Faculty Bibliography

BACKGROUND:, P-tau, and T-tau with sleep spindle density and other biophysical properties of sleep spindles in a sample of cognitively normal elderly individuals. METHODS:, P-tau and T-tau. Seven days of actigraphy were collected to assess habitual total sleep time. RESULTS:, P-tau and T-tau. From the three, CSF T-tau was the most significantly associated with spindle density, after adjusting for age, sex and ApoE4. Spindle duration, count and fast spindle density were also negatively correlated with T-tau levels. Sleep duration and other measures of sleep quality were not correlated with spindle characteristics and did not modify the associations between sleep spindle characteristics and the CSF biomarkers of AD. CONCLUSIONS:Reduced spindles during N2 sleep may represent an early dysfunction related to tau, possibly reflecting axonal damage or altered neuronal tau secretion, rendering it a potentially novel biomarker for early neuronal dysfunction. Given their putative role in memory consolidation and neuroplasticity, sleep spindles may represent a mechanism by which tau impairs memory consolidation, as well as a possible target for therapeutic interventions in cognitive decline.

Hutson and Vexler (2018) demonstrate an example of aliasing with the beta and normal distribution. This letter presents another illustration of aliasing using the beta and normal distributions via an infinite mixture model, inspired by the problem of modeling placebo response.

This study aimed to identify biomarkers of major depressive disorder (MDD), by relating neuroimage-derived measures to binary (MDD/control), ordinal (severe MDD/mild MDD/control), or continuous (depression severity) outcomes. To address MDD heterogeneity, factors (severity of psychic depression, motivation, anxiety, psychosis, and sleep disturbance) were also used as outcomes. A multisite, multimodal imaging (diffusion MRI [dMRI] and structural MRI [sMRI]) cohort (52 controls and 147 MDD patients) and several modeling techniques-penalized logistic regression, random forest, and support vector machine (SVM)-were used. An additional cohort (25 controls and 83 MDD patients) was used for validation. The optimally performing classifier (SVM) had a 26.0% misclassification rate (binary), 52.2 ± 1.69% accuracy (ordinal) and r = .36 correlation coefficient (p < .001, continuous). Using SVM, R² values for prediction of any MDD factors were <10%. Binary classification in the external data set resulted in 87.95% sensitivity and 32.00% specificity. Though observed classification rates are too low for clinical utility, four image-based features contributed to accuracy across all models and analyses-two dMRI-based measures (average fractional anisotropy in the right cuneus and left insula) and two sMRI-based measures (asymmetry in the volume of the pars triangularis and the cerebellum) and may serve as a priori regions for future analyses. The poor accuracy of classification and predictive results found here reflects current equivocal findings and sheds light on challenges of using these modalities for MDD biomarker identification. Further, this study suggests a paradigm (e.g., multiple classifier evaluation with external validation) for future studies to avoid nongeneralizable results.

Treatment response heterogeneity poses serious challenges for selecting treatment for many diseases. To better understand this heterogeneity and to help in determining the best patient-specific treatments for a given disease, many clinical trials are collecting large amounts of patient-level data prior to administering treatment in the hope that some of these data can be used to identify moderators of treatment effect. These data can range from simple scalar values to complex functional data such as curves or images. Combining these various types of baseline data to discover "biosignatures" of treatment response is crucial for advancing precision medicine. Motivated by the problem of selecting optimal treatment for subjects with depression based on clinical and neuroimaging data, we present an approach that both (1) identifies covariates associated with differential treatment effect and (2) estimates a treatment decision rule based on these covariates. We focus on settings where there is a potentially large collection of candidate biomarkers consisting of both scalar and functional data. The validity of the proposed approach is justified via extensive simulation experiments and illustrated using data from a placebo-controlled clinical trial investigating antidepressant treatment response in subjects with depression.

To date, there are no biomarkers for major depressive disorder (MDD) treatment response in clinical use. Such biomarkers could allow for individualized treatment selection, reducing time spent on ineffective treatments and the burden of MDD. In search of such a biomarker, multisite pretreatment and early-treatment (1 week into treatment) structural magnetic resonance (MR) images were acquired from 184 patients with MDD randomized to an 8-week trial of the selective serotonin reuptake inhibitor (SSRI) sertraline or placebo. This study represents a large, multisite, placebo-controlled effort to examine the association between pretreatment differences or early-treatment changes in cortical thickness and treatment-specific outcomes. For standardization, a novel, robust site harmonization procedure was applied to structural measures in a priori regions (rostral and caudal anterior cingulate, lateral orbitofrontal, rostral middle frontal, and hippocampus), chosen based on previously published reports. Pretreatment cortical thickness or volume did not significantly associate with SSRI response. Thickening of the rostral anterior cingulate cortex in the first week of treatment was associated with better 8-week responses to SSRI (p = 0.010). These findings indicate that frontal lobe structural alterations in the first week of treatment may be associated with long-term treatment efficacy. While these associational findings may help to elucidate the specific neural targets of SSRIs, the predictive accuracy of pretreatment or early-treatment structural alterations in classifying treatment remitters from nonremitters was limited to 63.9%. Therefore, in this large sample of adults with MDD, structural MR imaging measures were not found to be clinically translatable biomarkers of treatment response to SSRI or placebo.

Although there are changes in gene expression and alterations in neuronal density and afferent inputs in the forebrain of trisomic mouse models of Down syndrome (DS) and Alzheimer's disease (AD), there is a lack of systematic assessments of gene expression and encoded proteins within individual vulnerable cell populations, precluding translational investigations at the molecular and cellular level. Further, no effective treatment exists to combat intellectual disability and basal forebrain cholinergic neurodegeneration seen in DS. To further our understanding of gene expression changes before and following cholinergic degeneration in a well-established mouse model of DS/AD, the Ts65Dn mouse, we assessed RNA expression levels from CA1 pyramidal neurons at two adult ages (∼6 months of age and ∼11 months of age) in both Ts65Dn and their normal disomic (2N) littermates. We further examined a viable therapeutic, maternal choline supplementation (MCS), which has been previously shown to lessen dysfunction in spatial cognition and attention, and have protective effects on the survival of basal forebrain cholinergic neurons (BFCNs) in the Ts65Dn mouse model. Results indicate that MCS normalized expression of several genes in key gene ontology categories, including synaptic plasticity, calcium signaling, and AD-associated neurodegeneration related to amyloid-beta peptide (Aβ) clearance. Specifically, normalized expression levels were found for endothelin converting enzyme-2 (Ece2), insulin degrading enzyme (Ide), Dyrk1a, and calcium/calmodulin-dependent protein kinase II (Camk2a), among other relevant genes. Single population expression profiling of vulnerable CA1 pyramidal neurons indicates that MCS is a viable therapeutic for long-term reprogramming of key transcripts involved in neuronal signaling that are dysregulated in the trisomic mouse brain which have translational potential for DS and AD.

BACKGROUND: Deficits in N-methyl-d-aspartate-type (NMDAR) function contribute to symptoms and cognitive dysfunction in schizophrenia. The efficacy of NMDAR agonists in the treatment of persistent symptoms of schizophrenia has been variable, potentially reflecting limitations in functional target engagement. We recently demonstrated significant improvement in auditory mismatch negativity (MMN) with once-weekly treatment with d-serine, a naturally occurring NMDAR glycine-site agonist. This study investigates effects of continuous (daily) NMDAR agonists in schizophrenia/schizoaffective disorder. METHODS: Primary analysis was on MMN after double-blind crossover (60mg/kg/d, n=16, 6weeks) treatment with d-serine/placebo. Secondary measures included clinical symptoms, neurocognition, and the effects of open-label (30-120mg/kg/d, n=21) d-serine and bitopertin/placebo (10mg, n=29), a glycine transport inhibitor. RESULTS: Double-blind d-serine treatment led to significant improvement in MMN frequency (p=0.001, d=2.3) generation and clinical symptoms (p=0.023, d=0.80). MMN frequency correlated significantly with change in symptoms (r=-0.63, p=0.002) following co-variation for treatment type. d-Serine treatment led to a significant, large effect size increase vs. placebo in evoked alpha-power in response to standards (p=0.036, d=0.81), appearing to normalize evoked alpha power relative to previous findings with controls. While similar results were seen with open-label d-serine, no significant effects of bitopertin were observed for symptoms or MMN. CONCLUSIONS: These findings represent the first randomized double-blind placebo-controlled study with 60mg/kg d-serine in schizophrenia, and are consistent with meta-analyses showing significant effects of d-serine in schizophrenia. Results overall support suggest that MMN may have negative, as well as positive, predictive value in predicting efficacy of novel compounds. CLINICAL TRIALS REGISTRATION: Clinicaltrials.gov: NCT00322023/NCT00817336 (d-serine); NCT01116830 (bitopertin).

Latent class models are widely used to identify unobserved subgroups (i.e., latent classes) based upon one or more manifest variables. The probability of belonging to each subgroup is typically modeled as a function of a set of measured covariates. In this paper, we extend existing latent class models to incorporate matrix covariates. This research is motivated by a randomized placebo-controlled depression clinical trial. One study goal is to identify a subgroup of subjects who experience symptoms improvement early on during antidepressant treatment, which is considered to be an indication of a placebo rather than a true pharmacological response. We want to relate the likelihood of belonging to this subgroup of early responders to baseline electroencephalography (EEG) measurement that takes the form of a matrix. The proposed method is built upon a low rank Candecomp/Parafac (CP) decomposition of the target coefficient matrix through low-dimensional latent variables, which effectively reduces the model dimensionality. We adopt a Bayesian hierarchical modeling approach to estimate the latent variables, which allows a flexible way to incorporate prior knowledge about covariate effect heterogeneity and offers a data-driven method of regularization. Simulation studies suggest that the proposed method is robust against potentially misspecified rank in the CP decomposition. With the motivating example we show how the proposed method can be applied to extract valuable information from baseline EEG measurements that explains the likelihood of belonging to the early responder subgroup, helping to identify placebo responders and suggesting new targets for the study of placebo response.