Diagnosing idiopathic learning disability: a cost-effectiveness analysis of microarray technology in the National Health Service of the United Kingdom

Array based comparative genomic hybridisation (aCGH) is a powerful technique for detecting clinically relevant genome imbalance and can offer 40 to > 1000 times the resolution of karyotyping. Indeed, idiopathic learning disability (ILD) studies suggest that a genome-wide aCGH approach makes 10–15% more diagnoses involving genome imbalance than karyotyping. Despite this, aCGH has yet to be implemented as a routine NHS service. One significant obstacle is the perception that the technology is prohibitively expensive for most standard NHS clinical cytogenetics laboratories. To address this, we investigated the cost-effectiveness of aCGH versus standard cytogenetic analysis for diagnosing idiopathic learning disability (ILD) in the NHS. Cost data from four participating genetics centres were collected and analysed. In a single test comparison, the average cost of aCGH was ￡442 and the average cost of karyotyping was ￡117 with array costs contributing most to the cost difference. This difference was not a key barrier when the context of follow up diagnostic tests was considered. Indeed, in a hypothetical cohort of 100 ILD children, aCGH was found to cost less per diagnosis (￡3,118) than a karyotyping and multi-telomere FISH approach (￡4,957). We conclude that testing for genomic imbalances in ILD using microarray technology is likely to be cost-effective because long-term savings can be made regardless of a positive (diagnosis) or negative result. Earlier diagnoses save costs of additional diagnostic tests. Negative results are cost-effective in minimising follow-up test choice. The use of aCGH in routine clinical practice warrants serious consideration by healthcare providers. Copyright statement The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, an exclusive licence (or non exclusive for government employees) on a worldwide basis to the BMJ Publishing Group Ltd, and its Licensees to permit this article (if accepted) to be published in BMJ editions and any other BMJPGL products and to exploit all subsidiary rights, as set out in our licence (bmj.com/advice/copyright.shtml). Authorship The authors included on this paper fulfil the criteria of authorship and no one who fulfils the criteria has been excluded from authorship. The authors made a substantial contribution to the conception, design, analysis and interpretation of data. They were involved in drafting the article or revising it critically for important intellectual content and approving the version to be published. Contributorship Sarah Wordsworth (Guarantor): Planning, conducting and reporting work, interpretation of data, drafting and revising article. James Buchanan: Conducting and reporting work, interpretation of data, revising article. Regina Regan: Completing costing questionnaire, providing protocol details, other costing information, interpretation of data, information about learning disability and genome imbalance and revising article. Val Davison: Completing costing questionnaire, providing protocol details, sharing overall laboratory experience and drafting article. Kim Smith: Completing costing questionnaire, providing protocol details, drafting article. Sara Dyer: Completing costing questionnaire and providing protocol details. Carolyn Campbell: Completing costing questionnaire and providing protocol details. Edward Blair: Critical appraisal of article for clinical content and revising article. Eddy Maher: Completing costing questionnaire, providing protocol details, sharing overall laboratory experience and drafting article. Jenny Taylor: Planning and facilitating work between centres. Drafting and revising article. Samantha JL Knight: Completing costing questionnaire, providing protocol details, other costing information, interpretation of data, providing information about learning disability and genome imbalance, drafting and revising article. Jenny Taylor and Samantha JL Knight contributed equally to the work presented.     

Adipokine and insulin profiles distinguish diabetogenic and non-diabetogenic obesities in mice

Objective: To use longitudinal profiling of plasma adipokines to distinguish diabetogenic vs. non‐diabetogenic obesity syndrome in two new mouse models of polygenic obesity. Research Methods and Procedures: Male mice of the NONcNZO5 strain develop a polygenic obesity syndrome uncomplicated by diabetes, whereas NONcNZO10 males develop a comparable polygenic obesity that precipitates type 2 diabetes. A multiplex immunoassay for simultaneous measurement of insulin and a panel of mouse adipokines (leptin, resistin, adiponectin, interleukin‐6, tumor necrosis factor α, macrophage chemoattractant protein‐1, plasminogen activator inhibitor‐1) were used to profile longitudinal changes in these strains between 4 and 16 weeks of age that might distinguish the non‐diabetogenic vs. diabetogenic obesity (diabesity). Results: Both strains became adipose, with NONcNZO5 males attaining a higher mean body weight with a higher percentage fat content. Weight gain in NONcNZO5 was accompanied by a transient peak in plasma insulin (PI) at 8 weeks followed by a decline into normal range, with normoglycemia maintained throughout. In contrast, NONcNZO10 showed no early PI secretory response because both body weight and plasma glucose increased between 4 and 8 weeks. Only after 12 weeks, with hyperglycemia established, was a delayed PI secretory response observed. Neither plasma leptin nor adiponectin concentrations significantly differentiated the two syndromes over time. However, repeated measures ANOVA showed that NONcNZO10 males maintained significantly higher plasma concentrations of two adipokines, resistin and plasminogen activator inhibitor‐1, and the pro‐inflammatory cytokine/adipokine macrophage chemoattractant protein‐1. Discussion: Longitudinal profiling of PI and adipokines in two new mouse models developing moderate obesity demonstrated that specific marker signatures differentiated a non‐diabetogenic obesity from a diabetogenic obesity.     

Modifying ligand-induced and constitutive signaling of the human 5-HT4 receptor

G protein-coupled receptors (GPCRs) signal through a limited number of G-protein pathways and play crucial roles in many biological processes. Studies of their in vivo functions have been hampered by the molecular and functional diversity of GPCRs and the paucity of ligands with specific signaling effects. To better compare the effects of activating different G-protein signaling pathways through ligand-induced or constitutive signaling, we developed a new series of RASSLs (receptors activated solely by synthetic ligands) that activate different G-protein signaling pathways. These RASSLs are based on the human 5-HT(4b) receptor, a GPCR with high constitutive G(s) signaling and strong ligand-induced G-protein activation of the G(s) and G(s/q) pathways. The first receptor in this series, 5-HT(4)-D(100)A or Rs1 (RASSL serotonin 1), is not activated by its endogenous agonist, serotonin, but is selectively activated by the small synthetic molecules GR113808, GR125487, and RO110-0235. All agonists potently induced G(s) signaling, but only a few (e.g., zacopride) also induced signaling via the G(q) pathway. Zacopride-induced G(q) signaling was enhanced by replacing the C-terminus of Rs1 with the C-terminus of the human 5-HT(2C) receptor. Additional point mutations (D(66)A and D(66)N) blocked constitutive G(s) signaling and lowered ligand-induced G(q) signaling. Replacing the third intracellular loop of Rs1 with that of human 5-HT(1A) conferred ligand-mediated G(i) signaling. This G(i)-coupled RASSL, Rs1.3, exhibited no measurable signaling to the G(s) or G(q) pathway. These findings show that the signaling repertoire of Rs1 can be expanded and controlled by receptor engineering and drug selection.     

Evidence of compromised blood-spinal cord barrier in early and late symptomatic SOD1 mice modeling ALS

Background

The blood-brain barrier (BBB), blood-spinal cord barrier (BSCB), and blood-cerebrospinal fluid barrier (BCSFB) control cerebral/spinal cord homeostasis by selective transport of molecules and cells from the systemic compartment. In the spinal cord and brain of both ALS patients and animal models, infiltration of T-cell lymphocytes, monocyte-derived macrophages and dendritic cells, and IgG deposits have been observed that may have a critical role in motor neuron damage. Additionally, increased levels of albumin and IgG have been found in the cerebrospinal fluid in ALS patients. These findings suggest altered barrier permeability in ALS. Recently, we showed disruption of the BBB and BSCB in areas of motor neuron degeneration in the brain and spinal cord in G93A SOD1 mice modeling ALS at both early and late stages of disease using electron microscopy. Examination of capillary ultrastructure revealed endothelial cell degeneration, which, along with astrocyte alteration, compromised the BBB and BSCB. However, the effect of these alterations upon barrier function in ALS is still unclear. The aim of this study was to determine the functional competence of the BSCB in G93A mice at different stages of disease.

Methodology/Principal Findings

Evans Blue (EB) dye was intravenously injected into ALS mice at early or late stage disease. Vascular leakage and the condition of basement membranes, endothelial cells, and astrocytes were investigated in cervical and lumbar spinal cords using immunohistochemistry. Results showed EB leakage in spinal cord microvessels from all G93A mice, indicating dysfunction in endothelia and basement membranes and confirming our previous ultrastructural findings on BSCB disruption. Additionally, downregulation of Glut-1 and CD146 expressions in the endothelial cells of the BSCB were found which may relate to vascular leakage.

Conclusions/Significance

Results suggest that the BSCB is compromised in areas of motor neuron degeneration in ALS mice at both early and late stages of the disease.     

Decorrelation of the True and Estimated Classifier Errors in High-Dimensional Settings

The aim of many microarray experiments is to build discriminatory diagnosis and prognosis models. Given the huge number of features and the small number of examples, model validity which refers to the precision of error estimation is a critical issue. Previous studies have addressed this issue via the deviation distribution (estimated error minus true error), in particular, the deterioration of cross-validation precision in high-dimensional settings where feature selection is used to mitigate the peaking phenomenon (overfitting). Because classifier design is based upon random samples, both the true and estimated errors are sample-dependent random variables, and one would expect a loss of precision if the estimated and true errors are not well correlated, so that natural questions arise as to the degree of correlation and the manner in which lack of correlation impacts error estimation. We demonstrate the effect of correlation on error precision via a decomposition of the variance of the deviation distribution, observe that the correlation is often severely decreased in high-dimensional settings, and show that the effect of high dimensionality on error estimation tends to result more from its decorrelating effects than from its impact on the variance of the estimated error. We consider the correlation between the true and estimated errors under different experimental conditions using both synthetic and real data, several feature-selection methods, different classification rules, and three error estimators commonly used (leave-one-out cross-validation, -fold cross-validation, and .632 bootstrap). Moreover, three scenarios are considered: (1) feature selection, (2) known-feature set, and (3) all features. Only the first is of practical interest; however, the other two are needed for comparison purposes. We will observe that the true and estimated errors tend to be much more correlated in the case of a known feature set than with either feature selection or using all features, with the better correlation between the latter two showing no general trend, but differing for different models.[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22]