Jeffrey R. Sachs

Increasing the Power to Detect Causal Associations by Combining Genotypic and Expression Data in Segregating Populations

To dissect common human diseases such as obesity and diabetes, a systematic approach is needed to study how genes interact with one another, and with genetic and environmental factors, to determine clinical end points or disease phenotypes. Bayesian networks provide a convenient framework for extracting relationships from noisy data and are frequently applied to large-scale data to derive causal relationships among variables of interest. Given the complexity of molecular networks underlying common human disease traits, and the fact that biological networks can change depending on environmental conditions and genetic factors, large datasets, generally involving multiple perturbations (experiments), are required to reconstruct and reliably extract information from these networks. With limited resources, the balance of coverage of multiple perturbations and multiple subjects in a single perturbation needs to be considered in the experimental design. Increasing the number of experiments, or the number of subjects in an experiment, is an expensive and time-consuming way to improve network reconstruction. Integrating multiple types of data from existing subjects might be more efficient. For example, it has recently been demonstrated that combining genotypic and gene expression data in a segregating population leads to improved network reconstruction, which in turn may lead to better predictions of the effects of experimental perturbations on any given gene. Here we simulate data based on networks reconstructed from biological data collected in a segregating mouse population and quantify the improvement in network reconstruction achieved using genotypic and gene expression data, compared with reconstruction using gene expression data alone. We demonstrate that networks reconstructed using the combined genotypic and gene expression data achieve a level of reconstruction accuracy that exceeds networks reconstructed from expression data alone, and that fewer subjects may be required to achieve this superior reconstruction accuracy. We conclude that this integrative genomics approach to reconstructing networks not only leads to more predictive network models, but also may save time and money by decreasing the amount of data that must be generated under any given condition of interest to construct predictive network models.

Application of an End-to-End Biomarker Discovery Platform to Identify Target Engagement Markers in Cerebrospinal Fluid by High Resolution Differential Mass Spectrometry

The rapid identification of protein biomarkers in biofluids is important to drug discovery and development. Here, we describe a general proteomic approach for the discovery and identification of proteins that exhibit a statistically significant difference in abundance in cerebrospinal fluid (CSF) before and after pharmacological intervention. This approach, differential mass spectrometry (dMS), is based on the analysis of full scan mass spectrometry data. The dMS workflow does not require complex mixing and pooling strategies, or isotope labeling techniques. Accordingly, clinical samples can be analyzed individually, allowing the use of longitudinal designs and within-subject data analysis in which each subject acts as its own control. As a proof of concept, we performed multifactorial dMS analyses on CSF samples drawn at 6 time points from n = 6 cisterna magna ported (CMP) rhesus monkeys treated with 2 potent gamma secretase inhibitors (GSI) or comparable vehicle in a 3-way crossover study that included a total of 108 individual CSF samples. Using analysis of variance and statistical filtering on the aligned and normalized LC-MS data sets, we detected 26 features that were significantly altered in CSF by drug treatment. Of those 26 features, which belong to 10 distinct isotopic distributions, 20 were identified by MS/MS as 7 peptides from CD99, a cell surface protein. Six features from the remaining 3 isotopic distributions were not identified. A subsequent analysis showed that the relative abundance of these 26 features showed the same temporal profile as the ELISA measured levels of CSF A beta 42 peptide, a known pharmacodynamic marker for gamma-secretase inhibition. These data demonstrate that dMS is a promising approach for the discovery, quantification, and identification of candidate target engagement biomarkers in CSF.

Optimal Dosing for Targeted Therapies in Oncology: Drug Development Cases Leading by Example

One of the key objectives of oncology first-in-human trials has often been to establish the maximum tolerated dose (MTD). However, targeted therapies might not exhibit dose-limiting toxicities (DLT) at doses significantly higher than sufficiently active doses, and there is frequently a limited ability to objectively quantify adverse events. Thus, while MTD-based determination of recommended phase II dose may have yielded appropriate dosing for some cytotoxics, targeted therapeutics (including monoclonal antibodies and/or immunotherapies) sometimes need alternative or complementary strategies to help identify dose ranges for a randomized dose-ranging study. One complementary strategy is to define a biologically efficacious dose (BED) using an “effect marker.” An effect marker could be a target engagement, pharmacodynamic, or disease progression marker (change in tumor size for solid tumors or bone marrow blast count for some hematologic tumors). Although the concept of BED has been discussed extensively, we review specific examples in which the approach influenced oncology clinical development. Data extracted from the literature and the examples support improving dose selection strategies to benefit patients, providers, and the biopharmaceutical industry. Although the examples illustrate key contributions of effect markers in dose selection, no one-size-fits-all approach to dosing can be justified. Higher-than-optimal dosing can increase toxicity in later trials (and in clinical use), which can have a negative impact on efficacy (via lower adherence or direct sequelae of toxicities). Proper dose selection in oncology should follow a multifactorial decision process leading to a randomized, dose-ranging study instead of a single phase II dose. Clin Cancer Res; 22(6); 1318–24. ©2015 AACR.

High Resolution Discovery Proteomics Reveals Candidate Disease Progression Markers of Alzheimer's Disease in Human Cerebrospinal Fluid.

Disease modifying treatments for Alzheimer’s disease (AD) constitute a major goal in medicine. Current trends suggest that biomarkers reflective of AD neuropathology and modifiable by treatment would provide supportive evidence for disease modification. Nevertheless, a lack of quantitative tools to assess disease modifying treatment effects remains a major hurdle. Cerebrospinal fluid (CSF) biochemical markers such as total tau, p-tau and Ab42 are well established markers of AD; however, global quantitative biochemical changes in CSF in AD disease progression remain largely uncharacterized. Here we applied a high resolution open discovery platform, dMS, to profile a cross-sectional cohort of lumbar CSF from post-mortem diagnosed AD patients versus those from non-AD/non-demented (control) patients. Multiple markers were identified to be statistically significant in the cohort tested. We selected two markers SME-1 (p<0.0001) and SME-2 (p = 0.0004) for evaluation in a second independent longitudinal cohort of human CSF from post-mortem diagnosed AD patients and age-matched and case-matched control patients. In cohort-2, SME-1, identified as neuronal secretory protein VGF, and SME-2, identified as neuronal pentraxin receptor-1 (NPTXR), in AD were 21% (p = 0.039) and 17% (p = 0.026) lower, at baseline, respectively, than in controls. Linear mixed model analysis in the longitudinal cohort estimate a decrease in the levels of VGF and NPTXR at the rate of 10.9% and 6.9% per year in the AD patients, whereas both markers increased in controls. Because these markers are detected by mass spectrometry without the need for antibody reagents, targeted MS based assays provide a clear translation path for evaluating selected AD disease-progression markers with high analytical precision in the clinic.

Using Model-Based “Learn and Confirm” to Reveal the Pharmacokinetics-Pharmacodynamics Relationship of Pembrolizumab in the KEYNOTE-001 Trial

Evaluation of pharmacokinetic/pharmacodynamic (PK/PD) properties played an important role in the early clinical development of pembrolizumab. Because analysis of data from a traditional 3 + 3 dose‐escalation design revealed several critical uncertainties, a model‐based approach was implemented to better understand these properties. Based on anticipated scenarios for potency and PK nonlinearity, a follow‐up study was designed and thoroughly evaluated. Execution of 14,000 virtual trials led to the selection and implementation of a robust design that extended the low‐dose range by 200‐fold. Modeling of the resulting data demonstrated that pembrolizumab PKs are nonlinear at <0.3 mg/kg every 3 weeks, but linear in the clinical dose range. Saturation of ex vivo target engagement in blood began at ≥1 mg/kg every 3 weeks, and a steady‐state dose of 2 mg/kg every 3 weeks was needed to reach 95% target engagement, supporting examination of 2 mg/kg every 3 weeks in ongoing trials in melanoma and other advanced cancers.

The CRASSS plug-in for integrating annotation data with hierarchical clustering results

We describe an algorithm for finding the most statistically significant non-overlapping subtrees of a hierarchical clustering of gene expression data with respect to a set of secondary data labels on genes. The method is implemented as a Java plug-in for a commercial gene expression analysis program (GeneSpring).

Segmentation of arterial vessel wall motion to sub-pixel resolution using M-mode ultrasound

We describe a method for segmenting arterial vessel wall motion to sub-pixel resolution, using the returns from M-mode ultrasound. The technique involves measuring the spatial offset between all pairs of scans from their cross-correlation, converting the spatial offsets to relative wall motion through a global optimization, and finally translating from relative to absolute wall motion by interpolation over the M-mode image. The resulting detailed wall distension waveform has the potential to enhance existing vascular biomarkers, such as strain and compliance, as well as enable new ones.

Electrokinetics and electromechanics in controlled release from ionizable gels: theory & experiments

Hydrogels containing weak acid or base pendant groups swell and release drugs differentially as a function of environmental pH and ionic strength. Also, the identity and concentration of buffering species in the release medium can affect swelling and drug release kinetics. A full understanding of factors affecting swelling and release would be reflected in a quantitative mathematical model of these processes. The authors develop a finite element computer model which predicts swelling and release profiles from spherical gels given specified values of mechanical and chemical parameters of the hydrogel, drug, and release medium. The code contains electromechanical and electrochemical components. Numerical computations based on a mathematical model predict gel swelling and stresses, charge density, solute concentrations, and solute and solvent fluxes. The model is validated by comparing its predictions with an experiment on controlled release of sodium salicylate.<<ETX>>

Leveraging Digital Health Technologies and Outpatient Sampling in Clinical Drug Development: A Phase I Exploratory Study

Merck & Co, Inc (Kenilworth, NJ) is investing in approaches to enrich clinical trial data and augment decision making through use of digital health technologies, outpatient sampling, and real‐time data access. As part of this strategy, a phase I study was conducted to explore a few technologies of interest. In this fixed‐sequence two‐period trial, 16 healthy subjects were administered 50‐mg once‐daily sitagliptin packaged in a bottle that electronically captured the date and time study medication was dispensed (period 1) and in a traditional pharmacy bottle (period 2). Dried blood spot samples were collected for sitagliptin concentration analysis on select study days, both in clinic and at home, with collection time recorded using an electronic diary in period 1 and by clinic staff in period 2. Study results demonstrated the feasibility and subject acceptance of collecting digital adherence data and outpatient dried blood spot samples in clinical trials and highlighted areas for future improvements.

Differential mass spectrometry identifies candidate markers for Alzheimer's disease in humans

(TMT) A and B, denomination (DO80), digit span tests, battery of frontal function were performed in AD patients. All patients gave their informed consent. CSF biomarkers (Ab 1-42, tau and p181 tau) were assessed after lumbar puncture in the month following cognitive tests. Spearman correlation coefficients between CSF biomarkers and the different cognitive tests were calculated. Generalized linear models adjusted for age, sex and level of education were used for multivariable analysis. Results: MMSE scores were inversely associated with the levels of CSF Ab 1-42 (Spearman’s rho 1⁄4 0.28, p 1⁄4 0.03) and TMT B scores were associated with tau (Spearman’s rho1⁄4 0.37, p1⁄4 0.02) and p181-tau (Spearman’s rho1⁄4 0.45, p1⁄4 0.004) CSF levels. These associations remained after adjustment for age, sex, and level of education in multivariable analysis. No correlation was observed with the other cognitive tests. Conclusions: Our results show that in a cohort of recently diagnosed AD patients, CSF biomarkers levels can correlate with the results of cognitive functions. A b 1-42 levels are linked to MMSE scores reflecting the global cognitive functions and tau and p181 tau CSF levels are correlated with executive functions assessed by the TMT B test.