Methods and Issues to Consider for Detection of Safety Signals From Spontaneous Reporting Databases: A Report of the DIA Bayesian Safety Signal Detection Working Group.

Spontaneous reporting (SR) adverse event system databases, large observational databases, large clinical trials, and large health records databases comprise repositories of information that may be useful for early detection of potential harms associated with drugs, devices, and vaccines. All of the data sources include many different adverse events and many medical products, so that any approach designed to detect "important" signals of potential harm must have adequate specificity to protect against false alarms yet provide satisfactory sensitivity for detecting issues that really need further investigation. Algorithms for evaluating potential risks using information from these sources, especially SR databases, have been described in the literature. The algorithms may seek to identify potential product-event associations without any prior specifications, to identify events associated with a particular product or set of products, or to identify products associated with a particular event or set of events. This article provides recommendations for using information from postmarketing spontaneous adverse event reporting databases to provide insight into risks of potential harm expressed by safety signals and offers guidance regarding appropriate methods, both frequentist and Bayesian, to use in various situations as a function of the objective of the analysis.

Semiparametric Bayesian commensurate survival model for post-market medical device surveillance with non-exchangeable historical data.

Trial investigators often have a primary interest in the estimation of the survival curve in a population for which there exists acceptable historical information from which to borrow strength. However, borrowing strength from a historical trial that is non-exchangeable with the current trial can result in biased conclusions. In this article we propose a fully Bayesian semiparametric method for the purpose of attenuating bias and increasing efficiency when jointly modeling time-to-event data from two possibly non-exchangeable sources of information. We illustrate the mechanics of our methods by applying them to a pair of post-market surveillance datasets regarding adverse events in persons on dialysis that had either a bare metal or drug-eluting stent implanted during a cardiac revascularization surgery. We finish with a discussion of the advantages and limitations of this approach to evidence synthesis, as well as directions for future work in this area. The article's Supplementary Materials offer simulations to show our procedure's bias, mean squared error, and coverage probability properties in a variety of settings.

Recent Statistical Contributions to Medical Device Development.

Clinical trials in the development of new medical device products are in many ways analogous to clinical trials in the development of new drug or biologic products. However, the differences are important and not always intuitive to a statistician with only experience supporting development of drug and biologic products. In this paper we discuss some of the interesting differences with focus on the statistical innovation that is coming out of the medical device area. We discuss examples of the differences in clinical trial design and effects of these differences on clinical development programs.

Bayesian adaptive design for device surveillance.

BACKGROUND: Postmarket device surveillance studies often have important primary objectives tied to estimating a survival function at some future time $$T$$ with a certain amount of precision. PURPOSE: This article presents the details and various operating characteristics of a Bayesian adaptive design for device surveillance, as well as a method for estimating a sample size vector (determined by the maximum sample size and a preset number of interim looks) that will deliver the desired power. METHODS: We adopt a Bayesian adaptive framework, which recognizes the fact that persons enrolled in a study report their results over time, not all at once. At each interim look, we assess whether we expect to achieve our goals with only the current group or the achievement of such goals is extremely unlikely even for the maximum sample size. RESULTS: Our Bayesian adaptive design can outperform two nonadaptive frequentist methods currently recommended by Food and Drug Administration (FDA) guidance documents in many settings. LIMITATIONS: Our method's performance can be sensitive to model misspecification and changes in the trial's enrollment rate. CONCLUSIONS: The proposed design provides a more efficient framework for conducting postmarket surveillance of medical devices.

Cell death-inducing DFF45-like effector C is reduced by caloric restriction and regulates adipocyte lipid metabolism.

Members of the cell death-inducing DFF45-like effector (CIDE) gene family have been shown to regulate lipid metabolism. In this article, we report that the third member of the human CIDE family, CIDEC, is down-regulated in response to a reduced caloric intake. The down-regulation was demonstrated by microarray and real-time polymerase chain reaction analysis of subcutaneous adipose tissue in 2 independent studies on obese patients undergoing treatment with a very low calorie diet. By analysis of CIDEC expression in 65 human tissues, we conclude that human CIDEC is predominantly expressed in subcutaneous adipocytes. Together, these observations led us to investigate the effect of decreased CIDEC expression in cultured 3T3-L1 adipocytes. Small interfering RNA-mediated knockdown of CIDEC resulted in an increased basal release of nonesterified fatty acids, decreased responsiveness to adrenergic stimulation of lipolysis, and increased oxidation of endogenous fatty acids. Thus, we suggest that CIDEC is a regulator of adipocyte lipid metabolism and may be important for the adipocyte to adapt to changes in energy availability.

Relations of adipose tissue CIDEA gene expression to basal metabolic rate, energy restriction, and obesity: population-based and dietary intervention studies.

CONTEXT: Cell death-inducing DNA fragmentation factor-alpha-like effector A (CIDEA) could be a potential target for the treatment of obesity via the modulation of metabolic rate, based on the findings that CIDEA inhibits the brown adipose tissue uncoupling process in rodents. OBJECTIVES: Our objects were to investigate the putative link between CIDEA and basal metabolic rate in humans and to elucidate further the role of CIDEA in human obesity. DESIGN: We have explored CIDEA gene expression in adipose tissue in two different human studies: a cross-sectional and population-based study assessing body composition and metabolic rate (Mölndal Metabolic study, n = 92); and a longitudinal intervention study of obese subjects treated with a very low calorie diet (VLCD) (VLCD study, n = 24). RESULTS: The CIDEA gene was predominantly expressed in adipocytes as compared with other human tissues. CIDEA gene expression in adipose tissue was inversely associated with basal metabolic rate independently of body composition, age, and gender (P = 0.014). The VLCD induced an increase in adipose tissue CIDEA expression (P < 0.0001) with a subsequent decrease in response to refeeding (P < 0.0001). Reduced CIDEA gene expression was associated with a high body fat content (P < 0.0001) and high insulin levels (P < 0.01). No dysregulation of CIDEA expression was observed in individuals with the metabolic syndrome when compared with body mass index-matched controls. In a separate sample of VLCD-treated subjects (n = 10), uncoupling protein 1 expression was reduced during diet (P = 0.0026) and inversely associated with CIDEA expression (P = 0.0014). CONCLUSION: The findings are consistent with the concept that CIDEA plays a role in adipose tissue energy expenditure.

Dissociation between adipose tissue expression and serum levels of adiponectin during and after diet-induced weight loss in obese subjects with and without the metabolic syndrome.

The study aimed to examine if dysmetabolic subjects (MetS+) have lower adiponectin gene expression and lower circulating adiponectin levels than non-dysmetabolic obese subjects (MetS-) at baseline, if adiponectin expression and adiponectin concentration rise more in the dysmetabolic group during weight loss, and if v-SNARE Vti1a (vesicle transport soluble NSF attachment protein receptor vps10p tail interacting 1a) expression increases during the weight loss, as a mechanism for increased adiponectin secretion. Twenty-one obese MetS+ and 19 obese MetS- subjects underwent a very low-energy diet for 16 weeks followed by 2 weeks of refeeding. Abdominal subcutaneous adipose tissue biopsies and blood samples were taken before, during, and after dieting for DNA microarray, reverse transcriptase-polymerase chain reaction, and biochemical analyses. Serum adiponectin was also assessed in a sex- and age-matched healthy, nonobese reference group. Weight decreased by 26.3+/-9.8 kg in the MetS+ group and 28.2+/-8.4 kg in the MetS- group with concomitant reductions in insulin, hemoglobin A1c, and triglycerides that were more pronounced in the MetS+ group. Initially, the MetS+ subjects had lower serum adiponectin, but the differences disappeared at week 8, with a continuous increase in serum adiponectin throughout the study in both groups to a level that was higher than in the reference group. The expression of adiponectin and v-SNARE Vti1a did not differ between the groups or over time. In conclusion, obese subjects with the metabolic syndrome had lower circulating adiponectin than subjects without the syndrome. Weight loss increased serum levels of adiponectin without a parallel increase in adiponectin gene expression. The mechanisms involved in the regulation of adiponectin levels merits further investigation.

The expression of NAD(P)H:quinone oxidoreductase 1 is high in human adipose tissue, reduced by weight loss, and correlates with adiposity, insulin sensitivity, and markers of liver dysfunction.

CONTEXT: We have previously identified nicotinamide adenine dinucleotide phosphate:quinone oxidoreductase 1 (NQO1), an enzyme involved in the protection against oxidative stress, as a gene predominantly expressed in human adipocytes. Studies in mice deficient in NQO1 activity suggest that NQO1 may also play an important role in metabolism. OBJECTIVE: The aim of this study was to explore the expression and regulation of NQO1 in human adipose tissue (AT) and isolated adipocytes. PATIENTS AND RESULTS: The high expression of NQO1 in adipocytes was verified in human adipocytes and AT by real-time PCR. DNA microarray analysis showed that NQO1 was expressed at higher levels in large compared with small adipocytes, isolated from the same fat biopsy. Furthermore, NQO1 mRNA levels were positively correlated with adipocyte size (n = 7; P < 0.002). During an 18-wk diet regime (n = 24; mean weight loss 27 kg), the NQO1 expression in human sc AT was down-regulated (P < 0.0001), and mRNA levels correlated with body mass index (P = 0.0005), sc, and total abdominal AT areas, as determined by computerized tomography (P < 0.0001, both) and metabolic parameters. NQO1 mRNA levels were also positively correlated with aspartate aminotransferase (P = 0.0028) and alanine aminotransferase (P = 0.0219), markers known to be associated with severity of hepatic steatosis. CONCLUSIONS: NQO1 is highly expressed in human AT, particularly in large adipocytes. AT NQO1 expression is reduced during diet-induced weight loss, and the expression levels positively correlate with adiposity, glucose tolerance, and markers of liver dysfunction. Together, these findings indicate a role for NQO1 in the metabolic complications of human obesity.

Separation of human adipocytes by size: hypertrophic fat cells display distinct gene expression.

Enlarged adipocytes are associated with insulin resistance and are an independent predictor of type 2 diabetes. To understand the molecular link between these diseases and adipocyte hypertrophy, we developed a technique to separate human adipocytes from an adipose tissue sample into populations of small cells (mean 57.6+/-3.54 microm) and large cells (mean 100.1+/-3.94 microm). Microarray analysis of the cell populations separated from adipose tissue from three subjects identified 14 genes, of which five immune-related, with more than fourfold higher expression in large cells than small cells. Two of these genes were serum amyloid A (SAA) and transmembrane 4 L six family member 1 (TM4SF1). Real-time RT-PCR analysis of SAA and TM4SF1 expression in adipocytes from seven subjects revealed 19-fold and 22-fold higher expression in the large cells, respectively, and a correlation between adipocyte size and both SAA and TM4SF1 expression. The results were verified using immunohistochemistry. In comparison with 17 other human tissues and cell types by microarray, large adipocytes displayed by far the highest SAA and TM4SF1 expression. Thus, we have identified genes with markedly higher expression in large, compared with small, human adipocytes. These genes may link hypertrophic obesity to insulin resistance/type 2 diabetes.

The expression of inhibin beta B is high in human adipocytes, reduced by weight loss, and correlates to factors implicated in metabolic disease.

Adipose tissue is an endocrine organ that produces and secretes adipokines. The aim of this study was to identify genes predominantly expressed in human subcutaneous adipocytes. For this purpose, an algorithm was developed and DNA microarray expression profiles from 33 human tissues and cell types were used to select genes. Inhibin beta B (INHBB; coding for the activin betaB subunit) was identified and high expression in adipocytes was confirmed by real-time PCR and immunohistochemistry. INHBB expression in adipose tissue was down regulated by diet-induced weight loss (p<0.001). Furthermore, INHBB expression was positively correlated to total (p<0.001) and subcutaneous (p<0.01) adipose tissue areas and serum levels of fasting insulin (p<0.01) and cholesterol (p<0.05). In conclusion, INHBB expression was high in human adipocytes, reduced by weight loss and adipose tissue INHBB mRNA levels correlated to metabolic risk factors. This suggests that activin B produced in adipocytes may play a role in the metabolic syndrome.

Three definitions of the metabolic syndrome: relations to mortality and atherosclerotic morbidity.

BACKGROUND: Different definitions of the metabolic syndrome are used, and at least one of these does not include indices of glucose intolerance and/or insulin resistance as obligatory components. In this paper, we examine the predictive power of indices having and not having these obligatory components. METHODS: A total of 1135 men and women, aged 37-61 years, were randomly selected from the populations of Mölndal and Orebro, Sweden. Mortality rate and incidence of cardiovascular morbidity were analyzed in subjects with and without the metabolic syndrome according to the definitions of WHO (World Health Organization), EGIR (European Group for the study of Insulin Resistance), and ATPIII (Adult Treatment Panel-III Guidelines). Atherosclerotic morbidity was traced until December 2002 and mortality until December 2003. Due to lack of data, our WHO definition does not include information on micro-albuminuria. RESULTS: There were 17 deaths during the 3-8 year follow-up. As compared to subjects without the metabolic syndrome, all-cause mortality was increased significantly in subjects with the syndrome defined according to WHO(non u-alb) (hazards ratio [HR] 2.98, 95% CI 1.07, 8.28, p = 0.036) but not according to EGIR (HR 1.93, 95% CI 0.67, 5.55, p = 0.230) or ATPIII (HR 0.88, 95% CI 0.20, 3.89, p = 0.870). Incident cases of ischemic heart, cerebrovascular, and/or peripheral arterial disease (n = 18) were related to the metabolic syndrome according to WHO(non u-alb) and EGIR but not according to ATPIII. CONCLUSIONS: Inclusion of glucose intolerance and/or insulin resistance as obligatory criteria in the definition of the metabolic syndrome seems to be important for the ability to predict all-cause mortality and incident cardiovascular morbidity.

A microarray search for genes predominantly expressed in human omental adipocytes: adipose tissue as a major production site of serum amyloid A.

To identify genes predominantly expressed in omental adipocytes, microarray expression profiles from 33 human tissues or cell types were analyzed, using an algorithm developed for identification of transcripts predominantly expressed in a certain tissue. Both known adipocyte-specific and more unexpected genes were among the 28 genes identified. To validate the approach, adipocyte expression of three of these genes, acute-phase serum amyloid A (A-SAA), aquaporin 7, and transport secretion protein-2.2, was compared with 17 other human tissues by real-time PCR. The unexpectedly high expression of A-SAA in adipocytes was further verified by Northern blot and immunohistochemistry. The liver, reported to be the main production site for A-SAA, displayed the second highest expression using microarray and real-time PCR. In obese subjects, adipose tissue mRNA and serum A-SAA levels were down-regulated during an 18-wk diet regime (P < 0.05 and P < 0.0001, respectively). A-SAA serum levels were highly correlated to adipose tissue mRNA levels (P < 0.001) and to the total (P < 0.0001) and sc (P < 0.0001) adipose tissue areas, as analyzed by computed tomography. We show that adipose tissue is a major expression site of A-SAA during the nonacute-phase reaction condition. This provides a direct link between adipose tissue mass and a marker for low-grade inflammation and cardiovascular risk.

Mice deficient in apolipoprotein E but not LDL receptors are resistant to accelerated atherosclerosis associated with obesity.

The aims of this study were to determine whether mice induced to become obese also exhibited accelerated atherosclerosis, and to determine whether obesity itself or dyslipidemia associated with obesity enhanced atherosclerosis. Wild-type (C57BL/6) mice and mice deficient for the low density lipoprotein receptor (LDLR-/-) or apolipoprotein E (apoE-/-) were fed a low fat, rodent chow diet or a high fat, high sucrose (diabetogenic) diet to induce obesity. As compared with wild-type mice, diabetogenic diet-fed LDLR-/- mice became more obese and developed severe dyslipidemia. Consequently, atherosclerotic lesions were increased in the LDLR-/- mice 3.7-fold over chow fed values. ApoE-/- mice showed weight gain profiles similar to those observed for wild-type mice. However, no differences in plasma lipid levels, lipoprotein profiles or atherosclerotic lesion areas were observed between chow-fed and diabetogenic diet-fed apoE-/- mice. These data demonstrate that lipid storage and partitioning as mediated by the low density lipoproteins (LDL) receptor or apoE-/- have profound and opposing consequences for dyslipidemia and atherosclerosis susceptibility associated with obesity.

Adjusted P values for genome-wide scans.

Genome-wide scans for quantitative trait loci (QTL) have traditionally been summarized with plots of logarithm of odds (LOD) scores. A valuable modification is to supplement such plots with an additional vertical axis displaying quantiles of adjusted P values and labeling local maxima of the LOD scores with location-specific adjusted P values. This provides a visible gradation of genome-wide significance for the LOD score curve, instead of the stark dichotomy that a single threshold yields. Adjusted P values give genome-wide significance of individual LOD scores and are obtained through a straightforward modification of the familiar algorithm for generating permutation-based thresholds.

LDL receptor but not apolipoprotein E deficiency increases diet-induced obesity and diabetes in mice.

The aim of this study was to determine whether phenotypes associated with type 2 diabetes are altered in dyslipidemic obese mice. C57BL/6 wild-type, low-density lipoprotein (LDL) receptor-deficient (LDLR-/-), and apolipoprotein E-deficient (apoE-/-) mice were fed a high-fat, high-carbohydrate diet (diabetogenic diet), and the development of obesity, diabetes, and hypertriglyceridemia was examined. Wild-type mice became obese and developed hyperglycemia, but not hypertriglyceridemia, in response to this diet. LDLR-/- mice fed the diabetogenic diet became more obese than wild-type mice and developed severe hypertriglyceridemia and hyperleptinemia. Surprisingly, glucose levels were only modestly higher and insulin levels and insulin-to-glucose ratios were not strikingly different from those of wild-type mice. In contrast, diabetogenic diet-fed apoE-/- mice were resistant to changes in glucose and lipid homeostasis despite becoming obese. These data suggest that modifications in lipoprotein profiles associated with loss of the LDL receptor or apoE function have profound and unique consequences on susceptibility to diet-induced obesity and type 2 diabetic phenotypes.