Targeted learning in observational studies with multi-valued treatments: An evaluation of antipsychotic drug treatment safety.

We investigate estimation of causal effects of multiple competing (multi-valued) treatments in the absence of randomization. Our work is motivated by an intention-to-treat study of the relative cardiometabolic risk of assignment to one of six commonly prescribed antipsychotic drugs in a cohort of nearly 39 000 adults with serious mental illnesses. Doubly-robust estimators, such as targeted minimum loss-based estimation (TMLE), require correct specification of either the treatment model or outcome model to ensure consistent estimation; however, common TMLE implementations estimate treatment probabilities using multiple binomial regressions rather than multinomial regression. We implement a TMLE estimator that uses multinomial treatment assignment and ensemble machine learning to estimate average treatment effects. Our multinomial implementation improves coverage, but does not necessarily reduce bias, relative to the binomial implementation in simulation experiments with varying treatment propensity overlap and event rates. Evaluating the causal effects of the antipsychotics on 3-year diabetes risk or death, we find a safety benefit of moving from a second-generation drug considered among the safest of the second-generation drugs to an infrequently prescribed first-generation drug known for having low cardiometabolic risk.

Antipsychotics and the risk of diabetes and death among adults with serious mental illnesses.

BACKGROUND: Individuals with schizophrenia exposed to second-generation antipsychotics (SGA) have an increased risk for diabetes, with aripiprazole purportedly a safer drug. Less is known about the drugs' mortality risk or whether serious mental illness (SMI) diagnosis or race/ethnicity modify these effects. METHODS: Authors created a retrospective cohort of non-elderly adults with SMI initiating monotherapy with an SGA (olanzapine, quetiapine, risperidone, and ziprasidone, aripiprazole) or haloperidol during 2008-2013. Three-year diabetes incidence or all-cause death risk differences were estimated between each drug and aripiprazole, the comparator, as well as effects within SMI diagnosis and race/ethnicity. Sensitivity analyses evaluated potential confounding by indication. RESULTS: 38 762 adults, 65% White and 55% with schizophrenia, initiated monotherapy, with haloperidol least (6%) and quetiapine most (26·5%) frequent. Three-year mortality was 5% and diabetes incidence 9.3%. Compared with aripiprazole, haloperidol and olanzapine reduced diabetes risk by 1.9 (95% CI 1.2-2.6) percentage points, or a 18.6 percentage point reduction relative to aripiprazole users' unadjusted risk (10.2%), with risperidone having a smaller advantage. Relative to aripiprazole users' unadjusted risk (3.4%), all antipsychotics increased mortality risk by 1.1-2.2 percentage points, representing 32.4-64.7 percentage point increases. Findings within diagnosis and race/ethnicity were generally consistent with overall findings. Only quetiapine's higher mortality risk held in sensitivity analyses. CONCLUSIONS: Haloperidol's, olanzapine's, and risperidone's lower diabetes risks relative to aripiprazole were not robust in sensitivity analyses but quetiapine's higher mortality risk proved robust. Findings expand the evidence on antipsychotics' risks, suggesting a need for caution in the use of quetiapine among individuals with SMI.

Single channel and ensemble hERG conductance measured in droplet bilayers.

The human ether-a-go-go related gene (hERG) encodes the potassium channel Kv11.1, which plays a key role in the cardiac action potential and has been implicated in cardiac disorders as well as a number of off-target pharmaceutical interactions. The electrophysiology of this channel has been predominantly studied using patch clamp, but lipid bilayers have the potential to offer some advantages, including apparatus simplicity, ease of use, and the ability to control the membrane and solution compositions. We made membrane preparations from hERG-expressing cells and measured them using droplet bilayers, allowing measurement of channel ensemble currents and 13.5 pS single channel currents. These currents were ion selective and were blockable by E-4031 and dofetilide in a dose-dependent manner, allowing determination of IC50 values of 17 nM and 9.65 µM for E-4031 and dofetilide, respectively. We also observed time- and voltage- dependent currents following step changes in applied potential that were similar to previously reported patch clamp measurements.

Automated lipid bilayer and ion channel measurement platform.

Ion channels and transmembrane proteins play key roles in a wide range of physiological processes. Engineered ion channels have been explored as highly sensitive single molecule sensors. Scientific and sensing measurements of ion channel conductance often utilize artificial lipid bilayers, which have shortcomings limiting their application. We describe a fully automated lipid bilayer formation system that integrates the measurement electronics within the fluidic controls. Unattended operation of this system resulted in highly reproducible automatic bilayer formation and ion channel measurement over dozens of consecutive trials. The fully automated, closed-loop control algorithm enabled autonomous operation of the platform, a step toward applications of ion channel measurements for remote sensing and pharmacological studies requiring minimal operator involvement.

Automatable production of shippable bilayer chips by pin tool deposition for an ion channel measurement platform.

As a step towards an automated and operator-free ion channel measurement platform we have previously demonstrated a solution formulation for artificial lipid bilayers that enabled the indefinite storage and shipping of frozen bilayer precursors. In this work, the solutions were deposited by hand. Here, we have adapted pin tools to deposit the bilayer precursor solutions onto multi-element arrays, a popular method for microarray solution deposition. The pin tools have enabled the deposited volume to be applied highly repeatably and controllably, resulting in reduction of bilayer formation times to <1 h. The pin tools are also compatible with computerized motion control platforms, enabling automated and high throughput production. We discuss these results and the prospects of this technology to produce high density bilayer arrays for high throughput measurement of ion channels incorporated into artificial bilayers.

Ion channel and toxin measurement using a high throughput lipid membrane platform.

Measurements of ion channels are important for scientific, sensing and pharmaceutical applications. Reconstitution of ion channels into lipid vesicles and planar lipid bilayers for measurement at the single molecule level is a laborious and slow process incompatible with the high throughput methods and equipment used for sensing and drug discovery. A recently published method of lipid bilayer formation mechanically combines lipid monolayers self-assembled at the interfaces of aqueous and apolar phases. We have expanded on this method by vertically orienting these phases and using gravity as the driving force to combine the monolayers. As this method only requires fluid dispensation, it is trivially integrated with high throughput automated liquid-handling robotics. In a proof-of-concept demonstration, we created over 2200 lipid bilayers in 3h. We show single molecule measurements of technologically and physiologically relevant ion channels incorporated into lipid bilayers formed with this method.

Long-term storable and shippable lipid bilayer membrane platform.

The fragility and short lifetimes characteristic of conventionally formed lipid bilayer membranes has necessitated their preparation to be at the time and point of use. By using high freezing-point lipid-solvent mixtures, the process of lipid bilayer self-assembly may be reversibly arrested. In solid form, the bilayer precursor can be stored indefinitely and is sufficiently robust to withstand commercial shipping. Upon thawing, bilayer self-assembly resumes, resulting in a biologically functional membrane. Combination of this membrane precursor with an inexpensive chip results in a compact, practical, and disposable platform for ion channel measurements.

Black lipid membranes stabilized through substrate conjugation to a hydrogel.

Recent research in stabilizing lipid bilayer membranes has been directed toward tethering the membrane to a solid surface or contacting the membrane with a solid support such as a gel. It is also known that the solvent annulus plays an important role in lipid bilayer stability. In this work, the authors set out to stabilize the solvent annulus. Glass substrates with approximately 500 mum apertures were functionalized with 3-methacryloxypropyltrimethoxysilane to allow cross-linking with a surrounding polyethyleneglycol dimethacrylate hydrogel. The hydrogel makes a conformal mold around both the lipid bilayer and the solvent reservoir. Since the hydrogel is covalently conjugated with the glass substrate via vinyl groups, the solvent annulus is prevented from leaving the aperture boundary. Measurements of a membrane created with this approach showed that it remained a stable bilayer with a resistance greater than 1 GOmega for 12 days. Measurements of the ion channel gramicidin A, alpha-hemolysin, and alamethicin incorporated into these membranes showed the same conductance behavior as conventional membranes.