A computational framework for complex disease stratification from multiple large-scale datasets.

BACKGROUND: Multilevel data integration is becoming a major area of research in systems biology. Within this area, multi-'omics datasets on complex diseases are becoming more readily available and there is a need to set standards and good practices for integrated analysis of biological, clinical and environmental data. We present a framework to plan and generate single and multi-'omics signatures of disease states. METHODS: The framework is divided into four major steps: dataset subsetting, feature filtering, 'omics-based clustering and biomarker identification. RESULTS: We illustrate the usefulness of this framework by identifying potential patient clusters based on integrated multi-'omics signatures in a publicly available ovarian cystadenocarcinoma dataset. The analysis generated a higher number of stable and clinically relevant clusters than previously reported, and enabled the generation of predictive models of patient outcomes. CONCLUSIONS: This framework will help health researchers plan and perform multi-'omics big data analyses to generate hypotheses and make sense of their rich, diverse and ever growing datasets, to enable implementation of translational P4 medicine.

Efficacy and safety of olokizumab in patients with rheumatoid arthritis with an inadequate response to TNF inhibitor therapy: outcomes of a randomised Phase IIb study.

OBJECTIVES: The aim of this 12-week Phase IIb study was to assess the efficacy and safety of olokizumab (OKZ), a humanised anti-IL6 monoclonal antibody, in patients with rheumatoid arthritis (RA) with moderate-to-severe disease activity who had previously failed tumour necrosis factor (TNF) inhibitor therapy. The dose-exposure-response relationship for OKZ was also investigated. METHODS: Patients were randomised to one of nine treatment arms receiving placebo (PBO) or OKZ (60, 120 or 240 mg) every 4 weeks (Q4W) or every 2 weeks (Q2W), or 8 mg/kg tocilizumab (TCZ) Q4W. The primary endpoint was change from baseline in DAS28(C-reactive protein, CRP) at Week 12. Secondary efficacy endpoints were American College of Rheumatology 20 (ACR20), ACR50 and ACR70 response rates at Week 12. Exploratory analyses included comparisons of OKZ efficacy with TCZ. RESULTS: Across 221 randomised patients, OKZ treatment produced significantly greater reductions in DAS28(CRP) from baseline levels at Week 12, compared to PBO (p<0.001), at all the OKZ doses tested (60 mg OKZ p=0.0001, 120 and 240 mg OKZ p<0.0001). Additionally, ACR20 and ACR50 responses were numerically higher for OKZ than PBO (ACR20: PBO=17.1-29.9%, OKZ=32.5-60.7%; ACR50: PBO=1.3-4.9%, OKZ=11.5-33.2%). OKZ treatment, at several doses, demonstrated similar efficacy to TCZ across multiple endpoints. Most adverse events were mild or moderate and comparable between OKZ and TCZ treatment groups. Pharmacokinetic/pharmacodynamic modelling demonstrated a shallow dose/exposure response relationship in terms of percentage of patients with DAS28(CRP) <2.6. CONCLUSIONS: OKZ produced significantly greater reductions in DAS28(CRP) from baseline at Week 12 compared with PBO. Reported AEs were consistent with the safety profile expected of this class of drug, with no new safety signals identified. TRIAL REGISTER NUMBER: NCT01242488.

Safety and pharmacokinetics of olokizumab, an anti-IL-6 monoclonal antibody, administered to healthy male volunteers: A randomized phase I study.

Interleukin-6 (IL-6) is implicated in the pathophysiology of several inflammatory conditions. Olokizumab, a humanized anti-IL-6 monoclonal antibody, selectively blocks the final assembly of the IL-6 signaling complex. A randomized, double-blind, placebo-controlled, phase I dose-escalation study assessed the safety and tolerability of escalating single doses of olokizumab administered intravenously (iv) or subcutaneously (sc) to 67 healthy male volunteers. The pharmacokinetics, pharmacodynamics and immunogenicity of olokizumab were also assessed. Olokizumab was tolerated at doses up to 3.0 mg/kg sc and 10.0 mg/kg iv; the maximum tolerated dose was not reached. No serious adverse events or withdrawals as a result of treatment-emergent adverse events were reported. Pharmacokinetic analysis showed that both maximum serum concentration and area under the concentration-time curve increased linearly with increasing dose. Mean terminal half-life was 31.5 days (standard deviation 12.4 days). The bioavailability of the sc doses ranged from 84.2% to 92.5%. Rapid decreases in C-reactive protein concentrations were observed, with no dose dependency.

Leveraging systems biology approaches in clinical pharmacology.

Computational modeling has been adopted in all aspects of drug research and development, from the early phases of target identification and drug discovery to the late-stage clinical trials. The different questions addressed during each stage of drug R&D has led to the emergence of different modeling methodologies. In the research phase, systems biology couples experimental data with elaborate computational modeling techniques to capture lifecycle and effector cellular functions (e.g. metabolism, signaling, transcription regulation, protein synthesis and interaction) and integrates them in quantitative models. These models are subsequently used in various ways, i.e. to identify new targets, generate testable hypotheses, gain insights on the drug's mode of action (MOA), translate preclinical findings, and assess the potential of clinical drug efficacy and toxicity. In the development phase, pharmacokinetic/pharmacodynamic (PK/PD) modeling is the established way to determine safe and efficacious doses for testing at increasingly larger, and more pertinent to the target indication, cohorts of subjects. First, the relationship between drug input and its concentration in plasma is established. Second, the relationship between this concentration and desired or undesired PD responses is ascertained. Recognizing that the interface of systems biology with PK/PD will facilitate drug development, systems pharmacology came into existence, combining methods from PK/PD modeling and systems engineering explicitly to account for the implicated mechanisms of the target system in the study of drug-target interactions. Herein, a number of popular system biology methodologies are discussed, which could be leveraged within a systems pharmacology framework to address major issues in drug development.

Partitioning, diffusivity and clearance of skin permeants in mammalian dermis.

The partition coefficients (K(de)) and diffusivities (D(de)) of compounds in mammalian dermis were examined through an analysis of in vitro permeation data obtained from the literature combined with experimental results with the test permeant, (3)H-testosterone. The literature data involved 26 compounds ranging in molecular weight from 18 to 476 Da and four species-human, guinea pig, rat and mouse. Testosterone was studied by permeation and desorption measurements employing excised human dermis in the presence and absence of external serum albumin. Mathematical models for both K(de) and D(de) were developed. The K(de) model involved ionization, binding to extravascular serum proteins and partitioning into a small lipid compartment. The D(de) model employed a free diffusivity with a liquid-like size dependence multiplied by a binding factor derived from K(de). An additional analysis considered in vivo dermal concentration profiles of topically applied permeants. Literature data for 5 of 6 permeants were shown to be well described by a previously published model for capillary clearance in the dermis, which leads to an exponential decay of concentration with depth. Computed decay lengths (1/e values) ranged from 210 to 920 microm, and the corresponding clearance rate constants k(de) ranged from 0.9 x 10(-4) to 14 x 10(-4)s(-1) (n=8). Departures from the exponential decay profile are discussed in terms of non-uniform capillary clearance and incomplete attainment of a steady-state.

A geometrical model of dermal capillary clearance.

A new microscopic model is developed to describe the dermal capillary clearance process of skin permeants. The physiological structure is represented in terms of a doubly periodic array of absorbing capillaries. Convection-dominated transport in the blood flow within the capillaries is coupled with interstitial diffusion, the latter process being quantified via a slender-body-theory approach. Convection across the capillary wall and in the interstitial phase is treated as a perturbation which may be added to the diffusive transport. The model accounts for the finite permeability of the capillary wall as well as for the geometry of the capillary array, based on realistic values of physiological parameters. Calculated dermal concentration profiles for permeants having the size and lipophilicity of salicylic acid and glucose illustrate the power and general applicability of the model. Furthermore, validation of the model with published in vivo experimental results pertaining to human skin permeation of hydrocortisone is presented. The model offers the possibility for in-depth theoretical understanding and prediction of subsurface drug distribution in the human skin following topical application, as well as rates of capillary clearance into the systemic circulation. A simpler approach that treats the capillary bed as a homogeneously absorbing zone is also employed. The latter may be used in conjunction with the capillary exchange model to estimate measurable dermal transport and clearance parameters in a straightforward manner.

Glucose partition coefficient and diffusivity in the lower skin layers.

PURPOSE: This work aims to estimate the diffusivity and partitioning of glucose in the dermis and the viable epidermis of human skin. METHODS: The partition coefficient of glucose between phosphate-buffered saline and dermis, tape-stripped epidermis (TSE), stratum corneum (SC), and split-thickness skin, was measured in vitro using human cadaver skin. Glucose permeability across dermis and tape-stripped split-thickness skin (TSS) was measured using side-by-side diffusion cells. Glucose desorption from TSE and human epidermal membrane (HEM) was measured. All measurements were conducted at 32 degrees C. RESULTS: The partition coefficient for glucose [mean +/- SD (no. of samples)] was 0.65 +/- 0.09 (n = 25) for dermis, 0.81 +/- 0.06 (n = 10) for TSE, and 0.53 +/- 0.12 (n = 9) for SC. Glucose diffusivity in dermis was calculated to be 2.64 +/- 0.42 x 10(-6) cm2/s (n = 14). Glucose diffusivities in the viable epidermis estimated from TSS permeation, TSE desorption, and HEM desorption were 0.075 +/- 0.050 x 10(-6) cm2/s (n = 5), 0.037 +/- 0.018 x 10(-6) cm2/s (n = 4), and 1.0 +/- 0.6 x 10(-6) cm2/s (n = 4), respectively. CONCLUSION: The tissue/buffer partition coefficient of glucose in all skin layers was found to be less than unity, suggestive of excluded volumes in each layer. Glucose diffusivity in human dermis was found to be one third of its value in water, indicative of hindered diffusion related to the structural components of the tissue. A substantially lower value for glucose diffusivity in viable epidermis is suggested.

Distributed diffusion-clearance model for transient drug distribution within the skin.

Quantitative predictions of molecular transport rates through the skin are key to the development of topically applied and transdermally delivered drugs, as well as risk assessment associated with dermal exposure. Most research to date has focused on correlations for the permeability of the stratum corneum, and transient diffusion models that oversimplify vascular clearance processes in terms of a perfect-removal boundary condition at an artificially introduced lower boundary. Considerations of the spatially distributed nature and action of blood vessels have usually been limited to the steady-state case. This article describes a more comprehensive transient model of percutaneous absorption formulated in terms of volumetric dispersion and clearance coefficients reflecting the spatial distribution of vascular processes. The model was implemented through an analysis of published experimental results on in vivo permeation of salicylic acid (SA) in de-epidermized rat skin. With regard to the characterization of SA in rat dermis ("de") in vivo, it was found that: (i) SA is likely to have a dermal effective partition coefficient (relative to pH 7.4 aqueous buffer "pH7.4") around unity (K(de/pH7.4) = 0.9 +/- 0.3); (ii) vascular processes seem not to increase drug dispersion significantly beyond molecular diffusion [D(de) approximately (D(de))(mol) = (8 +/- 3) . 10(-7) cm(2) s(-1)]; and (iii) vascular clearance is characterized by a rate coefficient k(de) = (7 +/- 2) . 10(-4) s(-1). Application of a whole-skin variant of the model (including the stratum corneum and viable epidermis) allowed realistic predictions to be made of transient subsurface concentration levels after application from a finite dose.