Prevalence, risk factors and clinical implications of malnutrition in French Comprehensive Cancer Centres.

BACKGROUND: This epidemiological observational study aimed at determining the prevalence of malnutrition in non-selected adults with cancer, to identify risk factors of malnutrition and correlate the results with length of stay and 2-month mortality. METHODS: This prospective multicentre 1-day study conducted in 17 French Comprehensive Cancer Centres included 1545 patients. Body mass index (BMI), weight loss (WL) in the past 6 months and age were routinely recorded according to the French national recommendations for hospitalised patients; malnutrition was rated as absent, moderate or severe according to the level of WL and BMI. Age, sex, tumour site, type of hospitalisation and treatment, disease stage, World Health Organisation performance status (PS) and antibiotic therapy were the potential malnutrition risk factors tested. Follow-up at 2 months allowed to determine the correlation with length of stay and mortality. RESULTS: Malnutrition was reported in 30.9% of patients, and was rated as severe in 12.2%. In multivariate analysis, only pre-existing obesity (BMI> or =30), PS > or =2 and head-and-neck or upper digestive cancers were associated with increased risk of malnutrition. Antibiotics use was significantly higher in malnourished patients (35.5 vs 22.8%; P<0.001). Severe malnutrition was independently associated with mortality. The median length of stay was 19.3+/-19.4 days for malnourished patients vs 13.3+/-19.4 days for others (P<0.0001). CONCLUSION: In French Comprehensive Cancer Centres, one out of three cancer patients are malnourished and this was associated with a longer length of stay. Pre-existing obesity could be identified as a new risk factor for malnutrition in our cancer patient population perhaps because of a misidentification or a delay in nutrition support in this category of patients.

Hamiltonian description of a self-consistent interaction between charged particles and electromagnetic waves.

The Hamiltonian description of the self-consistent interaction between an electromagnetic plane wave and a copropagating beam of charged particles is considered. We show how the motion can be reduced to a one-dimensional Hamiltonian model (in a canonical setting) from the Vlasov-Maxwell Poisson brackets. The reduction to this paradigmatic Hamiltonian model is performed using a Lie algebraic formalism which allows us to preserve the Hamiltonian character at each step of the derivation.

Targeted mixing in an array of alternating vortices.

Transport and mixing properties of passive particles advected by an array of vortices are investigated. Starting from the integrable case, it is shown that a special class of perturbations allows one to preserve separatrices which act as effective transport barriers, while triggering chaotic advection. In this setting, mixing within the two dynamical barriers is enhanced while long range transport is prevented. A numerical analysis of mixing properties depending on parameter values is performed; regions for which optimal mixing is achieved are proposed. Robustness of the targeted mixing properties regarding errors in the applied perturbation are considered, as well as slip/no-slip and/or boundary conditions for the flow.

Mean value and fluctuations in a model of diffusion in porous media.

We consider a stochastic model for the diffusion in a porous media. For a case where the average satisfies an anomalous diffusion equation, we investigate the behavior of the realizations around the mean value. The most relevant result of our work is that, although the concentration corresponding to each realization diffuses normally for large times, it experiences large deviations from the mean value during intermediate times. As a consequence, the experimental measurements will always depart from the average value of the realizations (with respect to the stochastic process) for unpredictable times.

Chaotic advection and targeted mixing.

The advection of passive tracers in an oscillating vortex chain is investigated. It is shown that by adding a suitable perturbation to the ideal flow, the induced chaotic advection exhibits two remarkable properties which do not hold in the case of a generic perturbation: Particles remain trapped within a specific domain bounded by two oscillating barriers (suppression of chaotic transport along the channel), and the stochastic sea seems to cover this whole bounded domain (enhancement of mixing within the rolls).

Anomalous transport fluctuations in a model of irregular media.

We consider a diffusion model with stochastic porosity for which the average solution exhibits an abnormal transport. In this paper we investigate the relation of such an anomalous diffusive property of the mean value with the behavior of the solution corresponding to each realization of the stochastic porosity. Such a solution will correspond to the actual measurements in an experiment made on a particular tube. The most relevant result of our work is that, although the concentration corresponding to each realization diffuses normally for large times, it experiments on large deviations from the mean value during intermediate times.

Channeling chaos by building barriers.

Chaotic diffusion often represents a severe obstacle for the setup of experiments, e.g., in fusion plasmas or particle accelerators. We present a complete test of a method of control of Hamiltonian chaos, with both its numerical test and its first experimental realization on a paradigm for wave-particle interaction, i.e., a travelling wave tube. The core of our approach is a small apt modification of the system which channels chaos by building barriers to diffusion. Its experimental realization opens the possibility to practically achieve the control of a wide range of systems at a low additional cost of energy.