Danh V. Nguyen

Tumor classification by partial least squares using microarray gene expression data

MOTIVATION One important application of gene expression microarray data is classification of samples into categories, such as the type of tumor. The use of microarrays allows simultaneous monitoring of thousands of genes expressions per sample. This ability to measure gene expression en masse has resulted in data with the number of variables p(genes) far exceeding the number of samples N. Standard statistical methodologies in classification and prediction do not work well or even at all when N < p. Modification of existing statistical methodologies or development of new methodologies is needed for the analysis of microarray data. RESULTS We propose a novel analysis procedure for classifying (predicting) human tumor samples based on microarray gene expressions. This procedure involves dimension reduction using Partial Least Squares (PLS) and classification using Logistic Discrimination (LD) and Quadratic Discriminant Analysis (QDA). We compare PLS to the well known dimension reduction method of Principal Components Analysis (PCA). Under many circumstances PLS proves superior; we illustrate a condition when PCA particularly fails to predict well relative to PLS. The proposed methods were applied to five different microarray data sets involving various human tumor samples: (1) normal versus ovarian tumor; (2) Acute Myeloid Leukemia (AML) versus Acute Lymphoblastic Leukemia (ALL); (3) Diffuse Large B-cell Lymphoma (DLBCLL) versus B-cell Chronic Lymphocytic Leukemia (BCLL); (4) normal versus colon tumor; and (5) Non-Small-Cell-Lung-Carcinoma (NSCLC) versus renal samples. Stability of classification results and methods were further assessed by re-randomization studies.

A pilot open-label single-dose trial of fenobam in adults with fragile X syndrome

Objective: A pilot open label, single dose trial of fenobam, an mGluR5 antagonist, was conducted to provide an initial evaluation of safety and pharmacokinetics in adult males and females with fragile X syndrome (FXS). Methods: Twelve subjects, recruited from two fragile X clinics, received a single oral dose of 50–150 mg of fenobam. Blood for pharmacokinetic testing, vital signs and side effect screening was obtained at baseline and numerous time points for 6 h after dosing. Outcome measures included prepulse inhibition (PPI) and a continuous performance test (CPT) obtained before and after dosing to explore the effects of fenobam on core phenotypic measures of sensory gating, attention and inhibition. Results: There were no significant adverse reactions to fenobam administration. Pharmacokinetic analysis showed that fenobam concentrations were dose dependent but variable, with mean (SEM) peak values of 39.7 (18.4) ng/ml at 180 min after the 150 mg dose. PPI met a response criterion of an improvement of at least 20% over baseline in 6 of 12 individuals (4/6 males and 2/6 females). The CPT did not display improvement with treatment due to ceiling effects. Conclusions: Clinically significant adverse effects were not identified in this study of single dose fenobam across the range of dosages utilised. The positive effects seen in animal models of FXS treated with fenobam or other mGluR5 antagonists, the apparent lack of clinically significant adverse effects, and the potential beneficial clinical effects seen in this pilot trial support further study of the compound in adults with FXS.

Expanded clinical phenotype of women with the FMR1 premutation

Fragile X‐associated tremor/ataxia syndrome (FXTAS) is generally considered to be uncommon in older female carriers of premutation alleles (55–200 CGG repeats) of the fragile X mental retardation 1 (FMR1) gene; however, neither prevalence, nor the nature of the clinical phenotype, has been well characterized in female carriers. In this study, we evaluated 146 female carriers (mean, 42.3 years; range, 20–75 years) with and without core features of FXTAS (tremor; gait ataxia), and 69 age‐matched controls (mean, 45.8 years; range, 21–78 years). Compared with controls, carriers with definite or probable FXTAS had greater medical co‐morbidity, with increased prevalence of thyroid disease (P = 0.0096), hypertension (P = 0.0020), seizures (P = 0.0077), peripheral neuropathy (P = 0.0040), and fibromyalgia (P = 0.0097), in addition to the typical symptoms of FXTAS–tremor (P < 0.0001) and ataxia (P < 0.0001). The non‐FXTAS premutation group had more complaints of chronic muscle pain (P = 0.0097), persistent paraesthesias in extremities (P < 0.0001), and history of tremor (P < 0.0123) than controls. The spectrum of clinical involvement in female carriers with FXTAS is quite broad, encompassing a number of medical co‐morbidities as well as the core movement disorder. The remarkable degree of thyroid dysfunction (17% in the non‐FXTAS group and 50% in the FXTAS group) warrants consideration of thyroid function studies in all female premutation carriers, particularly those with core features of FXTAS.

Multi-class cancer classification via partial least squares with gene expression profiles

MOTIVATION Discrimination between two classes such as normal and cancer samples and between two types of cancers based on gene expression profiles is an important problem which has practical implications as well as the potential to further our understanding of gene expression of various cancer cells. Classification or discrimination of more than two groups or classes (multi-class) is also needed. The need for multi-class discrimination methodologies is apparent in many microarray experiments where various cancer types are considered simultaneously. RESULTS Thus, in this paper we present the extension to the classification methodology proposed earlier Nguyen and Rocke (2002b; Bioinformatics, 18, 39-50) to classify cancer samples from multiple classes. The methodologies proposed in this paper are applied to four gene expression data sets with multiple classes: (a) a hereditary breast cancer data set with (1) BRCA1-mutation, (2) BRCA2-mutation and (3) sporadic breast cancer samples, (b) an acute leukemia data set with (1) acute myeloid leukemia (AML), (2) T-cell acute lymphoblastic leukemia (T-ALL) and (3) B-cell acute lymphoblastic leukemia (B-ALL) samples, (c) a lymphoma data set with (1) diffuse large B-cell lymphoma (DLBCL), (2) B-cell chronic lymphocytic leukemia (BCLL) and (3) follicular lymphoma (FL) samples, and (d) the NCI60 data set with cell lines derived from cancers of various sites of origin. In addition, we evaluated the classification algorithms and examined the variability of the error rates using simulations based on randomization of the real data sets. We note that there are other methods for addressing multi-class prediction recently and our approach is along the line of Nguyen and Rocke (2002b; Bioinformatics, 18, 39-50). CONTACT dnguyen@stat.tamu.edu; dmrocke@ucdavis.edu

Effects of STX209 (Arbaclofen) on Neurobehavioral Function in Children and Adults with Fragile X Syndrome: A Randomized, Controlled, Phase 2 Trial

Administration of a selective GABAB agonist to individuals with fragile X syndrome improves their deficits in social avoidance, a core symptom of the disease. A Fragile Balancing Act A wide array of symptoms—including intellectual disability, anxiety, seizures, and autistic behavior—are associated with fragile X syndrome (FXS). Although some symptoms can be managed (or masked) with drugs or other therapies, treatments that target the fundamental impairments are not available. Henderson et al. and Berry-Kravis et al. now provide evidence that activation of a particular neuronal receptor can improve symptoms in both mice and humans. FXS is caused by silencing of the FMR1 gene, which encodes FMRP, an RNA binding protein that inhibits protein synthesis. In a mouse model of FXS, dendritic protein synthesis is abnormally high; FMRP is believed to regulate mRNAs important for neuronal development. Furthermore, these mice—and some humans with FXS—have an increased density of dendritic spines, which are dynamic structures that make neuronal connections. Dendritic spine plasticity is linked to learning and memory. Normally, FMRP may balance mRNA translation that is stimulated by activation of synaptic receptors that respond to glutamate, an excitatory neurotransmitter. Indeed, inhibitors of these receptors rescue many irregular phenotypes in the animal models but are not yet approved for human use. These mice also exhibit deficient signaling through a different set of receptors, which respond to the inhibitory neurotransmitter GABA—and for which clinically approved agonists already exist. Henderson et al. tested one such GABAB receptor agonist, STX209, in the mouse model and found that it decreased mRNA translation in the cortex and corrected the increased dendritic spine density. Berry-Kravis et al. studied the effects of STX209 in a double-blind, placebo-controlled crossover trial, in which 63 FXS patients received placebo or drug for 4 weeks and then switched to the other treatment. Although a measure of irritability and aggression was unchanged, social avoidance improved; the drug was well tolerated. Thus, this targeted approach, which may help restore the balance between excitatory and inhibitory neurotransmission, has promise for improving social function in FXS. Research on animal models of fragile X syndrome suggests that STX209, a γ-aminobutyric acid type B (GABAB) agonist, might improve neurobehavioral function in affected patients. We evaluated whether STX209 improves behavioral symptoms of fragile X syndrome in a randomized, double-blind, placebo-controlled crossover study in 63 subjects (55 male), ages 6 to 39 years, with a full mutation in the FMR1 gene (>200 CGG triplet repeats). We found no difference from placebo on the primary endpoint, the Aberrant Behavior Checklist—Irritability (ABC-I) subscale. In the other analyses specified in the protocol, improvement was seen on the visual analog scale ratings of parent-nominated problem behaviors, with positive trends on multiple global measures. Post hoc analysis with the ABC—Social Avoidance scale, a newly validated scale for the assessment of fragile X syndrome, showed a significant beneficial treatment effect in the full study population. A post hoc subgroup of 27 subjects with more severe social impairment showed improvements on the Vineland II–Socialization raw score, on the ABC—Social Avoidance scale, and on all global measures. STX209 was well tolerated, with 8% incidences of sedation and of headache as the most frequent side effects. In this exploratory study, STX209 did not show a benefit on irritability in fragile X syndrome. Nonetheless, our results suggest that GABAB agonists have potential to improve social function and behavior in patients with fragile X syndrome.

FMR1 CGG allele size and prevalence ascertained through newborn screening in the United States

BackgroundPopulation screening for FMR1 mutations has been a topic of considerable discussion since the FMR1 gene was identified in 1991. Advances in understanding the molecular basis of fragile X syndrome (FXS) and in genetic testing methods have led to new, less expensive methodology to use for large screening endeavors. A core criterion for newborn screening is an accurate understanding of the public health burden of a disease, considering both disease severity and prevalence rate. This article addresses this need by reporting prevalence rates observed in a pilot newborn screening study for FXS in the US.MethodsBlood spot screening of 14,207 newborns (7,312 males and 6,895 females) was conducted in three birthing hospitals across the United States beginning in November 2008, using a PCR-based approach.ResultsThe prevalence of gray zone alleles was 1:66 females and 1:112 males, while the prevalence of a premutation was 1:209 females and 1:430 males. Differences in prevalence rates were observed among the various ethnic groups; specifically higher frequency for gray zone alleles in males was observed in the White group compared to the Hispanic and African-American groups. One full mutation male was identified (>200 CGG repeats).ConclusionsThe presented pilot study shows that newborn screening in fragile X is technically feasible and provides overall prevalence of the premutation and gray zone alleles in the USA, suggesting that the prevalence of the premutation, particularly in males, is higher than has been previously reported.

Evidence of Increased Inflammation and Microcirculatory Abnormalities in Patients With Type 1 Diabetes and Their Role in Microvascular Complications

OBJECTIVE—Type 1 diabetes is associated with increased microvascular complications and inflammation. The monocyte-macrophage is a pivotal cell in atherogenesis. There are scanty data on noninvasive measures of microvascular abnormalities and inflammation in type 1 diabetic subjects with microvascular complications. Thus, we examined systemic and cellular biomarkers of inflammation in type 1 diabetic patients with microvascular complications (T1DM-MV patients) and type 1 diabetic patients without microvascular complications (T1DM patients) compared with matched control subjects and determined the microcirculatory abnormalities in the T1DM and T1DM-MV patients using computer-assisted intravital microscopy (CAIM). RESEARCH DESIGN AND METHODS—Fasting blood, 24-h urine, and CAIM measurements were obtained from the T1DM and T1DM-MV patients and matched control subjects. C-reactive protein, E-selectin, nitrotyrosine, monocyte superoxide, and cytokines were elevated in the T1DM and T1DM-MV patients compared with control subjects (P < 0.01). RESULTS—Severity index, as assessed by CAIM, was significantly increased in the T1DM and T1DM-MV patients compared with the control subjects (P < 0.001). There was a significant increase in C-reactive protein, nitrotyrosine, vascular cell adhesion molecule and monocyte superoxide anion release, and interleukin-1 release in T1DM-MV compared with T1DM patients (P < 0.05). T1DM-MV patients had significantly increased CAIM severity index and microalbumin-to-creatinine ratio compared with T1DM patients (P < 0.05). Furthermore, pp38MAPK, pp65, and pERK activity were significantly increased in monocytes from the T1DM and T1DM-MV patients compared with those from the controls subjects, and pp38MAPK and pp65 activity were significantly increased in the T1DM-MV compared with the T1DM patients (P < 0.01). CONCLUSIONS—T1DM-MV patients have increased inflammation compared with T1DM patients. CAIM provides an effective biomarker of microvascular complications, since it is significantly elevated in T1DM-MV compared with T1DM patients and can be monitored following therapies targeted at improving inflammation and/or microvascular complications of type 1 diabetes.

Chemopreventive n-3 Polyunsaturated Fatty Acids Reprogram Genetic Signatures during Colon Cancer Initiation and Progression in the Rat

The mechanisms by which n-3 polyunsaturated fatty acids (PUFAs) decrease colon tumor formation have not been fully elucidated. Examination of genes up- or down-regulated at various stages of tumor development via the monitoring of gene expression relationships will help to determine the biological processes ultimately responsible for the protective effects of n-3 PUFA. Therefore, using a 3 × 2 × 2 factorial design, we used Codelink DNA microarrays containing ∼9000 genes to help decipher the global changes in colonocyte gene expression profiles in carcinogen-injected Sprague Dawley rats. Animals were assigned to three dietary treatments differing only in the type of fat (corn oil/n-6 PUFA, fish oil/n-3 PUFA, or olive oil/n-9 monounsaturated fatty acid), two treatments (injection with the carcinogen azoxymethane or with saline), and two time points (12 hours and 10 weeks after first injection). Only the consumption of n-3 PUFA exerted a protective effect at the initiation (DNA adduct formation) and promotional (aberrant crypt foci) stages. Importantly, microarray analysis of colonocyte gene expression profiles discerned fundamental differences among animals treated with n-3 PUFA at both the 12 hours and 10-week time points. Thus, in addition to demonstrating that dietary fat composition alters the molecular portrait of gene expression profiles in the colonic epithelium at both the initiation and promotional stages of tumor development, these findings indicate that the chemopreventive effect of fish oil is due to the direct action of n-3 PUFA and not to a reduction in the content of n-6 PUFA.

A Randomized Double-Blind, Placebo-Controlled Trial of Minocycline in Children and Adolescents with Fragile X Syndrome

Objective: Minocycline rescued synaptic abnormalities and improved behavior in the fragile X mouse model. Previous open-label human studies demonstrated benefits in individuals with fragile X syndrome (FXS); however, its efficacy in patients with FXS has not been assessed in a controlled trial. Method: Randomized, double-blind, placebo-controlled, crossover trial in individuals with FXS, aged 3.5 years to 16 years (n = 55, mean age 9.2 [SD, 3.6] years). Participants were randomized to minocycline or placebo for 3 months and then switched to the other treatment. Results: Sixty-nine subjects were screened and 66 were randomized. Fifty-five subjects (83.3%) completed at least the first period and 48 (72.7%) completed the full trial. Intention-to-treat analysis demonstrated significantly greater improvements in one primary outcome, Clinical Global Impression Scale—Improvement after minocycline compared with placebo (2.49 ± 0.13 and 2.97 ± 0.13, respectively, p = .0173) and greater improvement in ad hoc analysis of anxiety and mood-related behaviors on the Visual Analog Scale (minocycline: 5.26 cm ± 0.46 cm, placebo: 4.05 cm ± 0.46 cm; p = .0488). Side effects were not significantly different during the minocycline and placebo treatments. No serious adverse events occurred on minocycline. Results may be potentially biased by study design weaknesses, including unblinding of subjects when they completed the study, drug-related side effects unblinding, and preliminary efficacy analysis results known to investigators. Conclusions: Minocycline treatment for 3 months in children with FXS resulted in greater global improvement than placebo. Treatment for 3 months appears safe; however, longer trials are indicated to further assess benefits, side effects, and factors associated with a clinical response to minocycline.

Partial least squares proportional hazard regression for application to DNA microarray survival data

MOTIVATION Microarrays are increasingly used in cancer research. When gene transcription data from microarray experiments also contains patient survival information, it is often of interest to predict the survival times based on the gene expression. In this paper we consider the well-known proportional hazard (PH) regression model for survival analysis. Ordinarily, the PH model is used with a few covariates and many observations (subjects). We consider here the case that the number of covariates, p, exceeds the number of samples, N, a setting typical of gene expression data from DNA microarrays. RESULTS For a given vector of response values which are survival times and p gene expressions (covariates) we examine the problem of how to predict the survival probabilities, when N << p. The approach taken to cope with the high dimensionality is to reduce the dimension using partial least squares with the response variable as the vector of survival times. After dimension reduction, the extracted PLS gene components are then used as covariates in a PH regression to predict the survival probabilities. We demonstrate the use of the methodology on two cDNA gene expression data sets, both containing survival data. The first data set contains 40 diffuse large B-cell lymphoma (DLBCL) tissue samples and the second data set contains 49 tissue samples from patients with locally advanced breast cancer in a prospective study.