A decision support system to facilitate management of patients with acute gastrointestinal bleeding

Objective: To develop a model to predict the bleeding source and identify the cohort amongst patients with acute gastrointestinal bleeding (GIB) who require urgent intervention, including endoscopy. Patients with acute GIB, an unpredictable event, are most commonly evaluated and managed by non-gastroenterologists. Rapid and consistently reliable risk stratification of patients with acute GIB for urgent endoscopy may potentially improve outcomes amongst such patients by targeting scarce health-care resources to those who need it the most. Design and methods: Using ICD-9 codes for acute GIB, 189 patients with acute GIB and all. available data variables required to develop and test models were identified from a hospital medical records database. Data on 122 patients was utilized for development of the model and on 67 patients utilized to perform comparative analysis of the models. Clinical data such as presenting signs and symptoms, demographic data, presence of co-morbidities, laboratory data and corresponding endoscopic diagnosis and outcomes were collected. Clinical data and endoscopic diagnosis collected for each patient was utilized to retrospectively ascertain optimal management for each patient. Clinical presentations and corresponding treatment was utilized as training examples. Eight mathematical models including artificial neural network (ANN), support vector machine (SVM), k-nearest neighbor, linear discriminant analysis (LDA), shrunken centroid (SC), random forest (RF), logistic regression, and boosting were trained and tested. The performance of these models was compared using standard statistical analysis and ROC curves. Results: Overall the random forest model best predicted the source, need for resuscitation, and disposition with accuracies of approximately 80% or higher (accuracy for endoscopy was greater than 75%). The area under ROC curve for RF was greater than 0.85, indicating excellent performance by the random forest model Conclusion: While most mathematical models are effective as a decision support system for evaluation and management of patients with acute GIB, in our testing, the RF model consistently demonstrated the best performance. Amongst patients presenting with acute GIB, mathematical models may facilitate the identification of the source of GIB, need for intervention and allow optimization of care and healthcare resource allocation; these however require further validation. (c) 2007 Elsevier B.V. All rights reserved.

Autologous non-myeloablative haemopoietic stem cell transplantation in relapsing-remitting multiple sclerosis: a phase I/II study

Background Autologous non-myeloablative haemopoietic stem cell transplantation is a method to deliver intense immune suppression. We evaluated the safety and clinical outcome of autologous non-myeloablative haemopoietic stem cell transplantation in patients with retapsing-remitting multiple sclerosis (MS) who had not responded to treatment with interferon beta. Methods Eligible patients had relapsing-remitting MS, attended Northwestern Memorial Hospital, and despite treatment with interferon beta had had two corticosteroid-treated relapses within the previous 12 months, or one relapse and gadolinium-enhancing lesions seen on MRI and separate from the relapse. Peripheral blood haemopoietic stem cells were mobilised with 2 g per m(2) cyclophosphamide and 10 mu g per kg per day filgrastim. The conditioning regimen for the haemopoietic stem cells was 200 mg per kg cyclophosphamide and either 20 mg alemtuzumab or 6 mg per kg rabbit antithymocyte globulin. Primary outcomes were progression-free survival and reversal of neurological disability at 3 years post-transplantation. We also sought to investigate the safety and tolerability of autologous non-myeloablative haemopoietic stem cell transplantation. Findings Between January 2003, and February, 2005, 21 patients were treated. Engraftment of white blood cells and platelets was on median day 9 (range day 8-11) and patients were discharged from hospital on mean day 11 (range day 8-13). One patient had diarrhoea due to Clostridium difficile and two patients had dermatomal zoster. Two of the 17 patients receiving alemtuzumab developed late immune thrombocytopenic purpura that remitted with standard therapy. 17 of 21 patients (81%) improved by at least 1 point on the Kurtzke expanded disability status scale (EDSS), and five patients (24%) relapsed but achieved remission after further immunosuppression. After a mean of 37 months (range 24-48 months), all patients were free from progression (no deterioration in EDSS score), and 16 were free of relapses. Significant improvements were noted in neurological disability, as determined by EDSS score (p<0.0001), neurological rating scale score (p=0.0001), paced auditory serial addition test (p=0.014), 25-foot walk (p<0.0001), and quality of life, as measured with the short form-36 (SF-36) questionnaire (p<0.0001). Interpretation Non-myeloablative autologous haemopoietic stem cell transplantation in patients with relapsing-remitting MS reverses neurological deficits, but these results need to be confirmed in a randomised trial.

A novel role for MNTB neuron dendrites in regulating action potential amplitude and cell excitability during repetitive firing

Principal cells of the medial nucleus of the trapezoid body (MNTB) are simple round neurons that receive a large excitatory synapse (the calyx of Held) and many small inhibitory synapses on the soma. Strangely, these neurons also possess one or two short tufted dendrites, whose function is unknown. Here we assess the role of these MNTB cell dendrites using patch-clamp recordings, imaging and immunohistochemistry techniques. Using outside-out patches and immunohistochemistry, we demonstrate the presence of dendritic Na(+) channels. Current-clamp recordings show that tetrodotoxin applied onto dendrites impairs action potential (AP) firing. Using Na(+) imaging, we show that the dendrite may serve to maintain AP amplitudes during high-frequency firing, as Na(+) clearance in dendritic compartments is faster than axonal compartments. Prolonged high-frequency firing can diminish Na(+) gradients in the axon while the dendritic gradient remains closer to resting conditions; therefore, the dendrite can provide additional inward current during prolonged firing. Using electron microscopy, we demonstrate that there are small excitatory synaptic boutons on dendrites. Multi-compartment MNTB cell simulations show that, with an active dendrite, dendritic excitatory postsynaptic currents (EPSCs) elicit delayed APs compared with calyceal EPSCs. Together with high- and low-threshold voltage-gated K(+) currents, we suggest that the function of the MNTB dendrite is to improve high-fidelity firing, and our modelling results indicate that an active dendrite could contribute to a `dual` firing mode for MNTB cells (an instantaneous response to calyceal inputs and a delayed response to non-calyceal dendritic excitatory postsynaptic potentials).

The Genome Sequence of Taurine Cattle: A Window to Ruminant Biology and Evolution

To understand the biology and evolution of ruminants, the cattle genome was sequenced to about sevenfold coverage. The cattle genome contains a minimum of 22,000 genes, with a core set of 14,345 orthologs shared among seven mammalian species of which 1217 are absent or undetected in noneutherian (marsupial or monotreme) genomes. Cattle-specific evolutionary breakpoint regions in chromosomes have a higher density of segmental duplications, enrichment of repetitive elements, and species-specific variations in genes associated with lactation and immune responsiveness. Genes involved in metabolism are generally highly conserved, although five metabolic genes are deleted or extensively diverged from their human orthologs. The cattle genome sequence thus provides a resource for understanding mammalian evolution and accelerating livestock genetic improvement for milk and meat production.

Coordinated regulation of follicle development by germ and somatic cells

The continuum of folliculogenesis begins in the fetal ovary with the differentiation of the oogonia and their isolation within the primordial follicles. Primordial follicle activation is an enigmatic process, whereby some follicles enter the growing pool to become primary follicles, thereby embarking on an irreversible progression towards ovulation or atresia. This process is under the coordinated regulation of factors from the oocyte itself, as well as from the somatic cells of the ovary, in particular the theca and granulosa cells, which are structural components of the follicle. These two influences provide the principal stimuli for the growth of the follicle to the late preantral or early antral stage of development. The endocrine effects of the gonadotrophins FSH and LH are essential to the continued progression of the follicle and most atresia can be attributed to the failure to receive or process the gonadotrophin signals. The peri-ovulatory state has received intensive investigation recently, demonstrating a coordinated role for gonadotrophins, steroids, epidermal growth factor family proteins and prostaglandins. Thus, a complex programme of coordinated interaction of governing elements from both germ and somatic cell sources is required for successful follicle development.