, salbutamol and triflusal, but not dimethadione or trazodone, significantly elevated FGF20 levels in the nigrostriatal tract. Preliminary examination in the unilateral 6-hydroxydopamine-lesioned rat revealed a modest but significant protection against nigral cell loss with both drugs. Our data demonstrate the power of targeted repositioning as a method to identify existing drugs that may combat disease progression in Parkinson's by boosting FGF20 levels.

Deposition of (236)U from atmospheric nuclear testing in Washington state (USA) and the Pechora region (Russian Arctic).

Stratospheric fallout-derived (236)U has been detected by sector field ICPMS at two field locations for which our laboratory possessed available archived samples: A) four soil cores from Washington state (northwestern USA) and B) sediment cores from three small lakes in the Pechora region (Russian Arctic). Four Washington state soil cores exhibit (236)U inventories of 8.1 ± 1.3, 11.1 ± 0.9, 18 ± 2, and 30.2 ± 3.9 Tatoms/m(2); the respective (239)Pu contents are 52.9 ± 3.5, 67 ± 3, 71 ± 2, and 151 ± 2 Tatoms/m(2). A (236)U/(239)Pu atom ratio of 0.19 ± 0.04 (1 SD) has been determined from the Washington state soil cores. The three Pechora region lake cores each exhibit coincident maxima in their (236)U and (239)Pu atom concentration profiles. The (236)U/(238)U atom ratios are controlled by two independent factors; (236)U is from fallout deposition and (238)U concentrations are a property of the geochemical distribution of naturally occurring U. A (236)U/(238)U atom ratio as high as 8.9 × 10(-6) has been observed for acid-leached soils containing Pu solely derived from bomb-test fallout. Accordingly, a non-zero (236)U background from stratospheric fallout must be recognized and taken into account when detectable (236)U is used to infer specific local or regional influences of reactor-irradiated U.