Artificial intelligence and machine learning for clinical pharmacology.

Artificial intelligence (AI) will impact many aspects of clinical pharmacology, including drug discovery and development, clinical trials, personalized medicine, pharmacogenomics, pharmacovigilance and clinical toxicology. The rapid progress of AI in healthcare means clinical pharmacologists should have an understanding of AI and its implementation in clinical practice. As with any new therapy or health technology, it is imperative that AI tools are subject to robust and stringent evaluation to ensure that they enhance clinical practice in a safe and equitable manner. This review serves as an introduction to AI for the clinical pharmacologist, highlighting current applications, aspects of model development and issues surrounding evaluation and deployment. The aim of this article is to empower clinical pharmacologists to embrace and lead on the safe and effective use of AI within healthcare.

Genetic Evidence for Protective Effects of Angiotensin-Converting Enzyme Against Alzheimer Disease But Not Other Neurodegenerative Diseases in European Populations.

Background and Objectives: Angiotensin-converting enzyme (ACE) inhibitors are a commonly prescribed class of medication used to treat heart failure, hypertension, and chronic kidney disease. However, previous observational studies have shown conflicting directions of associations between ACE inhibitors and risk of Alzheimer disease. Genetic evidence has supported a protective effect of cerebral ACE against Alzheimer disease (AD). However, it is unclear whether this effect is mediated through blood pressure and extends to other neurodegenerative diseases. Methods: We performed genetic colocalization investigating an effect of cortical ACE expression on AD risk in people of European ancestry. We further investigated whether any effect of ACE expression on AD risk is mediated through changes in blood pressure and whether effects extend to Parkinson disease, small-vessel disease, or cognitive function in a Mendelian randomization paradigm. Results: There was genetic evidence supporting a protective effect of cortical ACE expression on AD risk in people of European ancestry. Although higher cortical ACE expression was associated with higher blood pressure, there was no strong evidence to support that its association with AD was mediated through blood pressure nor that ACE expression affected risk of other neurodegenerative traits. Discussion: Genetic evidence supports protective effects of cerebral ACE expression on AD, but not other neurodegenerative outcomes in people of European ancestry. Further work is required to investigate whether therapeutic inhibition of ACE increases risk of Alzheimer disease.

Management of Heart Failure in Patients with Chronic Kidney Disease.

Chronic kidney disease (CKD) is increasingly prevalent in patients with heart failure (HF) and HF is one of the leading causes of hospitalisation, morbidity and mortality in patients with impaired renal function. Currently, there is strong evidence to support the symptomatic and prognostic benefits of ß-blockers, renin-angiotensin-aldosterone inhibitors (RAASis), angiotensin receptor-neprilysin inhibitors (ARNIs) and mineralocorticoid receptor antagonists (MRA) in patients with HF and CKD stages 1-3. However, ARNIs, RAASis and MRAs are often suboptimally prescribed for patients with CKD owing to concerns about hyperkalaemia and worsening renal function. There is growing evidence for the use of sodium-glucose co-transporter 2 inhibitors and IV iron therapy in the management of HF in patients with CKD. However, few studies have included patients with CKD stages 4-5 and patients receiving dialysis, limiting the assessment of the safety and efficacy of these therapies in advanced CKD. Interdisciplinary input from HF and renal specialists is required to provide integrated care for the growing number of patients with HF and CKD.

Elevated Urinary Biomarkers of Oxidative Damage in Photocopier Operators following Acute and Chronic Exposures.

Inhalation exposures to nanoparticles (NPs) from printers and photocopiers have been associated with upper airway and systemic inflammation, increased blood pressure, and cases of autoimmune and respiratory disorders. In this study we investigate oxidative stress induced by exposures to copier-emitted nanoparticles using a panel of urinary oxidative stress (OS) biomarkers representing DNA damage (8-hydroxydeoxyguanosine, 8-OHdG; 8-hydroxyguanosine, 8-OHG; 5-hydroxymethyl uracil 5-OHMeU), lipid peroxidation (8-isoprostane; 4-hydroxynonenal, HNE), and protein oxidation biomarkers (o-tyrosine, 3-chlorotyrosine, and 3-nitrotyrosine) under conditions of acute (single 6 h exposure, 9 volunteers, 110 urine samples) and chronic exposures (6 workers, 11 controls, 81 urine samples). Urinary biomarkers were quantified with liquid chromatography-tandem mass spectrometry after solid phase extraction sample cleanup. 8-OHdG, 8-OHG, 8-isoprostane, and HNE were significantly elevated in both the acute and chronic exposure study participants relative to the controls. In the acute exposure study, the geometric mean ratios post-/pre-exposure were 1.42, 1.10, 2.0, and 2.25, respectively. Urinary 8-OHG and HNE increased with time to at least 36 h post-exposure (post-/pre-exposure GM ratios increased to 3.94 and 2.33, respectively), suggesting slower generation and/or urinary excretion kinetics for these biomarkers. In chronically exposed operators, the GM ratios of urinary biomarkers relative to controls ranged from 1.52 to 2.94, depending on the biomarker. O-Tyrosine and 5-OHMeU biomarkers were not significantly different from the controls. 3-chlorotyrosine and 3-nitrotyrosine were not detected in the urine samples. We conclude that NPs from photocopiers induce systemic oxidative stress by damaging DNA, RNA, and lipids. Urinary levels of 8-OHdG, 8-OHG, HNE, and 8-isoprostane were orders of magnitude higher than in nanocomposite processing workers, comparable to nano titanium dioxide and fiberglass manufacturing workers, but much lower than in shipyard welding and carbon nanotube synthesis workers. Biomarkers 8-OHdG, 8-OHG, 8-isoprostane, and HNE appear to be more sensitive and robust urinary biomarkers for monitoring oxidative stress to NPs from photocopiers.

Sleep Disordered Breathing, Obesity and Atrial Fibrillation: A Mendelian Randomisation Study.

It remains unclear whether the association between obstructive sleep apnoea (OSA), a form of sleep-disordered breathing (SDB), and atrial fibrillation (AF) is causal or mediated by shared co-morbidities such as obesity. Existing observational studies are conflicting and limited by confounding and reverse causality. We performed Mendelian randomisation (MR) to investigate the causal relationships between SDB, body mass index (BMI) and AF. Single-nucleotide polymorphisms associated with SDB (n = 29) and BMI (n = 453) were selected as instrumental variables to investigate the effects of SDB and BMI on AF, using genetic association data on 55,114 AF cases and 482,295 controls. Primary analysis was conducted using inverse-variance weighted MR. Higher genetically predicted SDB and BMI were associated with increased risk of AF (OR per log OR increase in snoring liability 2.09 (95% CI 1.10-3.98), p = 0.03; OR per 1-SD increase in BMI 1.33 (95% CI 1.24-1.42), p < 0.001). The association between SDB and AF was not observed in sensitivity analyses, whilst associations between BMI and AF remained consistent. Similarly, in multivariable MR, SDB was not associated with AF after adjusting for BMI (OR 0.68 (95% CI 0.42-1.10), p = 0.12). Higher BMI remained associated with increased risk of AF after adjusting for OSA (OR 1.40 (95% CI 1.30-1.51), p < 0.001). Elevated BMI appears causal for AF, independent of SDB. Our data suggest that the association between SDB, in general, and AF is attributable to mediation or confounding from obesity, though we cannot exclude that more severe SDB phenotypes (i.e., OSA) are causal for AF.

Leveraging genetic data to investigate the effects of interleukin-6 receptor signalling on levels of 40 circulating cytokines.

Interleukin 6 (IL-6) is a circulating cytokine implicated in inflammatory processes. However, the broad effects of IL-6 receptor (IL-6R) signalling on other circulating cytokines is not known. Using summary-level data from genome-wide association studies, we leveraged genetic variants that proxy IL-6R signalling in two-sample Mendelian randomization analyses to investigate effects on levels of 40 circulating cytokines. Increased genetically proxied IL-6R signalling was associated with reduced levels of 10 circulating interleukins, chemokines and growth factors. The significant results include IL-10 (Mendelian randomization estimate -0.306, standard error [SE] 0.093), IL-4 (estimate -0.393, SE 0.1007), eotaxin (estimate -0.510, SE 0.1213) and Fibroblast growth factor (estimate -0.334, SE 0.1005). The findings from this study support the feedback effects of IL-6R signalling on reducing levels of some circulating cytokines and identify compensatory mechanisms that maybe modulating its inflammatory effects. These results provide insight into the mechanisms by which IL-6R signalling may be contributing to inflammatory and autoimmune diseases.

Metabolic Traits and Stroke Risk in Individuals of African Ancestry: Mendelian Randomization Analysis.

BACKGROUND AND PURPOSE: Metabolic traits affect ischemic stroke (IS) risk, but the degree to which this varies across different ethnic ancestries is not known. Our aim was to apply Mendelian randomization to investigate the causal effects of type 2 diabetes (T2D) liability and lipid traits on IS risk in African ancestry individuals, and to compare them to estimates obtained in European ancestry individuals. METHODS: For African ancestry individuals, genetic proxies for T2D liability and circulating lipids were obtained from a meta-analysis of the African Partnership for Chronic Disease Research study, the UK Biobank, and the Million Veteran Program (total N=77 061). Genetic association estimates for IS risk were obtained from the Consortium of Minority Population Genome-Wide Association Studies of Stroke (3734 cases and 18 317 controls). For European ancestry individuals, genetic proxies for the same metabolic traits were obtained from Million Veteran Program (lipids N=297 626, T2D N=148 726 cases, and 965 732 controls), and genetic association estimates for IS risk were obtained from the MEGASTROKE study (34 217 cases and 406 111 controls). Random-effects inverse-variance weighted Mendelian randomization was used as the main method, complemented with sensitivity analyses more robust to pleiotropy. RESULTS: Higher genetically proxied T2D liability, LDL-C (low-density lipoprotein cholesterol), total cholesterol and lower genetically proxied HDL-C (high-density lipoprotein cholesterol) were associated with increased risk of IS in African ancestry individuals (odds ratio per doubling the odds of T2D liability [95% CI], 1.09 [1.07-1.11]; per standard-deviation increase in LDL-C, 1.12 [1.04-1.21]; total cholesterol: 1.23 [1.06-1.43]; HDL-C, 0.93 [0.89-0.99]). There was no evidence for differences in these estimates when performing analyses in European ancestry individuals. CONCLUSIONS: Our analyses support a causal effect of T2D liability and lipid traits on IS risk in African ancestry individuals, with Mendelian randomization estimates similar to those obtained in European ancestry individuals.

Inhibition of interleukin 6 signalling and renal function: A Mendelian randomization study.

Inhibition of interleukin 6 (IL-6) signalling has been proposed as a potential cardioprotective strategy for patients with chronic kidney disease (CKD), but the direct effects of IL-6 inhibition on renal function are not known. A Mendelian randomization (MR) study was performed to investigate the association of genetically proxied inhibition of IL-6 signalling with estimated glomerular filtration rate (eGFR), CKD and blood urea nitrogen (BUN). Inverse-variance weighted MR was used as the main analysis, with sensitivity analyses performed using simple median, weighted median and MR-Egger methods. There was no evidence for an association of genetically proxied inhibition of IL-6 signalling (scaled per standard deviation unit decrease in C-reactive protein) with log eGFR (0.001, 95% confidence interval -0.004-0.007), BUN (0.009, 95% confidence interval -0.003-0.021) and CKD (odds ratio 0.948, 95% confidence interval 0.822-1.094). These findings do not raise concerns for IL-6 signalling having large adverse effects on renal function.

Comprehensive Assessment of Short-Lived ROS and H2O2 in Laser Printer Emissions: Assessing the Relative Contribution of Metal Oxides and Organic Constituents.

Inhalation exposure to nanoparticles from toner-based laser printer and photocopier emissions (LPEs) induces airway inflammation and systemic oxidative stress, cytotoxicity, and genotoxicity (such as DNA damage). Recent evidence from human and in vitro studies suggests a strong role for oxidative stress caused by free radicals, such as reactive oxygen species (ROS), in the toxicity of laser printer emissions. However, the amount of ROS generated from laser printer nanoparticle emissions and the relative contribution of various fractions (vapors, organics, metals, and metal oxides) have not been investigated to-date. In this study, we aim to quantify short-lived ROS and H2O2 laser printer emissions, as well as the relative contribution of various fractions of LPEs in ROS generation. An aerosol chamber with HEPA filtered air was used to generate LPE emissions from one representative printer. In separate experiments, size fractionated LPEs were collected on filters (particles) or impingers (particles and vapors). The nanoscale fraction of LPEs (PM0.1) was further separated into the organic fraction and inorganic (transition metals/metal oxides) following a sequence of extraction with solvents and centrifugation. The short-lived ROS and H2O2 generated from each fraction were quantified with an acellular Trolox-based liquid chromatography-electrospray-tandem mass spectrometry (LC-ESI-MS/MS) method recently developed in our lab. The particulate fraction of LPEs PM0.1 generated 2.68 times more total ROS (sum of short-lived ROS and H2O2) than the vapor fraction. In tested LPEs, transition metal oxides, which constituted 3% by mass, produced 69× and 202× times more short-lived ROS and H2O2, respectively, on a mass basis, than the organic fraction. Furthermore, fresh PM0.1 generated 282× and 32× times more short-lived ROS and H2O2, respectively, than aged and processed PM0.1. We conclude that transition metal oxides, albeit a minor constituent of the LPE PM0.1 emissions, are the species responsible for the majority of acellular ROS in this printer. A larger range of printers should be tested in the future. Because transition metal oxides in toners originate primarily from engineering nanomaterials (ENMs) in printer toner powder, reformulation of toner powders to contain less of these ROS active metals is recommended.

Early screening and treatment of gestational diabetes in high-risk women improves maternal and neonatal outcomes: A retrospective clinical audit.

AIMS: Evidence suggests that screening for gestational diabetes (GDM) occurs too late in pregnancy, when changes in glucose metabolism and fetal growth rates can already be detected. In August 2016 NHS Lothian began screening women with risk factors for GDM during early pregnancy (11-13 weeks). We hypothesised that an earlier identification and treatment of dysglycaemia would improve pregnancy outcomes compared to previous standard care. METHODS: We compared management and outcomes for singleton pregnancies with GDM delivering at Royal Infirmary Edinburgh, UK, diagnosed through routine or early screening from 01/01/2015-31/10/2017 (routine screening n = 335, early screening n = 241). RESULTS: Early screening increased the proportion of women diagnosed before 24 weeks' gestation (n = 59/335, 17.6% vs n = 103/241, 42.7%, p < 0.001) but did not change the average monthly rate of diagnosis. Early screening increased the median duration of GDM during pregnancy (71 vs 93 days of gestation, p < 0.001) with no significant changes in the pharmacological management. Early screening improved the primary composite outcome (emergency caesarean section, neonatal hypoglycaemia and macrosomia; n = 138/335, 41.2% vs n = 73/241, 30.3%, adjusted Odds Ratio [95% confidence interval] 0.62 [0.43-0.91]. CONCLUSIONS: There is a role for early screening and management of GDM however it is unclear whether this represents a cost-effective intervention.

Characterizing the interaction between uranyl ion and fulvic acid using regional integration analysis (RIA) and fluorescence quenching.

The development of chemometric methods has substantially improved the quantitative usefulness of the fluorescence excitation-emission matrix (EEM) in the analysis of dissolved organic matter (DOM). In this study, Regional Integration Analysis (RIA) was used to quantitatively interpret EEMs and assess fluorescence quenching behavior in order to study the binding between uranyl ion and fulvic acid. Three fulvic acids including soil fulvic acid (SFA), Oyster River fulvic acid (ORFA) and Suwannee River fulvic acid (SRFA) were used and investigated by the spectroscopic techniques. The EEM spectra obtained were divided into five regions according to fluorescence structural features and two distinct peaks were observed in region III and region V. Fluorescence quenching analysis was conducted for these two regions with the stability constants, ligand concentrations and residual fluorescence values calculated using the Ryan-Weber model. Results indicated a relatively strong binding ability between uranyl ion and fulvic acid samples at low pH (log K value varies from 4.11 to 4.67 at pH 3.50). Fluorophores in region III showed a higher binding ability with fewer binding sites than in region V. Stability constants followed the order, SFA > ORFA > SRFA, while ligand concentrations followed the reverse order, SRFA > ORFA > SFA. A comparison between RIA and Parallel Factor Analysis (PARAFAC) data treatment methods was also performed and good agreement between these two methods (less than 4% difference in log K values) demonstrates the reliability of the RIA method in this study.

Toxicity of the 13 priority pollutant metals to Vibrio fisheri in the Microtox chronic toxicity test.

The Microtox Acute Toxicity Test has been successfully used to measure the toxicity of metals and other pollutants at high concentrations (ppm) in selected environmental samples. However, metals and other toxicants are often found in much lower concentrations (ppb) in many municipal wastewaters and receiving waters. In order to assess the toxicity of these pollutants in these samples, a more sensitive toxicity assay is needed. The Microtox chronic toxicity test has been developed to measure the sublethal effect of toxicants over multiple generations of the test species, Vibrio fisheri. In this study, the toxicity of the 13 priority pollutant metals [i.e. As, Se, Cd, Cr (III and VI), Cu, Pb, Sb, Ag, Tl, Zn, Be, Hg and Ni] to V. fisheri was evaluated using the Microtox chronic toxicity test. In this test, the inhibitory concentration (IC), lowest observable effect concentration (LOEC), and no observable effect concentration (NOEC) were obtained after 22-h of incubation at 27+/-1 degrees C, by comparing the light output of the control to that of the test sample. Among the 13 priority pollutant metals, beryllium (Be) was found to be the most toxic in the test (LOEC=0.742-1.49 microg/l) while thallium (Tl) was the least toxic (LOEC=3840-15300 microg/l). The LOECs for copper (as Cu) and lead (Pb) in reagent (ASTM Type I) water were 6.78-13.6 microg/l and 626-1251 microg/l, respectively. The toxicity of copper sulfate (as Cu) in reagent water was shown and significantly reduced with the addition of natural organic matter (fulvic acid) or EDTA to the sample. The LOEC values for the 13 priority pollutant metals in this test were comparable to or lower than those reported for commonly used aquatic toxicity tests, such as the Ceriodaphnia dubia assay.

A modified multisite Stern-Volmer equation for the determination of conditional stability constants and ligand concentrations of soil fulvic acid with metal ions.

In this work, we modify the multisite Stern-Volmer (MSV) equation for fitting fluorescence titration curves. Under the condition of a static quenching mechanism, the MSV postulates an underlying 1:1 fulvic acid (FA)/copper coordination ratio at multisites. Approximates of six fitting parameters characterize the stability constants (K1 and K2) of FA ligands with Cu2+, micromolar ligand site concentrations (CL1 and CL2), the unquenched, steady-state fractional fluorescence contributions (fx), and the residual fluorescence intensity (IRES). Prior to its application to actual FA titration data, the MSV function is simulated, and its predictive ability is confirmed by titrating a mixture of model fluorophores, glycyl-L-tryptophan and L-tryptophan with Cu2+ at pH 6. Molecular fluorescence measurements of FA are acquired at a fixed spectral position (lambda(ex) = 335 nm; lambda(em) = 450 nm), and FA is titrated with copper in triplicate at three pH values-5, 6, and 7. An objective analysis of log K(1) and K(2) values supports several site organization schemes, including (i) subtle, cooperative interaction, (ii) interfering molecular conformations, and (iii) aggregate forms. Site densities (CL1 and CL2) are consistent across varied pH. The f(1) is indicative of a pH-induced spectral shift of a fluorophore and convincingly associates with a transect in the Deltalambda = 25 synchronous fluorescence spectrum and with the preexponential terms describing the time-dependent fluorescence decay. The MSV and its parent one-site version are equivalent for data fitting but are only simple approximations of a FA ligand system with more complex molecular fundamentals.

Multi-wavelength fluorescence-quenching model for determination of Cu2+ conditional stability constants and ligand concentrations of fulvic acid.

In this work, a multi-wavelength model (MWM) is developed. It uses fluorescence bands in the fulvic acid (FA) spectrum that quench upon binding of inorganic Cu2+ to FA. Quenching data at pH values of 5, 6, and 7 are placed in sets, containing fluorescence measures at select wavelengths versus added copper (CM). Intensity data of wavelength set 1 are obtained from 25 nm constant offset synchronous fluorescence spectra (SyF), in which are observed distinct peaks (lambda(ex) = 415 nm, lambda(em) = 440 nm; and lambda(ex) = 471 nm, lambda(em) = 496 nm). Wavelength set 2 intensity data are obtained from the FA fluorescence excitation and emission maxima (lambda(ex) = 335 nm, lambda(em) = 450 nm; and X(ex) = 471 nm, lambda(em) = 496 nm). Application of MWM shows that the multi-wavelength data sets characterize ligands of different binding strength (log K(x)) and concentration (C(Lx)). Corresponding to pH values of 5, 6, and 7, mean and standard deviation values for wavelength set 1 are log K(415/440) = 4.66 (0.12), 5.03 (0.12), and 5.05 (0.08), log K(471/496) = 4.93 (0.06), 5.27 (0.11), and 5.39 (0.09), C(415/440) = 3.1 (1.5), 10.9 (4.5), and 7.9 (3.9) microM, C(471/496) = 14.3 (3.0), 1.7 (0.6), and 1.4 (0.5) microM. And for wavelength set 2, log K(335/450) = 4.50 (0.03), 4.96 (0.27), and 5.22 (0.08), log K(471/496) = 5.02 (0.04), 5.42 (0.32), and 5.71 (0.09), C(335/450) = 8.8 (0.5), 21.9 (7.9), and 18.7 (0.3) microM, C(471/496) = 21.0 (2.5), 7.17 (1.2), and 7.09 (0.3) micrpM. The ability of the 415/440 nm SyF transect to characterize the main excitation and emission maximum of FA at 335/440 nm is evaluated. Relatively low concentration values returned by the model for this transect (415/440 nm) suggest that it is not entirely illustrative of the maximum. The model predictive capability is verified at pH 6 with two fluorescing Cu2+ chelating organic compounds, L-tyrosine and salicylic acid. This test confirms that the model is capable of providing good estimates of equilibrium binding parameters from multi-wavelength measurements of a mixed ligand system.