Search

In this paper, the problem of best subset selection in logistic regression is addressed. In particular, we take into account formulations of the problem resulting from the adoption of information criteria, such as AIC or BIC, as goodness-of-fit measures. There exist various methods to tackle this problem. Heuristic methods are computationally cheap, but are usually only able to find low quality solutions. Methods based on local optimization suffer from similar limitations as heuristic ones. On the other hand, methods based on mixed integer reformulations of the problem are much more effective, at the cost of higher computational requirements, that become unsustainable when the problem size grows. We thus propose a new approach, which combines mixed-integer programming and decomposition techniques in order to overcome the aforementioned scalability issues. We provide a theoretical characterization of the proposed algorithm properties. The results of a vast numerical experiment, performed on widely available datasets, show that the proposed method achieves the goal of outperforming state-of-the-art techniques.

The present pilot study investigates whether an abnormal miRNA profile in NIPT plasma samples can explain the finding of a low cell-free DNA (cfDNA) fetal fraction (cfDNAff) in euploid fetuses and non-obese women. Twelve women who underwent neoBona® NIPT with a normal fetal karyotype were studied. Six with a cfDNAff < 4% were matched with a control group with normal levels of cfDNAff > 4%. Samples were processed using the nanostring nCounter® platform with a panel of 800 miRNAs. Four of the maternal miRNAs, miR-579, miR-612, miR-3144 and miR-6721, had a significant abnormal expression in patients. A data filtering analysis showed that miR-579, miR-612, miR-3144 and miR-6721 targeted 169, 1, 48 and 136 placenta-specific genes, respectively. miR-579, miR-3144 and miR-6721 shared placenta-specific targeted genes involved in trophoblast invasion and migration pathways (IGF2R, PTCD2, SATB2, PLAC8). Moreover, the miRNA target genes encoded proteins localized in the placenta and involved in the pathogenesis of pre-eclampsia, including chorion-specific transcription factor GCMa, PRG2, Lin-28 Homolog B and IGFBP1. In conclusion, aberrant maternal miRNA expression in circulating plasma could be a source of dysregulating trophoblast invasion and migration and could represent a novel cause of a low cfDNAff in the sera of pregnant women at the time of NIPT analysis.

Treatment with fingolimod for multiple sclerosis (MS) reduces the efficacy of COVID-19 vaccination. We evaluated by a multivariate linear regression model whether main lymphocyte subsets and demographic feature correlated to the subsequent increase in anti-SARS-CoV2 antibodies following the third dose of COVID-19 vaccination in fingolimod-treated MS patients. We found that number and proportion of peripheral blood CD19+ B lymphocytes before the third dose of vaccination in MS patients treated with fingolimod, predict the subsequent increase of anti-SARS-CoV2 antibodies (respectively p = 0.013; p = 0.015). This work suggests that evaluating the numbers of CD19+ B cells may be important to identify patients at risk of not producing SARS-CoV-2 antibodies, with possible reduced protection from COVID-19.

Coronavirus disease 2019 (COVID-19) resulted in several psychological consequences. Past epidemiological experiences already showed the deep albeit heterogeneous psychological repercussions of pandemics. Nevertheless, little is known about COVID-19 outbreak and the possible strategies for boosting resilience in patients with chronic diseases such as Multiple Sclerosis (MS). Therefore, we designed a study aiming to assess the changes in mental distress during COVID-19 outbreak in patients with MS and to identifyfactors contributing to resilience's development.We enrolled 106 patients (69 relapsing-remitting, 20 secondary-progressive, and 17 primary-progressive) whose neuropsychological assessment before the COVID-19 pandemic (1 January 2019-1 March 2020) was available. It consisted of Brief International Cognitive Assessment for MS (BICAMS), Hospital Anxiety and Depression Scale (HADS) and patient-reported MS Neuropsychological Screening Questionnaire (MSNQ-P). All patients were re-tested during Italian lockdown through an online survey, comprehensive of sociodemographic information, HADS self-rating Scale, MSNQ-P Questionnaire and finally Connor-Davidson Resilience self-rating Scale (CD-RISC 25), in order to evaluate resilience.No significant changes in HADS and MSNQ-P scores were detected during COVID-19 pandemic in our population. Though, pre-existing lower HADS and MSNQ-P scores but not demographic, disease- and treatment-related elements were found significantly (p

Invasive ductal carcinoma (IDC) constitutes the most frequent malignant cancer endangering women’s health. In this study, a new spontaneously immortalized breast cancer cell line, DHSF-BR16 cells, was isolated from the primary IDC of a 74-years old female patient, treated with neoadjuvant chemotherapy and disease-free 5-years after adjuvant chemotherapy. Primary breast cancer tissue surgically removed was classified as ER−/PR−/HER2+, and the same phenotype was maintained by DHSF-BR16 cells. We examined DHSF-BR16 cell morphology and relevant biological and molecular markers, as well as their response to anticancer drugs commonly used for breast cancer treatment. MCF-7 cells were used for comparison purposes. The DHSF-BR16 cells showed the ability to form spheroids and migrate. Furthermore, DHSF-BR16 cells showed a mixed stemness phenotype (i.e. CD44+/CD24−/low), high levels of cytokeratin 7, moderate levels of cytokeratin 8 and 18, EpCAM and E-Cadh. Transcriptome analysis showed 2071 differentially expressed genes between DHSF-BR16 and MCF-7 cells (logFC > 2, p-adj

In this manuscript, we consider the problem of minimizing a smooth function with cardinality constraint, i.e., the constraint requiring that the "Equation missing"-norm of the vector of variables cannot exceed a given threshold value. A well-known approach of the literature is represented by the class of penalty decomposition methods, where a sequence of penalty subproblems, depending on the original variables and new variables, are inexactly solved by a two-block decomposition method. The inner iterates of the decomposition method require to perform exact minimizations with respect to the two blocks of variables. The computation of the global minimum with respect to the original variables may be prohibitive in the case of nonconvex objective function. In order to overcome this nontrivial issue, we propose a modified penalty decomposition method, where the exact minimizations with respect to the original variables are replaced by suitable line searches along gradient-related directions. We also present a derivative-free penalty decomposition algorithm for black-box optimization. We state convergence results of the proposed methods, and we report the results of preliminary computational experiments.

Soma and Neurite Density Image (SANDI) is an advanced diffusion magnetic resonance imaging (dMRI) signal model devised to probe in vivo microstructural information from both grey and white matter. However, this model requires multi-shell acquisitions that include b-values that are at least 6 times higher than those used in clinical practice. Here we present a 10-minute acquisition protocol that enables to acquire the necessary images for SANDI modelling on a clinical 3T scanner. We demonstrate the feasibility and assess the repeatability and reproducibility of our approach by computing microstructural metrics of SANDI and other state-of-the-art models on five healthy subjects and we present its potential clinical impact on five subjects affected by multiple sclerosis with relapsing-remitting course. Our results suggest that SANDI is a repeatable, reproducible, feasible, and practical method to characterize both white and grey matter tissues in both the healthy brain and in neurological diseases.

to investigate, for the first time, the incidence of cancer (years 2009-2015) and geographical distribution among children and adolescents with cancer diagnosis in Lazio Region (Central Italy).to compute incidence rates of childhood cancers from Lazio Region Childhood Cancer Registry (LRCCR) database, established in 2015, and to compare results with national figures for 2012 provided by the Italian cancer registries network (AIRTUM).all new cases of malignant tumours (behaviour: /3 of ICD-O-3 classification) and all central nervous system tumours were selected, regardless of behaviour (/0, /1, /3) in children and adolescents (0-19 years) registered in the LRCCR data base.it was computed: • the raw and the direct standardised rates for the 0-14-year and the 15-19-year age groups for total malignant tumours of the ICCC-3 classification by area (province level and municipality of Rome); • Relative Risks (RR) for area-specific rate compared with that of the Lazio Region and 95% Confidence Intervals (95%CI).a total of 1,782 incident cases were recorded in 2009-2015; of these, 91.4% were confirmed by a pathology report. Standardized Incidence Rate for all malignant tumours is 207.2×1,000,000 (95%CI 195.5-219.5) in children and 335.1×1,000,000 (95%CI 308.9-361.2) in adolescents. Compared to the Lazio Region, a higher incidence of tumours is observed in Rome municipality (RR 1.09; 95%CI 0.98-1.20) and in the Frosinone province (RR 1.07; 95%CI 0.91-1.25) for the whole 0-19-year age group.compared to the pooled AIRTUM figures for 2003-2008, Lazio Region showed a higher incidence for all cancers, both in children and adolescents, and for specific tumours, such as leukaemia in children and thyroid carcinoma in adolescents. Apart from the diverse observation period, these differences may be due to a higher registry sensitivity of the childhood specialized registry compared to general population registries. The observed incidence excesses for specific geographical areas and tumours deserve further investigations. Overall, in its first seven years of activity, the Lazio childhood cancer registry was able to provide reliable epidemiological figures of cancer incidence in children and adolescents in the Italian context.

Local and global anthropogenic pressures due to climate change and to local uses and activities are exerting significant cumulative impacts to greater extents of the oceans and seas. Coastal ecosystems are particularly threatened by the intensity and coexistence of several marine uses and pressures, including sewage and urban constructions, tourism, ship traffic, fisheries and aquaculture. Assessment of pressures and the identification of mitigation measures are key urgent actions, as already highlighted by the EU Marine Strategy Framework Directive and the United Nations Sustainable Development Goal 14. The aim of this work, developed within the Interreg-Med project SHAREMED, is to systematize existing knowledge on threats and pollution, including those of transboundary origin, for long term strategies and common action marine spatial planning, jointly developed with stakeholders. The quest is to assess coexisting environmental threats, and their propagation in space and time, at proper spatial and temporal scales, according to the type and action of each stressor (i.e. global vs. local). Cumulative pressures are tackled within a dedicated Atlas comprising three sub-basinsins of the Mediterranean Sea: the North Adriatic Sea, the Sicilian Channel and the North-Western region. The Atlas integrates information generated at the best available resolutions by 1) in-situ sampling, 2) remote observations, 3) numerical models, and 4) focusing on target ecosystems and habitat forming species. These sub-basins are subjected to multiple local and larger scale (e.g. climate) pressures that propagate in space and time, and across political boundaries, that need to be addressed through coordinated actions, based on evidence-rooted common understanding. Interactions with relevant Stakeholders, solicited through an online survey, and meetings, were used to select target ecosystems and to identify the key relevant pressures. The Atlas is based on open-access databases and portals, literature reviews and from ad-hoc model simulations concerning marine heatwaves, ship traffic, oil pollution, marine litter and fishing efforts. We will present the main preliminary results and needs and gaps in observations related to marine ecosystems threats.

Marine plastic litter is one of the most significant signal of the impact of human activities on the marine environment. Therefore, improved methods and models are needed to better understand the distribution pattern of plastics (mainly microplastics) on the sea surface, along the water column, and on the seabed. So far, most plastics sampling campaigns have collected sea surface data, but these data were very scattered and mostly unrepresentative of the seasonal variability of their distribution. A comprehensive overview of the presence of plastic litter in marine enviroment must rely on models having the skill to better represent spatial patterns, interactions with marine ecosystems, and even predict the possible presence of plastic clusters at a specific time and position. Numerous studies adopted models of plastic transport which consider some sources of pollution (rivers, ports, ship routes) to determine plastic distribution in the open sea due to meteo-marine forcing. However, most of these models have not been validated against field data.In this presentation we show the results of a validation procedure of the modelled marine debris distributions expected in the North-Western Mediterranean between May and September 2019, through the comparison with field observations on the sea surface from campaigns carried out within the Interreg Med project Plastic Busters MPA. Marine debris observations show a significant variability, especially along the coasts, highlighting the need to employ a hydrodynamic model with a resolution much higher than that of basin-scale models. In the comparison between the observed and modelled surface plastic concentrations, the effect of model resolution will be specifically addressed.

Introduction In the last years, many new drugs have been developed targeting different oncology pathways, overall improving both quality of life and survival in several malignancies. However, the increase of those therapies is associated with novel toxicities, mainly immune-related adverse events (irAEs), never observed before. Different irAEs are now well characterized, and, among them, neuromuscular complications, following immune checkpoint inhibitor (ICPi) therapy, are increasingly studied and described. However, there are also neurological complications related to the use of other targeted therapies, less known and probably underestimated. Herein we describe two oncological patients who developed neuromuscular diseases after administration of targeted therapies, different from ICPi. Case reports The first patient was treated with the combination of Vemurafenib and Cobimetinib, BRAF and MEK inhibitors, respectively, for a cutaneous melanoma. One year after the beginning of the combined treatment, she developed a sub-acute motor neuropathy with predominant cranial nerve involvement. She was successfully treated with methylprednisolone. The second patient received therapy with Imatinib, tyrosine kinase inhibitor and precursor of the targeted therapy, for a gastrointestinal stromal tumour. Few days after the first administration, he developed generalized myasthenia gravis with respiratory failure. Clinical remission was obtained with plasma-exchange, intravenous immunoglobulins and steroids. Discussion and Conclusion We strengthen the relevance of neuromuscular complications which may occur long after treatment start or in patients receiving not only the latest ICPi but also “older” and apparently better-known targeted therapies. Also in the latter cases, an immune-mediated “off-target” pathogenic mechanism can be hypothesized, and consequences can be life threatening, if not promptly diagnosed and appropriately managed.

BACKGROUND AND PURPOSE The best therapeutic approach for aggressive relapsing-remitting multiple sclerosis remains unknown. The objective was to compare the efficacy and safety of autologous haematopoietic stem cell transplantation (aHSCT) and alemtuzumab in aggressive relapsing-remitting multiple sclerosis. METHODS The time to first relapse, time to confirmed disability worsening, time to first evidence of magnetic resonance imaging (MRI) activity and time to first evidence of disease activity were compared between the two treatment groups. Secondary outcomes included the 12, 24 and 36 month annualized relapse rate (ARR) and the 6-month confirmed Expanded Disability Status Scale (EDSS) changes at months 12 and 24. RESULTS Fifty-seven patients treated with aHSCT (n = 25) or alemtuzumab (n = 32) were included. At baseline, aHSCT patients had a higher EDSS (median score 6 vs. 3; P

The benefit of adjuvant chemotherapy in the early stages of colorectal cancer (CRC) is still disappointing and the prediction of treatment outcome quite difficult. Recently, through a transcriptomic approach, we evidenced a role of PNN and KCNQ1OT1 gene expression in predicting response to fluoropyrimidine-based adjuvant chemotherapy in stage III CRC patients. Thus, the aim of this study was to validate in an independent cohort of stages IIIII CRC patients our previous findings. PNN and KCNQ1OT1 mRNA expression levels were evaluated in 74 formalin-fixed paraffin-embedded tumor and matched normal mucosa samples obtained by stages IIIII CRC patients treated with fluoropyrimidine-based adjuvant chemotherapy. PININ, the protein encoded by PNN, was immunohistochemically evaluated in 15 tumor and corresponding normal mucosa samples, selected on the basis of a low, medium, or high mRNA expression tumor/mucosa ratio. PNN and KCNQ1OT1 mRNA mean expression levels were significantly higher in tumor compared with normal tissues. Patients with high PNN or KCNQ1OT1 tumor mRNA levels according to ROC-based cutoffs showed a shorter disease-free survival (DFS) compared with patients with low tumor mRNA gene expression. Also, patients with tumor mRNA expression values of both genes below the identified cutoffs had a significantly longer DFS compared with patients with the expression of one or both genes above the cutoffs. In a representative large cohort of stages IIIII CRC untreated patients retrieved from GEO datasets, no difference in DFS was observed between patients with high and low PNN or KCNQ1OT1 gene expression levels. These data confirm our previous findings and underscore the relevance of PNN and KCNQ1OT1 expression in predicting DFS in early stages of CRC treated with fluoropyrimidine-based adjuvant chemotherapy. If further validated in a prospective case series, both biomarkers could be used to identify patients who benefit from this treatment and to offer alternative chemotherapy regimens to potential unresponsive patients. In relation to the suggested biological role of PNN and KCNQ1OT1 in CRC, they might also be exploited as potential therapeutic targets.

To retrospectively evaluate the different performances of T1-SE and T1-GE sequences in detecting hypointense lesions in multiple sclerosis (MS), to quantify the degree of microstructural damage within lesions and to correlate them with patient clinical status. Sixty clinically isolated syndrome (CIS) and MS patients underwent brain magnetic resonance imaging (MRI) on 1.5-T and 3-T scanners. We identified T2 fluid-attenuated inversion recovery hyperintense lesions with no hypointense signal on T1-SE/T1-GE (a), hypointense lesions only on T1-GE (b), and hypointense lesions on both T1-SE and T1-GE sequences (c). We compared mean lesion number (LN) and volume (LV) identified on T1-SE and T1-GE sequences, correlating them with Expanded Disability Status Scale (EDSS); fractional anisotropy (FA) and mean diffusivity (MD) values inside each lesion type were extracted and normal-appearing white matter (NAWM). Thirty-five patients were female. Mean age was 39.2 (± 7.8); median EDSS was 3 (± 2). There were 23 CIS, 21 relapsing–remitting (RR), and 16 progressive MS. T1-GE and T1-SE LN and LV were significantly different (p

ATP Binding Cassette (ABC) transporters, widely studied in cancer for their role in drug resistance, have been more recently also considered for their contribution to cancer cell biology. To date, many data provide evidences for their potential role in all the phases of cancer development from cancer susceptibility, tumor initiation, tumor progression and metastasis. Although many evidences are based on correlative analyses, data describing a direct or indirect role of ABC transporters in cancer biology are increasing. Overall, current available information suggests a relevant molecular effector role of some ABC transporters in cancer invasion and metastasis as reported in experimental tumor models. From a therapeutic point of view, due to the physiological relevant roles that ABC transporters play in the organism, the capability to selectively inhibit the function or the expression of ABC transporters in cancer stem cells or other tumor cells, represents the main challenge for researcher scientists. A detailed and updated description of the current knowledge on the role of ABC transporters in cancer biology is provided.

This paper provides a theoretical and numerical investigation of a penalty decomposition scheme for the solution of optimization problems with geometric constraints. In particular, we consider some situations where parts of the constraints are nonconvex and complicated, like cardinality constraints, disjunctive programs, or matrix problems involving rank constraints. By a variable duplication and decomposition strategy, the method presented here explicitly handles these difficult constraints, thus generating iterates which are feasible with respect to them, while the remaining (standard and supposingly simple) constraints are tackled by sequential penalization. Inexact optimization steps are proven sufficient for the esulting algorithm to work, so that it is employable even with difficult objective functions. The current work is therefore a significant generalization of existing papers on penalty decomposition methods. On the other hand, it is related to some recent publications which use an augmented Lagrangian idea to solve optimization problems with geometric constraints. Compared to these methods, the decomposition idea is shown to be numerically superior since it allows much more freedom in the choice of the subproblem solver, and since the number of certain (possibly expensive) projection steps is significantly less. Extensive numerical results on several highly complicated classes of optimization problems in vector and matrix paces indicate that the current method is indeed very efficient to solve these problems., Comment: Comput Optim Appl (2023)