Search

ObjectivesThe use of preventive medication in palliative oncology patients may be inappropriate due to limited life expectancy. Deprescribing tools are available but time-consuming and not always tailored to this specific population. Our primary goal was to identify potentially inappropriate medications (PIMs) in palliative oncology patients with a life expectancy of up to 2 years using an adapted deprescribing tool. Our secondary aim was to identify patient characteristics associated with the presence of PIMs.MethodsOncology patients with a life expectancy of up to 2 years were included cross-sectionally. An adapted deprescribing tool was developed to identify PIMs. Logistic regression was used to identify factors associated with having PIMs.ResultsA total of 218 patients were included in this study of which 56% had at least one PIM with a population mean of 1.1 PIM per patient. Most frequently defined PIMs were antihypertensive drugs and gastric acid inhibitors. Identification of PIMs by review took an estimated 5–10 min per patient. Polypharmacy, age >65 years and inpatient/outpatient status were found to be associated with having at least one PIM.ConclusionsDeprescribing is possible in more than half of palliative oncology patients with a life expectancy of up to 2 years. The adapted deprescribing tool used is non-time consuming and suitable for palliative oncology patients, regardless of age.

The large spatial/temporal/frequency scale of geoscience and remote-sensing datasets causes memory issues when using convolutional neural networks for (sub-) surface data segmentation. Recently developed fully reversible or fully invertible networks can mostly avoid memory limitations by recomputing the states during the backward pass through the network. This results in a low and fixed memory requirement for storing network states, as opposed to the typical linear memory growth with network depth. This work focuses on a fully invertible network based on the telegraph equation. While reversibility saves the major amount of memory used in deep networks by the data, the convolutional kernels can take up most memory if fully invertible networks contain multiple invertible pooling/coarsening layers. We address the explosion of the number of convolutional kernels by combining fully invertible networks with layers that contain the convolutional kernels in a compressed form directly. A second challenge is that invertible networks output a tensor the same size as its input. This property prevents the straightforward application of invertible networks to applications that map between different input–output dimensions, need to map to outputs with more channels than present in the input data, or desire outputs that decrease/increase the resolution compared to the input data. However, we show that by employing invertible networks in a non-standard fashion, we can still use them for these tasks. Examples in hyperspectral land-use classification, airborne geophysical surveying, and seismic imaging illustrate that we can input large data volumes in one chunk and do not need to work on small patches, use dimensionality reduction, or employ methods that classify a patch to a single central pixel.

Metoprolol is among the most frequently prescribed β-blockers for the treatment of various cardiovascular diseases. Genetic polymorphism within CYP2D6has been shown to affect the rate of metabolism of metoprolol. Whether metoprolol dose adjustments are indicated in CYP2D6poor metabolizers (PMs) has thus far not well been studied. The aim of this study was to determine the effect of the CYP2D6genotype on the metoprolol maintenance dose in a chronic Dutch patient population. Patients were included if they were treated with metoprolol and in whom CYP2D6genotype status was known. Patient and treatment characteristics were obtained retrospectively from the electronic healthcare records. Metoprolol maintenance dose was the primary endpoint and was defined as the last known dose that the patients had been treated with. Genotype data were categorized into four phenotypes, that is, PMs, intermediate metabolizers, extensive metabolizers, and ultra-rapid metabolizers (UMs). The endpoints were analyzed as PM versus non-PM. A total of 105 patients were included. The mean ± SD maintenance dose in PMs (n= 12) was significantly lower compared with non-PMs (n= 93), that is, 48 ± 20 versus 84 ± 53 mg, respectively (P= 0.019). No association of the CYP2D6genotype with the incidence of side effects was observed, although there was a trend for increased risk of drowsiness (P= 0.053). The results of this study show that the CYP2D6genotype is associated with the maintenance dose of metoprolol. Patients with the CYP2D6 PM phenotype may benefit from a lower metoprolol starting dose, followed by further dose titration to provide patient-tailored therapy and thereby increase the effectiveness of treatment.

Convolutional neural networks (CNN) have recently seen tremendous success in various computer vision tasks. However, their application to problems with high dimensional input and output, such as high-resolution image and video segmentation or 3D medical imaging, has been limited by various factors. Primarily, in the training stage, it is necessary to store network activations for back-propagation. In these settings, the memory requirements associated with storing activations can exceed what is feasible with current hardware, especially for problems in 3D. Motivated by the propagation of signals over physical networks, that are governed by the hyperbolic Telegraph equation, in this work we introduce a fully conservative hyperbolic network for problems with high-dimensional input and output. We introduce a coarsening operation that allows completely reversible CNNs by using a learnable discrete wavelet transform and its inverse to both coarsen and interpolate the network state and change the number of channels. We show that fully reversible networks are able to achieve results comparable to the state of the art in 4D time-lapse hyper-spectral image segmentation and full 3D video segmentation, with a much lower memory footprint that is a constant independent of the network depth. We also extend the use of such networks to variational auto-encoders, where optimization begins from an exact recovery and we discover the level of compression through optimization.

The SLCO1B1c.521T>C polymorphism is associated with statin plasma levels and simvastatin-induced adverse drug reactions. We studied whether the c.521T>C polymorphism is associated with dose decreases or switches to other cholesterol-lowering drugs during simvastatin and atorvastatin therapy, because these events are indicators of adverse drug reactions.

Automation robots have recently come to the market as an alternative for manual compounding of drugs for intravenous administration. Our aim was to assess whether robotic compounding can be performed with monoclonal antibodies (mAbs) without influencing the aggregation state of the proteins. Three frequently used mAbs were studied: infliximab (Remicade®, Janssen Biotech) and trastuzumab (Herceptin®, Roche) in lyophilised form, and bevacizumab (Avastin®, Roche) as a liquid formulation stored at 2°C to 8°C. The effects of different procedures to prepare the patient doses on antibody aggregation were evaluated. Remicade®and Herceptin®were reconstituted both manually and by a robotic arm (i.v.STATION®, Health Robotics). Additionally, the influence of vigorous shaking during reconstitution was investigated. The effects of rapid aspiration and dispensing on antibody aggregation were investigated for all three mAbs. Aggregation state was assessed by UV-Vis absorbance, 90° light scatter, fluorescence spectroscopy, Nile red fluorescence microscopy, and field flow fractionation without cross and focus flow. Robotic reconstituted samples showed similar findings compared with manual reconstitution if performed exactly according to the summary of product characteristics (SPC). Vials that were vigorously shaken showed a significant increase in aggregates. Similarly, rapid aspiration/dispense cycles resulted in a strong increase in the number and sizes of aggregates for all three mAbs; this result was observed after just one rapid aspiration/dispense cycle. Our study showed that robotic compounding of mAbs is feasible if the robot is exactly programmed according to the SPC, indicating that robotic compounding can be used to achieve reproducible high-quality compounding for delicate formulations.

High plasma HDL cholesterol is associated with reduced risk of myocardial infarction, but whether this association is causal is unclear. Exploiting the fact that genotypes are randomly assigned at meiosis, are independent of non-genetic confounding, and are unmodified by disease processes, mendelian randomisation can be used to test the hypothesis that the association of a plasma biomarker with disease is causal.

To investigate the influence of tagging single-nucleotide polymorphisms (SNPs) within candidate genes involved in the putative anti-inflammatory effects of statins on the effectiveness of statins in reducing the risk of myocardial infarction (MI).

Pharmacogenetic research has shown that genetic variation may influence statin responsiveness. Statins exert a variety of beneficial effects beyond lipid lowering, including antithrombotic effects, which contribute to the risk reduction of cardiovascular disease. Statins have been shown to influence the expression of coagulation factors II, V, VII, XII and XIII.

The GNB3 C825T polymorphism has been shown to affect lipid parameters, atherosclerosis progression, and incidence of myocardial infarction (MI). Therefore, we assessed whether the effectiveness of statins in reducing the risk of MI was modified by the GNB3 C825T polymorphism.