Search

The high glycolytic activity of multiple myeloma (MM) cells is the rationale for use of Positron Emission Tomography (PET) with18F-fluorodeoxyglucose ([18F]FDG) to detect both bone marrow (BM) and extramedullary disease. However, new tracers are actively searched because [18F]FDG-PET has some limitations and there is a portion of MM patients who are negative. Glutamine (Gln) addiction has been recently described as a typical metabolic feature of MM cells. Yet, the possible exploitation of Gln as a PET tracer in MM has never been assessed so far and is investigated in this study in preclinical models. Firstly, we have synthesized enantiopure (2S,4R)-4-fluoroglutamine (4-FGln) and validated it as a Gln transport analogue in human MM cell lines, comparing its uptake with that of3H-labelled Gln. We then radiosynthesized [18F]4-FGln, tested its uptake in two differentin vivomurine MM models, and checked the effect of Bortezomib, a proteasome inhibitor currently used in the treatment of MM. Both [18F]4-FGln and [18F]FDG clearly identified the spleen as site of MM cell colonization in C57BL/6 mice, challenged with syngeneic Vk12598 cells and assessed by PET. NOD.SCID mice, subcutaneously injected with human MM JJN3 cells, showed high values of both [18F]4-FGln and [18F]FDG uptake. Bortezomib significantly reduced the uptake of both radiopharmaceuticals in comparison with vehicle at post treatment PET. However, a reduction of glutaminolytic, but not of glycolytic, tumor volume was evident in mice showing the highest response to Bortezomib. Our data indicate that [18F](2S,4R)-4-FGln is a new PET tracer in preclinical MM models, yielding a rationale to design studies in MM patients.

High glycolitic activity of multiple myeloma (MM) cells is the rationale for the use of Positron Emission Tomography (PET) with 18F-fluorodeoxyglucose ([18F]FDG) to detect both medullary and extramedullary disease. However, FDG-PET has some limitations, since there is a good portion of MM patients who are false-negative. Besides enhanced glycolysis, glutamine (Gln) addiction has been recently described as a metabolic feature of MM by our group. To sustain high Gln demand, MM cells increase the expression of several Gln transporters (ASCT2, SNAT1, LAT1) and are endowed with fast Gln uptake. Yet, at variance with other Gln-addicted cancers, the possible exploitation of Gln as a PET tracer in MM has never been assessed and was investigated in this study. To this purpose, we have firstly synthesized enantiopure (2S,4R)-4-Fluoroglutamine (4-FGln) and validated it as a Gln analogue in human MM cell lines (RPMI8226 and JJN3) comparing its uptake with that of 3H-labelled Gln. The intracellular levels of 4-FGln were determined by HPLC-MS/MS employing a HILIC gradient separation and multiple reaction monitoring (MRM) detection. Both Gln and 4-FGln were actively accumulated by MM cells and exhibited a strong reciprocal competition, pointing to shared transporters. Inhibition analysis revealed that ASCT2 was the major entry route of both compounds, with minor contributions from the other transporters. However, compared with Gln, 4-FGln exhibited higher affinity for both ASCT2 and LAT1 transporters. On the basis of these results, we then tested [18F]4-FGln uptake for MM detection by Positron Emission Tomography (PET) in two different in vivo murine models. Firstly, to investigate sensitivity of human MM to [18F]4-FGln in vivo, JJN3 cells were subcutaneously injected in immunodeficient NSG mice In this xenograft model, [18F]4-FGln- and[18F]FDG-PET scans were performed after plasmacytomas became palpable and repeated after one week. All the tumours were positive for [18F]FDG and displayed [18F]4-FGln uptake with Standard Uptake Values (SUV) of 1.21±1.9 and 0.99±0.07 after 2 weeks, respectively. Thereafter, the effect of bortezomib (BOR) was investigated to evaluate the potential use of [18F]4-FGln to monitor anti-MM treatment. Ten NGS mice were injected with JJN3 cells and, after 14 days, treated twice weekly with BOR, 1mg/kg, or vehicle for two weeks. PET scans were performed before and after 5 and 12 days of BOR treatment. As expected, BOR reduced tumour size as compared to vehicle. At the first post-BOR PET scan, [18F]4-FGln (SUV mean: pre 0.85±0.31; post 0.45±0.10, P Thereafter, to mimic BOR-resistant MM in a syngeneic mouse model, C57BL/6 mice were injected intravenously with Vk12598 cells obtained from transgenic Vk*MYC mice repeatedly treated with sub-optimal doses of BOR. Upon injection into C57BL/6 mice, Vk12598 cells colonize the BM without lytic lesions and extensively colonize the spleen generating an aggressive MM that brings animals to death within five weeks. PET scans were performed with [18F]4-FGln and [18F]FDG before Vk*MYC MM cells injection and after three, four and five weeks. Blood samples for M-spike evaluation were obtained in parallel. Four weeks after MM cells injection a significant increase of both [18F]4-FGln and [18F]FDG uptake was detected in spleens (SUV mean: 1.14±0.23, P=0.018; 0.94±0.24, P= 0.005). In both MM models, the volume of distribution of [18F]4-F-Gln did not overlap that of [18F]FDG. In conclusion, our data indicate that [18F]-(2S,4R)-4-Fluoroglutamine is a new potential PET tracer in pre-clinical MM models especially of extramedullary disease, either in a BOR-sensitive or in a BOR-resistant context, supporting the exploitation of Gln addiction for diagnostic purposes in MM patients. Disclosures Giuliani: Janssen: Research Funding.

In this paper we present a Decision Support System (DSS) aimed at forecasting high demand of admission on health care structures due to environmental pollution. The computational engine of the DSS is based on Autoregressive Hidden Markov Models (AHMM). We estimate the forecasting model by analyzing the historical daily average concentrations of pollutants and the number of hospital admissions, collected from multiple data sources. Given the actual concentration of different pollutants, measured by a network of sensors, the DSS allows to forecast the demand of hospital admissions for acute diseases in the following 1 to 6 days. We tested our system on cardiovascular and respiratory diseases in the area of Milan. The performances of our system, compared with multiple linear regression, show that AHMM are a robust approach to capture the connections among health and environmental indicators.

This paper presents a framework for managing data from sensor of poor quality, withthe objective to reduce at the same time the communication load and hence energy consumption.Each node in a wireless sensor network maintains a simple local model of the data it is collectingand sends its parameters to a central location (sink), where it is executed the global monitoring.Local models are used to simulate sensor's readings, minimising the need of communication withsensors and hence the consumption of their battery; they are updated locally, when sensorreadings differ excessively from simulated data. At the sink the global model (a BayesianNetwork) is learnt on the simulated data. It is used to identify and replace anomalous readings(outliers) that a sensor should have produced and to detect anomalies missed by any single node(when communication with a sensor is interrupted). © 2010 Inderscience Enterprises Ltd.

Emergency management involves collaboration among different operators (e.g. policemen, firemen, medics) on critical and dangerous situations (e.g. fires, floods). Real-time elaborations of a large amount of information and knowledge are needed to automate control and decision making processes. In this scenario raw data are captured and processed through a chain of activities (filtering, fusion, classification) generating abstract information that is selected, diffused and presented to interested users. These activities can be turned into the interpretation of executable transformation models, abstract representations of input-output transformations which can be coded on a calculator. Executable transformation models might not be static; but they may be dynamically generated by observing and processing incoming data. From the architectural point of view the dynamic model generation implies the existence of mechanisms for the dynamic deployment of the executable transformation models. In this paper we present a software architecture aimed at providing these mechanisms. Copyright © 2009 ACM.

We consider the problem of modelling processes sequentially changing behaviour and unexpected changes that can hinder finding the best approximation function. These dynamics cannot be observed directly either because they are masked by observational noise or because the process generating them is too complex and involves too many variables. In this paper, the problem of modelling financial time series has been approached using hidden Markov models (HMMs), which have been shown to be suitable for sequential data analysis and in particular for financial time series modelling and forecasting. HMMs are essentially data-driven models that allow us to focus attention on the observation generation process, which is indeed final objective. The goal of our time series analysis model is the generation of scenarios to be included in decision models. Therefore, our focus will not be on determining the best forecast but in capturing the generation process behaviour in order to characterize its possible evolutions

The WWW is actually the most dynamic and attractive information exchange place. Finding useful information is hard due to huge data amount, varied topics and unstructured contents. In this paper we present a web browsing support system that proposes personalized contents. It is integrated in the content management system and it runs on the server hosting the site. It processes periodically site contents, extracting vectors of the most significant words. A topology tree is defined applying hierarchical clustering. During online browsing, viewed contents are processed and mapped in the vector space previously defined. The centroid of these vectors is compared with the topology tree nodes' centroids to find the most similar; its contents are presented to the user as link suggestions or dynamically created pages. Personal profile is saved after every session and included in the analysis during same user's subsequent visits, avoiding the cold start problem.

This paper presents a machine learning system to handle traffic control applications. The input of the system is a set of image sequences coming from a fixed camera. The system can be divided into two main subsystems: the first one, based on Artificial Neural Networks classifies the typology of vehicles moving within a limited image area for each frame of the sequence; the second one, based on Genetic Algorithms, takes as input the frame-by-frame classifications and reconstructs the global traffic scenario by counting the number of vehicles of each typology. This task is particularly hard when the frame rate is low. The results obtained by our system are reliable even for very low frame rate (i.e. four frames per second). Our system is currently used by a company for real-time traffic control.